Handbook of Exploration Geochemistry
VOLUME 7
Geochemical Remote Sensing of the Sub-Surface
Handbook of Exploration Geochemistry
VOLUME 7
Geochemical Remote Sensing of the Sub-Surface
H A N D B O O K OF E X P L O R A T I O N G E O C H E M I S T R Y G.J.S GOVETT (Editor) 1. 2. 3. 4. 5. 6. 7.
ANALYTICAL METHODS IN GEOCHEMICAL PROSPECTING STATISTICS AND DATA ANALYSIS IN GEOCHEMICAL PROSPECTING ROCK GEOCHEMISTRY IN MINERAL EXPLORATION REGOLITH EXPLORATION GEOCHEMISTRY IN TROPICAL AND SUB-TROPICAL TERRAINS REGOLITH EXPLORATION GEOCHEMISTRY IN ARCTIC AND TEMPERATE TERRAINS DRAINAGE GEOCHEMISTRY GEOCHEMICAL REMOTE SENSING OF THE SUB-SURFACE
Handbook of Exploration Geochemistry
VOLUME 7 Geochemical RemoteSensing of the Sub-Surface
Edited by M. HALE International Institute for Aerospace Survey and Earth Sciences and Delft University of Technology Delft, The Netherlands
2000 ELSEVIER Amsterdam- Lausanne- New York-Oxford-Shannon-Singapore-Tokyo
E L S E V I E R S C I E N C E B.V. S a r a B u r g e r h a r t s t r a a t 25 P.O. B o x 2 1 1 , 1 0 0 0 A E A m s t e r d a m ,
The Netherlands
92 0 0 0 E l s e v i e r S c i e n c e B.V. A l l r i g h t s r e s e r v e d . This work is protected under copyright by Elsevier Science, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier Science Rights & Permissions Department, PO Box 800, Oxford OX5 1DX, UK; phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail:
[email protected]. You may also contact Rights & Permissions directly through Elsevier's home page (http://www.elsevier.nl), selecting first 'Customer Support', then 'General Information', then 'Permissions Query Form'. In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (978) 7508400, fax: (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 171 631 5555; fax: (+44) 171 631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of Elsevier Science is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier Science Rights & Permissions Department, at the mail, fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.
First edition 2000 Library of Congress Cataloging
in P u b l i c a t i o n
Data
A c a t a l o g r e c o r d f r o m t h e L i b r a r y o f C o n g r e s s h a s b e e n a p p l i e d for.
ISBN: 0-444-50439-7
GThe
p a p e r u s e d in t h i s p u b l i c a t i o n
(Permanence
of Paper).
P r i n t e d in T h e N e t h e r l a n d s .
meets the requirements
of ANSI/NISO
Z39.48-1992
V
EDITOR'S F O R E W O R D
In my Foreword to the first volume in the Handbook of Exploration Geochemistry, published in 1981, I rashly gave a list of forthcoming volumes that would be published ".... over the next few years .... ". Indeed, the titles were listed on the flyleaf opposite the title page. Volume 7 was listed as Volatile Elements in Mineral Exploration. The advances in concepts since that time are reflected in the title metamorphosing into Geochemical Remote Sensing of the Subsurface. It is worth recalling that when I first proposed the idea of the Handbook Series to Elsevier in 1974 my list of titles did not even include a volume dealing with gaseous and volatile elements and compounds. The concept of expanding the original scope of this volume and providing a new focus that the words "Remote Sensing" imply, was a bold and prescient step by Professor Hale. Notwithstanding that some chapters do not deal with gases or volatile elements, it signifies a different dimension for geochemical exploration methods. The theme of all volumes in the Handbook Series is ore-finding; in this volume "ore" clearly encompasses hydrocarbons. A particular objective is that the contents are presented in such a way as to be easily understood by the practising exploration geologist as well as the specialist exploration geochemist. One of the aims of the original concept of a series of volumes, each devoted to some particular aspect of exploration geochemistry, was to allow sufficient space to examine and describe the basic underlying scientific principles of the techniques in sufficient detail to be useful as a reference source for researchers. This volume amply fulfils all objectives. Not only is it replete with case histories and "how to" information, it also has a greater emphasis on theoretical scientific principles than other volumes in the Series. This is an inevitable reflection of the relatively lower level of development of the techniques, and the complexity of controlling mechanisms of gaseous migration and long distance epigenetic dispersion of elements. Professor Hale is to be congratulated in bringing together such an impressive international team of authors. He is also to be congratulated in allowing, indeed encouraging, the inclusion of negative data. I refer especially to the chapters on mercury and helium. Both these elements have tantalised exploration geochemists for more than a generation as having all the attributes to detect deeply-buried mineralisation. The comprehensive reviews in this volume suggest that, in fact, neither has much
VI
Editor's foreword
demonstrated use in mineral exploration. This type of negative data is just as useful as strongly positive information. Notwithstanding the long gestation period of this volume, I am confident that it will nevertheless be not only a valuable guide for exploration geologists, but also the definitive source book on remotely-sensed exploration geochemical techniques for many years. G.J.S. GOVETT Pen-y-Coed, Moss Vale, NSW, Australia October 1999
VII
PREFACE
Once, all of the Earth resources needed to meet the needs of society were either clearly evident at the surface (river-bed placers) or had characteristic visible surface manifestations (gossans, oil seeps). Growth in demand for metals and fossil fuels prompted prospecting and exploration on a scale that has ensured that the endowment of such (near-)surface deposits has been discovered and evaluated in all but the most inaccessible places on Earth. The principle of using subtler non-visible clues in prospecting and exploration was taking shape in the early decades of the 2 0 th century. Its practical value was demonstrated with the introduction of new instrumental techniques (especially in chemical analysis) that were able to furnish the appropriate data. This marks the origins of what we now call mineral exploration geochemistry. It had gained widespread acceptance by mid-century and went on to account for countless new discoveries in a period of unprecedented growth in metal demand. Once again, however, we face the problem of exploration-technique exhaustion. Most deposits amenable to discovery by drainage geochemistry and soil geochemistry may well have been discovered. Innovations in analytical chemistry and geographic information systems improve data quality and data interpretability, but these represent refinements of an established technique rather than a new technique. As early as the opening decades of the 2 0 th century the petroleum industry was searching for subsurface resources that had no conventional expression at surface. The minerals industry found itself in a similar position in the closing decades. So far, for prospecting, both industries have relied mainly on well-constructed geological models and remote sensing of the subsurface of target areas by geophysical techniques, most obviously seismic surveys in petroleum exploration, conductivity and gravity surveys in mineral exploration. Alongside these, however, are thoroughly-researched and fieldtested techniques for detecting, near the surface, geochemical expressions of subsurface petroleum reservoirs and mineral deposits. Gases play an important role in this geochemical remote sensing of the subsurface. Some are indicators of major or trace components of the subsurface resource: light hydrocarbons leak from petroleum reservoirs; sulphur gases are generated by sulphide mineral oxidation; and volatile mercury is released by sulphide oxidation. Others with an indirect link to the resource act as pathfinders: radiodecay of uranium generates radon and helium; sulphide oxidation consumes oxygen and generates sulphuric acid, which
VIII
Preface
reacts with carbonate minerals to generate carbon dioxide. In other cases, gases from depth are simply carriers of trace quantities of metals collected as the gases pass through a mineral deposit. On the other hand, gases are not involved if such trace quantities of metals are transported upward by means of geoelectrochemical potentials. This volume sets out to document the techniques for geochemical remote sensing of the subsurface, to present case-history evidence of their successes and limitations, and to consider their further potential. The chapters in Part I focus on the mechanisms and models of dispersion that give rise to the patterns we attempt to detect. Those in Part II deal with the detection of dispersion pattems that owe their origins to processes (such as leakage) that are allied to resource emplacement. Those in Part III describe the detection of dispersion pattems that are generated by processes (such as radiodecay and oxidation) taking place in deposits after their emplacement. If I generalise, the particular strength and attraction of the techniques that are presented is their potential to detect a chemical signature at surface genetically-related to a parent petroleum or mineral resource in the subsurface. Their weakness is poor signal reproducibility due to a plethora of chemical, biological and meteorological factors at play in the near-surface environment. The obstacle to their wider application has been this poor signal reproducibility coupled with the lack of a universally-accepted migration model. Nevertheless, every chapter brings a fresh perspective. Radon has met with much success in uranium exploration, whilst thorough research studies on helium and mercury lead to conclusions that tend to discourage use of these gases in mineral exploration. The case for light hydrocarbons is one of compelling simplicity whilst elaborate mathematical and electrochemical models are advanced for metal migration. The volume has taken an inexcusably long time to assemble and I must register here an apology to those contributors who had quite reasonably expected earlier publication of their work. Most have shown unending patience and have even been kind enough to update their reviews; two withdrew and their work, though a loss to this volume, has appeared elsewhere. The other side of this coin has been the opportunity to include recently-drafted chapters on geoelectrochemistry. This subject has experienced something of a resurgence of interest in recent years and it gives me particular pleasure to be able to include it in this volume as a compliment to the much earlier work of the series editor. I thank, of course, the contributors and note that they represent expertise from Australia, Canada, China, the Netherlands, Russia and the USA. I thank the Intemational Institute for Aerospace Survey and Earth Sciences for resources and many individuals for assistance. In particular I thank my graduate students, John Carranza, Asadi Haroni and Alok Porwal, for helping me with the not inconsiderable task of producing the first camera-ready volume of the Handbook of Exploration Geochemistry. MARTIN HALE, Delft October 1999
IX
LIST OF CONTRIBUTORS
Charles R. M. Butt, BA (geology and chemistry), University of Keele, DIC, PhD (applied geochemistry), Imperial College, London, joined the CSIRO in Perth, Western Australia, in 1971 to initiate research in exploration procedures for deeply-weathered terrain, and worked with gas geochemistry between 1979 and 1985. He then returned his attention to regolith geochemistry, co-editing Volume 4 of the Handbook of Exploration Geochemistry on this subject, which was published in 1992. He is a Chief Research Scientist in the Division of Exploration and Mining of CSIRO and Programme Leader in the Cooperative Research Centre for Landscape Evolution and Mineral Exploration. He was an Associate Editor of the Journal of Geochemical Exploration from 1976 to 1999. Graham R. Carr, BSc, University of New South Wales, PhD, University of Wollongong. Following his early studies on the genesis of sediment-hosted massive sulphide deposits, he joined CSIRO in 1979 to research the applications of mercury geochemistry in mineral exploration. In 1983 he joined the lead isotope group of CSIRO and is currently a Principal Research Scientist in its Division of Exploration and Mining, actively involved in researching and applying isotopic exploration techniques. Willy Dyck, MSc, University of Saskatchewan. During his career with the Geological Survey of Canada, he carried out intensive research into the geochemistry and geology of uranium deposits and their detection using soil gases, radon and helium. A chemist and engineer by training, he was especially adept at devising instrumentation to sample gases, groundwaters, lake-bottom sediments and waters. Now retired to a farm where he brews beer and produces a miscellany of fruit wines, "Radon" Dyck continues to retain his old interests. Fei Qi is a graduate in petroleum geology, China University of Geosciences, Wuhan. Following a period in the geological exploration team in Hebei Province, she joined Wuhan College of Geology where she was engaged in education and research in petroleum exploration. She was appointed Vice-President of the Department of Petroleum Geology in 1986. Since 1992, she has been a senior director of the Institute of Petroleum Geology in the China University of Geosciences, Wuhan. She is an active member of the American Association of Petroleum Geologists.
X
List of contributors
Martin J. Gole completed his BA at Macquarie University, Sydney, Australia, in 1972, and initially worked as an exploration geologist before completing his PhD on Archaean banded iron formations at the University of Western Australia in 1979. He then spent 2 89 years in the USA, undertaking postdoctoral research at Indiana University and Northwest Illinois University and teaching at Georgia State University. He joined the CSIRO in 1981 to work on the use of helium in exploration and, from 1984 to 1988, worked on komatiitehosted nickel sulphide deposits. Since 1989 he has been a consultant geologist. Martin Hale, BSc (geology), Durham, was a mineral exploration geologist in central Africa before completing his PhD (applied geochemistry) at the Royal School of Mines, Imperial College, London, and subsequently entering academic life there. He is now Professor of Mineral Exploration at the International Institute for Aerospace Survey and Earth Sciences, the Netherlands, and Professor of Geochemistry at Delft University of Technology. He coedited Volume 6 of the Handbook of Exploration Geochemistry, Drainage Geochemistry, which was published in 1994. Stewart M. Hamilton, BSc (geology), Laurentian, Sudbury, MSc (hydrogeology), Carleton, Ottawa, has worked since 1984 in several fields of earth sciences including geology, hydrogeology and aqueous geochemistry. In the period 1990 to 1994 he worked as a hydrogeologist on a wide variety of geochemical and hydrogeological projects for the environmental engineering firm Jacques Whitford Ltd. He joined the Ontario Geological Survey in 1994 as an aqueous geochemist and has spent much of the last five years investigating mechanisms controlling metal mobility in thick glacial overburden. Margaret E. Hinkle, BSc (chemistry), Wayne State University, Detroit, Michigan, MS (geology), University of Michigan, Ann Arbor, Michigan, joined the US Geological Survey in 1962. From 1972 until her retirement in 1995 she worked on methodologies and applications of soil-gas geochemical surveys to geothermal and mineral resource studies. Hu Zhengqin is a geological engineer carrying out geochemical and geophysical surveys with the 814 Geochemical and Geophysical Survey Company of Huadon, China. He has taken many new initiatives in geochemistry research, including his work on thermal release of mercury, which has been widely applied in exploration in China. He has made significant contributions to many aspects of geochemistry research and published several exploration geochemistry handbooks. lan R. Jonasson, BSc, BSc (hons.), PhD (chemistry), Universities of Melbourne and Adelaide. Following research fellowships from the Nuffield Foundation (Adelaide) and the National Research Council of Canada (Geological Survey of Canada), he joined the staff of the GSC in Ottawa in 1971. During ten years in the Exploration Geochemistry section he
List of contributors
XI
was engaged in all aspects of surficial geochemistry applied to mineral exploration and to environmental contamination studies. He then transferred to the Mineral Deposits SubDivision to work on sediment-hosted base-metal sulphide deposits. His current diversions lie with the study of volcanogenic massive sulphide deposits in ancient terranes and the modem seafloor of the Pacific Ocean, where he engages in "submersible" mapping. Victor T. Jones, III, BSc (physics), University of Southwestern Louisiana, MS, PhD (physics) Texas A&M University. Upon completion of a two-year National Research Council of Canada Postdoctoral Fellowship at the Chemistry Department of the University of Western Ontario (1969-71), he initiated a career in the petroleum and minerals exploration industry as a physicist at the research laboratory of Superior Oil Company. His subsequent research activities in pathfinder techniques were extended to include hydrocarbons at the Pittsburgh laboratory of the Gulf Research and Development Company, initially as a research geochemist and ultimately as the Director of the Physical Geochemistry and Minerals section. Following 12 years of experience at major oil company laboratories, he became the manager of the Exploration Geochemistry Division at Woodward Clyde Oceaneering, before founding Exploration Technologies, Inc. (ETI) in 1984. As President and CEO of ETI, he has continued to be actively involved in the research and development of surface geochemical techniques for both exploration and environmental applications. John X Lovell has a BSc in geology from Southampton University and a PhD in applied geochemistry from Imperial College, London. He worked as a mineral exploration geologist in west Africa and Australia prior to carrying out his PhD research on vapour geochemistry in mineral exploration, for which he conducted field tests in Chile, Saudi Arabia, South Africa, Namibia, the USA, Spain, Ireland and the UK. He joined Barringer Geoservices, Colorado, in 1979 and has been involved in mineral and petroleum exploration programmes throughout the world. Martin D. Matthews, BS, Allegheny College, MS, West Virginia University, PhD (geology) Northwestern University. After being an Assistant Professor of Geology at Washington State University, he joined Gulf Research and Development Company, where he progressed to Director of Geological Research, Manager of Geochemical Research and was a Senior Staff Geologist for Gulf Oil US. He joined Texaco's Exploration and Production Technology Department as a Senior Scientist and is currently a Consulting Explorationist in Texaco's Central Exploration Department. He is also an adjunct professor in the Department of Computational and Applied Mathematics at Rice University and a member of the Earth Science Advisory Board at the Savannah River Laboratory. He has worked in surface and subsurface geochemistry, remote sensing, diagenesis, fractures, fluid flow, basin modelling, depositional systems and global cyclostratigraphy.
XII
List of contributors
Freek D. van der Meer, BSc (geology), Free University of Amsterdam, BSc (geophysics), University of Utrecht, MSc (structural geology), Free University of Amsterdam, PhD (geology and mineralogy), Wageningen University, began his professional career as a geophysicist at Delft Geotechnics, processing and interpreting ground radar survey data. In 1989 he joined the International Institute for Aerospace Survey and Earth Sciences where he researches algorithm development and geologic applications of hyperspectral remote sensing. In 1999 he was appointed to the chair of Imaging Spectrometry at Delft University of Technology. Oleg F. Putikov, Eng., Doctor of Geological and Mineralogical Sciences, began his career in 1961 by joining the Krasnokholmskaja expedition in Tashkent, Uzbekistan, to take part in logging of ore deposits. He then worked for two years on airborne geophysical surveys with the Western Geophysical Trust, Leningrad. Since 1965 he has worked at the St. Petersburg State Mining Institute (SPSMI), Russia, where he gave his attention to the theory of geothermal investigations and subsequently (since 1967) to the theory and practice of geoelectrochemistry. He is now Professor of Exploration Geophysics at SPSMI and a Chief Research Scientist in the VIRG-Rudgeofizika Institute in St. Petersburg. He is an Associate Editor of Russian Geophysical Journal. David M. Richers, BS, Pennsylvania State University, MS, University of Kentucky, PhD (geology) University of Kentucky, joined the Basin Studies Group of Cities Service Oil Company, Tulsa, Oklahoma, in 1979 as a petroleum geochemist performing work in surface prospecting and remote sensing. In 1983 he joined Gulf Research and Development Company to continue his work on surface methods and in 1985 moved to Marathon Oil Company to perform further work in surface geochemical methods, remote sensing and geologic computer applications. In 1990 he became the Assistant Director of Computer Graphics at Syracuse University and managed the Advanced Graphics Research Laboratory, developing image processing, GIS, and virtual reality methods for use in the geologic sciences. Since 1993 he has been a Principal Scientist at the Savannah River Laboratory, applying geochemical methods to solve geologic and environmental problems. Ruan Tianjian is a graduate in applied geochemistry, China University of Geosciences, Wuhan. In the early part of his career he was engaged in education and research in geochemical exploration for minerals at Beijing College of Geology. Following his appointment in 1985 as Head of the Department of Geochemistry in the China University of Geosciences, Wuhan, he has been involved in geochemical exploration for petroleum. Sun Xiangli is a graduate in geochemistry and began his career as an analyst in the laboratory of the geophysical prospecting team of the Geology Bureau of Chinese Heavy Metal Industry in 1957. He has been engineer-in-chief of geochemical exploration since 1960 and engaged in research in exploration techniques and instrument trials since 1974.
List of contributors
XIII
He began to research gas geochemical surveys for mineral exploration in 1978 and has worked on geological techniques for metallic ore, petroleum and natural gas exploration since 1987. He is a member of the Chinese Society of Metals and has been a member of the 1st and 2nd Geological Society of China Commissions on Geochemical Exploration.
Wen Baihong has BEng. and MEng. degrees in exploration geophysics from Central South University of Technology (CSUT), China, and a PhD in geology and mineralogy from St. Petersburg State Mining Institute, Russia. Between 1987 and 1994 he researched exploration for mineral resources by means of magnetic, gravity and geoelectrical studies at CSUT. From 1994 he tumed his attention to geoelectrochemistry and carried out experimental studies and physico-mathematical modelling in Russia and in China. He now applies geoelectrochemical techniques to hydrocarbon exploration for the China National Petroleum Corporation (CNPC).
John R. Wilmshurst, Dip RMIT, BSc, PhD, University of Melbourne, commenced work on exploration methods for base and precious metals within the CSIRO Division of Mineral Chemistry at North Ryde in 1972. His particular interest was the weathering of metalliferous minerals and he was involved with developing and applying an in-house mercury detector for base-metal and precious-metal exploration. He is presently working in the CSIRO Division of Petroleum Resources, developing tools for source-rock maturity estimation.
Yang Hong, BSc (geography), Peking University, began work at the Institute of Remote Sensing Applications, Chinese Academy of Sciences, as a research assistant. Her work focused on detecting hydrocarbon microseeps in the Tarim and Junggar Basins using remote sensing techniques. In 1995, she obtained her MSc in structural geology at the Intemational Institute for Aerospace Survey and Earth Sciences, the Netherlands, where she subsequently carried out her PhD research in using imaging spectrometer data to detect hydrocarbon microseepage. This research received the Merit Award of the American Association of Petroleum Geologists in 1997. She is now a remote sensing specialist at Shell International in the Netherlands.
Zhang Jianzhong, BSc, MSc (geography), Peking University, began his career as a research assistant at the Institute of Remote Sensing Applications, Chinese Academy of Sciences, using image processing for detecting hydrocarbon microseeps. In 1996, he obtained a second MSc at the International Institute for Aerospace Survey and Earth Sciences, the Netherlands, on the applications of shortwave infrared spectra of rocks to areas in China affected by coal fires. He now specialises in aerospace image processing and GIS applications at the Institute of Remote Sensing Applications.
XIV
List of contributors
Zhang Meidi is a graduate of the Geological College of Guilin, Guangxi, China. In 1976 she joined the Research Institute of Geology for Mineral Resources in Guilin to research geochemical methods of mineral exploration, including the application of pathfinder elements and gas dispersion. During the 1980s she studied heat-released CO2 as a tool for discovering mineral deposits under exotic overburden. She is now deputy general manager of the Geological Academic Embassy of the Research Institute of Geology for Mineral Resources in Guilin.
XV
CONTENTS
E d i t o r ' s foreword ............................................................................................................................. Preface .......................................................................................................................................... List o f contributors .........................................................................................................................
V VII IX
P A R T I. G E N E T I C M O D E L S OF R E M O T E D I S P E R S I O N P A T T E R N S
Chapter 1. Genesis, behaviour and detection o f gases in the crust ................................................... 3 M . Hale Introduction ...................................................................................................................
3
The g e o c h e m i c a l b a c k g r o u n d ........................................................................................
4
T h e a t m o s p h e r e .........................................................................................................
4
The soil air ................................................................................................................
6
Indicator and pathfinder gases for exploration .............................................................. 6 Gases c o n t e m p o r a n e o u s with resource e m p l a c e m e n t ............................................... 7 Gases o f post-mineralisation p r o v e n a n c e .................................................................. 7 M e c h a n i s m s o f gas migration ........................................................................................
8
Diffusion ...................................................................................................................
8
Mass flow ...............................................................................................................
11
Gas streaming .........................................................................................................
12
Indicator and pathfinder gas data acquisition ..............................................................
13
C o n c l u s i o n s .................................................................................................................
14
Chapter 2. Geoelectrochemistry and stream dispersion .................................................................
17
O. F. Putikov and B. W e n G e o e l e c t r o c h e m i c a l prospecting .................................................................................
17 17
P h y s i c o - c h e m i c a l basis ...........................................................................................
17
Introduction .................................................................................................................
Partial extraction o f metals ( C H I M ) ....................................................................... 36 Diffusion extraction o f metals ( M D E ) .................................................................... 46 O r g a n o m e t a l l i c ( M P F ) and t h e r m o m a g n e t i c ( T M G M ) patterns ............................. 49 G e o e l e c t r o c h e m i c a l exploration ..................................................................................
53
P h y s i c o - c h e m i c a l basis ...........................................................................................
53
Contact polarisation (CPC) .....................................................................................
60
Contactless polarisation ( C L P C ) .............................................................................
69
Polarographic logging (PL) .....................................................................................
73
Discussion and conclusions .........................................................................................
78
Contents
XVI
Chapter 3. Spontaneous potentials and electrochemical cells ........................................................ 81 S.M. Hamilton Introduction ................................................................................................................. Geochemical transport mechanisms ............................................................................ Diffusion ................................................................................................................. Advective groundwater transport ............................................................................ Gaseous transport .................................................................................................... Electrochemical transport .......................................................................................
81 82 82 82 83 85
Voltaic Cells ................................................................................................................ 86 Spontaneous potential in Earth materials .................................................................... 91 Measurement o f spontaneous potential ................................................................... 92 Sources of spontaneous potential ............................................................................ 95 Redox stratification in the Earth ............................................................................. 98 Spontaneous potential cells ....................................................................................... 1O0 O h m ' s law in the development of cells ................................................................ 100 Cells associated with electronic conductors in bedrock ........................................ 101 Cells in the absence o f electronic conductors ....................................................... 107 Field evidence for the presence of cells ............................................................... 112 Geochemical response to spontaneous potential cells ............................................... 113 lon mobility .......................................................................................................... 113 Geochemical anomalies ........................................................................................ 115 Conclusions ............................................................................................................... 118 PART If. R E M O T E D I S P E R S I O N P A T T E R N S OF C O - G E N E T I C P R O V E N A N C E
Chapter 4. Carbon dioxide dispersion halos around mineral deposits ......................................... 123 M. Zhang Introduction ............................................................................................................... Method ...................................................................................................................... Case histories ............................................................................................................ Discussion ................................................................................................................. Speciation of carbon dioxide ................................................................................. Formation o f carbon dioxide dispersion patterns ................................................. Factors affecting carbon dioxide anomalies .........................................................
123 123 124 127 127 130
131
Conclusions ............................................................................................................... 131
Chapter 5. Light hydrocarbons for petroleum and gas prospecting ............................................. 133 V.T. Jones, M.D. Matthews and D.M. Richers Introduction ............................................................................................................... Origin of light hydrocarbon gases ............................................................................. Origin of petroleum .............................................................................................. Origin of light hydrocarbon gases in the near-surface .......................................... Laboratory and field evidence of biogenic C2-C4 hydrocarbons ........................... Distinguishing petrogenic and biogenic hydrocarbons ......................................... History .......................................................................................................................
133 134 134 137 137 140 140
Basic concepts ...................................................................................................... 141 Methods o f geochemical prospecting .................................................................... 141
Contents
XVII Physical basis for migration o f h y d r o c a r b o n s to the surface ..................................... 143 Basic assumptions .................................................................................................143 Physical transportation by effusion ....................................................................... 144 Physical transportation by diffusion ...................................................................... 144 H y d r o c a r b o n residence sites at surface ...................................................................... 148 Free gas .................................................................................................................149 Bound gas .............................................................................................................150 Choice o f free gas or bound gas ........................................................................... 150 Factors influencing near- surface h y d r o c a r b o n flux .................................................. 152 Microbial activity ..................................................................................................152 Barometric p u m p i n g .............................................................................................152 Earthquakes ...........................................................................................................153 Sampling and m e a s u r e m e n t methods ........................................................................ 155 A t m o s p h e r i c techniques ........................................................................................ 155 Soil gas .................................................................................................................159 Dissolved gas ........................................................................................................170 Headspace gas .......................................................................................................172 Disaggregation ......................................................................................................173 Acid extraction ......................................................................................................176 Fluorescence .........................................................................................................179 Sampling strategy ...................................................................................................... 180 Data interpretation ..................................................................................................... 182 Preferential p a t h w a y model .................................................................................. 182 Geochemical populations ...................................................................................... 187 Case histories ............................................................................................................192 Neuquen Basin, Argentina .................................................................................... 192 High Island area, G u l f o f Mexico ......................................................................... 195 Great Basin, Railroad Valley, N e v a d a .................................................................. 197 Overthrust Belt, W y o m i n g - U t a h ...........................................................................208 Conclusions ...............................................................................................................211
Chapter 6. Gas geochemistry surveys for petroleum ....................................................................213 Y. Ruan and Q. Fei Introduction ...............................................................................................................213 Theoretical principles ................................................................................................214 Indicator gases ......................................................................................................216 Gas migration ........................................................................................................217 Surface expressions o f hydrocarbon migration ......................................................... 218 Gas anomalies .......................................................................................................218 Alteration ..............................................................................................................218 M o d e s o f occurrence o f gases in microseeps ............................................................. 219 Free molecules ......................................................................................................220 Adsorbed molecules ..............................................................................................220 Microbubbles ........................................................................................................221 Mineral constituents ..............................................................................................221 Practical methods ......................................................................................................222 Soil air ..................................................................................................................222
XVIII
Contents Soil .......................................................................................................................223 Case histories ............................................................................................................225 Ordos Basin ..........................................................................................................225 Lixian Depression ................................................................................................. 229 Conclusions ...............................................................................................................231
Chapter 7. Aerospace detection o f hydrocarbon-induced alteration ............................................ 233 H. Yang, F.D. Van der M e e r and J. Z h a n g Introduction ...............................................................................................................233 H y d r o c a r b o n m i c r o s e e p a g e .................................................................................. 233 Induced surface manifestations o f m i c r o s e e p a g e .................................................. 234 R e m o t e detection o f induced surface manifestations ................................................ 235 Bleached red beds .................................................................................................236 Kaolinisation .........................................................................................................238 Carbonate enrichment ........................................................................................... 238 Vegetation stress ...................................................................................................240 Other anomalies ...................................................................................................244 Problems and future trends ........................................................................................ 244 P A R T III. R E M O T E D I S P E R S I O N P A T T E R N O F P O S T - G E N E T I C P R O V E N A N C E
Chapter 8. Sulphur gases ............................................................................................................. 249 M.E. Hinkle and J.S. Lovell Introduction ...............................................................................................................249 C h e m i s t r y and g e o c h e m i s t r y o f sulphur gases .......................................................... 250 Experimental techniques ...........................................................................................256 Sample collection .................................................................................................256 Analysis by gas c h r o m a t o g r a p h y .......................................................................... 259 Other methods o f analysis ....................................................................................267 Reference materials ..............................................................................................267 S a m p l e storage and preparation for analysis ......................................................... 269 Degassing soils and molecular sieve adsorbents ................................................... 269 Injection o f samples ..............................................................................................270 Recording analytical results .................................................................................. 270 Case histories ............................................................................................................270 Johnson Camp, Arizona ........................................................................................271 North Silver Bell, Arizona ....................................................................................283 Crandon, Wisconsin ..............................................................................................285 Kazakhstan ...........................................................................................................286 Ireland ...................................................................................................................286 Discussion .................................................................................................................287 Conclusions ...............................................................................................................288
Chapter 9. Sulphide anions and compounds ................................................................................ 291 X. Sun Introduction ...............................................................................................................291 Experimental investigations ......................................................................................291
Contents
XIX Soil adsorption o f h y d r o g e n sulphide ................................................................... 291 H y d r o g e n sulphide transport through soil ............................................................. 293 R e d o x conditions for h y d r o g e n sulphide persistence ............................................ 294 C o n c l u s i o n s o f experimental investigations ......................................................... 295 Field investigations ................................................................................................... 295 Mineralisation beneath thick transported o v e r b u r d e n ........................................... 297 Mineralisation beneath thick lithic cover ..............................................................
299
Mineralisation beneath m i x e d e l u v i u m and transported s e d i m e n t ........................ 300 Discussion ................................................................................................................. U n d e t e c t e d mineralisation ....................................................................................
301 301
False anomalies ....................................................................................................
302
C o n c l u s i o n s ...............................................................................................................
302
C h a p t e r 10. H e l i u m ......................................................................................................................
303
C.R.M. Butt, M.J. G o l e and W. D y c k Introduction ...............................................................................................................
303
O c c u r r e n c e ................................................................................................................
304
D i s c o v e r y ..............................................................................................................
304
A b u n d a n c e and origin ...........................................................................................
304
A b u n d a n c e in the Earth and atmosphere ...............................................................
305
Isotope ratios ........................................................................................................
306
Properties and migration ....................................................................................... S a m p l i n g ...................................................................................................................
307 310
Soil and overburden gases ....................................................................................
310
Soils ......................................................................................................................
311
Waters ...................................................................................................................
312
A n a l y s i s ....................................................................................................................
313
Mass s p e c t r o m e t r y ................................................................................................ Gas c h r o m a t o g r a p h y .............................................................................................
313 315
Portable helium analysers .....................................................................................
316
Determination o f helium isotope ratios .................................................................
316
Analysis o f waters ................................................................................................
316
Variations o f helium concentrations .........................................................................
317
Soil and overburden gases ....................................................................................
317
G r o u n d w a t e r s .......................................................................................................
319
Biological activity and soil gas composition ........................................................ 319 H e l i u m surveys in mineral exploration .....................................................................
320
Rationale ...............................................................................................................
320
H e l i u m in gases and soils .....................................................................................
321
Soil and overburden gas surveys ......................................................................
321
Soil surveys ......................................................................................................
326
Discussion o f soil-gas and soil survey techniques ............................................ 327 Helium in waters ...................................................................................................
328
G r o u n d w a t e r surveys in u r a n i u m - m i n e r a l i s e d areas ......................................... 328 G r o u n d w a t e r s u r v e y s in n o n - m i n e r a l i s e d areas ................................................ 331 Surface water s u r v e y s .......................................................................................
336
Discussion o f water survey techniques ............................................................. 338
XX
Contents Helium surveys in petroleum exploration ................................................................. 338 Helium surveys in geothermal resource exploration ................................................. 343 Helium in thermal waters and gases ...................................................................... 343 Helium surveys for geothermal resources ............................................................. 344 Helium associated with faults .................................................................................... 346 Faults as secondary sources o f helium .................................................................. 346 G r o u n d w a t e r surveys ........................................................................................ 346 Soil gas surveys ................................................................................................ 348 Helium monitoring in earthquake prediction ........................................................ 349 Conclusions ............................................................................................................... 350 Migration o f helium in the near-surface environment ........................................... 350 Application o f helium surveys .............................................................................. 352
Chapter 11. Radon ........................................................................................................................ 353 W. Dyck and I.R. Jonasson Introduction ............................................................................................................... 353 Physical and chemical properties o f radon ................................................................ 354 Definitions ................................................................................................................. 356 Geochemistry o f radon .............................................................................................. 357 Concentrations o f radon and radium in natural environments ............................... 358 Disequilibrium in the uranium decay series .......................................................... 365 Emanation and mobility o f radon .......................................................................... 367 Analytical methods .................................................................................................... 382 Principles o f methods ............................................................................................ 383 Instantaneous mode ............................................................................................... 384 Semi-integrating mode .......................................................................................... 386 Fully-integrated mode ........................................................................................... 388 Field methods ............................................................................................................ 389 Determination o f radon in natural waters .............................................................. 389 Determination o f radon in soil emanations ........................................................... 391 Comparison studies and case histories ...................................................................... 392 Future needs .............................................................................................................. 394 Chapter 12. Mercury ..................................................................................................................... 395 G.R. Cart and J.R. Wilmshurst Introduction ...............................................................................................................
395
Geochemistry o f mercury .......................................................................................... 396 Behaviour o f mercury in the primary environment ................................................... 399 Low temperature epithermal base-metal deposits ................................................. 399 Volcanogenic massive sulphide deposits ............................................................... 400 Sediment-hosted massive sulphide deposits ......................................................... 402 Gold deposits ........................................................................................................ 403 The role o f metamorphism .................................................................................... 404 Behaviour o f mercury in the secondary environment ................................................ 406 Outcropping mineral deposits in dry climates ....................................................... 406 Outcropping mineral deposits in wet climates ...................................................... 412 Blind and buried mineral deposits in dry climates ................................................ 417
Contents
XXI Blind and buried mineral deposits in wet climates ................................................ 422 Sampling m e d i a .........................................................................................................427 A t m o s p h e r i c air ....................................................................................................427 Soil gas .................................................................................................................430 Soil ........................................................................................................................431 C o m p a r i s o n o f soil and soil gas ............................................................................433 R o c k s ....................................................................................................................434 R e c o m m e n d e d analytical procedures ........................................................................435 Conclusions ...............................................................................................................437
Chapter 13. Discrimination o f mercury anomalies ....................................................................... 439 Z. Hu Introduction ...............................................................................................................439 Method ......................................................................................................................440 Case histories ............................................................................................................440 Traverses over known mineral deposits ................................................................440 Regional traverse ..................................................................................................444 Regional grid ........................................................................................................445 Discussion .................................................................................................................447 Conclusions ...............................................................................................................448
Chapter 14. Oxygen and carbon dioxide in soil air ...................................................................... 451 J.S. Lovell Introduction ...............................................................................................................451 O x y g e n and carbon dioxide in the subsurface ........................................................... 452 Sampling and analytical methods ..............................................................................457 O x y g e n and carbon dioxide analysers ................................................................... 457 Orstat gas analyser ................................................................................................459 Draeger tubes ........................................................................................................459 Gas c h r o m a t o g r a p h y and mass spectrometry ........................................................ 460 Case histories ............................................................................................................461 Russia ....................................................................................................................461 Azerbaijan .............................................................................................................462 K y r g y z s t a n ............................................................................................................463 N a m i b i a ................................................................................................................463 Johnson Camp, Arizona ........................................................................................464 Colorado Plateau, Arizona ....................................................................................466 Discussion .................................................................................................................468 Conclusions ...............................................................................................................469 References ....................................................................................................................................471 A u t h o r index .................................................................................................................................513 Geographical index .......................................................................................................................529 Petroleum and mineral deposit index ............................................................................................533 Subject index ................................................................................................................................537
This Page Intentionally Left Blank
P A R T 1.
G E N E T I C M O D E L S OF REMOTE DISPERSION PATTERNS
This Page Intentionally Left Blank
Geochemical Remote Sensing of the Subsurface Edited by M. Hale Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) @2000 Elsevier Science B.V. All rights reserved.
Chapter 1
GENESIS, BEHAVIOUR AND DETECTION OF GASES IN THE CRUST
M. HALE
INTRODUCTION The search for minerals can be traced back for millennia and the search for petroleum for more than a century. Despite this difference, the early days of both relied on clear surface manifestations of the commodities sought in the near-subsurface. Over the course of time, countless gossans and oil seeps have acted as the spur for the discovery of resources at depth. Ultimately all such surface indications are exhausted, and new clues are needed if further subsurface resources are to be discovered. Enter remote sensing, the science of gathering data describing distant objects. Geophysical techniques have contributed a wealth of data for suggesting the presence of mineral deposits and petroleum (including natural gas) accumulations in the subsurface, and are set to continue to be vital exploration tools. There has been considerable scientific interest in extending geochemical methods to the acquisition of data that aid the search for deep mineral deposits and petroleum reservoirs. This involves studying the genesis and geochemical behaviour of elements and compounds that are naturally associated with these resources at depth and are able to migrate to the surface. This chapter considers those elements and compounds that are gases at ambient temperatures. Models of the dispersion of less volatile species are put forward in Chapters 2 and 3. Gases exhibit a high degree of geochemical mobility and their dispersion is unconstrained by gravity. These dispersion characteristics represent a potentially powerful combination of attributes in exploration. If mineral deposits or petroleum accumulations are judged or can be shown to liberate a gas into a porous medium such as overlying rock, overburden or soil then, in the simplest case, the gas will form a broad spherical halo. According to Oakes (1984) the parameters of such a halo are described by the formula:
p - 2m / 37tr3p where a mass of gas m, released into a medium of porosity p, produces a dispersion halo of radius r with a mean partial pressure O 9As the value of r increases, that of p decreases, and vice-versa. The resulting hemisphere of gas above the source is a particularly appealing
4
M. Hale
exploration guide because it is potentially broad in extent and occurs at absolute elevations above its source. The gaseous constituents of the Earth are of course not conf'med to gases exuded by exploration targets. Gases have accumulated near the outer rim of the Earth and tend to occur in mixtures, the constituents reacting with one another, with solids and liquids with which they come into contact, and responding to changes in pressure and temperature. The contribution of mineral deposits and petroleum accumulations to these mixtures is extremely small. Any attempt to recognise them with the confidence required for exploration investment prompts careful consideration of the occurrence and behaviour of not only those gases that might prove suitable for exploration but also the gas mixtures in which they have to be detected. A further complication is that the ideal dispersion hemisphere of a gas is prone to distortion. The source may not liberate gas uniformly over time, producing fluctuations in m. The rock and overburden column above the source may comprise lithologies of variable porosity, which may be cut by faults and fractures, and these various voids may be (partially) occupied by liquids, thus producing several different values of p in the column. The voids themselves may be occupied at different times by liquid (usually water) or by gas (usually soil air) of variable barometric pressure, with the result that the capacity of the voids to disperse gases from depth changes with time.
THE GEOCHEMICAL BACKGROUND
The atmosphere The most widespread gas mixture in the Earth is the atmosphere. The atmosphere is estimated to weigh 5.1 x 10 ~5 tonnes. It comprises 0.016% of the mass of the combined crust, hydrosphere and atmosphere, and less than 0.0001% of the mass of the whole Earth (Henderson, 1986). Despite being poorly represented as a proportion of the composition of the Earth, or even its outer shell, the atmosphere is omnipresent at the surface of the crust, and partially permeates it, occupying faults, pores and other voids in rocks and overburden. This is the very zone in which almost all exploration takes place. Consequently the atmosphere is a major part of the geochemical background against which gases employed in exploration must be recognised. The atmosphere is a physically and chemically dynamic system. A mixture of primeval gases was expelled to the outermost shell of the Earth during its exothermic accretion. Over geologic time the chemical composition of this mixture has changed, mainly as a result of photochemical dissociation, interaction with water and its dissolved constituents (e.g., the oxidation of Fe 2§ to Fe 3+ during the formation of the Proterozoic banded iron formations) and biogenic processes (especially photosynthesis). The composition that the atmosphere has now reached is shown in Table 1-I. Although at least 17 gases are regarded as
Genesis, behaviour and detection of gases in the crust
constituents of the present atmosphere, 99.96% of its overall composition is made up of only t h r e e - N2, O2 and Ar. The atmosphere is still experiencing continual change in its bulk composition. For the most part these bulk changes are on such a small scale, or are so slow, that they are of no consequence in exploration. Localised changes to the flux of gases in the near-surface crust are more pronounced and of much more significance to the application of gas geochemistry to exploration. Continuing out-gassing of the mantle brings gases to the surface through conduits such as deep faults and volcanic vents. These gases include HE, He, At, N2, O2, CH4, CO, CO2, SO2, H2S, S:, COS, HF, HC1, CHaC1, CHaBr and CHaI.
TABLE 1-I Composition of the atmosphere Gas Nitrogen Oxygen Argon Carbon dioxide Neon Helium Methane Krypton Hydrogen Nitrous oxide Sulphur dioxide Xenon Ozone Nitrogen dioxide Ammonia Carbon monoxide Iodine
Formula N2 02 Ar CO2 Ne He CH4 Kr H2 N20 SO2 Xe 03 NO2 NH3 CO I2
ppm by volume 780840 209460 9340 330 18.18 5.22 2.0 1.14 0.5 0.5 0-1 0.087 0.05 0-0.02 0-trace 0-trace 0-trace
ppm by weight 755100 231500 12800 460 12.5 0.72 0.74 2.9 0.035 0.8 0.36 -
Although the bulk composition of the atmosphere may be experiencing barely perceptible changes, some of its physical characteristics, such as temperature, pressure and turbulence, fluctuate locally with periods varying from annual to diurnal or even shorter. At ground level these properties greatly influence the degree and rate of change of atmospheric aeration of the pore voids in the uppermost layers of the lithosphere.
6
M. Hale
The soil air
Some rocks have a natural porosity and, at the interface of the atmosphere and the lithosphere and given the presence of moisture, all rocks tend to weather to a relatively porous soil. In the simplest case, the pore spaces of soil are occupied by atmospheric air, but the very moisture that enhances soil formation is also highly supportive of flora and fauna, which interact with the air in the pores and modify its composition. Perhaps the most obvious way in which the composition of soil air differs from that of atmospheric air is through plant respiration, which reduces the 02 content of the soil air and raises the CO2 content. Gases almost absent from the atmosphere are added to the soil air by biogenic activity. According to Enhalt (1974), 80% of CH4 in soil air is of recent biologic origin. Most H2S is biogenic (Schlegel, 1974), resulting from the bacterial reduction of sulphate under anaerobic conditions. Both CH4 and H2S are only meta-stable in the soil air, but biogenic activity generates them on a more-or-less continuous basis, so at any time they may be present in significant concentrations. Gases migrating from depth are also constituents of the soil air, their supply to the soil air varying according to proximity to sources and conduits. Sources of these gases include, but are not exclusively, mineral deposits and petroleum accumulations. The resulting soil air, unlike atmospheric air, has no fixed or stable composition. It is, however, generally regarded as carrying the most distinct expression of gases escaping from mineral deposits and petroleum accumulations; once such gases escape to the free atmosphere they experience extremely rapid dilution. Thus the soil air is an important sampling medium for gases used in exploration, but the diversity of its source gases and its variable physical properties induced by changes in atmospheric aeration make it a difficult medium in which to obtain reproducible measurements. Nonetheless, it is in this variable background that most samples and measurements for gas geochemical remote sensing of the subsurface are acquired.
INDICATOR AND PATHFINDER GASES FOR EXPLORATION There are no gases uniquely associated with mineral deposits or petroleum accumulations. Even those that are perhaps the most obvious indicator gases (sulphur gases and volatile hydrocarbons) can sometimes be generated by biogenic reactions in the soil. A few gases, notably Hg and Rn, have the advantage of being naturally concentrated in or associated with ore minerals whilst playing only a minor role in biogenic activity and occurring in only trace amounts in the atmosphere. At the other extreme, widespread gases such as CH4, CO2, He and 02 have exploration value, but their anomalous concentrations have to be recognised against their relatively high background partial pressures in the atmosphere.
Genesis, behaviour and detection of gases in the crust
7
Gases contemporaneous with resource emplacement
Some gases are physically trapped in mineral deposits and petroleum accumulations at depth but escape in trace quantities and migrate to the surface. The emplacement of many hydrothermal mineral deposits is accompanied by the introduction of large quantities of CO2 into the surrounding host rocks. Much of this CO2 is either trapped in fluid inclusions or incorporated into carbonate minerals. Its detection may act as a guide to the presence of the mineral deposit with which its introduction was associated (Chapter 4). Petroleum and natural gas accumulations require a physical trap for their preservation. Such traps are rarely gas-tight and the more volatile hydrocarbons (indeed, m some cases, heavier hydrocarbons) may escape to the surface, producing microseeps. Attempts to detect light hydrocarbon microseeps in the 1930s mark the origins of gas geochemistry. Progressive sophistication has yielded techniques to chartacterise effectively microseeps both onshore and offshore (Chapter 5). Regional surveys involving the determination of light hydrocarbons adsorbed onto soil have contributed to successful petroleum prospecting (Chapter 6). These light hydrocarbons are the near-surface expression of a flux of gas leaking from the reservoir and creating towards the surface a reduction chimney in an otherwise aerated and oxidising environment. The effects induced m rocks and vegetation can sometimes be detected from satellites (Chapter 7).
Gases o f post-mineralization provenance
Many metalliferous mineral deposits formed at depth are in the reduced state. Where they interface with the near-surface oxidising environment, there is considerable chemical reactivity. This typically takes the form of sulphide oxidation, which includes the generation of several meta-stable sulphur gases that have been shown to be useful in mineral exploration (Chapter 8). Incompletely oxidised sulphide anions and compounds are transported away from mineral deposits at depth by the groundwater, and can be mapped at surface as dispersion patterns of H2S (Chapter 9). Uranium deposits, by virtue of their radiogenic constituents, present a special case in mineral exploration. Many of the disintegrations in the radiodecay chains of U and Th liberate alpha particles (He nuclei). These are quickly stabilised as atoms of He gas, making He a potential guide to U (Chapter 10). Amongst the daughter elements in the radiodecay chains of U and Th, the only gas is Rn. Owing to a combination of its conveniently limited half-life and relative ease of measurement, Rn has been used extensively as a guide to U (Chapter 11). Strictly speaking it is a guide only to its immediate parent, Ra, which may have become geochemically separated from earlier members of its radioactive decay series, including U. Just as trace constituents of mineral deposits can act as conventional geochemical pathf'mders, trace volatile constituents are potentially gaseous pathfinders. Some sulphide minerals, in particular sphalerite, accommodate trace quantities of Hg. When liberated into
8
M. Hale
the atomic state by sulphide oxidation, this Hg has a very high vapour pressure. In addition, Hg readily lends itself to analytical detection at extremely low concentrations, and so it has been widely used as a gaseous pathf'mder in mineral exploration (Chapter 12). Unique to Hg is the possibility of discriminating between anomalies derived from mineralisation and those of anthropogenic origin (Chapter 13). Finally, the very process of sulphide oxidation at depth can provide geochemical signals at surface. Sulphide oxidation consumes O2 that is ultimately drawn from the aerated rock and soil voids in the immediate vicinity. The chemical reactions of oxidation create a low pH environment in which any carbonate minerals break down with the liberation of CO2, some of which finds its way into the neighbouring rock and soil voids. Thus anomalous concentrations of O2 and CO2 in the near-surface soil air provide an indication of oxidising mineralisation at depth (Chapter 14).
MECHANISMS OF GAS MIGRATION The way in which gases migrate after their generation has a bearing on the detection of indicator and pathfinder gases and the rate at which they experience dilution in background gas mixtures such as the atmosphere. For migration in the gas phase, diffusion and mass flow are both well-established mechanisms and each has a role to play in gas transport in the crust. Their relative contributions to gas migration at different positions in the crust is less certain. A more controversial mechanism, applicable to gas transport through groundwater, is gas streaming. Finally, most gases are to some extent soluble in groundwater and may experience dispersion in solution.
Diffusion Diffusion is the most fundamental mechanism of gas migration in that it requires only a partial pressure (concentration) gradient. The rate of diffusion of a gas is then determined by the medium in which diffusion takes place, its temperature and absolute pressure, and the diffusion coefficient of the gas. The diffusion coeffiecient is a function of molecular weight, the shape of molecules, and their intermolecular attraction. Every gas thus has a different diffusion coefficient. The influence of the medium in which gas diffusion occurs is related to the density of the medium: gases diffuse less quickly through a solid than through another gas. The rate at which a gas diffuses in a specified medium is sometimes termed its diffusivity. In rocks and soils, the only appreciable diffusion of gases occurs in the voids or pores, which may be occupied by air, water or a mixture of both. Migration over any appreciable distance is possible only if the soil pores are continuous with each other. Collisions of the gaseous molecules with liquids or solids impede their progress, so that diffusion in a porous medium is slower than in a free space. The important factors are the shape, size, tortuosity
Genesis, behaviour and detection of gases in the crust
9
and unevenness of the pores, the shape, orientation and size distribution of the solid phase and the degree of water-saturation of the system. Experimental investigations have determined the diffusivity of different gases in various porous soil and overburden materials. The periods calculated for gases to travel a particular distance vary considerably (Table 1-II). Mercury vapour diffuses in 15 days through 10 m of sand whereas 5.7 years are required for Kr to pass through 10 m of fine-grained playa sediments. TABLE 1-II Diffusion rates for gases through porous overburden Gas Kr Rn Hg Hg
Overburden Playa sediment Desertsoil Clay Sand
Apparent diffusion coefficient (cm 2 S-1) 0.00387 (Robertson, 1969) 0.036 (Tanner,1964a) 0.05 (Ruan et ai., 1985a) 0.56 (Ruan et al., 1985a)
Transit time, 10m 5.7 years 225 days 162 days 15 days
In an ideal case a source at depth liberating a gas into a homogeneous porous medium that, at some distance from the source, is open to the atmosphere establishes by diffusion a hemispherical halo in the porous medium. The time taken to establish the halo varies up to many years, depending upon the thickness of the medium and the diffusivity of the gas in it. Once established, however, the halo is persistent provided the supply of gas from the source is maintained. Gas concentration in the hemispherical halo falls with increasing distance from the source such that, near to its intersection with the ground surface, the halo presents as a broad symmetrical zone with peak concentrations directly above the source. To compare this ideal case with more complex settings, Ruan et al. (1985b) employed numerical modelling techniques based on the alternating direction method for the solution of finite difference equations. Their results allow comparison of the shapes of halos of the same gas diffusing through media of different homogeneity from sources of different sizes (Fig. 1-1). The ideal hemisphere (Fig. 11A) is perturbed whenever the diffusing gas comes into contact with a medium of different porosity. The vertical boundaries of the model imply contact with rocks of zero permeability bounding a porous medium in which gas is diffusing. When the gas flux is sufficiently strong for gas to reach these contacts, gas is then channelled upward (Fig. 11B). Horizontal boundaries in the model separate media of different porosity. Where a low-porosity medium (e.g., clay) lies above a more porous medium (e.g., sand) gas diffuses from a source in bedrock with relative ease until it reaches the inter-layer boundary, along which it can more easily migrate laterally than vertically (Fig. 1-1C). Preferential lateral gas migration might proceed as far as an impermeable medium, at the boundary of which the gas largely retained in the more porous medium at depth is diverted upward (Fig. 1-1D). In this case, gas does not reach the surface directly above
10 A
M. Hale B
[3
F
Fig. !-I. Gas plumes established by diffusion in porous media above a gas source: (A) small source, homogeneous medium; (B) large source, homogeneous medium, impermeable vertical boundaries; (C) small source, medium of low porosity overlying medium of high porosity; (D) large source, medium of low porosity overlying medium of high porosity, impermeable vertical boudaries; (E) small source, medium of high porosity overlying medium of low porosity; (F) large source, medium of high porosity overlying medium of low porosity, impermeable vertical boundaries (from Ruan et al., 1985a).
the source. Where a high-porosity medium (e.g., sand) lies above a less porous medium (e.g., clay) the gas flux reaching the boundary is weak and, once dispersed in the more porous upper medium, may barely be detectable (Fig. 1-1E). Even a large source emitting gas into porous media bounded by impermeable media yields only a broad weak dispersion halo near the surface (Fig. I-IF). These numerical modelling results are supported by Hg data from in vitro experiments by Ruan et al. (1985a) and by a range of field observations. For example, Ball et al. (1983b) note that anomalous concentrations of 02, CO2 and Rn in soil air over a fault-hosted sulphide mineral deposit in England tend to occur over the steep vertical walls to the mineralisation and in juxtaposition to the boundary faults. Diffusion represents an important mechanism for gas migration in the porous uppermost crust and seemingly produces interpretable near-surface dispersion pattems. At depth in rocks of much lower porosity, however, diffusion rates are likely to be exceedingly slow, bringing into question the significance of the contribution of diffusion
Genesis, behaviour and detection of gases in the crust
11
to development of gas dispersion patterns. At the other extreme, close to the ground surface and along conduits in the upper crust, mass flow plays a prominent role.
Mass flow
Whereas diffusion of gas at depth is widespread, mass flow is often localised (near the ground surface, in faults) or intermittent (volcanic eruptions). Mass flow is an important consideration in the application of gas geochemistry to mineral because of its significant role in the interchange of atmospheric air and soil air, and therefore its influence on gas composition in the shallow subsurface from where most samples and measurements are taken. Lovell (1979) reviewed soil aeration in this context. Mass flow through a porous medium is influenced by the porosity of the medium in much the same way as diffusion is influenced by porosity. Thus, mass flow proceeds faster in a high-porosity sand than in a low-porosity clay. In addition, many of the physical properties of atmospheric air influence the aeration of soil and porous overburden by mass flow. Baver et al. (1972) estimate their contributions (Table 1-III). In regions that experience regular precipitation, rain draining downward through the soil induces most of the mass flow of gases in soils. In extreme cases water may dislodge soil air from the soil pores themselves, but typically water displaces soil air from the inter-crumb matrix into neighbouring macropores (Currie, 1960). Seasonal rises in the water table might displace soil air upward. Variations in saturation of the soil also affect the extent to which gases are dissolved in soil moisture.
TABLE 1-III Mechanisms of soil aeration Source of soil aeration Rainfall flushing of soil air Barometric pressure variations Wind action Temperature gradient, soil-atmosphere Diurnal temperature variation in soil Total mass flow Balance (mainly diffusion) Total
Percentage of total air exchange 6.25-8.3 1 0.1 0.2-0.4 0.13 7.68-9.93 90.07-92.32 100
Aeration of the soil due to absolute pressure changes also leads to mass flow of gases in soil. Continuous meteorological pressure variations in the atmospheric air above the soil are the principal driving force. This barometric pumping causes atmospheric air to
12
M. Hale
flow into the soil in response to a pressure increase and to leave the soil in response to a fall in pressure. These pressure changes are relatively slow and the effects in the soil tend to show little detectable time lag. An increase in barometric pressure compresses downward the soil air originally occupying the pores whilst a decrease in barometric pressure induces egress of soil air into the atmosphere. Turbulent wind blowing as gusts across the surface of soil produces slight but numerous changes in pressure and adds to the pumping effects of longer-amplitude meteorological pressure changes. Wind speed has been shown to influence the rate of loss of water vapour from soil (Acharya and Prihar, 1969) and the same is likely to apply to the rate of loss of other gases. Temperature affects the volume that air occupies and hence its pressure. Diurnal temperature variations are rapid but confined to the near-surface zone; seasonal variations are more pervasive. In the near-surface, where gas geochemistry samples and measurements are acquired, mass flow is a source of background variations that tend to obscure any signal arriving from depth. The interplay of the many different causes of variation has proved a serious impediment to the provision of interpretable gas data in exploration and this has prompted a number of field investigations (Hinkle, (1990). In comparatively elaborate studies, Klusman and Webster (1981) and Klusman and Jaacks (1987) monitored many of the sources of variation along with emissions of Hg, Rn and He. By stepwise multiple regression they found that air temperature, soil temperature, barometric pressure, relative humidity and soil moisture exerted most influence on gas concentrations. However, even if such monitoring could be used for gas data noise reduction, it is not practical to monitor so many sources of variation as part of an exploration programme. Rather, in practice, the problems tend to be alleviated by sampling as far as possible below the ground surface and/or integrating the signal over a considerable period of time.
Gas streaming The relatively slow gas diffusion rates in rocks of low porosity at depth have brought the contribution of diffusion to long-distance gas migration into question. The half-life of Rn is so short that its persistence and detection after transport by diffusion over tens or hundreds of metres is extremely unlikely. Kristiansson and Malmqvist (1982) and Malmqvist and Kristiansson (1984, 1985) hypothesise that, in the zone of saturation, pressure gradients and pressure shocks cause over-saturation, leading to the formation of gas bubbles. These stream upward at a comparatively rapid rate until they reach the water table and mix with the soil air. The resulting mixture is then driven slowly further upward by the pressure gradient caused by the bubble stream. Any gas that dissolves in groundwater could, given the appropriate conditions, migrate by streaming. Groundwater is most likely to be saturated in gases dissolved in meteoric water, i.e., N2, 02, Ar, CO2. These then are the gases from which bubble streams may form.
Genesis, behaviour and detection o f gases in the crust
13
By their very provenance, these are not gases indicative of mineral deposits or petroleum accumulations. Rather a bubbles acts as a carrier for atoms of other elements which attach to the surface of the bubble. The atoms that attach to bubble surfaces include not only indicator and pathfinder gases such as Rn but also non-gaseous species such as metals. Streams of gas bubbles therefore have the capacity to deliver to the near surface minute geochemical samples from considerable depth.
INDICATOR AND PATHFINDER GAS DATA ACQUISITION By virtue of their physical state, dispersion halos of indicator and pathf'mder gases are difficult to measure compared with dispersion pattems in solids and liquids. The halos are formed by gases migrating upwards from depth, and these usually need be intercepted before they experience catastrophic dilution in the open atmosphere. The near-surface soil suggests itself as being the most accessible medium in which to detect the dispersion halo, although its atmospheric aeration and biogenic activity create undesirable levels of background noise. Procedures that have been devised for making measurements of gas dispersion halos may be initially divided according to the measurement substrate, for example, atmospheric air, free soil air or adsorbed gas. The period over which the sample is accumulated is an additional important consideration, because it has a bearing on the representativity of the measurement. The atmospheric air immediately above the ground surface is clearly a convenient medium in which to obtain measurements of gases emanating from depth, but the likelihood of catastrophic dilution of the signal is very large. Limited success was achieved with a vehicle-mounted Hg detector which collected large atmospheric air samples whilst on the move. More success has been achieved by taking advantage of the exceptional olfactory sense of dogs. Their use in prospecting, however, has been confined to detecting concealed sulphide-bearing boulders in glacial dispersion trains (Kahma, 1965; Nilsson, 1971; Brock, 1972), and their capabilities do not seem to be readily translated into an instrumental technique. Measurements made on free soil air are obtained on samples extracted through probes. A probe can be driven manually to a depth of 1-2 m below the surface and soil air extracted through it with a hand pump. If entrainment of atmospheric air is suspected, holes can be drilled mechanically to greater depths and sealed well below the surface; soil air is drawn out after the hole has equilibrated with the surrounding soil air. The resulting soil air sample may be passed directly to a portable analytical instrument or may be trapped for analysis later. On-site measurement systems range from back-pack insmmaents (e.g., for Rn, O2, CO2) to a vehicle mounted mass spectrometer (McCarthy and Bigelow, 1990). The fieldwork requirement can be reduced if measurement of gas concentration is performed at a field or central laboratory. This can be achieved by transporting samples of soil air in gas-tight containers, or by selectively depositing the gas of interest onto a convenient substrate (e.g., Hg vapour onto Au film).
14
M. Hale
Samples obtained through probes reflect the soil air composition at a particular time and the composition of soil air is prone to fluctuation. Other soil air sampling methods take advantage of a time-integrated measurement of the soil air flux by leaving a simple collection device at the sample site for a period of days or weeks. Inverted cups placed just under the surface have proved the most popular design. An adsorber (e.g., activated charcoal) or detector (e.g., film that is scarred by particles emitted through radiodecay) fixed in the uptumed base of the cup effectively collects or records the amount of one or more gases that find their way into the cup from the underlying soil air. After the cup is recovered from the sample site, quantitative measurement is carried out in a laboratory. Active surfaces on soil particles are able to adsorb some of the gases with which they come into contact. These surfaces are normally in equilibrium with the contents of the pores that surround them and their adsorbed gas concentrations are therefore representative of the gas concentration in the pores. Soil samples are a particularly convenient medium for collection and transport, but they must be treated with care to avoid losses or additions of gases during transport and storage. After transport to a laboratory, gases are introduced into an analytical instrument for quantitative determination of the constituents of interest. Soil air in a container is introduced directly to the instrument, whilst adsorbed gas is released by thermal of chemical desorption. The instrumental methods most widely used for gas analyses include gas chromatography, mass spectrometry and atomic absorption spectrophotometry. For quantifying the radiation scars on film, image analysis methods are employed. Gas concentration measurements are most usually reported as a volume ratio, that is, the volume of the measured gas as a fraction (typically ppm) of the volume of the gas mixture on which the measurement was made. Since, by virtue of the ideal gas law, equal volumes of any gas at constant temperature and pressure contain equal numbers of molecules, the volume ratio is also the molecular ratio. If necessary, the weight of gas can be obtained from the relation that one mole occupies 22.4 litres at 0~ and 1 atm. When gas concentration measurements are made by soil desorption, they are more conveniently reported as a weight ratio. Radioactive gases are usually quantified in terms of "counts" of radio-decay events, and more rarely in terms of the curie, which is the amount of the radioactive element that produces 3.7 x 10 ~~disintegrations per second.
CONCLUSIONS Gases commonly occupy the pore voids m rocks, overburden and soil. Elements existing as (components of) gaseous molecules possess, in principle, a high degree of geochemical mobility compared to elements in solids and liquids. However, the ways in which gases experience dispersion in the subsurface natural environment are more diverse and less well characterised than mechanisms of dispersion in the solid and liquid phases. Nevertheless, the application of gases in the search for deeply-buried resources is attractive.
Genesis, behaviour and detection of gases in the crust
15
This volume goes on to review research investigations and case history studies of every gas that might act as an indicator or pathfinder and to elaborate the models developed to explain the resulting observations. A pervasive source of ambiguity in the interpretation and understanding of the data of gas geochemistry proves to be the magnitude and diversity of background variations. Only where these are overcome effectively can the detection of trace quantifies of gases derived from mineral deposits and petroleum accumulations at depth provide a reliable means of geochemical remote sensing of the subsurface.
This Page Intentionally Left Blank
Geochemical Remote Sensing of the Subsurface Edited by M. Hale Handbook of Exploration Geochemistry, VoL 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
17
Chapter 2
GEOELECTROCHEMISTRY AND STREAM DISPERSION
O. F. PUTIKOV and B. WEN
INTRODUCTION In conventional geochemical methods, geochemical signatures are the dispersion halos of elements in mineral form, but the geochemistry of the ground surface does not reflect the actual chemical content of the sources in the subsurface. In conventional geophysical methods, geophysical fields are directly related to the physical properties of rocks, but indirectly to their chemical compositions. The ambiguity of interpretation of conventional geophysical and geochemical data led, in Russia, to research into geoelectrochemistry, which was begun in the 1960s by Y.S. Ryss and his colleagues (I.S. Goldberg, V.P. Korostin, S.G. Alekseev and others). Some geoelectrochemical methods and the general physico-mathematical theory of the geoelectrochemical methods were developed in the St. Petersburg State Mining Institute by O.F. Putikov, N.N. Uvarov and others. The results are essentially a family of physico-chemical methods, in which the physical fields of the rocks are utilised, but their chemical compositions rather than their physical properties are studied. In this chapter these methods are divided into: (1) prospecting methods (aimed at finding undiscovered deposits); and (2) exploration methods (investigating poorly characterised deposits).
GEOELECTROCHEMICAL PROSPECTING
Physico-chemical basis According to Antropova (1975) and Antropova et al. (1992), the forms in which heavy metals are present in rocks and their weathering products are: (1) in mineral lattices; (2) dissolved in groundwater; (3) dissolved in capillary moisture; (4) sorbed on solid surfaces; (5) co-precipitated by iron-manganese hydroxides; (6) as metallo-organic compounds; and (7) in the gaseous and quasi-gaseous states. In mineral lattices heavy metals are constituents of ore minerals (oxides, sulphides, sulphates, arsenates and others) and to a lesser extent of rock-forming minerals. In groundwater, heavy metals
18
O.F. Putikov and B. Wen
occur as ions, complex ions and compounds. Their absolute concentrations are of the order of 1-20 • 10.2 ~tg/l and their relative concentrations as compared with their total contents in the rocks usually ranges from 0 to 0.2%, and is rarely greater than 2% (Antropova, 1975). In capillary moisture the absolute concentrations of Pb and Cu are about 10-2-102 ~tg/1 and the relative concentrations of these metals are about 0.1-1% (Antropova, 1975). Absolute concentrations of sorbed metals are 10-1-102 ~tg/l for Pb and 6 • 10.3 to 4.7 • 102 ~tg/1 for Mo, and their relative concentrations are about 1% and for Pb and 0.01-0.1% for Mo. Antropova (1975) showed that after extraction of heavy metals soluble in groundwater, capillary moisture and sorbed forms, a significant fraction of metals is still held by chemical bonds of different strengths in oxides and hydroxides of Fe and Mn. The relative content of this form of metals is usually 1-10% for Cu and Ni, up to 90% (frequently 16.5-86%) for Pb and up to 93% for Mo. A metal may also form compounds with natural organic acid, making metallo-organic compounds (MOC). Antropova (1975) found that fulvic and humic acids are special concentrators of heavy metals and the metals they concentrate exist not as cations but as constituents of humate and fulvate complexes. In these complexes the concentrations of metals are usually higher than the a~erage metal concentrations in the total organic fraction. Several researchers (Krat, 1983; Malmqvist and Kristiansson, 1984; Kristiansson and Malmqvist, 1986; Krchmar, 1988; Dukhanin, 1990; Putikov and Dukhanin; 1994) have pointed out that, in the upper crust and in the near-surface atmosphere, heavy metals exist in the gas phase. But the concentration of elements in this phase is very low, of the order of 10.4 pg/1 (Dukhanin, 1990; Ozerova and Mashianov, 1989). In physico-chemical terms, metals in these different forms are held by chemical bonds of different strengths. According to the strength of the chemical bonds, the forms of metals may be divided into four groups: 9 9 9 9
strongly confined (strong chemical bonds within minerals); moderately confined (MOCs and Fe-Mn hydroxides); weakly confined (metals in capillary moisture, sorbed on surfaces); mobile (soluble in groundwater, quasi-gaseous and gaseous).
The capacity of metals to disperse in space is clearly related to this classification: weakly-confined metals are able to disperse for greater distances than strongly-confined metals, and so on. In particular, mobile forms are able to migrate far from their sourcesprovided that some natural or artificial physical field is exerted to effect migration. However, all forms are in a state of dynamic equilibrium and can transform into other forms. One consequence is that, with increasing concentration of metals in the more mobile forms, there is a simultaneous increase in concentration of metals in the forms with stronger chemical bonds. For example, dissolved Pb, Mo and Cu will interact with oxides and hydroxides of Fe and Mn under certain physico-chemical conditions of pH and Eh. First, due to adsorption of these metals from solution, the metals are
Geoelectrochemistry and stream dispersion
19
C C,"~,,T X
.....
B
Fig. 2-1. Schematic distribution of metal concentration C in: (A) lithogeochemical dispersion halos; and (B) stream halos; hl,h2,h3- different depths of the ore body (reproduced with permission from Putikov, 1993). transformed into sorbed forms. Then through the process of diagenesis of the oxides and hydroxides, the metals penetrate into crystalline lattices and may partly replace Fe and Mn in the crystalline structure, possibly forming new minerals (Antropova, 1975). Mobile and weakly-confined forms of metals make up only a minor part, less than 2%, of the total content of heavy metals in rocks and their weathering products. However, it is these mobile forms of metals that can migrate for significant distances from sources, and thereby convey information about deep ore bodies and oil and gas reservoirs. Weakly-confined forms have direct and steady equilibrium with mobile forms, and thus to a certain degree acquire the same property. Moderately-confined forms also share this property, but only to a small extent. Conventional geochemical exploration rests largely upon the determination of socalled total concentrations of metals, which include a high proportion of stronglyconfined forms. The concentrations of metals found at surface by these methods are highly dependent upon the depth of the source. The amplitude of anomalies, Cmax, decreases and their width, b, increases with the depth, h, of the source (Fig. 2-1A). This relation determines the shallow effective prospecting depth of conventional geochemical methods, which is limited to sources less than 15 m deep (Solovov, 1985). On the other hand, investigation of the mobile and weakly confined forms of metals has disclosed a new type of dispersion halo, called the stream (or jet) halo (Ryss et al., 1987b). The main features of the jet halo (Fig. 2-1B) are as follows:
O.F. Putikov and B. Wen
20 Pb, ~g/ml 10
/a
c 50 m
5 0 AB 0 iiiili:~i~ii~i~ii!ii~i::~ii!::i~i~!::~ilili~ii~i:. ~iiiiiiiiii~iiiiiiiiii~iiiiii I00200 \,, "x,
300 ~il
[x~ w 14
Fig. 2-2. Jet halo of lead, over a blind polymetallic ore body, overlain by allochthonous clays of thickness 30 to 100 m with different intervals between measurement points (a- 50 m, b- 20 m, c- 5 m). Schematic geological section: 1- silts, 2- clays, 3- clay-siliceous siltstones, 4- quartzites, 5mudstone with pyrite, 6- mudstones, siltstones, 7- pyrite-polymetallic ore (reproduced with permission from Ryss et al., 1987b).
9 the shape of anomalies in profile is more variable 9 anomaly amplitude, Cm~x, and width, b, are only loosely related to the depth of the source, h 9 the halo extends nearly vertically from the source, so that the halo width, b, corresponds to the vertical projection of the source to the surface These features of the jet halo enhance the prospecting depth of the geoelectrochemical methods. A number of field experiments have verified that the prospecting depth for an ore body attains some hundred metres and for oil and gas reservoirs several kilometres. Similar results have been obtained with data for relativelyconfined forms of metals, although the widths of halos are greater than those for mobile and weakly-conf'med forms of metals. Detailed studies of the distributions of concentrations of metals in jet halos reveal apparent non-uniformity of anomaly structure (Fig. 2-2). Maximum concentrations of anomalies on the diumal surface correspond to the zones of enhanced concentrations of mobile fo~ms of metals at depth, which extend almost vertically and have a complicated
Geoelectrochemistry and stream dispersion
21
Pb, ~tg/ml 20
~ h = 0 m
5l o
5 0
'l
01 A
, h=2.5m -
>
B
~1 Fig. 2-3. Jet halo of lead for the A-B segment of Fig. 2-2 at the surface (h = 0 m), at depths 0.5, 1.0 and 2.5 m, and a schematic depth section of the halo of the mobile forms of lead: 1- halo of the mobile forms of lead; 2- streams of lead migration (reproduced with permission from Ryss et al., 1987b).
structure, representing a stream of migrating metals (Fig. 2-3). Shigaev (1997) studied the structure of the stream of Mn in an oil field to depth of 140-160 m. The principal difference between jet halos and diffusion halos is the vertical prolongation of the former. The two important factors determining this vertical prolongation are temperature and pressure in the Earth's crust. The numerical solution of the differential diffusion equation for the concentration of a mobile form of metal in a non-isothermal rock shows that even a localised temperature gradient results in vertical prolongation of halos. The pressure gradient influences the migration of gases of the least density, which perhaps migrate in the form of bubbles. A physico-chemical model of jet halos formed by bubble-facilitated transport of metals is proposed and illustrated in Fig. 2-4. In this model the following conditions need to be satisfied:
O.F. Putikov and B. Wen
22 B
C
x ji~
I I l ,
Li
I Zone 3
~ iiii iiii i iiiiiiiiii i i iliii"
.
~
.
.
.
.
.
.
.
.
I [ IZone2 z
........ ~
. . . . i . . . . ~. . . . r . . . . . . . . .
t_..~._.~__~.._~__i_.~.__~i, ~, ~ , ~ ..........
i iiii!iiii
iiiiill!i!iilil
,_ ,.__~___~._~..__~__..~...~.
...................
i'.'!
-,.--i-----4 I ! I I /
,i,
~
Fig. 2-4. Scheme of the formation of stream halo of mobile forms: (A) water table coincides with the ground surface; (B) water table is below the ground surface.
9
9 9 9
9
sources of metals related to ore bodies, oil and gas reservoirs, or the surrounding rocks, exist in a water-saturated porous system with pores, faults, fractures, or microfractures; a regional flow of gaseous bubbles exists, migrating in the porous system by Archimedes force (Fig. 2-4, zone 1); in the process of migration in the porous system, metals are captured on and in the bubbles, forming quasi-gaseous and gaseous forms of metals (Fig. 2-4, zone 2); by bubble-facilitated transport various forms of metals accumulate in the porous system above the source (Fig.2-4A, zone 3, the black circles); if the water table is lower than the ground surface, zone 3 has an aeration area (Fig. 2-4B), in which bubble-facilitated transport of metals is absent, and the dispersion of mobile forms of metals in this area is completed by other mechanisms of migration.
The porosity and fracture content of rocks determine the maximum possible volume of gases in rocks and the gas permeability of rocks determines the speed of migration of these gases. On the basis of these parameters, typical geological structures may be divided into closed and open structures. Igneous and metamorphic rocks tend to have closed structures, whereas sediments have open structures. The porosity of igneous rocks is typically 0.5-2% and changes little with depth. Their gas permeability is typically less than 10.5 lam2 and mainly depends on fracture content. The porosity of sediments decreases from 30-35% at surface to 10-20% at a depth of 2 kin. Their gas permeability varies from 10 .7 lam2 to 3 l.tm2 (Fridman, 1970; Dortman, 1992).
Geoelectrochemistry and stream dispersion
23
Studies of the behaviour of water in capillaries (Derjaguin et al., 1980) reveal that in a capillary of diameter 4 x 10 -3 ~tm water still remains as a Newtonian fluid, that is to say, the start of water movement m the capillary does not demand an initial pressure gradient. Therefore it is possible for gaseous bubbles in capillaries of diameter up to 4 x 10.3 ~tm to migrate by Archimedes force. In addition experimental data show that for rocks of low porosity, for instance, limestone, having porosity 1.14% and permeability 1.1 x 10.5 ~tm2, the diameter of the pores in the rock ranges from 0.016 - 0.2 lam (mainly 0.020-0.032 ~tm) (Kalinko, 1987). Consequently, gaseous bubbles of corresponding diameter possibly penetrate this kind of rock. The following points are of great significance m evaluating the possibility of migration of gaseous bubbles in the Earth's crust. 9 The first super-deep drilling to 12.8 km in the Kola peninsula of Russia verified the effect of de-consolidation (the reverse of consolidation) in rocks at a depth >5 km (Dortman, 1992). This effect results from the increase of fracture content and porosity of rocks at depth. 9 Modem analysis techniques reveal in the upper crust, concealed loose structures, which can not be observed macroscopically (Favorskaya and Tomson, 1989). These may serve as channels for penetration of gaseous bubbles with diameters of the order of microns. 9 Isotope analysis in gas fields in northeast China has revealed the presence of biogenic He, H2 and CH4 from depth (Go and Wang, 1994). It is reported that R - 0.19RA0.5RA (where R = 3 H e / 4 H e in natural gas, RA = 3He/4He in the atmosphere), which means that some gases come from depth in the crust and even from the upper mantle. 9 Experimental measurements have determined the great speed of development of gas anomalies over man-made underground gas reservoirs. 9 Many field and experimental measurements have shown that gas flow can penetrate the cap of gas reservoirs. These observations suggest that, at least up to several kilometres in the crest, gas flow (i.e., flow of gaseous bubbles in a water-saturated porous system) exists. The quantity of gases depends on their solubility at the temperature and pressure at depth (Fridman, 1970; Kalinko, 1987). Experimental studies show that, within a large area, there exists at depth in considerable concentrations many different soluble gases, including N2, CO2, CH4, H2, H2S, He and others (Shvets, 1973; Kalinko, 1987; Kiriukhin et al., 1988). These gases may be divided into poorly soluble and highly soluble gases. At a temperature of 20~ and a pressure of 1 atmosphere the highly soluble gases (CO2, H2S) have solubilities of 878-2588 ml/1 whilst the poorly soluble gases (He, H2, N2, CH4) have solubilities of 9.333.1 ml/1. Laboratory modelling under high temperature and pressure conditions demonstrates that H2, CO: and CH4 at temperatures of 600-800~ and pressure of 20-30
24
O.F. Putikov and B. Wen
Kb, equivalent to conditions at depths of 60-90 km, remain stable and prevail as gases (Wang, 1994). In regions in which there are considerable concentrations of gases at depth, those that are poorly soluble escape preferentially from the water in the form of free bubbles and migrate upward in the water-saturated porous system. On the whole, rocks of low porosity are able to exude free gaseous bubbles of a corresponding diameter. As stated by Fridman (1970), in the case of rocks of low effective porosity and insignificant sorption capacity (thus, especially igneous and metamorphic rocks but not organic-rich sediments), natural gases are mainly in the free state in fractures and faults, and are even dissolved in underground water. A wide distribution of a free gaseous phase in rocks is supported by numerical estimates. For example, the background concentration of H2 in underground water is 0.10.4ml/1 and anomalous concentrations are 3-50 ml/1, whilst concentrations of up to 501500 ml/1 and higher are found with hydrogen flow of 105 m3/day (Scherbakov and Kozlova, 1986). In an underground mine tunnel at a depth of 252 m a flow of gas bubbles containing 76% C O 2 o r 90% CO2 + Hz by volume was observed from 1961 to 1975. It was estimated that average flux of methane was 60-80 cm3/m2 in one year and that the source of the methane was at a depth of 15-20 km (Hitarov et al., 1979). Regional sources of hydrogen may be situated at great depth, even in the mantle (Larin, 1980; Scherbakov and Kozlova, 1986), and some metallic elements, such as Mn, Fe, Ni, Co, Cr and rare earths, are thought to be carried with the gas and form sulphide minerals near the surface, for example in the vicinity of mid-oceanic ridges (Goriainov et al., 1989). In the laboratory of geoelectrochemical methods of St. Petersburg State Mining Institute, a series of experimental studies has been carried out on the physico-chemical mechanism of penetration of gaseous bubbles through the porous system (Putikov and Wen, 1997; Wen, 1997a). In the experiments the porous system consists of a long wide robe containing small particles of silicates or gravel and water with different concentrations of metals and organic substances. The particles of silicates have different fractions with diameters 1-2, 2-3 and 3-5 mm, and the particles of gravel have diameters of 5-7 and 7-10 mm. Groups of gaseous bubbles of diameters from 5 • 10-5 m to 2 • 10.4 m were introduced into the bottom of the tube, and the average speeds of the leading and rear fronts of every group of bubbles penetrating the porous system were determined. Three forces act on a gaseous bubble in free liquid (without a solid phase): gravitational force (G - mg - VOog); Archimedes force (F = Vog) and the resistant force of the medium defined by Stoke's law (R = 6rtqr0v0), where, g = acceleration due to gravity, r0 = radius of bubble, V = volume of bubble, 190 = density of gases in bubble, 9 = density of liquid, rl = dynamic viscosity of liquid, v0 - speed of bubble at equilibrium of the three forces. The speed of the bubble can be calculated by the following equation:
25
Geoelectrochemistry and stream dispersion V/Vo 1.
_-.7 r---
0.8
f
0.6 0.4
/
f 0
?
0.2 0 0
f
g
10
El
""
a(1)
0
a(2)
I
b(1) 9 b(2)
-----3
.11
--
20
n
30
40
50
60
70
80
90
100
r/ro Fig. 2-5. Dependence of the relative speed of air bubbles v/v0 (ratio of speed of bubbles in porous system to that in free liquid) on the relative radius of the fraction of particles r/ro (ratio of radius of particles to that of bubbles): a- front speed; b- rear speed; (1) r0--0.085mm; (2) ro= 0.06mm; (3) empirical curve.
Vo - kro 2
where,
k-
2(p-Po)g
9n
As shown in the experimental results (Fig. 2-5), the relative speed of the front of a group of bubbles, V/Vo (ratio of speed of bubbles in porous system to that in free liquid), increases with the relative radius of the fraction of particles, r/r0 (ratio of radius of particles to that of bubbles), if r/ro >6.4. The relationship may be represented by the following empirical formula, v
Vo
= 1
-
6.4 ~
(r / ro)
= 1-6.4(r
o /r)
From this equation it can be seen that if the relative radius of bubbles r0/r < 0.156, the relative speed of the front of bubbles v/v0 increases with the decrease of relative radius
26
O.F. Putikov and B. Wen
V l ( k r 2) 0.000
,,
1
0.005
0.004 0.003 0.002
0.001 0 0
I 0.1
0.05
0, 0.15
0.2
r~/r Fig. 2-6. Dependence of the ascending speed of the gas bubbles v on their radius ro: 1- in free liquid; 2- in porous medium with solid particles of radius r.
of bubbles ro/r. If ro/r >__0.156, the bubble will not penetrate the porous system in the experiment. The equation may be rewritten as follows (Fig. 2-6)"
kr 2
-
From this equation a maximum speed F "~nlax
k
~
will be found in romax (Fig. 2-6, curve 2):
2
9.6 2 •
1
ro max i
r
Vma x
9.6
- 0.104
In this case, ro Vmax
1
- (1 - 6.4--)Vomax [~o=ro r
m.
= --VOmax
3
Geoelectrochemistry and stream dispersion
27
Vm~t~m/year 10 7 L 10 6 .
105 10 4
103 102 10 1 10-! 10.2 10-3 10 .4
10-s 0.01
i 0.1
i 1
i 10
i 100
r, lxm
Fig. 2-7. Dependence of the maximum speed of the gas bubbles in the porous rock model Vmaxon radius r of the rock particles.
where, Vomax is the estimated speed of bubbles in free liquid when these bubbles have the m a x i m u m speed in a porous system of particles of radius r. If we take P = 103 kg/m 3, 90 - 0 kg/m 3, g = 9.8 m/s 2, r I = 10 .3 kg/(m• we obtain,
Vmax
-
-
7 . 9 x 103 r 2
where, [r] = m,
Vmax
--
[Vmax]-"
m/S, or,
2 . 4 9 x 1 0 zz r 2
where, [r] = m, [Vmax] ---- m/year For comparison we extrapolate the experimental data by the equation to those in a porous system of particles of very small radius (Fig. 2-7), for instance, r = l~tm. Then the m a x i m u m speed of the front of the bubbles may be Vmax-- 2.49 X 10 -I m/year. When r - 0.1 ~tm, then Vma x = 2.49 x 10 .3 m/year. For prospecting, the requirements for bubble-facilitated transport of metals are that: (1) the ore bodies or oil and gas reservoirs contain heavy metals of higher concentrations than the surrounding rocks; (2) the metals may transform into mobile forms of metals in the vicinity of prospection targets; and (3) the metals may be captured by gaseous bubbles and be transported upwards through the overlying rocks. First consider ore deposits in which metal concentrations are raised to some degree. Transfer of these
28
O.F. Putikov and B. Wen
metals from the solid phase to the liquid phase means transformation of confined forms of metals to mobile forms. Many studies have shown that in the water in the vicinity of ore bodies concentrations of metals is higher than elsewhere. For example, background concentrations (Cb~) and anomalous concentrations (C,,) in oxidised polymetallic sulphide deposits a r e C b a -" 8-50 pg/1 Cu 2+, 5-8 pg/1 Pb 2+ and 10-30 pg/1 Zn 2+, C a n - - 5 0 0 20000 pg/1 Cu 2+, 10-20 lag/l PbZ+and 50-1500 pg/1 Zn 2§ In the case of unoxidised polymetallic sulphide deposits, Cb~ = 5-7 btg/1 Cu 2+, 4-6 btg/1 Pb 2+ and 5-35 pg/1 Zn 2+, Ca. = 10-140 ~tg/l Cu 2+, 12-30 ~g/1 Pb 2§ and 35-700 btg/1 Zn 2§ (Goleva, 1977). Oil and gas reservoirs contain micro-components including heavy metals (Table 2-I). As Figs. 2-8 and 2-9 show, concentrations of some metals in oil significantly exceed their Clark values. The greatest concentration coefficients (ratio of concentration of metals in oil to their Clark) are 102-103 attained by Au, PGE, Re and Hg, whilst Mo, Ag, Sb, V and Ni have concentration coefficients of 3-102.
()ll. C. I 0 ' M O I / g I I1~
ill ~
i
lit ~ .....
lW
!
: .....
CI
i
..............
..... :' ...............
~
'.
i
. . . . ! ...............!...... : "
io' lip ,,,,
V rr
TI : 4K
:
!. . . . ~ ~ ? v +
; "NI':
.... ,,.(
A~"
....
:
i i !: . . . . . . . :: . . . . . .
: cd ' / 4 . , c~ isc, / i ' n,': I / ' . : : 0?~, i: . .i . . . . W! . ,..~. . Z - . .!A, : : . , C.... ' . _ :..... , ~ . > i }~ . . . . :!. . . . .
!: . . . . . . . . . . . 9.
i
iA , ~ ~ ' C ~ b
!
.
, ............. S
!
i
i.5"~
i
.
i
IO"
lO 4'
' " ~ ' o ' 4 ~
9 ltf I
! :
~"
~:.+"
,4. (':,
~,i
I()"
!
....
!.
Ill ~
1[I ~
i i
:
:
i
:
!
!
!
:
i
:
i
!
i;
Ill "
.......
:
Ill i
lllU
......
lllJ
lilt
("lark. alOlllJfe~
Fig. 2-8. Correlation of average concentration of trace elements in oil samples from Kaliningrad district, Russia, with their Clark concentration in the Earth's crust.
K ll)J
'Re
* 10~
"
Au
tO' I II (' lid
-lll
!
*S~ Mo~ *Ag ,Sb ,B i *J ,Br ,As .ln : ................. ,U !,Co 9 9~u
. . . . .
,V
*N[ .............................
,Zn
Cs- . C . !
............................
,Cr
$ ~ i ' i ~ i .................
-Eu
: :
,Hf
IO
eLl
- .........................
-~
: .........................
*St : ,BaoRb '
-PI~
u)
-CI
*Ns ,Cs
-Mn
;iii;i~;;s;;,~,:
-K
.......................
iOa
-MI :
Ill"
...................................
i ....................................
I0~ 1010"7
! 0 "'~
10 .4
Clark,
I0 J
10 4
!. . . . . . . i ' -St i.......
,TI Ill'
.,AI
IO-I
:
I00
Ill I
102
%
Fig. 2-9. Dependence of the concentration coefficient K of elements in oil (ratio of the maximum concentration of an element in oil to its Clark in lithosphere) on Clark of elements in lithosphere.
Geoelectrochemistry and stream dispersion
29
TABLE 2-I Average content of chemical elements in oil (Punanova, 1974) Element Na Mg AI Si C1 K Ca Sc V Cr Mn Fe Co Ni Cu Zn Ga As Se
Content (mg/kg) 13.2 9.1 8.0 8.7 45.8 11.1 26.0 2.91 32.6 0.58 0.31 23.3 0.37 12.8 0.38 2.98 0.078 0.22 0.285
Number of Samples 234 124 267 39 126 27 117 116 1442 411 560 418 503 1311 461 255 60 102 39
Element Br Rb Sr Sb I Cs Ba La Ce Sm Eu Yb W Au Hg Pb Th U
Content (mg/kg) 2.43 0.34 0.42 0.022 1.93 0.06 0.44 0.0064 few few 0.0075 few few 0.00051 2.56 0.00065 few 0.02
Number of Samples 203 8 190 112 55 21 193 34 10 64 82 477 32
Chemical analyses of gas condensates reveal concentrations of Cr, Sb, Eu and U greater than those in oil. But for C1, K, Mn, Cu, Zn, Br, Rb, Ba, La and Au, concentrations in gas condensates are 2-3 times lower than those in oil. Concentrations of just a few elements, Sc, Fe, Ni, As and Hg are an order of magnitude lower in gas condensates than in oil. On the whole, gas condensates tend to have increased concentrations of rare and trace elements. In the vicinity of oil reservoirs, concentrations of practically all elements in underground water are greater than those in oil (Fig. 2-10). For instance, alkaline elements and alkaline earth elements, halogen elements, As, Se, Mn and others have concentrations in underground water that are approximately five orders of magnitude higher than those in oil (upper line on Figs. 2-10A, 2-10B), whilst S, V, Ni, Cu, Co, Be and others are two orders of magnitude higher. Other elements such as Cd and Bi have similar concentrations in oil and in water. Consequently, in the lithosphere-oil-water system, elements of greatest concentration in oil are Au, PGE, Re and Hg, while those of greatest concentration in nearby water are elements such as Se, Mn, V, Ni, Cu and Co.
30
O.F. P u t i k o v a n d B. Wen
Water, C, 10 -6M/g 10 7
A
.......
t.1!
%/i
10 6 10 5
Brs /
10 4 1o 3
..............
!
..............
i ...............
!
:................
.~,~
: ...............
.
v.,
. . . . . . . . . .
... . . . . . . . . . .
/
i
.g "! .......
10 z 1o I lO o
i i
10-1 . . . . . . . . . . . . .
~9
/ /~. -
....
. . . . . . . . . . . . . . . . . . . . . .
':c ~ . i ,y v i
~
i
10 .2
co
.........
.Sb~
:
i ~:
i
!/~eAsw
!
10 -3 ............~.,.~.~..~._~i17;i
~. . . . . . .
!
i
i
i
.....i.............]...............i......
i0 4 ..............................................
i0/0-5
10-4
Water 10 7
10-3
'10-2
.. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10-1
10 0 101 10 2 oil, C , 10 9 MOI/g B
C, 106g/!
, .......
10 3
10 4
10 5
[I 6
: ~1
lO 6
i
lOs . . . . . . . . . . . . . .
: ...............
.a
~. . . . . . . . . . .
: i.
.
.
.
1o 4 ,0-'
..............i...............i. . . . . . .
i
102
,o,
..............
iOcl
i
~ b ~AI
i. . . . . . . . . . . . . . .
i ..............
,
Z
nSr
~~
i
!:
.........
~r ii
~ ........
i .... Nf . . . . . ! . . . . . . . . . . . . . .
..... ; : ~ ~ '
:
!0 -I
~
i
.....
i!
",'. . . . . . .
..... i .............. i .....
_
...................
10-2
,o-., .............. ~.... T.....~.' 10 -4
.......... i ............. i .............. i ......
Bi .............................................
1~r~0-5
10-4
io-.X
10-2
10-~
10 0
9
9
:. . . . . . . . . . . . . .
7 ..............
101
10 z
10 3
10 4
' .......
10 -~
10 6
oil, (2, 10 .9 Mol/g
Fig. 2- I0. Correlation of the element composition of oil and oil water in oil deposits in Russia at: (A) Severo-Krasnoborskoe, borehole 1" (B) Deyminskoe, borehole 45.
Since the gaseous bubbles that have the maximum speed of migration in a porous system have a radius of about one tenth of the radius of the particles of the system, it is rational to suggest that bubbles of this size are able to pass through the porous channels in rocks. In this case the dispersion of metals in mobile forms depends mainly on the mechanism of bubble-facilitated transport of metals within and on the surface of bubbles. Dispersion of metals (Sb, Cu, Zn, Pb, Fe and others) in a gaseous phase in rocks was established directly by experimental studies of hydrogeothermal vapour (Krat, 1983) and volcanic gases (Putikov and Dukhanin, 1994), and indirectly by studies of subsurface air (Kristiansson and Malmqvist, 1986; Krchmar, 1988; Dukhanin, 1990).
Geoelectrochemistry and stream dispersion A 0
25
C-Cf, mg/l 0.2
H20
B 0.4
C-Cf, rag/1 0.2
0
2~
0.4
:li .....
50 [.1 ",1
50
31
" , ;~
CuIO4 air
air 75 h, cm
75 h cm
Fig. 2-11. Concentration distribution of metals along a vertical tube during introduction of air bubbles into the bottom of a tube: (A) copper; (B) manganese. Experiments: 1- without surface active agents (SAA); 2- with SAA of 1% solution of acetic acid; 3- with SAA of 1% solution of acetic acid and sodium nitrate (reproduced with permission from Putikov and Dukhanin, 1994).
Hydrogeochemical studies show that there are various organic substances in underground water and concentrations of organic acids (such as formic, acetic, propyonic and other acids) reach 20-60 mg/1 (Shvets, 1973). Several salts of these acids function as anionic surface-active agents (SAA). Molecules of anionic SAA concentrate on the surface of bubbles and are oriented with their negatively-charged poles outwards into the liquid phase. Consequently they attract positively-charged metal ions. In this way, as bubbles penetrate through water with anomalous concentrations of metals and natural soluble organic substances, they adsorb the metals on their surface and transport them into overlying porous rocks. The process may be defined as natural ionic flotation of metals. In order to understand the mechanism of the process a number of physico-chemical modelling experiments have been carried out under laboratory conditions (Putikov and Dukhanin, 1994; Wen, 1997a). In the first series of experiments the porous system is modelled by a vertical glass tube of height 79 cm. There are five openings on the side of the tube for sampling. The lower part of the tube is filled with water and a solution of KMnO4 (concentration of Mn, 700 mg/1) and CuSO4 (concentration of Cu, 800 mg/l). In the experiments Cu is in the form of the simple cation Cu 2+ but Mn is in the form of complex anion MnO4-. A flow of bubbles (radius 0.01-0.1 mm) is introduced into the tube from the bottom. After several hours the concentration of Cu is 0.021 mg/1 in the upper part of the tube is about the same as the background concentration of 0.001-0.032 mg/l, but the concentration of Mn is increased to 0.11 mg/l compared to a background of 0.025-0.060 mg/1. When a solution of 1% acetic acid (and in some experiments NaNO3), is added to simulate the presence of SAA, the concentration of Cu is increased 1.5-17.5 times, but the concentration of Mn is essentially unchanged (Fig. 2-11).
32
O.F. Putikov and B. Wen u, pg/I
100000
10000
1000
100
3 10
0
50
J
J
I
100
150
200
250
mm
Fig. 2-12. Concentration distribution of uranium along a vertical tube. Concentration of fulvic acid Cfa=200mg/l, amplitude of mechanical vibration A=0. l mm, frequency 50 Hz. Duration of air bubble flow x, min: 1- 0; 2- 30; 3- 120.
In another series of experiments the porous system is simulated by a wide tube which is filled with water and particles of quartz of diameter 1-5 mm and, in the lower part, with a layer of a solution of UO2(NO3)2 with a concentration of 40 mg/1 U. A flow of air bubbles of radius 0.01-0.12 mm is introduced from the bottom. In order to intensify the penetration of bubbles through the porous system a mechanical vibration of amplitude 0.1-0.5 mm and frequency 50 Hz is applied to the tube. Variation of the concentration of U is monitored by a laser luminescence detector on the surface of the water and five samples from different heights of the tube are taken periodically for chemical analysis of U. Figure 2-12 shows the concentration distribution of U at different heights in the tube. Concentration in the lower part of the tube decreases with time, but that in the upper part increases. After sufficient time the concentration distribution of U reaches a maximum near the surface of the water in the upper part of the tube. This occurs as a result of bubble-facilitated transport of U through the porous system. Concentration of U on the surface of water increases non-linearly with time. This is probably due to the intensity of bubble flow, interaction of metallic ions and the surfaces of bubbles and adsorption on particles of the porous system. Adding fulvic acid to the solution of UO2(NO3)2 results in an increase of the concentration of U on the surface 1.5-3 times faster than without fulvic acid (Fig. 2-13). These experimental results verify the importance of soluble organic substances in bubble-facilitated transport of metals in a porous system. They suggest that such substances contribute to the formation of jet halos. In this way bubble-facilitated transport of metals through rocks occurs in a quasigaseous phase in bubbles and as complex ions on the surface of bubbles, in effect by natural ionic flotation in overlying rocks (Fig. 2-4a, zone 3). In zone 3 there is interaction between gaseous, liquid and solid phases with different concentrations of
Geoelectrochemistry and stream dispersion
33
U, ~ g / l 80
70 60 50 40 30 20 10 0
.......
0
t
I . . . . .
l
,,
20
40
60
~
80
~......
100
120
min
Fig. 2-13. Accumulation of uranium concentration at the surface of a liquid in a tube as a function of time. Concentration of fulvic acid C f a and amplitude of mechanical vibration A: 1- Era=200 rag/I, A=0.5 mm; 2- Cf~=200 mg/1, A=0.1 mm; 3- Cfa=0, A=0.5 mm; 4- Cfa=0, A=0. l mm.
metals, leading to the presence of all forms of occurrence of metals in the surrounding and overlying rocks. From these studies it is suggested that the general mechanism of bubble-facilitated transport of metals can be represented by a non-linear integro-differential equation for concentration distribution of soluble components in underground water, considering interactions of metals at phase interfaces of gas-liquid and liquid-solid under conditions of constant radius of bubbles (Putikov et al., 1994). Taking account of some simplified conditions the equation can be reduced to the following diffusion-quasi-convective equation,
V 2C
]2eft #C D
Oz
q m a x Ce -[~ [tCJo(x,y,z,~)d~ _ D
aC = 0
u ~ l
(2.1)
Dat
where, C - volume concentration of soluble components, e.g., metals in water in porous rocks, veff = effective speed of quasi-convection, related to the penetration of gaseous bubbles, D = coefficient of hydrodynamic diffusion of soluble components in porous rocks, qm~x= maximum concentration of components in solid phase, 13 = kinetic constant of chemical sorption of components, t = time of transportation of the components, x,y,z - spatial co-ordinates of origin. If bubble flow is sufficiently intensive, diffusion may be neglected and equation (2.1) can be rewritten,
O.F. P u t i k o v a n d B. Wen
34
-[3 ~oc(x,y,z,~ )d~
OC
+ qmaxCe
Ver ~
OC
+--Ot -- 0
(2.2)
This non-linear integro-differential equation is equivalent to the following system of non-linear differential equations, OC
Velf
--+
Ot
OC
Oq
OZ
Ot
.... +
= 0
(2.3)
Oq
c3t
(2.4)
= ~C(qma x - q)
where, q
-
qmax
1 - e -~
(2.5)
C ( x , y , z~ )d~ o
Solution of equations (2.3) and (2.4) under the following boundary and initial conditions, CIz_o - C O
(2.6)
Cl,=o - 0
(2.7)
q],-o - 0
(2.8)
where, Co is the concentration of the soluble components at source and is considered constant in geological time, with the form (Panchenkov, 1964),
C(z,t) CO
C(z,t)
1 (e f~q..... "/"~ -- 1)e ~c~
=0,
, z < v~Lrt
(2.9)
+1
(2.9')
z > v~fft
Co 1-e
qmax
e f~c~
[SCo( z / v,,rt -t )
.... z/v"rr - e f~C~
~Z < V eff t
(2.10)
Geoelectrochemistry and stream dispersion
q
= O,
35
v~t
z >
(2.10')
qmax
When z and t have values such that (t >_z/v~ff, z/v~ff--~oo), we obtain, C
1
=
Co
e
[3 (q ....
+C~176
I(C~
)) I Velr
,z < --
Veff t
(2.11)
From equations (2.9) and (2.11) and Fig. 2-14, it is obvious that the shape of the concentration-front distribution is stabilised with time and the concentration front moves with a certain constant speed, vr, determined by the following equation,
C~v~ff Vf = Co +qmax
(2.12)
For example, if Vef f -- 10 .7 V0 and qmax/C0 = 100, from (2.12) we obtain vf = 10 .9 v0. That is, under the given conditions of speed of quasi-convection, veff makes up to 10 .7 of the physical speed of movement of bubbles, v0, and the speed of the concentration front
C/Co 1
0.8 0.6 0.4 0.2
0
50
1O0 crn
150
200
Fig. 2-14. Concentration distribution of the mobile element forms in a I D convection stream halo for different moments of time x, hours: 1- 100; 2- 300; 3- 500; 4- 1000; 5- 1500; vefr~lcm/hour; ~qmax/Veff=0.1 cm -1", 13Co/verr=0.1 cm l
O.F. Putikovand B. Wen
36
vf is 10 .9 v0. In the general case a one dimensional equation can be written,
O 2C D~Oz 2
OC V ~ ~g Oz
-~j'~c(~,n)~n kzqmaxCe
OC Ot
=
0
(2.13)
Under boundary and initial conditions,
Clz=o -Co,
(2.14)
(V~yyC- D OC ] -
Oz ) z=,-,
-
~ C]z=H
(2.15)
C],_ ~ - 0
(2.16)
The equation is solved by means of iteration and fully-implicit finite difference in the following scheme (Wen, 1997b),
_(0.25c+0.5s)Ci'+l,,+(1.O+s+~qmax~e
,
~=,
~ )C~i~,+~+(0.25C--0.5S)C++,'m =
(0.25C + 0.5slC'_ , +(1.0--SIC' + (--0.25C + 0.5S)C:'+, where, c = verrAt/Az, s = DAt/(Az) 2. Figure 2-15 shows that with parameters Co = 1 1/m ~, V e f f • 10 .5 m/s, D = 10 -7 m2/s, q,,,,x = 1 1/m ~, [3 = 10 .6 m3/s, the concentration front extends with constant speed and keeps its distribution shape for a certain time. In this case concentration distribution is also (as in Fig. 2-14) stabilised with time without diffusion.
Partial extraction of metals (CHIM) In the CHIM method of prospecting a direct current is introduced into the ground by means of a current electrode and an element collection electrode. This facilitates the extraction and accumulation of ions of the metals in zones near electrodes. The subsequent analysis of the extracted metals yields information about their distribution in the rocks in the zone of investigation.
37
Geoelectrochemistry and stream dispersion ClCo 1.2
0.8
5
0.6
6
0.4 0.2
t 0
20
40
60
80
100
120
140
meters
Fig. 2-15. Concentration distribution of the mobile element forms in a l D diffusion-convection stream halo for different moments of time x, days: 1- 1.45 2- 2.89; 3- 14.5; 4- 28.9; 5- 86.8" 6144.7.
During current flow through the rocks several physico-chemical processes take place, including dissolution of the solid phase, ion transfer in the electric field and accumulation of ions in the vicinity of the electrodes. It is necessary to recognise certain conditions for the dissolution of the solid phase. For example, dissolution of electronconducting minerals in a salt solution containing the same metals as these minerals gives rise to a potential difference corresponding to the potential of the required anodic electrochemical reaction (Table 2-II, next section). To increase the speed of ion flow, a relatively large electric field is used (greater than, for example, that used in the CPC method, described below). Under these conditions, we can neglect the diffusion and convection components of current compared to the migration component. Then the density of current jn+ due to cations with number n+ is: j , + - F z ,+ u ,+ C,+ E
(2.17)
where, F = Faraday constant, z,+ = charge of the n+ cations, u.+ = movability of the n+ cations, C,+ = concentration of the n+ cations in moles/m 3, E = the electric field strength. The speed of the ion motion (speed of migration) in an electric field v is,
- u)EI where u is the movability of the ion. According to Ryss (1983) the ion movability in rocks is small, about 0.01-1 cm/(v.hour). 2 This means a speed of ion motion of 0.01-1 cm/hour under an electric field strength E = 1v/cm.
38
O.F. Putikov and B. Wen
~Cathode
/ / /"l'\ \x
I
v
I V
I
I V
t
[ w
I ~.
I v
Me2++2Olff
.[v
I
I v
i IOi.][-
v
>Me2~(OH)2
0,1TTl Tl Tl Fig. 2-16. Movement of ions close to cathode (reproduced with permission from Putikov, 1993).
The motion of metallic ions to the cathode and their accumulation on its surface are accompanied by interaction of metallic ions with the products of the cathodic electrochemical reactions. In particular, discharge of hydrogen ions at the cathode leads to the accumulation of the hydroxyl ions in this region, according to the activities of the hydrogen ions [H +] and hydroxyl ions [OH]: [H+].[OH -] = kw where kw is the water dissociation coefficient. The activity of the hydrogen ions [H +] decreases during the cathodic reduction reaction 2H § + 2e- ---~ H2~. This leads to increasing activity of the hydroxyl ions [OH] because the product kw is a constant. As a result there is a flow of hydroxyl ions by diffusion and migration in the electric field against the flow of the metal cations. At some distance from the cathode these hydroxyl ions meet the common metal ions (those of Pb, Zn, Cu, Ni, Fe and others) and produce insoluble hydroxides (Fig. 2-16). This process prevents metallic ions from further accumulating in the vicinity of the cathode. Me 2+ + 2OH----~ Me(OH)2 In order to avoid this undesirable effect and to promote metal ion accumulation in the liquid phase, Ryss and Goldberg (1973) developed a special element-collector. This consists of a vessel containing a metallic electrode and a semi-permeable membrane, on one side of which is a solution of nitric acid (Fig. 2-17). The semi-permeable membrane prevents egress of the acid solution and allows ionic exchange between the elementcollector and the surrounding environment. The acid neutralises the hydroxyl ions and thereby maintains the solubility of metal ions in the vicinity of the cathode.
Geoelectrochemistry and stream dispersion
39
Fig. 2-17. An element collector: l- titanium rod electrode; 2- solution of nitric acid; 3polyethylene vessel; 4- semi-permeable membrane (reproduced with permission from Putikov, 1993). The acid in the element-collector dissociates according to the equation, HNO3 ~
H ++ NO3.
Due to its large diffusion coefficient the hydrogen ions, H +, pass into the surrounding environment through the membrane and form an excess positive charge on its outer surface. The NO3 ions have a lower speed and form an excess negative charge on the inner surface of the membrane (Fig. 2-18). Thus with time a double electrical layer with a descending electromotive force gd~(~) is formed on the membrane. The electric field strength of this layer on the membrane is Edl = gd~(~)/A1, where AI is the thickness of the membrane. If C and Co are the ionic concentrations of metal in the solution of the elementcollector and in the surrounding environment respectively, then metal ions pass through the membrane as a result of (1) diffusion (concentration difference C-C0), (2) migration in the external electric field with strength E and (3) migration in the internal electric field with strength Ed~. Putikov (1993) shows that the differential equation for the metal concentration C in the element-collector is,
dC
h dr
- - k ( C - C o) + u E C o + uEa~ ('r.)C o
(2.18)
where, h = height of layer of solution in element-collector, 9 = time, k = membrane coefficient, u = movability of metal ions in the membrane.
40
O.F. Putikov and B. Wen
l
.,$+
*H+ \ S = I
\3
12o
Fig. 2-18. Scheme of concentration distribution of ions in and out of an element collector: C, concentration of metal in element collector; Co, concentration of metal outside element collector; I- solution of acid in element collector; 2- elementary pillar of solution in element collector with unit bottom surface S = I and height h; 3- semi-permeable membrane (reproduced with permission from Putikov, 1993).
According to Ohm's law, E = 9J, where E is electrical field strength, P the resistivity of the membrane and j is the current density. Taking this into consideration, we can rewrite equation (2.18),
dC dx
--h
k (c
-
C ) + UpJ c o
h
o
+
U
~
E,j/(T.)C
o
(2 19)
Equation (2.19) lays down the theoretical basis of the element-collector function for the CHIM methods (and for the MDE method, described below). For the high values of current and current density in the CHIM method we can suppose,
lu ,JCol >> I- (C-Co>l
lupjCol >> luE.,(= >Col
Then equation (2.19) takes the form,
dC
upj
d~
h
~C
o
(2.20)
Solution of equation (2.20) for the initial condition,
C 1~=o - 0
(2.21)
Geoelectrochemistry and stream dispersion
41
C~m
2
3
Co
"'~
.
.
.
.
.
>I;
0
Fig. 2-19. Dependence of concentration C and mass m of accumulated metal in an element collector on time ~: 1- pure diffusion accumulation (j=0; Ed~=0); 2- migration under the action of an external electric field (j>>0); 3- diffusion accumulation with action of a double electrical layer of semi-permeable membrane (Em~:0)(reproduced with permission from Putikov, 1993).
is a linear function of time (Fig. 2-19, curve 2):
C-
upy
h C~
(2.22)
In the case of MDE the extemal current is absent (j=0) and the equation (2.19) takes the form,
dC
da: - - h
k (c _ c )+ o
U
~
Eo,('c)C
o
(2.23)
If we neglect the influence of the double electrical layer (Edl=_0) we can obtain the following solution to equation (2.23), k
C - C o (1 - e
)
(2.24)
This is the case of the pure diffusion accumulation of metal, its concentration in the element-collector increasing smoothly with time up to the equilibrium value Co (Fig. 219, curve 1 ). In the general case, solution of equation (2.23) considering the influence of the double electrical layer, metal concentration in the element-collector is timedependent (Fig.2-19, curve 3). In its usual configurations CHIM is applied in ground and borehole modes. It is possible to use different kinds of installations in the ground mode, including those shown Fig. 2-20. In the case of homogeneous rocks with a constant concentration of metal, C1, the mass of this metal accumulated in the element-collector, m, is a function of time x,
42
O.F. Putikov and B. Wen
Fig. 2-20. Scheme of a field installation of CHIM: 1- current source; 2- ore body; 3- halo of dispersion; 4- host rocks; AI-As- element collectors; B- auxiliary earth electrode (reproduced with permission from Ryss et al., 1987a).
m = Sql: where, S = square of the membrane, q = the flow density of metal. Taking into account that q = uC,E, we have, m = SuC~Ex. This means that SuC,E is a constant for a homogeneous rock and that the accumulated metal mass is a linear function of time. The relationship is represented as a geoelectrochemical hodograph (Fig. 2-21, branch I) in which the angle of inclination of the curve to the time axis, x, depends on the concentration, C,. Another case represents an inhomogeneous medium with different metal concentrations at different depths, for example, a dispersion halo with concentration C, and an ore body with concentration C2. The dependence of metal accumulation, m, as a function of time, x, is shown in Fig. 2-21. It is possible, in principle, after determination of the time xt and the angles (~1 and o~2 to estimate depths and concentrations of metals in the different layers of the inhomogeneous medium. But because of the low movability of ions in rocks, only branch I of the geoelectrochemical hodograph is used for ground (or halo) mode CHIM. Branch II of the geoelectrochemical hodograph is used in borehole (or basic) mode CHIM.
Geoelectrochemistry and stream dispersion
43
m
II (X2 m
,,.
m
w
..o
|
0
(Zl
.
T
'1;1 Fig. 2-21. Geoelectrochemical hodograph (metal accumulation, m, versus time, x) for a two-layer medium.
Different types of equipment have been designed in Russia for the CHIM method. The CHIM-k installation has a capacity of 10 Kw and allows the use of 40 elementcollectors with a current and voltage stabiliser. Another installation consists of two modules and allows the use of 60-120 element-collectors simultaneously. Ground mode CHIM is usually used for detailed investigations on the scale 1:10000 and larger. Observation spacing depends on width and inhomogeneity of the anomaly targets and is usually is 20-25 m or less. It is necessary to place all element-collectors in the same ground layer. Logging mode CHIM allows the determination of concentrations of metals in ore with an error +100% at intervals that cannot be investigated by other borehole logging methods (in the fractured zones, in cavities and so on). The time of accumulation is 10-20 hours for the ground mode of CHIM and 1-2 hours for the logging mode. Ground mode CHIM can be applied under condition, in which conventional methods of geochemical exploration are not effective, for example: 9 detection of deep-seated mineralisation, beneath as much as 500 m of unconsolidated overburden or clay sediments; 9 evaluation of geophysical anomalies and indication of the material composition of their sources; 9 more accurate determination of the position of sources of anomalies first detected by other geoelectrochemical methods; 9 indication of the contours of oil-gas reservoirs.
44
O.F. Putikov and B. Wen
PI: cI-nM, gg 25
Pbtotai, 10 -3 %
/ "
~
~
I I ll l ,lltl
Pbto~. "\'" 0
"
I
I
II I
I I
I
I
I
I
0pm,
I Fig. 2-22 Results obtained by the CHIM method over polymetallic mineralisation in Rudny Altay, Russia: Pbcl~lM- concentration of mobile forms of lead (CHIM); Pbtotal- total concentration of lead (lithogeochemistry); 1- unconsolidated sandy-clay overburden; 2- volcaniclastic strata; 3- polymetallic ore; 4- low-grade disseminated ore (reproduced with permission from Ryss, 1983).
The first publication on CHIM (Ryss and Goldberg, 1973) contains some examples of the successful applications of the method. For ground mode CHIM these include investigations of known polymetallic ore bodies at Altay and the copper-nickel composition at depths of 10-100 m of ore bodies in the Kola peninsula. The detection of copper-nickel ores in boreholes by logging mode CHIM is demonstrated. The polymetallic sulphide deposit at Rudny Altay comprises a number of nearly vertical ore bodies at depths of 450-500 m in a tuff-slate formation. The tuff-slate is covered by dense Mesozoic-Cenozoic clays 40-50 m thick. The lithogeochemical survey does not yield clear anomalies. But the CHIM results for lead delineate satisfactorily the position of the ore bodies and their approximate projection to the surface (Fig. 2-22). In the far east of Russia a cassiterite stockwork at a depth of 700 m lies between sandstone and aleurolite (Fig. 2-23). The results of the conventional geochemical survey and rock samples from trenches fail to reveal the position of the ore body (Fig. 2-23, curves A, B). The CHIM survey, however, gives a good expression with up to 16 lag Sn compared with a background of 0.5-1 lag (Fig. 2-23, curve C). This anomaly has a width of 250 m and practically coincides with the projection of the ore body to the surface.
Geoelectrochemistry and stream dispersion
45
Fig. 2-23. Results obtained by the CHIM method over tin stockwork ore in Primorsky Kray, Russia: results of lithogeochemicai survey (A) at surface, (B) in rocks in trenches, and (C) results of CHIM survey; 1- Quaternary sediments; 2- aleurolites and sandstones; 3- tin stockwork ore; 4small veins of tin mineralisation; 5- small veins of polymetallic mineralisation (reproduced with permission from Alekseev et al., 1981).
In Byelorussia the early Proterozoic crystalline basement, represented by biotitegranite-gneisses, migmatites, micaceous slates and quartzites with acid intrusions and (gabbro-)diabase dikes, is concealed beneath Quaternary fluvio-glacial sediments 30-150 m thick. Beryllium mineralisation, the product of regional and local metasomatism, occurs in the concealed basement. The Be data of a CHIM survey expressed very well the ore body known from boreholes 1 and 2 and revealed a further anomaly. Drilling of boreholes 3 and 4 on this anomaly confirmed the existence of a previously-unknown ore body (Fig. 2-24). The value of CHIM in prospecting for gold deposits is shown by the "Prijutinsky" section near the Enisey river in Eastem Siberia. This section comprises phyllite slates beneath proluvium sediments. The known placer gold mineralisation is in phyllites covered by gravels and clays 5-7 m thick (Fig. 2-25). Mining has indicated the distribution of gold (Fig. 2-25A). The curve of gold mass extracted by CHIM (Fig. 225B) corresponds satisfactorily to the gold concentration distribution obtained from mining. The deposit was further explored by drilling only of the CHIM anomalous zones, with a 3-4-fold reduction in expenditure.
O.F. Putikov and B. Wen
46 .
,
+
^, i_
^.+ _
;I"
H' ~/~.." /. lllot--~ole/~ V ) ~ [ '
r+ (^,//1^ .
400
^'.
9 320 uo
240
~
160
~
+ /. /-
..J - /
+
+
~
J
80
B
i
0
Borehole
1
2
3
^
~--m.... ~IW 9
"~"s +
',,+ ^
^
4
m.
l
.~
I+
",,. 4-[ ^+'~.~+
3 I'--12 -] 4 F ' L - ' ] ' 5 [~z ~
T
T
,+
+
+
.~ + + "'., ~
+
+ "
Fig. 2-24 9Results obtained by the CHIM method over beryllium mineralisation in Byelorussia: (A) plan of beryllium anomaly, and (B) distribution of beryllium along profile and schematic geological section; 1- sands, clays; 2- diabases; 3- granites; 4- tectonic disjunctions; 5- ore bodies; 6- zones of mineralisation (reproduced with permission from Bensman et al., 1982).
The role of CHIM in oil prospecting is illustrated in Fig. 2-26. The perimeter of the oil deposit at depth of about 2500 m is satisfactorily delineated by the contour of maximum lead mass extracted by the CHIM method.
Diffusion extraction o f metals (MDE) The MDE method relies on an element-collector analogous to a simplified CHIM element-collector, but without the inner metallic electrode. The application of MDE does not involve an external current. These differences underlie the essentially different physico-chemical basis of MDE. The transfer of the movable forms of elements (generally ions) through the semi-permeable membrane occurs mainly because of two processes, diffusion and migration in the electric field of the double electric layer of membrane. In the absence of an external electric field, as in the MDE method, there is increased diffusion of hydrogen ions through the membrane into the surrounding environment, promoting dissolution of the solid phase. Dissolved elements pass through the membrane into the element-collector but, in contrast to the CHIM method, the
Geoelectrochemistry and stream dispersion
47
/ku~ 0,8 0,6 0,4 0,2 0
"[ 0.04 ~0"020I.
B
0.06
Fig. 2-25. Results obtained by the CHIM method at the Prijutinsky section: (A) concentration of gold in placer, (B) results of extraction of gold by means of CHIM; 1- clays; 2- gravels; 3phyllites; 4- crystal slates; 5- placer (reproduced with permission from Karar and Sakovich, 1989).
accumulated mass (concentration) of metal in the element-collector is not normalised to any electric current. Two modes of MDE are used, ground mode and air mode. In the latter case the gaseous forms of elements are under investigation. Because ion accumulation in the MDE element-collector is due to diffusion and migration of ions in the field of the double electrical layer across the membrane, concentration reaches a maximum over time (Fig. 2-19, curve 3). The maximum time (Xmax) and maximum concentration (Cmax) depend on many factors and therefore measurement time (Xmes) is selected on the basis "lTmes > Xma x. As a rule "l~mes is 20-24 hours. The MDE element-collectors are filled with nitric acid solution and placed in the ground much the same as ground mode CHIM element-collectors. Diffusion of acid from the element-collector into the surrounding environment induces interaction of this acid with the rocks and minerals and brings sorbed forms of metals and the constituents of some minerals (e.g., carbonates) into solution. This results in two effects: (1) higher element concentrations in the MDE element-collector compared to the CHIM elementcollector; and (2) greater anomaly width with the MDE method compared to the CHIM method. Absence of current normalisation in MDE leads to a greater influence on results of the homogeneity in rocks (porosity, humidity, texture and so on) compared to CHIM. A possible solution is normalisation to macrocomponents (Na, Ca and so on) of MDE
48
O.F. Putikov and B. Wen
Pb, ~tg
I,
l~176 0
A 2
II"~..,
4
6
II',.,II'~.II _ ' 7 ~ . I I ~ U " ~ I I " ~ . I I " ~ I I ~
-1 ~ . , - - x ~ - - x . , - - x J - - ~ -
- "-~~"--
-2 -"-~x- '. ,~-'--x' .~,--'-~~- -
---x..,** "~,**-x.,t, "~%~.'~r -3 --- ,-~00 ,.,,..,0, -,,.,0. --.-,-o ................
8
10kin II ~II'%..II",-.II
-,~ --x., - -x.,- - x ~ -
""~~---''~~--"~~''~--" ~ -
~'-
-x.,** -x...o--'~ro-~....~'x.. -oo ,-~,0,;.k-_..~"-~
li
li "x-,Ill 2t'x~--i 3 ['x.,**l 41--::=i 5 ~ Fig. 2-26. Results obtained by the CHIM method over an oil deposit in Byelorussia and schematic geological section: 1- Permian-Cretaceous-Quaternary clays, sands, coals; 2- marly siliceous clay formations; 3- Carboniferous sand-clay formations; 4- middle-late Devonian sandstones, aleurolites, marls; 5- oil deposit (reproduced with permission from Ryss et al., 1990).
data (Fig. 2-27). The advantage of an MDE survey compared to a CHIM survey is its low cost. Consequently MDE is applied in reconnaissance surveys at scales 1:250001:10000 and more detailed follow-up is performed with a CHIM survey. The results of MDE investigations along a profile over the Mirona copper-nickel sulphide ore body in the Pechenga ore field (Kola peninsula) are shown in Fig. 2-28. The ore body, grading 0.4-1% Ni, is related to an ultrabasic intrusion in tuffaceous sedimentary rocks. These rocks are covered by a moraine 10-15 m thick. The MDE element-collectors over the ore body have values up to 12.5 mg/1 Ni, 10 mg/l Cu and 15.8 mg/1 Fe, compared to background concentrations of 1-2.5 mg/1 Ni, 0.5-1.5 mg/l Cu and 2.5-3 mg/1 Fe. This survey was carried out in order to determine the nature of gravimetric anomalies and induced polarisation and other electrical prospecting anomalies. A number of geophysical anomalies were considered non-prospective as a result of low concentrations of nickel and copper in the MDE survey. Subsequent drilling has verified this conclusion. However, the MDE survey at Karic-Gavr, on the edge of the Pechenga structure, produced high-contrast nickel and copper anomalies in the region of an IP anomaly (Fig. 2-29) and a subsequent borehole revealed coppernickel sulphide mineralisation, grading 2.4% Cu and 0.12% Ni, at a depth of 120 m (Fig. 2-30). The stockwork-type porphyry-copper ore bodies at Kyzyl-Tu in central Kazakhstan are covered by allochthonous sediments 20-30 m thick. The conventional geochemical
Geoelectrochemistryand stream dispersion
49
C/~m~ 1
. . . . . . . . . . . . . . . . . . .
,a .....
io
0
iv I
10
I :1 I
20
c/~
0.5 A I : ~ ~ I 0
10
I
20
I
30 40 x,hour
I .....
[ ............
.....
Ca
9
0.5
=
50
.-~ I
30 40 x,hour
I
60
='....... I 50
iCo 60
Fig. 2-27. Curves of accumulation of the macrocomponent (Ca) and microcomponent (Co) in a MDE element collector from samples of surface sediments of different composition: 1- watersaturated peaty sediments; 2- clay rocks; 3- loamy sands (from Testury, 1996).
survey does not detect these ore bodies (Fig. 2-31A) whilst the MDE survey yields goodcontrast copper anomalies over the ore bodies (Fig. 2-31B). The Korbalikhinskoe deposits (Rudny Altay) are blind cupriferous pyrrhotite and polymetallic ore bodies in bedrock patchily covered by soft autochthonous sediments 510 m thick. The ore bodies take the shape of ribbons and lie on the contact of acid tufts, tuff-sandstones and aleurolites at depths of 75-350 m in the southeast of the area and 500-1000 m in the northwest. The MDE results for copper and lead concentration distribution allow the detection of the ore bodies at depths of 75-450 m. A cupriferous ore body corresponds to a copper anomaly (Fig. 2-32, southeast), whilst the polymetallic ore bodies correspond to anomalies of copper and lead (Fig. 2-32, northwest).
Organometallic (MPF) and thermomagnetic (TMGM) patterns The MPF and TMGM prospecting methods are based on the use of metallo-organics (fulvates and humates of metals) and oxides of iron and manganese (metals bound in oxides and hydroxides of iron and manganese). These forms of metals are the result of the secondary fixation of the movable forms in rocks and have features such as (1) increased concentration coefficient and (2) only a weak bond with their initial geological source (in comparison, for example, with the movable forms collected in CHIM and MDE). Samples for MPF are taken from the humus-enriched layer at a depth of 5-10 cm, and samples for TMGM are taken from the sand-clay layer at a depth of 15-20 cm,
50
O.F. Putikov and B. Wen
C ~tg/ml 12 9
Cu
Fe l
/ 6 3 ' - ~ - ~ ~ ~
i
o 1
10
5
15 0 O0
llf
I 21J-' I ~
~
,50m,
Fig. 2-28. Results obtained by the MDE method over a sulphide copper-nickel ore body in Pechenga ore field, Kola peninsula, Russia: 1- phyllites; 2- peridotites; 3- diabases; 4- poor and rich ores. A
'V'I
C, mg/1
B
9
1o
1~ 1 % ,
4 j16
,,
18Prof.tOC, mg/I
.,.'
100m t-----t i
10
I
15
i
20
Fig. 2-29. Results obtained by the MDE method at the Karik-Javr section in Pechenga ore field, Kola peninsula, Russia: (A) plan of MDE profiles with location of traverses and contour (I) of MDE anomaly for nickel and copper; (B) distribution of nickel and copper along profiles 9 and 10.
enriched in iron and manganese. Processing and analysis of samples are accomplished in the laboratory. In the case of MPF the fulvates and humates of metals are extracted from samples using sodium pyrophosphate solution. After determining the content of metals (Me) and carbon (C) in the extract, a concentration coefficient (normalisation of content of metal to carbon, Me/C) is used to determine the concentration distribution of metals from depth. In the case of TMGM it is necessary to measure magnetic susceptibility of samples before and after annealing, to separate the magnetic fraction and to analyse both
Geoelectrochemistry and stream dispersion Geol. 3.~u
section
O. 10
20
30
Diagrams of logging C~zrent.mA ~ . CGS o. ;o ;o o o.~.,o-,
9
51
diameter, r r ~
60
100
de
o ~ ~,-*:,-
.-+.-
f
._<--
i
25 ~ _ . ----~
50 p'v ~ v
75
......
!
i
h,m
Fig. 2-30. Geological column and logs of borehole 3427 at the Karik-Javr section in Pechenga ore field, Kola peninsula, Russia: 1- moraine; 2- garnet-biotite gneisses; 3- shadow migmatites; 4alumina gneisses; 5- plagio-amphibolites; 6- biotitc-chlorites; 7- biotite gneisses; 8- copper-nickel sulphide ore.
the magnetic fraction and the initial sample for 25-30 elements. The resulting MPF and TMGM patterns show anomalies that are wider and smoother than CHIM and MDE anomalies. For this reason and the low cost of MPF and TMGM, these methods are applied at the regional investigation stage at scales of 1:500000-1:1000000 (Ryss et al., 1988) and for prospecting at scales of 1:50000-1:200000 (Antropova et al., 1986). Figure 2-33 shows the results along a profile sampled at intervals of 5-20 m across known sulphide mineralisation in quartz veins in late Palaeozoic igneous rocks covered by 60 m of Quaternary sediments. Conventional geochemical survey data fail to detect the mineralisation (Fig. 2-33A). The MPF results, on the other hand, give well-defined anomalies for gold, arsenic, silver and copper (Fig. 2-33B). These methods also have applications in oil prospecting. Figure 2-34 shows that high Ni
Co
values of the multiplicative coefficient -6- x -~- calculated from MPF data delineates an oil reservoir of 4 • 6 km. At the Kristalnoe oil deposit TMGM and MPF outline the boundary of the deposit by high concentrations of
Zn
Mo
Co
Pb
Co
-~-, c ~-~-, ~ - ~-6- ~
Ni
determined by MPF and Ni, V, Cr determined by TMGM; the multiplicative coefficients Cu
Ni
Co
-'6-" x --~ x --~ and -if- x -~- x ~- calculated from MPF data are also effective. By comparison a conventional geochemical survey for Cr does not give positive results (Fig. 2-35).
52
O.F. Putikov and B. Wen Cu,%
A
0.03'~
0.02~ o,o1|.
.
6
_
260
meters
400
600
400
600
Cu, I~g/ml ~
0
200
meters
200
-I-
~F'cl 2~'--;1 3,,~'~rl,, ~
5|
Fig. 2-31. Results obtained over the Kyzyl-Tu porphyry-copper deposit, central Kazakhstan, by (A) surface lithogeochemical survey and (t3) MDE method: 1- clays; 2- granites; 3- limestones; 4- sites of MDE measurement; 5- ore bodies (reproduced with permission from Dukhanin, 1989).
Fig. 2-32. Results obtained by the MDE method over the Korbalihinskoe polymetallic deposit in Rudny Altay, Russia: 1- clays; 2- lavas; 3- lava-breccias; 4- diabases; 5- aleurolites; 6- tuffstones; 7- ore bodies (reproduced with permission from Dukhanin, 1989).
Geoelectrochemistry and stream dispersion
53
A
Ag,% 1 "10 5
1"10~ Mo,% 1.10 ~ I "10 5
Sn,% 1-10~ 2"10 5 Cu,% 1 "102 2 "103 I "10"~
l.lej Au,g/t I "10 2
/
1 "lO j
n.10" % Cu/C,As/C n-10"~% Ag/C n.10"~% Au/C
B A //
rlyl /
-
21T-~
16Oral
3m
~ Y y
Yy
Fig. 2-33. Results obtained over quartz-poor sulphide ore zone by (A) lithogeochemical survey and (B) by the MPF method: 1- unconsolidated overburden; 2- andesite-dacite porphyrites; 3- ore zone (reproduced with permission from Abaturova and Shkarupa, 1981).
GEOELECTROCHEMICAL EXPLORATION
Physico-chemical basis The methods of geoelectrochemical exploration are based on investigation of the polarisation curve (voltammogram), which records the dependence of a current through an electronic conductor on its electrode potential. The electrode potential of an electronic conductor is its potential measured with respect to an auxiliary standard non-polarisable electrode which is immersed in the host ionic conductor in direct contact with the surface of the electronic conductor. Two of these methods, the Contact Polarisation Curve (CPC) and the Contactless Polarisation Curve (CLPC) are used to investigate ore bodies with electronic
54
O.F. Putikov and B. Wen
km (Ni/C ~2o/C)'10" 10'-
,~ott 8
6-
4-
2-
0
i
I
i
i
/
2
4
6
8
1o
kin
Fig. 2-34. Application of MPF method for locating oil deposit contours: 1- contour of the YuznoRadovskoe oil deposit in the Voiga-Ural oil-gas province, Russia; 2- traverse lines of the MPF survey; 3- result along traverse lines of the multiplicative coefficient (reproduced with permission from Vasilieva and Voroshilov, 1995).
conductivity (sulphides, magnetite, graphite and so on). To understand polarisation curves of ore bodies as polymineralic bodies it is necessary above all to investigate polarisation curves of the separate minerals that have electronic conductivity. Such investigations were made in the 1960s under the leadership of Ryss (1969). An apparatus is used to study both cathodic processes, when an electronic conductor is a cathode and has a negative potential, and anodic processes, when an electronic conductor is an anode and has a positive potential (Fig. 2-36). The polarisation curves of minerals are obtained using the galvanodynamic polarisation mode, in which current is a linear function of time. The main results of these studies (Fig. 2-37) are that: (1) different samples of the same mineral give more-or-less similar (stable) forms of polarisation curves; (2) polarisation curves of minerals have a non-linear (stepped) form, each step reflecting a definite electrochemical reaction in terms of potential (q~, q~2.... ) and limiting current (Iliml, llim2,--.), and (3) values of the electrochemical reaction potentials for each mineral are more or less stable (characteristic). This work further showed that there are two types of polarisation curves of minerals, one for cathodic processes and another for anodic processes, with two electrochemical reactions at potentials q~ and q~2 (Fig. 2-37, curve I) and one electrochemical reaction at the potential q~! (Fig. 2-37, curve II). When recording the polarisation curves of minerals,
Geoelectrochemistry and stream dispersion
55
TMGM
5t
/,./
\~Ni
~_V\ I//
"x,.~
_
Cr.lO-2% 7"5 5f.~" 2.5 0
rrM
Cr ,t~
It
Zn/C M~C
Fb/C
Cu/C
oCl o,Tc .... 0
3km
,,oo
~f I ~
2~
3[m][ml]] ~
5~
Fig. 2-35. Results obtained by the TMGM and MPF methods over an oil deposit in the VolgaUral oil-gas province, Russia: 1- Late Devonian clay, fractured limestones and dolomites; 2- Late Devonian sandstones, aleurolites and argillites; 3- oil pool of the D3 seam; 4- profiles of the MPF survey; 5- curves of concentration distribution of elements (from Vasilieva and Voroshilov, 1995).
samples are taken simultaneously and analysis of the products of electrochemical reactions in solid, liquid and gaseous phases is carried out. In this way it was established that a number of reactions of an electrochemical nature take place on the mineral surface. In particular, studies of the cathodic processes have shown that the majority of minerals have two wave steps on the polarisation curve. The first wave step, as a rule, corresponds to potentials in the interval -0.3 to -1.2 V. For divalent metal sulphides this step corresponds to the following electrochemical reaction,
56
O.F. Putikov and B. Wen I A
q~
A(M)
Fig. 2-36. Scheme of laboratory installation for recording polarisation curves of electronic conductors: 1- electronic conductor (mineral or metal); 2- solution of electrolyte; A- current electrode; B- auxiliary current electrode; M- measuring electrode; N- non-polarisable measuring (reference) electrode; q~- potentiometer; CS- electric current source; I- ammeter (reproduced with permission from Putikov, 1993).
II
0
q~l
q)l
I
q0z
q~
Fig. 2-37. Typical shapes of polarisation curves of electron-conducting minerals: I- current; q0potential (reproduced with permission from Putikov, 1993).
M e S + 2e----~ M e ~
S 2"
(2.25)
This reaction reflects the cathodic sedimentation (deposition) of the pure metal. The second wave step lies mainly in the interval-1.3 to -1.5 V and corresponds to the hydrogen emission reaction, which is related to dissociation of water molecules, H20 r
H + + OH-
2H + + 2e---~ H2']"
(2.26)
Geoelectrochemistry and stream dispersion
57
For anodic processes there are also two wave steps on the polarisation curves at potentials q~ and cp2 corresponding to two electrochemical reactions. The first anodic electrochemical reaction for sulphide at the potential q~t corresponds to its anodic decomposition (transition of ions of metals in solution), M e S --~ M e 2 + + S ~
2e.
(2.27)
The second anodic reaction for sulphides reflects the further oxidation of sulphur. For example, in the case of galena, PbS + 4 H 2 0 ~
Pb 2+ +
8042--F
8H + + 8e-
(2.28)
As a result of these studies on the cathodic and anodic mineral processes, the electrochemical reaction potentials for the main electron-conducting minerals were compiled (Table 2-I1). Processing of CPC and CLPC data includes determination of the electrochemical reaction potentials that take place on the surface of an ore body. For CPC data this is done directly using the polarisation curves from the ore body; for CLPC it is necessary to use additional processing and calculations. Matching these potentials with the corresponding tabulated potentials affords estimation of the mineral composition of the
TABLE 2-1I Electrochemical reaction potentials of main ore minerals relative to a saturated calomel electrode (Ryss, 1983) Mineral
anodic processes q~l, V
Magnetite Pyrrhotite Pyrite Chalcopyrite Chalcocite Sphalerite Galena Pentland ite Molybdenite Graphite
1.60+0.10 0.60+0.10 0.60+0.05 0.15+0.10 0.15+0.10 -0.05+0.10 0.30+0.10 0.40+0.05 0.80+0.05 1.50+0.05
qh, V -0.90+0.05 sometimes 0.90-1.20 0.70+0.10 m 2.30+0.10 1.70+0.10 m --
cathodic processes q~l, V
tp2, V
sometimes -0.70 -0.50+0.05 -0.50+0.10 -0.60+0.10 -0.60+0.05 - 1.20+0.05 -0.75+0.10 -0.35 +0.05 - 1.25+0.05 - 1.55+0.05
-1.45+0.10 - 1.50+0.10 -1.30+0.10 1.40+0.10 - 1.00+0.05 -2.10+0.10 -1.50+0.10 - 1.10+0.05 -
58
O.F. Putikov and B. Wen
ore body. This answers certain questions related to the quality of the ore body. Quantitative problems are addressed by using the limiting current values of the electrochemical reactions (wave height) to determine the total surface area, size, concentration and mass of minerals (metals) of an ore body. First consider the limiting current density, because the limiting current is the product of the limiting current density and the surface area. In the general case the current density is, J=jo+jM+jc
(2.29)
w h e r e , jD = diffusion current density, jM = migration current density, jc = convection
current density. When the electric field intensity is low and the liquid phase is immovable we can neglect jM and jc such that, j =jD
(2.30)
As is evident from Fig.2-37, only the first anodic and cathodic reactions (equations (2.25) and (2.27) for sulphides) have finite limiting currents. According to equations (2.25) and (2.27) some ion-carriers of current (sulphide ions or metal ions ) transfer from the solid phase into the liquid phase. For a planar ore body the concentration C of the ion-carriers of current in the ion-conducting host rocks satisfies the one-dimensional differential diffusion equation,
02C ~2
1 c~
D Or
-0
(2.31)
where, x = distance from the mineral surface of the ore body, D = diffusion coefficient of the ion carriers of current in the host rocks, x = time. The initial condition can be written as,
C I,-=o - C o
(2.32)
where Co is the initial concentration of ion-carriers of current in the host rocks. For the galvanodynamic polarisation mode, in which current density is a linear function of time, from (2.30) we can write,
Jl,=o- Jo Ix=o = C lx~oo -> C o
-zFD- l
x=O
-ar
(2.33) (2.34)
Geoelectrochemistry and stream dispersion
59
where, z = charge of the ion-carriers of current, F = Faraday constant, a = rate of current density change with time (current density rate). Applying the integral Laplace transformation relative to time x, Putikov (1993) obtains a solution of equation (2.31) under conditions (2.32) to (2.34),
a C-C~ +z F ~
x
2 ~e~Cf~d~,_ x --~ ~
1
(2.35)
where, Erf(y) = 1-erf(y), and erf(y) is the probability integral,
erf (y) = -~K ~e-~dzl 2
At the mineral surface for x = 0 we have, from (2.35), 4 CIx=o -Co +~ zF 4a ~
r
3/2
(2.36)
At the limiting time for a given electrochemical reaction x = 171im,the ion-carriers of current near the mineral surface reach a maximum value, Cmax, which depends upon the dissolution production of the corresponding compound. Taking these points into account we can obtain from (2.36) for x = Xlim,
C max
4 a ~ = Co + -~ zF 4nO r)jm
V3 (c
- L
m.x - C o
(2.37)
)zF ~-~l~a -~
(2.38)
With respect to (2.33) it is possible to write,
Jlim --
a g-,i m
E3
'~- (Cma x - C 0 )
zF~
l
~ a '~
Taking logarithms of equation (2.39) we obtain,
(2.39)
O.F. Putikov and B. Wen
60
J lim, mA/cm2
I
i
10"
A 0.1
"~ 1-10 -4
!
II
j IIi
!
1"10 -3
aj ,
1"10 1
1-10 -l
1
mA
,
0,17 ) crn
s
Fig. 2-38. Dependence of limiting current density on rate of current density excitation. Surface of spherical electrodes, cmZ: o- 1.5; A- 3.0; v- 10; x- 40; [-~- 120; I- experimental data for spherical electrodes; II- theoretical dependence for a plane electrode (reproduced with permission from Korostin, 1976).
1 Iog~o J,~rn - 3 IOg~o a + Iog~o b
(2.40)
where,
-
]
This means that for the bi-logarithmic co-ordinates of a planar ore body there is a linear dependence of the limiting current density on the current density rate. The corresponding slope coefficient is 1/3. Results of laboratory physico-chemical simulation (Korostin, 1976) confirmed this dependence for the interval of current density rate 'B' (Fig. 2-38).
Contact polarisation (CPC) The essence of the CPC method consists of recording and interpreting polarisation curves obtained when the polarisable electrode is an electron-conducting ore body (Fig. 2-39). One pole of the current source, electrode A, is connected to the ore body by means of a special device (e.g., in a borehole intersection through the ore body). The
Geoelectrochemistry and stream dispersion
N
61
irlil
Fig. 2-39. The basic mode of the contact polarisation curve (CPC) field installation: 1- ore body; 2- borehole; 3- compensation resistor; 4- electric current source; 5- resistance of shunt; 6potentiometer; 7- recorder; A- current electrode; B- auxiliary current electrode; M- measuring electrode; N- non-polarisable measuring (reference) electrode (reproduced with permission from Ryss, 1983).
other pole, electrode B, is placed in the host rocks beyond the ore body. Recording comprises the simultaneous measurement of current, which flows across the ore body surface, and the electrode potential of the ore body. The current is supplied as a linear function of time. In order to measure electrode potential of the ore body, one measuring electrode, M is positioned inside the ore body and another measuring electrode, N, is placed in the host rocks directly in contact with the ore body. In practice we cannot put electrode N at the desired point, so it is impossible to measure directly the electrode potential of the ore body in the field. However Ryss (1969, 1973) found that by putting another measuring electrode on the ground surface in the host rocks beyond the ore body, it is possible to eliminate the difference between the measured voltage and the electrode potential of the ore body by means of the compensation method (Fig. 2-39). In this case the difference of potentials AUMNbetween the two measuring electrodes is,
AUMN=(a+~AU; where, q~ is the electrode potential of the ore body (the useful signal), and )-".AUi is the sum of the potential differences in different parts of the measuring circuit (in the ore body, in the host rocks and so on) which comprises the disturbance. Usually l yAu, I>> and this does not allow us to observe the polarisation effects without using a special technique of potential separation, which is based on the different natures of the potentials q~ and ~AUi. Since q~ is produced by electrochemical reactions, whereas ~AUi satisfies Ohm's law, q~ is a non-linear function of current, but Y.AUi is a linear function. This difference between q~ and ~AUi is used in CPC for elimination of the
62
O.F. Putikov and B. Wen
T
~9
II i
;2
1
............i
.... ...i
,t__ .....
.x.
I
13
/
4"" ~i ./" "'~, l~.]".." i /" 0
q)l
(P2
q)3
Fig. 2-40. Influence of linear disturbances on the shape of the polarisation curve: I- ideal polarisation curve; 2- linear disturbances with inadequate compensation; 3- polarisation curve under condition of inadequate compensation of linear disturbances; 4- linear disturbances with over-compensation; 5- polarisation curve under condition of over-compensation of linear disturbances (reproduced with permission from Putikov, 1993).
~AU~ disturbance. To do this a compensation voltage AUcom is applied to the measuring potential set. The compensation voltage Ucom is proportional to the current, but has an opposite sign to the voltage ZAU~. The compensation voltage is taken from a special resistor (3) in the current circuit (Fig. 2-39). Then at the measuring potential set inlet we have, AU r,,~ - AU MN + AU :or. - ~n + ~
AU ; + AU :or. - ~ + a ( A U )
where AUmes is the measuring voltage and 6(AU) is the disturbance voltage, which is a linear function of current. In the case of full compensation 5(AU) = 0, ZAU~ = -AUcom and AUmes - q). For proper compensation it is necessary to change the compensation resistor value, then we obtain an ideal polarisation curve (Fig. 2-40, curve 1). With inadequate compensation or over-compensation we obtain distorted polarisation curves (Fig. 2-40, curves 3, 5). It is not essential to achieve full compensation of disturbances for the satisfactory determination of the electrochemical reaction potentials (q)l, q)2 ...) and corresponding limiting currents (Iliml, Ilim2 ...). Using polarisation curves with some inadequate compensation or with some over-compensation may even be successful. The compensation operation is necessary only for good representation of the polarisation wave steps. The current source and current circuit must also satisfy definite requirements (Ryss et al., 1978: Ryss, 1973), including: (1) the possibility to change the current over time in different ways up to the maximum values (160-250 A); and (2) minimal resistance in the current circuit (including cable, electrode A, electrode B and all other parts). For
Geoelectrochemistry and stream dispersion Mineral Chalcopyrite Pentlandite
I 9, V [ +0.17 [s
q~, V Penflandite Chalcopyrite Pyrrhotite ..... Penflandite Pyrrhotite
63 Mineral Pyrrh.o.tite Pyrrhotite
-0.32 -0.5
~,v
II~,v ]1 -0.5 1-1.5
+0.6
+0.9
Mineral
Pyrrhotite Pyrrhotite
-0.14
-0.15
I,A
I, A 15
15
1[
10
B
A
! !
\ \ +1
at _,,4,, V~," .,4",' ,' _ 0
// / / ,/ I/ _ I" .-1 ~ V v
J./ +1
0
_ -1
q~,V
r
Fig. 2-41. The CPC polarisation curves of sulphide ore bodies, Kola peninsula, Russia: (A) copper-nickel ores; (B) pyrrhotite mineralisation.
example, to introduce 160 A using the 12-channel Russian SPK instrument, it is necessary to have the current circuit resistance less than 0.56 f2. In order to meet this requirement a special cable with a resistance of 0.7f2/km is used. The SPK instrument can then investigate ore bodies from metres to kilometres in size and at depths up to 1000 m. The best CPC results are obtained on ore bodies with favourable texture (massive, veined, banded, laminated, etc.), when the ore body as a whole is an electronic conductor. For such ore bodies CPC is able to: 9 determine the mineral composition of ore bodies; 9 estimate the total surface area and size of ore bodies; 9 indicate the mineral concentration and reserves of ore bodies; 9 assist correlation of different ore intersections; 9 assist investigation of ore body location. Determination of mineral composition of the ore body requires finding the electrochemical reaction potentials that take place on the ore body surface due to the current that is introduced. Values of the electrochemical reaction potentials correspond to the cross points of lines tangential to the flexure of the polarisation curve and the potential axis (Figs. 2-41 and 2-42). The average mineral composition of an ore body is
64
O.F. Putikov and B. Wen
mineral " l o w ] r mineral Chalcopyrite ]+0,18I -0,53Pyrite -0,6 Chalconvrite -1,18Sphalente - 1,45Chalcopyrite Pyrite I,A
Ji
~100
g~,V + 1.0
0
- 1.0
q~,V
Fig. 2-42. The CPC polarisation curves of the great polymetallic ore body in Rudny Altay, Russia (reproduced with permission from Ryss, 1973).
found from comparison of these electrochemical reaction potentials with the standard values of potentials for minerals (Table 2-II). The CPC method gives information not simply in the vicinity of a borehole but about the whole ore body, for example, at a distance 200-300 m from the measurement borehole. Thus CPC is considered a remote sensing method (Ryss, 1973; Putikov, 1995). For example, Fig. 2-43 shows two boreholes, 2012 and 2014, originally drilled to check an electrical survey anomaly. Both intersected only the pyrite ore. However, CPC results obtained from borehole 2012 show the presence of polymetallic mineralisation. This polymetallic ore body was subsequently intersected by underground borehole 326, at a distance of 200 m from borehole 2012. The limiting current Ilim i of the electrochemical reaction of a mineral i satisfies the expression, Ilim i = Si j lim
(2.41)
where S~ is the mineral surface on the contact between the electron-conducting ore body and the ion-conducting host rocks and jlim is the limiting current density of the electrochemical reaction. For a concentration C~ of the mineral i we can write, Ci=Si/S0 '
(2.42)
Geoelectrochemistry and stream dispersion
65
I,A
q),V -0,44 -0,58 -0,78 -1,20
, 15
mineral ,Pyrite Chaleopyrite Galena Sphalerite
1o 5 o 0,5 1[ xr-z'] 2 [ " ~
3 ~
4[~
1.0
%
5["#~ 6['V'] 7 ~
Fig. 2-43. The CPC polarisation curves of a zoned polymetallic ore body at Rudny Altay, Russia, with current electrode in borehole 2012: 1- unconsolidated overburden; 2- acidic tuffites; 3sericite-quartz slates; 4- dacitic porphyrites; 5- limonitisation; 6- polymetallic ore; 7- pyrite ore.
where, So' is the total physical surface of the ore body as an electronic conductor. Comparison of expressions (2.41) and (2.42) gives,
S~
'-
1
J li~
[ lira i 9
C
(2.43)
But it is difficult to use equation (2.43) in practice because of difficulty in determining values of So' and j~m. TO avoid this difficulty, Ryss suggested the analogue empirical correlation,
limi S~176176C~
(2.44)
where So is the square of the conventional (geological) surface which bounds the calculated mineral reserves of an ore body, and kl0o is an empirical coefficient (the transient coefficient). If the conventional surface So coincides with the physical surface So', from equations (2.43) and (2.44) we obtain,
O.F. Putikov and B. Wen
66
h" 100
-
1 J lira
[mZ/A]
(2.45)
In this case the transient coefficient kl00 has a clear physical meaning, the inverse value of the limiting current density. In the general case the relationship between kl00 and ji~n is more complicated and has a statistical character. For determination of transient coefficient k~00 values, Ryss (1973, 1983) used tens of known deposits with different mineral compositions, at different depths and in different geological situations, but with known geological surface So, mineral concentration Ci and Ii~ i values, which were obtained by CPC polarisation curves for each ore body. It was found that for cathodic electrochemical reactions, on average, k~00 = 500 m2/A and for the anodic electrochemical reactions kl00 = 200 m2/A for pH>5 (in dry regions) and kl00 = 100 m2/A for pH<5 (in wet regions). The Ryss formula (2.44) is the basic equation of the CPC method for determination of the quantitative parameters of an ore body. The polarisation curve of the ore body reflects a number of electrochemical reactions (i = 1, 2,..., n). Thus we can obtain the limiting current value Ilim i for each reaction and concentration of the corresponding mineral (as a result of the core investigation). With respect to the dependence on current that is analogous to (2.44), we can calculate a partial meaning of the total ore body surface that corresponds to the reaction number i,
Soi - klo o
Ilimi
Ci
100
(2.46)
where, Ilim i = limiting current for the reaction of mineral i in A, reduced to the standard excitation velocity a'st = 15 mA/s, kl00 = the Ryss transient coefficient, m2/A, Ci = concentration of the mineral i in %. In accordance with equation (2.39) the reduced limiting current Ilim i is defined as,
Ilim i - - / l i m
mes.i
ast 9 a
(2.47)
where Ilim mes i is the limiting current of the reaction i, measured from the polarisation curve for the excitation current rate a ~ It is sufficient, in principle, to perform the CPC investigations only in one borehole, in which case the exact concentration value for the mineral i, as a rule, is unknown. Deflections from their exact values of the known concentrations Ci for different minerals are accidental and the calculated surface S0~ is also accidental. The most probable value
Geoelectrochemistry and stream dispersion
67
of the ore body surface is an average S0av of n values that were found from equation (2.46),
l
Soav -
2 . Soj
n
(2.48)
i=1
After substitution of S0i by S0av in equation (2.46) it is possible to define the concentration of the separate minerals Ci. The error can be reduced by using the average borehole concentration Ci,v instead of concentration Ci. Using CPC, for each ore cross-section we obtain n independent values of S0i ( i = 1, 2,...,n) according to equation (2.46). Then the number of boreholes required to calculate with the given accuracy the total ore body surface S0av (or concentration Ciav ) is less than relative to the number of boreholes required without using the CPC method. This demonstrates the economic effectiveness of the CPC method for evaluation and preliminary exploration of ore deposits. If an ore body of thickness H has a lens shape of 1~ x 12 and H<
/2
After substitution of the average values of concentration C~av and total ore body surface S0av in equation (2.46) we obtain, So~ v C i~v - kloo I ,im ~ " 1 0 0 where S0avCiavis the reserve of the i-th mineral in m 2 %. Then the reserves Mi of the i-th mineral in mass units are (Ryss, 1973, 1983),
M j-
Soav 2
CiavH~vd~v
-
klool tim i H avdav 2
" 100
(2.49)
where Hav and dav are the average thickness of the ore body and ore density respectively determined by means of drilling. An example of CPC polarisation curves obtained at a sulphide copper-nickel ore body on the Kola peninsula are shown in Fig. 2-44 and the interpretation of these curves is given in Table 2-111 (Ryss, 1973). The geological data used are the concentrations of minerals and metals in relative mass units (Cpnt~0.03; Ccp~0.015; Ccp+pyr~0.07;Cy'~0.l; Cni~0.01; Ccu~0.005) and the size parameters of the ore body (Hav~2m, dav=3500 kg/m3, S0~15• m2). These are used in equations (2.46) and (2.49) with kt00 = 500 m2/A for
O.F. Putikov and B. Wen
68 mineral tg,V Chalcopyrite +0.17 Pentlandite +0.4
,V mineral -0.32 Pentlandite -0.5 Chalcoowite -0.5 P~hotite -1.14 Penflandite
I, A
n~t,A
1.0
.0
0
-1.0
~o,V
Fig. 2-44. The CPC polarisation curves of a copper-nickel sulphide ore body, Kola peninsula, Russia (reproduced with permission from Ryss, 1973).
the cathodic reactions and with k~oo = 100 m2/A for the anodic reactions for calculation of the total ore body surface So and the masses of pentlandite, Mpnt, chalcopyrite, Moo, and the combined mass of pentlandite, chalcopyrite and pyrrhotite, My. The results of the calculation are Mpnt = 1.3 x 106 kg, Mcp = 0.65 x 106 kg and My = 4.5 x 106 kg. The ore body surface values So for different electrochemical reactions are approximately equal to the average value, S0av = 12.8 x 103 m 2.
TABLE 2-III Interpretation of CPC polarisation curves from Kola peninsula (Ryss, 1973) q~, V 0.17
Mineral
llim, A
Chalcopyrite
So, 103m2
1.8]
12.0] 5.5
0.40 -0.32
Pentlandite Pentlandite
-0.50
Chalcopyrite ]
-0.50 1.14
Pyrrhotite Pentlandite
-
3.7 J 1.01 ]
J
2.54 1.53 J
~ 12.2 12.3J 16.8 ] ~ 12.7 11.0 J
Geoelectrochemistry and stream dispersion ,
2,,..~
16
i
:J. 7
A
Mt
M2
M3
69
M4
Ms
! ,
M6
M7
M8
B
v
cathodic t egion
anodic region
Fig. 2-45. Scheme of the CLPC field installation: 1- transducer of compensation; 2- ammeter; 3galvanic decoupling unit; 4- potentiometer; 5- recorder; 6- electrical current source; 7measurement channels switch; A, B- current electrodes; Mi,...M8 -measuring non-polarisable electrodes; N- remote non-polarisable reference electrode (reproduced with permission from Ryss, 1983).
Where it is necessary to correlate different ore-body cross-sections or to investigate ore-body morphology, it is possible to use special CPC field installations, the correlation CPC mode and the prospecting CPC mode.
Contactless polarisation (CLPC) The CLPC method is also intended for investigations of electron-conducting ore bodies. It differs from the CPC method in that the CLPC method does not require direct contact (earthing) of one electrode in the ore body. Both current electrodes, A and B, are placed in the host rocks on ground surface (Ryss, 1973, 1983). Current, as a linear function of time, is introduced into the ground by means of these electrodes (Fig. 2-45). If there is an electron-conducting ore body at depth this current flows into the ore body at one end (the zone of cathodic polarisation) and flows out of the ore body at the other end (the zone of anodic polarisation). This results in a potential difference with different signs and values in different parts of the ore body. This double electrical layer creates a secondary electrical field in the host rocks that can be measured by means of the measuring electrodes, M and N. For this measurement electrode M is moved along a profile to successive positions M~, M2, ..Mn whilst electrode N is placed at "infinity" (Fig. 2-45). For each position of electrode M the CLPC polarisation curve is recorded. This curve is dependent on current I in the
O.F. Putikov and B. Wen
70
Ilim I a
lliml
C,,i e
|
....
Fig. 2-46. Shapes of polarisation curves of the same surface of an electronic conductor for: (1) CPC method: and (2) CLPC method (reproduced with permission from Putikov, 1993).
current circuit of electrodes A and B and on the potentials of electrodes M, measured with respect to electrode N. In the CLPC method there is a large distance between the measuring electrodes and the double electrical layer compared with that in the CPC method. Therefore the CLPC method does not measure true potentials of the electrochemical reactions, q)~, q~2, ... and their differences Aq~, but apparent potentials %~, qga2, ...and apparent differences of potentials A%, where [A% ]< I Aq, I (Fig. 2-46). The total current I in the CLPC method is the sum of the internal current I~n, which flows through the ore body, and the external current Iext, which flows through the host rocks. The corresponding total current for the CPC method is only the internal current I~,. Therefore there is a link between limiting currents of the CLPC method and the CPC method (Putikov, 1993),
IlimCLPC = Ilimi,, + Iext -- IlimCeC + I,..,r IlimCLP C > IlimCPC The CLPC polarisation curves have an additional peculiarity arising from the fact that in the CLPC method both measuring electrodes M and N are in the host rocks, whilst in the CPC method electrode M is in the ore body, and the auxiliary nonpolarisable electrode N is placed in the host rocks. As a result the signs of the measured potentials of CLPC and CPC are opposite. For example, if under cathodic polarisation in CPC we have q)<0, then in the vicinity of the cathodic zone for CLPC we have q)>0 and so the signs of the potentials are opposite (Fig. 2-47). Values of the apparent potentials qh~ and their differences A(Da depend upon distance from electrode M~ to the nearest polarisation zone of the ore body and they decrease with distance (Fig. 2-47).
Geoelectrochemistry and stream dispersion
ACI
A ii
Ill|re.C] ...............
A0
A
T'?
iI . . . . I . . . .1. . . . . . ---q,-~
'"
71
A =.A/
,
....................
4%
C
i
+'
!
t
A
Fig. 2-47. Shapes of different types of CLPC polarisation curves in a profile over an ore body under polarisation: CA- cathodic-anodic; AC- anodic-cathodic; C- cathodic zone of polarisation of ore body; A- anodic zone of polarisation of ore body; o~, co2, oJ4, co5- angles of visual range of polarisation zones for points M~, M2, M4, M5 respectively (reproduced with permission from Putikov, 1993). There are two modes of CLPC polarisation curves: CA, where the cathodic zone of the ore body is on the left and the anodic zone is on the right; and AC, where the position of the zones is reversed. In CA mode the polarisation curves that correspond to points on the left to centre of the ore body indicate in general the cathodic reactions. These reactions correspond to the defmite sign of the potentials (Pal, (Pa2, "" and to the definite limiting currents I~m c 9The polarisation curves on the right to centre of the ore body (Fig. 2-47, points M4 and Ms) indicate the anodic reactions with the opposite sign of the potentials (P.~, (P.~2and with values of the limiting current Ilim a- Above the centre of the ore body (Fig. 2-47, point M3), due to the compensation effect of the cathodic and anodic fields, we have measured potentials of the secondary electric field close to zero. The AC polarisation curves differ from CA polarisation curves in the sign of the potentials and in the values of the limiting currents of the electrochemical reactions at each position Mi. In summary, the main features of CLPC polarisation curves (and their dissimilarities with CPC polarisation curves) are as follows. 9 In CLPC the apparent difference of potentials A(p, is less than the true difference of potentials %, measured in CPC. This calls for accurate measurements in CLPC. 9 There are two simultaneous polarisation zones of the same ore body for the CLPC method, cathodic and anodic zones, and it is not possible to investigate cathodic and anodic processes separately. The CLPC field techniques must decrease the reciprocal influence of these two zones to the greatest degree. To satisfy this condition the direction of the profile in the CLPC method is arranged along the strike of the ore body (along its maximum dimension). 9 In the CLPC method two types of polarisation curves CA and AC, are recorded, corresponding to the opposite directions of current. In order to increase productivity it is desirable to record simultaneously all polarisation curves at positions Mi, or at
72
O.F. Putikov and B. Wen
Ac Is A
I,A
I,A
o~2
0
A
ZSZ 0.2
-0,5 r
0 ~
-0,5 ~0,V
,l'x%lf mV
sialce'l'l
1:2
1'3
14
1'5
16
B
1~/
18
IA%l,mV
~301 50
. . ~~l~ro ,. . . ~. .. 100 200 300 Y'~,7"
Fig. 2-48. Example of CLPC data interpretation: (A) anodic-cathodic (AC) and cathodic-anodic (CA) polarisation curves for stake 13 and stake 15; (B) curves of ]Aq~a[ along profile for the second cathodic process at the southern part (continuous line) and at the northern part (dotted line) of ore body; (C) interpretation results of the experimental curves I Aq)a ] by means of the theoretical curves for h/r0= 1 (reproduced with permission from Ryss, 1983).
least most of them. This requires the use of a multi-channel device such as the Russian 12-channel SPK instrument. For interpretation of CLPC data two relationships are used: (1) the dependence of the apparent differences of potentials of two sequential reactions Aq~a at a distance 'y' along a profile, Aq~a = f(y); and (2) polarisation curves. Using these it is possible to determine the size of an ore body projection to the surface measuring profile, the mineral composition of an ore body and reserves of minerals. Ryss (1983) describes an example of CLPC data interpretation on a section at Rudny Altay (Fig. 2-48). The end faces of the horizontal ore bodies correspond to extremes of Aq)a(y) (Fig. 2-48B). The group of ore bodies is considered as a horizontal cylinder with radius r0, depth of axis h. Also we
Geoelectrochemistry and stream dispersion
73
consider that the end faces of this cylinder carry a double electrical layer with a potential difference equal to the difference of the potential of two sequential electrochemical reactions Aq~. Then from the coincidence of the experimental curve [Aq~a] =f( [ Y-Yext1) with the theoretical curve (Fig. 2-48C) it is possible to estimate values of r0, h and Aq). For the second cathodic reaction at the southern part of the ore body we obtain h = r0 = 320 m, Aq~ = -170 mV. For the same electrochemical process at the northern part of the ore body, h = r0 = 220 m, Aq) = -160 mV. The estimated values of h and r0 are somewhat greater than the real values, probably because of the complicated shape of the ore bodies. For an ore body that contains pyrite, the potential of the second cathodic process can be calculated. For this it is necessary to add the potential of the first cathodic reaction of pyrite, q~ =-0.50 V, and the potential difference between the second and first cathodic reactions, Aq~ = -0.16 V. Thus q)2 ---- q ) l + m q ) - - -0.66 V. This value of potential corresponds to the first cathodic reaction of chalcopyrite (-0.60 + 0.10 V), which is actually present in the ore body (Ryss, 1983). Application of the CLPC method is effective at the detailed prospecting stage for checking geophysical and geochemical anomalies and for locating mineral-enriched zones. The accuracy of determination of mineral composition, concentration and reserves of ore bodies by means of the CLPC method is much lower than in the case of the CPC method.
Polarographic logging (PL) Polarographic logging belongs to a group of non-linear polarisation geoelectrochemical methods that are based on the acquisition and interpretation of voltammograms (in the case of PL, polarograms). These describe the non-linear dependence of the current on voltage between two special electrodes immersed in the medium under investigation (Heyrovsky and Kuta, 1965; Ryss, 1973). For obtaining borehole water polarograms, a dipping sonde is used (Putikov, 1977). The sonde consists of a mercury-dropping working electrode (WE), an auxiliary lead electrode and a mercury container which prevents mercury from escaping and causing pollution. In contrast to laboratory polarography (Heyrovsky and Kuta, 1965), the PL method does not require an additional supporting neutral electrolyte of high concentration. The PL sonde accomplishes in situ qualitative and quantitative analysis of water in boreholes, lakes and seas up to depths of 2 km. Two modes of polarographic logging have been developed: direct current polarographic logging (DCPL) (Uvarov, 1981); and pulse polarographic logging (PPL) (Uvarov, 1984). The potential of the mercury-dropping electrode is a linear function of time in the DCPL mode. In the PPL mode an additional pulse is introduced by means of a synchronisation system which enhances sensitivity approximately ten-fold in comparison with the DCPL mode.
O.F. Putikov and B. Wen
74
I~
t'\
"-'
j:~ "-[ 9
0%
~ ,, ,3/;
1
ImIx
/
:
///-" \,
...'/
%-1
% -2 %,o
o.V Fig. 2-49. Typical shape of natural water cathodic polarograms: TDS- total dissolved solids or salinity in gram/litre (g/l); 1- TDS>5" 2- I
(Pi"
02 + 2H + + 2 e - - + H202
(P2"
H202 + 2H + + 2 e - - + 2 H 2 0
(03:
M e 2+ + 2e- --+ M e ~
(~H 20
"
2H + + 2 e - - + H21"
At the potentials tot and q)2 a two stage reduction of dissolved gaseous 02 takes place. At the potential % the reduction of heavy metal ions Me 2+, such as Fe 2+ or Mn 2§ occurs. The potential q~.:o is the threshold for the electrochemical decomposition of water with the release of gaseous hydrogen. With a decrease of salinity there is an increase in the shift of the apparent potentials q~.~, ..., q).2o of the electrochemical reactions. If the salinity is <5 g/1, the first polarographic wave of oxygen at potential q), is complicated by a polarographic maximum of the first kind, Im,x. On anodic polarograms of chloride water (Fig. 2-50, curve 1), the CI polarographic wave at the potential q), corresponds to the reaction 2C1- --+ C12'I" + 2e-, whilst on anodic
Geoelectrochemistry and stream dispersion
75
1
o~,
i
~,v
Fig. 2-50. Typical shape of natural water anodic polarograms with principal anions: 1- chloride, 2- bicarbonate, 3- sulphate-bicarbonate (reproduced with permission from Putikov, 1993).
polarograms of bicarbonate and sulphate-bicarbonate water (Fig. 2-50, curves 2, 3), there are no polarographic waves. Polarographic recording is performed in either discrete mode (point by point) or continuous mode whilst the logging tool is going down the borehole. During the discrete mode of operation, the tool is stopped at specific depths and complete cathodic and anodic polarograms are recorded, making use of the full analytical capacity of polarographic logging. During the continuous mode of operation, the logging tool is lowered at a constant speed and the limiting current for one specific reaction is recorded; the WE is maintained at a constant potential corresponding to this particular reaction. Thus by continuous polarographic logging we obtain a concentration distribution of one dissolved component as a continuous function of depth. Using the DCPL mode it is possible to determine dissolved gaseous oxygen, 02 (detection limit 0.5 mg/1 ), C1- (>3 mg/l), Fe 2+ (>5 mg/1) and Mn 2§ (>1 mg/1). Using the PPL mode it is possible also to determine Zn 2§ Ni 2§ Cu 2+, Cd 2+, S 2-, UO2 2+, VO2 +, etc. with detection limits of approximately 0.1 mg/l for the majority of metals. The advantages of PL include real-time operation, increased reliability of analysis for volatile (02) and unstable (Fe 2+) components, and high productivity. Its main applications are: 9 9 9 9
hydrochemical investigation of underground water; hydrochemical prospecting for ore deposits; monitoring during underground leaching of ore deposits; analysis of industrial contamination of groundwater, rivers, lakes and seas.
Hydrochemical investigations in deep boreholes and hydrochemical prospecting for ore deposits on the Karelian isthmus, Kola peninsula, and in northern Tajikistan have revealed very low concentrations of ore metals in groundwater. Usually, by means of the
76
O.F. Putikov and B. Wen
/ :
/i"
1
/
, /:
'5
:
// . /
5
0
-1
-2 cp, V
Fig. 2-51. Cathodic logging polarograms, Karelian isthmus, Russia, with depth investigation in metres: 1- 100; 2- 160; 3- 240; 4- 260; 5- 300; 6- 320; 7- 340; 8- 360 (reproduced with permission from Putikov, 1993).
I
~
850m
llmi MnZ+ 30C~200 m
0
0
.............................................................
-0.5
-lt.0
....
I
-1.5
, ,,
q),V Fig. 2-52. Cathodic logging polarograms in Pechenga ore field, Kola peninsula, Russia, with lines (arrowed) showing electrochemical reaction potentials for the corresponding ions (reproduced with permission from Putikov, 1993). DCDL mode, it is possible to locate the anomalous concentrations of Fe 2§ and Mn 2+. But only at low dissolved oxygen concentrations (as a rule, at depths of more than 300 m) is it possible to determine Zn z+, Ni 2+ and other heavy metals (Figs. 2-51, 2-52, 2-53). Use of the more sensitive PPL mode increases the effectiveness of hydrochemical prospecting for deep-seated ore deposits. Laboratory experiments and practical field experience have demonstrated the use of PL in underground leaching of uranium ores. The method has been shown to be effective for all phases of a mining operation, i.e., preparation of a section for exploitation, exploitation itself and subsequent remediation. The main components that need to be determined during preparation for exploitation of a mineral deposit are O2, Fe 2§ C1- and
Geoelectrochemistry and stream dispersion
77
Fig. 2-53. Results obtained by polarographic logging (PL) at the Altyn-Topkan polymetallic deposit, northern Tajikistan: (A) cathodic logging polarograms of the discrete mode of PL; (B) distribution of manganese with depth obtained with the discrete mode of PL; (C) distribution of manganese with depth obtained with the continuous mode of PL; 1- ore interval (galena, sphalerite, pyrite); 2- potential of the polarographic waves (reproduced with permission from Klochkov et al., 1989).
S2-. Using PL it is straightforward to obtain the levels and pattems of the background distribution of these ions in groundwater, including their hydrochemical zonation. Figure 2-54 shows polarograms from such an investigation: polarogram 2, which has two polarographic waves of dissolved gaseous O2, is typical for boreholes crossing the oxygenated part of a deposit; polarograms 1 and 3, on which the corresponding waves are absent, pick out the non-oxygenated zones, and here polarographic waves reveal that Fe 2+ and S 2 are present; polarogram 4 shows that the concentration of C1- is independent of zonality and practically constant at 200-300 mg/1. Indicative polarographic waves on polarograms have been noted in an area where industrial contamination includes sulphuric acid. By means of correlation with chemical analyses of water samples, it was established that these are waves of U022+, V02 +, Fe z+ and Mn 2+ (Fig. 2-55). It is necessary to point out that concentrations of iron and manganese in the industrial effluent are tens to hundreds of times higher than concentrations in natural waters.
78
O.F. Putikov and B. Wen
FeZ+
Q2
0
+1.0
+2.0
q),V Fig. 2-54. Logging polarograms of natural water at the site of underground leaching of uranium ore: 1-3- cathodic polarograms; 4- anodic polarogram (reproduced with permission from Putikov, 1993).
A
I
I
B
~
Mn2+ V02
V02,~
f 0
.- I
-0.5
I
-1.0
q~,V
....
~
0
!
-0.5
-1.0
.
I
-1.5
q~,V
Fig. 2-55. Dependence of shape of cathodic PL polarogram on depth at site of underground leaching of uranium ore: (A) borehole 7 (sensitivityl0 6 A) at 1- 90 m, 2- 95 m, 3- 103 m; (B) borehole 3H (sensitivity 10.6 A) at 4- 90 m, 5- 98 m, 6- 104 m, 7- 107 m and (sensitivityl0 5 A) at 8- 98 m.
DISCUSSION AND CONCLUSIONS Geoelectrochemistry found wide application in the 1970s and 1980s in the exploration of copper-nickel sulphides, lead-zinc sulphides and polymetallic deposits in the Rudny Altay, Kola Peninsula, Caucasus, Orenburg and Pacific coastal regions of Russia, and in Kazakhstan, Uzbekistan, Tajikistan. By using the CPC method for evaluation of geophysical anomalies, some hundreds of small mineral occurrences were very quickly eliminated and exploration expenses on promising targets were reduced by
Geoelectrochemistry and stream dispersion
79
more than 50%. Some new ore deposits were revealed by the CPC method in Rudny Altay, the Kola Peninsula, the Pacific coast region and in Tajikistan. The CLPC method was tested on the polymetallic deposits of Rudny Altay. The CHIM, MDE, MPF and TMGM methods were used for regional and detailed surveys in regions of exotic cover, where traditional geochemical methods are not effective. By using these methods new discoveries were made of copper deposits (in Kazakhstan, Azerbaijan and the Ural region of Russia), polymetallic deposits (in the Baikal region of Russia and Uzbekistan), gold deposits (in the Russian Far East region, Siberia), tin deposits (in the Pacific coast region of Russia, Khabarovsk Kray) and rare metal deposits (in Byelorussia and the Kola Peninsula). The PL method has been used for monitoring underground leaching of uranium deposits in Uzbekistan, Kazakhstan and Tajikistan, and for groundwater monitoring in the Leningrad district. After the 1980s, geoelectrochemistry began to be used in other countries, initially in Canada and Australia, with participation of Russian specialists and then, in China, USA, India and elsewhere independently. From the successful case histories presented in this chapter, it is evident that geoelectrochemical methods are very effective and economical tools for prospecting and exploration of ore deposits, especially deep-seated ore bodies. With increasing demand for mineral products and the decreasing opportunities to discover new mineral resources at surface, it is timely to make use of the theory and application of geoelectrochemistry. Due to the presence of trace elements in oil and natural gas accumulations and gas condensates, it is possible to use some geoelectrochemical methods (CHIM, MDE, MPF, TMGM) for prognosis and prospecting of oil and gas. Extending the application of geoelectrochemical methods beyond the former USSR into other countries will undoubtedly have similar benefits, such as reducing costs of exploration and increasing exploration productivity. Development of interpretation theory and improvement of methodology in the application of geoelectrochemical methods are two factors that help to solve practical exploration problems. In the near future it is hoped that raised sensitivity and accuracy of MPF, TMGM, CHIM and MDE data will lead to the development of criteria to determine the depth, size and reserve of anomaly sources. The immediate objective for the CPC method is to make investigations on non-equipotential, disseminated ore bodies in host rocks with low porosity and high resistivity. Through the study of geoelectrochemical processes in rocks, variations of or new directions in geoelectrochemical methods may be developed. For example, based on the phenomenon of interaction of elastic waves and electromagnetic fields in rocks, it may be possible to develop seismogeoelectrochemistry. Further research, within the framework of international collaboration, is clearly desirable.
This Page Intentionally Left Blank
Geochemical Remote Sensing of the Subsurface Edited by M. Hale Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
81
Chapter 3
SPONTANEOUS POTENTIALS AND E L E C T R O C H E M I C A L CELLS S.M. HAMILTON
INTRODUCTION Selective leach techniques have become popular in mineral exploration for the treatment of geochemical soil samples. Their popularity stems from the fact that they are considered to extract selectively a particular hydromorphically-transported component of metals in the sample and, as such, show better anomaly-to-background contrasts than do conventional strong acid digestions which dissolve most of the chemical matrix of the soil. A number of case studies have been published involving selective leaching of samples taken over known mineralisation. It is apparent from this work that these techniques have some capability to detect a geochemical response due to mineralisation and other geological features through a significant thickness of rock or overburden. What is less apparent, however, is the transport mechanism that moves elements to surface from the source. Recent work in Quaternary glaciated terrain (Bajc, 1998; Jackson, 1995; Hamilton and McClenaghan, 1998) has shown selective leach anomalies, apparently related to bedrock features, above as much as 45 m of overburden sediments. The young age (<10 Ka) and thickness of these deposits require that, at least in glacial environments, a very fast transport mechanism operates. During the weathering process, elements can disperse from source mineralisation by a variety of chemical processes. For reasons discussed below, electrochemical processes are increasingly thought to be the primary transport mechanism in environments of thick, young, exotic (i.e., transported) overburden. They are also likely to operate in other environments but their dominance as a transport mechanism is less certain. This chapter presents the principles behind electrochemical mas~ transport and discusses the role of natural geoelectrochemical processes in the formation of selective leach and conventional geochemical soil anomalies.
This chapter is published with permission of the Senior Manager, Sedimentary Geoscience Section, Ontario Geological Survey.
82
S.M. Hamilton
GEOCHEMICAL TRANSPORT MECHANISMS Transport processes resulting from the chemical weathering of mineralisation involve the dispersal of elements down gradients of one kind or another, including chemical, temperature, piezometric (both gaseous and aqueous) and electrochemical gradients. These mechanisms have been invoked as components of the numerous transport and dispersion concepts that have been used by researchers to account for soil geochemical anomalies overlying mineralisation. The models used in the past can be grouped broadly into those that rely on: (1) diffusion; (2) advective groundwater transport; (3) gaseous transport; and (4) electrochemical transport. Combinations of these are also possible.
Diffusion Diffusion along concentration gradients has been named as a possible cause for observed soil anomalies over mineralisation (Govett and Chork, 1977; Smee, 1979, 1983). Smee (1979) calculated diffusion rates and total diffusion distances for a number of ions through glaciolacustrine clay for an 8000-year period. The calculations were based on measured parameters or reasonable approximations. None of the ions was calculated to travel more than 10 m, with the exception of H § which would have travelled 28 m. Smee (1979) concluded that diffusion was too slow to account for soil anomalies in Quaternary glacial units thicker than 5 m, whereas anomalies spatially related to mineralisation have been noted overlying thicknesses of up to 45 m of young glacial sediments. Hydrogen ions and other ions are likely to be the product of oxidation of mineralisation, which requires the downward diffusion of oxygen. This would decrease the speed of the process by limiting the formation of ions at the bedrock surface and thereby limiting the development of a concentration gradient between bedrock and ground surface. This further precludes chemical diffusion as a major contributor to the formation of selective leach anomalies in thick, young, exotic overburden. Diffusion becomes a more likely transport mechanism with increasing age and/or decreasing thickness of the overlying material.
Advective groundwater transport Bolviken and Logn (1975) and Smee (1983) include groundwater transport as a possible mechanism for element dispersion from mineralisation. Webber (1975) pointed to the much higher potential migration rate of groundwater as compared with that of diffusion or electrochemical transport and concluded that advective groundwater transport is likely to be the most important dispersal mechanism.
Spontaneous potentials and electrochemical cells
83
Groundwater transport may indeed disperse large quantities of elements but in stratified geological materials it tends to do so laterally. This is particularly true with respect to stratified glacial overburden, including glaciolacustrine varved clays, for which horizontal hydraulic conductivities are typically orders of magnitude higher than those that occur in the vertical direction (Freeze and Cherry, 1979). This results in horizontal groundwater flow being favoured over vertical flow by a similar factor. Selective leach anomalies that have been attributed to bedrock features are most commonly reported as either apical (single peak) or rabbit-ear (twin peak) anomalies lying almost directly over the feature. If groundwater transport was a significant contributor to the mobility of the elements involved, one would expect anomalies occurring down a long dispersal plume in the down-gradient direction of groundwater flow. This has not been reported in surface soils. Furthermore, selective leach anomalies are often noted in surface soils far above the water table. Since groundwater obviously could not play a significant role in the formation of many of these anomalies, it is reasonable to conclude that it plays a relatively minor role in the formation of selective leach anomalies as a whole.
Gaseous transport The measurement of soil gases has been used to identify the presence of mineralisation by a number of workers (e.g., Lovell et al., 1983; McCarthy et al., 1986). Gaseous transport of metals and/or other species has been suggested as a possible mechanism in the development of geochemical soil anomalies (Klusman, 1993; Clark, 1997; Smee, 1998). Case studies in support of gaseous transport are largely cited from arid or semi-arid areas where the water table exists 10s to 100s of metres below ground surface. Large scale and rapid gas transport in these environments is plausible because of the very thick vadose (unsaturated) zone. Gases in the phreatic zone (i.e., below the water table) in the zone of meteoric groundwater (envisioned here as the zone where lithostatic load adds nothing to the fluid pressure) typically exist in dissolved form and therefore their transport occurs largely by diffusion or groundwater advection. Some authors have suggested a gaseous transport mechanism below the water table in which gaseous carriers move metals in a separate gas phase (Clark, 1997). This is extremely unlikely. In the vadose zone, gases can exist in the gaseous phase at partial pressures that are considerably below atmospheric pressure. This is not the case below the water table where the sum of partial pressures of all dissolved gases must exceed the fluid pressure for a separate gas phase to exist. By definition the "water table" in a saturated geological medium is the point at which the fluid pressure is exactly equal to atmospheric pressure (Freeze and Cherry, 1979). Below the water table, the fluid pressure exceeds the vapour pressure and, above it, the vapour pressure exceeds the fluid pressure. In the zone of meteoric groundwater, gases typically exist only as a vapour phase below the water table in the following circumstances.
84
S.M. Hamilton
1.
In extremely sluggish groundwater flow environments, e.g., as methane pockets in shale where the gas is in pressure-equilibrium with surrounding groundwater and where the only dispersal mechanism is diffusion. In environments where the fluid pressure is decreasing, i.e., where groundwater is moving up, such as at discharge zones. In environments where the vapour pressure is increasing or at least being maintained by ongoing processes at a continuously high level, i.e., where gas is being actively generated or being supplied from below.
2. 3.
Scenarios 1 and 2 can be ruled out because, as mentioned, neither diffusion nor groundwater discharge can account for the majority of selective leach anomalies. If scenario 3 were applicable as a transport mechanism for metals and gases below the water table, a large source of gas would be necessary. If it were applicable as a cause of rabbit-ear anomalies in all environments, a large source of gas would be necessary at every mineral deposit that produces such anomalies. It follows, therefore, that the source of gas must be genetically associated with the weathering of the mineral deposit. Finally, if a separate gas phase were present, it would require that gas be generated at mineralisation in high enough concentrations to partition into a vapour phase. If it did not, the rate of diffusion of elements from mineralisation would be in a similar range to that of other dissolved species, which is far too slow to account for many of the anomalies noted in thick, young, glacial terrain. Elevated CO2 has been shown to be coincident with rabbit-ear metal anomalies over mineralisation in arid terrain and will be used as an example. For any gas in equilibrium with water, the concentration of that gas in solution is proportional to total pressure. At a depth of 30 m below the water table and assuming CO2 is the dominant dissolved gas phase, the partial pressure required to exsolve CO2 would be about 4 atmospheres, which is an extremely high concentration (exceeding the pCO2 of soft drinks). Concentrations exceeding 1 atmosphere are very rare in natural meteoric groundwater and usually occur only in thermal springs and spring discharges along seismically active faults (Barnes et al., 1978; Hamilton et al., 1990). Concentrations this high in areas of abundant carbonate would dissolve vast quantities of carbonate, which would re-precipitate when the CO2 degasses from near-surface groundwater. The general lack of such deposits in association with rabbit-ear anomalies over conductors suggests that transport of elements by CO2 or CO2-saturated groundwater cannot be responsible for many of the reported selective leach anomalies. It is difficult to conceive of any naturally-occurring gas other than CO2 that could be called upon to transport metals or other species from sulphide mineralisation as a widespread process. Similarly, other gases would also require high concentrations to exsolve as depth and hydrostatic pressure increase, and are therefore unlikely to be responsible for transport of metals below the water table. Furthermore, in stratified geological environments, gases do not typically move straight up but are trapped by
Spontaneous potentials and electrochemical cells
85
zones of fine-grained sediments. As such, gases below the water table are likely to be dispersed by groundwater and are not expected to form tight, well-centred anomalies immediately over mineralisation nor slightly offset rabbit-ear anomalies, which often show a depletion over mineralisation. Carbon dioxide and elevated carbonate content in soils have been used to account for rabbit-ear anomalies in arid environments but have been attributed to direct oxidation of sulphide (Smee, 1998) in what is presumably an unsaturated environment. These anomalies are discussed further below.
Electrochemical transport Characteristic element dispersion patterns have long been recognised over mineralisation at certain sulphide deposits and other electronic conductors in bedrock. These patterns include apical anomalies overlying mineralisation or rabbit-ear anomalies which flank mineralisation. Anomalous parameters reported include H § organic carbon, electrical conductivity and metals. Many authors (e.g., Govett, 1973, 1976; Govett and Chork, 1977; Bolviken and Logn, 1975; Smee, 1983; Govett et al., 1984; Clark, 1996; Jackson, 1995; Bajc, 1998; Hamilton, 1998; Hamilton and McClenaghan, 1998) have attributed these patterns to electrochemical processes. Although the electrochemical transport models presented in the literature are sometimes conflicting, most are based on the same underlying idea. The conduction of electrons upward, along mineralisation, from deep, more chemically-reducing areas results in anomalous electrochemical gradients in the surrounding country rock and overlying overburden. The movement of mass and charge in the form of ions results in the development of geochemical anomalies in the overlying overburden. Different names have been attached to this process, including geobattery, oxygen concentration cell, natural galvanic corrosion, oxidation cell and natural voltaic cell. The primary appeal of using electrochemical transport to explain selective leach anomalies is that it can account for their development in almost any environment, including thick, young, exotic drift. The rate of movement of charged species in an electrochemical field can far exceed that of chemical diffusion, provided the gradient is high enough. Hamilton (1998) determined theoretical transport times through 30 m of clay along a realistic electrochemical gradient to be less than 2000 years for most species. The presence of clay confining units, groundwater flow or saturated/unsaturated conditions should not impede the movement of charge provided an electrochemical gradient exists through overburden as a whole. All four of the mechanisms described above probably operate to some degree at every site where mineralisation is buried and it is possible that any one of them could dominate in selected environments. However, the first three are precluded as major contributors to transport in thick, water-saturated Quaternary glacial environments. Since selective leach anomalies are now commonly reported in glacial terrain, electrochemical processes are likely to dominate in at least this environment.
86
S.M. Hamilton
VOLTAIC CELLS Since various authors have referred to the electrochemical mechanism occurring around ore bodies as a geobattery or natural voltaic cell, it is appropriate to introduce the concept of voltaic cells. Voltaic cells are man-made electrical circuits in which the impetus for current flow comes directly from the chemical energy of partially-separated reactants within the cell. All batteries are voltaic cells. The basic components of a voltaic cell are a wire, two electrodes and two partiallyseparated solutions. When the electrodes are placed in their respective solutions and the wire is used to connect them, a spontaneous flow of electrons occurs in the wire from one electrode to the other. The impetus for current flow comes from the difference between the oxidation potentials of the electrodes and the solutions, or between the electrodes themselves; or between the two solutions in which the electrodes are immersed. A chemical redox reaction occurs between these separated species such that the oxidation half of the reaction occurs in one solution and the reduction half occurs in the other. The partial separation of the solution can be accomplished by a membrane or a salt bridge, which allows an electrolytic connection but does not allow a general mixing of the two solutions. Within the cell, electrical current moves in the form of free electrons in the wire and as ions in the electrolyte. At the electrodes, ions or neutral species change their oxidation state and either accept electrons from, or discharge electrons into, the electrodes. At the anode, electrons are discharged into the electrodes by conversion of the metallic electrode materials to a positively-charged ion or by conversion of a negatively-charged species to a neutral species, gas or ion with a higher oxidation state. At the cathode electrons are usually discharged into solution by either conversion of a positively-charged ion to a neutral species or gas, or by conversion of a neutral species to an anion with a lower oxidation state. In all voltaic cells there must be reduction of species at the cathode and oxidation at the anode and ions must be discharged, or attenuated, or both, at the electrodes. In order to maintain electrical neutrality, the creation of an ion at an electrode must be accompanied by the simultaneous removal of an oppositely-charged ion at the other electrode. To prevent local charge imbalances this process also requires the net migration of ions away from the electrode at which they are liberated into solution and toward the electrode at which they are attenuated. In general, anions move toward the anode and cations toward the cathode, and therefore charge can be carried by reduced anions toward the anode and oxidised cations toward the cathode (Hamilton, 1998). However, it is actually the movement of negative and positive charge that results in electrical current between the electrodes and, in some cases, reduced cations can move toward and deposit negative charge at the anode, or oxidised anions can move toward and deposit positive charge at the cathode. Examples of this include the movement of Fe 2+ toward an anode where it loses an electron to become Fe 3+ and the movement of Ag(CN)2 toward a cathode where it is reduced to Ag(s). This apparently paradoxical process of cations moving toward an anode and anions
Spontaneous potentials and electrochemical cells r
87
Cu Cathode Cu .......2+ 9
.. ..
..........
. . . . . . . . . .......
... ....
9- . .
. ..................
.............
Semi-permeable
membrane ,\
~Na.
CuCI~
Solution
CI" NoCI
.,_ ..
...
. . . .
Solution
.. .....
', ,.
[
i
F--1 Zn'2"i ! e-
equipotenttal lines
| Zn Anode
1
e-r~ electron flow
Fig. 3-1. An example of a voltaic cell spontaneously generating current in a wire. The impetus for electron movement in the wire comes from the difference in oxidation potential between Zn(s) and Cu2+. The reducing agent, Zn(s), gives up electrons at the anode to become Zn2+. The oxidising agent, Cu2~, acquires electrons at the cathode and plates-out on the copper electrode as Cu(~). The semi-permeable membrane allows ions to move between the two solutions preventing charge imbalances and completing the electrical circuit (from Hamilton, 1998).
toward a cathode arises because, in addition to electrical attractive forces acting upon charged species, diffusive forces also play a role according to the Nemst-Planck flux equation (Bockris and Reddy, 1970, p. 398). As charge-carrying species are consumed there is a strong diffusive gradient toward the electrode that results in their movement. There are many species, both positive and negative, to which these forces of opposite attraction can apply, but iron species are probably the most significant from a geological point of view. Figure 3-1 shows an example of a voltaic cell. A zinc electrode is immersed in a solution of NaCI and a copper electrode in a solution of CuCl2,with a semi-permeable membrane separating the two solutions. If a wire connects the two electrodes, electrons flow spontaneously from the zinc electrode to the copper electrode because Cu 2+ is a stronger oxidising agent than Zn(~). At the copper cathode, Cu 2+ in the solution is reduced to Cu(~) by electrons that are the product of the simultaneous oxidation of Zn(s) to Zn 2+ at the zinc anode. The difference in oxidation potential of the two metals results in a differential of approximately 1.10 volts between the two electrodes (assuming equal concentrations of Cu 2+ and Zn2+). Across the membrane, C1~ ions must move toward or
88
S.M. Hamilton
Na + ions must move away from the anode to redress the loss of negative charge from solution around it and the gain in negative charge around the cathode. In this particular cell, there is a net increase in ionic strength around the anode and a decrease around the cathode. The redox reaction would still occur without the membrane but, in the absence of a physical separation of the solutions, Cu~s) would plate-out on the surface of the zinc electrode and no current would flow through the wire. An electrolytic cell is similar to a voltaic cell except the electrochemical reactions involved do not occur spontaneously but require the input of current from an external source. Wires connected to each end of a battery and submerged in a suitable electrolyte can represent an electrochemical cell. As with voltaic cells, the creation and/or removal of ions at the electrodes facilitates the transfer of current into and out of solution. If the electrolytes in solution are redox-inert within the stability field of water (e.g., Na + and CI ~ and the voltage is over 1.2 volts, the hydrolysis of water may transfer current at the electrodes: - at the anode 89H20 zz> H + + 1/402(g) + e at the cathode H + + e ~ 89H2(g) -
Hydrogen ions are created at the anode and removed at the cathode and during the process oxygen and hydrogen gases are produced at the respective electrodes. If the voltage of the cell is significantly below 1.2 volts then water cannot be hydrolysed. If the electrodes are inert and there are no redox-active electrolytes (i.e., those that cannot change their oxidation state and thereby accept or lose electrons at the electrodes) in solution, then virtually no current will flow in the cell. Many voltaic cells can act as electrolytic cells if a power source is used to reverse the direction of spontaneous electron flow. A lead storage battery, for instance, is a voltaic cell when discharging and an electrolytic cell when being recharged. The two types of cell also differ in a number of other ways. The most fundamental difference is that in a voltaic cell conditions are more chemically reducing around the anode than around the cathode whereas in the electrolytic cell the situation is reversed (Fig. 3-2). Electrons are "pulled" into solution at the cathode of a voltaic cell by the oxidising agents in solution whereas they are "pushed" in at the cathode of an electrolytic cell by an external power source. Another difference is the nature of their electrical fields within the electrolyte. In the electrolytic cell, the electrical field is applied; if the current is turned off, the field will dissipate. In most voltaic cells, either the oxidants or reductants or both are dissolved species in separate solutions. As such, the potential field between the two electrodes is not applied but is semi-permanent and results from the differences in oxidation potential between species in solution, or between these and the electrode materials. Allowing the flow of current between the two electrodes will modify the shape of equipotential lines in the field but does not create the field. Table 3-I shows the standard electrode potentials of a number of redox half-reactions. Tables of standard voltages can give an approximate idea of the likelihood of redox reactions occurring spontaneously. Standard voltages are a measure of the oxidation
Spontaneous potentials and electrochemical cells
89
Fig. 3-2. Differences in oxidation potential around the electrodes in an electrolytic cell (requires external power) and a voltaic cell (spontaneous) (from Hamilton, 1998).
potential of half reactions measured against that of the H+-H2 half cell. Any reducing agent in the centre column is capable of spontaneously reducing any oxidising agent in the left-hand column that is lower in the table. Standard voltages are usually calculated for 25~ and assume molar concentrations of all species in solution and partial pressures of 1 atmosphere of all gases involved in the reaction. Concentration changes of up to several orders of magnitude and temperature changes up to several 10s of degrees Celsius have relatively small effects on the voltages in a given reaction. The effect of concentration can be calculated using the Nernst equation. For the reaction: aA + b B ~ cC + dD the Nernst equation takes the following form (at 25~ E = E~
{0.0591/n} 9 log,0{([C] c , [D]d)/([A]a,[B]b)}
where E = reaction voltage; E ~ standard electrode potential of the reaction; and n = number of electrons involved in the redox reaction. The Nernst equation demonstrates that a change in concentration of a species involved in the reaction does not change the final voltage if the concentrations of all species change by a similar factor. For example, a hundred-fold dilution of the species involved in the Cu-Zn reaction to 0.01 molar still produces 1.10 volts because the ratio of [ZnZ+]/[Cu2+] is the same. It takes very large changes in the ratio to change voltage significantly and therefore it is often acceptable to use standard voltages to qualitatively predict the spontaneity of a reaction. However, a problem arises with the application of
90
S.M. Hamilton
this principle in the case of natural reactions that involve either H + or O H . In nature, a molar concentration of either H + or OH- is exceptionally high relative to a molar concentration o f many of the other species that they are likely to react with. As such, the standard conditions used in the determination o f standard electrode potentials result in an exaggeration of the potential concentrations o f H + and O H relative to m a n y other naturally-occurring reactants.
TABLE 3-I Standard potentials for a number of half reactions in aqueous solutions at 25~
EOred(V)
Oxidising agent
Reducing agent
Al3+(aq) + 3 e Zn2+(aq) + 2 e Fe2+(aq) + 2 e pb2+(aq) + 2 e 2 H+(aq) + 2 e S(s) + 2 H+(aq) + 2 eCu2+(aq) + e SO42(aq) + H+(aq) + 8 eSO42(aq) + 4 H+(aq) + 2 e SO42(aq) + 4 H+(aq) + 2 e Cu2+(aq) + e
~ :::> :::> :z> ::> ::> :::> :z> ~ ~ ::z,
Al(s) Zn(s) Fe(s) Pb(s) Hz(g) H2S(aq) Cu+(aq) S 2- + 4 H20 H2SO3 SO2(g) + 2 H20 Cu(s)
-1.66 -0.76 -0.44 -0.13 0.00 0.14 0.15 0.16 0.17 0.20 0.34
CIO4(aq) + H20 + 2 e H2SO3(aq) + 4 H+(aq) + 4 eCIO3(aq) + 3 H20 + 6 e Fe3~(aq) + e-
=:, ~ ::::, ::::,
CIO3(aq) + 2 0 H ( a q ) S + 3 H20 Cl(aq) + 6 0 H ( a q ) Fe3+(aq)
0.36 0.45 0.62 0.77
Ag+(aq) + e CIO-(aq) + H20 + 2 e O2(g) + 4 H+(aq) + 4 e Cl2(g) + 2 eCIO3-(aq) + 4 H+(aq) + 4 eAu3+(aq) + 3 e
::, :::::, ~ ::> :::, :::,
Ag(s) Cl(aq) + 2 0 H ( a q ) 2 H20 2 Cl(aq) 89Cl2(g) + 4 H20 Au(s)
0.80 0.89 1.23 1.36 1.47 1.50
The results of this are particularly apparent when considering the half-reactions involving the oxidation of C I to more oxidised forms such as C I O , C103 and C104. Since these substances are above the 02 - H20 half reaction (Table 3-I), it appears that they are less oxidising than oxygen. This suggests that dissolved oxygen is capable o f spontaneously oxidising C I and producing H C I O , CIO3 and ultimately C104. In fact, this is not the case within the chemical realm of natural waters. The only naturally-
Spontaneous potentials and electrochemical cells
91
occurring form of chlorine within the Eh-pH regime of shallow terrestrial waters is CI (Stumm and Morgan, 1970, p. 320). The confusion results from the over statement of the concentrations of reactants in this equation which is inherent with standard electrode potentials. In this case, the concentration of O H is overstated by at least three orders of magnitude, since the upper limit of pH in terrestrial waters is about 11 (Bass Becking et al., 1960). The pO2 is overstated by a factor of five, since the maximum pO2 in the groundwater environment is the same as the mole ratio of oxygen in the atmosphere, which is about 0.2. Notwithstanding this, provided the user is aware of these problems, Table 3-1 can be useful in the estimation of the approximate voltages of natural reactions. In addition, it can be used to show the range of reactions that could occur in nature and that might be responsible for spontaneous potential currents in the Earth. In the terrestrial environment, materials such as C12 or Fe(s), which are capable of oxidising or reducing water to 02 or H2 respectively, rarely occur in large quantities. Therefore, reducing agents that occur in half-reactions above the H+-H2 reaction in Table 3-I and oxidising agents that occur below the O2-H20 reactions are not likely to participate in natural voltaic cells. In the preceding and following discussions, the effect of kinetics and rate-limiting factors on redox reactions are largely ignored. Electrode potentials have been treated as the sole factor that will determine whether one reaction is favoured over another. In reality, there are many processes that affect rates of reaction, such as diffusion of species across phase boundaries, high resistance to current flow between electrodes and solution, high activation energies and slow intermediate reactions. Such processes may result in one reaction being favoured over another that has a much higher reaction potential, or in negligible reaction rates for reactions that might otherwise occur according to their electrode potentials. In the Earth, it is more justifiable to make extrapolations from thermodynamic data than would be the case in a sterile laboratory setting because of the ubiquitous occurrence of biological catalysis. In nature, thermodynamic reactions that can occur typically do occur because organisms have developed systems for "piggybacking" on these reactions to obtain metabolic energy from them. Examples abound of biologically-mediated reactions in nature that have reaction rates that are orders of magnitude faster than they would be in a sterile laboratory setting. As a result, in a very general sense, kinetic inhibitors to thermodynamic reactions tend to be minimised in natural environments relative to the laboratory.
SPONTANEOUS POTENTIAL IN EARTH MATERIALS The existence of spontaneous potentials in the Earth has been known for at least 150 years and their measurement has been used systematically in the search for buried ore deposits since the 1920s (Parasnis, 1979). Spontaneous or "self' potentials (SP) are natural voltage differences between two points in the Earth which result in electrical
92
S.M. Hamilton
currents in Earth materials. They arise largely because of differences in the oxidationreduction (redox) potential of Earth materials. Spontaneous potentials are useful in mineral exploration because SP anomalies are often associated with electronically-conductive mineralisation. The vast majority of these anomalies are negative over mineralisation relative to surrounding terrain, which suggests that mineralisation acts as a source or conduit of electrons (Sato and Mooney, 1960). Burr (1982) reports that sulphide mineralisation in Canada produces negative anomalies of up to 350 mV, whilst anomalies of over 450 mV are usually due to the presence of graphite, which is a far better conductor of electrons. These ranges are in contrast to typical background variations of only a few millivolts to a few tens of millivolts where mineralisation is absent (Parasnis, 1973). In thick overburden the contrast between anomalies and background readings decreases to the extent that 100 m or more of overburden can render the SP response due to bedrock features too weak to interpret (Lang, 1956).
Measurement o f spontaneous potential Theoretically, SP can be measured in two ways. One way involves the use of an oxidation-reduction potential (ORP) meter on samples of Earth materials taken from two areas; the measurement should provide voltage differences that approximate the spontaneous potential between the two areas. An ORP probe, in contact with a moist sample, represents a reversible voltaic cell. The probe is connected to a millivolt meter that can measure the voltage difference between the sample and a half-cell such as AgAgCI, which is usually located inside the probe. An inert metal, usually platinum, on the outside of the probe serves as an electrode in direct contact with the sample. A semipermeable junction of porous glass or ceramic maintains electrolytic contact between the sample and the half-cell. If the sample is more reducing than the Ag-AgCI half-cell (standard electrode potential = 222 mV) the platinum wire behaves as an anode and accepts electrons from the sample. Inside the probe, electrons are simultaneously provided to Ag + to form Ag(s) on the Ag electrode. If the solution is more oxidising than the Ag-AgCI half-cell, the platinum behaves as a cathode and the reverse reaction occurs inside the probe. Thus, the ORP readings on the millivolt meter represent a relative voltage difference between the half-cell and the sample. If desired, the Eh of the solution can be obtained from these results by adding to the readings the voltage difference between the reference cell and the standard hydrogen electrode. In practice, ORP probes have inherent limitations that often render this approach to SP measurement unworkable. Because platinum is effectively inert, the gain or loss of electrons on the electrode surface necessitates the attenuation of ions from the sample by their conversion to neutral species or to charged species of a higher or lower oxidation state. For the readings to be reproducible, one or more reversible redox-couples, such as Fe 2§ / Fe 3+, H S / $ 0 4 2 or Cu 2+ / Cu +, must be present in solution near the electrode.
Spontaneous potentials and electrochemical cells
93
Since not all redox-active solutions contain such reversible redox couples (Drever, 1982), ORP probes may not always provide reproducible or accurate results (Bartlett and James, 1995). A related problem occurs due to the build-up of deposits on the platinum during these reactions, leading to memory effects in the probe when testing other solutions. These cause long-term drift in the readings, particularly when testing a sample of low ionic strength after one with high ionic strength. Furthermore, many natural redox couples have slow rates of reaction, which result in slow response and instrument drift during ORP measurement. Consequently, ORP probes are often used to obtain a qualitative idea of the redox composition of waters or sediments but are seldom used to obtain quantitative SP measurements between two areas. By far the most common way to measure SP is to use a millivolt meter connected between two non-polarising electrodes placed at different points on the ground. A measured voltage difference would usually equal the spontaneous potential between the two points. Non-polarising electrodes are important because they must be equally capable of receiving and discharging electrons. The type that is almost invariably used for SP surveys is the Cu-CuSO4 electrode. It consists of a copper rod in contact with a saturated solution of CuSO4 that is in electrolytic contact with the ground. The latter contact is accomplished through the use of a wet membrane of asbestos or porous ceramic. If the ground at point "A" is more reducing than at point "B" (Fig. 3-3), it will have the capability of providing negative charge. Across the membrane of the electrode at "A", anions move up from the ground and/or cations down. Inside the electrode, electrons are provided to the copper by oxidation of Cuts) to Cu2+(aq). At point "B", which is relatively oxidising, there will be a tendency for negative charge to discharge into the
e-
k
cu
i
c
l
CuSO4~q~
CuSO4~,~
~)1
........ , .............................
(-)
A
(.4-) I~ ,ql--(-)
B
Fig. 3-3. Non-polarising Cu-CuSO4 electrodes for measurement of spontaneous potentials in surface materials. If the ground at "A" is more reducing than the ground at "B", a potential difference exists between the two electrodes and current in the wire is possible.
94
S.M. Hamilton
ground and positive charge to discharge into the electrode by the conversion of Cu2+(aq) to Cu(s) at the surface of the copper, and a corresponding migration of cations up through the membrane and anions down through the membrane. There is, therefore, an electrical potential difference between the two areas which is measurable with the millivolt meter and which would result in current in the wire if a direct connection were made. An important interference often encountered during the measurement of SP using Cu-CuSO4 electrodes occurs due to variable moisture conditions. These can cause false anomalies which can seriously complicate the interpretation of survey data, especially in areas of thick overburden. Wet areas have often been observed to cause high positive readings relative to adjacent areas (e.g., Parasnis, 1979; Burr, 1982). This is perhaps counterintuitive because it appears to imply that wet soils are more oxidising than unsaturated soils. Burr (1982) attributed higher SP readings in swamps and wet soil to pH effects. The Eh, and therefore SP, have a general pH dependency because many natural redox reactions involve hydrolysis. For example, for the half-reactions that involve either the oxidation or reduction of water, a decrease of 1 pH unit will result in an increase in the Eh of the reaction of approximately 60 mV. Therefore, the higher Eh of peat and moist humus layers is consistent with the fact that these materials are usually more acidic than are mineral soils. Lower electrode-ground resistance has also been noted in moist areas and in humus soil layers relative to underlying mineral soils, and has been suggested as a possible contributor to moisture-related anomalies (R. Chaplain, pers. comm., 1998). Streaming potentials as a result of groundwater discharge in lowlying areas has also been suggested as a possible source of SP (Dobrin and Savit, 1988) and therefore could be a source of false anomalies in low-lying areas. Streaming potentials are generated by the movement of water through a porous medium that is capable of ion exchange, such as clay or oxides on sand. Other factors that can affect SP surveys tend to be less significant than the problems due to moisture. Magnetic storms (Burr, 1982), radar and other electromagnetic radiation can cause induction in the long SP wire, particularly when it is fully extended. Telluric currents, which are global-scale electrical currents in the Earth induced by the Earth's magnetic field, could conceivably affect SP surveys but typically result in a SP difference of only a few millivolts per kilometre. The use of SP surveys as an exploration tool has waned since the 1950s with the increasing sophistication of other electrical geophysical techniques such as induced polarisation (IP) and ground resistivity. Part of the reason is that the interpretation of these non-passive techniques is easier because electrical theory and electronics theory can be applied. Since the causes of natural SP above mineralisation are still widely misunderstood (Hamilton, 1998), the interpretation of the results of SP surveys is difficult.
Spontaneous potentials and electrochemical cells
95
Sources of spontaneous potential All moist Earth materials contain redox-active species, such as O2(aq), Fe 2+, HS, OHand H +, which impart a bulk redox capacity, or Eh if measured against the H+-H2 halfcell, to the soil or groundwater. The variable chemical composition of overburden, groundwater and rock therefore results in variable redox capacity which, in turn, results in spontaneous potential voltages and currents between different points in the Earth. The multitude of potential processes that affect the composition of Earth materials and might thereby affect redox processes can be divided into primary lithological processes and surficial processes. Primary lithological processes are defined as all those that contribute to the variable composition of rock. They result in specific lithologies and mineral accumulations in various places in the upper crust. The mineral assemblages in these materials can impart a redox potential to the groundwater/rock matrix with which they are in contact and can result in a characteristic redox signature for various rock types and mineral accumulations. For example, the presence of large amounts of gypsum in carbonate rock can fix the equilibrium Eh of the groundwater/rock environment to no lower than approximately -275 mV, which is the Eh of the S O 4 2 " - H S - half-cell (at pH 8 and using certain other assumptions; Garrels and Christ, 1965, p. 215). In this environment, species that can impart a lower Eh are unlikely to be present because most of the reducing agents capable of reducing 8042 to HS-would already have been consumed. However, the Eh can be greater than -275 mV because SO42- is the only geologically-important sulphur species in oxidised environments (Krauskopf, 1979), where its redox behaviour is relatively inert (Bartlett and James, 1995), and therefore it adds little to the redoxbuffering of oxidised systems. Processes such as these can occur due to mineral assemblages or dissolved species in many other rock types. In unweathered pyritic rocks, Eh is typically maintained below the sulphide oxidation half-cell. Ferrous olivines and clinopyroxenes in ultramafic rocks undergoing weathering can produce very reducing conditions that approach the limits of water stability (around -400 mV at pH 11; Barnes et al., 1978). The presence of dissolved oxygen in oxygenated terrestrial waters typically maintains Eh at a fairly high empirically-observed level of over 200 mV. The empirical limit is used because there is no perfect correlation between dissolved oxygen concentration and Eh, probably due to the myriad biological and inorganic processes that involve oxygen. There is, however, a general relationship between the presence of dissolved oxygen and high Eh to the extent that most oxygenated terrestrial waters are found to have an Eh of between 200 and 400 mV. The upper limit of geologically observed Eh is around 800 mV and the theoretical electrical potential of oxygen is higher still at 1000 mV at pH 4.0 (Fig. 3-4). The presence of water itself restricts the Eh of aqueous systems to a well-defined stability field. Figure 3-4 shows the theoretical and empirical stability fields for water in natural environments. As shown in Table 3-I, reducing agents that are more reducing than H2(g) rarely exist in the shallow terrestrial environment because their oxidation by
96
S.M. Hamilton
1.2 x ~ O ~ . _ _ Theoretical upper limit of water stability !.(i
JiO -- r
"r ~/..,.,
9 I
0.8 : ; , " 0.6
I
0.4
II
(1.2
~
~
-0.4 -0.6
x I ~ Extreme empirical "x,,,.Ld"" limits tbr terrestrial I ~ ters .. as,..--
i
,
,,0
-0.2
..0%
'
II2N "~
., , / ,
" O.~ ~
Jncorctical lower limit of water stability
2
4
-,ojt ,~
6
8
i0
12
13
pl!
Fig. 3-4. Theoretical and empirical stability fields for water (reproduced with permission from Bass Becking et al., 1960, Journal of Geology, v.68, copyright by the University of Chicago Press).
abundant water would have occurred long ago. However, as crustal thickness increases, water and other volatile fluids are excluded from the geological environment and this allows more reduced Eh conditions. The mineral assemblages of rocks formed in these environments often reflect their low-Eh origins. The rare production of H2(g,s) due to the reduction of water by minerals has been noted in groundwater interacting with ophiolite sequences (Barnes et al., 1978; Clark, 1987) and demonstrates the very reducing nature of some rocks that form in water-poor environments. Groundwater from kimberlites (author's unpublished data) shows Eh and pH conditions that also border the lower limits of water stability. This process of mineral-water reactions fixing the Eh of groundwater is loosely referred to as redox buffering (Drever, 1982). However, the slow rates of reaction of many redox processes rarely result in mineral-water solutions that approach chemical equilibrium, as do most pH buffering reactions. Consequently, most natural redox processes are in a state of disequilibrium and therefore one can only generalise about the outcome of most redox-buffering processes. Non-equilibrium kinetics play a major role in almost all natural redox processes, especially those involving oxygen, the most geologically-important oxidising agent. Surficial processes that affect the redox composition of Earth materials include weathering, drainage, groundwater movement, mechanical mixing and dispersion of rock material, soil formation, the accumulation of organic material and biological processes. There is an almost unlimited number of ways in which these factors can combine to affect the composition of Earth materials and therefore to affect local redox conditions. However, the processes that are most likely to affect redox locally can be simplified.
Spontaneous potentials and electrochemical cells
97
The principal oxidising agent on Earth is free oxygen (02(g)) , for which the only significant terrestrial source is plant photosynthesis. The only appreciable source of oxygen for the geological subsurface is the atmosphere, which contains 21% oxygen. Since redox variability in shallow Earth materials cannot be due to variations in the primary source of oxidising agents, it must be due to: (1) kinetic processes that limit the transfer of oxygen into the subsurface; and/or (2) processes that control the consumption of oxygen (i.e., that control the availability of reducing agents). The primary limitations on the transfer of oxygen into the subsurface are its solubility in water and its slow rate of aqueous diffusion. In the fully-saturated groundwater environment below the water table, the concentration of oxygen has an upper limit from which it can only decrease. Its solubility limits the concentration to a maximum of about 10 ppm at 25~ In contrast, the maximum concentration of gaseous oxygen in the vadose zone is limited only by its molar ratio in the atmosphere and can therefore reach 210,000 ppm. This enormous disparity demonstrates why moisture content is the single most important factor controlling the availability of oxygen in a geological environment (it also demonstrates why subaqueous disposal of sulphidic mine tailings is so effective in preventing their oxidation). In many subsurface environments, the water table represents a sharp redox boundary between abundant oxidising agents above and abundant reducing agents below. This is particularly true in young exotic overburden. Most of the other surficial factors controlling local redox conditions involve processes that control the availability of reducing agents. Chief amongst these is the accumulation of organic matter. All forms of organic matter are reducing relative to oxygen and most can be oxidised fairly quickly in Earth materials by microorganisms. As such, the oxidation of organic matter is one of the primary consumers of oxygen in the shallow subsurface. Typically, higher concentrations of organic matter in soils lead to more reducing conditions, particularly in saturated environments. Also important is the lability (i.e., availability to microorganisms as a source of metabolic energy) of the organic matter. Well-humified peat is not as chemically reducing as methanogenic organic muck in a perpetually submerged portion of a bog because the former has already undergone oxidation of its most labile organic components. The other major consumer of oxygen in the shallow subsurface is the oxidation or weathering of mineral matter and particularly of metallic sulphides. Areas of unusually active weathering, due either to large accumulations or to more reactive minerals, typically result in an enhanced consumption of oxidising agents relative to surrounding areas and to more negative redox conditions. The dispersal of the dissolved products of weathering, such as Fe z+, can affect redox conditions at some distance from the source. All forms of mechanical dispersion of rock material can affect the availability of reducing agents. Continental glaciation results in the widespread deposition of relatively unoxidised rock, till, clay and other drift materials over vast areas. Different deposit types (e.g., sand, clay) resulting in different permeabilities in transported materials can also lead to higher or lower water tables and variable rates of percolation of oxygenated groundwater. This can result in a poorer availability of oxidising agents in fine-grained
98
S.M. Hamilton
deposits as compared to coarse-grained material. Generally, in older terrain with deep weathering profiles, such as laterite, there is a greater availability of oxidising agents in shallow areas because most reducing agents have already been consumed. Finally, both temperature and pH variations can also affect redox reactions. Significant horizontal temperature variations in the Earth are rare over short distances but vertical temperature gradients are ubiquitous. However, the effect of temperature on Eh is fairly small and, since vertical geothermal gradients are locally quite uniform, their effects on redox reactions will not be considered here. On the other hand, pH does vary significantly over short distances. Redox reactions that involve either H + or O H in either the reactants or products are affected by pH, and this is the case in many, though not all, natural redox reactions. Surficial and bedrock processes control the local balance of oxidising and reducing agents in the shallow subsurface. The result, in young surficial environments, is an electrochemically inhomogeneous shallow subsurface with local redox gradients almost everywhere. These represent fields of electrical potential (SPs) between the local sources of oxidising and reducing agents along which ions have a tendency to move. This movement of redox-active ions is consistent with the universal tendency of chemical systems to approach maximum entropy. In the long term, and therefore in older deposits, it results in increasing local homogenisation of redox conditions in the shallow subsurface and a tendency for local conditions to approach the larger redox trend that overprints all redox processes, i.e., the redox stratification of the Earth's crust.
Redox stratification in the Earth An upward increasing redox stratification exists in the Earth's crust (Bass Becking et al., 1960; Bolviken and Logn, 1975). This redox field results from the process of oxygen re-supply by the atmosphere over-riding the general tendency toward redox homogeneity (maximum entropy). It establishes an overall vertical gradient between the oxygenated surface and mineralogical reducing agents deep in bedrock. Subject to the limitations described, the upper limit of this Eh field is fixed by the electrical potential of oxygen at the lowest geologically reasonable pH (around +1000 mV; Fig. 3-4). The lower limit is usually considered to be the lower limit of water stability at the highest geologically reasonable pH (around-400 mV; Fig. 3-4). Indeed, rocks from the lower crust and upper mantle appear to have formed in Eh environments that are at or slightly below the Eh stability field of water. This 1400 mV spread represents the maximum potential redox differences to be expected due to naturally-occurring redox-active substances in the Earth's crust. It is consistent with the vast majority of spontaneous potential measurements, which are below 1500 mV (Sato and Mooney, 1960). Almost any natural redox-active substance that can exist in the Eh stability field of water could potentially contribute to this redox gradient. Natural terrestrial materials more oxidising than oxygen are virtually non-existent and consequently natural redox
Spontaneous potentials and electrochemical cells
99
reactions that oxidise oxygen in water to 02 do not occur. Oxidising agents are therefore likely to be restricted to oxygen and electrochemically-weaker oxidative species such as Fe 3+, Mn 4+ and SO42. Geological materials more reducing than water do exist but the natural reduction of hydrogen in water is rare in the zone of meteoric groundwater and occurs only under exceptional circumstances (e.g., Barnes et al., 1978; Clark, 1987). As a result, reducing agents are likely to be less reducing than H2(g) and could include reduced aqueous sulphur species (e.g., HS), reduced sulphide minerals (e.g., pyrrhotite), mafic or ultramafic minerals (e.g., ferrous olivines and pyroxenes) or their dissolved products, and hydrogenous organic matter. An electrochemical gradient such as occurs in the Earth's crust represents a field of electrical potential in an electrolyte. This gradient induces the movement of ions (Fig. 35) and results in an electrolytic charge transfer between deep and shallow areas (Bolviken and Logn, 1975; Hamilton, 1998). Negative charge-carrying redox-active ions tend to move upward toward more oxidising Eh conditions and positive charge-carrying ions tend to move downward toward a more reducing environment. The ion migration is analogous to movement of charge-carrying ions toward the electrodes of a voltaic cell. In order for charge transfer to occur, ion movement must be accompanied by redox reactions that attenuate some or all of the migrated species. All of the charge-carrying species have a particular Eh range within which they are stable in groundwater, and a particular Eh limit beyond which they are likely to become attenuated and thereby pass on charge to other species (Fig. 3-5). As such, the reactions that transfer charge from the migrating ions are likely to occur all down the gradient from deep in the crust to ground surface. The movement of redox-inert species (such as Na § and CI) is also likely to occur in the Earth's redox field, as it does in a voltaic cell, to prevent local charge imbalances. However, this movement is in response to the migration of redox-active species and the resultant redox reactions, and therefore does not cause the transfer of electrical charge but rather results from it (Hamilton, 1998). This electrical current that is inferred to exist between shallow and deep areas in the Earth's crust must be subtle but ubiquitous. The upward movement of negative charge is a kinetic process and counteracts, to some extent, the continuous supply of oxidising agents to the shallow subsurface from the atmosphere. However, deep weathering profiles in arid environments, in which the majority of mineralogical reducing agents have been consumed, suggests that this process is not static but favours the long-term consumption of mineralogical reducing agents.
1O0
S.M. Hamilton lz;arth'S' s u r t a c e
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
"." (mV) . . . . .
4-600
.
u,I O
i I 7'
'
+400
iii iii ; iiiiiiiiiiii !iiii)ill iiiiiiill iiiii. -400 Earth's Lower Crust _
_
.
.
Increasing Eh H < d (t~< c 0~< b 9 < a (-~
SPONTANEOUS POTENTIAL CELLS
Ohm's Law in the development o f cells Current density at low voltages in an electrical m e d i u m is governed by O h m ' s Law: E=r,j where E is electrical field strength, or voltage gradient, in V/m; r is resistivity, or electrical resistance per unit length o f medium, o h m e m ; and j is current density, or electrical current per unit cross-sectional area, in amp/m 2. This simple equation states that in an electrical field, current density will increase if either the voltage gradient or the electrical conductivity (i.e., the inverse of resistivity) increases within the medium. The term cell, when used in a laboratory electrochemical context, refers to a system through which electrical current passes and which has two connectable electrodes
Spontaneous potentials and electrochemical cells
101
immersed in an electrolytic conductor. In a voltaic cell, electronic current is spontaneous and therefore to induce current through the wire joining the electrodes there must be physical separation of oxidising and reducing agents in the solution. The redox field in the Earth represents an electrolyte within which there is vertical separation of oxidising and reducing agents. For the purpose of the following discussion, an SP cell is defined here as a natural system that induces the spontaneous, long-term flow of electrical current in a focused area within the Earth, i.e., the current flux has to be anomalous with respect to background current. Although there must be constant ionic current in the Earth's redox field, cells as just defined can only develop due to major inhomogeneities in either the electrical conductivity of Earth materials or the SP gradient in a particular area. Indeed the development of Earth cells of this type should be predictable by Ohm's law. If an area of increased electrical conductivity occurs across an otherwise uniform voltage gradient, an increase in current density in that zone must also occur. As discussed below, a verticallyoriented geological conductor crossing horizontal redox equipotential lines in the Earth is one example of such a system. Likewise if an area of increased voltage gradient occurs in a medium of uniform electrical conductivity, an increase in current density must also occur. Examples of this type of cell are also discussed below. The term conductor, when used in a geological context, usually refers to electronically-conductive or semi-conductive materials in bedrock, such as graphite or metallic sulphide mineralisation. These substances conduct electricity far better than low-porosity silicate or carbonate bedrock. However, they are typically poorer conductors than most groundwater-saturated overburden. Part of the reason why this fact is not widely recognised among geologists is that the electrically-conductive properties of groundwater and bedrock have typically been expressed in inverse ways. The electrical conductivity of groundwater is most commonly reported in conductivity units whereas the conductive properties of bedrock are reported in resistivity units. Table 3-11 shows the electrical conductivities of some typical groundwaters and those of a number of common bedrock materials converted from their usual resistivity units. It is evident that very few common bedrock conductors begin to approach the electrical conductivity of most groundwaters. Overburden is usually more electrically conductive than the groundwater within it because of the additional conductivity imparted by clay minerals and oxide surfaces (Keller and Frischknecht, 1966). This demonstrates that the electrical conductivity of overburden far exceeds that of low-porosity bedrock and usually also exceeds that of most bedrock conductors.
Cells associated with electronic conductors in bedrock Geobatteries centred on electronically conductive, steeply-dipping mineralisation have been described for many years and most geologists are aware of their existence. However, several widely-referenced geochemical models developed to account for SP
102
S.M. Hamilton
cells in the Earth were based on a fundamental false premise. The problems stem from the fact that most authors (e.g., Bolviken and Logn, 1975; Govett, 1976; Sivenas and Beales, 1982; Smee, 1983; Clark, 1997), inadvertently or otherwise, saw the conductor itself as the voltaic cell. This led to models involving self-polarising conductors that generated their own electrical field within the Earth's redox field (Fig. 3-6). When considering the conductor and the Earth electrolyte as a whole, the result was an electrolytic cell, not a voltaic cell. Hamilton (1998) provides an in-depth treatment o f the theoretical problems with some of these models.
TABLE 3-II Electrical conductivities of typical shallow groundwater in glacial terrain (Shoot Zone gold property, northwestern Ontario) and those of common bedrock materials converted from the usual resistivity units (from Keller and Frischknecht, 1966)
Conductor
n
Electrical conductivity (~tS/cm) Range
Mean
Groundwater Peat Glaciolacustrine Clay Sand Aquifer Glacial Till Crystalline Rock
3 7 23 7 15
Bedrock material" Arsenopyrite Chalcopyrite Native copper Graphite (parallel to cleavage) Galena Magnetite Marcasite Molybdenite Pyrrhotite Pyrite Pyrolusite Sphalerite ("non-conductive")
297357138 254 159-
347 714 558 536 581
328 508 387 417 370
0.33 - 5 0.011 - 0.67 330 - 8300 100 - 280 0.0011 15 1.9 0.00067 - 0.10 0.000013 - 0.00083 0.63 - 50 0.00017 - 0.083 0.0000033 - 0.014 0.0000000083 - 0.037
1.3 0.086 1700 167 0.13 1.9 0.0082 0.00011 5.6 0.0037 0.00022 0.000018
-
* Means for bedrock material are geometric means
There are two models for the development o f SP cells around electronicallyconductive mineralisation that are considered here to be theoretically sound. They are
Spontaneous potentials and electrochemical cells
103
t ..... \, ......... I .........................
-200 Equipontentiallines, mY. --'- (Positive}Current flow Direction of ion movement (c)
Cathode
~
Electronflow
~{J
Sulfide
(4)
Anode
Fig. 3-6. Diagram showing the equipotential line and ionic current flow line distribution around an electronic conductor in bedrock, after the model of Govett (1973) and Bolviken and Logn (1975). The conductor is shown to be sel[-polafising and to generate an electrical field. As such, this current and equipotential line configuration represents that of"an electrolytic cell, not a voltaic cell (from Govett et al., 1976).
the reactive groundwater model and the reactive conductor model. Each was developed separately for different geological environments and in neither case do the authors address both sets of processes when developing their models. However, they are in no way mutually exclusive and probably represent end-members of the types of cells that can occur around conductive ore bodies. Sato and Mooney (1960) proposed a reactive groundwater model for the development of SP cells that, in effect, put to rest several decades of debate on the origins of negative SP centres over conductive mineralisation. They carried out a critical review of existing hypotheses and used geophysical theory to establish a mechanism that accounted for the commonly-observed spontaneous potential phenomena around singlephase mineralogical conductors. The greater oxidation potential in the near-surface groundwater-bedrock environment relative to the deep environment was postulated to result in upward movement of electrons through conductive mineralisation (Fig. 3-7). At the surface of the conductor, oxidation of reducing agents in deeper areas provides electrons to the conductor, allowing the simultaneous reduction of oxidising agents in
104
S.M. Hamilton
[ .t t.- ~ ..... J
...
\\
+200
\
T
"
"
/
-." 7 ( ~ ) - - )
~
+100
..-
--
~
4 .....
/
'
........
"~
. . . . . . . . . . . . . . . .
..~.
, -
I
"
I.
"
§ J
.. mV
" 'ki
. . . . . . .
E.I .'3"-
O
(e)
i
,--
l J
. . . . . . . . . .
~
.........
.
_~. ~ . . .
/ 1 O0
-
/
.~,
" :/
~
/
-200
,
i
.
(5)
//
..
;
j .
.
.
.
.
.
..
..-L0~.. Equipotentiai lines .,- Negative Current flow ~,q:
,
.-
.
.
Positive/negative ion movement Cathode
~,
Electron flow
Im r.X)
Sulfide Anode
Fig. 3-7. Interpretation of the equipotential lines and ionic current flow lines around a singlephase sulphide ore body in a uniform redox field, after the model of Sato and Mooney (1960). Equipotential lines are labelled to depict an upward increasing gradient and are not intended to be an actual representation of the Earth field (from Hamilton, 1998).
shallower areas where electrons are received from the conductor. For a single-phase conductor, oxidation or reduction of the electronic conductor itself does not contribute directly to the process that results in spontaneous potential currents, i.e., does not contribute to the remote transfer of electrons from one area of the conductor to another. If the conductor were the reducing agent it would oxidise in its upper part, where oxidising agents are more abundant, negative current would not move along its length from depth and there would be no resulting spontaneous-potential phenomena associated with the conductor. Oxidation of the conductor takes place as a local detached redox phenomenon that does not contribute directly to spontaneous potentials. The movement of electrical current along the conductor necessitates the mass transport of ions in groundwater to, or away from, the electrodes in order to deposit charge and/or to prevent local charge imbalances caused by the production or deposition of charged species. In general there will be a migration from surrounding areas of positive charge toward (i.e., negative charge away from) the upper part of the conductor and negative charge toward (i.e., positive charge away from) the lower part of the conductor. Thus the conductor functions as both the electrodes and the wire in a natural voltaic cell, connecting the cathodic part of the conductor near surface with the anodic part in the deeper environment. The reactants are solid-phase and dissolved constituents in the low-Eh and high-Eh environments that respectively surround the anode and cathode. The difference in oxidation potential of the reactants arises from the ubiquitous redox
Spontaneous potentials and electrochemical cells
105
gradient that exists in the Earth's crust. The conductor provides a less resistive route to upward flow of negative current from depth and concentrates the background current, thereby short-circuiting the redox field. The conductor and redox field combined represent the voltaic cell: without the redox field, there would be no reactants; and without the conductor there would be no focusing of current and hence no cell. The consumption of oxidising agents around the upper part of the conductor results in more reducing conditions immediately around the top of the conductor than in adjacent areas (Hamilton, 1998). Likewise, locally-anomalous oxidised conditions develop around the bottom of the conductor because of the consumption of reducing agents. However, conditions can never become as reducing at the top of the conductor as they are at the lower end or all current would cease. The result of the process is to modify the shape of the redox field around the conductor (Fig. 3-7). This also modifies the lines of current flux since, in isotropic media, current moves perpendicular to lines of equal potential. The specific redox half-reactions that result in current flow through the conductor are of less importance than their aggregate effect on voltage. The stability field of water restricts the overall voltage of the cell to less than 1500 mV. Redox reactions involving the reduction or oxidation of water, such as occur in most electrolytic cells, are unlikely to be part of the SP charge-transfer mechanism despite contrary statements by some authors (Bolviken and Logn, 1975; Sivenas and Beales, 1982; Clark, 1997). Electrochemical cells after the model of Sato and Mooney (1960) develop due to zones of anomalously-high electrical conductivity in Earth materials in what would otherwise be a roughly uniform vertical redox gradient. The issue of the conductor being electronically conductive is, perhaps, a red herring. A zone of fault gouge made up of water-saturated rock flour and phylosilicates could conceivably have a very high electrolytic electrical conductivity, especially relative to surrounding poorly-fractured rock. This should also develop a significant electrical current within it, provided an upward-increasing redox gradient also exists. Two hidden assumptions implicit in the model of Sato and Mooney (1960) are: (1) that the conductor consists of a single phase, such as graphite or pyrite; and (2) that oxidation of the conductor would result in its conversion to a non-conductive phase. Thornber (1975a, 1975b) presents a reactive conductor model in which the conductor itself is the reducing agent, which is in apparent contrast to the model of Sato and Mooney. However, the reactive conductor model is based on the presence of one oxidised phase and at least one reduced phase relative to the first phase, all of which are electronically conductive. Such scenarios have been noted in terrain with deep weathering profiles due to the phase conversion of reduced sulphide minerals to more oxidised forms. A sulphide body containing pyrrhotite in its lower part and pyrite in its upper part is an example of such a cell (Fig. 3-8). Pyrrhotite at the boundary between the two phases oxidises to form pyrite, which remains electronically conductive. Electrons liberated into the sulphide mineralisation move up toward the more oxidising environment and allow
106
S.M. Hamilton
+300 l ................. .....
.....
....... - .....
]
ii
~ - - N e g a t i v e Current Flow Lines .~.
,,. ........
........................ 3', .....
.... -.. ....
~
........... O
,.,
.0
9
~
-.
..,..."
"~.. r
Pyrite
..
~
+ ]ool- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ,.,
Fe S 1__~ 4FeS + 3Fe + 6e }"/" j''~Jr ,
,
.
" -id
" .......
9...
~
" ....
:
t.~ r ~,
....................
.... _
0"Q
.
r,...,.
0 mV
.
.
........
.
.
.
.
-
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Z
,,. ...............................
-]00
Fig. 3-8. Interpretation of the equipotential lines and ionic current flow lines around a two-phase sulphide body (pyrite/pyrrhotite), after the model of Thornber (1975a). The purpose of the labelled equipotential lines is as in Fig. 3-7 (from Thornber, 1975a, 1975b).
the reduction of oxidising agents in the area surrounding the upper part of the conductor. The anodic half-reaction at the pyrrhotite/pyrite interface is (Thornber, 1975a): FeTS8 ~ 4 F e S / + 3Fe 2++ 6eElectrons pass into the conductor and positive charge moves into the groundwater environment in the form of Fe 2§ For every mole of pyrrhotite oxidised at the anode, three moles of Fe 2§ are released into the groundwater electrolyte. Groundwater around the anode acts merely as an electrolyte that accepts positive charge from the sulphide anode. The redox potential of groundwater in contact with the anode does not directly affect the process but must be sufficiently low to prevent the oxidation of Fe 2§ to Fe 3§ If it were not, the groundwater would directly oxidise the pyrrhotite or even the pyrite, there would be no separation of oxidising and reducing agents and hence no SP cell. At the upper, cathodic end of the cell, groundwater plays a similar role in the SP process to its role in the reactive groundwater model, i.e., it provides oxidising agents that consume negative charge originating from the conductor. Therefore, the movement of ions should be similar to that occurring around the top of the reactive groundwater model. At the anodic pyrite/pyrrhotite interface, Fe 2+ moves away carrying charge upward along the increasing redox gradient and presumably oxidises, forming a gossan when it reaches the water table. It is unlikely that Fe 2§ generated at the anode would migrate toward the cathode in any redox-active role. There are two reasons for this. First, in the groundwater environment, Fe 2§ is redox-inert with respect to reduction and only
Spontaneous potentials and electrochemical cells
107
by oxidation can it pass on charge. Second, the cathode, like the anode, must be at least as reducing as Fe 2+, otherwise the sulphide would be locally oxidised in its upper end and no SP phenomenon would result. The pyrrhotite-to-pyrite reaction is just one example of a process that could generate such a cell. Another is pentlandite-to-violarite (Thornber, 1975a) and there are many other possibilities. The differentiation of a sulphide into oxidised and reduced phases in the upper and lower portions respectively is, in itself, a product of the operation of this type of cell. If subjected to long term oxidation, a deep pyrrhotite body will slowly oxidise to pyrite as the pyrrhotite/pyrite boundary progresses downward. These cells therefore require a deep weathering profile where: (1) there has been abundant time for the oxidised/reduced sulphide front to migrate downward; and (2) most species in the bedrock/groundwater environment surrounding the conductor that are more reducing than the reduced species (e.g., pyrrhotite) have already been consumed. As such, one would expect the development of cells such as these in old geological terrain but not in younger terrain such as continentally-glaciated areas. However the reactive groundwater and reactive conductor models may be end members and probably both operate to varying degrees wherever conductor-based cells occur.
Cells in the absence o f electronic conductors Virtually all discussion in mineral exploration regarding SP cells and associated electrochemical phenomena assumes the presence of an electronic conductor. There has been little discussion of voltaic cells that involve no electronic conduction, but these cells undoubtedly exist. The nervous systems and muscles of organisms use the transfer of purely ionic current with no electronic conduction. Spontaneous potentials in the absence of electronic conductors have long been recognised in the petroleum industry and result from salinity and redox differences between strata. The presence of spontaneous potentials has also been noted in relatively thick overburden overlying mineralisation in the absence of an overburden conductor of electrons (Burr, 1982). Since electrons cannot move freely in an electrolyte solution, many of these cases must involve electrochemical cells of sorts in which current is transferred exclusively in the form of ions. Two types of SP cells have been postulated to develop in media with presumably homogenous resistivity. These are SP cells over bedrock mineralisation and deep hydrocarbon-based cells in bedrock. These cells are not centred on zones of elevated electrical conductivity but rather on zones of elevated SP (voltage) gradient. In both the reactive groundwater and reactive conductor models, the impetus for electronic current flow in mineralisation comes from the redox differential between the oxidised groundwater environment surrounding the upper part of the conductor and reducing agents in contact with its lower part. The upward movement of electrons consumes oxidising agents in basal overburden and results in the development of a
108
S.M. Hamilton
negative redox anisotropy in groundwater-saturated basal overburden over mineralisation relative to the surrounding overburden over country rock. Whilst electronic current continues, there must be continual outward dissipation of negative charge into surrounding overburden; otherwise the build-up of reducing conditions around the top of the conductor would eliminate the voltage differential and current would cease. Once solid-phase oxidising agents are consumed, this dissipation of negative charge must take place by the outward migration of reduced ions away from, or the inward migration of oxidised ionic species toward, the cathode. This process is necessarily coupled with oxidation-reduction reactions, occurring between the top surface of the conductor and the water table, that involve both reduced and oxidised species. The dissipation of negative charge away from the conductor is accomplished as the oxidation states of reduced species change during this process, resulting in dissolved, gaseous or precipitated products that have a higher redox potential than the reduced reactants. Figure 3-9 provides a hypothetical example of the possible outcome of this process in young, exotic sediments. Figure 3-9A depicts a fine-grained glacial material, shortly after deposition, in which a background redox differential of 150 mV exists between groundwater at the water table and in basal overburden units. An electronicallyconductive, steeply-dipping mineral deposit occurs in bedrock. More reducing conditions immediately above the conductor result in a redox differential of 300 mV between the top of mineralisation and the water table. Spontaneous potential contrasts exceeding 150 mV between conductive mineralisation and adjacent rock have often been reported (e.g., Pflug et al., 1996; Bolviken and Logn, 1975). At the time of overburden deposition, a very strong vertical redox gradient exists just above the bedrock conductor along which ions have a tendency to move. The outward movement of reduced ions such as HS, Fe 2+ or $2032- results in the migration of a reduced front away from mineralisation. At the front, reduced ions come into contact with oxidising agents and redox reactions take place, thereby dissipating negative charge away from the conductor. Once the front reaches the water table, a reduced column will have developed in the groundwater-saturated overburden above mineralisation. If the capacity for electrical current in the conductor is high, the production of reduced species might exceed the capacity of the unsaturated zone above the column to provide oxygen across the water-table phase boundary. In this case, the column would probably widen out. As it widens, the surface area exposed to oxidising agents along the entire reducing front increases. Once the column diameter is sufficiently large, the capability of the water-table phase boundary and surrounding overburden to provide oxidising agents equals that of the conductor to provide negative charge. The supply of oxidising and reducing agents is therefore balanced and a steady-state kinetic equilibrium is established between the two processes within the cell. The equipotential lines around the column above the conductor cease to move outward and their previously-horizontal configuration becomes nearly vertical in the vicinity of the reduced column. The end result is a permanent Eh anisotropy in overburden between the
Spontaneous potentials and electrochemical cells
109
Fig. 3-9. The progressive modification of redox equipotential lines in saturated overburden overlying an electronic conductor in bedrock. Negative current flow lines depict the movement of negative charge-carrying species such as Fe2+, 82032" and Co2§ Positive charge-carrying ions such as UO22§ MoO4 2, SO42 and dissolved oxygen radicals have similar flow lines but in the opposite direction. The purpose of the labelled equipotential lines is as in Fig. 3-7 (from Hamilton, 1998).
conductor and the water table, due to the upward propagation of the redox anisotropy from the bedrock surface. In the fully-developed column (Fig. 3-9B), electrical current must become focused at the flanks of the column because the SP gradient is stronger in that direction than it is in the vertical direction. For every electron passed upward along the conductor, a corresponding amount of reduced species must move away from, or oxidised species move toward, the conductor. This continual migration of redox-active species must be coupled with redox reactions in order to transfer charge. If redox equipotential lines are totally static, the production of reduced species at the conductor must be accompanied by the simultaneous consumption of reduced species somewhere between bedrock and the water table. This would result in the almost instantaneous transfer of electrical current despite the much longer time required for mass transport of reduced species to the ground surface (see discussion on ion mobility, below). There are many redox-active ions that could potentially carry charge including abundant reduced anionic sulphur species and ferrous iron. The migration of cations and anions occurs simultaneously and must be exactly balanced, after accounting for precipitation and other fixation reactions, in order to maintain macroscopic charge
110
S.M. Hamilton
balance in all parts of the groundwater environment. The migration of redox-inert ions (such as Na + and CI) may play a role in maintaining charge balance in some part of the cell. The model just described was proposed by Hamilton (1998) to account for selective leach geochemical anomalies over conductive mineralisation in high water-table Quaternary glaciated terrain. It should, however, be applicable in most overburden environments that meet the main criterion for the development of such a cell, i.e., overlying a redox anisotropy on the bedrock surface. The presence of any geological unit that is strongly reducing relative to surrounding rock should result in an overburden cell. Indeed, the association of reduced bodies with field data that suggest electrochemical activity has been made in the past (Clark, 1996, 1997). Probable reduced sources of such cells are discussed later in this chapter. Since at least the late 1970s, very large-scale, low-voltage SP current fluxes have been thought to occur in close association with petroleum reservoirs. Pirson (1981) attributed these to the reducing action of hydrocarbons in seepage areas above reservoirs. All petroleum reservoirs have overlying seepage areas of varying size, as evidenced by drill cuttings. These areas are attributed to the movement of fluids (mainly water) carrying trace amounts of hydrocarbons upward through microfractures in the shale as part of the early compaction process and later due to lithostatic pressure. Core logging technology and experimentation (Tomkins, 1990) have revealed that these areas are strongly reducing due to the cracking of long-chain hydrocarbons by natural zeolites in the shale. This creates around most reservoirs an envelope or bag of negative charge that extends over a hundred metres above and a few metres below the petroleum-bearing horizons (Tomkins, 1990). Pirson (1981) provided core-log evidence that large-scale electrical gradients are established between the reduced zone above the reservoir and the more oxidised zone in the shallow subsurface and referred to the process as a "redox cell". Tomkins (1990) further developed this model by reviewing many of the commonly-associated physical features of petroleum reservoirs and unifying them into a single model based on a central redox-cell overlying the reservoir. Such physical features include areas of darker surface discoloration relative to the more oxidised plain surrounding the hydrocarbon zone, metal anomalies in a halo around the zone, magnetite and magnetic sulphide mineralisation above the zone, calcareous cementation above the reservoir and uranium deposits flanking it. He also attributed leakage of gases such as CO2, He and Rn to redox processes but did not clearly indicate how these related to the cell. Tompkins considered current in these reservoir-based cells to be maintained by ongoing seepage and cracking processes. He cites other sources as having determined that depletion of the reservoir results in a reversal of hydrostatic gradients above the reservoir which cuts off the upward seepage of hydrocarbons and effectively "shuts off" current flow. This redox cell of Pirson (1981) and Tomkins (1990) has sources of oxidising and reducing agents that are separated in space and also has a zone of apparently-elevated current between the two sources, although, to the author's knowledge, this current
Spontaneous potentials and electrochemical cells
111
Fig. 3-10. Interpretation of the equipotential lines and ionic current flow lines around a petroleum reservoir, after the model of Pirson (1981) and Tompkins (1990). The purpose of the labelled equipotential lines is as in Fig. 3-7.
cannot be measured directly. Despite the enormous scale (kilometres in depth), this clearly represents a SP cell of the type based on an area of elevated voltage gradient, not increased electrical conductivity. The model implicitly assumes an upward increasing redox gradient in place above the petroleum reservoir, along which negatively-charged species move. Figure 3-10 is a diagrammatic representation of this model superimposed on a background redox field. The redox environment in the country rock surrounding the reservoir is likely to be less reducing than the reservoir itself; otherwise there would be no redox differential and no cell could develop. It must, however, be significantly more reducing than the ground surface or current from the reservoir would not flow up but rather would flow laterally outward. Since the lower limit of the aqueous redox field in the Earth is limited only by the redox stability of water, the downward decreasing redox field should re-establish itself at some point below the reservoir. This scenario should produce equipotential lines and ionic current patterns similar to those shown in Fig. 3-10. It should be possible for both anionic and cationic species to move upward along this gradient provided they carry negative charge. Figure 3-10 is a hypothetical representation and it should be noted that the ionic current configuration is not similar to that of Pirson (1981, figs. 13 and 14), nor is the equipotential line distribution the same as that which he inferred. Tomkins (1990, fig. 1) showed no equipotential lines and no current flow lines per se.
112
S.M. Hamilton
Field evidence for the presence o f cells The four models of SP cells described above were developed independently in different disciplines of Earth science for different reasons. However, they all address field evidence that is best explained using electrochemical processes. The model of Sato and Mooney (1960) was developed to resolve many conflicting theories in the geophysical community as to the origin of spontaneous potential phenomena occurring over conductive sulphide deposits and graphite. The model of Thornber (1975a, 1975b) was developed partly on mineralogical evidence from supergene-enriched nickel deposits to: (1) better understand the process of supergene enrichment; and (2) improve exploration techniques for this type of deposit. The model of Hamilton (1998) was developed to account for selective leach geochemical anomalies that occur over gold and base metal deposits in thick (>30m), young (-8 Ka), water-saturated glacial deposits and that can not be explained by any other transport mechanism. The model of Pirson (1981) and Tomkins (1990) was developed to unify observed geochemical and geophysical phenomena that had been known for many years to occur over petroleum reservoirs. All four of these models involve a redox cell localised over a geological source of reducing agents. For many years, the formation of "rabbit-ear" anomalies has been attributed to electrochemical processes. Rabbit-ear anomalies were first recognised over conductive mineralisation in shallow glacial overburden environments (Govett 1973, 1976; Bolviken and Logn, 1975; Govett and Chork, 1977; Nuutilainen and Peuraniemi, 1977; Smee, 1983). They have since been recognised over much thicker glacial overburden (Jackson, 1995; Bajc, 1998). They also occur in temperate non-glacial and tropical areas and are very commonly reported in arid to semi-arid environments (e.g., Govett et al., 1984, Govett and Atherden, 1987; Clark, 1996, 1997; Smee, 1998). Where rabbit-ear anomalies are reported, usually apical anomalies of other elements are noted also. Anomalies in metals are most commonly reported because it is usual to determine metal concentrations in mineral exploration surveys. However, where H +, electrical conductivity and anions have been measured, these parameters have often been found to be anomalous. Apical and rabbit-ear patterns are almost invariably centred directly over mineralisation or other geological features such as faults (Clark, 1997; Bajc, 1998). Where corresponding SP data are available at sites, rabbit-ear anomalies usually correlate with central SP lows over mineralisation (Bolviken and Logn, 1975; Govett, 1976; Govett and Chork, 1977; Smee, 1983; Hamilton and McClenaghan, 1998). Until recently, the third-dimensional shape of such SP anomalies was not known and this led to various interpretations of current flow-patterns and ion movement. A recent crosssectional study (Hamilton and McClenaghan, 1998) involved down-hole SP in thick (30 m) glacial overburden overlain by a peat-bog and underlain by graphitic argillite hosting gold mineralisation. The results show equipotential lines and inferred ion flow patterns that closely follow the model of Hamilton (1998) and that correlate with apical and rabbit-ear geochemical anomalies in shallow groundwater and peat. It was concluded
Spontaneous potentials and electrochemical cells
113
that these cells should occur wherever the presence of conductive mineralisation results in redox anomalies on the bedrock surface. Part of the reason overburden cells such as these have not been recognised in the past might be that they would be extremely difficult to detect in the vadose zone. Much higher contrasts are expected between redox conditions inside and outside the column below the water table than would occur above the water table where oxidising agents are abundant and far more mobile. As such, the cells would develop best in areas of high water table and in these areas SP surveys have traditionally not been carried out because of false anomalies that often occur due to variable moisture content and pH. This is not to suggest that the cells cannot exist in the unsaturated zone above the water table, just that they are likely to be subtler. The presence of SP lows on surface above mineralisation in areas of a thick unsaturated zone (Parasnis, 1979; Burr, 1982; Govett et al., 1984) suggests that oxidising agents are consumed and that cells, as defined above, can also exist in the unsaturated zone. There is abundant field evidence for the existence of SP cells over petroleum reservoirs. Indeed this evidence itself was used to find petroleum reservoirs for almost 100 years before the cells were recognised. It includes gaseous and liquid hydrocarbon seeps, paraffin deposits, halo-type metal anomalies, iron and manganese deposition, carbonate cementation, hard-drilling areas, magnetic anomalies (associated with magnetite mineralisation) and elevated uranium concentrations (Tomkins, 1990). Chapters 5-7 of this volume and references provided therein document many of these features.
GEOCHEMICAL RESPONSE TO SPONTANEOUS POTENTIAL CELLS
Ion mobility One of the problems with the understanding of surface selective leach anomalies is their development in thick, young deposits. Glacial materials in Canada only 8000 years old have well-developed surface anomalies spatially associated with underlying mineralisation. As argued above, diffusion, gaseous carriers and advective groundwater transport can be ruled out as the primary transport mechanisms in these environments because they are too slow to account for most anomalies. The rate of movement of charged species in an electrical field, however, can far exceed the rate of other dispersal mechanisms. Ionic migration rates in an electric field are described by the equation (Webber, 1975; Glasstone and Lewis, 1960): s = ()~+/F)-(V/d)
114
S.M. Hamilton H§
1O0
~
x-
~ ~0 "0
0
I
0
2000
4000
6000
800C
Time (yeors} for ton migrotion through overburden
Fig. 3-11. Theoretical ion migration time as a function of overburden thickness for H+ and a hypothetical anion, X, under the influence of a potential difference of 300 mV (from Hamilton, 1998).
where, s = ion velocity, ~.+ = ion conductance, F = Faraday's constant (96500 Coulombs), V = potential difference and d = distance. Ion conductance is specific to species. Most major ions have similar ionic conductance ranging between about 50 and 70 (ohms-lcm2), although H + has an ionic conductance of 350 ohms-lcm 2, the highest of any ion. Figure 3-11 shows theoretical ion migration distances as a function of overburden thickness for H § and a hypothetical anion, X-, with k+ = 60. A voltage difference of 300 mV is used. Differences in Eh of this magnitude and greater are not uncommon between mineralisation and ground surface (Bolviken and Logn, 1975; Pflug et al., 1996; Hamilton and McClenaghan, 1998). Figure 3-11 shows that since glaciation, H + and other ions have had ample time to migrate through 30 m of non-reactive saturated overburden along an electrochemical gradient. Voltage differences as low as 10 mV for H + and 60 mV for X would be sufficient to move these species through a 30 m thick electrolyte in 8000 years. The calculation shown above assumes that groundwater behaves as a perfect electrolyte and that only the ions in question are present in solution to carry charge. In fact, there are many geochemical processes that could occur between bedrock and ground surface that would affect the migration rate. For charge to be transferred between mineralisation and ground surface, it is not necessary that individual ions move the entire distance. As a given ion moves upward along the redox gradient, it may move beyond its pH-Eh stability range and oxidise (Fig. 3-5) thereby passing on its charge to another species. If the new species carries negative charge it will also migrate upward, ultimately passing on charge itself, possibly by reaction with dissolved oxygen radicals at surface.
Spontaneous potentials and electrochemical cells
115
Sharply-defined Eh zones within the overburden stratigraphy tend to promote the replacement of charge-carrying ions with others. An actively-flowing and oxygenated aquifer, for example, might force a change in the charge carrier to a less reduced species, or might short-circuit the cell altogether so that the cell operates between mineralisation and the base of the aquifer instead of the ground surface. The water table is also a sharp redox boundary and almost certainly forces changes in the species that carry charge. A series of processes that involve the electrical double layer on clay and oxide surfaces could limit the rate of mass-transport of ions to the surface. Electro-filtration is the exclusion of some ions during their transport through double-layer media. Electrofiltration may limit the movement of certain ions though thick overburden thereby slowing the whole charge transfer process. Ion exchange and sorption are also doublelayer processes that could reduce the rate of mass and charge transfer. Ion exchange of a redox-active ion in solution for a redox-inert ion will effectively stop the transfer of charge at that exchange site and could exchange migrating species that originated from mineralisation with others that originated in overburden. This could slow down both mass and charge transfer until a significant proportion of the sorption or exchange sites are occupied by redox-active species. It should be noted, however, that double-layer processes limiting mass transfer are also likely to limit charge transfer. The commonlyreported surface selective leach anomalies in thick overburden above mineralisation are evidence that charge transfer does occur, even in young deposits. Therefore it appears that double-layer processes are not major inhibitors of the development of SP cells in overburden. Not all geochemical processes in overburden necessarily slow down mass transport and charge transfer. Groundwater in overburden above mineralisation may have higher ionic strength than that in surrounding areas due to the attraction of ions to, and/or generation of ions in, the reduced column. As such, there may be an area of increased electrical conductivity within the cell that should enhance current flow. Furthermore, clay surfaces impart a significant electrical conductivity to groundwater in overburden (which, incidentally, is also a double-layer process) and this might also enhance current relative to that which would occur in a pure electrolyte.
Geochemical anomalies The issue of replacement of the charge carrier as just described is extremely important to the use of selective leach procedures for geochemical exploration, particularly in younger terrain. In older geological terrain, the process may have operated for long enough to inundate the overburden with elements and species transported from depth and to give the overburden an elemental signature that is partially reflective of that of bedrock. In younger terrain, however, the charge could just as easily have been transported by exotic species liberated from overburden materials and, as such, surface anomalies would be partially or wholly reflective of the geochemistry of overburden. In
116
S.M. Hamilton
general there has been insufficient research into the issue of mass transfer between bedrock and ground surface by electrochemical processes. Therefore, despite the paramount importance of this issue to exploration geochemistry, it is not considered further here. Rather, attention is given to the theoretical development of surface geochemical anomalies above a reduced feature in bedrock without addressing the issue of where the ions that form the anomaly actually come from. The four types of redox cells described in this chapter should have similar surface geochemical phenomena associated with them because they share similar ionic flow patterns in the uppermost portion of the cells. They are all based on a reduced feature as a source of negative charge and the oxygenated surface as the ultimate source of positive charge, and they all must involve transfer of ionic current between the two sources. The
Fig. 3-12. The development of anomalies that are (A) directly related to electrochemical processes and (B) related to secondary processes occurring as a result of mobility and oxidation of ferrous iron.
Spontaneous potentials and electrochemical cells
117
result will be a general outward movement of ions that are capable of transporting negative charge, such as Fe z+, Co z+, Cu +, H S and $2032-, and an inward movement of ions that are capable of transporting positive charge, such as UO22+, MoO4 2-, VO43, SeO42, AsO43", $042 and dissolved oxygen radicals. The final transfer of negative charge at the top of the cells involves redox reactions that, in many cases, attenuate the transported reduced species into the solid phase, thereby forming geochemical anomalies. One outcome of the migration of ions from one redox region into another should be a zonation of elements in relation to the reduced column. Element zonation is a reported feature of selective leach anomalies (Clark, 1996). Zonation could occur due to a variety of processes, the most important of which would be progressive deposition of redoxactive species as they migrate into or out of the reduced column. The migration paths of reduced and oxidised ions are predictable provided the current flow patterns can be inferred and therefore, if the redox behaviour of a particular ion is known, the shape of anomalies can be inferred. Two of the factors that should control where mobile elements are deposited in relation to the reduced column are: (1) the Eh at which a redox-active species converts to a species of a different oxidation state; and (2) the mobility of the new species. For example, inward migrating species that are highly oxidising and show low mobility in reducing environments, such as UO22+, are expected to become reduced early and form anomalies at the outermost edges of the reduced column (Fig. 3-12A). At the other extreme, inward migrating, weakly oxidising species that can show high mobility in reducing environments, such as SO42, do not become reduced until they reach the inner part of the column, and even then might not form anomalies in soils because of their high mobility. Similar but opposite processes occur with reduced species in the outward direction. One of the more important reduced species capable of forming anomalies in soils is Fe 2+. Both the abundance of and the secondary processes associated with iron reactions suggest that Fe 2+ will have a major impact on the geochemistry of surface soils. It is moderately-to-strongly reducing and Fe 3+ has very low mobility. It is therefore expected to form anomalies at the redox front near the inner edge of the reduced column. When most reduced metals (and in particular, Fe 2+) oxidise, they hydrolyse water to form insoluble metal hydroxides. In the process, large amounts of acid are generated (Fig. 312B). Anomalies of H § are commonly reported in association with rabbit-ear anomalies over mineralisation. In the phreatic zone H + is most likely to be generated, not by the downward diffusion of oxygen and the oxidation of the sulphide itself, as in the past had always been assumed, but by the upward transport of reduced iron along an electrochemical gradient to oxidising agents at the redox front (i.e., the reducing agents are brought to the oxidising agents rather than the other way around). Secondary process related to iron oxidation may cause CO2 production in areas where carbonate is present (Fig. 3-12B). Acid produced at the edge of the reduced column by Fe 2+ oxidation should produce dissolved CO2 as H2CO3 (Fig. 3-12B) by dissolution of carbonate in rock or
1 18
S.M. Hamilton
overburden material. In adjacent areas near or above the water table, dissolved C O 2 c a n degas causing carbonate supersaturation and deposition. From here, CO2 and carbonatecharged soil moisture could also disperse upward through the unsaturated zone by capillary action. As moisture evaporates, near-surface CO2 would degas and carbonate would precipitate, forming both soil carbonate and CO2 gas anomalies (Smee, 1998). Smee (1998) attributed rabbit-ear anomalies of Ca, Mg, Sr and, possibly, Au and As to their transport in bicarbonate complexes and their precipitation in shallow soils due to processes such as these occurring in the unsaturated zone. The thousands of possible redox reactions that facilitate charge transfer away from a reduced source could result in a net loss of cations or anions from solution in the vicinity of some reactions. As such, the movement of redox-inert species is likely to be continuously occurring from one part of the overburden electrolyte to another in order to prevent local charge imbalances. However, it is difficult to predict the transport paths of these ions because their movement depends on the specific nature of the reactions occurring at a given site. Empirical observation is probably the most viable method for determining the behaviour of a redox-inert species. Clark (1996) reports that an empirically-observed oxidation suite of elements that often forms rabbit-ear anomalies flanking ore deposits includes C1, Br, I, As, Sb, Mo, W, Re, Se, Te, V, U and Th. Most of the elements in this suite are far more mobile in their oxidised forms than in their reduced forms and as such probably migrate inward to form reduction anomalies (notwithstanding the name applied to the suite). However, several elements are likely to be mobile as redox-inert species (e.g., CI, Br, I) or perhaps as ligands in metallic complexes, and therefore their movement may be a secondary result of other species migrating and subsequent reactions occurring inside and outside the column.
CONCLUSIONS Despite the different origins of the four SP cell models described in this chapter, their underlying principles are similar and can be summarised as follows. Within the redox field of the Earth, Ohm's Law suggests that a cell will form if either the redox gradient or electrical conductivity is anomalously high with respect to that of surrounding Earth materials. An Earth cell is defined as an area of spontaneously elevated electrical current between two separated sources of oxidising and reducing agents. Resulting similarities in the geochemical surface expression of the cells would include: a reduced SP centre above the cell; the outward movement of negatively charged ions; the inward movement of positively charged ions; and an element dispersal pattern that is characteristic of a redox cell. The most typical anomaly morphologies are the centred rabbit-ear and apical patterns which are often coincident in different elements. Thus, the primary criterion for development of a cell is the existence of a buried reduced feature. Such features of potential economic interest include the following.
Spontaneous potentials and electrochemical cells 9 9
9 9 9
9 9
119
Conductive sulphide mineralisation. Non-conductive but oxidisable sulphide mineralisation (e.g., sphalerite). Ultramafic dikes, diatremes and flows, including komatiites, kimberlites and lamprophyres. Geological contacts between two units with strong redox contrast, such as carbonatite and granite. Shear-zones containing fault gouge. Graphitic-hosted gold mineralisation. Bitumen, coal and natural gas seeps.
There are, however, a considerable number of sources of reducing agents that are of no economic significance. These include the following. 9 9 9 9 9
Barren sulphide mineralisation. Geological contacts between two units with strong redox contrast, such as diabase and granite Minor barren graphite horizons. Methane pockets in shale or overburden. Gas hydrates.
The study of electrochemical cell processes that result in surface geochemical anomalies is still in its infancy. Many apparently unrelated geochemical processes and phenomena occurring over chemically-reduced geological features may prove to be directly or indirectly related to electrochemical processes. Although electro-geochemical techniques have enormous potential, a great deal of additional work must be done if geochemical mapping over deeply buried features is to be a reliable technique yielding easily interpretable data. The most pressing issues are to determine mass transport rates from bedrock, whether a near-surface signature due to bedrock is possible within the age period of young deposits, and whether the replacement of charge carriers between bedrock and ground surface results in spurious overburden-related anomalies. Another important issue is the understanding of the specifics of mass and charge transfer from the reduced source to the groundwater environment.
This Page Intentionally Left Blank
P A R T II.
REMOTE DISPERSION PATTERNS OF C O - G E N E T I C P R O V E N A N C E
This Page Intentionally Left Blank
Geochemical Remote Sensing of the Subsurface Edited by M. Hale Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
123
Chapter 4
CARBON DIOXIDE DISPERSION HALOS AROUND MINERAL DEPOSITS M. ZHANG
INTRODUCTION The carbonate gangue associated with many types of mineral deposits is evidence that large quantities of CO2 often accompany the emplacement of mineralisation. Detection in soil of traces of the dispersion pattern of this CO2 is postulated to produce anomalies indicative of concealed mineralisation. Therefore a method for determining this CO2 was developed. It has been tested in China over various types of mineralisation covered by different thicknesses of soil in the provinces of Gansu, Shandong, Anhui and Jiangsu and in the region of Shanghai.
METHOD Soil samples are normally taken at intervals of 20 m along traverses. Over known mineralisation or zones of structural control, the interval has been reduced to 10 m in the case histories described here, whereas far from mineralisation the interval has been increased to 30-40 m. Two methods of sample preparation have been tested. In the first, the sample is placed in a plastic bottle immediately and the bottle is sealed. The contents are allowed to dry for 40-80 days prior to sieving. This method minimises contamination, but sample transportation is inconvenient and samples may still be damp after the period allowed for drying, making sieving difficult. In the second method, samples are air dried soon after they are collected, then sieved through a 80-120 mesh sieve. In orientation studies the first method gave slightly sharper anomalies than the second, but the difference was insufficient to warrant the additional inconvenience. Thus the air-drying method has been used is the case histories described below. Analysis involves boiling the sample in distilled water to release CO2 into solution from carbonates that break down at temperatures up to 100~ Dissolved CO2 is then determined by titration using a colour indicator and photoelectric device to control the termination of titration.
124
M. Zhang
CASE HISTORIES The skarn copper deposits at Huaitongshan, Gansu province, which are accompanied by some minor lead-zinc veins, lie of depths of 5-80 m. The area is arid and the surface is covered by Quaternary sediments on which a poor sandy and rudaceous soil, 3-50 cm thick, has developed. A traverse of soil samples revealed CO2 anomalies reaching 285 ppm over the main copper mineralisation, a contrast of about 4.4 (Fig. 4-1). The weaker anomaly further north is thought to lie over minor lead-zinc veins. The iron ore deposit at Wang-wang, Shandong province, has been described as a pneumatolytic high-temperature skarn. The ore minerals are sulphur-rich magnetite and minor pyrrhotite. The deposit is about 300 m below surface. The Quaternary cover is 130-180 m thick. The area is semi-arid, with soils under cultivation. Along a soil traverse, a CO2 anomaly reaching 285 ppm occurs over the mineralisation, and against background levels has a contrast factor of 3.6 (Fig. 4-2). This result is of particular interest for two reasons: (a) the deposit is not a sulphide ore body, but a high-sulphur iron deposit; and (b) other methods for detecting concealed mineralisation (geoelectrochemistry and Hg) did not yield anomalies here. The Shanbaidu polymetallic deposit, Jiangsu province, lies between 5 and 40 m below the surface and is overlain by eluvium and alluvium 5-15 m thick. The area is semi-arid and the soils are cultivated. Soil samples collected at three depths along a
Fig. 4-1. Relation between carbon dioxide in soil and copper mineralisation at Huaitongshan, Gansu province, China.
Carbon dioxide dispersion halos around mineral deposits
125
traverse reveal broadly similar CO2 patterns (Fig. 4-3). In the centre of the traverse, over the mineral deposit, soils from 30 and 60 em exhibit anomalies reaching 400 ppm CO2, whilst anomalies in soils from 100 cm are rather weaker. All of these anomalies stand out clearly against the low background in the southern part of the traverse, where the overburden is eluvium. However, they have, at best, poor contrast relative to the higher background in the northem part of the traverse, where the overburden is alluvium. The Qixiashan lead-zinc deposit is a hydrothermal lead-zinc mineralisation in a cataclastic fault zone in limestone. The ore body is 300 m below surface, and the overburden comprises 30 m of Quatemary alluvium. Soil samples were collected along a traverse over the mineralisation in three successive years and analysed for CO2 (Fig. 44). The samples from Year 1 yielded the highest COz concentrations and broadest anomaly, but contrast is poor. In Year 2 the anomaly is about 200 ppm CO2 and is best developed around a fault that cuts the mineralisation at depth. Anomaly contrast is put at 4.2. Samples were taken from a depth of 80 cm in Year 3, compared to 30 cm in earlier years; the resulting CO2 pattern, however, is closely similar to that found in Year 2. Analysis of bore hole samples for CO2, Hg, Pb, Zn, Cu and Ag showed that the Qixiashan deposit is vertically and horizontally zoned (Fig. 4-5). The highest COz concentrations are found close to the richest ore. The CO2 halo extends upward on the hangingwall side of the deposit and spreads laterally. This halo is much better developed above the mineralisation than those of the base and precious metals; only the Hg halo extends as far upward, and it is also rather wider.
Fig. 4-2. Relation between carbon dioxide in soil and iron mineralisation at Wang-wang, Shandong province, China.
M.Zhang
126 c02[
(ppm) ]
30 cm soil
sampling depth
60 cm soil
sampling depth
400 l 300
]
200 I
C02 (ppm) 400 300 200 100 0
C02
(ppm) 400
100 cm soil s a m p l i n g
depth
300 200 100 0
Fig. 4-3. Relation between carbon dioxide in soil and polymetallic mineralisation at Shanbaidu, Jiangsu province, China.
The Xuanchengshan cupriferous pyrite deposit, Anhui province, occurs in the cataclastic zones between granodiorite and limestone, and is of hydrothermal origin. It is overlain by 300 m of basic rocks and 10 m of alluvial sediments. Samples from a traverse at surface above the mineralisation reveal a broad, clear CO2 anomaly with peak values in excess of 200 ppm CO2 and a contrast factor of 6 (Fig. 4-6). Of the other elements determined along this traverse, Mo, Sn and W yielded no anomalies, and a Cu anomaly is confined to a single point. However the pattern for thermally-released Hg matches that of CO2. The primary halo of CO2 at Xuanchengshan is similar to that at Qixiashan. The highest concentrations of CO2 occur within the ore body, and the halo extends outward and upward. Nearer the surface is a second lens enriched in CO2, related to ore mineral formation and chloritisation. The halo around this extends towards the surface and its suboutcrop exactly matches the extent of the anomaly on the surface traverse.
Carbon dioxide dispersion halos around mineral deposits
127
Y e a r 1 (30 c m s o i l s a m p l i n g d e p t h )
,.., 400 E Q. 300 200
O" lOO (,..1 0
"~
n. 250 n. 200
,
9
'
9
.
.
.
.
Y e a r 2 (30 cm s o i l s a m p l i n g d e p t h )
" " 160 ,
A
E 2so
,
,
,
,
,
,
Y e a r 3 (80 cm s o i l s a m p l i n g d e p t h )
Q" 200 15o . 100 0 6O (.) 0
Fig. 4-4. Relation between carbon dioxide in soil and lead-zinc mineralisation at Qixiashan, China.
DISCUSSION
Speciation o f carbon dioxide What is the parent or form of CO2 detected in these studies? Theoretical considerations suggest that, of the various carbonates and bicarbonates likely to occur in rocks and soils,, NaHCO3 releases the most CO2 at relatively low temperatures; at 100~ it liberates 32% of its CO2 content, whereas metal carbonates, CaCO3, K2CO3 and NazCO3 do not begin to break down until much higher temperatures are reached (Fig. 4-7). The thermal stability of these carbonates was confirmed by laboratory experiments in which they were boiled in distilled water and the CO2 released into solution was determined by titration (Table 4-I).
M.Zhang
128
Fig. 4-5. Lithogeochemical halos at the Qixiashan lead-zinc deposit, China.
TABLE 4-I Release of CO2 into solution from pure carbonates at 100~ Carbonate CO2 release (%)
NaHCO3 9.62
n.d. - not detected
Na2CO3 0.31
K2CO3 0.15
CaCO3 n.d.
Carbon dioxide dispersion halos around mineral deposits
129
Fig. 4-6. Relation between carbon dioxide in soil and lithogeochemical halos at Xuanchengshan cupriferous pyrite deposit, Anhui province, China.
Some CO2 could also be present in soils in an adsorbed form, having previously been transported in solution. This hypothesis was investigated using five soil samples, of which three were from a humid area and two were from as arid area (Table 4-1I). A sample from a background site in the humid area was found to contain 84 ppm CO2 in the adsorbed form, compared with only 14% that was released by boiling; thus 86% of the CO2 was in the adsorbed form. In anomalous samples from the same area, 34% and 52% of the CO2 was in the adsorbed form. In the arid area, however, the two anomalous samples contained only 26% and 27% of their CO2 in the adsorbed form. These experiments suggest that in arid areas, CO2 related to mineral deposits occurs in soils mainly as salts of bicarbonates (which break down upon gentle heating). In humid areas, by contrast, CO2 related to mineral deposits coexists as bicarbonates and adsorbed CO2.
M.Zhang
130 #
100
v
76
60
32
25
: I '/ I
100 2 0 0
//"5,
!
400
600
/ 800
oo-
1000 (C ~)
Fig. 4-7. Thermal stability of some simple carbonates and bicarbonates.
TABLE 4-II Speciation of CO2 (ppm) in soils from two areas in China Form of CO2
Adsorbed Boil 100~ Total % adsorbed
Background
Humid area Anomaly 1
Anomaly 2
84 14 98 86
168 152 320 52
48 92 140 34
Arid area Anomaly 1 Anomaly 2 126 364 490 26
126 334 460 27
Formation of carbon dioxide dispersion patterns The most likely source of most of the anomalous CO2 detected in these studies is the volatile fraction associated with the emplacement of mineralisation. These volatiles are rich in CO2, which can migrate along fractures above the main mineralisation. Some of this CO2 may form carbonates, while some may pass directly into circulating groundwater. Some carbonates may hydrolyse at a later stage, releasing CO/ into solution. Where circulating groundwater brings dissolved CO2 near to the surface, some CO2 exists as bicarbonates of low thermal stability, whilst some is adsorbed onto the
Carbon dioxide dispersion halos around mineral deposits
131
surfaces of soil particles. These two forms of CO2 (especially the former) therefore give a good indication at surface of buried and blind ore deposits. Some contribution to these patterns may come from CO2 released by postmineralisation fracturing of fluid inclusions that trapped CO2 during metallogenesis. A third source is the release of CO2 from carbonates as a result of sulphide oxidation. Where mineralisation is deeply buried and below the water table, however, any contribution from this process seems likely to be small.
Factors affecting carbon dioxide anomalies In wet soils, background levels of CO2 are increased due to the hydrolysis of carbonates. A traverse over mineralisation at Zhanglian was sampled once when the soil was wet and again in early spring when the soil was relatively drier. Sampling of wet soils produces a number of false anomalies of CO2, whilst the dry soils produced a satisfactory low contrast CO2 anomaly over the mineralisation only. Therefore sampling along river banks or in soils of varying moisture content should be avoided in order to obtain clear anomaly definition. Thin soils underlain by limestone can also be the source of false anomalies. On a traverse across a molybdenum deposit at Tongshan, Jiangsu province, which is covered by 0.3-1.0 m of eluvium, CO2 anomalies were found not only above the ore body but also on a nearby ridge. Here soils are particularly thin and poorly developed, and the underlying limestone creates an environment in which the pH can be as high as 8.5, allowing carbonate enrichment which in turn produces a false anomaly of CO2. The natural organic carbon content of the soil can affect the amount of CO2 detected, and evidence from one traverse suggests that proximity to the exhaust residues of road traffic could also be a source of CO2 in soil. The ratio of CO2 to organic carbon can be used to resolve this problem.
CONCLUSIONS Carbon dioxide, probably introduced at the time of emplacement of mineralisation, occurs in a form that is easily released at 100~ as halo around and above many mineral deposits. Along soil traverses over nine mineral deposits in China, determination of this CO2 yields good-contrast anomalies, even when the mineralisation is deeply buried. There is also an elevated CO2 expression in soils over faults in mineralised areas. In general, over those mineral deposits in China where both methods have been tested, there is considerable agreement between CO2 patterns and thermally-released Hg patterns (Chapter 13). The continuity and intensity of CO2 anomalies tends to be a little poorer. On the other hand, in some cases CO2 anomalies are present over mineralisation where thermally-released Hg anomalies are absent.
132
M.Zhang
The recognition of CO2 anomalies in soils can be used, along with other methods, for the detection of buried and blind mineral deposits. Samples are easy to obtain and prepare, and the same samples can be used for other routine geochemical analyses. The method described here for the determination of CO2 is simple, efficient, inexpensive and can be used to analyse large numbers of samples.
Geochemical Remote Sensing of the Subsurface Edited by M. Hale Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
133
Chapter 5
LIGHT HYDROCARBONS FOR PETROLEUM AND GAS PROSPECTING V.T. JONES, M.D. MATTHEWS and D.M. RICHERS
INTRODUCTION Surface geochemical prospecting for hydrocarbons includes a myriad of techniques ranging from the direct detection of hydrocarbons escaping from subsurface accumulations and source beds to identifying secondary responses in the soils, rocks and biota in proximity to such accumulations or source beds. In the historical sense, the observation of visible seepage of hydrocarbons is the oldest method of prospecting for petroleum. Drake's historic well near Titusville, Pennsylvania, was drilled on the basis of a seep in the adjacent creek bed. The relationship of such "macroseeps" to reservoirs was well established by Link (1952), who stated: "A look at the exploration history of the important oil areas of the world proves conclusively that oil and gas seeps gave the first clues to most oil-producing regions. Many great oil fields are the direct result of seepage drilling". In this respect, few would argue that the presence of a macroseep indicates the presence of petroleum migration or surface source beds. Microseeps, or smaller scale macroseeps, also occur because of the physical continuity necessarily imposed by nature. These are invisible seeps, usually detectable only by sensitive instruments or by the visible result of their effect on the near-surface environment. These microseeps, although perhaps not as obvious or dramatic as macroseeps, are just as valid for the exploration of undiscovered reserves. This chapter presents the conceptual and practical application of microseepage detection and interpretation in the evaluation of areas for their subsurface hydrocarbon potential. Five factors are necessary to form a hydrocarbon reservoir. These are: (1) a source; (2) a reservoir in which the hydrocarbons can collect or concentrate; (3) a means of trapping these fluids in this reservoir (a seal); (4) a pathway to the reservoir (migration); and (5) the proper timing such that the source, reservoir, seal and migration pathway are present when required. Near-surface seepage of thermogenic hydrocarbons indicates the subsurface presence of a mature source and migration pathway. It also suggests that, if the hydrocarbons are reservoired, the seal is imperfect. This is true of both macroseepage and microseepage. Patton and Manwaring (1984) found that even in an area of extensive evaporites (Hugoton Field, Kansas), the seal was not perfect, and that microseepage could be detected in the vicinity of the Syracuse Fault.
134
V.T. Jones, M.D. Matthews and D.M. Richers
Basically, surface geochemical prospecting is a source-rock tool applied at the surface. The magnitude of a microseep from a reservoir is related to the permeability of the migration pathway (and not to the economic worth of the reservoir). A surface geochemical survey is not currently, and perhaps never will be, a stand-alone prospect tool. However, with judicious use, this technology can provide information on the maturity of source beds in a basin and the composition of subsurface hydrocarbons. In addition, detection of surface microseepage allows mapping the surface expression of the migration patterns created by the expulsion of fluids as a basin compacts and matures. When used in conjunction with geophysical and geological information, geochemical data can refine subsurface models of hydrocarbon trapping and migration configurations. It is only through careful analysis and integration with other exploration tools that one can achieve the optimum benefits from this technology. Near-surface hydrocarbon detection techniques have been shown in both the former USSR and the United States to be capable of distinguishing basins (or large portions of basins) that are unproductive from those that are productive, and of distinguishing the type of production (oil, gas, or mixed oil and gas). This ability has been independently recognised by Jones and Drozd (1979), Mousseau and Williams (1979), Janezic (1979), Weismann (1980), Drozd et al. (1981), Jones and Drozd (1983), Richers (1984), McCrossan et al. (1971), Richers et al. (1982, 1986), Horvitz (1985) and Klusman and Voorhees (1983). Surface geochemical techniques can select which of several frontier basins has the greatest chance of containing reservoired hydrocarbons, and the expected composition (gas, oil, mixed), in addition to high-grading portions of these basins that have the highest potential. The premise that microseeps occur and that they provide useful information for exploration is no longer questionable.
ORIGINS OF LIGHT HYDROCARBON GASES
Origin o f petroleum The formation of petroleum and natural gas from organic matter through increasing depth of burial and temperature has been very well established by many geochemical studies (Tissot and Welte, 1978; Hunt, 1979). As shown in Fig. 5-1, the generation of the light hydrocarbon gases, methane (C~), ethane (C2), propane (C3) and the butanes (C4), occurs in three main stages: diagenesis (<50~ catagenesis (50-200~ and metamorphism (>200~ in which only dry gas and ultimately graphite are formed. During the first stage bacteria acting under reducing conditions on organic substrates in sediments form predominantly methane. According to Hunt (1979), about 82% of the methane and practically all the heavier hydrocarbon gases are formed in the next, catagenic stage. Ethane, propane and the butanes are formed in the temperature range from 70 - 150~ with peak generation occurring around 120~ As shown in Fig. 5-1, a very large thermal methane peak occurs near 150~
Light hydrocarbons for petroleum and gas prospecting Sapropellc Source 20Diagenesls
~/ .........
13 5
Humic Source
#41
-
50 -
68
- - 122
100 -
-
212
-
302
-
392
Catagenesis
150--
200 Metamorphism
Or
Relative Yield o f G a s from Organic Matter in Fine-Grained Sediments
Fig. 5-1. Generation of gases with depth: C2+ represents hydrocarbons heavier than CH4; N2 is generated initially as NH3.
I
Ro (%)
10 I
I----hydrocarbonsgenerated ---~l
-70
~-70
- Imr,,.)~"60 ol.5
~-60
1.2
2.0 3.0 20 3~0
-40
20 I I
C2+[%1--~ 30 50
] I I l',*r-Q
/ --e--~iZI
I,
-40 ~ . t:~,/f/,,
9
xe -20 100
I 90
9 , 80
?,V-r 70 50
Fig. 5-2. Genetic characterisation of natural gases by compositional and isotopic variation: (a) formation of natural gas and petroleum in relation to maturity of organic matter; (b) relative concentration of C2+ hydrocarbons in relation to ~3C in CH4, with arrows indicating compositional changes due to shallow migration (Ms) and deep migration (Md) (reproduced with permission of the American Association of Petroleum Geologists, whose permission is required for future use, from Schoell, 1983, AAPG Bull., vol. 67, no. 12, Fig. l, pp. 2226-2227, AAPG 9
136
V. T. Jones, M.D. Matthews and D.M. Richers
-3
O(Oa)
5 DcH4[ppt]
-250 -200
-150
(b) -lO0
-70
-70 t
-60
-60
8 13Ccthane[ppt ] -50
-40
I
I
-30
-20 I
I
/
-50
d
r..)
m
cO -40
-30
-20
-40
3~
l
~(h~ ~
I
I
Fig. 5-3. Genetic characterisation of natural gases by isotopic variation: (a) relative concentration of deuterium and ~3C in C|t4; (b) relative concentration of '3C in CH4 and in C2H5, with arrow Md indicating the result of mixing with isotopically-positive CH4 from depth, and arrow Ms indicating the result of shallow mixing with isotopically-negative CH4 ol" biogenic origin (reproduced with permission of the American Association of Petroleum Geologists, whose permission is required for future use, from Schoell, 1983, AAPG Bull., vol. 67, no. 12, Fig. 1, pp. 2226-2227, AAPG~) 1983).
In addition to time, the quantity of gaseous hydrocarbons formed varies with the type of organic source material, which can be broadly classified as sapropelic (marine) or humic (terrestrial). As shown in Fig. 5-1, considerably more C2-C4 and other oil-type hydrocarbons are generated from sapropelic sources than from humic sources. In addition to the different volumes and types of petroleum (oil versus gas) produced from the two source materials, their carbon isotope compositions are different; terrestrial organic matter is reported to have lower ~3C concentrations than marine organisms (Galimov, 1968; Silverman and Epstein, 1958). The carbon isotope concentration of ~3C, as compared to t2C, is also very useful for classifying natural gases as to their source type and/or maturity. Maturity is generally proportional to the depth of generation. Schematic diagrams published by Schoell (1983a, 1983b) distinguish the major natural gas types into three end members, as shown in Fig. 5-2. Schoell suggests that most natural gases are admixtures of these three basic end members. As shown in Fig. 5-3, further classification of reservoir gas types can be made from data for deuterium in methane and ~3C in methane and ethane.
Light hydrocarbonsfor petroleumandgas prospecting
137
TABLE 5-I Literature review ofbiogenic light hydrocarbon production (C1-C4) Reference Coleman, 1979 Davis and Squires, 1954 Voytov et al., 1975 Kim and Douglas, 1972 Stahl, 1974 Bukova, 1959
Study Methane in glacial till gases: ~i3c1; CI/C 2 > 1000 Cellulose, ethanol and sewage fermentation: C2-C-4 < 2 x I0-5%; Cl/C2 > 105 Swamp gases, glacial till gases, subsoil bacterial gases: C1 > 99%; C2-C4 < 10-4'~'; C1/C2 > 106 Cellulose fermentation: C~ > 99%; C~/C2 > 500; C2-C4 <10-2% Bacterial gases: C~ >99%; 813CI <-60% Anaerobic bacterial decomposition, soil gases, sewage, silt: C~ > 99%; Ci/C2 >1000
Origin of light hydrocarbon gases in the near-surface The near-surface occurrence of ethane through butanes is of fundamental importance to the purpose of this chapter and to the usefulness of these gases as prospective indicators of buried natural gas and petroleum deposits. An extensive review of the literature suggests that C2-C 4 hydrocarbons can be generated biogenically; however, solid proof exists only for methane and ethylene as major products of bacteria (McKenna and Kallio, 1965). A review of the literature shown in Table 5-I provides conflicting evidence for the biogenic occurrence of the C2-C 4 hydrocarbons, although most of the literature suggests an abiogenic, thermocatalytic origin for these gases. Compositionally, however, these gases display large variations and do not resemble compositions characteristic of petroleum gases. All these studies are further characterised by methane:ethane ratios in excess of 1000 and a percent methane composition >99%, and are quite uncharacteristic of petrogenic gases. Some of the results reported before the invention of the gas chromatograph must be regarded with suspicion due to limitations of the analytical methods employed and possible sampling collection at locations contaminated by mixed biogenic and petrogenic gases. Russian researchers have illustrated that some of the earlier analytical methods, such as the combustion technique of Kartsev et al. (1959), can measure gases that are mistaken for hydrocarbons.
Laboratory and field evidence of biogenic C2-C4 hydrocarbons Studies were conducted at Gulf Research & Development Company by Janezic (1979) to investigate the anaerobic microbial evolution of C~-C4 hydrocarbons upon decomposition of various organic substrates including green plant branches, grass
138
V.T. Jones, M.D. Matthews and D.M. Richers
1 - Liter Modified SOHNGEN Flask
Nutrient Medlunl
Substrate Added
NH4CI
1.0 g
K.z*O4
1.og
Autoclave
e/
Nutrient Added
MgSO4 -7H20 0.5g Innoculate
Incubate Chromatographic Analysis
CaCI 2 Water
0.5 g 1000 ml
Deionlzed, Distilled, & Degassed
Fig. 5-4. Experimental scheme for anaerobic decomposition of natural organic matter (from Kim and Douglas, 1972).
clippings, plant roots, decayed wood and pure cellulose. These substrates were chosen to compare results with those of Davis and Squires (1954), Bukova, (1959), Smith and Ellis (1963), Kim and Douglas (1972) and Voytov et al. (1975), who used similar substrates. The experimental scheme for anaerobic decomposition is shown in Fig. 5-4. Exactly 1.5 g of each substrate was added to a modified 1 litre Sohngen flask and autoclaved at 120~ and 15 psi to ensure sterility, after which each flask was filled to capacity with a sterile inorganic nutrient medium and pH adjusted. Next 50 ml of a heterogeneous innoculum prepared from muds from a local lake was injected into each flask as "inoculates", while 50 ml. of sterile nutrient medium was used for control samples. Headspace C~-C4 hydrocarbons were measured prior to incubation to provide baseline concentrations. Minimum detection limits were 3 ppb on a volume basis using a highsensitivity gas chromatograph equipped with a flame ionisation detector. Samples were incubated at 25~ and 36~ over a five week period. The results are summarised in Table 5-II and Fig. 5-5. Of the organic substrates fermented, green plant branches, grass clippings and plant roots evolved significant quantities of gas after a few days of incubation. Of the C~-C4 hydrocarbons determined, only methane was observed in copious quantities, with minor amounts of ethylene coproduced. Ethane, propane and butanes were not evolved, in good agreement with the work of Kim and Douglas (1972) and Voytov et al. (1975). Peak concentrations of methane and ethylene exceeded 25,000 ppm by volume and 8 ppm by volume, respectively. Ethane evolution would be masked on the chromatographic trace by these quantities of methane, but ethane above background levels was not observed after seven days of incubation. No measurable C2-C4 gases were found in the remaining 30-day incubation period. This experiment suggests that biogenically-generated gases do not initially contain any C2-Ca hydrocarbons.
Light hydrocarbons for petroleum and gas prospecting
13 9
Fig. 5-5. Anaerobic microbial evolution of CH4 upon decomposition of various organic substrates.
Another approach was that taken by Coleman (1979), who studied both the chemical and isotopic composition of glacial till gases in Illinois. Coleman obtained the same result, finding that C2-C4 gases are not present in the glacial till gases in Illinois. Coleman also determined 14C age dates on the gases and showed that the biogenic methane varied from about 10,000 years to as much as 40,000 years in age. This is particularly significant since it suggests that no bacterial generation of C2-C4 hydrocarbons occurs, either initially in test tubes or even within the first 40,000 years in glacial till.
TABLE 5-11 Generation of C1-C4 hydrocarbons in vitro (average concentrations, ppm) 36~ Gas Methane Propane i-butane n-butane Ethylene Propylene
Inoculate 14,862.000 0.081 0.022 0.055 5.101 0.086
Control 3.416 0.073 0.022 0.060 1.670 0.089
25~ Inoculate 4,814.000 0.051 0.014 0.038 2.310 0.057
Control 2.003 0.050 0.009 0.033 0.434 0.045
140
V.T. Jones, M.D. Matthews and D.M. Richers
Distinguishing petrogenic and biogenic hydrocarbons The results of the studies by Janezic (1979) and Coleman (1979) strongly suggest that C2-C4 hydrocarbons are not generated biogenically. Most of the previous studies cited appear to be compromised because they were conducted in natural environments in which migrated petrogenic gases might have also been present. Even assuming that small quantities of C2-C4 gases are generated in biological environments, a methane:ethane ratio greater than 500 appears sufficient to delineate anaerobic gas production from thermocatalytic gases, since such ratios do not occur in petrogenic natural-gas deposits. As shown in the test-tube experiments (Table 5-11 and Fig. 5-5), this value is achieved within two to three days for all substrates studied and exceeds 100,000 after seven days of incubation. Similar values are cited in the literature (Frank et al., 1970; Swinnerton and Lamontagne, 1974; Bernard et al., 1976; Sackett, 1977; Reitsema et al., 1978) as the biogenic threshold in marine geochemical prospecting (Table 5-11I).
HISTORY The first attempt to relate soil-gas hydrocarbon concentrations to oil and gas deposits was made in 1929 in Germany by Laubmeyer (1933). Surveying a known oil deposit, he collected samples of soil gas from systematically-located boreholes 1-2 m deep, after sealing them from the atmosphere for 24-48 hour periods. Using portable analytical equipment, he demonstrated that the samples over the deposits were enriched in methane. Soil-gas investigations were initiated shortly after this time in the then Soviet Union by Sokolov (1933), who verified Laubmeyer's results (Kartsev et al., 1959), but measured both methane and heavier hydrocarbons. Research in the area of surface prospecting was also carried out in the United States during the 1930s beginning with Teplitz and Rodgers (1935), Rosaire (1938) and Horvitz (1939). These investigations entailed the collection and analysis of the soils
TABLE 5-III Literature review of methane/ethane ratios diagnostic of biogenic origin Reference Davis and Squires, 1954 Reitsema et al., 1978 Frank et al., 1970 Bernard et al., 1976 Sackett, 1977
Diagnostic of biogenic origin C1/C2> 500 C~/C2> 500 C1/(C2+C3) > 1000 CI/(C2+C3) > 1000; 8C 13-50% C1/(C2+C3) > 1000; 8C 13-50%
Light hydrocarbonsfor petroleum and gas prospecting
141
themselves for hydrocarbon gases. The use of adsorbed gas on soils was regarded as an important improvement upon soil gas, as short-term diurnal variations in soil-gas flux could be avoided by the assumption that soil would have a tendency to establish over time a metastable equilibrium with the regional flux.
Basic concepts In the years following these early studies, the basic concepts have remained largely the same, except that detection limits have been improved with technological advances. Recent work has focused on compositional ratios or signatures of the light hydrocarbon gases and their relationship to known hydrocarbon products in the investigated area (Weismann 1980; Jones and Drozd, 1983). Emphasis has also been placed on the fundamental principles of surface seepage, and the interpretation of the data. It is the opinion of the authors that the overall acceptance of microseep technology in the West has been hindered not only by the emphasis and success of seismic methods but also because of the lack of a comprehensive and public surface geochemistry database. There are, by comparison, more publications on geochemical survey data and basic concepts in the Soviet and Russian literature. As a consequence, many of our discussions rely on experience gained in the private sector in the West, supplemented by literature published in the East. Although the Soviet/Russian literature is clearly positive about surface microseep technology, the Western literature is strongly divided. Debnam (1969) has reviewed several cases crediting geochemical prospecting with petroleum discoveries. Overall success rates are in the range 25-75%. Duchscherer (1980) reports a success rate of 25%, slightly over the industry average, of which 58% are stratigraphic traps. Sealey (1974a, 1974b) reported a success rate of 80% in Texas using a microbiological technique.
Methods of geochemical prospecting Geochemical methods of prospecting are classified as direct or indirect. The direct methods involve detecting the presence of dispersed oil components in the form of hydrocarbon gases or bitumens in the soils, waters or rocks in the vicinity of oil and gas accumulations. The indirect methods involve detecting any chemical, physical, or microbiological changes in the soils, waters, rocks or vegetation spatially associated with the oil and gas deposits. Figure 5-6 is a schematic diagram outlining most of the direct and indirect methods currently in use (Kartsev et al., 1959). Identifying secondary responses generated by leakage of hydrocarbons at the surface has merit and has been reported by many investigators. These include the use of (1) soil microbes (Soli, 1954, 1957; Kartsev et al., 1959; Sealey, 1974a; Sealey, 1974b); (2) reduction effects (Pirson et al., 1969; Donovan, 1974; Ferguson, 1975); (3) carbon and
142
V.T. Jones, M.D. Matthews and D.M. Richers ,,,,,
Geochemical Methods of Prospecting and Exploration I
I
I
! I Indi,ootl ! I i , I :i_1Soil-Salt H;drochemical J Microbiological!
[ Direct t I
--~Free soil 'gases.] ~ Fluorescence] -~ Gas logging ]
oroox --. I
~ Marinesurvey] method Surfacecore [ w to, wo. -q.o no I Deepcores I UI seismic shotholes __~ Chloride] [-GypsumI woH,og g 1 Formationbrines [
....
-~ Waterclassificationi Fig. 5-6. Geochemical methods of prospecting for petroleum and natural gas (reproduced with permission from Kartsev et ai., 1959, Geochemical Methods of Prospecting and Exploration for Petroleum and Natural Gas, copyright by the University of California Press).
oxygen isotopes (Donovan et al., 1974); and many other effects as reviewed by Matthews (1985). As an exploration tool, the identification of hydrocarbon seeps is particularly useful when coupled with remotely-sensed images and photographs. Case studies by researchers in the West have shown that secondary indicators of microseepage are often present in the near-surface environment. Examples noted by Horvitz (1972), Donovan (1974), Donovan and Dalziel (1977), Matthews (1985) and Ferguson (1975) have indicated the presence of diagenetic alteration of soils above or adjacent to hydrocarbon accumulations. Work by Rock (1985), Matthews et al. (1984) and Patton and Manwaring (1984) has shown that these effects may often be reflected in the health and type of vegetation over the seep, which also alters the spectral response detected by satellite and airborne sensors. These methods of geochemical prospecting for oil and gas are reviewed in more detail in Chapter 7. Others have noted changes in resistivity or radioactive signatures above accumulations due to the seepage and possible interaction of ascending fluids and solutions with the encapsulating medium. In some cases the actual removal or addition of soluble chemical species has been noted.
Light hydrocarbonsfor petroleum and gas prospecting
143
It appears therefore that the direct detection of hydrocarbon gases is not the only means of identifying areas of active microseepage, but that a myriad of other possible secondary techniques can be used either as adjuncts, or as solitary techniques in themselves, to infer the presence of hydrocarbons in the subsurface environment. Most of these utilise the detection and subsequent analysis of gaseous hydrocarbons, while other methods employ the detection and analysis of liquid hydrocarbons, nonhydrocarbon gases, the presence and relative concentration of bacteria, and even the presence (or absence) of inorganic compounds and elements. For the most part, however, methods that directly measure the hydrocarbon content of soils or soil atmospheres have met with the most acceptance.
PHYSICAL BASIS FOR MIGRATION OF HYDROCARBONS TO THE SURFACE
Basic assumptions The fundamental assumption of near-surface hydrocarbon prospecting techniques is that thermogenic hydrocarbons generated and trapped at depth leak in varying quantities towards the surface of the Earth. That these hydrocarbons present in the near-surface environment represent the products of generation and migration from subsurface points of origin is a necessary conclusion that is universally accepted with respect to hydrocarbon macroseepage. Examples abound, such as the Santa Barbara Channel seeps, the La Brea Tar pits of Los Angeles, the Athabasca Tar Sands, etc. The same relationship has been equally well established, although less commonly accepted, for microseepage. A further assumption is that the pattern and intensity of this leakage also provides information on preferential pathways that the leakage follows, and as such can be combined with additional geologic information to predict broad subsurface hydrocarbon fairways. In fact, in some instances it has been claimed that such data can identify areas of reservoired hydrocarbons. This last claim is often the subject of heated debate, however, commonly depending in which camp (for or against geochemistry) the explorationist resides. The physical state of the hydrocarbons during transport is not well known; see Matthews (1996a) and Matthews (1996b) for a full discussion. Nevertheless, most of the models proposed for the transport of these fluids from source to reservoir (aqueous transport, micellular, discrete oil-phase transport, gaseous transport, etc.) are applicable to the continued transport of hydrocarbons from these source beds and/or reservoirs to the near-surface environment. An additional constraint on land is that the last stage of transport is generally above the water table. The physics of transport can be subdivided into two categories, effusion and diffusion.
144
v.T. Jones, M.D. Matthews and D.M. Richers
Physical transportation by effusion Effusion transport is believed to be the dominant mode of moving hydrocarbons to the reservoir and to the near-surface environment. The sharp localised nature of many anomalies associated with microseepage and macroseepage is more consistent with an effusion model rather than a diffusion model. The experience of the authors in monitoring leakage from gas storage reservoirs and controlled experiments where subsurface gas pressures were typical of true reservoirs suggests vertical transport rates of several metres (tens of feet) per day, clearly greater than the distances of migration dictated by the diffusion mechanism alone (Jones and Thune, 1982). The sharp and often linear nature of anomalies suggests that faults and fractures play an important part in the movement of these gases. Major linear features discernible on satellite images, as well as other remotely-sensed media, from Patrick Draw, Wyoming, show such a relationship (Richers et al., 1982). The Lost River, West Virginia, Geosat study (Matthews et al., 1984) shows anomalously-high soil-gas values in relation to linear features on imagery. There are anomalously-high gas values along faults in the San Joaquin Basin and in the Wyoming-Utah Overthrust Belt (Jones and Drozd, 1983). The Russians have shown that the magnitude of soil-gas values on faults increases dramatically shortly after an earthquake in which fault movement is involved (Zorkin et al., 1977). An extensive study, involving 105 observation wells, 3-5 m deep, was set up over the Mulchto oilfield in northeastern Salchalin. A total of 3,700 samples was collected and analysed over a four-month period with the most active wells sampled daily (Table 5-IV). The results from this study provide impressive evidence for the tectonic relationship of this leakage gas flux (Fig. 5-7). This study leaves no doubt that faults and fractures provide the main control on the effusion of gases from the subsurface.
Physical transportation by diffusion Diffusion, on the other hand, is a slow and widely-dispersive process. Antonov et al. (1971) measured hydrocarbon diffusion coefficients for a variety of rock types from several hydrocarbon provinces in the former USSR. They discovered that the coefficients of diffusion vary over a wide range (10-3-10 -s cm2/s) depending on the particular lithology and geologic conditions. The time required for diffusion to occur can sometimes be restrictive. Indeed the time required not only often exceeds the age of the hydrocarbon accumulation but also quite often exceeds the age of the host rock. If this were the dominant process for migration, then the appearance of soil-gas anomalies in the near subsurface would indicate only very shallow accumulations. If a non-steady state exists, where the hydrocarbon signal observed represents only 0.001 times the steady-state signal, then diffusion times could be reduced by a factor of 25 compared to that of the steady-state model. Table 5-V
145
Light hydrocarbons for petroleum and gas prospecting
TABLE 5-IV Gas concentrations in the near-surface rocks before earthquake and (in italics) after earthquake Date
09.09.1974 08.04.1975 24.05.1975
Strength of shock
Distance Well Time of from No. sampling CPI days* center to deposit (km)
K=9 M=4 K=6.2
100 12 25
8 8 8
104 Vol percent (ppm)
Vol percent
Percentage of hydrocarbon fraction of gas
6
135.40
1.90
0.00
98.56
3
283.70
4.20
0.00
98.50
2
73.60
0.81
0.53
98.35
4
213.50
2.34
1.22
98.96
98.50
2
188.50
3.18
3. ! 0
1
525.00
2.52
7.10
99.50
1
152.00
7.36
17.80
95.50
04.07.1975
K=7.2
9
8
08.07.1975
K=7.5
25
11
05.10.1975
K=9.5
100
8
1
100
61
5 !
273000.00
:g
10-4 Vol percent (ppm)
2
852.00
8.85
15.70
98.97
5
935000.00
11908.70
0.09
98.70
2
954000.00
12465.00
0.26
98.70
6
58.80
3.80
26.80
93.90
396.00
5.80
33.60
98.60
256000.00
1273.00
3.60
99.54
1399.00
4.20
99.54
From the onset of shock
shows some of the times that this scenario would require. However, diffusion can still be considered as a potential secondary process in microseepage. Sokolov (1965) calculated diffusion to be sufficient to have resulted in the dissipation of oil fields formed in the Palaeozoic, although to what extent, if any, this has occurred is not known. Furthermore, if any such fields had leakage along faults and fractures or due to erosion of the seal, diffusion might not be able to bring about accumulation before much faster effusive loss caused depletion. Diffusion of benzene into brines adjacent to accumulations has been demonstrated and used as an exploration tool by Zarella et al. (1967). In productive basins the process of diffusion from both source rocks and reservoirs may be responsible for observed elevated background concentrations that have no apparent relationship to the known accumulations. Alternatively, the presence of free hydrocarbons effusing outward and upward in areas of microfractures and dispersed by groundwater flow could similarly account for this background. If diffusion were the
146
V.T. Jones, M.D. Matthews and D.M. Richers
TABLE 5-V Hydrocarbon diffusion times (minimum years) through sediments of different thickness (Antonov, 1971) Diffusion coefficient (crn/sec) 5 x 10-5 1 x 10.5 5 x 10-6 1 x 10.6 5 x 10.7 1 x 10.7 5 x 10.8 1 x 10-8
1000 m 4.9 24 49 244 488 2440 4880 24400
Steady state 2000 m 20 98 195 976 1950 9760 19500 97600
3000 m 44 220 440 2200 4400 22000 44000 224000
1000 m 0.2 0.9 1.8 9.8 18 90 180 900
Non-steady state 2000 m 3000 m 0.7 1.6 3.6 8. I 7.2 16 36 81 72 162 360 810 720 1620 3600 8100
responsible mechanism, then one might expect broad anomalous zones, with localised effusive "spikes" superimposed on the background. Starobinetz (1983) listed as typical examples of diffusion the studies of Driepro-Douetsk and Anuddria grabens. Aside from the potential of diffusion for producing a broad dispersive background, it would also be expected to alter the composition of the gases detected in surface methods. Starobinetz (1983) notes that not only can diffusion affect composition, but two additional processes have a similar effect. These are chromatographic separation and selective adsorption. An example of such chromatographic separation is shown in Fig. 5-8 (Sokolov, 197 lb), which shows the results of a mixture of methane and benzene injected into the bottom of a hand-bored 6-metre deep well. Samples of subsoil air were taken periodically from observation wells 1-2 m deep, resulting in the obvious separation shown in Fig. 5-8. Indeed these processes have been cited by detractors of surface prospecting as evidence that the technique is not a valid means of searching for subsurface hydrocarbon deposits, arguing that pulses (non-steady state) of gases will have a different composition from their source because of the chromatographic separation. The example shown in Fig. 5-9, taken from an artificial underground coal gasification experiment near Rawlins, Wyoming (Jones and Thune, 1982), shows that such effects are only temporary. In this experiment, a pulse of gas travels from a retort at a depth of 180 m (600 feet) and migrates vertically and laterally to a series of observation wells 5.5 m (18 feet) deep. As shown in Fig. 5-9, although the first gas to be seen in high concentrations is methane, the compositional separation does not last more than a few days before equilibrium is achieved, when all the migrating gases have ultimately reached the surface.
Light hydrocarbonsfor petroleum and gas prospecting
147 Key CH4 ZHC
CH 4 Y'HC
vol%
C H 4 ] ~ H C 10 -4 v o i %
30 I - o.3o I-26 22
60 40 20
0.25
1000
).20 D.15
600 200
).10
.-,--,- ,..-,"
i
_
I'"I::-,-:"~ ,;: Ir
12 4
J CH 4
CH 4
c.,
Well
o,
/
o.4
shY,
~f
I
Z.c~0 "4 vo. % $ ~ ..
~
30
/~ 1000
f-
!
I
I
!
CH 4
y"H C 10 .4
t
t
I
I
I
!
~o
I 60
l /
5
~J~
31- t
t -Fi
10
20
September
! I
I
". I
I
I
I /
K-T
j
,..';.i
/
, ,..., ,
w.,
/I ~
~rIt
I
!
I!ll
" ~I I" ' -I
!
I
I
I
!
I
C H 4 T ' H C 10 .4 v o l %
vol %
W e l l 104
i i t t
10 8
100
I
/
140
"" -"
L-
800 600 400 200
1. I " - ~ ' I ~
i
c . z . c ~ 0 4vo, % ,,o
10"40[
i
Y ' H C l 0 "4 v o l %
0.5
i
it
i
!
30
5
i
t 15
October
i
5
0.4
4 3 2
0.3
1 J
Well 79
I
0.2 I i i i I I I I 27 31 3 16 25 August September
I
5
I
15
October
Fig. 5-7. Methane and total hydrocarbon gases in subsoil before and after an earthquake (reproduced with permission from Zorkin et al., 1977).
As to the second point, if selective adsorption is occurring, the volumes of material escaping over geologic time should ultimately saturate (poison) the adsorber such that no additional material can be adsorbed, or at best, material is exchanged in a steady-state. The result will be a gradual return of the signal to the original composition. This is clearly shown in a study by Zorkin (1977a). There is, however, one important area where diffusion may be responsible for compositional changes; near the soil-air interface. Methane should, due to its lightness and zero net dipole moment, be preferentially lost (followed perhaps by ethane). This would possibly result in an oilier gas signal at the surface. This could be countered by the production ofbiogenic methane, which might partially compensate for this loss.
148
V.T. Jones, M.D. Matthews and D.M. Richers
0.1
0
I
0
5
.......
i
i
10
15
Time (days)
Fig. 5-8. Differentiation of methane (1) and heavy hydrocarbons (2) during migration from an artificial source (from Sokolov, 1971b).
10 5_
E I~.
10 4 -
~'IO
3Cl
,,
C2 Ioo.......--'- . . . . I
C3
,~176176176
oO o o o ~
C4
10/19/81
11118181
.::.-'.;'
10 0
ld 1 .
u
07/21/81
08120/81
09/19181
I
I
12/18/81
I
12131/81
DATE
Fig. 5-9. Arrival at a surface well of hydrocarbon gases following a subsurface coal-bum experiment at Rawlins, Wyoming (reproduced with permission of the Society of Petroleum Engineers from Jones and Thune, 1982, Surface detection of retort gases from an underground coal gasification reactor in steeply dipping beds near Rawlins, Wyoming, SPE 11050).
HYDROCARBON RESIDENCE SITES AT SURFACE The most important of the direct techniques shown in Fig. 5-6 involve the measurement of light hydrocarbons, methane through butane. Because of their volatility, these light hydrocarbons are generally found in the free pore space. The seepage of
Light hydrocarbonsfor petroleum and gas prospecting
149
hydrocarbons into the near-surface environment above the water table must involve transport through both water-filled and air-filled pores. Sampling these pore gases is obviously one of the most fundamental concepts. However, gases can be bound in the sediment matrix. This latter possibility leads to the development of some disaggregation and desorption extraction techniques. Discussion of sampling techniques must involve both "flee" and "bound" gases. To facilitate this discussion the collection, measurement and analysis of light (C1-C4) hydrocarbons will be broken into two main categories each with two subcategories: (1) free gas, which can be vapour or dissolved gas; and (2) bound gas, which can be adsorbed gas or chemi-adsorbed gas.
Free gas Gases in the free pore space can be found either in the vapour state or dissolved in water. Extensive research at Gulf Research and Development Company has demonstrated that the "free" and "dissolved" gas seeps yield comparable compositional results, both to one another and to their associated reservoirs when they are properly collected and analysed (Teplitz and Rodgers, 1935; Jones, 1979; Janezic, 1979; Mousseau and Williams, 1979; Weismann, 1980; Drozd et al., 1981; Williams et al., 1981; Jones and Drozd, 1983; Richers, 1984; Price and Heatherington, 1984; Matthews et al., 1984; Jones et al., 1984). This documentation even extends to numerous observations over artificial underground gas generation and storage reservoirs (Jones and Thune, 1982; Jones, 1983; Pirkle and Drozd, 1984). Sampling of vapour can be extended to any depth above the water table by analysing the exhaust air from an air-drilled well. Complications occur because of dilution effects by the air injected for drilling and by the additional fact that the drill bit disaggregates and liberates rock or matrix gas in the process of drilling the hole. Dissolved gases must be extracted from the aqueous system before analysis. This is usually accomplished by a simple gas-water partition into a vapour phase followed by standard headspace measurement techniques (McAuliffe, 1966). Alternatively a socalled "stripper" continuously partitions the dissolved gases into a carrier gas which is then sent to a gas chromatograph for analysis (Mousseau and Williams, 1979; Aldridge and Jones, 1987). These separations are aided by the very low solubility of the light hydrocarbon gases. Standard mud gas logging is one variant of dissolved gas analysis conducted on deeper drill holes. A gas trap is deployed in the return mud system for extracting the dissolved and free gases. Compositional information obtained from mud logging gas is useful for predicting the composition of a potential reservoir (Pixler, 1969). These same ratios have been found to be indicative of oil versus gas potential from surface seeps observed from 4 m (12 feet) deep soil-gas measurements or from analysis of gases dissolved in the shallow groundwater (Jones and Drozd, 1983).
150
v.T. Jones, M.D. Matthews and D.M. Richers
Bound gas Bound gas, which is adsorbed on both the organic and inorganic matter contained in the sediment by means of physicochemical binding, introduces new complexities into defining the appropriate sample for analysis. The difficulty with defining this bound gas is forced by the reality that rocks and/or sediments contain gases of multiple origins. By their very nature, sediments contain both migratory (epigenetic) and indigenous (syngenetic) gases. Migratory gases (biogenic and thermogenic) have migrated to the surface from a deeper, more concentrated source. Indigenous gas is related to biogenic, diagenetic and thermogenic generation within the rock sampled at the surface and to recycled materials which may contain some physically-transported hydrocarbons tightly bound in inclusions or other interstitial sites within the sediment matrix. The nature of the bonding of the hydrocarbons to the grain surfaces leads to two categories, adsorbed and chemi-adsorbed. These form an important part of this discussion because of misnomers involved with the use of the word "adsorbed". True adsorbed gases are by definition bound to the surfaces of sediment or rock particles. As defined by Greenland (1981) adsorption is the process by which a chemical species passes from one bulk phase to the surface of another, where it accumulates without penetrating the structure of the second phase. Because the light hydrocarbons are so labile, they do not strongly adhere to surfaces and are easily desorbed if the source of these gases is removed. The gas must be replenished by continuous migration in order to maintain the presence of adsorbed gases on the available surfaces. Bound within the rock matrix, or within certain minerals (calcite, oxide coatings, etc.) gases are chemi-absorbed. They can be removed only by a chemical attack that completely dissolves the rock or sediment matrix. Sometimes these more tightly-bound gases not only include indigenous gases, but also might integrate the signal over time, mixing the products of "dead" or "non-active" seepage with those gases actively migrating today. The non-active seeps are often coupled to the lithologies of transported, non-residual sediments (Richers et al., 1986). These last considerations provide two of the main reasons why "free" and "chemi-adsorbed" gases are often found to have no obvious spatial correlation.
Choice o f f r e e gas or bound gas Any prospector would generally agree that it is desirable to measure only the gas which has migrated from depth, since this is clearly the gas signal which is related to buried reservoirs. The difficulty in doing this begins with choosing the method of sample collection, because there are few sample-collection techniques that do not mix the syngenetic and epigenetic gases. Both "free" and "adsorbed" hydrocarbons can often be related to a migratory source, and thus can yield useful exploration information. The free
Light hydrocarbons for petroleum and gas prospecting
151
gases appear to be dominated by the migratory gases, unless samples are taken within an outcropping source rock. In addition, the free gases also contain any biological gases which, because of their recent generation, also occur in the free state. If source rocks or recycled source-rock materials are present near surface, then the "adsorbed" gases can obtain a major contribution from these sources. Exclusions are often provided by sampling in areas where calcite concretions have been deposited from carbon dioxide generated by biological oxidation of seepage hydrocarbons. This is one reason why adsorbed gas has been successful in marine offshore environments. A good example is provided by studies of the Green Canyon macroseeps (Anderson et al., 1983; Pirkle, 1985). If one can assure that only migratory gas is measured, then the type of gas measured is unimportant. Including indigenous (syngenetic) gas results in misleading measurements. This is believed by the authors to be one of the primary causes of failure in the application of surface geochemical prospecting. Failure to collect a properlydistributed data set can be equally misleading and result in an incorrect interpretation, since interpretations will always be the educated guesses of an explorationist. Any measurement on a real-world sample is always a combination of the free and bound gas sample types. This is because the process of taking the gas sample generally requires that the sediment or rock system is disturbed by some mechanical means which creates the mixing of these sample types. Because of this unavoidable interaction, we have recognised the need to consider an intermediate sample-collection technique that measures the more loosely-bound gases liberated into a container containing the core sample. Sampling gases that accumulate within the gas-filled "headspace" of a core-sample container is potentially flawed because of the obvious losses encountered in transferring a sample to a container. This is further compounded by the difficulty in achieving a rapid and total equilibration of the core gases into the headspace. An alternative technique for measuring the loosely-sorbed gas has been proposed by Hunt and Whelan (1979), in which the headspace equilibrium is obtained mainly by mechanical disaggregation and heat. In our opinion, this disaggregated gas should more properly be called "adsorbed" gas. The truly "free" gas is always lost (or at least greatly diminished in volume) from any sample of core that is brought to the surface for collection and handled before being put into a sample container (Sokolov, 1971 b). Typical losses are shown in Table 5-VI. This mechanically-disaggregated gas has been usefully applied as a bridge to relating the free and bound gas (Richers et al., 1986). Simple mechanical disaggregation always liberates a considerable volume of gas which, if handled properly, has a predictably oilier composition than the associated free gas. This change in composition, created by fractionation of the lighter components, is demonstrated later in examples under case studies.
V.T. Jones, M.D. Matthews and D.M. Richers
152 TABLE 5-VI
Generation of C1-C4 hydrocarbons in vitro (average concentrations, ppm) Rock type
Depth (m)
Sandstone Shale Shale Sandstone
385 575 620 640
Hydrocarbonconcentration, C (10-4 cm3/kg) Sealed kc sampler Open-holesampler 106243 119 2431 52 1610 35 36473 69
Relative loss, Ckr / Co~n 893 47 46 529
FACTORS INFLUENCING NEAR- SURFACE HYDROCARBON FLUX The hydrocarbon flux near to the surface varies according to the supply of hydroca~rbons and whether local chemical and biological conditions favour their preservation or breakdown. In addition, hydrocarbon magnitudes at any given location vary with time because of displacement by wind, rain and barometric pumping (Wyatt et al., 1995).
M i c r o b i a l activity
In a very extensive review, Price (1985) suggested that surface bacterial activity can totally obliterate the gases in a microseep. That this is not typically the case has been demonstrated by extensive research over both macroseeps and microseeps (Jones, 1984). However, bacterial activity does probably contribute to the noisy appearance of soil-gas seepage.
Barometric pumping
An example of gas flux related to barometric pumping has been demonstrated over an underground propane-storage reservoir. This mined cavern is about 60 metres (200 feet) deep. In order to observe the gas flux related to atmospheric phenomena, plastic ground sheets about 1.5 x 1.5 m (5 x 5 feet) were buried along their edges to contain any gas flux. The variation with rainfall is shown as vertical bars in Fig. 5-10. A very large seepage anomaly is shown by the dashed line at the right edge of the first bar. The rain probably displaced the gas in the ground and caused it to come up underneath the ground sheet. However, the same effect is not repeated every time it rains. Around the 19th, 20th, 21st and 22nd days of the month very small barometric changes were observed. Nevertheless, small barometric lows have clearly-expressed gas-flux increases. Thus falls in barometric pressure lead to a gas flux that escapes into the atmosphere. This
Light hydrocarbonsfor petroleum and gas prospecting
NlmW
1,4" RAIN.
2 ~ RAIN
N
\
g I~o
~ m m
1 1'112
,.
RAIN
2 , 3 " RAIN. 5 ~ RAIN.
MI|
W N
J
_, ..-..-N"
I 13 1 IAr 1 Is I I~ 1 1T
~8
153
N N
"
Ig12012rl
',e
4 . 8 " RAIN
/ ,//<. N
2zi 2312412s12~12r
128
Fig. 5-10. Changes in flux of propane concentration (dashed lines) with barometric pressure (solid line) and rainfall (shaded bars) at two surface sample sites over an underground propane-storage reservoir; horizontal scale shows days of month.
escape occurs despite the extensive microbiological activity that has developed over this cavem. As shown in Fig. 5-11, a propane profile collected over the top of the cavem requires a log scale to illustrate the enormous range in gas leakage flux. An interesting secondary observation taken from this example is the obvious colour changes noted on the soil cores. These chemical changes are related to hydrocarbon seepage and might be used as an additional exploration tool to provide evidence of where the gas leakage has occurred around any type of storage cavern. The soil changes from red-brown to green-black directly over the top of the cavern, where the largest seepage anomalies occur. Thus the main difficulty with atmospheric sampling is created by meteorological changes which can greatly displace and dilute the seepage emissions. In addition, it is clear that the stress fields in the Earth can also influence this gas flux significantly.
Earthquakes The fact that earthquakes may sometimes be preceded by geochemical anomalies was discovered at about the same time in Japan (Okabe, 1956) and the then USSR (Fursov, 1968). Earthquake prediction studies in Russia, Japan and China include extensive geochemical measurements. Chinese geochemical data are reported to have contributed, at least partly, to the successful prediction of several strong earthquakes (Allen et al., 1975). In contrast, the Earthquake Hazards Reduction Program in the United States emphasises mainly geophysical data.
154
V.T. Jones, M.D. Matthews and D.M. Richers
PPM 10 6
..... "~
~"
/
~
.,.
"
, :-.
U .
!
~ .t
~
[]
. ,
. .
. .
.
'~
.
.
.
'1o: 10 3
',,, _
-
~o 2
-
.,nl
48O l
~I~
.
.
.
.
.
.
.
.
475
.
.
.
.
.
.
.
.
an..~. illlH~
465
,so . 45s 450
,,
. .
47,o
-"" gllg_.~__iUg , .... ..agl
.
. .
,_
,
| H U | ~ . _
. .
.
[] r
....
.
. .
.
\ ~
9= = ! @ l l l t . m
"
[]|
u .
A
' ,
'
I
"--7-
T
.......
~
~..
f " ~
. . ,,
.
445 .
.
ltl
u ,
440 435 430 425
H o r i z o n t a l s c a l e 20 ' Vertical s c a l e 5'
Red-Brown Green-Black
Fig. 5-1 I. Propane concentrations and soil colour of augered samples along a traverse of sample sites over an underground propane storage reservoir.
Limited programmes using radon for earthquake prediction began in the United States about 1975, at about the same time as Gulf Research and Development Company first made measurements on light hydrocarbons, helium and hydrogen on the San Andreas Fault in the Cholame Valley in Califomia (Jones and Drozd, 1983). This study confirmed that helium is a deep-basement or tectonic indicator which is commonly independent of oil and gas deposits. This is clearly illustrated in Fig. 5-12, in which helium anomalies appear to be associated with the San Andreas fault and with two other deep-basement faults. The proposed deep fault on the west flank of the Lost Hills oil field also acts as a common migration pathway for hydrocarbon gases (Fig. 5-13). This initial study, and the joint research programme subsequently initiated by Gulf Research and Development Co. with the Cal-Tech earthquake radon programme, was designed to obtain data concerning the rates of change of gas flux associated with tectonic stress in the Earth. Numerous other examples of gas flux related to earthquakes have been reported, for example, by Kartsev et al. (1959), Fursov et al. (1968), Elinson et al. (1970), Sokolov (1971b), Eremeev (1972) Ovchinnikov et al. (1972), Zorkin (1977b), Melvin et al. (1978), Wakita et al. (1978, 1980), Barsukov et al. (1979), Borodzich et al. (1979),
Light hydrocarbonsfor petroleum and gas prospecting
155
Fig. 5-12. Relation of near-surface gases to deep faults and oil fields along a traverse in the Cholame Valley, California (reproduced with permission of the American Association of Petroleum Geologists, whose permission is required for future use, from Jones and Drozd, 1983, AAPG Bull., vol. 67, no. 6, Fig. 12, p. 942, AAPG 9 Mamyrin (1979), King (1980b), Reimer (1980), Shapiro et al. (1981, 1982), Mooney (1982) and Pirkle and Jones (1983). Particularly intriguing examples have been published by Antropov (1981) of atmospheric methane flux related to petroleum deposits (Fig.5-14) and seismic shock (Figs. 5-15 and 5-16). These measurements were made with adsorption-type gas lasers: one type makes point measurements of the sample in an adsorption tube (Iskatel-2); the other (Luch) measures the specific gas adsorption along a path length (1-100 m).
SAMPLING AND MEASUREMENT METHODS There are a variety of sample collection and hydrocarbon analysis methods use in geochemical surveys for oil and gas deposits. In the case of free gas, samples are collected either in the atmosphere or, more usually, within the soil. For bound gas soil or rock is collected and the gas is liberated by one of several methods. In practice, however, it is rarely possible to determine solely free gas or solely bound gas.
Atmospheric techniques The detection of hydrocarbons above the ground surface offers obvious advantages: continuous sampling, no permit requirements, access over rough and hostile
156
V.T. Jones, M.D. Matthews and D.M. Richers
200
--
Methane
Hydrogen Propane
22--
...-. D.
z
T 12oo
LU
a>-r
I A
~
I
~
I
SW
~
O. W
~"
z
Z ~
"1-
NE
Feet Metem
SEA
1000
,.--
-
9
ooo-i_o/, ,-,,..,,, E R , . , , . , , 2100
9O00
Fig. 5-13. Relation of near-surface gases to proposed deep fault adjacent to Lost Hills oil field, San Joaquin Basin, California (reproduced with permission of the American Association of Petroleum Geologists, whose permission is required for future use, from Jones and Drozd, 1983, AAPG Bull., vol. 67, no. 6, Fig. 13, p. 943, AAPG 9
environments, large areas covered rapidly. A drawback is that diffusive and convecting mixing in the atmosphere decreases the signal strength with distance from the sediment or soil surface. Nevertheless, the capability of detecting gases in the atmosphere has seen significant developments over the past 10-15 years. Research has resulted in the development of approaches based on microwave energy, infrared lasers and adsorbed hydrocarbons on aerosols carried into the atmosphere by thermals. The microwave approach has been developed by Owen (1972), Goumay (1979) and Thompson (1981). Although Thompson (198 l) has stated that "conclusive proof does
Light hydrocarbonsfor petroleum and gas prospecting 4
15 7
-
!
0
=
3
< 7-
2
-
W
I
I
0
IOO0
20OO
Meters
Fig. 5-14. Variations in methane concentration in air above a petroleum reservoir (from Antropov, 1981).
C H 4 , x 10"4% 6-
4-
O O
~ 0
I I I
~.~,,.~ ~
2-
0 i .... 10
i..I...l 12
.....j
0
to o
I
14
,1 16
I
18
_o/ .1
20
.
f
---
22
T ~ e , hr Fig. 5-15. Variations in methane concentration in air as a result of seismic shock to the ground (from Antropov, 198 ! ).
not exist that the gases being detected by the sensor are low molecular-weight hydrocarbons and nothing else", he has published numerous positive case studies relating the response of one of these instruments to soil gas probe anomalies (Burson and Thompson, 1985). Additional technical difficulties result from the fact that microwave adsorption energy levels represent rotational energy in the molecule. Deactivation of rotational energy by collisions can occur rapidly at atmospheric pressure, causing the molecule excited by the microwave energy to lose its adsorbed energy in a non-emission mode, thus reducing the signal-to-noise ratio. This coupled with the low concentrations of hydrocarbons in the atmosphere has meant that the technique has not been extensively tested as an exploration tool.
158
V.T. Jones, M.D. Matthews and D.M. Richers
CH4xl 0-4 %
3f .................... 0
8L
m -5o1-
6V
8V
......... ',
10V
12V
,, ,,h, '
I'~V
'!
16V
18V
20V
22V
2,~V
26V
28V Date
t
l"
'
,,I , l ' l t
"
Fig. 5-16. Variations in methane concentration in air with crustal displacement in a seismicallyactive region (from Antropov et al., 1981).
Remote monitoring of the gas composition of the atmosphere with laser sources has been actively pursued for over a decade, with systems actually built and used for nitrogen dioxide, sulphur dioxide, ozone, carbon dioxide, ethylene, ammonia, hydrazine, hydrogen fluoride and methane. A small mobile laser system capable of measuring methane and ethane in the atmosphere has been developed (at Stanford Research Institute for the Gas Research Institute) for detection of natural gas pipeline leaks (Van de Laan et al., 1985). Another laser technique, based on established physical principles, is LIDAR, which stands for light detection and ranging. The technique uses light from a tuneable infrared monitored. The development of an airborne or truck-mounted system CO2 laser to selectively detect methane and heavier gases by adsorption. The technology was reviewed by Grant and Menzies (1983). Briefly, laser light is pulsed into the atmosphere and aerosols, liquid droplets and gaseous molecules scatter or adsorb the light in different ways. Some portion of the scattered light returns to its point of origin, where a telescope-like receiver channels it to a photodetector, which produces an electrical signal proportional to the optical radiation received by the telescope. The length of time between transmission and reception indicates from what distance the light was scattered and the intensity of the electrical signal indicates the concentration of the particles or molecules being capable of range resolving the location and concentrations of an atmospheric gas cloud will provide an extremely efficient and cost-effective exploration tool for detecting both macroseeps and microseeps in frontier regions. The third atmospheric technique analyses the residual liquid and/or condensate hydrocarbon traces on aerosols carried into the atmosphere by thermals (Barringer, 1981). The aerosols are created by gas bubbles which exsolve into the atmosphere from the sea in areas where microseeps create gas bubbles which reach the sea surface. The aerosols are concentrated from large volumes of air and collected by an airborne cyclone sampler carried aboard an aircraft which is flown at 30 m (100 feet) above the sea surface. Hydrocarbons adsorbed on the aerosols are measured by a flame ionisation detector which yields a total hydrocarbon signal. This system is claimed to produce
Light hydrocarbonsfor petroleum and gas prospecting
159
detector which yields a total hydrocarbon signal. This system is claimed to produce direct vertical anomalies over reservoirs at depth. This technology appears reasonable for detection of seepage which is large enough to produce free gas bubbles, but for feeble seepage (i.e., below water solubility levels) the effectiveness would seem to be reduced by dispersion due to underwater currents.
Soil gas The hydrocarbon gases migrating through soil pore spaces are not dissipated and diluted to the same extent as those in the atmosphere. There are, however, problems posed by the very low levels of hydrocarbon gases and by the diurnal "breathing" of many near-surface soils. In order to overcome these problems, soil-gas techniques which integrate the hydrocarbon signal were introduced by Pirson (1946), Horvitz (1950), Kartsev et al. (1959), Karim (1964), Heemstra et al. (1979), Hickey (1983), Hickey et al. (1983) and Klusman and Voorhees (1983). Karim (1964) published data on laboratory adsorption studies for light hydrocarbons using activated charcoal, molecular sieve (diatomaceous earth) and silica gel. As shown in Table 5-VII, these substrates greatly increase the concentrations available for analysis, but selective adsorption severely affects the relative compositions of the individual gases. The lightest gases are obviously not as effectively trapped by adsorption techniques as are the heavier, less volatile components. This is particularly true for methane and ethane. The adsorption capacities of the substrates are also strongly reduced by moisture content, which may vary from site to site, particularly since the sampling is conducted in the ground where moisture content varies more rapidly than in the atmosphere. Klusman and Voorhees (1983) introduced a variation of this technique which uses sample collection on charcoal wire over extended collection times, followed by analysis using a quadrupole mass spectrometer. The advantages cited are lower field expenses, increased field mobility, improved signal-to-noise ratio and negation of barometric and other meteorological factors. Major drawbacks are that the most mobile light gases are not collected by the charcoal wire, so that the samples comprise mainly the intermediate to heavier molecular-weight components, which include butane through gasoline and diesel. Multivariate statistical techniques are required to interpret the large number of mass peaks recorded, which includes both parent and multiple daughters. In some cases qualitative information based on fragment patterns of the adsorbed compounds is possible (Fig. 5-17). However, different molecular species and their fragment patterns overlap; for example, propane and carbon dioxide have identical masses (44) and thus cannot be separated. The exploration value of these data lies in the demonstrated presence of reservoir-type hydrocarbons at the surface and the composition noted in the lighter to heavier fragment patterns.
160
V. T. Jones, M.D. Matthews and D.M. Richers
TABLE 5-VII Concentrations of hydrocarbons adsorbed by different adsorbents Tube Length Methane Ethane Activated carbon Columbia G 3 in 5 l0 8 in 11 21 12 in 19 36 18 in 29 53 Molecular sieve 4A 3 in 8 40 8 in 12 60 12 in 13 67 18 in 14 71 Molecular sieve 5A-13X 3 in 8 44 8 in 12 67 12 in 15 81 18 in 16 83 Silica Gel 3 in
8 in 12 in 18 in
8
4
33 61 109
17 32 65
Hydrocarbon concentration (ppm) Propane i-butane n-butane Pentane
Total
Ratio to source*
21 43 72 109
30 59 99 143
33 64 108 160
35 71 120 178
134 269 454 672
22.3 44.9 75.7 112.0
52 80 89 93
3 5 6 7
67 100 110 129
73 110 130 152
243 367 415 466
40.5 61.2 69.2 77.7
62 100 127 129
78 120 143 151
82 125 149 150
97 122 146 153
371 546 661 682
61.9 91.0 110.2 113.7
14 60 110 193
11 43 81 141
13 61 109 190
12 50 94 163
62 264 487 861
10.3 40.4 81.2 143.5
* Total adsorbed hydrocarbon concentration / hydrocarbon concentration in source gas
The difficulty in interpreting this particular type of data is further compounded by its application in the upper soil zone where the most active plant and microbiological activity takes place. Many organic and inorganic compounds (humic acids CO2, N20, NO2, etc.) are produced in this zone, all of which are rapidly adsorbed by activated charcoal. These compounds are present in macro concentrations (parts per thousand to percent) and produce fragment patterns which overlap the much lower concentrations of hydrocarbons, which are generally in the ppm range. Another consideration in using adsorbers is the residence time required for the collector in the soil medium. Care must be taken to ensure that the entire survey area is sampled for the same time interval. Also, each region has its own unique flux rate which will affect the results. In a region with a low flux, the collectors should be left buried in the soil for a longer period of time than collectors in a region of higher flux. An orientation survey should always be designed to establish the proper length of time required to obtain valid data prior to conducting a large scale survey.
Light hydrocarbonsfor petroleum and gas prospecting
161
Typical Gas Spectrum o rYo
,;
2'
'
'
:1o
,
,;o
'
1~0
'
M/Z
1~o
'
2~o
'
~o
'
Typical 011 Spectrum
i
!
-
,,ll ....,,Ih t , , ,
, I , .... ,, 140
,~o
'
2~o
'
~0
M/Z
Fig. 5-17. Typical mass spectra of gas and oil.
Although the concept and approach of this technique are excellent, it does not integrate the flux of hydrocarbons heavier than butanes during the one to two weeks for which the collectors are left in the soil. Hydrocarbons heavier than butanes are liquids, and do not migrate more than a few centimetres during the short collection period. It may be equally effective to place a soil sample in a jar with the collection wire; the collection efficiency could probably even be increased by heating the sample jar. Direct sampling of free soil gas requires that a sampling probe be inserted into the ground to collect a soil gas sample. The deeper the penetration, the more difficult and expensive the procedure becomes, eventually requiring that analysis be conducted on drilling fluids or rock samples recovered from a hole. Deeper holes almost always encounter water, which also influences the collection of free gases, forcing one to analyse the gas content of some type of recycled water or mud system which is used to drill the hole. Although sampling from holes of any depth is possible, for simplicity two free soilgas techniques will be discussed and compared (as case studies): shallow probes (Matthews et al., 1984) which penetrate to 1.2 m (4 feet); and auger holes (Jones and Drozd, 1983) which are 3.5 m (12 feet) deep. These methods differ mainly in terms of the resulting soil-gas sample. The shallow-probe samples are influenced more by closer
162
V.T. Jones, M.D. Matthews and D.M. Richers
proximity to the atmosphere and the soil/air interface, where the boundary conditions change. Numerous sample collection methods have been devised for extracting near-surface soil-gas samples. Any suitable mechanical device having a small internal volume can be used to collect the sample. Because the probe sampling port must be forced into the soil, some soil grains are shattered by the necessary mechanical force; many laboratory studies have shown that gas is almost always liberated by this process (Collins, 1983). If the probe volume is very small relative to the dimensions of the sample hole, then the magnitude of the collected sample will be dominated by the gas liberated by crushing. In such cases the volume of available gas will rapidly deplete as the soil gas is aspirated from the hole. This effect can be reduced by collecting a larger volume of soil gas, thereby incorporating a large portion of the natural free soil gas into the sample measured, as compared to that gas liberated by forcing the probe into the ground. One method of collecting gases with a shallow-probe system that has proven to be simple and relatively reliable was developed by Burtell (1988). This probe system consists of separate devices for sampling and for creating the probe hole. The device used to make the hole is a pounder bar 1.2 metre (4 feet) long and 1.3 centimetre (1/2 inch) in diameter, with a sliding hammer that is used to pound the bar into and out of the ground. The soil gas probe consists of a short hollow tube, tightly enclosed by a concentric sealing tube of the same diameter as the pounder bar, which is inserted into the ground through the hole made by the pounder. A hand pump or syringe is used to evacuate the residual atmospheric gases from the hollow probe before the soil-gas sample is collected. The soil-gas sample is collected in a 125 ml glass serum bottle with an aluminium crimp top securing a butyl-rubber stopper. The sample bottle is evacuated just before the sample is collected in order to reduce the possibility of contamination and to eliminate atmospheric dilution effects. A sample of the soil gas is drawn into the evacuated bottle. Additional soil gas is then pumped under pressure into the sample container. Probe sampling using this or any similar portable design can be used in a variety of geologic terrains within the limits of surface geologic features. Since an effective soil gas survey measures gas concentrations which have migrated into the soils, it is important that sample locations be placed in areas with at least one metre of residual soil. Alluvial and glacial deposits can also be sampled in most areas, provided there is not active, high-volume sediment deposition (which would require a deeper sampling method). Water-saturated soils and mud should be avoided because the wet sediments clog the sampler and if the open pore spaces normally present in the soil are reduced by water, then the amounts of free soil gas are much lower than in non-saturated soils. Shallow probe techniques are prone to near-surface lithologic, meteorological and barometric effects. This means that one must be careful in interpreting background values since the absence of an anomaly in a prospective or producing area may be related to lithology, rainfall, meltwater or barometric pumping. Areas containing
Light hydrocarbonsfor petroleum and gas prospecting
163
Fig. 5-18. Location of major basins in the USA (shaded) and surface geochemical surveys (black dots) carried out by Gulf Research and Development Company.
anomalously high gas contents, on the other hand, are almost always real seeps, since active flux is necessary to overcome these dilution effects. Shallow probes have been used successfully at Lost River in Hardy County, West Virginia, Patrick Draw in Sweetwater County, Wyoming (Matthews et al., 1984; Richers et al., 1982), Arrowhead Hot Springs in San Bemardino County, Califomia (Burtell, 1988) and on a large number of surveys conducted throughout the industry. Limited tests by Williams (1985) in the west Texas Permian Basin suggest that shallow probes are difficult to use in this area because of impermeable deposits of caliche and thick salt and anhydrite beds at a depth of about 300 m. An example of a halo-type anomaly reported by Williams (1985) is included in his thesis. Despite these limitations, shallow-probe sampling is still worthy of consideration because of the low sampling cost and ease of access in rugged areas with limited roads. With this method, small crews of only one or two persons can obtain large numbers of samples at minimal expense. In addition obtaining a permit (if required) is usually relatively simple because permitting authorities tend to classify such surveys as causing minimal environmental impact. The mobility of the soil gas probe sampling technique opens up large areas to geochemical exploration that are otherwise difficult to explore. Another means of obtaining free soil gas data is from auger holes drilled to 3.5 m (12 feet). These holes generally yield higher hydrocarbon concentrations than shallow probes. A fairly extensive research programme at Gulf Research and Development
V.T. Jones, M.D. Matthews and D.M. Richers
164
CHARCOAL TRAP RADON
IR DETECTOR
Amblerd Air Inlal
CO2
DUAL GAS CHROMATOGRAPH l~ - C4 Hydrocarbon Pneumatic Valves & Flame Ionization Detector Support
II Flame \ \ II ton o \ \
j
-
RECORDER INTEGRATOR
Carrier Gas
~ectr~ IIi PUvl,~ I,LI
2.1
I!Lxx\'\'\\\'''~ <
Fig. 5-19. Soil gas sampling procedure used by Gulf Research and Development Company.
Company established a database for geochemical exploration using auger holes comprising more than 21,000 analyses covering 16,000 line km (10,000 line miles) (Jones and Drozd, 1983). The locations of some of the research surveys are shown by black dots on a map of the major US basins (Fig. 5-18). An important aspect of this technique is that the data contain compositional information that not only can be tied to known fields but also are capable of predicting the oil versus gas potential of an unknown area before drilling. This predictive capability has proven to be applicable to several other techniques as well. A diagrammatic representation of the soil-gas sampling procedure used by Gulf Research and Development Company is shown in Fig. 5-19. Soil gas measurements are made in an auger hole, at least 4 m (13 feet) deep and typically 8.9 cm (3.5 inches) in diameter. A probe jacketed with an inflatable rubber packer is placed in the hole. When inflated, the packer effectively isolates the bottom of the hole from the atmosphere, so that the sealed base of the hole effectively serves as the sample container for the liberated gases. Soil gases are then either pumped into evacuated steel bombs or glass bottles for later analysis, or pumped directly into an on-site dual-column gas chromatograph for determination of the light hydrocarbons, helium and hydrogen. A 1 metre alumina-packed column coupled to a flame ionisation detector (FID) is used to determine the hydrocarbon content and a 3 metre molecular sieve column coupled to a thermal conductivity detector is used for the hydrogen and helium determinations. Carbon dioxide is analysed continuously using infrared adsorption techniques.
Light hydrocarbons for petroleum and gas prospecting
165
TABLE 5-VIII Composition range of soil-gas hydrocarbons over different reservoir types
Dry gas Gas condensate or oil and gas Oil
CI/C n
CI/C 2
(C3/C!) x 1000
95-100 75-95 5-50
20-100 10-20 4-10
2-20 20-60 60-500
TABLE 5-IX Average composition ratios of soil-gas hydrocarbons over different reservoirs Reservoir type Dry gas
Location Sacramento Basin
Oil and gas
San Joaquin Basin
Gas condensate
Southwest Texas
Oil
Western Overthrust Uvalde, Texas Permian Basin Utah Overthrust, Pineview Appalachians, Rosehiil Uinta Basin, Duchesne
Date 1972 1974 1975 1972 1974 1975 1975 1976 1978 1975 1976 1976 ! 978 1976
C1/C n
95 95 94 82 84 82 89 90 88 77 75 77 73 68
CI/C 2
55 49 55 8 7 8 12 11 12 5 5 5 4 4
(C3/CI)x 1000 6 8 I1 46 61 56 33 30 30 77 64 83 141 171
The auger hole technique yields excellent compositional information, even though the magnitudes are influenced slightly by the mechanical disaggregation associated with the drilling process. Compositional results for auger holes are sufficiently important to warrant further discussion here. An empirically-determined range of soil-gas data is shown in Table 5-VIII and a small selection of auger hole survey results is shown in Table 5-IX. The geochemical distinction between gas-type basins and oil-type basins was first noted from surveys in the Sacramento and San Joaquin Basins in California. Initial compositional data were gathered in these two basins in three separate years with excellent repeatability (Table 5-IX). Additional surveys conducted in southwest Texas supported the differences noted in California. Final confirmation on the oil versus gas predictions was obtained when numerous surveys were carried out in all three types of productive areas: gas, gas-condensate and oil. Soil-gas data from the Sacramento drygas, Alberta gas-condensate, and Permian Basin oil areas were used to establish statistically-valid populations based on histograms that demonstrate a close association with reservoir gases and gas shows in drilling fluids.
166
V.T. Jones, M.D. Matthews and D.M. Richers
TABLE 5-X Composition (mole fractions of C1-C4) of typical reservoir types (Katz and Williams, 1952) Reservoir type Dry gas High pressure gas High pressure oil Low pressure oil
Methane 0.91 0.81 0.77 0.37
Ethane 0.05 0.07 0.08 0.21
Propane 0.03 0.07 0.08 0.21
Butanes 0.01 0.05 0.07 0.21
Some typical percentages of methane and relative amounts of ethane through butanes in different types of deposits are given in Table 5-X. These data, taken from Katz and Williams (1952), show clearly that methane decreases in the trend from a dry-gas deposit to a typical low-pressure undersaturated oil deposit containing only dissolved gas but no gas cap. A better demonstration of this relationship comes from the study by Nikonov (1971), who compiled gas-analysis data from 3,500 different reservoirs in the United States, Europe and the then USSR, and grouped them into the populations shown in Fig. 5-20a. Gases from basins containing only dry gas (designated NG) contain less than 5% heavy homologs, whereas gases dissolved in oil pools (designated P) contain an average of 12.5% - 15% heavy homologs. The heavy homologs include ethane, propane, butane and pentane. Three of the near-surface data sets from Table 5-VIII are particularly convincing because the soil-gas measurements were made in basins that contained only one type of production. As shown by Fig. 5-20b, they are the dry-gas production of the Sacramento Basin (more than 450 sites), the gas-condensate production in the Alberta foothills (more than 650 sites), and the oil production of the Permian basin (more than 450 sites). Figures 5-20c, 5-20d and 5-20e show methane content (%C~), the methane:ethane ratio (C~/C2), and the propane:methane ratio (1000 x C3/C~) from the soil-gas populations over these three basins. These data clearly demonstrate that the chemical compositions of the soil gases from these three different areas form separate populations that appear to reflect the differences which exist in the subsurface reservoirs in these three basins. This correlation is particularly striking when compared with the data of Nikonov (1971), shown in Fig. 5-20a. The use of hydrocarbon compositions in soil gas prospecting requires enough data to allow statistically-valid and separate populations to be defined, so that a particular geochemical anomaly can be related to a geologic or geophysical objective or province. A percentage composition based on only two or three sites having 85% or 95% methane is not sufficient to define a population. As shown in Fig. 5-20a, considerable overlap exists among the intermediate gas-condensate and oil-type and gas-type deposits. In basins having mixed production, prediction of a reservoir gas-to-oil ratio (GOR) is clearly impossible.
Light hydrocarbonsfor petroleum and gas prospecting
167
Fig. 5-20. (a) Frequency distribution of the sum of methane homologs in 3,500 samples from different types of reservoirs (from Nikonov, 1971). Gas, oil and condensate surveys: (b) location; frequency distributions of hydrocarbons in soil gas over different basins, (r methane:ethane ratio, (d) propane:methane ratio, (e) methane content, (f) Pixler ratio diagram (Pixler, 1969), (g) soilgas data plotted on Pixler diagram. (h) Reservoir gas analyses of Verbanac and Dunia (1982) plotted on Pixler diagram.
168
V.T. Jones, M.D. Matthews and D.M. Richers
Where seeps contain gases from more than one reservoir, their compositions may not match those of any of the underlying reservoirs. Mixing of a shallow oil and a deep gas will generally yield an oily but intermediate-type composition. Without some knowledge of the reservoir possibilities, this type of signature cannot be recognised. Nevertheless, the intermediate nature of the seep will indicate some liquid potential at depth. Thus, dry-gas basins can be distinguished from basins that have at least some liquid oil or condensate potential. As suggested by Bernard (1982), the presence of fairly large ethane-propane-butane anomalies strongly suggests an oil-related source. Pixler (1969) found that the gases observed during drilling could distinguish the type of production associated with the hydrocarbon show during mud logging and published the graph shown in fig. 20f. Pixler's data were obtained by monitoring the C j-C5 hydrocarbons collected by steam-still reflux gas sampling during routine mud logging. Individual ratios of the C2-C5 light hydrocarbons with respect to methane provided discrete distributions that reflect the true natural variations of formation hydrocarbons from oil and gas deposits. Ratios below approximately 2 or above 200 indicated to Pixler that the deposits were non-commercial. The upper range for these ratios for dry-gas deposits has been enlarged by Verbanac and Dunia (1982), who studied more than 250 wells from 10 oil and gas fields. Their data, shown in fig. 5-20h, suggest the following upper limits for dry-gas reservoir ratios: C~/C2 <350, C~/C3 <900, CI/C 4 <1,500, C~/C5 <4,500. These ratios clearly aid in defining transition between thermogenic and biogenic gases. Another empirical rule suggested by Pixler is that the slope of the lines defined by these ratios must increase to the right; if they do not, the reservoir will be water-wet and therefore non-productive. Verbanac and Dunia (1982) suggested that a negative slope connecting individual ratios may result from fractured reservoir zones of limited permeability. Auger-hole soil-gas data for the surveys over the three basins described above are plotted on a Pixler-type diagram of reservoir gases in Fig. 5-20g. Direct comparison of these two independent data sets is very striking and proves the concept of migration of reservoired hydrocarbons to the surface. It is important to note that amounts of migrated gases almost always decrease in the following order: methane > ethane > propane > butane. Thus, in a Pixler-type diagram, soil gas-data, like reservoir data, generally plot as line segments of positive slope for the soil gases to represent a typical migrated seep gas. Exceptions to this order have been noted where surface source rocks were drilled, which thus far have yielded ratios with lighter gases depleted in relation to heavier gases. According to Leythaeuser et al. (1980), this would be expected if gases in the boundary layer very near the surface followed a diffusion model. Thus, compositional changes related to diffusion might be expected at or very near a boundary layer where the hydrocarbon gas concentration approaches zero. This behaviour has been observed when comparing soil gas probe data measured at very shallow depths (0.3-0.6 m, 1-2 feet) with the corresponding data from 4 metre (13 feet) auger holes. The shallow probe data are always "oilier", indicating preferential loss of methane and implying diffusion from the 4-m (13-feet) level to the surface. If diffusion were the dominant migration mechanism,
Light hydrocarbonsfor petroleum and gas prospecting
169
TABLE 5-XI Composition ratios of soil-gas hydrocarbons over Pleasant Creek gas storage area Date May 1975 July 1975 July 1976
C1/Cn
CI/C2
90 89 89
20 18 16
(C3/CI)x 1000 19 24 20
a chromatographic effect would be expected for gas that migrated through the Earth. The fact that the compositions of the soil-gas data from auger holes match the underlying reservoirs confirms that the major migration mechanism to the near-surface must be via faults and fractures, rather than by diffusion. The percent-methane compositions from the auger hole surveys conducted over the Sacramento and San Joaquin Basins are plotted in Fig. 5-21. There is a decrease from 98% methane in the north of the Sacramento basin to 90% in the south part, whilst the soil gas over the San Joaquin Basin has 82% methane. These data imply that a soil-gas grid would have defined local differences regionally. Furthermore these geochemical data are repeatable (Table 5-XI); the percent-methane values on Fig. 5-21 were all determined at least two or three times over a three-year period and found to be repeatable. Compositional data have remained repeatable throughout our experience with soil-gas surveys.
L-...,
f
Fig. 5-21. Percent-methane in soil gas over the Sacramento and San Joaquin Basins, California.
170
V.T. Jones, M.D. Matthews and D.M. Richers
O0
~T,~I 9
! wcJ, e,~., u g ( wca~ ewmo ~ewcme
Md ee ~. 1N
ETI04 9 .z,,e9
Elm
~1'~
~W
Delta 9
9 ~ ~, G r a n d I. worn e 4 _ I ,rt-._r--,__.r-" ~ ~,.r-~--.rd.J" I - - - ! S TImballer _ '--7 lYermillion I Eugenel. I " W. u a m e r o n I --. S. Marsh :~nnp E. C a m e r o n Shoal
Fig. 5-22. Well locations offshore Louisiana for which gas compositions are published (from Rice, 1980).
Dissolved gas
In offshore prospecting, "sniffers" have been used to detect anomalous hydrocarbon concentrations in bottom waters. An extensive review of the literature was published by Philp and Crisp (1982). Some of the most significant results reported by Williams et al. ( 1981) are highlighted here. Gulf Research and Development Company designed and operated several marine seep detectors that were employed aboard various research vessels, such as the RV Hollis Hedberg and its predecessor the RV Gulfrex. These ships circumnavigated the globe and conducted extensive detailed surveying in areas such as the Gulf of Mexico (Mousseau and Williams, 1979). The RV Hollis Hedberg system employed three separate water inlets which, whilst the ship was underway at normal seismic survey speeds, continuously supplied sample streams from the near surface, intermediate depths to 135 m (450 feet) and a deep-towed sample inlet at a depth of nearly 180 m (600 feet). Each sample stream is analysed for seven hydrocarbon gases once every three minutes with a sensitivity that depends upon the hydrocarbon and, for example, is about 5 x 101~ litres of propane at STP per litre of seawater. By using multiple depth inlets, surface contamination can be demonstrated to have no effect on seeps observed by the deep inlet. At sea, sniffer geochemical data from a deep tow inlet were superimposed to scale on a seismic section to aid the explorationist in making real-time evaluations of hydrocarbon potential of structurally-significant areas. As for surface soil gases, a powerful confirmation of the validity of marine geochemical data can be shown by the very close agreement between the composition of component hydrocarbons in production gases and the composition of seep anomaly gases in the same areas. Figure 5-22 shows the well database used for this confirmation
Light hydrocarbonsfor petroleum and gas prospecting
171
Fig. 5-23. Compositional crossplots of Rice's reservoir gas analysis. The underlying color code was chosen to distinguish oil, oil-condensate, gas condensate and gas within Rice's Gulf of Mexico production data.
in the Gulf of Mexico (Rice, 1980). For each of the 32 fields shown on this figure, the USGS has published the composition of gases produced from predominantly gas fields, oil fields and combined oil and gas fields or condensate fields. A crossplot of the compositions of gases from all field types is shown in Fig. 5-23 (Williams et al., 1981). The underlying colour code on this figure was chosen to distinguish oil, oil-condensate, gas-condensate and gas production using the Rice well analysis data as a standard. The log of the ratio of ethane to propane-plus-butane is plotted against the log of the ratio of methane to ethane-plus-propane. A distinctive compositional clustering of gas anomalies signifies different kinds of production: oil anomalies occur near the origin and become gassier as the points move up and to the right in Fig. 5-23. A crossplot of 146 sniffer geochemical anomalies from the same part of the Gulf of Mexico is plotted in Fig. 5-24b for direct comparision with the Rice well data shown in Figures 5-23 and 5-24a. As shown, the overall distribution is similar to the well data. Figures 5-24c and 5-24d illustrate the contrast in composition of dissolved hydrocarbon anomalies from a gas area and an oil area in the Gulf of Mexico. This type of regional separation was found to be typical of surveys conducted throughout the world. The fa'-t that production and surface anomaly gases correspond both onshore and offshore is significant. It proves that the observational techniques are valid despite the great variation in these surface environments.
172
V.T. Jones, M.D. Matthews and D.M. Richers
Fig. 5-24. Crosspiots of the compositions of gases from offshore Louisiana: (a) well gases throughout the area; (b) marine sniffer gases throughout the area; (c) marine sniffer gases from a gas area; (d) marine sniffer gases from an oil area.
Headspace gas A headspace sampling technique is commonly employed for the analysis of canned samples from drilling returns and from shallow sediments. In this technique a controlled volume of sediment is placed in a can or jar filled with a measured volume of degassed brine. The can is sealed and a measured volume of brine is displaced with nitrogen to create a known volume headspace. The can is then allowed to come to equilibrium. The concentration of light gases can then be measured by syringe injection of a headspace sample into a gas chromatograph equipped with an flame ionisation detector.
Light hydrocarbonsfor petroleum and gas prospecting
173
In order to maintain reproducibility it is important to measure all volumes accurately. In a typical operation using 500 ml (one pint) cans, the procedure is to place 300 ml of degassed salt-water brine into the 500 ml can and add sediment until the can is filled to the brim, giving 200 ml of sediment and 300 ml of brine. The can is sealed and then zero-grade nitrogen is injected through a prepared septum to displace 100 ml of brine and leaving the can with a 2:2:1 mixture of 200 ml brine, 200 ml sediment, and 100 ml headspace. Experiments have shown that a fairly long time is required for the adsorbed sediment gases to completely equilibrate with the headspace. This equilibrium time is shortened by heating and shaking the cans before analysis. A generally accepted procedure is to heat the cans for about 12 hours at 60-70~ followed by shaking in a paint mixer for five minutes. After heating and shaking, the cans are allowed to stand for at least five further minutes to ensure that dissolved gases return to the headspace. One of the drawbacks of using this technique is the need to freeze the canned samples if they cannot be analysed within one or two weeks of their collection. Failure to follow this procedure can create problems because of the generation of biogenic gas in the cans or the bacterial oxidation of the hydrocarbon gases to carbon dioxide. Hydrocarbon concentration values are reported in terms of ppm by volume in the nitrogen headspace or as ppm or ppb by weight, normalised to the weight of sediment. Gases concentrations reported by weight are not truly representative of the actual gas migrating from depth because some of the free gas has been allowed to escape during collection and sample preparation. Furthermore, the sorbed gas is never completely extracted into the headspace, and may not always reflect the true gas content of the soil. The headspace sampling technique can yield useful results if sufficient numbers of samples can be collected to use statistical populations to suggest anomalous areas. One should always exercise caution, however, with respect to characterisation of gas composition, since evaporation during the collection stage always occurs, resulting in the relative depletion of the lighter gases.
Disaggregation Extensive soil gas sampling programmes carried out by the petroleum exploration industry have demonstrated that the crushing and/or disaggregation of soils (including the action performed in drilling auger holes) is an important component part of the extraction of gas from the soil. This suggests that it would be advantageous to employ a soil core disaggregation technique which would closely mirror the effect of auger hole drilling. A device developed at Citco and commonly used in both industry and academia for analysing well cuttings appears suitable for accomplishing this objective (Whelan, 1979; Hunt and Whelan, 1979; Whelan et al., 1980). In fact, Richers (1984) has demonstrated successfully that in some instances, such as at Rose Hill, Virginia, and in
V.T. Jones, M.D. Matthews and D.M. Richers
174 '--
?.5cm
L,
L I C O N E R U B B E R SEAL 5 c m O U T S I D E DIA. 5 c m I N S I D E DIA.
6.
f
INJECTION PORT
/ - - THREADED P I PE I crn D ~A. ~ SEPTUM CAP FROM VARI AN / J S I L I CONE RUBBER / ~ /SEPTUM, ,crn DIA. /
I~
r
"i'-i~,NJECT,ON PO~T
llllilllllilllll i, 2cm
~
- -'
'--
-
Fig. 5-25. Stainless steel disaggregation mill used in headspace analysis
the Western Overthrust Belt, the results obtained by this technique are in very good agreement with data from auger holes. The device used in this technique is a small stainless steel ball-mill containing two stainless steel or ceramic balls which crush and disaggregate the sample when the ballmill is shaken (Fig. 5-25). This approach concentrates the loosely-bound adsorbed gases into the headspace of the ball-mill. Because of the equilibrium problem mentioned above under headspace techniques, this sampler was adapted by Whelan (1979) and Whelan et al. (1980) to ensure that lithified sediments and cuttings are completely broken up during analysis. Basically, the technique is as follows. A small (but constant) volume of sediment, soil or cuttings is placed into the mixer cell along with two ceramic or stainless steel ball-bearings, and water is added to bring the remaining headspace to 10 cc. The mill is sealed and placed in a SPEX/Mixer-Mill and agitated for about five minutes. The cell is then immersed into a hot-water bath at 90~ for three minutes. A 1 ml aliquot of gas-free water is injected into the cell through a septum-sealed side arm on the cell, and then a 1
Light hydrocarbonsfor petroleum and gas prospecting
175
ml aliquot of the headspace is sampled using a locking gas-tight syringe. The sample is then hand injected into a gas chromatograph equipped with a flame ionisation detector for analysis of the disaggregated gases. It is assumed that these gases represent micopore gas, some free gas and lightly adsorbed gas on the sample-medium surface. This technique (or modifications of it) has been used in the analysis of well cuttings and deep sea cores (Hunt and Whelan, 1979), in addition to surface geochemical prospecting (Richers et al., 1986; Richers and Weatherby, 1985). Initial tests of this method were conducted at Gulf Research and Development Company for comparison with the auger hole technique and to gain a better understanding of the relationship between free gas and adsorbed gases liberated by the drilling process. To be an effective and viable technique, the disaggregation desorption method must be able to distinguish between oily and gassy areas. An area known to be predominantly oily, Rose Hill in Lee County, Virginia, and another known to be predominantly gassy, the Gulf Research Facility in Pittsburgh, Pennsylvania, were chosen as initial test sites. Both areas had been sampled previously using the auger hole technique, allowing the new data to be compared with the established data sets (Richers, 1984). The Rose Hill test site includes 126 soil cores of which 51 fall within 300 m (1,000 feet) of the earlier auger holes. Despite differences in the sample locations and depths, both techniques correctly identify the area as oil-prone. Table 5-XII shows the relationship between the diagnostic gas ratios (Jones and Drozd, 1983) and the results of the two surveys (Richers, 1984). It is obvious that the ball-mill technique accurately describes the oil-prone nature of the Rose Hill oil field. However, the data of Table 5XII suggest a slight difference in the composition of the hydrocarbons detected by the two techniques. In the auger holes the soil gas is slightly drier (methane-rich) than the soil gas obtained by ball-mill disaggregation-desorption. This shift may reflect the preferential loss of methane from the shallow cores compared to the deeper auger holes and the difference between core samples and free gas measurements. The other gases are essentially the same in both techniques: the C4/nC4 ratio for the disaggregation technique is 0.34, and the auger hole technique yields a value of 0.40; the C2/C 3 ratios are comparable at 1.84 for the disaggregation technique and 1.76 for the auger hole technique. In addition, the intercorrelation of the various hydrocarbon gases in the disaggregation data set is higher than that for the auger hole data. This high degree of
TABLE 5-XII Comparison of results of free soil-gas and disaggregated soil-gas surveys, Rosehill, Virginia (Richers, 1984) Survey method Free soil gas Disaggregated gas
No. of sites 145 128
% methane 72 70
CI/C 2
(C3/Ci)x 1000
7 7
110 117
176
V.T. Jones, M.D. Matthews and D.M. Richers
correlation among the gases may reflect a near-equilibrium condition achieved through time for the adsorption-desorption process in soils. Hence, the signal seen by the desorption technique may effectively integrate and smooth rapid changes one might expect to see with a free-gas technique such as auger holes. At the Gulf Research Facility in Pittsburgh, Pennsylvania, there are two producing gas wells, and 38 sites were selected to test the ability of the disaggregation technique to define gassy areas. Not only did the test yield gassier results than those obtained at Rose Hill, but also the results were again comparable to those obtained using the auger hole technique. Table 5-XIII is a compilation of these results. Clearly, the two data sets reflect a more gas-prone area for the Gulf Research Facility than for the Rose Hill area. Although the data set for the disaggregation technique is only half of the size of the data set from the auger holes, it still yields useful information regarding composition of the subsurface reservoirs.
A c i d extraction
A technique which measures only the most tightly-bound gas was originally developed by Horvitz (1939, 1945, 1950, 1954, 1957, 1965, 1969). In this technique the sample is subjected to acid digestion under vacuum at an elevated temperature of about 80~ Further developments by Debnam (1969) and Horvitz (1972, 1980, 1981) involved corrections for lithology to reduce the effect of acid-soluble minerals biasing the data. Debnam (1969) noted that soil samples could be dried, pulverised and sieved without affecting their hydrocarbon content. He also noted that sieving sand samples to <200 mesh gave analytical values comparable with those produced by clay samples from the same location. Horvitz developed a wet-sieving technique to concentrate the analysis on only the clay fraction of the sediment. McCrossan et al. (1971) evaluated the acid-extraction technique in the western part of Alberta. This extensive survey of over 4561 samples covering 15 townships concluded that the distribution of anomalous points was random and was strongly biased by samples rich in carbonate minerals. Adequate corrections for amounts of acid-soluble material were not successful and it was concluded that this method could not be used in areas covered by glacial till.
TABLE 5-XIII Diagnostic soil-gas ratios at the Gulf Research Facility, Pittsburgh, Pennsylvania Survey method Free soil gas Disaggregated gas
No. samples 73 38
% methane 89 91
CI/C 2
18 25
(C3/CI)XI000 21 22
Light hydrocarbonsfor petroleum and gas prospecting
177
As early as 1940, Sanderson had discussed a number of factors that affected adsorption of hydrocarbon gases by soils. He noted that the ability of the soil to adsorb any gas depends upon the type of gas, the characteristics of the soil and the conditions under which the soil is exposed to the gas. Adsorption will depend upon the type and surface area of particles and their chemical composition. The surface reactivity will be modified considerably by the presence of previously-adsorbed molecules, such as carbon dioxide, water and mineral ions. The condition of adsorption is complicated by temperature and pressure and length of exposure time in addition to concentrations and species of gases present. Adsorbed-gas data can, at best, be only approximations of the original mixture of migrated gases. Another possible problem lies in the quantitative desorption of the gases from the mineral components of the soil. Sanderson (1940) observed up to six-fold differences in the ability of soils to adsorb hydrocarbons in his laboratory. He also noted that the adsorptive characteristics of the colloidal soil systems would vary slowly with moisture content, time and season. Of particular significance was his observation that the adsorptive capacity for hydrocarbons on wet soil was only a small fraction of that for dry soil. A further complication is created by near-surface biological activity that creates wide variations in the content of carbon dioxide, nitrous oxide and other biological gases. Overcoming all these problems is probably impossible; however, it will suffice if the gases are liberated in proportion to the amounts present so that the analytical results bear some relationship to one another, and allow identification of potentially prospective areas. Various other approaches have been devised in attempts to overcome this problem. Bays (US patent no. 2,165,440) suggested correcting for the sorptive power of the soils and McDermott (US patent no. 3,120,428) suggested correcting for the surface area. An alternative technique proposed by Thompson (1971) used ethylenediaminetetracetic acid (EDTA) at about pH 7 and slightly heated in order to decompose the carbonate minerals under conditions that do not release such large quantities of carbon dioxide. Thompson reports that a comparison on duplicate samples shows that the EDTA technique consistently releases from 94-99.5% of the hydrocarbon gases released by the standard strong-acid treatment. A further refinement of this method by Thompson et al. (1974) separates a critical carbonate mineral before analysis. This critical mineral was almost always found to be dolomite, but occasionally is other carbonate minerals, such as iron or calcium carbonate. The ratio of hydrocarbons per unit of critical mineral is then plotted to form a geochemical prospecting map. This technique was reported to highlight a salt dome in the Gulf of Mexico on which a major oil discovery was made after the survey was conducted. Poll (1975) addressed this problem of lithologic corrections by dividing data according to desorption efficiencies based on their physicochemical properties. The first step is to prepare a detailed lithological description of the samples. This involves a differentiation on sediment lithology, sample coherence, structure, cementation and mineral types, including carbonate and sand percentages. This information is used as shown in Fig. 5-26 to classify the samples into homogeneous sets for each of which the
178
K T. Jones, M.D. Matthews and D.M. Richers
10~
< SAND
~,01t
4(MM~ CARB ~;0%
~ i:+:i:+:'. o . "-,'-''+''''+'"o o % ~,,,
l 3~ Content
~
Gas Content
> Ca~RB- O~
T
"~
CAM - I~
" Gas Content
i Gas Content
i.9 ~[
T
I' I
I
I
o.
501,1(SAND '~"0~
I Gas
+ +-
O~ C.,ARB- 0%
u.
m3
I
D
<.<,
Gas Q~tent
~tent---'--~
>
Fig. 5-26. Scheme for lithological classification of samples prior to interpretation of gases released by acid extraction (reproduced with permission of Australian Petroleum Production and Exploration Association from Poll, 1975).
average, or background, gas content is computed. The gas content in each group is assumed to be distributed according to a Laplace-Gauss law. Each subset is then assumed to have a uniform efficiency of desorption and its own background and anomaly threshold. As shown in Fig. 5-26, for calcareous sediments these are very high, due to the effectiveness of the acid attack. The mean normal standard can be computed for each set yielding dimensionless values that can be added together for mapping, regardless of the sediment type. This technique has been applied by Poll (1975) in the Gippsland Basin and by Devine (1977) and Devine and Sears (1985) in the Cooper Basin in Australia. Reasonably positive results were reported in all three cases. The acid-extraction technique relies on the ability of soil and minerals to retain hydrocarbons that migrate past them through the soil pore system. It is therefore not subject to the fluctuation involved in the soil-air system but hopefully represents some averaged or integrated signal over time. As noted above, the samples must be corrected for lithologic efti~cts by only making comparisons within a given lithology or by specifically analysing certain minerals. Corrections must always be applied because adsorption occurs in both the fine-grained fractions and in carbonates, which often release disproportionately large amounts of hydrocarbons.
Light hydrocarbonsfor petroleum and gas prospecting
179
Fig. 5-27. Comparative fluorescence spectra of nine crude oils of different (from Purvis et al., 1977)
Fluorescence As an extension of light hydrocarbon gas analysis, UV fluorescence spectroscopy can be used to measure the oil potential of near-surface sediments by analysing their aromatic hydrocarbons. This is highly sensitive and selective method for the analysis of oil components, particularly those containing one or more aromatic functional groups. Using spectroscopic scanning, complex molecular aggregates, such as those found in crude oils, can be rapidly characterized and quantified on the basis of their combined intensity wavelength distribution or "fingerprint". The fluorescence spectra of nine crude oils of different gravity are shown in Fig. 5-27 (Purvis et al., 1977). These two-dimensional fluorograms were produced by exciting at 265 nm and scanning from 250 nm toward the red end of the spectrum. The accepted procedure for illustrating the change in the emission spectrum associated with differentgravity crude oils is to measure the intensity of fluorescence at two wavelengths: 320 nm for light aromatic compounds; and 365 nm for the heavier, multiple-ring aromatic compounds. The intensity of the fluorescence emission is proportional to the quantity of aromatics in the extracted sample. The standard field method employs a rapid wet extraction process which dissolves loosely-bound trace aromatics into hexane. This extract generally favours the heavy oil fraction, which is in hydrophobic association with the sediment.
180
V.T. Jones, M.D. Matthews and D.M. Richers
A second phase of in-depth, total scanning fluorometric analysis is often performed on selected anomalous samples identified by the field fluorescence, adsorbed-gasdata or interstitial-gas data. These samples undergo freeze drying followed by a thorough cyclic extraction in hexane to optimise recovery of associated sedimentary aromatics (Brooks et al., 1986). The oil type is then determined by total scanning fluorescence which employs step-wise scanning of excitation and emission wavelengths to produce a threedimensional fingerprint fluorogram (Fig. 5-28).
SAMPLING STRATEGY Spatial pattems of near-surface hydrocarbon composition and concentration are prime factors when interpreting the survey results. Results from a poorly-designed or an uncontrolled survey can be difficult or impossible to interpret, and can lead to a completely erroneous assessment of the hydrocarbon potential of an area. An improperly-spaced grid with sample spacing in excess of target size can result in only the most cursory assessment of potential, with anomalous areas appearing as localised single-point anomalies. The distribution of sample sites in a geochemical survey is largely governed by the purpose and budget of the survey. For regional surveys a sampling density of one sample per 2-5 km 2 seems adequate. Such a density still allows for the discrimination of regional ambient backgrounds from secondary backgrounds. Detailed diagnostic work requires a close-spaced grid, sometimes with a sample interval of only a few tens of m. Regional sampling is generally performed using a modified grid because a regular grid, on which samples are taken at the intersections of a straight lines, does not minimise cost or maximise information. We recommend that sample positions be chosen within grid cells according to ease of access (minimum cost) and along zones of known or inferred fracturing and faulting (maximum information). Satellite imagery, aerial photography, seismic data and other data are useful when attempting to site samples on or near fractures and faults. The analytical results from a regional survey should yield some indication of compositional and/or magnitude "sweet-spots", either as isolated data points or small clusters. If the objective is merely to evaluate whether a basin has a source section, and general trends of where it is mature and focused to the surface, a regional study may be all that is required. A more detailed follow-up survey, however, is recommended if the objective is to highlight the zones of higher hydrocarbon potential. One method commonly employed for detailed surveys is to sample seismic shot holes, further providing a means to easily tie the geochemistry to subsurface structure. Because seismic lines are not normally placed on a close-spaced grid, infill sampling between seismic lines is usually recommended. It should be emphasised that in order to define a target adequately, approximately 70% of the data should be collected in presumed background areas beyond the immediate target area. An embarrassingly large number of surveys have been performed in which sample locations do not extended
Light hydrocarbonsfor petroleum and gas prospecting OIL/CONDENSATE lu..
FU.,.,o
SEDIMENT EXTRACT
ao it p , ~ t . . , a 4 P ~ Y a o l - ) i ,. .=o.. Son j~
~s, i~o(e)mo(ot}l
300 - ~ En#ldon 400-
181
~
3K)O F.xdMlion --
T1 - 12448.0
440
"r2 911067.9
EfflklMoll
400-
-300
~
TI -- 19~3a.2
IExciletlm
'r2 .. lllOeO.7
m
6oo
JmuwV f~
,,,
I
200 mO
' * *,,' 404) iJO0 r:mJim (mumms)
800
200
I
dO0
I am
I
,
,
400 (nlm ometem)
MO
HIGH ISLAND AREA 2-ring aromatic compounds Spectra characterizes gas/condensate OIL
SEDIMENT EXTRACT
Ua. = ,~s I,lm~m)rJoo(uN
T 1 , a1743L4
o~o
m
!
N Max.. ~
200 T2..3a41i3.6 i
i
1 rJJa~
|
T! -~3104.3
"r2,1644.1
"
I~
!:!
(a7o(emyaao(~)l
/
9
I v
m
(n,,mmnn)
GREEN CANYON AREA 3 and 4 ring aromatic compounds Spectra characterizes upward migrated oil Fig. 5-28. Fluorograms of reservoir hydrocarbons and corresponding hydrocarbons extracted from sediments in the Gulf of Mexico.
more than one or two sites beyond the anomaly. The result of this misplaced desire to save money is often an ambiguous survey interpretation. The selection of a technique that is inappropriate for the surface geologic conditions in part of the survey area can also lead to erroneous results. An example is the use of the
182
V.T. Jones, M.D. Matthews and D.M. Richers
acid extraction technique on glacial till or acid soils, which normally yield low results. Without regard to the particular constraints on such data, one could easily overlook a favourable area. In this case another technique such as free gas would be more representative.
DATA INTERPRETATION There are many ways to analyse hydrocarbon gas data with no one particular method being correct or incorrect. Common sense and a deterministic approach to sound geologic models are the best guidelines. Integration with other data such as structure, lithology, soil types and hydrogeology, to name a few, can be most fruitful.
Preferential pathway model The lack of a model explaining the mechanisms and constraints of hydrocarbon leakage is often an obstacle to the acceptance of surface geochemical prospecting (although a similar lack of understanding of the migration of hydrocarbons from source beds to reservoir has not precluded the acceptance that migration occurs). Assimilation of the data, however, suggests that much can be explained by a relatively simple model. The conclusion that effusion is the dominant mode of migration enables us to use the visual patterns associated with macroseeps as a basis for our microseepage model. Link (1952) and Levorsen (1967) have summarised the geologic conditions and controls on macroseepage. There is no reason to expect that these controls should not apply as well to microseepage; the only real difference should be a matter of scale. In addition to seepage directly from exposed source beds, controls on surface seepage include: (1) the surface exposure of reservoir beds or porous carrier facies; (2) porosity associated with unconformities; and (3) surface expressions of faults and fracture systems that are pervasive to depth. These controls may be summarised as the focusing of migration along preferred permeability pathways. Horizontal migration along the pathway is dominated by grain or bed permeability (including old erosion surfaces and other unconformities), whilst vertical migration is controlled by cross-stratigraphic discontinuities. Horizontal pathways deflect the surface location of the anomaly laterally away from its subsurface origin. Thus if an anomaly is associated with the surface expression of a porous formation, one should suspect a down-dip source (or down-groundwater gradient source). The same conclusions can be inferred for anomalies associated with unconformities, low angle faults and listric faults. Vertical pathways are dominated by the intersection of high angle faults and fractures with reservoir and carrier beds. In this case the surface expression of the source of the hydrocarbons will lie directly above, or only slightly displaced from the source. The
Light hydrocarbonsfor petroleum and gas prospecting
183
presence of multiple, stacked porous zones also often results in a surface geochemical expression that is approximately vertically above its subsurface origin. The role of faults and fractures is particularly important for microseepage and some further comment is in order. The close association of near-surface geochemical anomalies with faults and fractures has been pointed out by, amongst others, Horvitz (1939), Sokolov (1971b), Richers et al. (1982), Jones and Drozd (1983) and Matthews et al. (1984). McCrossan et al. (1971) point to the close association of high concentrations of hydrocarbons in the surface environment with photolineaments. McDermott (1940) suggests that the permeability of shale is dominated by microfractures and that these fractures are preferentially normal to the bedding plane. This potentially important role of microfractures is emphasised by Rosaire (1938), who correctly points out that the failure to observe displacement does not eliminate the existence of a fault or fracture. The high permeability of fractures causes them to preferentially focus fluid flow. The effectiveness of fractures as mass transport systems for fluids is evident from a casual examination of mineralisation in fractured rocks and leakage of groundwater at fracture outcrops. Similarly, these fractures act as preferential hydrocarbon pathways, focusing their flow from source beds to surface. Faults and interconnected fracture systems have a significant effect on the magnitude and, less commonly, composition of the near-surface gases. The effect on magnitude is generally to increase concentrations in fractured areas, whilst the effect on composition theoretically should be preferential loss of lighter gases compared to heavier gases. In practice, gas compositions on faults are often lighter or heavier than those at neighbouring sites. This is believed to be controlled primarily by the depth of the fault and the composition of the subsurface gases it conducts. Thus deep, basement-related faults are often gassy because they tap deep over-mature sediments. Shallower faults are often oily because large molecules migrate more easily than the lighter compounds. The increase in magnitude in fracture systems can often be abrupt and localised. It commonly spans several orders of magnitude, going from nil to macroseep levels in the extreme cases. In an area where there is no significant source of subsurface hydrocarbons, there are no high-magnitude soil-gas signals, even on faults and fractures. In a hydrocarbon-bearing environment, however, overall high variance in the data is more often the case, but the anomaly-to-background ratio is smaller in non-producing areas than in producing areas. Some of these anomalous zones are associated with preferential leakage directly from a source bed, while others are from reservoirs. Since some faults and fractures are sealed locally along their lengths, high-magnitude signals do not occur everywhere along their length. Thus, we often observe "hydrocarbon spots", similar to the "helium spots" discussed by Wakita (1978). Naturally, those faults penetrating only source beds will show a signal that reflects the source beds, whereas those penetrating a reservoir or both reservoir and source beds will exhibit a larger anomalous signal. It is not known, however, if one can truly distinguish between the two types in all instances, although extremely high magnitudes are felt to be more diagnostic
184
V.T. Jones, M.D. Matthews and D.M. Richers
Lineament Center 9
..,h ...ue Station
///
B
/ / Fracture /,/~X//r x
Background Station
/x
x
x/
D i s t a n c e from Lineament canter ~
t /
i oOoo ,.;o.o"'o"'" ".-;" D i s t a n c e from Lineament center v
Fig. 5-29. Relation between fracture intensity and gas leakage: (A) plan showing lineament, fractures and gas sample sites; (B) distribution of fracture intersections with distance from lineament; (C) distribution of anomalous gas sample sites with distance from lineament (reproduced with permission of the American Association of Petroleum Geologists, whose permission is required for future use, from Richers et al., 1986, AAPG Bull., vol. 70, no. 7, Fig. 13, p. 885, AAPG 9 1986). of reservoirs, as seepage volumes are expected to be larger from reservoirs than from a source bed (Hunt, 1981). The expectation that all samples in a leaking fracture zone are higher than those outside the zone is simplistic, and is not always realised in practice. A fault or fracture is rarely one discrete plane, but zones of broken or disrupted strata, separated by relativelyunaffected competent strata. It is analogous to a fractured pipe: certain portions of the conduit are solid, whereas the fractured section is composed of both intact fragments and cracks. Fluids flowing through the pipe are going to leak in the fractured areas of the pipe but not in the solid-walled portions. Similarly, even in the fractured zones, the fragmented areas will leak only through the fractures, not through the fragments of pipe between the fractures. Extrapolating this model up to geologic scales, sampling outside the fracture zone is expected to give values that are typical of the background of the area. Within the fractured sample zone, sample sites may intersect discrete fractures or encounter coherent blocks between the fractures (Fig. 5-29A). The intensity of fracturing, and hence the probability of the fractures interconnecting, increases toward the centre of the fracture zone, as shown in Fig. 5-29B. Therefore, samples taken near the centre would be expected to be a mixture of high values (intersecting fractures that connect), median values (intersecting fractures that do not connect) and low values (not
Light hydrocarbonsfor petroleum and gas prospecting
185
intersecting fractures). Further from the centre of the fracture zone, the maximum values fall until they merge with those typical for the background of the area. This distribution of free soil-gas magnitudes as a function of distance from the centre of the fracture zone is shown in Fig. 5-29C (Richers et al., 1986). Disaggregation data from Patrick Draw exhibit a similar pattern, although the increase near the centre of the fracture zone is not as great; acid extraction data from this example show no obvious relationship, clearly suggesting that different analysis techniques are extracting gases from different sources. The following examples illustrate the means of interpreting what are often referred to as direct anomalies using preferential pathway models. These direct anomalies may be either vertically over their subsurface source, or laterally displaced by varying amounts (Sokolov, 1971b; Pirson, 1969; Laubmeyer, 1933). What is generally not realised is that most areas contain microfractures to the extent that they allow gases to escape vertically. Using a coal-bum experiment in the central Wyoming coal region, Jones and Thune (1982) showed that a definite vertical-migration component could be identified. In that experiment, gases formed during combustion appeared both in soil gases directly above the retort and up-clip along the bedding planes of the strata involved in the burn. Thus, vertical signals from a known subsurface origin were shown to exhibit crossstratigraphic migration, presumably due to the presence of fractures in the system. A second horizontally-displaced component also migrated along the bedding planes at the same time. An example of the use of direct anomalies and the preferential pathway model is shown in Fig. 5-30, which shows an idealised subsurface cross-section through the Lost River field in West Virginia along with a propane profile (Matthews et al., 1984). From this profile and with some knowledge of the geology, it can be seen that a large anomaly is probably caused by updip leakage of the fractured Devonian Oriskany reservoir at depth. This outcrop anomaly is due to updip leakage along the bedding plane of the reservoir facies. A smaller but significant anomaly is related to leakage from a fault that strikes along and to the east of the crest of the producing anticline. Blind drilling on the outcrop anomaly would have resulted in a dry hole, whereas drilling just west of the fault anomaly would have encountered the producing structure. Appropriate geological modelling identifies the location at which to drill. An alternative to the direct anomaly interpretation method relies on identifying one of two types of halo: (1) local lows, source background areas surrounded by highs; or (2) extremely low areas, surrounded by areas of moderate concentration. These halos are consistent with the initial results obtained with soil-gas analysis techniques (Rosaire, 1938; Horvitz, 1939, 1945, 1954, 1985; McDermott, 1940; Rosaire, et al., 1940), which indicated that adsorbed and occluded hydrocarbons occur in greater quantities around the edges of production, whereas relatively lower values are found directly above production. Halo anomalies have been recognised in many regions of the former USSR (Kartsev et al., 1959). Horvitz (1969, 1980) has emphasised that although other hydrocarbon distribution patterns are recognised, including direct anomalies, the halo
186
V.T. Jones, M.D. Matthews and D.M. Richers
4
E
Q. O.
3 LU Z
2~ rr
Q. I
,
SOUTH 40~176 .~
," ~
2000, z
g
WHIP COVE EAST ANTICLINE No. 9254 ~
0
NW--"
~
WEST LIMB HANGING ROCK ANTICLINE
LOST
RIVER
0
>-2000 W W
-4000' -6000
Oriskany Sandstone VERTICAL EXAGERAT)ON 2 : 1
g Metws
Fig. 5-30. Cross-section through the Lost River oil field, West Virginia, and profile of propane in soil-gas (reproduced with permission of Veridian-ERIM International from Matthews et al., 1984).
pattern continues to be the most common type found in conjunction with important oil and gas accumulations. Numerous explanations have been put forth as to why halos form around hydrocarbon accumulations. Most of these link the phenomena to the impedance effect of a diagenic mineralisation zone overlying the main part of the petroleum accumulation. Such a zone would tend to reduce the ability of gases to seep vertically, except along well-pronounced fracture systems. Hence, most transport would be deflected around the edges of the occluded zone. The occluded zone could form by any number of diagenic processes. Rosaire (1940) suggested that the greater solubility of carbon dioxide in petroleum, as compared to water, results in the conversion of bicarbonates to less soluble carbonates over an accumulation. An initial chimney effect would result in a greater supply of bicarbonate being present above an accumulation resulting in the cementation. Rosaire (1940) also proposed the reduction of sulphates to sulphides over an accumulation. Fenn (1940) reintroduced another process that was first introduced by Mills and Wells (1919). This model is based on the evaporation of ground moisture as the result of gas expansion which results in the subsequent precipitation of minerals at shallow depths. The origin of the blocked central portion over an accumulation implies that gas-induced evaporation occurs more effectively over an accumulation than along its margins. This model is consistent with results on the variations in unusual chemical
Light hydrocarbonsfor petroleum and gas prospecting
187
and isotopic compositions of carbonate-cemented surface rocks over oil and gas fields (Donovan, 1974; Donovan and Dalziel, 1977). Stroganov (1969) has confirmed that the deeper distribution of hydrocarbons only rarely yields a halo pattern, suggesting the halos have a near-surface origin. Matthews (1985) suggested that diagenetic blockage related to hydrocarbon emplacement may originate at intermediate depths and then be exhumed by erosional processes. Although direct anomalies and halos have conflicting explanations, both appear to be valid. Indeed, the controversy is significant only if it is assumed that lateral displacement has not occurred during subsurface leakage. This is certainly a valid assumption in some, but definitely not all, cases. If the halo pattern is interpreted as a subset of several preferential pathways, one can assume that at least one major flowpath could become blocked by diagenetic cement, resulting in a bias of leakage, with a false halo forming as the gases are diverted around this blockage in an area that previously yielded a direct anomaly. In one study the occurrence of halos was suggested by adsorbed soil-gas samples, whilst direct anomalies were observed using free soil-gas samples (Richers et al., 1986). One must speculate that these techniques measure different aspects of the leakage phenomena. For this reason, it is felt prudent to always collect both types of samples whenever economically feasible. In addition one would be well advised to incorporate geological and geophysical data into the model. A significant portion of near-surface hydrocarbon survey results appear to be compatible with the mechanisms of macroseepage, particularly leakage occurring along preferential pathways. Those anomalies seemingly not coincident with known faults, fractures, unconformities, bedding planes or other obvious pathways may lie on pathways unrecognised due to limited or incorrect mapping. Alternatively, some occurrences may represent processes not completely understood, or processes not validly extrapolated from macroseepage to microseepage. The preferential pathway model summarises the movement of hydrocarbon fluids through the subsurface to their final destination as a surface seep, either directly or by way of an intermediate trap. It is certainly not definitive nor complete, but illustrates some of the challenges confronting the petroleum geologist in his quest for new resources.
Geochemical populations An altemative to modelling hydrocarbon gas migration as a basis for data interpretation is to decompose data into geochemical populations. On this basis surface geochemical data can be interpreted with respect to both composition and magnitude. The goal of compositional analyses is to be able to characterise the type or types of subsurface accumulations present and to be able to predict the location at which they occur. This can be achieved through using ratios of the various hydrocarbon constituents that are detected in the soil-gas sample. In general, gas reservoirs are commonly
188
V.T. Jones, M.D. Matthews and D.M. Richers
TABLE 5-XlV Concentrations of unsaturated hydrocarbons (10 -4 vol. %) generated during oxidation of gaseous hydrocarbons by a culture of Myc. Flavum incubated at 30-32~ (Telegina and Cherkinskaya, 1971) Day 0
8-10 30
Experimental conditions Myc. Flavum present Control no bacteria Myc. Flavum present Control no bacteria Myc. Flavum present Control no bacteria
C2= 0
Aerobic (21.2% O2) C3-Ca-0
C2-0
Anaerobic (1.4% 02) C3= Ca= 0 0
0
0
0
0
0
0
0.004 0 0 0
0 0 0.003 0
0 0 0 0
0.227 0.158 0.064 0.500
0.062 0 0.114 0
0 0 0.003 0
dominated by the presence of methane, whereas oil reservoirs usually contain additional quantities of hydrocarbon gases heavier than methane (Nikonov, 1971). There are three potential origins for gases detected in the near-surface environment: biogenic, thermogenic (or katogenic) and igneous (including mantle degassing); and irrespective of the origin, the gases tend to migrate towards the surface due to pressure and buoyancy effects. Gases from several sources may mix or undergo other compositional changes such as chromatographic separation, during this migration. Thus the measured compositions may not always reflect the original subsurface composition. In most areas mixing presents little problem because gases of thermogenic origin are by far the most abundant. Furthermore, the tendency for gases of biogenic and igneous origin to be extremely dry and of a different isotopic composition from thermogenic gases enables recognition of their presence. Extreme chromatographic separation may only be recognised by careful isotopic analysis and through the close comparison of near-surface gas with known reservoir gas in the region. The presence of gas of igneous origin generally indicates the occurrence of deep, pervasive faulting, and/or the presence of igneous activity in the area. This association, as well as the extremely methane-rich character of such gases, allows for the facile distinction between gases from thermogenic and igneous sources. Telegina and Cherkinskaya (1971) found that the olefin content of soil gases decreased relative to saturated hydrocarbons until depths of about 300 m. Experimentally, as illustrated in Table 5-XIV, olefins can be formed from saturated compounds in areas of low oxygen content (0.5-3.2 %). The presence of these olefins may be biogenic (Smith and Ellis, 1963), although Starobinetz (1976) showed a linear relationship between the concentrations of saturated and unsaturated gases derived from the thermogenic alteration of organic matter. Sokolov (1971 b), among others, suggested a relationship between the generation of unsaturated compounds and drilling activity.
Light hydrocarbonsfor petroleum and gas prospecting
189
Gleezen (1985) showed that there is promise in using the olef'm contents of soil gases as a scaling factor to separate seep signals from ambient signals. He was able to define areas with signatures similar to those of the reservoired gases. It would appear that in some cases the presence of olefins may merely represent the breakdown of saturated hydrocarbons by some yet-undetermined process during the migration of gases to the surface and/or some activity such as biogenic degradation of the saturates in the nearsurface environment (Telegina and Cherkinskaya, 1971). Compositional information in soil gases has been related to subsurface accumulations through the application of specific ratios (Jones and Drozd, 1983). Methane-dependent ratios (Table 5-VIII) are reliable unless multiple sources of gas are present in the area. An independent methane-rich source biases an oilier composition toward a drier gas composition. This can sometimes be overcome by plotting histograms of the compositional data and noting multiple populations in the data. Another set of diagnostic ratios that are not methane dependent has also been defined and further aid in properly defining the true potential of an area (Drozd et al., 1981; Williams et al., 1981). In general, the agreement between the surface compositions with reservoir compositions is the strongest evidence that surface prospecting can accurately define the potential of an area. In addition to compositional information, soil-gas data can yield useful information according to the presence or absence of anomalously-high magnitudes. To understand the concept of anomalously-high magnitudes, one must understand the general distribution of gases in nature. Basically these can be reduced to three main populations for any given region. 1) An ambient background population (which represents a detectable level of nonsignificant hydrocarbon concentrations). This includes mantle-derived hydrocarbons, contamination, instrumental noise, sampling error, etc. 2) A source background population representing hydrocarbons derived from the presence of organic-rich source beds in a region. These are generally areal in extent, and they may or may not be relatively consistent throughout the area depending on local geologic variations, regional trends or multiple sources. 3) An anomalous population of higher-than-normal concentrations of hydrocarbons that represent the subsurface presence of concentrated hydrocarbons such as those found in reservoirs. Ambient levels, by their very nature, are encountered everywhere, and are always a component of the total soil-gas signal regardless of the overall hydrocarbon potential of an area. Their presence may be due to natural catagenesis of organically-poor rocks during the processes of diagenesis and lithification, and can be thought of as being syngenetic. Another source is the biogenic alteration of organic matter in the near-
190
V.T. Jones, M.D. Matthews and D.M. Richers
TABLE 5-XV Hydrocarbon content of Palaeogene formations in productive and non,productive areas of western Siberia (Starobinetz, 1983)
Nort_h Varegau (productive) No. samples C~-C8 (I 0"4cc/kg) C5-Cs / Cz-C4 Cs-Cs % No. samples MCA l 0"4 Aromatics, C6+C7 (I 04 cc/kg) C6 / C7 Pokrpvskaya (non-productive) No. samples C5-C~ (10"4cc/kg) Cs-Cs / C2-C4 Cs-Cs % No. samples MCA 10-4 Aromatics, C6+C7 (10.4 cc/kg) C6 / C7
Nekrasovskaya
Cheganovskaya
Dulinvorskaya
7 129 42 2.73 7 50 41 0.6
10 131 47 1.90 4 21 17 0.5
20 133 21 0.08 9 19 13 0.4
10 90 65 0.24 5 5 4 0.3
14 16 4.4 0,01 6 6 4 0.3
17 24 4.3 0.01 9 11 7 0.3
surface. Typically, ambient background areas contain a little methane and virtually no other hydrocarbon gases. An illustration of ambient background levels is shown after Starobinetz (1983) in Table 5-XV. Zinger et al. (1983) provide data that are typical of a sourced background from the Kuybyshev oil-bearing area of the former USSR. Here the methane content varies between 20% and 57%, with heavier homologs consistently present. The backgrounds that occur in such areas are considered to be sourced backgrounds because the effects of the pooled hydrocarbons are superimposed on the lower ambient background signal. The anomalous population comprises only a very small portion of the overall data set, typically only a few percent. Values for these samples generally are 2-3 times the magnitude of the sourced background. In some instances, concentrations may reach the percentage level, in which case the locations border on the macroseepage rather than microseepage. At the other end of the spectrum are those samples that are 5 or 10 times above the background concentration. These may represent either a separate population from the sourced background, or merely high-frequency fluctuations in the sourced background. There are two fundamentally-different approaches to defining anomalous magnitudes. The traditional technique focuses solely on the distribution of hydrocarbon
Light hydrocarbons for petroleum and gas prospecting
191
concentrations, regardless of location. A magnitude threshold, or series of thresholds, is chosen and those locations with values above this threshold (anomalously-high concentrations) are identified on a map. The second technique focuses on the spatial clustering of anomalous stations. This is accomplished by the identification of regions where the number of stations with magnitudes above a threshold is statistically significant. The traditional method of identifying a magnitude threshold has been accomplished by a variety of techniques. These include: (1) the mean plus two standard deviations of a normally-distributed data set; (2) arbitrarily selecting the 90 th percentile or 95 th percentile, etc., of the data; (3) identifying the inflection point on a cumulative frequency plot that deviates from a straight line (Sinclair, 1976). In the opinion of the authors it is dangerous to select any hard-and-fast rule for defining an anomalous population, although the approach of Sinclair (1976) is the most appropriate for a mixed mode data set. Sample populations should be normal, or at least log normal, for many of the statistical tests to be valid, and bias in sample sites should be avoided if possible. Ideally a training set made up of a data subset with known hydrocarbon potential should be employed. This gives a means to tie-in data to a known feature, whether it be a source bed, a reservoir or a barren area. Once the results are available, a first step is to construct histograms to determine the spread of the data. The data can then be plotted on cumulative frequency plots to determine the different populations. Scatter plots of key components, such as methane versus propane, or isobutane versus normal butane, often yield multiple trends for multiple populations in the data. Pearson correlation analysis also yields useful information on the "cleanliness" of the data, with single populations generally showing a high degree of inter-correlation. Filtering or screening the data according to composition prior to applying statistics is also an effective means of determining areas of favourable potential. The method of identifying regions of anomalously-high leakage by clustering (Dickinson and Matthews, 1993) is accomplished by first identifying a magnitude threshold and a search area. The magnitude threshold, which is somewhat arbitrary, is used to transform the distribution of magnitudes into a binomial population (above the threshold "heads" and below the threshold "tails"). The size of the search area (the "cell") is such that it includes 20 or more sample stations regardless of its location within the surveyed area. Once these parameters are chosen, the cell is placed at one position on the map, usually in one comer, and the percentage of heads and the total number of stations within the cell are recorded. The cell is then translated to a new location and the same parameters are recorded. This process is repeated until the entire survey area has been examined. Because the properties of the binomial distribution are well known, statistical tests of the chance of a particular cell having a particular percentage of heads can be made and probability maps contoured. Thus, regions of anomalously-high frequency of magnitudes above the threshold can be identified, and their chance of arising due to random processes, instead of focused leakage, can be estimated. There is,
V. 7". Jones, M.D. Matthews and D.M. Richers
192 TABLE 5-XVI
Success rates of surface gas geochemistry in petroleum prospecting in Azerbaijan (Zorkin et al., 1982) Province Areas Apsheronskaya N izh n iek uri n skaya Kirovobadskaya Kobystano-Shemahinskaya Total
Positive prognosis Correct % 8 7 87 4 3 75 10 7 70 4 2 50 26 19 70
Areas
Negative prognosis Correct % 8 7 87 6 6 100 14 13 90
however, the risk that information about spatial variability within a cell is lost and so is the information about the absolute magnitude of individual samples. On the basis of anomalous magnitudes, Zorkin et al. (1982) showed that, in 90% of cases, the soil-gas technique correctly identified areas lacking hydrocarbon potential in Azerbaijan, and correctly identified areas with hydrocarbon potential in 70% of cases (Table 5-XVI). Although the distinction of ambient from secondary background is often relatively straightforward, the distinction becomes ambiguous in areas with effective seals, such as stable intracratonic salt basins.
CASE HISTORIES Numerous case histories illustrating the relationship of surface seeps to their associated production are given in the cited references. Four surveys, three onshore and one offshore, are selected here to demonstrate and confirm the compositional relationships defined above. The first onshore example consists of calibration grids conducted over two fields in the Neuquen Basin of Argentina, and the second example is a sniffer survey conducted for calibration purposes over gas-productive areas in the High Island area of the Gulf of Mexico. The two other onshore surveys are located in the Great Basin of Nevada and in the Overthrust Belt of Wyoming-Utah.
Neuquen Basin, Argentina Detailed soil gas geochemical surveys were conducted for calibration purposes over two fields, Filo Morado and Loma de La Lata, in the Neuquen Basin in Argentina. These two fields were chosen for this calibration study because of their differences in both reservoir composition and entrapment mechanisms. Filo Morado is an anticlinal oil field producing from the Agrio Formation at a depth of 3000 m (10,000 feet). Loma de La Lata consists of two stratigraphically-trapped
Light hydrocarbonsfor petroleum and gas prospecting
193
Fig. 5-31. Spatial distribution of soil gas hydrocarbons at Filo Morado, Argentina (arbitrary coordinates): dot size indicates ethane concentration; dot colour indicates CI/C2 ratio, such that green = low (oil), yellow = intermediate, red = high (gas).
reservoirs formed on a homocline which dips to the northeast. Oil production comes from the Quintuco Formation at 2000 m (6600 feet). This reservoir is partially underlain by a separate gas to gas condensate reservoir producing from the Sierras Blancas formation at 3000 m (10,000 feet). The three separate reservoirs from these two fields provide two oil reservoirs and one gas to gas condensate reservoir for calibration of the soil-gas geochemical data. The geochemical data come from 239 shallow probe (1.2 metre, 4 feet) soil-gas samples collected on 500 - 1000 m grids placed directly over these two fields, with 95 sites over Filo Morado and 144 sites over Loma de La Lata. The free soil gases were analysed for methane, ethane, ethylene, propane, propylene, iso-butane and normal butanc by gas chromatography using a flame ionisation detector. In order to illustrate the distribution and compositions of the light hydrocarbon seepage, compositional dot maps which combine both the light hydrocarbon magnitudes and compositional information are shown in Fig. 5-31 for Filo Morado and in Fig. 5-32 for Loma de La Lata. Each dot is coloured according to the C~/C2 ratio to reflect the composition of soil gases as indicative of oil (green), gas (red) or intermediate (yellow). The dots, including those at localities with only background magnitudes, vary in size according to their ethane magnitudes.
194
V.T. Jones, M.D. Matthews and D.M. Richers
Fig. 5-32. Spatial distribution of soil gas hydrocarbons at Loma de La Lata, Argentina (arbitrary coordinates): dot size indicates ethane concentration; dot colour indicates CJC2 ratio, such that green = low (oil), yellow = intermediate, red = high (gas).
The compositional subdivisions are derived from the published literature (Nikonov, 1971; Jones and Drozd, 1983) and are the same as those shown in Table 5-VIII. The shade of each of the anomaly clusters suggests the oil versus gas potential of the anomaly according to these empirical divisions alone. Ratios of methane/ethane, methane/propane and methane/total butanes for all sites that exceed the median of the data are also shown in Figs. 5-31 and 5-32, in order to provide a visual illustration of the composition of the anomalous data. The bimodal nature of the Loma de La Lata soil-gas data is clearly shown by the red (gas) and green (oil) populations whilst Filo Morado stands in stark contrast, with its unimodal oily (green) population and lack of gas-type anomalies. Examination and comparison of these ratio plots and dot maps for each of the two fields indicate that the more anomalous magnitude sites (large dots) match the composition of the known underlying reservoirs. The areal groupings and Pixler ratio plots of these specific components with their appropriate reservoirs lends confidence to the deduction that these soil-gas anomalies are the result of migration of petrogenic hydrocarbons from the underlying sedimentary sources. The geochemical soil gas data exhibit clearly defined compositional sub-populations which match the composition of the underlying reservoirs and change in direct response to the major structural and/or stratigraphic features that control the location of the subsurface reservoirs. Predictions of oil versus gas from these soil gas data are in
Light hydrocarbonsfor petroleum and gas prospecting
195
excellent agreement with published soil gas and reservoir data (Jones and Drozd, 1983; Nikonov, 1971). A single oil source is predicted at Filo Morado, in agreement with the known oil field. Much gassier soil-gas data is noted over the Loma de La Lata Field, where there exists an oil field underlain by a gas to gas condensate reservoir. However, a very striking change to fairly large magnitude oil-type compositional anomalies occurs directly over the northwest portion of the Loma de La Lata Field where the Quintuco oil reservoir is the only known producing horizon. This change in composition from oil to gas condensate signatures over the Loma de La Lata Field occurs across a permeability pinchout at depth, which controls the updip limits of the deeper gas condensate reservoir.
High Island area, Gulf of Mexico A marine hydrocarbon seep-detection survey was completed over High Island Blocks A-152 and A-198 and surrounding areas in the Gulf of Mexico on 22-23 April 1988 (Fig. 5-33a. This study, consisting of 399 km (239 miles) of sniffer data, was conducted aboard the RV GYRE by Texas A&M University in conjunction with Exploration Technologies Inc. using a marine hydrocarbon analytical system originally designed by Gulf Oil Corporation for use on the RV Hollis Hedberg. Light hydrocarbon data were collected continuously along seismic lines of interest from a water-sampling system towed about 9 m (30 feet) above the bottom of the sea floor. A total of 87 km (52 miles) of gridded data (259 analyses) were completed over Block 152A and a total of 51 km (31 miles) of gridded data (129 analyses) were completed over Block A-198. Samples were taken at 3-minute intervals giving an approximate sample spacing of about 450 m (1500 feet) Anomaly compositions are plotted on a marine crossplot in Fig. 5-33b for comparison with the calibration crossplots in Fig. 5-24. Three regional profiles are presented in Fig. 5-34 to show the magnitude variations along the survey lines. Survey tracks, as shown on Fig. 5-33a, include a 90 km (54 miles) regional southnorth line which extends from Block A-198 to Block A-321 in the High Island South extension. The results from this regional line, plotted in Fig. 5-34b, provide both a calibration data set over the known gas fields and a background data set which extends between the two gridded blocks. As shown by Fig. 5-34b, background values are observed in Blocks A-237, A-224 and A-223, where concentrations are about 100 nl/1 methane, <0.70 nl/ethane and <0.50 nl/1 propane. These concentrations correspond with typical backgrounds found in previous studies in the Gulf of Mexico (Mousseau and Williams, 1979. The largest magnitude anomalies observed on this entire survey are also noted on this regional line (Fig. 5-34b), where it crosses the centre of Block A-268 and traverses the major trend of the known gas-producing fields. Within this producing trend, methane values exceed 500 nl/1, ethane ranges from 1-5 nl/1 and propane rises from 0.501 nl/1. In addition, iso-butane and normal butane reach a combined total of about 1 nl/1 in anomalies associated with these known gas fields.
196
V.T. Jones, M.D. Matthews and D.M. Richers
Light hydrocarbonsfor petroleum and gas prospecting
197
The presence of butanes in the sniffer data clearly separates the southern gasproducing trend from the area to the north of Block A-252. The grids over both Block A-152 and Block A-198 and the profile data north of Block A-252 exhibit a clear lack of propane and butane anomalies. The presence of mainly methane in the northern area suggests that these anomalies are derived from biogenic gas sources. Marine compositional crossplots from the anomalies observed in Block A-152 and A-198, and from the regional profile are compared in Fig. 5-34d. All three areas fall exactly as expected, based on the known oil and gas producing reservoirs within this survey area. Blocks A-152 and A-198 are similar to the fairly dry gas type Pleistocene reservoirs found in West Cameron, in the Louisiana offshore area, and are indicative of only gas potential. Block A-198 anomalies appear to contain even drier gas data than those in Block A-152. Data from both of these blocks plot below the cluster associated with the major Pliocene gas-producing trend that lies to the south of Blocks A-152 and A-198. The increase in ethane, propane and butanes in this southern gas-producing area suggests that these gas fields contain Pliocene gas from a more petrogenic source, whereas the areas to the north appear to be dominated by biogenic gas sources that do not contain C2+ components. It should be noted that the new field discoveries (A-129, A154 and A-200), highlighted in Fig. 5-33a, were made after the sniffer survey was completed.
Great Basin, Railroad Valley, Nevada The third example in this chapter is an onshore survey conducted in Railroad Valley, Nevada. This example is abstracted from an integrated two-year (1984-85) remotesensing and surface-geochemical research project which provides an excellent example of the exploration value of combined remote-sensing and geochemical studies in frontier basins (Jones et al., 1985; Burtell et al., 1986). The variability of sample spacing used over this two-year programme, coupled with repeated, even-closer detailed sampling on grids in 1985, allows a demonstration of the sampling artifacts that can be created by over-interpreting a low-density regional survey. In addition, the repeatability of soil-gas surveys is demonstrated. The compositional correlations predicting oil versus gas, as shown in the two previous examples, are extended to differentiating non-commercial heavy-oil deposits from their lighter counterparts, neither of which contains any significant gas production.
Fig. 5-33. High Island geochemical sniffer survey 1988: a) sniffer track map; b) marine compositional crossplot.
198
V.T. Jones, M.D. Matthews and D. Richers
Fig. 5-34. Profile of dissolved hydrocarbon data from High Island sniffer survey; a) east-west line A-B; b) south-north line C-D; c) north-south-north line E-F; d) marine compositional crossplot for 1988 sniffer anomalies.
The first-year study in Railroad Valley, conducted in 1984, consisted of a regional lineament evaluation made from Landsat Thematic Mapper (TM) imagery, Synthetic Aperture Radar (SAR) imagery and regional soil-gas probe sampling to identify areas of significant hydrocarbon seepage (Jones, et al. 1985). Railroad Valley was chosen for this research study because of the excellent surface expression of structural features,
Light hydrocarbonsjbr petroleum and gas prospecting
199
Fig. 5-35. Contour maps of 1984 regional soil-gas data, Railroad Valley, Nevada: a) methane contour map; b) propane contour map.
including both lineaments and circular geomorphic anomalies, which have been used by Dolly (1979) and Foster (1979) to locate drainage anomalies, interpreted as reflecting differential subsidence of subsurface structural blocks. The first three producing fields discovered in Railroad Valley (Eagle Springs, Trap Spring and Grant Canyon) occur in circular features mapped by Dolly (1979) and Foster (1979). Subsequent discoveries in Railroad Valley have not discounted this proposed association, but the discovery of Paleozoic reservoirs has added considerably to the complexity of the model. Contour maps of the regional methane and propane soil-gas data gathered in 1984 are shown in Fig. 5-35, along with major structural and geomorphic features mapped by Dolly (1979) and Foster (1979). Both components exhibit large-magnitude geochemical anomalies that clearly originate at the basin-bounding fault and extend updip onto the adjacent pediment block. A very simplified cartoon explains how this updip migration might occur through fractures and/or draped sand lenses contained within the Tertiary fill (Fig. 5-36). An alternative approach to contour maps is to generate a colour compositional dot map (Fig. 5-37), in which the size of each dot is proportional to the ethane magnitude and the colour is selected from the Pixler ratio plot (Fig. 5-37, inset, upper left).
200
V. 7". Jones, M.D. Matthews and D. Richers
Fig.5-36. Geological/geochemical seep model illustrating possible migration pathways for Railroad Valley, Nevada.
Choosing the standard empirical classes from Table 5-VIII for these data shows that the producing oil fields fall within the yellow, rather than within the green areas, as would be expected for the heavy oils produced in Railroad Valley. This colour compositional dot map suggests that it is possible to differentiate between hydrocarbon types from the relative position of each site on these Pixler ratio plots. Eagle Springs, Trap Springs and Grant Canyon fields have well-controlled intermediate compositions (yellow dots), while the Currant well area exhibits much lower, oilier ratios (green dots). Thus hydrocarbon seep compositions observed in Railroad Valley appear to differentiate productive or potentially-productive reservoirs from non-productive heavy-oil accumulations at depth. These compositional changes are spatially closely related, suggesting that the compositional changes may occur across geologic boundaries, which control both the hydrocarbon reservoirs and their associated surface seepage. The study in Railroad Valley also showed that a large number of high-magnitude seeps occur near to, or on, lineaments and lineament intersections (Jones, et al. 1985). This classic relationship reflects one of the most valuable uses of remote-sensing lineament studies in frontier basins. Preferential location of geochemical samples in the vicinity of active structural zones and their intersections will usually locate a large number of the hydrocarbon seeps in any basin. In addition, regions of intense fracturing
Light hydrocarbonsfor petroleum and gas prospecting
201
Fig. 5-37. Ethane colour compositional dot map for 1984 regional soil-gas data, Railroad Valley, Nevada; inset, Pixler ratio composition.
that do not exhibit hydrocarbon seepage strongly suggests a lack of source potential at depth in such areas. This study in Railroad Valley demonstrates a unique surface geochemical expression of one particular lineament, herein named the Currant Lineament, which appears to have a dramatic effect on the commercial possibilities for a subsurface oil deposit. The noncommercial Currant No. I well is located just to the southeast of a NE-SW linear feature, which crosses the valley through the town of Currant in northern part of Railroad Valley. The location of this lineament is obvious on all of the regional remote-sensing products. Although the lineament is dramatically expressed both northeast and southwest of Currant, it is not as obvious in the centre of the study area. Even more detailed aerial
Y.T. Jones, M.D. Matthews and D. Richers
202
1
Mapped Photolinears, Currant Detail Area, [ Railroad Valley, Nye County, Nevada
/
~~, / f--/
/
~
.
""
/
'
"
~'////"
J ~ S '
~\
~~
,/.- /"
"~
x /
,
l"
/"
I
,
"~ !------
/i
~
-'-
; ,
f-
i
-z~j_..~~,
t
"
'4
/'i
N
L~/1
,
LEGEND t',?,, beach
/ /_
J __
I C~--,possible
>//~
/,I
9
'
,
t_.-
....
"
//
/ Y
-
/
i/,'i
..
' S/.L~"
"
//'/
-,-
"Z / /
7/
./~.//.~~---/ .OS_ / / ~
/
, fault
down $1de)
_
t
0
,.
t
~0
I Mi J
~ 1 Km
Fig. 5-38. Mapped photolinears, Currant detail area (from Jones et al., 1985).
photography (Fig. 5-38) yields only a series of fairly short photolineaments, most of which are drainage segments not obviously related to the regional lineament. As shown by the colour compositional dot map in Fig. 5-37, hydrocarbon seeps to the northwest of the Currant lineament have compositions, as defined by Pixler ratio plots, that are quite similar to the productive fields in Railroad Valley. Sites to the southeast of the lineament
Light hydrocarbonsfor petroleum and gas prospecting
203
have a much oilier signature, suggesting a relative depletion of volatiles from the sources of measured soil gases. A 400-site detailed grid geochemical survey on 300-metre (1000-feet) centres was carried out the second study year (1985) over this section of the Currant lineament and further supplemented by aerial-photographic studies in an attempt to characterise the local expression of this regional lineament. The location of this second-year study with respect to the 1984 regional survey is shown in Fig. 5-37 (inset). High- and low-altitude aerial photographic studies reveal that, although the regional feature is well expressed in TM data, it does not dominate the length, azimuth, or density of the small-scale linear features. The lineament appears only as a minor group of parallel and subparallel linear features, which are easily lost in the clusters of more local fracture zones (Fig. 5-38). The Currant lineament is, however, expressed in gravity data as a trough, suggesting that it is, in fact, a deep-sourced feature of regional significance, which may influence subsurface fluid and gas migration along or across its strike. Light hydrocarbon soil-gas data from the Currant grid area show some alignment of anomalous values along the strike of the lineament, but it is apparent that the lineament does not control hydrocarbon magnitudes in this area (Fig. 5-39). Hydrocarbon magnitudes appear to be controlled, to a greater degree, by the north-south and east-west small-scale linear features (shown in Fig. 5-38), which probably reflect, to some extent, the location of subsurface structural faults and fault-related fracture systems. This relationship is quite important because structures identified by lineament zones are generally not the sole controlling factor for light hydrocarbon seepage; rather they simply provide enhanced pathways of migration for gases and fluids. The local geologic framework and source potential are the most important factors for interpreting the relationship of hydrocarbon seeps and lineaments. Compositional data from the Currant grid area adds great insight into the effect a regional lineament can have on the sources of migrating gases, even though the lineament is not directly mappable at the local scale. Regional data from the 1984 Railroad Valley programme indicated a compositional shift across the lineament zone. A methane-ethane crossplot of the 1985 Currant detailed grid data shows two distinct populations (Fig. 5-40c). Pixler ratio plots of anomalous sites support this subdivision and actually show two distinct populations, with the vast majority of the gassier sites plotting to the northwest of the Currant lineament (Fig. 5-40a), on the basin side where deeper sources exist. These two compositional subpopulations are clearly shown by the yellow-to-green colour change in the crossplot in Fig. 5-40b. This was first noted in the 1984 survey data and confirmed by the 1985 survey data. It should be kept in mind that the colour code used for both the dot maps (Fig. 5-40a) and the crossplot (Fig. 5-40b) is determined by plotting the raw soil-gas data on the Pixler diagram. Thus the colour compositions are selected to be similar to analyses of samples from actual producing reservoirs. The combination of both spatial and compositional clustering clearly demonstrates the stability, and repeatability of light soil-gas data that are properly collected and analysed.
204
V.T. Jones, M.D. Matthews and D. Richers
Light hydrocarbonsfor petroleum and gas prospecting
205
Fig. 5-40. a) Ethane colour compositional dot map for 1985 Currant detailed soil-gas data, Railroad Valley, Nevada; b) Pixler-type diagram characterisation of anomalous soil-gas hydrocarbons associated with known oil fields in Railroad Valley, Nevada; c) methane/ethane scatter plot for 1985 Currant detailed soil-gas data, Railroad Valley, Nevada.
The distinct compositional change associated with this regional lineament suggests that subsurface hydrodynamic processes related to the lineament may control not only the sources at depth but also the light-hydrocarbon seepage compositions associated with these sources. The lineament may form a barrier to subsurface water flow and divert fluid flow to the east of the lineament. Oil accumulations east of the lineament could, therefore, be water-washed, resulting in the non-commercial heavy oil observed in the Currant No. i well. Potential petroleum reservoirs west of the lineament may be protected from water-washing, retaining their volatile constituents, and providing a gassier soil-gas signature at the surface. If this interpretation is correct, it proves the
Fig. 5-39. Railroad Valley, Nevada: a) methane contour map of 1985 Currant detailed soil-gas data; b) propane contour map of 1985 Currant detailed soil-gas data.
206 V.T. Jones, M.D. Matthews and D. Richers
Light hydrocarbonsfor petroleum and gas prospecting
207
local significance of this regional lineament system, even though the feature is not immediately obvious from small-scale remote-sensing data alone. It is also important to note that the regional geochemical study conducted in 1984 would not have been sufficient to support this interpretation, and that close detailed data gathered in 1985 were required to properly confirm the relationships between lineaments and hydrocarbon seepage in this case. It is very important to realise that a regional geochemical survey on 1.6-km (1-mile) or even 5-km (3-mile) grids represents a low-resolution approximation to the actual size or shape of any geochemical anomaly. As shown in Fig. 5-39, the C1 and C3 patterns from the survey on 300-metre (1000-feet) centres are very different from those in the 1984 regional contour maps. The sharp geochemical boundaries observed in the detailed study cannot be mapped from the regional geochemical data. This is shown very clearly in Fig. 5-41, which expands the regional survey propane contour for direct comparison with the detailed grid data. Comparison of the magnitudes and compositions in these two data sets using the colour compositional dot maps (Fig. 5-42) proves that the regional data are valid and of good quality. However, using the regional data to draw contours is a serious mistake, which results in an erroneous interpretation in terms of the location of this complex anomaly. The data of the more detailed survey are essential before comparison with seismic data is attempted.Fracture orientations from the aerialphotography overlay define and control the sharp boundaries of the geochemical anomalies (Jones et al., 1985). The Currant lineament cuts through the centre of this major seep anomaly and appears to have some influence on fluid flow at depth (it appears to control the economics of the potential reservoirs). The shape of the geochemical anomaly is controlled by the bounding fractures, which are obviously not controlled by this regional lineament. A comparison of the regional lineament interpreted from satellite remote-sensing data with the detailed composite interpretation from aerial photography shows that the azimuth of the Currant lineament is expressed only in the short photolineaments. However, the regional lineament is not obvious from only the short photolineaments. Based on only the aerial photography we might suggest that this lineament is not real; the geochemical data, however, clearly shows otherwise and clearly shows the value of integrated multidisciplinary interpretations.
Fig. 5-41. Comparison of propane contour maps for soil-gas data, Currant area, Railroad Valley, Nevada, illustrating importance of sample spacing: a) 1984 regional survey; and b) 1985 detailed survey.
208 V.T. Jones, M.D. Matthews and D. Richers
Light hydrocarbonsfor petroleum and gas prospecting
209
Overthrust Belt, Wyoming-Utah The final example is one of the largest regional applications of light-gas surface studies ever published (Dickinson and Matthews, 1993). Some 3300 km 2 (1280 square miles) of the Wyoming-Utah overthrust belt, including the Clear Creek, Ryckman Creek and Whitney Canyon-Carter Creek fields, plus several small fields, was investigated using 1890 free soil-gas measurements (Fig. 5-43). The effective source rocks in the area are believed to be within the subthnast Cretaceous (Warner 1982). The maturity of these source rocks increases westward and appaears responsible for the change in production from mixed oil, condensate and gas in the east, to dry gas, wet gas and some condensate in the middle, to dry gas in the west. The compositional information derived from the surface gas study falls within the gas/condensate-mixed oil/gas classification of Jones and Drozd (1983). Further, there is a trend towards a more gas prone character from east to west, in agreement with both the production trends and increasing source rock maturity. A comparison of the light gas analysis of produced hydrocarbons with the surface free gases shows that the (C2/C3)xl 0 values are in very good agreement for the Ryckman Creek and Clear Creek fields and in general agreement with respect to the ranges of values for the multiple reservoirs in the Whitney Canyon-Clear Creek field. The (C3/Cl)xl000 ratios, however, are considerably more methane rich in the surface than in the subsurface at Ryckman Creek and Clear Creek. This suggests that there is an independent source of methane in the region which is mixing with the leakage of the Cretaceous-reservoired gases. This independent source is either absent or much less effective at Whitney Canyon-Carter Creek. In designing this study, Dickinson and Matthews (1993) decided that a sampling density of two samples per 2.5 km 2 (1 square mile), with approximately uniform distribution of locations, would represent a good compromise between the need for detail and cost. The regional focus of this study precludes the identification of all but very broad regions of interest because of the possibility of the occurrence of single point anomalies due to the coarse sample spacing. As a result, Dickinson and Matthews (1993) developed their cell technique, which we have previously described as an anomaly-probability map. Figure 5-44 shows a composite cell map in which the technique has been applied to methane, ethane and propane; the regions where all three of these gases are above their respective medians has been highlighted. The average number of sites within a cell was 18. Thus, binomial theory suggests that cells with more than 75% of the values above the median would be expected to occur only 5% of the time. The 75% contour line clearly identified several large areas that occupy more
Fig. 5-42. Comparison of ethane colour dot maps for Currant area, Railroad Valley, Nevada, illustrating repeatability of soil-gas compositional data: a) 1984 regional survey; and b) 1985 detailed survey.
210
V. T. Jones, M.D. Matthews and D. Richers
Fig. 5-43. Oil fields and gas fields of the Overthurst Belt, Wyoming-Utah.
than 5% of the total area. These regions are statistically anomalous, suggesting the occurrence of above-average seepage in these areas. Note the association of these anomalous areas with the Whitney Canyon-Carter Creek, Ryckman Creek and Clear Creek fields and with the surface trace of the major thrust faults. If this information had been available prior to the discovery of these fields, exploration could have been concentrated in the currently-productive region, saving costs. In addition, the general type of production and trend of composition would have been correctly predicted.
Light hydrocarbonsfor petroleum and gas prospecting
211
Fig. 5-44. Composite cell map in which the Monte Carlo simulations have been applied to methane, ethane and'propane anomalies in the Overthrust Belt, Wyoming-Utah, highlighting the regions where all three of these gases are above their respective medians.
CONCLUSIONS Surface and near-surface hydrocarbon occurrences arise as a result of a complex series of events and interrelationships. Except in rare instances, surface prospecting cannot reveal the outline of subsurface accumulations, nor indicate the potential commercial worth of a prospect. It can allow the explorationist a means to high-grade prospects, but should never be used as the sole criteria for delineating drilling locations. Surface soil-gas anomalies exist for many understandable reasons, although some do appear rather random. The interpretation of such data is derived from the general ability to extrapolate from macroseepage to microseepage, and the fact that often the surface
212
V.T. Jones, M.D. Matthews and D. Richers
signal detected is directly correlated to gross subsurface hydrocarbon composition. Thus surface soil gas prospecting techniques utilising hydrocarbons can be a reliable test for indicating the presence of subsurface hydrocarbon source and/or accumulation. Present day exploration for oil and gas requires a coordinated effort based on all useful techniques of geophysics, geology, and geochemistry. The above discussion on geochemical-prospecting techniques is useful for exploration geologists and geophysicists who wish to enhance their exploration activities through the use of surface geochemistry. We must avoid hailing each new technological advance as a panacea, because there is no direct method for finding oil and gas. Each exploration tool has its positive and negative points, and it is up to the explorationist to use these tools properly. The basic programme is one of economics in an era of rising exploration, developing and marketing costs. The function of an exploration geologist is to increase the odds of drilling a producing well by every economic means at his command. Given appropriate limitations, established geochemical prospecting techniques can be applied to aid a rational exploration program in any basin in the world.
Geochemical Remote Sensing of the Subsurface Edited by M.Hale
Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9
Elsevier Science B.V. All rights reserved
213
Chapter 6
GAS GEOCHEMISTRY SURVEYS FOR PETROLEUM T. RUAN and Q. FEI
INTRODUCTION Gas geochemical surveys appear inherently suitable for petroleum exploration because oil and gas fields are natural accumulations of volatile hydrocarbons. They were first attempted some sixty years ago by Laubmeyer (1933) and Sokolov (1935). Nevertheless, there has since been only rather limited use of them. The reasons for this are fourfold: (1) the fundamental concept of hydrocarbon gases escaping through considerable thicknesses of "impermeable" cap rock has been slow to gain acceptance; (2) gas geochemical sampling devices and detection methods have often been insufficiently sophisticated and have therefore produced unsatisfactory data; (3) gas anomalies are unstable and have often proved difficult to repeat; and (4) the petroleum industry has achieved a high success rate with seismic exploration and hence has not felt an acute need for an alternative exploration tool. By the 1980's, however, the situation was beginning to change. A series of direct observations was persuading petroleum geologists that volatile hydrocarbons related to petroleum at depth are detectable at the surface (Jones and Drozd, 1983). Drilling proved the significance of a few hydrocarbon-gas anomalies found some forty years earlier (Horvitz, 1969; Horvitz, 1981). Technical advances in analytical instrumentation (especially in gas chromatography and mass spectrometry) meant that concentrations and isotopic compositions of hydrocarbons could be determined with the required sophistication in terms of sensitivity and accuracy, and also comparatively quickly and at reasonable cost. The importance of subtle non-structural oil and gas traps was recognised by the petroleum industry, along with the difficulty of detecting them by conventional seismic exploration. The consequences of these developments in China were that petroleum companies, geological institutions and the Sinica Academy carried out a series of gas geochemistry research programmes on land in areas with different climate conditions and offshore. This chapter outlines the background, some of the methods that were developed and illustrates some of the results.
214
T. Ruan and Q. Fei i~~ :-!;r~;;, ~, ;::IF,i:iiz, :;::~=::r~: ::~:s: :,,~,:J:sa~;:::~.:~= ~::n= ~:!:!~ ~.~.~.:!~.~!i!i~~!!"y .~!~:~ !~~:~ !~.~: ~ ~, ~ ,~i~~': ~ ~!r.~: ~i~i~ ~ 9 ~ ~ r
SURFACEGEOCHEMICAL ANOMALIES~~ .....
i{
DESTRUCTION ~Ii!i
....... '............... -
-
..........................
I
DISPERSION I .......... ~ ................................... 1 .............................................
~
Fig. 6-1. Conceptual model ot'surtace gas geochemical anomalies.
'!'t IF.ORETICAI. PRINCIPI~I']S The basis for gas geochemical surveys for petroleum lies in modern concepts of petroleum genesis. Petroleum originates mainly from degradation and thermal cracking of disseminated organic matter in sedimentary rocks. Oil and gas fields are formed by the migration of the resulting liquid and gaseous hydrocarbons and their accumulation in favourable settings, termed traps. Accumulated petroleum is constantly dispersing laterally and vertically, albeit slowly, and this can lead to disappearance or destruction of the accumulation. Petroleum generation, migration, accumulation, dispersion and destruction are, to varying degrees, taking place simultaneously in a continuous evolutionary process. A surface geochemical anomaly can develop from any and all stages in this process, and should be considered as an integral part of the oil or gas field (Fig. 6-1). Conventional petroleum exploration tends to focus on the first three stages. Exploration geochemistry, on the other hand, concentrates on the dispersion stage. Surface macroseeps of oil and gas, which can be seen by the naked eye, led to the discovery of some of the most famous oil fields in the world (Link, 1952). Today, almost all of the visible macroseeps have been tested and drilled. Microseeps are the extension of macroseeps into the non-visible range. Their detection therefore requires the use of alternative methods, most obviously gas geochemical surveys.
Gas geochemistry surveys for petroleum
215
[ ~ c ~ ~oc.~,c~ ~~..~ ] DETECTION TECHNIQUES AND INSTRUMENTATION
iNDICATORS AND PARAMETERS
.-~
~0 ~
l"x
/
1_.., iDIAPIRISM AN] ICLINE LITitOLOGY FAUI.T -
I
PALEO- [ THRUSTS ttYDROCONF'ORMITYMORPitOLOGY DYNAMIC
~
""
ttYDROCARBON SOURCES
SHALLOW BIOGENIC GAS COALBED GAS
THERMAL GAS
OVER MATURED GAS
Fig. 6-2. l:actors that intlucnce surthcc gcochelnical patterns of hydrocarbons.
Whilst the execution of such surveys calls for appropriate methods of sampling and sufficiently sensitive analytical techniques, the planning and interpretation stages require a conceptual model to guide thinking. The model must be consistent with the principles of petroleum geology and geochemistry and provide a strategy for undertaking the survey and a reasonable explanation of the results. Since the very beginning of gas geochemical surveys, various models have been proposed; Price (1986) has given a critical review of the literature and proposed a working model which stresses the importance of surface conditions. Probably there will never be a perfect model, because there are so many interrelated factors influencing the relation between surface anomalies and their underlying oil and gas fields (Fig. 6-2).
216
T. Ruan and Q. Fei
Indicator gases
Microseeps comprise a number of different gases that are evolved as a product or by-product of the generation, migration, accumulation, dispersion and destruction of petroleum, along with other gases that follow the same migration route. Methane (C~) is by far the most abundant indicator gas in microseeps by virtue of its correspondingly high relative abundance in petroleum, its low molecular weight and its chemical stability. However, as an indicator of oil and gas it suffers interference from methane of near-surface biological origin (Philp and Crisp, 1982). Thus care must be exercised when interpreting the C~ patterns of gas geochemical surveys. Discrimination between thermogenic methane (from oil and gas) and near-surface biogenic methane is possible by means of carbon isotope determinations, but the method is relatively expensive. Ethane (C2), propane (C3) butane (C4) and pentane (C5) are useful for characterising the thermogenic origin of microseeps because they are less likely to form in the near surface by biological processes. Some higher molecular-weight hydrocarbons (C5 up to C22) are found in microseeps either because they possess appreciable vapour pressures or because they occur in the form of an aerosol, which essentially behaves as a gas. Data for a combination of the foregoing gases provide a fingerprint from which the origin and significance of surface anomalies can be deduced with greater confidence than is possible using only data for a single gas (Klusman and Voorhees, 1983). In addition to hydrocarbons, a number of inorganic gases that follow similar migration pathways are useful in gas geochemical surveys for petroleum, ltelium produced by the radiodecay of U and Th in rocks may accumulate in gas traps to thousands of times of its average concentration of 5.24 ppm in atmospheric air, and being light and inert, may escape to the surface. Mercury is absorbed from water by phytoplankton, the very raw material of petroleum" therefore Hg is intrinsically related to the generation of oil and gas. In remote sedimentary basins where Hg of igneous and anthropogenic origins is rare, Hg detection can prove useful in gas geochemistry surveys for petroleum exploration. Since uranium seems to concentrate in the low Eh oil-water contact zone in oil fields, radon and its daughters find applications in surface prospecting tbr oil fields. Some natural gas contains H2S and SO2, which are produced by bacteria attacking sulphates or by thermal alteration of amines (Hunt, 1979). These sulphur gases are therefore possible surface indicators of natural gas fields, but their ready solubility in groundwater leads to weak and unstable anomalies. Nitrogen is evolved during thermal maturation of organic matter, but N2 anomalies are difficult to detect in the near-surface because of the high background concentration of N2 in the atmosphere. Thus oil and gas fields are associated with many types of gases. During migration many of these gases, especially the hydrocarbons, react with their surroundings to alter the environment, usually by consuming free oxygen and producing more reducing conditions. The formation of pyrite, magnetite, siderite and other carbonates, as well as the proliferation of certain bacteria, are amongst the near-surface consequences of hydrocarbon gas migration. Whilst this chapter is concerned with the direct detection of
Gas geochemistry surveysfor petroleum
217
microseeps through gas geochemical surveys, detection of these alteration products by surface geophysical and microbial methods of prospecting is also possible, as, in some cases, is their detection by satellite remote sensing (Chapter 7). Indeed, gas geochemical survey data should be integrated with data obtained using these other methods and interpreted by combining all available geological, geophysical and geochemical data.
Gas migration Mechanisms of gas migration and the medium through which migration takes place should be considered simultaneously, although in practice they tend to be discussed separately. The mechanisms of gas migration can be divided broadly into three categories, with migration in each category governed by a well-established physical law: (1) diffusion, governed by Fick's Law; (2) buoyancy, governed by Archimedes' Law; and (3) mass flow, governed by Darcy's Law. These forces are not discussed in detail here; only a brief account is given of the contribution they make to the formation of surface gas anomalies. Diffusion, by definition, is a spontaneous tendency to eliminate a concentration gradient. Because it involves no external forces, it is a constant but slow process. On the whole, however, diffusion is a fringe force in terms of the speed and scale of gas migration, and other more rapid processes are superimposed upon it. Buoyancy originates from the density difference between water and oil or gas bubbles. MacElvain (1969) proposed an ascending-microbubble theory to explain the presence of microseepages. He noted that "colloidal-size gas bubbles are readily displaced upward by the surrounding water at rates up to several millimetres per second, regardless of any sedimentary particles that may intrude in the way of their upward zig-zag Brownian path. Such exceedingly small bubbles can quickly ascend hundreds and even thousands of feet in a manner not available to large gas bubbles or to individual gas molecules. Large bubbles have too large a surface area to be able to demonstrate kinetic or Brownian oscillation, whilst individual molecules of dissolved gas possess insufficient difference of density for gravitational displacement". This theory explains the sharp edges and limited lateral displacement of surface gas anomalies, their quick response to reservoir pressure changes, and the general absence of hydrocarbons heavier than C5 in surface anomalies. Mass flow is a bulk movement of hydrocarbons in monophase or dissolved in water. It requires an external force (pressure gradient or structural stress) and a well-defined conduit or plumbing system. The microseepage resulting from mass flow is therefore characteristically confined to a small area or belt, within which the anomalies are of high contrast and often contain some high molecular-weight hydrocarbons.
218
T. Ruan and Q. Fei
SURFACE EXPRESSIONS OF HYDROCARBON MIGRATION
Gas a n o m a l i e s
According to Kartsev et. al. (1959), Sokolov (1970), Horvitz (1986) and Fei and Ruan (1991), gas microseepage associated with petroleum has three characteristic styles of surface anomaly: (1) apical; (2) annular; and (3) linear. One or any combination of these surface expressions may occur over a given oil or gas field. An apical anomaly comprises either a continuous zone of elevated gas concentrations or an area with erratic elevated gas concentrations directly over the oil or gas field. An annular or halo anomaly has the form of a doughnut of continuous or discontinuous high gas concentrations surrounding a central zone of lower or background values, the latter overlying the surface projection of the oil or gas field. In a linear or belt anomaly, high gas concentrations are found continuously or intermittently along a line or confined in a belt, usually associated with faults, fracture zones, or matured source beds. Both apical and annular anomalies can be explained by diffusion models (although diffusion alone is unlikely to be responsible for them), whilst linear anomalies are readily attributable to mass flow through an elongated conduit, such as a fault or fracture zone. Some investigators have argued that diffusion can only lead to the formation of an apical anomaly (Price, 1986; MacElvain, 1969), but this argument is only valid if diffusion only takes place in a homogeneous medium. This is rarely the case in the natural environment and through inhomogeneous media diffusivity is anisotropic and variable up to several orders of magnitude. Indeed with certain combinations of media, annular anomalies can be obtained by diffusion (Ruan et. al., 1985a, 1985b). It is vital, but not necessarily easy, to recognise whether an anomaly is apical, annular or linear, because the interpretation placed upon each of them leads to quite different courses of action. The model used to explain an apical anomaly favours drilling the anomaly peak, whilst the model for an annular anomaly favours the area within the ring of high values. Follow-up of linear anomalies needs to take into account structural or lithological information. All types of anomalies can be present over a given oil or gas field depending on the relative depth of the field, its caprock lithology and structural control, the gas species in the microseepage, their modes of occurrence, near-surface lithologies and soil types, climate and even the time of year. This wide range of factors renders gas anomaly interpretation difficult.
Alteration
Migrating hydrocarbons may also interact with the rock column through which they pass. These interactions become most pronounced as the ascending hydrocarbons approach the surface and strive to equilibrate with changing conditions characterised by
Gas geochemistry surveys for petroleum
219
free oxygen and abundant bacteria. The chemical reactions that take place in this zone have been discussed by Rosaire (1938), Horvitz (1950), Kartsev et al. (1959) and Price (1986), who consider that the most probable reactions are: 9 oxidation of hydrocarbons (catalysed by bio-enzymes or clay minerals), accompanied by the production of CO2 and consumption of oxygen, which leads to a reducing environment and the formation of new minerals, often manifested by changes in physical properties and colour; 9 release of organic acids which may attack the common rock-forming minerals leading to increased concentrations of Ca 2+ and Mg 2+ in solution; and 9 precipitation of carbonates in which some of the migrating gases may be occluded, sometimes with the result that the rock becomes undulating. All these reactions are interrelated and evolving in time. Once they reach an advanced state, an alteration "chimney" is formed. This chimney, although formed through the microseepage of hydrocarbon gases, may itself obstruct their further migration. Any apical gas anomalies are subsequently lost. The microseeping gases find new pathways around the margins of the chimney and surface anomalies then develop above these new pathways. These anomalies are annular in outline. Thus, since the shape of a surface anomaly recorded in a gas geochemical survey may depend upon the presence or extent of an alteration chimney, integrated interpretation of gas survey data with geological observations and geophysical measurements is particularly important.
M()I)I,~S ()F ()CCURRI.~NCI" ()F (;ASI.'S IN MICR()SI'I:~PS The hydrocarbon gases found at surface may have different origins and migration histories, and hence different exploration implications. The origins and migration histories of gases can often be deduced from their modes of occurrence at surface. In particular, it is important to distinguish epigenetic concentrations from syngenetic concentrations, and this can often be achieved by careful research into the modes of occurrence of hydrocarbon gases in soils. In practice, the methodology for a gas geochemical survey is underlain by a rational selection of the appropriate mode of occurrence. It is possible to recognise four distinctly different ways (some with sub-divisions) in which gases occur at surface: (1) free molecules; (2) adsorbed molecules; (3) microbubbles; and (4) mineral constituents (Fig. 6-3).
220
o
T. Ruan and Q. Fei
o
.''.
Soil particle Adsorbed molecule
I [~
() "" 0 0 , o
o'~
Free molecule Interstitial soil Occluded gas Microbubbles
Fig. 6-3. Principal modes of occurrence of hydrocarbons in surface microseeps (from Ruan and Cheng, 1991 ).
Free m o l e c u l e s
Free gas molecules in the interstitial soil air (and atmospheric air) represent the most mobile and most recently-arrived portion of the gas. Usually, this portion accounts for less than 0.1% of the total amount of gas in the near-surface environment. Gases in the interstitial soil air can or should be determined either: (1) on-site, with a portable instrument or buried sampling device; or (2) in the laboratory after on-site pre-concentration onto an artificial absorbent. The advantage of this mode of gas occurrence is that the composition of the gases has remained virtually unchanged since their upward migration began. The disadvantage is that the anomalies are unstable therefore frustrating to the investigator (Devine and Sears, 1977). The instability may be reflected in diurnal, weather or seasonal variations. However, although absolute gas concentrations may prove irreproducible, the concentration ratios between individual constituents are repeatable over several years (Jones and Drozd, 1983). The eventual dispersion of free hydrocarbon gases from the soil air into the atmosphere provides the basis for airborne gas geochemical surveys, which are capable of covering large areas of difficult access (Barringer, 1981 ; Sandy, 1988).
A d s o r b e d molecules
Non-polarised gases such as methane and helium can be adsorbed onto soil particle surfaces via molecular attraction. This attraction is weak and the gases can be separated from the soil by gentle heating, pressure reduction, or the combination of these two procedures. This mode of gas occurrence is less influenced by meteorological conditions than the free-molecule mode, and so the results are more repeatable. The samples are easier to handle but the desorption procedures need to be strictly controlled: complete desorption is not necessary, but artifacts in the data due to variations in desorption
Gas geochemistry surveysfor petroleum
221
conditions must be avoided. As the nature of adsorption onto and desorption from soil is poorly understood, the procedures are basically empirical. Polarised molecules such as higher molecular-weight hydrocarbons can be adsorbed via electrostatic forces. Organic or inorganic colloids (humic matter and clay minerals) are electrically charged and possess high adsorption capacities. Hydrocarbon gas molecules adsorbed in this way can be released by the addition of more competitive molecules (e.g., dichloromethane) to substitute in the occupied sites.
Microbubbles Gases occur as microbubbles in soil moisture (MacElvain, 1969). Gases occur in this mode only in humid climates; where soils contain no moisture, as in arid climates, microbubbles are also absent. The concentration of gases in this mode of occurrence is much higher than that in the free molecule and adsorbed modes, and the gas composition is more representative of the sources. Provided soil samples are kept moist, microbubbles can be driven out by a stream of more active molecules and their gas composition determined. Again, absolute concentrations are undefined, and the determinations are relative.
Mineral constituents Gases occur as inclusions in secondary minerals, especially carbonates. Being protected by the host minerals from dissipation and microbial attack, hydrocarbons in this form are very stable. As a result, sampling and sample handling are simplified and the shelf life of samples is long, which is convenient for large surveys extending over long periods. The analytical procedures, which usually involve gas release by acid treatment, are easily standardised and data obtained at different times can be plotted on the same map (Horvitz, 1986). Also the gas composition may have altered considerably, usually becoming enriched in heavier hydrocarbons, which can lead to over-optimistic interpretations. Furthermore, the existence of carbonates of different origins may complicate the data. Soils derived from matured source rocks, particularly carbonates, always contain very high concentrations of hydrocarbons with compositions of matured oily sources. If the parent rock of the soil was transported by a surface agency such as water, wind or ice, gas anomalies reflect the source area of the sediments rather than the oil or gas potential of the underlying strata. The western part of the Bayinhaote Basin of west China is characterised by Archaean metamorphic rocks covered by thousands of metres of Cenozoic red clastic sediments with no oil or gas potential. Acid treatment of soils revealed very high C~-C5 concentrations (up to 600 ~tl/kg of C~ and 70 ~tl/kg C2). However, it was found that the anomalous soils were derived from carbonate fragments
222
T. Ruan and Q. Fei
of Palaeozoic age, transported by a modern river from an area of outcrop tens of kilometres to the west. Similarly, Price (1986) cites an example from the oil fields of western Alberta, where high concentrations of gases released by acid treatment from soils were attributed to carbonate fragments glacially transported from the Canadian Rocky Mountains to the west, and not to the underlying oil fields. Finally, gases are incorporated within the lattices of clay minerals, from which they can be extracted only by destroying the mineral structure. This mode of gas occurrence had attracted little interest because the lattice positions are likely to be syngenetic rather than indicative of underlying sources.
PRACTICAL METHODS For hydrocarbon gases, the concentration measurement tool is exclusively gas chromatography with a packed column and an FID detector, capable of precisely determining 10.`7 levels of C~-C5 in less than 5 minutes. Heavier hydrocarbons are sometimes determined using a quadrupole mass spectrometer. As these instruments and the techniques for loading gas samples onto them are described elsewhere (e.g., Chapter 5), only field methods and, where applicable, sample pre-treatment methods for releasing gases from soils are discussed here.
Soil air
Interstitial gases in soil air are first extracted by driving a probe into the soil and sucking the soil air into a measurement device or a pre-concentration medium. A variety of procedures have been reported (Devine and Sears, 1977; Jones and Drozd, 1983; Richers and Jones, 1986). The main operational parameters are the probe depth, isolation of the soil air from the atmospheric air, the amount of negative pressure applied, the protection of the probe from blockage and the reduction of the dead volume of the tubing. These have to be optimised in terms of both cost and effectiveness. Of course it is not possible to extract interstitial soil air from waterlogged soils. If the soil air is not pumped directly into an instrument for in situ determination of gas concentrations, it is passed through an artificial adsorbent in order to pre-concentrate the gases to be determined later in the laboratory. Several kinds of artificial adsorbents have been tested and activated charcoal has been found to be the most suitable, although it is not entirely satisfactory for C~ and C2. The activated charcoal is commonly used in the form of a thin film coated on a ferromagnetic wire (Klusman et al., 1986) or in the form of commercially-available fine-grained activated charcoal. Any adsorbent should be pre-treated in an inert environment (vacuum or N2) at 400~ to release all possible adsorbed gases before use. After sample absorption in the field, the adsorbed gases are
Gas geochemistry surveys for petroleum
223
evolved in the laboratory by heating to high temperatures and their concentrations are determined instrumentally.
Soil Soil samples are usually collected at the surface proper or at depths of up to several metres below the surface. As far as possible, samples are kept in their original condition until they are prepared for analysis in the laboratory. The analytical procedures are invariably partial extraction techniques. This approach has been widely used in other branches of exploration geochemistry, but assumes particular importance in gas geochemical surveys for petroleum. For gases adsorbed on soils, the essential requirement is to regulate desorption parameters, so that variations in these parameters do not introduce variations into the concentrations desorbed. In order to prevent production of new gas species during the thermal desorption, the heating temperature is usually kept under 200~ or applied only for milliseconds (Kiusman et al., 1986). Stable thermal desorption conditions can be attained by using a metal heating block maintained at a fixed temperature, into which a tube containing the soil sample is placed for a fixed time. The gases released are then condensed in a loop held in liquid N2, which subsequently acts as the input to the gas chromatograph (Fig. 6-4). If the soil samples are in sealed containers, simple vibration at elevated temperature produces an equilibrium concentration of gases in the container headspace that can be taken as a measurement proportional to the concentration of adsorbed hydrocarbons. For the extraction of hydrocarbons in carbonate minerals and microbubbles, the method widely used is acid treatment under vacuum, for which most laboratories in China use the apparatus shown in Fig. 6-5. The soil sample is heated to 40~ in tile flask, then 1:6 HCI is added until its reaction with the warm sample ceases. The mixture of gases evolved is filtered through NaOH solution to remove CO2 (which makes up the bulk of tile evolved gases) and the remaining gases are injected into a gas chromatograph. A selective method to flush out microbubbles is hydrogen stripping. This was first proposed by Schaefer et ai. (1978) and has been developed by Ruan and Cheng (1991). Soil samples are kept in a natural wet state and an aliquot of saturated NaCI solution is added to make a slurry, through which H2 is passed. The gases stripped from the sample are condensed in a liquid N2 cold trap, which is then used as the gas sample loop in the carrier gas line to a gas chromatograph (Fig. 6-6).
224
T. Ruan and O. Fei
Carrier gas "I
Gaichr~176 /
6-portvalve Pump
__ Sample loop Liquid nitrogen
Fig. 6-4. Apparatus for dcsorption of hydrocarbons from soil by elevated temperature and reduced pressure.
N2
H2
va'ves6-port valve,,,
l
) '
02
9
I FID 1 ,P~'-I
Outlet Sample loop Liquid nitrogen Sample slurry
......
1
l:ig. 6-5. Apparatus for release of hydrocarbons l'ronl soil by acid treatment under vacuum and elevated temperature (from Ruan and Chcng, 1991 ).
Gas geochemistry surveysfor petroleum
Pump ~ V3
HCi
Gauge Vl(~ I'T" ~
225
w........ Septum
NaOH ~
v4
II
V2 Sample ......
H-FWater bath
'
i:ig. 6-6. Apparatus lbr flushing hydrocarbons t'ronl soil by hydrogen stripping.
CASI'~ ! IIS'I'()RIES
Ordos Basin Tile Ordos Basin is located about 300 km to the north of Xi'an, central China. With an area of 260,000 km 2, it is the second largest sedimentary basin in China. Its central part is characterised by nearly horizontal strata some 4,000 m thick. The sedimentary column can be divided into three major units: Lower Palaeozoic marine carbonates: Upper Palaeozoic coal-bearing series: and Mesozoic terrestrial sediments. Oil and gas fields occur throughout these sediments. The gas is believed to originate from over-matured Ordovician carbonates and coal beds, whilst the source of the oil is thought to be lacustrine sediments of Triassic age. The region is semi-arid and the soils are thin and poorly developed. Jingbian is located in the central part of the Ordos Basin, where the oil and gas potential has long been considered to be poor because favourable structures are not evident in the seismic data. In 1988, however, the Changqing Oil Company proposed a scientific research well. This was preceded by a surface gas geochemical survey along several traverses, one of which passed through the site of the proposed well. The surface material, which is mainly aeolian sand, dry river-bank sediments and loess, was sampled at depths of 2-3 m at intervals of one km along traverse lines. At each sample site, Hg in
226
T. Ruan and Q. Fei 5
................................................................................................................................................
25 "r" 0 60 00
80
100
(m)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
60
80
100
(m)
1000 ........................................................................................................................................... CI _,.,.~500 " 0
disco
B
0
60
80 eolian sand
100 river sediments
-J (m)
eolian sand
Fig. 6-7. Geochemical traverse of C~, C2+ in soils and Hg in soil air, Jingbian gas field, Ordos Basin.
soil air was determined in the field using a portable gold-film Hg detector. The soil samples were analysed in the laboratory for C,-C5 using the acid treatment method described above. The results from the traverse over the site of the proposed scientific research well (Fig. 6-7) aroused considerable interest amongst geochemists and geologists. There are pronounced anomalies in the traverse but they are apparently reflections of variations in lithology: hydrocarbon gases are low in aeolian sand and high in bank sediments; Hg in soil air is low in bank sediments but high in the flanking aeolian sand. However, there were several factors suggesting that the anomalies were significant and indicative of an underlying gas accumulation. First, examination under a binocular microscope revealed that the bulk of the samples is composed of rounded quartz grains, implying long-distance transport. Adhering to the surfaces of these quartz grams in samples with high values of hydrocarbons were tiny well-crystallised rhombic minerals; these were dissolved by acid treatment, indicating that they were carbonates. Carbonates are less resistant than quartz, and would not have adhered to the surface of the quartz grams during long-distance transport; therefore the carbonates formed in situ. Furthermore, the source material of the river bank sediments is loess that contains 300-400 ~tl/kg of C, and 20-30 ~tl/kg of C2+, much lower
Gas geochemistry surveysfor petroleum
227
concentrations than those found in the anomalies, so hydrocarbons must have been added. Both of these observations can be explained by ascending hydrocarbons being trapped during the formation of secondary carbonates, and released and detected by the acid treatment method used to analyse the samples. The river valley, filled with loose sediments, might follow a good conduit for ascending hydrocarbon gases. Still other factors point to the significance of the anomalies. The hydrocarbon anomalies in the river bank sediments include the complete range from C1 to C5, which excludes the possibility of the hydrocarbon anomaly being of surface biological origin. The aeolian sand is a comparatively poor medium for the development and detection of hydrocarbon anomalies in soil, but is highly suitable for the detection of Hg in soil air, for which there is no obvious source other than natural gas at depth. Subsequently the scientific research well was drilled as planned. When it penetrated Lower Ordovician carbonates it produced 16.3 x 104 m 3 per day of natural gas. The scientific research well thus became the discovery well in this region. Since the well was drilled directly on the surface hydrocarbon gas anomaly, the soil gas survey traverse was thought to cross an apical or linear anomaly. Further work, however, was to reveal that the gas field has an annular anomaly. Encouraged by the gas discovery, the Changqing Oil Company decided to explore the full extent of the field. In 1990, a 5035 km 2 area was covered by a gas geochemical survey with a target sample density of 1-2 per km 2 which yielded a total of 6162 sample sites. At each site C~ and He were measured in the interstitial soil air, and soil samples were analysed in the laboratory by the acid treatment method for Cj-C5 and by thermal desorption at 250~ for Hg. The random data points were gridded by inverse distance squared weighting, smoothed by means of a 3 x 3 moving average and contoured. The resulting hydrocarbon gas patterns and Hg pattems are presented in Figs. 6-8 and 6-9, respectively (which also show the location of the traverse discussed above and shown in Fig. 6-7). The regional pattems comprise an arc of anomalous values extending from the southwest to the northeast. This arc is interpreted as part of an annular anomaly that is still open to the north and west. The exploration implication is that a gas field underlies the low values in the northwest of the survey region. In fact the earlier gas survey traverse and the scientific research well that became the discovery well were already located in this part of the region. Subsequently the Changqing Oil Company drilled more than twenty wells, most of them commercial, in the prospective area (Figs. 6-8 and 6-9). The highest production from a well in this area is more than 120 x 104 m 3 per day. The gas reserves in the Jingbian field are likely to exceed 10 ~ m 3 as the field is not yet closed off to the north and west by surface exploration. The gas production layer is a Lower Palaeozoic karstic plateau at a depth of 3000 m, surrounded by palaeo-valleys filled with a Permo-Carboniferous coal series (which could be the source of the natural gas). The annular surface geochemical anomaly may be related to the edge of the buried plateau. In the production layer, the gas is under a pressure of 120 atmospheres. Therefore the average pressure gradient between the gas
228
T. Ruan and Q. Fei
iii!i!i!ii!ii!!i!ii~iii!!!ii!!ii!i~iiiii!!!!ii~!i!iii!iiiiiiiiiiiii!ii~~ ~i~!~:!~i~iiii:::i~iii!~i~!~i~i~i~ii~iiiiiiii~sii~ii~i~i~i~ ::~: ii!!~;!i )il;!i~iii~ii~ilil...... ::::~.... ::::::iii~iiii~ilili!!!!!!!iiiiii~,
!!i!!?!??!!?!!?!ii?i?!!iii!?i~iiii[iiiiiii;2!!!!!!!!!ii?::
l:::l Oos W,li
0
r
(~)c,:>,oo.cl<.~oo
~
cl>;~OOui/i
O,
lOkml
Fig. 6-8. Contour map of C~ in soils, Jingbian gas field, Ordos Basin.
iiiiiiiiiiiii!iiiii',iiiiii~ii!~i~!!i~ii~!ii',i',',!!i!ii!~!~i~!~!!~i~
"
~Di,00w~y~Wei ii ! : !i !? ili i i~!~ ~: if:! ;:~ ~ ,: ~......
A i:::::i,l.i-i:!!.i-:.::.~i:::i:iii:~ii~iii~iiiiii::i!: ::i:: :~,:
~2~Gas Well
0
Hg<8 ppb
~)
9..... " ....
. . . . .
Hg >8 ppb
:i~: ~::i::iii i! i~?ii~i::
O,
_ l Okra!
Fig. 6-9. Contour map of Hg in soils, Jingbian gas field, Ordos Basin.
field and the surface is 0.04 atm/m. This pressure gradient could be the main driving force for vertical gas migration. The Pinqiao oil field occurs in sandstone lenses of Triassic age, at a depth of less than 1000 m. A gas geochemical survey was carried out over a 1500 km 2 area in order to test for microseepage from the known oil field and detect any neighbouring fields. Soil samples were collected on a 1 x 1 km grid and analysed by the acid treatment method for
CI. The surface material in the region is loess that is quite homogeneous and possesses high concentrations of syngenetic hydrocarbons (450 ~tl/kg of C!). This has the effect of
Gas geochemistry surveysfor petroleum
Q Cl>400ul/kg
@ Cl>600ul/k9
229
0 I
1Okra ,I
Fig. 6-10. Contour map of C~ in soils, Pinqiao oil field, Ordos Basin.
making the anomaly-to-background contrast very low, so the contour intervals used to map the results have to be selected carefully. Using contours at 480, 500 and 520 ~tl/kg of C~ the known oil field is clearly outlined by the discontinuous patches constituting a low relief annular anomaly (Fig 6-10). A small area in the southeast comer of the survey area was also interpreted as an annular anomaly, and the presence of oil there has since been confirmed by drilling.
Lixian Depression The Lixian Depression, some 500 km south of Xi'an, is filled by about 10,000 m of Palaeozoic and Mesozoic sediments. Mature source rocks occur throughout the stratigraphic sequence, but the complicated geological structure has impeded exploration and there are as yet no commercial discoveries of oil or gas. Located in southern China, the region has a very wet climate and is largely covered by rice fields, so that soils in the region are invariably moist.
230
T. Ruan and Q. Fei
~
\ "~\ "-.:. ~ i / "
)
/ "
~] ~:~ \16,0, - x ' Uv-
\
0,
\
.i
/I
/
I
-
/
~-~
'-- 1600
f
:-~
\
..~/ ./ /q.,~/ \N / / / (lalk g-1 ) \ ""
1, km
L~
I
t
~/"
/
Fault
[----7] Cz + bF i,hydrocarbonsL.__~ ]T9 depth (m)
Fig. 6-11. Contour map of C2+ in soils determined by acid treatment method, Lixian Depression (from Ruan and Cheng, 1991).
-"
~
R_ing anomaly
@
~Hydrogen stippingC,
Fig. 6-12. Contour map of C 4 in soils determined by hydrogen stripping, Lixian Depression (from Ruan and Cheng, 1991).
Ruan and Cheng (1991) describe two gas geochemical surveys in the region. In the first survey, a pronounced C2+ anomaly was found in soils analysed by the acid treatment method (Fig. 6-11). In a subsequent survey at a higher sampling density, soil samples were analysed for hydrocarbon gases by the hydrogen stripping method. The results for C4, shown in Fig. 6-12, confirm the general patterns of the earlier survey and
Gas geochemistry surveys for petroleum
231
additionally reveal an annular anomaly in the west of the area and a linear anomaly running southwest-northeast across the area. Seismic data have demonstrated three sub-surface southwest-northeast trending faults in the area. The annular gas anomaly partially overlies an uplifted block, a structurallyfavourable setting for an oil or gas trap. The linear anomaly follows the trend of another of the subsurface faults.
CONCLUSIONS Gas geochemical surveys of light molecular-weight hydrocarbons (C~-C5) and Hg have proved to be useful methods of detecting oil and gas fields in the Ordos Basin of northern China and of indicating potential oil or gas fields in the Lixian Depression of southern China. The annular anomaly is the type most usually found in these regions, although apical and linear anomalies occur under particular conditions. If gas geochemical surveys are to be applied effectively, proper attention should be paid to the mode of occurrence of gases in the soil. Using available geological and geophysical data, a microseepage migration model should be developed for each survey area, and the sampling and analytical techniques should then be chosen to suit the model. Lithology, soil type and surface conditions should also be considered. These can play important roles in the distribution and intensity of anomalies, as the pre-discovery traverse over the Jingbian gas field demonstrates. Finally, if the entire field is not covered by the gas geochemical survey, anomalies may be missed or their shape and significance misinterpreted.
This Page Intentionally Left Blank
Geochemical Remote Sensing of the Subsurface Edited by M. Hale Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
233
Chapter 7
AEROSPACE DETECTION OF HYDROCARBON-INDUCED ALTERATION H. YANG, F.D. Van der MEER and J. ZHANG
INTRODUCTION The continuing development of remote sensing has been of considerable significance to earth scientists in general and exploration geologists in particular. Beginning in the 1930s, aerial photography was used to map structure and stratigraphy. Since the 1970s, imagery from satellite platforms equipped with the multispectral scanning system (MSS), the thematic mapper (TM), SPOT and radar systems has been used to obtain synoptic views of the geology of an area and conduct basin-wide assessments (Halbotay, 1980; Moore and Anderson, 1985). Airborne platforms carrying multispectral scanning systems and imaging spectrometers, have been used routinely to provide imagery for geological mapping and identification of prospect-scale structures. The use of sidelooking radar images and synthetic aperture radar (Goetz and Rowan, 1981; Froidevaux, 1980) provided the possibility to enhance subtle expression of subsurface geologic structures. In general the key to the successful application of remote sensing imagery in hydrocarbon exploration is its integration with other exploration data such as seismic, well, gravity and magnetic data. Researchers have also attempted to use remote-sensing imagery to detect the distinct spectral characteristics of surface manifestations of hydrocarbon microseepages, originating from oil and gas reservoirs at depth. This chapter provides an overview of the background to the direct detection of onshore hydrocarbon microseepage by remote sensing techniques and the results achieved to date.
Hydrocarbon microseepage The occurrence at surface of hydrocarbon seeps suggests that an oil or gas reservoir leaks even though it acts as a trap for hydrocarbons. Macroseepage is the visible presence of oil and gas seeping to the surface. Macroseeps have been documented in various parts of the world (Davidson, 1963; Sittig, 1980; Hunt, 1981; Davis, 1967; Tedosco, 1995). Microseeps are invisible trace quantities of hydrocarbons seeping to the
234
H. Yang, F.D. Van der Meer and J. Zhang
surface. The most persuasive evidence for microseepage is the measurement, sometimes over many years, of statistically-significant anomalous amounts of light hydrocarbons in soil gases and soils directly over oil and gas reservoirs (Price, 1986). In these cases the hydrocarbons in the soil gas or soil have very similar carbon isotope ratios to those in the underlying reservoirs, whereas the hydrocarbons of near-surface biogenic origin have different carbon isotope ratios. There is also good compositional correlation between the hydrocarbons of a microseepage and those in the underlying reservoir (Saunders et al., 1991). The occurrence of hydrocarbon microseepage directly above reservoirs points to vertical migration of hydrocarbons, despite the fact that groundwater movement might be expected to militate against this. Indeed, the cross-sectional shape of the hydrocarbon leakage pattern has been termed a "chimney", and most chimneys are nearly vertical (Tedosco, 1995). Vertical migration through the strata has been attributed to at least four mechanisms: effusion; diffusion; solution; and gas bubbles. Effusion as free hydrocarbon gases is thought to be the principal mechanism leading to macroseepage. It arises as a result of the very large pressure differential that exists across a petroleum reservoir. Diffusion of hydrocarbon gases that are usually dissolved in groundwater has been observed through seemingly impermeable barriers (Rosaire et al., 1940). This form of migration is thought to contribute to microseeps. Also dissolved low molecular weight hydrocarbons in groundwater migrate through capping shales as a result of hydrodynamic or chemical potential drive (Duchscherer, 1980). Vertical ascent of ultra-small (colloidal size) gas bubbles through a network of inter-connected, groundwater-filled microfractures is advocated by Price (1986). Buoyant colloidal gas bubbles are readily displaced upward at rates of up to several millimetres per second. This fast ascent explains the rapid development of light hydrocarbon anomalies in soil gas over newly-filled gas storage reservoirs, and their rapid disappearance after a reservoir is depleted. Although microseeps (and macroseeps) represent leakage from a temporarily stationary source of petroleum, they do not necessarily indicate the presence of economically-recoverable hydrocarbons at depth. The economic viability of the underlying reservoir can only be established by further exploration.
Induced surface manifestations o f microseepage The surface manifestation of hydrocarbon microseepage in not necessarily confined to the presence of trace quantities of hydrocarbons. Schumacher (1996) made a thorough review of the major hydrocarbon-induced changes affecting soils and sediments and their implications for surface exploration. Schumacher (1996) contended that long-term leakage of hydrocarbons can establish locally-anomalous redox zones that favour the development of a diverse array of chemical and mineralogical changes. The bacterial oxidation of light hydrocarbons can
Aerospace detection of hydrocarbon-induced alteration
235
Fig. 7-1. Surface alterations caused by migrating hydrocarbons (reproduced with permission from Duchscherer, 1982).
directly or indirectly bring about significant changes in the pH and Eh of the surrounding environment, thereby influencing mineral stability and chemical reactivity. Such oxidation in the chimney above a leaking petroleum accumulation leads to dissolution or precipitation of minerals and the mobilisation or immobilisation of certain elements in the chimney, which thereby becomes mineralogically and chemically different from laterally-equivalent rocks (Pirson, 1969; Oehler and Stemberg, 1984; Price, 1986). The resulting alteration includes: the formation of calcite, pyrite, uraninite, elemental sulphur, and certain magnetic iron oxides and iron sulphides; bleaching of red beds; clay mineral alteration; electrochemical changes; radiation anomalies; and biogeochemical and geobotanical anomalies (Fig. 7-1). Where such changes can be measured and mapped at surface, they provide the basis for a number of surface exploration methods for petroleum. Some of these changes are in principle amenable to measurement and mapping by remote sensing techniques.
REMOTE DETECTION OF INDUCED SURFACE MANIFESTATIONS Remote sensing has the potential to detect hydrocarbon-induced alteration in rocks, soils and vegetation. Extensive studies have been performed on the reduction of ferric iron (red-bed bleaching), the conversion of feldspars and mixed-layer clays to kaolinite, the increase of carbonate content and the anomalous spectral reflectance of vegetation.
236
tl. Yang, F.D. Van der Meet and,1. Zhang
The attraction of remote sensing is that it offers a rapid and cost-effective means of conducting reconnaissance for hydrocarbon-induced alteration.
B l e a c h e d r e d beds
The presence at surface of bleached and discoloured red sandstones above petroleum accumulation has been widely noted, but detailed studies are few. Bleaching occurs whenever acidic, reducing fluids dissolve the ferric oxide (hematite) that gives the red bed its characteristic colour. Reducing conditions also favour the formation of pyrite and siderite from the iron that is released during the dissolution of hematite. Leakage from petroleum accumulations of reducing agents such as hydrocarbons, H2S and CO 2 could be responsible for bleaching overlying red beds (Schumacher, 1996). The reflectance characteristics of various ferric and ferrous iron minerals, clay minerals and calcite are shown in Fig. 7-2. Ferric iron (in hematite) exhibits its strongest reflectance at wavelengths greater than 1.0 pm; at progressively shorter wavelengths there is first a distinct absorption feature at 0.9 [am, then an increase in reflectance at 0.8 ~tm and finally, at still shorter wavelengths, reflectance falls off sharply (Hunt et al., 1973). On the other hand, the ferrous iron in non-transparent minerals such as pyrite and magnetite shows a near-uniform low total reflectance, although transparent minerals such as siderite have broad shallow reflectance at 1.0-1.1 ~tm (Hunt, 1970). These
'-
~
_,
80
'
""9 ~
~
80
,o
""~ ~ "", 9
:-
---..
Kaolinite
.....
Calcite
:"-''.
- .... i"..
Montmorillonite
.f./
.......... lURe
.
.
20
o 40(
'''
"T"l"l
"r "1" ]'" I'" I " l " l 1000
"] "l" "i" i ' l 1500
I"1 2000
I
Hematite
---
Pyrite
....
Magnetite
2500
Wavelength (nm)
Fig. 7-2. Spectra of minerals associated with hydrocarbon microseepage.
Aerospace detection of hydrocarbon-induced alteration
23 7
characteristics can be used in remote sensing data-processing to separate bleached red beds from their unbleached equivalents. In the Cement oil-field area of Oklahoma, Donovan (1974) reported that the colour of the Permian Rush Springs formation grades from reddish-brown for unaltered sandstone adjacent to the field, to pink, yellow and white along the flanks of the Cement anticline, then to white along the flanks of the anticlinal axis. The white colour reflects maximum bleaching and maximum iron loss. Similar changes are observed at the nearby Velma, Eola, Healdton, and Chickaska fields (Ferguson, 1979a, 1979b; Donovan et al., 1981). Townsend (1984) was able to detect the red-bed bleaching at the Velma oil field using an infrared colour-composite and a colour-ratio composite of Landsat 4 TM imagery (Table 7-I).
TABLE 7-I Landsat 4 TM imagery used to detect bleached red beds at Veima, Oklahoma (compiled from Townsend, 1984)
Band
Colour composite Window (~tm) Colour
Bands
Ratio composite Windows (I.tm)
Colour
1 4 5
0.45 - 0.52 0.76 - 0.90 1.55 - 1.75
3/4 5/ ! 5/ 7
0.63 - 0.69 / 0.76 - 0.90 1.55 - 1.75 / 0.45 - 0.52 !.55 - 1.75 / 2.08 - 2.35
Blue Green Red
Blue Green Red
In the Sheep Mountain anticline of the Bighorn Basin, Wyoming, areas of red-bed bleaching within the Chugwater Formation correspond spatially with known hydrocarbon deposits. Malhotra et al. (1989) outlined bleached areas from Landsat TM imagery by a decrease in the ratio of bands 3:1 in conjunction with an increase in total reflected radiance. In the Lisbon Valley field of southeastern Utah the distribution of the bleached outcrops of the Triassic Wingate formation approximates the geographic limits of the oil and gas reservoirs at depth. The red colour of the unbleached Wingate results from a pervasive hematite-clay mixture which coats virtually all sand grains, whereas the bleached Wingate is white or grey due to the absence of these hematite grain coatings. Some hematite is present in the bleached Wingate, but as pseudomorphs of pyrite and siderite rather than as grain coatings (Segal et al., 1984, 1986; Conel and Alley, 1985). Segal and Merin (1989) used the ratio of Landsat TM bands 2:3 to delineate variations in ferric iron content and applied density slicing to map the bleached rocks.
238
H. Yang, F.D. Van der Meer and J. Zhang
Kaolinisation
The acidic conditions resulting from the oxidation of hydrocarbons in near-surface soils and sediments promotes the diagenetic weathering of feldspar to clay and the conversion of smectite clay to kaolinite. The kaolinite thus formed remains chemically stable unless the environment is changed (Schumacher, 1996). Kaolinite exhibits a very strong absorption feature centred at 2.2 ~tm along with a subordinate absorption feature at 2.16 ~tm (Fig. 7-2), forming a diagnostic doublet. This can be picked out in remote-sensing imagery and used to indicate areas enriched in kaolinite. According to Segal et al. (1984, 1986) the bleached portions of the Wingate Sandstone directly overlying the Lisbon Valley oil field in Utah contain three to five times more kaolinite than the unbleached sandstone located away from the field, which contains more plagioclase and muscovite. Using Landsat TM imagery, Segal and Merin (1989) found that kaolinite-poor unbleached sandstone exposures have relatively low ratios of bands 5:7, moderately kaolinite-rich sandstone (e.g., in an area known as Three Step Hill) have higher ratios of bands 5:7 and kaolinite-rich bleached sandstone exposures that overlie the Lisbon Valley field exhibit the highest ratios of bands 5:7.
Carbonate enrichment
The formation of diagenetic carbonates and carbonate cements, especially porefilling and replacement calcite, are amongst the most common alteration features induced by hydrocarbon microseepage. These carbonates are a product of the oxidation of hydrocarbons such a methane to carbon dioxide, which in groundwater hydrolyses to bicarbonate anions. Dissolved calcium (and magnesium) in groundwater reacts with this bicarbonate to precipitate as carbonate minerals or carbonate cement. One of two reaction pathways applies, depending in the redox conditions, viz., aerobic anaerobic
CH4 + 202 + C a 2+ = CaCOs CH4 + SO42- + Ca 2+ = C a C O 3
+ 2H + HzS + H20
+ H20 +
The resulting accumulation of carbonate at or near surface can be exploited as an indicator of a hydrocarbon reservoir at depth (Patton and Manwaring, 1984; Duchscherer, 1982; McDermott, 1940; MacElvain, 1963). The carbon in this carbonate carries the isotopic signature of its parent hydrocarbon(s). The carbon of most carbonate minerals is derived from the atmosphere, freshwater or the marine environment and has a ~3C isotopic value of about -10 to +5 per mil relative to the PDB standard (Fairbridge, 1972; Anderson and Arthur, 1983). The 13C content of most crude oil ranges from about -20 to -32 per mil, and that of methane ranges from -30 to -90 per mil. Thus carbonate formed from hydrocarbon oxidation
Aerospace detection of hydrocarbon-induced alteration
239
incorporates carbon that typically has a 13C content more negative than -20 per mil. Depending on the proportion of carbon derived from hydrocarbon oxidation, the 13C content of the resultant carbonate can range from-10 to-60 per mil (Schumacher, 1996). Carbonates with isotopically anomalous carbon derived in part from hydrocarbons have been reported from the Cement, Chikasha, Velma, and other southwestern Oklahoma oil fields (Donovan, 1974; Lilburn and Alshaieb, 1984), the Ashland gas field in the Arkoma basin of southeastern Oklahoma (Oehler and Stemberg, 1984), the Recluse oil field in Wyoming (Dalziel and Donovan, 1980), the Gulf of Alaska (Barnes et al., 1980), the Davenport oil field in Oklahoma (Donovan et al., 1974), the Ocho Juan field in Texas, the Fox-Graham field in Oklahoma (Duchscherer, 1984), and from the carbonate hardgrounds formed around modem gas seeps on the California continental rise (Paull et al., 1995) and near Fredrikshavn, Denmark (Dando et al., 1994). These isotopically anomalous carbonates are often present in both on-field and off-field wells, but their ~3C content is more anomalous (more negative) on-field. Remote sensing cannot detect the isotopic signature of carbonates, but it can detect the increase in carbonate formation and carbonate cement that hydrocarbon oxidation induces. Ground truth investigations can subsequently establish if this carbonate is indeed isotopically anomalous. In relatively arid regions, such as west Texas, carbonates may be visible on aerial photographs or on satellite imagery as light tonal anomalies indicating an excessive development of caliche in surface soils (Thompson et al., 1994). On the Landsat TM imagery of the Hugoton gas field in southwestern Kansas, Patton and Manwaring (1984) found a tonal anomaly that corresponded to a slight increase in calcium content. Carbonate enrichments can also be determined by selecting appropriate wavelengths sensitive to variations in absorption and reflectance of calcite, which are most pronounced in the 1.8 to 2.6 ~tm range (Fig. 7-2). The remote sensing bands used to map carbonate (calcite) concentration are typically those at 1.8 ~tm, 2.0 ~tm, 2.16 ~tm, 2.35 ~tm and 2.55 ~tm (Hunt, 1971). Laboratory spectra of brown sandstone from the Palm Valley gas field of the Amadeus Basin, central Australia, show only about half the brightness of laterallyequivalent red sandstone, with a reflectance maximum at 1.85 ~tm. Laboratory spectra of calcrete, with a weak to moderate absorption feature at 2.32 ~tm caused by the carbonate (calcite) content, are similar to those of the brown sandstone (but with increased brightness towards the visible wavelengths), suggesting that the colour change of the sandstone is due to the addition of carbonate. Based on these field spectral reflectance data, a distinctive colour discrimination was obtained by digital image processing of NASA NS-001 aircraft-borne thematic mapper simulator data. Using calibrated ratios of bands 3:7 (in blue), 1:5 (green) and 7:6 (red), carbonate-enriched areas are picked out in yellow. The carbonate can also be identified as a pale-dark magenta colour using uncalibrated Landsat TM images with ratios of bands 4:1 (in blue), 7:1 (green) and 3:1 (red) (Simpson et al., 1991). In the Junggar Basin, Xinjiang, China, airborne short-wave infrared split spectral scanner data revealed an anomaly that proved to reflect a major increase in total
240
1-1. Yang, F.D. Van der Meer and J. Zhang
carbonate in soil. Four channels, at the wavelength ranges of 1.550-1.650 ~tm, 1.9852.085 ~tm, 2.037-2.137 ~tm and 2.039-2.193 ~tm, were used for anomaly extraction. The anomaly overlies a subsurface heavy-oil reservoir (Zhu and Wang, 1991).
Vegetation stress
Hydrocarbon microseepage creates a reducing environment in the soil and overburden at depths shallower than would be expected in the absence of microseepage. The presence of hydrocarbons stimulates the activity of hydrocarbon-oxidising bacteria, which decreases oxygen content of the soil whilst increasing its contents of carbon dioxide and organic acids. These changes affect pH and Eh of soil, which in turn affect the solubilities of elements that are plant nutrients and consequently their availability to vegetation (Schumacher, 1996). This may affect the root structure of vegetation and ultimately influence its vigour and hence its spectral reflectance (Feder, 1985). Remote sensing of anomalous (or stressed) vegetation takes two forms. One is the mapping of the distribution of different species of vegetation and the differences in vigour and morphology within each species (Brooks, 1972; Siegel, 1974). Vegetation that is typically prolific is often stunted or absent in areas of unusual soil environments. On the other hand, some species thrive in environments that are toxic to most other species and are recognised as geobotanical indicators. The second approach is to determine differences in spectral characteristics between healthy and stressed vegetation. The spectral signatures of vegetation associated with hydrocarbon microseepage have been extensively studied. The main targets of attention are the green peak (at 0.56 ~tm), the red trough (at 0.67 ~tm), the shift of position of red edge and the height of the infrared shoulder. In both cases, normal or background variability in the distribution and vigour of various species presents complications that need to be taken into account. Factors such as bedrock geology, soil type, slope, soil moisture and climate can have a more pronounced effect than that due to the presence of hydrocarbons (Rock, 1984; Klusman et al., 1992). Nevertheless, numerous accounts have been published of the detection of hydrocarbon-induced vegetation anomalies by remote sensing. McCoy and Wullstein (1988) analysed leaves of sagebrush and greasewood from the Blackburn oil field in Nevada and reported a halo anomaly of high Mn:Fe ratios surrounding the productive part of the field. McCoy et al. (1989) revisited the Blackburn field and determined that the spectral reflectance of sagebrush from the anomalous area was lower than that of sagebrush from background areas. Bammel and Birnie (1993) evaluated reflectance in the visible and near infrared regions (0.45 to 1.1 ~tm) of sagebrush in five areas in the Bighorn Basin of Wyoming to determine its usefulness in hydrocarbon exploration. The most effective indicator of hydrocarbon-induced stress in sagebrush proved to be a consistent blue shift (to shorter wavelengths) of the green peak and red trough. This shift is only detected in areas where sagebrush is prolific and
Aerospace detection of hydrocarbon-induced alteration
241
perhaps in areas of profuse vertical microseepage. Spectral reflectance intensity was found to have no significant correlation with the presence or absence of surface or subsurface hydrocarbons. In the Stoney Point oil field, Michigan, Fluorescence Line Imager (FLI) data, covering the spectral range of 0.430 to 0.805 ~tm in 288 channels, were collected in both spatial mode and spectral mode. With the advantage of the high spatial and spectral resolutions of the two modes, three features of vegetation were detected: (1) a decrease in the height of the infrared shoulder (Rs) due to structural damage (Rock, 1988); (2) an increase in the reflectance at the chlorophyll absorption maximum (Ro) due to decreased leaf chlorophyll (Singhroy et al., 1986); and (3) a shift in the position of the red edge (~,p) towards shorter wavelengths (Fig. 7-3). A stress image with low digital numbers representing anomalous areas was created by multiplying Rs by kp. Reid (1988) reported encouraging relationships between C2-C4 hydrocarbons in soil and anomalies in the stress image. At the Mist gas field in Oregon, high-resolution reflectance spectra data of Douglas Fir trees (obtained with a spectroradiometer built by Geophysical Environmental Research Inc.) revealed spectral differences seemingly related to chlorosis. Douglas Firs off-field have higher reflectivity in the 0.750 to 0.125 ~tm range and a stronger absorption feature at 1200 ~tm than those on-field. Some Douglas Firs on-field exhibit higher reflectivity in the 0.550 to 0.650 pm region and a shift towards shorter wavelengths of the sharp rise in reflectivity at 0.700 ~tm. Locally, however, chlorosis is poorly correlated with commercial gas production. Microscopic examination of needles from Douglas Firs on-field revealed the presence of stress effects at the cellular level. Trees characterised by both chlorosis and cell damage may be indicative of more severe stress conditions. Crawford (1986) suggested that the poor health of Douglas Firs onfield may be caused by geochemical changes in soil and groundwater of their root zone induced by methane above the gas reservoirs. Natural gas seeps occur in the central portion of the Sao Francisco Basin of central Brazil and, in areas of soil gas anomalies, eucalyptus trees show clear signs of nutritional deficiency. Oliveria and Crosta (1996) simulated Landsat TM data for bands 1-4 by flying a non-imaging airborne data acquisition system (SADA) at a height of 150 metres over healthy and nutritionally-deficient eucalyptus stands. The resulting data were analysed alongside the corresponding Landsat TM data. Eucalyptus in areas of hydrocarbon microseepage exhibits higher reflectance in TM band 3 (chlorophyll absorption) compared to eucalyptus outside the anomalous areas. This is attributed to a reduction in chlorophyll pigmentation due to hydrocarbon-related stress. The presence of hydrocarbons seems also to produce lower reflectance in TM band 4, probably due to individual trees imposing differential conditions on the internal structure of the canopy. There may also be an effect on the spectral contribution from underlying soil energy; in areas of microseepage, this effect is more pronounced due to the lower vegetation density.
242
H. Yang, F.D. Van der Meer and J. Zhang 50
Blue = 4 0 0 - 5 0 0 n m
- -
Green = 500-600 nm -
Red
= 600-700 nm
Infrared = 7 0 0 - 1 1 0 0 n m
40
o-e
--
v
CD O t---
,s S''~--!
. . . . . . . . . .
30
"6
II I
CD
20
(Lp)
rv'
Blue
shift
10
i
0
500
400 -
-
600
Ro
i
i
I
700
800
~J 900
1 1000
1100
Wavelength (rim) Fig. 7-3. Generalised manifestations of stress on deciduous leaves (reproduced with permission of Veridian-ERIM International from Reid et al., 1988).
The Aircraft Thematic Mapper Simulator (NS-001) was used to examine spectral reflectance over the Patrick Draw oil field in southwest Wyoming (Lang et al., 1985a) and over the Lost River gas field in the Appalachian Mountains of West Virginia (Lang et al., 1985b). The spectral characteristics of the seven visible and near-infrared NS-001 bands are summarised in Table 7-11. In the Patrick Draw area sagebrush is the predominant vegetation. Over the gas cap of the oil reservoir, there is strong light hydrocarbon microseepage (Meyer et al., 1983), the sagebrush is stunted and an associated tonal anomaly is visible on Landsat TM imagery (Richers et al., 1982, 1986). According to Arp (1992) the sagebrush anomaly results from the upward migration of gases and waters injected to maintain reservoir pressures in the field; these produced anoxic, low Eh, high pH and high salinity soils that are toxic to sagebrush. Lang et al. (1985a) used the mixing model of Siegel and Goetz (1977) to separate the vegetation and soil/rock components for spectral analysis studies with the NS-001 data and found that the area of stunted sagebrush is characterised by a decreased in vegetation cover of at least 10% compared to the healthy sagebrush of the surrounding region. The principal vegetation anomaly over the Lost River gas field is the presence of maple trees, whereas the typical vegetation of the surrounding area is oak-hickory
Aerospace detection of hydrocarbon-induced alteration
243
climax. The maple trees occur in an area of maximum methane seepage and minimum soil oxygen content. The symbiotic mycorrhizal fungi on the root hairs of maple trees appear better able to tolerate these anaerobic soils than their counterparts on the root hairs of oaks. Lang et al. (1985b) searched the NS-001 data for information on the geobotanical effects. They concluded that: an image of bands 2, 3 and 4 supplied optimal species and species community discrimination; the ratio of bands 4:3 measured the amount of chlorophyll; the ratio of bands 4:5 showed green vegetation biomass; and the ratio of bands 5:7 indicated general vegetation vigour. Based on these findings, two supervised vegetation classifications were found to be optimal, viz: (1) bands 2, 3 and 4, band-ratio 4:5 and band-ratio 5:6; and (2) bands 2, 3 and 4, band-ratio 4:5 and band-ratio 5:7. Good accuracy is provided by these supervised classifications whereas single bands, band-ratios and composite band-ratios yield inadequate information for discriminatory mapping.
TABLE 7-1I Spectral characteristics of Aircraft Thematic Mapper Simulator (NS-001) Bands 1-7 Band 1
Centre and window ( l a m ) 0.49 0.45 - 0 . 5 2
2
0.56
0.52 - 0.60
3
0.66
0.63 -0.69
4
0.83
0.76 - 0.90
5
1.15
1.00 - 1.30
6
1.65
1.55 -I.75
7
2.22
2.08 - 2 . 3 5
Features Absorption features due to chlorophyll and carotenoid Chlorophyll reflectance peak; sensitive to green pigmentation Maximum chlorophyll absorption Reflectance peak sensitive to leaf area, vegetation density, biomass Fine spectral structure, sensitive to leaf area Absorption characteristic of leaf water mass Absorption attributed to water molecules within a leaf
Applications Deciduous / coniferous differentiation Assessment of vegetation vigour Species differentiation and assessment of vegetation vigour Species identification and assessment of vegetation vigour
Moisture status Moisture status
Biogeochemical analysis of leaves of trees over the Pollard oil field, Alabama, indicates anomalous concentrations of Mn. Correlation analysis reveals a statistically significant relationship between biogeochemical Mn anomalies and producing wells.
244
H. Yang, F.D. Van der Meer and J. Zhang
Stress symptoms such as leaf chlorosis are also observed. Cwick et al. (1995) searched for these biogeochemical effects using from an altitude of 1667 m, an approximate spatial resolution of 2 m and with narrow bandwidth filters in the spectral ranges of yellow-green (0.543-0.552 ~tm), red (0.656-0.664 ~tm) and near-infrared (0.815-0.827 ~tm). Although single-band video data were generally not sensitive to biochemical variations, transformed video data, in particular those in the near-infrared/red, exhibited significant correlations with biogeochemical Mn concentrations and Mn / Fe ratios. These results suggest that multispectral video data may have the potential to detect vegetation stress associated with hydrocarbon microseepage.
Other anomalies
Everett and Petzel (1973) reported that some oil and gas fields in the Anadarko Basin of the Texas Panhandle and western Oklahoma are manifested on imagery data as unique features, appearing to be smudged or erased, which they term "hazy anomalies". Moore and Anderson (1985) found a circular tonal anomaly on a Landsat TM image of the Herg oil field of the western Hardeman Basin, north Texas, which did not correlate with any single soil type or groups of soils, and suggest that it is possibly related to vertical migration of hydrocarbons causing chemically-altered soils or anomalous vegetation growth. Carter and Koger (1988) processed MSS and Landsat TM data from an area with hydrocarbon prospects. They suggest that, in the near-surface and surface rocks, soil and vegetation, structure and alteration (possibly due to hydrocarbon microseepage) can be detected by tonal and spectral anomalies in various filtered, contrast-stretched and edge-enhanced formats, in ratios of various bands in both colour and black-and-white, and in false colour composites.
PROBLEMS AND FUTURE TRENDS Oil and gas exploration methods based on what are assumed to be hydrocarboninduced alterations to rock and soil date back several decades. However, the processes that produce the observed effects are not well documented. Hydrocarbons are but one of the possible causes of the alterations and may not always be the most probable cause. The nature and extent of the alteration can vary significantly not only laterally and vertically but also temporally. Schumacher (1996) concludes that considerable research is needed before we understand the many factors affecting the formation of theses alterations in the near surface. We must evaluate seemingly "significant" alteration anomalies carefully to determine if they are related to hydrocarbon seepage. This requires answers to the following questions. Is the anomaly a function of geology or an artifact of culture'? If geology, is the observed alteration syngenetic or authigenic? If authigenic, is the anomaly seep-related or of non-seep origin'? If seep-related, does the
Aerospace detection of hydrocarbon-induced alteration
245
anomaly result from an active hydrocarbon seep or a palaeoseep? Finally, if the anomaly at the surface is to be related to a drilling objective at depth, does the anomaly result from mainly vertical migration or does the migration path follow a more complex route? The laboratory and field spectra of hydrocarbon-induced alteration products have been characterised, but the spectral resolution of MSS and TM are not high enough to be compared with the laboratory or field spectra. Although there has been some success in the detection of vegetation stress and alteration of mineral assemblages, the broad bandwidths of existing satellite platforms cannot characterise all the absorption features caused by hydrocarbon microseepage, regardless of the type of enhancements employed or the type of information extraction method applied. In short, high spectral resolution imaging data are needed for the recognition of the hydrocarbon-induced alterations and related vegetation changes. Remote sensing for direct detection of hydrocarbon microseepage holds great promise as an airborne geochemistry-biogeochemistry method that can complement seismic exploration by helping to recognise marginal and submarginal low relief structural prospects and stratigraphic traps. Through improved understanding of hydrocarbon microseepage and its surface manifestations along with imaging data of higher spatial and spectral resolution, we have a better chance to delineate the surface anomalies associated with subsurface hydrocarbon reservoirs.
This Page Intentionally Left Blank
PART III.
REMOTE DISPERSION PATTERNS OF POST-GENETIC PROVENANCE
This Page Intentionally Left Blank
Geochemical Remote Sensing of the Subsurface Edited by M.Hale Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
249
Chapter 8
SULPHUR GASES M.E. HINKLE and J.S. LOVELL
INTRODUCTION The odour of oxidising sulphides is a well-known geological phenomenon. This observation, combined with the existence of a wide range of volatile sulphur species, has led to a number of attempts to develop methods using sulphur gases as an aid to exploration. The potential application of sulphur-gas surveys was first discussed in the literature by Glebovskaya and Glebovskii (1960). Their attempts to detect sulphur gases were unsuccessful, despite the odour of sulphur. This failure was ascribed to the poor sensitivity of the (unspecified) analytical system. In the next few attempts to detect sulphur gases associated with mineralisation (Kravtsov and Fridman, 1965; Elinson et al., 1970), the detection systems used were either not described or appear to be of insufficient sensitivity. Shipulin et al. (1973), in a major article, report the association of sulphur dioxide (SO2), hydrogen sulphide (H2S) and carbonyl sulphide (COS) with mineralisation. The profiles illustrated in that paper are persuasive; however, the levels of H2S reported are as high as 1% in the soil air at a depth of 1.5-2.0 m. The data must be regarded with some caution as this highly poisonous gas is inimical to virtually all forms of life, including plants, at these concentrations. Furthermore, it is at least one million times the concentration at which H2S is detectable as an odour by humans. This suggests that, if the reports are accurate, then vapour geochemical techniques would have been the oldest-established method of mineral exploration. Rouse and Stevens (1971) described the potential of sulphur dioxide as a means of locating concealed mineralisation. They analysed both the overlying free atmosphere and soil gas, and reported low-contrast anomalies associated with sulphide mmeralisation, along with an ability to detect ore beneath many metres of overburden. However, the analytical technique has never been disclosed and there is some doubt that the sulphur species detected was, in fact, sulphur dioxide. Meyer and Peters (1973) and Peters (1973) reported anomalous concentrations of sulphur dioxide in the air and soil air in a very humid environment in Newfoundland. However, many of the anomalous values and all of the background concentrations are much below the analytical detection limit of 4 ppb claimed by the originators of the colorimetric method used (West and Gaeke, 1956; Scaringelli et al., 1967).
250
M.E. Hinkle and J.S. Lovell
Davy and Stokes (1977) also reported the determination of sulphur dioxide in the soil air overlying sulphide mineralisation. The levels measured were low, the contrast was poor, the samples were few and the traverses were not repeated. Despite very long sampling times (24-96 h) and deep sample holes (1.5-7.6 m), the results proved to be ambiguous. Byman (1977) attempted to detect sulphur dioxide and hydrogen sulphide in soil air over oxidising sulphides and in the interstitial air and free atmosphere associated with sulphide-beating mine dumps. Despite many refinements aimed at improving the sensitivity of the specific colorimetric detection methods used, he was unable to detect sulphur dioxide levels above 20 ppb, insufficient to account for the readily-detectable odour near the dumps, and hydrogen sulphide was not detected. Furthermore he was able to show that sulphur dioxide was not released by sulphide minerals oxidising in vitro. The principal gases released and identified with certainty by gas chromatography were carbonyl sulphide (COS) and carbon disulphide (CS2). The possibility of sulphur-beating gases being released by sulphide mineralisation has led to various attempts to utilise the inherent sensitivity of the sense of smell of dogs. This work has been apparently confined to glaciated areas, initially in northern Europe (Kahma, 1965; Nilsson, 1971, 1973) and the former USSR (Orlov et al., 1969), and has been mainly directed towards boulder tracing. Attempts were made to identify the gases to which the dogs responded (Kahma, 1975). Brock (1972) reported at length on a programme intended to train dogs in boulder tracing in Canada. However, the success rate was unimpressive in areas of known mineralisation. Subsequently, a period of particularly active research into the exploration value of sulphur gases yielded promising results. Hinkle (1978), Hinkle and Harms (1978) and Hinkle and Kantor (1978) used molecular sieves as passive adsorbents of sulphur gases. Lovell (1979), Lovell et al. (1980), Hale and Moon (1982), Hinkle and Dilbert (1984) and Oakes and Hale (1987) used the soils themselves. McCarthy et al. (1986) used soil gases and Hinkle (1986) used both soils and soil gases. The objective of this chapter is to present the rationale and the results of the investigations from this period.
CHEMISTRY AND GEOCHEMISTRY OF SULPHUR GASES Sulphur occurs as a primary constituent in three major inorganic mineral species: sulphides, sulphosalts and sulphates. It is found in crude oils in the range 0.1-7% sulphur, where it occurs in aliphatic and cyclic mercaptans and sulphides (Davis, 1967). Trace quantities of free sulphur and hydrogen sulphide are sometimes present in crude oil, and hydrogen sulphide occasionally occurs in appreciable concentrations in natural gas. Sulphur dioxide, hydrogen sulphide and carbonyl sulphide have been reported in the gases released by volcanic activity (Rose et al., 1979) and hydrogen sulphide is frequently associated with geothermal areas. Sulphur is also an essential element in the biosphere and is incorporated in numerous compounds, particularly proteins. Anaerobic bacterial decomposition of this material can
Sulphur gases
251
+0.5
G,
0.0
H2S
-0,5
-
1
2
.
4
0
6
8
~
10
12
14
pH
Fig. 8-1. Equilibrium distribution of stable sulphur species in water; total dissolved sulphur 0.001M, 25~ total pressure = 1 atm (reproduced with permission of Economic Geology, v. 64-2, p. 166, Fig.7, Granger and Warren, 1969).
lead to the production of the volatile by-products dimethyl sulphide, methyl mercaptan and hydrogen sulphide (Rasmussen, 1974; Banwart and Bremner, 1975). Long ago, Gottschalk and Buehler (1912) were able to demonstrate that, even under ideal conditions, it was not possible to produce detectable concentrations of either H2S or SO2 during the oxidation of sulphides in laboratory experiments which continued for several months. These conclusions were supported by the findings of Byman (1977). It is possible that Gottschalk and Buehler (1912) carried out the oxidation with an excess of oxygen which would not occur in nature at the oxidation interface of a mineral deposit. Granger and Warren (1969) conducted experiments to investigate the oxidation of iron sulphides by an aerobic aqueous phase in a sterile system. The objective was to study the formation of unstable intermediate ions during the oxidation process. The intermediate sulphur species detected were sulphite and thiosulphate. In general these products are more easily oxidised than the metallic sulphides and are stable only if removed from the oxidising environment. The oxidation of sulphide to sulphate proceeds according to the fields of stability shown in Fig. 8-1. If the reaction, FeS2 + 402 + H2 = Fe 3+ + 2H + + 2S042proceeds to sulphate as the sole sulphur species, there is no further oxidation-reduction in an abiogenic environment.
252
M.E. Hinkle and J.S. Lovell
.10.5
HSO~ 0.0
H2S ,4= UJ
- 0.5 I -
-1.O
0
-Co~
2
4
6
8
HS-
10
12
14
pH
Fig. 8-2. Equilibrium distribution of metastable sulphur species in water; activity coefficients of all species arbitrarily set to unity, total dissolved sulphur exclusive of sulphate species - 0.001 M, 25~ total pressure = Iatm (reproduced with permission from Economic Geology, v. 64:2, p. 166 fig.7, Granger and Warren, 1969).
The intermediate reactions that may occur in an environment with an inadequate supply of oxygen are: FeS2 + 2'/202 + H20 = Fe 2+ + 2HSO 3FeS2 + 1'/202 = Fe 2+ + $203 FeS2 + VzO2 + 2H + = Fe 2+ + H20 + 2S Thiosulphate ($2032-) and sulphite were the principal ions detected by Granger and Warren (1969) in the effluent from their experiment. The stability fields of these labile intermediates are shown in Fig. 8-2. The major relevance of these results to gas geochemistry is the formation of the bisulphite ion, a possible source of SO2 gas" HSO3 + H += H20 + 502 However, bisulphite ions are unstable and in dilute acidic solution the favoured reaction is decomposition by disproportionation: 3HSO3 -= H20 + H § + S + 2SO4 z
Sulphur gases
253
Furthermore, SO 2 is a highly soluble gas (23 ml/100 ml H2O at 0~ and is only released from concentrated solutions. In an oxidising, near-surface environment it is more easily oxidised than the sulphide minerals. Numerous authors have noted the odour of oxidismg sulphides and generally ascribe it to SO2. Unfortunately, the levels reported by those authors who have measured the concentrations of this gas are below the human odour threshold of 500 ppb, usually by more than an order of magnitude. The odour, therefore, remains unexplained, although it is possible that the gas could have been lost by adsorption onto the surfaces of the collection equipment. The natural environment is not, however, abiogenic, and it is possible that microbiologically-induced transitions could be the major source of volatile sulphur species. Rasmussen (1974) reports the bacterial generation of dimethyl sulphide from an inorganic source (MgSO4). In an anaerobic environment, sulphate may be autotrophically reduced to hydrogen sulphide using carbon dioxide as a carbon source and free hydrogen as the energy source (Shturm, 1952). Even in an oxidising soil there are, within soil crumbs, completely anaerobic micro-environments where these transitions may occur (Smith, 1977). It is feasible, therefore, that a local excess of sulphate may release an anomalous supply of volatile sulphur species into the environment and the overlying soils. Byman (1977) detected volatile sulphur species in the headspace gas over sulphides oxidising in vitro. The major gases detected were, initially, carbonyl sulphide and carbon disulphide. As the experiments continued dimethyl sulphide, (CH3)2S was detected, presumably as the result of bacterial action. Consideration of the stability of these gases by Rose et al. (1979) suggests that the gases are not in equilibrium. Figure 8-3 shows the relative equilibrium concentrations of a number of gaseous sulphur species. It suggests that, theoretically, HzS, even in an oxidising environment, should be by far the most abundant species. The evidence suggests that equilibrium is not attained. Further studies on the stability of sulphur gases in the natural environment were conducted by Taylor et al. (1982). They calculated the thermodynamic equilibrium abundances of sulphur gases that should be given off by decomposing sulphide minerals and compared the theoretical concentrations with actual concentrations of sulphur gases, which they measured by gas chromatography, evolving from pyrite, chalcopyrite, sphalerite, and galena. According to their calculations, the following gases should, theoretically, be observed, under equilibrium conditions, in order of decreasing abundance: HzS, COS, CS2, CH3SH, (CH3)282 or SO2 or 82 (depending on Eh and pH). Of these gases, only H2S and COS attain a concentration in nature (10 .9 atm) high enough to be measured by gas chromatography (Fig. 8-4).
254
M.E. Hinkle and J.S. Lovell
i I
I
I pH-2 Pco,~ Pc., " 10"1 lIH2s Or 88o~ -1O'~
I I 1
CONCn.
/
I
natlvo S
t -2sOc a.= o- 1
I
C~
10 "m ,
.-l____Z___t 03 0.2
/ (11
C2HaS Eh
0
-0.!
-0.2
Fig. 8-3. Equilibrium concentrations of a variety of sulphur species at pH 2 (reproduced with permission of Academic Press from Rose et al., 1979, Geochemistry in Mineral Exploration).
In contrast, their experiments showed that decomposing sulphides evolved only CS2 and COS, in order of decreasing abundance. In addition, measurements of the Eh and pH of the final solutions from moist and saturated sulphide mineral experiments showed that CS2 and COS were formed under metastable oxidising conditions (Fig. 8-5). Pyrite produced the largest amounts of sulphur gases, more than chalcopyrite, sphalerite and galena (Fig. 8-6). Moist (not saturated) and non-sterile (as compared to sterile) conditions enhanced sulphur-gas evolution from pyrite samples. Sulphur gas evolution from saturated pyrite samples is shown in Fig. 8-7. The authors concluded that CS2 and COS are inorganically generated from decomposing sulphide minerals under disequilibrium conditions. They went on to suggest that while CS2 and COS can be derived from soils as well as from sulphide minerals, soils also generate quantities of the organic sulphur compounds dimethyl sulphide, dimethyl disulphide and methyl mercaptan, and the presence or absence of these organic sulphur compounds may indicate whether or not the CS2 and COS in a sample evolved from the action of soil micro-organisms or from decomposing sulphide mineralisation at depth. Subsequent to the experiments of Taylor et al. (1982) on sulphur gases derived from decomposing sulphide minerals, Hinkle et al. (1990) performed simulated weathering experiments on ground drill cores from the Santa Cruz porphyry copper deposit, Casa Grande, Arizona. Fresh, fmely-ground samples were placed in sealed tubes, either in contact with a small quantity (0.8 ml) of water or exposed to water vapour but not in
Yt
Sulphur gases 0.4
c.. 'i'
I
I
i
i
0.3
0.2
001]
0.1
255
--~'~> -0.1 7
~]
-0.2
-0.4
-0.5
-0"60 I I
2
4
II 6
pH
8
10
tk 12
14
Fig. 8-4. Stability fields of H2S (vertical shading) and COS (diagonal shading) greater than 10-'~ atm partial pressure and stability field for solid sulphur (solid shading) in relation to natural limits of Eh and pH (hachured border, Bass Becking et al., 1960); total dissolved sulphur = 0.01 M, Pco2 + Pctt4 = 10-~ atm (from Taylor et al., 1982).
contact with water. The samples were separated into various mineral fractions and analysed for contents of metals and sulphide minerals. The principal sulphide mineral present was pyrite. The gases produced in these experiments were analysed by gas chromatography: CO2, 02, COS, SO2 and CS2 were found; H2S, organic sulphides and mercaptans were not detected. In general gas production depended more on 02 concentration that on any variable related to the sample material, such as metallic element content, sulphide mineral content or mineral fraction (oxide or sulphide). The various volatile species appeared to be interactive, so some may have formed through gas reactions. In summary, the sulphur gases most likely to be related to sulphide mineralisation in the natural environment are CS2, COS, H2S and (CH3)2S. Many chemical reactions can occur between the time a sulphur compound (volatile or non-volatile) leaves a deposit and the time a volatile sulphur compound appears near the ground surface above the deposit. Bacterial action probably plays a large role in the formation of sulphur gases as they react with minerals in the deposit, with bedrock, with groundwater and with soil en route to the surface. Therefore, while gaseous sulphur compounds over or peripheral to sulphide mineralisation may be related to the mineralisation, the compounds may or may not have originated directly from the mineralisation.
256
M.E. Hinkle and J.S. Lovell 1.0
-
0
*
0.8 0.6
0.4 r
o> 0.2 J= uJ
0.0
-0.2
-0.4
-0.6
-
%
I
0
I
2
I 4
e u x~n iC~ ~ , , ~ . ~ ' ~ 0 ~ .'~,0,6.. marine ~ ....Oe ;'P~'e~e environment ~ I,,~t~" l
6
DH
.
I
8
~
10
c,,,.~ ~
12
_
Fig. 8-5. Eh-pH measurements of final solutions from experiments using moist sulphide minerals (open circles) and saturated sulphide minerals (filled circles) in which CS2 and COS were detected; natural environments according to Garrels and Christ (1965); outlined area depicts natural limits of Eh and pH according to Bass Becking et al. (1960) (from Taylor et al., 1982).
EXPERIMENTAL TECHNIQUES
Sample collection Hollow probes driven into the soil to various depths have been used effectively to collect soil-gas samples. Probes have the advantage of being a dynamic sampling method, collecting soil gas as it exists in the ground at a given moment. Disadvantages to the use of probes arise from possible changes in concentration of soil-gas components due to changes in environmental conditions during the course of sample collection. Diurnal fluctuations of temperature and barometric pressure cause changes in concentrations of soil-gas components. Rain, depending on the depth of water penetration into the soil, may either flush gases from the soil or form a wet layer of surface soil, which prevents subsurface gases from rising and thus concentrates them. Disadvantages due to the soil itself arise from rocky or cemented soils that cannot be penetrated by the probe. Another disadvantage is the possible reaction of sulphur gases with metal surfaces of the probe.
Sulphur gases
257
>
A
1500
Z 0
EXPLANATION ,
<
,,
CS~
COS
Z z 0
/,o.., ...--o~..,,o~. --.-o,-,.,,,,,o.._..,o.__ _...o~ .,,. ~
< 50O
0
2
6
4
I
DAYS
8
~
I
10
I
,
12
Fig. 8-6. Sulphur gas concentrations (ppb by volume) from the decomposition of 20 g of 40-80 mesh sulphide minerals under saturated conditions: pyrite (circles); chaicopyrite (triangles); galena (squares); concentrations of sulphur gases <100 ppbv have been omitted for the sake of clarity (from Taylor et al., 1982).
Z O I--
1500
EXPLANATION CS,
,,.,..
i--
ZI~. O~ U m
<
COS 1000
....~.~---o---
500
.....o~~'~o
r
--.-o~ ~......... o
--.-
.J 2
4
6
8
10
12
DAYS Fig. 8-7. Sulphur gas concentrations (ppb by volume) from the decomposition of 20 g of 40-80 mesh pyrite under saturated conditions: sterile (open circles); non-sterile (filled circles) (from Taylor et al., 1982).
258
M.E. Hinkle and J.S. Lovell
One type of probe is a hand-held stainless-steel hollow auger, which has soil-air vent holes drilled into the shaft near the bottom and is fitted with a gas sample port near the top (Lovell, 1979). Another type of hollow probe has a straight shaft with a gas sample port at the top and soil-air vent holes at the bottom; this probe is pounded into the soil by means of a captive hammer that slides up and down the shaft to either drive the probe into or remove it from the soil (Dyck, 1972; Chemical Projects Limited, Toronto, Ontario, Canada, written commun., 1972; Lovell, 1979). Both the hollow auger and hollow hammer-probe are efficient dynamic soil-gas samplers in light soils. Neither sampler works well in stony or hard, compacted soils, or in soils containing layers of caliche; attempts to use probes in these soils may either bend the probe or disturb the soil to the extent that the soil-gas sample is diluted by atmospheric air. Soil gas can not be sampled in wet soils with these probes. Variations of a third method of soil-gas sampling have been used for many years. A hole is bored by a hand auger or gasoline-powered auger, a hollow sample tube is placed in the hole and the hole is sealed with rubber or soil. Later, a sample is collected of the gases in the hole that have equilibrated with gases in the soil surrounding the hole (Boynton and Reuther, 1938; Dyck, 1972; De Camargo et al., 1974; Lovell, 1979). The advantage of the sealed sample hole is that large volumes of equilibrated soil gas can be collected, depending on the size of the hole and sampling tube. The disadvantages of the sealed sample hole are the difficulty in forming an air-tight seal over the hole and the delay of several hours that is necessary for the soil gas to reach equilibrium before sampling the site. The passive technique of burying an adsorbing material in the soil to collect soil gases eliminates problems of climatic changes in concentrations of soil gases, as well as site-to-site problems of ground penetration that are encountered by dynamic gas sampling with hollow probes. Molecular sieves, used by the chemical industry to remove undesired components from natural gases and process streams, have proved to be good soil-gas adsorbents. Molecular sieves are synthetic dehydrated crystalline zeolites. These materials have a high internal surface area available for adsorption due to the channels or pores which uniformly penetrate the entire volume of the solid. Zeolite molecular sieves have pores of uniform size (3-10A) which is uniquely determined by the unit structure of the crystal. These pores will completely exclude molecules that are larger than the diameter of the pores. The extemal surface of the adsorbent particles contributes only a small amount of the total available surface area (Breck, 1974). Molecular sieves were used to collect and concentrate sulphur compounds in soil gases over buried sulphide mmeralisation at Johnson Camp, Arizona, in the Roosevelt Hot Springs Known Geothermal Resource Area, Utah, and in the Puhimau thermal area, Hawaii (Hinkle, 1978; Hinkle and Harms, 1978; Hinkle and Kantor, 1978). The molecular sieve adsorbers were placed in containers and buried in the ground for periods of 6-8 weeks. This method of collecting gas samples has the advantage of concentrating trace quantities of gases that might otherwise be below the detection limit of the analytical method used, as well as averaging the effects of climatic changes on the concentration of
Sulphurgases
259
the soil gases collected. However, the method has two distinct disadvantages: the necessity for two field visits to bury and collect the adsorbers: and the preferential adsorption of water vapour over other gases. A more effective use of molecular sieve adsorbers is in an environment sheltered from rain or soil moisture, such as a covered drill hole where the molecular sieves may collect volatiles rising from mmeralisation penetrated by the hole. Hinkle and Kantor (1978) used molecular sieves to collect trace sulphur compounds from the air inside drill holes at Johnson Camp, Arizona. Increased sensitivity of analytical techniques has made concentration of the soil gas unnecessary in many cases. The soil itself acts as an adsorbent, and does not appear to reflect diurnal variations as much as the soil gas, although variations of temperature and rainfall will affect the concentrations of volatiles in the soil. Soil samples are taken to a laboratory for analysis of degassed volatiles. Soil samples for degassing are usually collected from depths between the ground surface and 50 cm. A surface sample consists of soil from which the top 0.5-1.0 cm of soil and surficial debris has been scraped away. The deep sample is usually collected at 30-50 cm depth using a hand auger or gasoline-powered auger. Another type of soil sampling in arid areas collects only the top 1 mm surface microlayer which, after sieving to remove coarse debris, consists of dust and particulates. Surface microlayer samples have shown sulphur-gas anomaly patterns that are more closely related to known mineralisation than sulphur-gas patterns from deeper soil samples (Lovell, 1979). At Johnson Camp, Arizona, mineralisation is best expressed by sulphur compounds from the surface microlayer, while sulphur compounds from 0-5 cm reflect the same mineralisation to a lesser extent, and sulphur compounds from 30-40 cm show the least expression of the mineralisation. A comparison of concentrations of COS, CS2, and SOE degassed from soils collected at depths of 0.5-2 cm and 30-40 cm at the same sites near Casa Grande, Arizona, showed almost identical patterns of sulphur-gas concentrations over a 150 km 2 (58 square miles) area. Average concentrations of COS, CS2 and SO2 were slightly higher in the shallow samples than in the deeper samples (Hinkle, unpublished data, 1981). These data indicate that, at least in arid areas, surficial soil and microlayer samples are superior to deeper augered samples.
Analysis by gas chromatography Gas chromatography (GC) is the most common and successful method of soil-gas analysis. The detection limits are about 1-10 ppb by volume. The basic components of a gas chromatographic system are a carrier gas and a flow control system, a column packed with a gas-separating material, an oven for temperature control of the column, a sample introduction device, a detector and a recording system (Fig. 8-8). In the operation of a GC system, a gas mixture is injected onto the column and is carried through the column and detector by the carrier gas at a specific flow rate and temperature (either a constant temperature or a reproducibly programmed temperature range). The components of the gas mixture pass through the column at different rates, and
260
M.E. Hinkle and J.S. Lovell
FLOW CONTROLLER COLUMN OVEN ! CARRIER GASSUPPLY
t II omc~o.
9
SAMPLE " ~ ! INTHODUCTION I 1 1 COLUMN
tt
AMPLfFIER
I,.~
TEMPERATUIIE CONTROL
Fig. 8-8. Schematic gas chromatographic system (from Lovell, 1979).
are eluted from the end of the column at distinct elapsed times after injection. The retention time of each component on the column is a qualitative feature of that particular component and distinguishes it from the other components in the mixture. The eluted components are detected and quantitatively measured by a suitable detector system as they leave the column. The retention time and a peak height or area related to the concentration of the component eluted are recorded either by a strip-chart recorder or by a recorder-integrator system. Various commercially-packed columns and packing materials are available and have been used for the separation and analysis of mixtures of sulphur gases (Bremner and Banwart, 1974; De Souza et al., 1975; Supelco, 1977). The most commonly-used columnpacking materials for sulphur-gas analysis are Chromosil 310, acetone-washed Porapak-QS and three types of Carbopak B. Selection must be based on the analytical data desired. Each of these packings separates different components and has advantages and disadvantages. The performance and limitations of columns commonly used in sulphur-gas analysis are summarised in Table 8-I. Chromosil 310 isothermally separates COS, H2S, CS 2 and SO 2, but not organic sulphur compounds. Acetone-washed Porapak-QS separates several organic and inorganic sulphur compounds, but the required gradual heating of the column during the separation may cause erratic retention times from day-to-day, resulting in confusing data. Carbopak B-HT and Carbopak B-HT-100 separate several compounds, but they do not separate COS and SO:. Carbopak-B coated with polyphenyl ether and phosphoric acid separates COS and SO:, but is sensitive to moisture (S.E. Kesler, oral commun., 1981).
261
Sulphur gases
TABLE 8-I Column packings for gas chromatographic determination of sulphur gases Column packing material
Sulphur gases separated
Disadvantages
Chromosil 310
COS, HzS, CS 2, 802
Does not separate organic sulphur compounds
Acetone-washed Porapak-QS
COS, HzS, CS2, SO2, CH3SH, (C2H5)28, (CH3)282
Repeated column heating during operation may lead to erratic retention times and generation of organic sulphur compounds
Carbopak B-HT and Carbopak B-HT-100
H2S, SO2 + COS, CH3SH, CzHsSH, (CH3)28, other organic sulphur compounds
S01 and COS are not separated
Carbopak B coated with H3PO4 and a polymer
COS, H2S, CS2, CHsSH
Column is sensitive to moisture
Glass capillary column (no packing material)
COS, H2S, C82, CHsSH, CH3)2S, (CH3)2S2
Polar glass surface may adsorb sulphur compounds
Instead of packed chromatographic columns, Farwell et al. (1979) used wall-coated open-tubular glass-capillary columns to separate sulphur compounds. Various organic compounds were used to coat the interior walls of the different types of glass open-capillary tubes, in order to de-activate the polar glass surface, which otherwise adsorbs sulphur compounds (Farwell, 1980). A cryogenic enrichment system coupled with these wall-coated open-tubular glass-capillary columns was used to analyse biogenic sulphur gases that were collected from soils by sample chambers placed over the ground surface (Adams et al., 1981 ). Helium, argon, nitrogen and compressed air are the carrier gases used to move gas mixtures through chromatographic columns. Inert gases are often preferred over air as a carrier gas to avoid the possibility of oxidation of sulphur compounds in the soil-gas sample. However, because most soil-gas samples are already in contact with the atmosphere at the time of collection, they are not likely to oxidise further in the chromatographic column unless the column is heated to high temperatures. The choice of carrier gas depends on which gas yields the largest and sharpest elution peaks under the operating conditions of the gas chromatographic column.
262
M.E. Hinkle and AS. Lovell
Total (non-speciated) sulphur analysis may yield all the information needed or desired for some sampling programmes. When only a total sulphur measurement is required, the gas sample is passed through an empty Teflon column, by-passing the chromatographic column, and is sent directly into the detector (Tracor, 1973; Lovell, 1979). Total sulphur analyses can be very rapid, requiting less than 2 min. per analysis. When extremely sensitive sulphur compound detectors are used, the carrier gases must be of ultra-high purity grade. In addition, gases for reaction, for the flame and for mixing must also be of ultra-high purity grade. A detector that is especially sensitive to sulphur compounds is critical to the gaschromatography system. There are three types of detectors that are sensitive to ppb concentrations of sulphur compounds: the flame photometric detector (FPD); the Hall electrolytic conductivity detector (HEED); and the photoionisation detector (PID). The advantages and disadvantages of each are summarised in Table 8-11.
TABLE 8-11 Gas chromatographic detectors Type of detector
Advantages
Disadvantages
Flame photometric detector (FPD)
Sulphur-specific (although C02 causes a small peak) Sturdy components
Non-linear response (GC must have built-in lineariser or standard concentrations must be run and standard plots prepared)
Hall electrolytic conductivity detector (HEED)
Sulphur specific Linear response More sensitive than FPD to organic sulphur compounds and equally sensitive to inorganic sulphur compounds
Very temperamental (ultra-high purity carrier and reaction gases are required)
Photoionisation detector (PID)
Sample not destroyed by analysis Same sensitivity as FPD and HECD
Not sulphur specific (responds to a variety of compounds)
263
Sulphur gases EXHAUST TUBE
"0"
PHOTOMULTIPLIER HOUSING
IGNITOR
WINDOW HOUSING RADIATOR
ION COLLECTOR
TEFLON "O'RING W/ METAL BUSHrN6
FILTER
IGNITOR PLUG
BURNER AIR INLET "
BULKHEAD
,,
FILTER HOUSING HEATER
H2 INLET INLET FITTING
CHROMATOGRAPHIC COLUMN
Fig. 8-9. Cross section of flame photometric detector (from Tracor, 1980a).
The FPD (Fig. 8-9) is widely used for the detection of various sulphur-containing compounds. The principle of detection by the FPD is based on the formation of a luminescent $2 molecule when a sulphur compound is mixed with air and burned in a hydrogen-rich flame. This luminescent $2 molecule emits brightly at a wavelength of 396 nm. An optical filter placed between the flame and the photomultiplier tube permits only light of 396 nm to pass, and makes the FPD specific for sulphur compounds (although a small CO2 peak is often detected at the beginning of a chromatographic run). The output from the photomultiplier tube is amplified and passed to a strip-chart recorder or integrator. The response of the FPD is not linear, but is proportional to the square of the concentration of the sulphur-containing compound. In addition, the response of peak areas to concentrations becomes fiat above a certain sulphur concentration (generally in the 2-4 ppm range). Standard curves must be plotted for each compound, and extrapolation is difficult. Some gas chromatography systems are fitted with electronic filters which automatically convert the FPD output to a linear function (Brody and Chaney, 1966; Tracor, 1980a). Gases required for a gas chromatograph with a FPD detector are: (1) a carrier gas (compressed air, nitrogen, or helium); (2) hydrogen, for the flame; and (3) make-up air, for the flame. The FPD is a sturdy detector and may be transported to field laboratories. However, field use is not usually practical because cylinders of compressed gases must also be carried to operate the equipment. Catalytic hydrogen generators, which
264
M.E. Hinkle and J.S. Lovell
~[~j~,Reactiontube Heating element Reactor
1/16"Tubanut Reactor bas Ferruh Reaction o f
3.25"
Vent~-..~ pte
Column Fig. 8-I0. Cross section of Hall electrolytic conductivity detector reactor (from Tracor, 1980b).
produce H2 from distilled water, are commercially available and are recommended for field use in preference to cylinders of H2 which are dangerous to transport; even so, a catalytic H2 generator is another item of cumbersome equipment to transport. The HECD is another sulphur-specific detector. It comprises four basic components: a heated reactor (Fig. 8-10), a differential conductivity cell (Fig. 8-11), a signal processing module and a solvent reservoir module (Tracor, 1980b). The HECD is not a "stand alone" type of detector; a gas chromatograph must be specially adapted. As the separated sulphur components of the gas sample are eluted from the chromatographic column, they are mixed with air at the base of the reactor and converted to SO2 and SO3 at 800-1000~ Experiments by Hinkle and Ayres (1985) show that 900~ is the optimum reactor temperature for conversion of CS2 and COS to SO2 and SO3. Other products formed in the oxidation reaction are NO2, CO, CO2, halogen gases and H20. The reaction products are carried into the differential conductivity cell where they are mixed with a deionised solvent (reagent-grade methanol). The electrolytic conductivity of the resulting solution is compared to the electrolytic conductivity of the pure deionised solvent. The difference in
265
Sulphur gases
Top Electrode
r~
Insulator
Outer "J" Electrode"'~ i i ~ ~ Gas-Liquid Contactor ~ ,
~
G
S~
Inlet
1 aSB1Inlocket
Gas-Liquid ~i,~' Separator "~ '~& So,ve
Hole [~ ' ~
Insulator
L~ --ExiBott0m t Line Electrode Fig. 8-11. Cross section of Hall electrolytic conductivity detector differential conductivity cell (from Tracor, 1980b).
conductivity is then transmitted electronically to the strip-chart recorder or integrator. The response to concentration is linear. Under normal operating conditions H20 and CO give no response, while CO2 and NO2 are formed in very low concentrations and cause little interference. A scrubber is used to remove halogen gases. Therefore, the HECD is almost entirely selective for sulphur. The HECD has about the same sensitivity to inorganic sulphur compounds as the FPD; however, it is much more sensitive to organic sulphur compounds than the FPD (Farwell et al., 1981). Ultra-high purity grade helium for the carrier gas and air for oxidising the gas chromatograph effluent in the reactor are required for a gas chromatograph with a HECD. Both the flow rate of the ultra-pure air for oxidation and the flow rate of the reagent-grade methanol through the conductivity cell must be finely adjusted to obtain the maximum sensitivity of electrolytic conductivity. A gas chromatograph with a HECD is not practical for transporting to a field laboratory. The fittings on the reactor and conductivity cell of the HECD are very fragile. In addition, the ultra-high purity operating conditions and sensitive settings of flow rates confine a HECD system exclusively to laboratory use. The PID (Fig. 8-12) can be used with any gas chromatograph and has a sensitivity at least equal to those of the FPD and HECD. The sample is not destroyed by PID analysis; hence, the PID may be connected in series with other detectors or with a mass spectrometer for further analysis of the sample.
266
M.E. Hinkle and s
Lovell
Sealed E V
,
Detect~ ~ _, 1
9
Glass Line ~ Inlet
~
Source
I ~ ' ~ ~ 1 ' [/-- CeramicIonization . II L Chamber
rl.
_
/J
_
]i. ~
U Fig. 8-12. Schematic of high-temperature photoionisation detector (reproduced with permission of American Laboratory from Driscoll et al., 1978).
The principle of operation of the PID is photoionisation. The absorption of ultraviolet light by a molecule leads to ionisation of the molecule: R+h=R++e where R is the ionised species and h is a photon which has an energy greater than the ionisation potential for the species. The detector consists of a sealed, interchangeable ultraviolet lamp that emits a selected energy line, for example 9.5, 10.2, or 11.7 electron volts. A chamber adjacent to the ultraviolet light source contains a pair of electrodes. A positive potential applied to the accelerating electrode creates a field that drives ions formed by absorption of ultraviolet light to the collecting electrode where current, proportional to concentration, is measured. The response is linear (Driscoll et al., 1978). The PID responds to a wide range of organic and inorganic compounds. Therefore, a careful comparison must be made of the response of the detector to sulphur compounds of interest and to compounds with similar ionisation potentials. The only gas needed for a gas chromatography system with a PID is the carrier gas, usually helium or nitrogen. Therefore, the system appears fairly compact and transportable and has a good potential for field laboratory use.
Sulphur gases
267
Other methods of analysis Detector tubes are widely used in air pollution studies in industrial settings and mines to collect and measure toxic sulphur gases, such as CS2, CH3SH, H2S and SO2, on reactive absorbents sealed m glass tubes. The ends of the tubes are snapped off to expose the absorbents to the atmosphere. A measured volume (usually 100-500 ml) of air is drawn through the tube by a pump and sulphur compounds in the air react with the absorbents to form a coloured stain that can be compared to a chart of coloured stains produced by known quantifies of the sulphur gases. Generally, the concentration levels of sulphur compounds that can be measured by detector tubes have to be greater than the concentration levels of sulphur compounds found in a geochemical exploration programme. Several wet chemical methods of gas analysis are available for studies of sulphur compounds in soil gases. The West-Gaeke reaction is the best known wet chemical procedure for the collection and analysis of SO2. Sulphur dioxide in the atmosphere is removed and concentrated by scrubbing through 0.1 M sodium tetrachloromercurate. The SO2 in solution is then determined by the addition of a mixture of p-rosaniline hydrochloride-hydrochloric acid and formaldehyde. The red-violet complex of SO2 is measured at 560 nm. The method is sensitive to 0.005-0.2 ppm when a 38.2 litre air sample is scrubbed through 10 ml of sampling solution (West and Gaeke, 1956). Hydrogen sulphide may also be collected by bubbling air through a cadmium hydroxide scrubbing solution. The concentration of H2S in the solution is then measured colorimetrically by development of the methylene blue colour (Bamesberger and Adams, 1969). Both of these procedures and others were reviewed by Forrest and Newman (1973). Although no use is presently made of sulphur compound analyses on soil gases pumped from the ground and collected in aqueous absorbers or on chemically impregnated filters, this procedure may be useful in some cases for geochemical exploration. A major disadvantage would be the amount of chemicals and equipment needed to collect the sample and the time involved in pumping and concentrating the soil gas at each sample site. Other collection and analytical methods for soil-gas measurement involve the use of a portable mass spectrometer and the use of trained dogs (Kahma et al., 1975). Analytical methods for sulphur gases are summarised in Table 8-III.
Reference materials Sulphur gases in permeation devices are the usual means for calibrating the response of a gas chromatographic detector. Liquefied gases are sealed in Teflon tubes, or in glass or stainless steel tubes with Teflon "windows" on the ends. Gases permeate through the Teflon at a constant rate until the enclosed supply is exhausted. The rate of permeation is
268
M.E. Hinkle and J.S. Lovell
dependent on the temperature of the permeation device and the amount of Teflon exposed to the atmosphere. For calibration purposes, the permeation devices are kept in glass ovens or other suitable enclosures at a constant temperature (usually 30~ A dilution gas such as nitrogen, helium, or air is passed through the oven at different measured flow rates; the higher the flow rate, the less concentrated is the mixture of sulphur compound in the dilution gas. The sulphur-gas mixture is withdrawn from the oven at different flow rates of dilution gas, and is injected into the gas chromatograph for calibration and for preparing plots of concentrations versus response (O'Keefe and Ortman, 1966; Stevens et al., 1969; Stellmack, 1982). Permeation devices containing sulphur compounds of interest in geochemical exploration are available commercially in wide ranges of concentrations. A soil-gas sample collected by hollow probe needs no further preparation for injection into a GC system. However, the results obtained from sulphur analysis of a soil gas may be inaccurate representatives of their original state in the field. The time elapsed between collection of the sample in the field and analysis in the laboratory permits many changes in the sample. Some sulphur components may adsorb on the walls of the glass, plastic, or stainless-steel syringes or sample carriers used for sample transportation. Non-adsorbed components may interact with each other. The most reliable soil-gas analyses are those made almost immediately after collection. Even these analyses may not be completely accurate.
TABLE 8-111 Summary of analytical methods for sulphur gases Method
Advantages
Disadvantages
Gas chromatography
Several sulphur compounds determined simultaneously
Difficult and, in some cases, impossible to use in the field
Detector tubes
Easy to use in field Inexpensive Collects and analyses gases
Limits of detection too high for geochemical exploration
Wet chemical (solutions and filters)
Widely used for H2S and S O 2 air pollution studies
Tedious
Portable mass spectrometer
Several volatile constituents can be determined
Expensive, experimental
Sulphurgases
269
Sample storage and preparation for analysis Methods of storing and preparing samples collected in sulphur-gas survey programmes for analysis vary according to the type of sample. Soil samples should be sieved before analysis to assure homogeneity of the sample. Sieving to 200-600 mm size assures good homogeneity. Further sieving is not advised, as it may cause loss of soil gases trapped between the soil particles. If the soil is dry, sieving in the field assures an adequate sample size for the desired analyses. If the soil is wet, it must be allowed to air-dry before sieving. Samples may also be sieved in the laboratory immediately prior to analysis. Soil samples should be stored in containers that are as air-tight as possible. Loss of moisture from the sample also leads to loss in sulphur-gas concentrations. Screw-capped glass bottles or jars are good soil-storage containers, especially if the caps are taped to the glass to prevent interchange with the atmosphere. Plastic bags are easier to carry in the field than glass jars and, if the sample is well taped within the plastic bag and then wrapped in a second plastic bag, the gases will remain intact for 1-2 months at cool temperatures. Freezing will maintain the sample intact for several months (Hinkle, unpublished data, 1981). Soil-gas samples collected on adsorbents should be stored in the same manner as soil samples to prevent drying out and loss of sulphur compounds. Samples collected in solutions or on filters are generally analysed immediately or shortly after collection.
Degassing soils and molecular sieve adsorbents Soil and molecular sieve adsorbent samples must be degassed for analysis. Useful analytical results may be obtained by simply heating the sample in an air-tight system and driving the evolved gases into the gas chromatograph. However, this technique does not yield very reproducible results, a limitation which is sometimes minimised by the analysis of several replicate samples. Two other techniques may be used to obtain the gas sample: headspace analysis (for soil samples only); or flushing the sample with a stream of air or carrier gas. Headspace analysis consists of removing a sample of air in equilibrium with a weighed quantity of soil from a sealed container maintained at a constant temperature, generally ambient or slightly elevated (Lovell, 1979), although higher temperatures are sometimes used (Van den Boom and Poppelbaum, 1982). A sample that is flushed by a stream of carrier gas must be heated in order to release the adsorbed gases. In a desorber manufactured by Century Systems, Inc. (Arkansas City, Kansas), a weighed amount of soil or molecular sieve sample is heated in a metal tube while being flushed by air or carrier gas. The desorbed gases are mixed with air or carrier gas and are carried into an empty cylinder of fixed volume. Multiple injections of equal volumes of the gas mixture are injected from the cylinder into the gas chromatograph by means of a piston. Samples may be degassed at temperatures ranging from 50-350~ Above 350~ there is the possibility of breakdown of any sulphide
270
M.E. Hinkle and J.S. Lovell
minerals present, and determination of sulphur gases from sulphides in the sample is not normally desired. The disadvantage of the Century Systems device is that extremely small concentrations of soil gases may be undetected because samples greater than 1 gram in size are too large for the sample holder.
Injection of samples Gas samples are injected on to the chromatographic column by one of two methods: a gas-sample loop of 1-10 ml volume, which is attached to a valve that is turned to flush the contents of the loop onto the column; or a 0.5-1.0 ml gas-tight syringe and needle, which is pushed through a septum for direct injection on to the chromatographic column. Of the two methods, the gas-sample loop is preferred for soil-gas analysis because: (1) a gassample loop provides a reproducible injection volume; (2) with a gas-tight syringe, back pressure is exerted on the sample by the carrier gas, and some of the sample may be blown out by the pressure of the carrier gas; and (3) the large volume of gas sample injected with a loop is often necessary to measure the trace quantities of sulphur gases present in many samples.
Recording analytical results Output of the detector may be recorded by a strip-chart recorder or by an integrator. Strip-chart recorders are difficult to work with when analysing mixtures of compounds because of the need to manually change attenuation and record retention times whilst compounds of different concentrations are rapidly eluted. Peak heights or areas must also be measured or computed manually. Integrators are manufactured by many companies (for example Hewlett-Packard, Spectro Physics, Varian). Many integrators plot the chromatography run, record the retention times of peaks, calculate the areas under the peaks automatically and print this data at the end of the run. Recorders alone are adequate for total sulphur analysis, where the relative retention times of multiple peaks are not recorded. However, an integrator is necessary for detection and measurement of concentrations of sulphur species in a mixture.
CASE HISTORIES The case history section of this chapter is severely limited by the absence of welldocumented reports of the use of sulphur gases in mineral exploration. The case histories that follow are demonstrations of the ways in which sulphur gases delineate known mineral
Sulphur gases
271
occurrences. There are no accounts of concealed mineralisation having been discovered by sulphur-gas methods.
Johnson Camp, Arizona Johnson Camp has been the site of three independent surveys carried out by the authors of this chapter. The area lies in southeastern Arizona, in the Basin and Range province of the western United States. The survey area is a fiat, sparsely vegetated semi-desert lying to the east of the Little Dragoon Mountains. Mineralisation consists of pyrometasomatic zinc and copper sulphides which have been introduced into Palaeozoic limestones by a Tertiary quartz monzonite lying to the south. The sulphide mineralisation consists of sphalerite and chalcopyrite with dominant pyrite. The overall sulphide concentration is about 3%. The metalliferous bodies form tabular masses and chimneys in the plane of the beds. Three mineralised zones have been outlined by drilling (Fig. 8-13). Zone I, the most southwesterly, is about 30 m wide; Zone II, about 180 m wide, is the main ore zone and lies some 750 m to the northeast of Zone I; the third body of mineralisation, Zone III, is located approximately 350 m further to the northeast and is about 60 m wide. These zones, which strike northwest-southeast and dip at 40 ~ to the northeast, have been partially oxidised near their suboutcrops and adjacent to innumerable minor fractures for much of their downdip extensions. They are overlain by pediment gravels and alluvium that thicken to the northeast. Thus at suboutcrop, Zone I is concealed by about 10 m of overburden, Zone II has 90-150 m of cover and Zone III is about 225 m below the surface. The land surface is nearly level, with a gentle northeasterly slope. The soil is alkaline, consisting of mixed sand, clay and caliche, and grades imperceptibly into the underlying alluvium. The caliche varies between 0.5-2 m in thickness and may be encountered anywhere between the surface and about 2 m. The first sulphur-gas survey was carried out in this area by Hinkle and Kantor (1978) using molecular sieve adsorbents to collect and concentrate the sulphur gases. Seventy-one samples of soil air were collected on three parallel traverses running southwest-northeast (Fig. 8-13). On the principal traverse, B-B' (37 sites), the sample spacing varied between 15 m in the southwest to approximately 70 m in the northeast. The sampling depth was 3045 cm. The molecular sieve samples were left in the ground for eight weeks. The two subsidiary traverses comprised 14 sample sites (A-A') and 22 sites (C-C') with similar spacings to the main traverse. The molecular sieve samples were analysed within four weeks of retrieval for adsorbed sulphur gases (COS, H2S, CS2, and SO2) using a Tracor 270 HM sulphur analyser. The gases CS2 and SO2 were not detected. The average concentration of COS was 43 ppb with a range of 0-300 ppb. The majority of values were between 30 and 60 ppb and concentrations above 55 ppb were considered to be anomalous. Hydrogen sulphide concentrations ranged from 0-1000 ppb with the majority being below the threshold value of 300 ppb. The average H2S concentration was 140 ppb, and the lower
272
M.E. Hinkle and J.S. Lovell
limit of detection was 50 ppb. Anomalous concentrations of COS and H2S were located above and peripheral to the ore zones on all three traverses (Figs. 8-14 to 8-16). The shallowest ore zone (Zone I) is characterised by an exceptionally-large number of anomalous samples. Anomalous values detected in an undrilled area at the northeastern extremities of all three traverses could, perhaps, be ascribed to the presence of an extension of Zone III.
110o03,12,, 9
ZONE
EXPLANATION
i
III
Down-dip extension Z O N E II 126 ~ 1 1 4 ~ o f ore 125c~
Q
0
Drill hole B---
.'...,,B
7
Sample locality along traverse
T
15
S
32 ~
ZONE
I
07'
.~,.------~7~-
33'
9
r
...
R 22 E 0 1 !
0
1
I
1000 I
!
!
I
2500 FEET I
500 METERS
Fig 8-13 Ore zones, numbers of sampled drill holes and soil air sampling traverse locations at Johnson Camp, Arizona, for survey performed by Hinkle and Kantor (1978).
Sulphur gases
273
A 40O
AverageHIS,
z~ 3:
A
/
0
m 160 o~ ~.
I
~176
50
z_ o
u
i
-- ~
_
Average C O S .
I
|
AverageCO,
10 r 0
Fig. 8-14. Cross section along Hinkle and Kantor's traverse A-A' at Johnson Camp, Arizona, showing concentrations of H/S, COS and CO2 in soil air (from Hinkle and Kantor, 1978).
In order to confirm that the sulphur gases were derived from the mineralisation, bags of molecular sieve adsorbent were lowered into the casings of drill holes which intersected the ore, and left for nine weeks to equilibrate. The analysis of these molecular sieves showed that sulphur compounds had been adsorbed. The results are summarised in Table 8-IV. These data demonstrate the advantages of a technique which integrates the gaseous flux over an extended period, as tests showed that no sulphur gases were detected after one week of burial and only trace quantities after two weeks. A different approach was used by Lovell (1979) and Lovell et al. (1980) who used the soils themselves as sulphur-gas adsorbents. The field programme consisted of three soil sampling traverses with a total of 185 sites (Fig. 8-17). The sample interval was generally 30 m, but where traverses A-A' and C-C' overlay Zone I, the interval was narrowed to 10 m
2 74
M.E. Hinkle and J.S. LoveIl
z _o __JS00
1000
=0 n,-
., 400
~
j;: 3o0 200
A
_z
___]
,_
z o
7
ge H=S
_J.
r
,
a. 120 a: i-
80 i- []
A
Average C O S . . J
0-
O
(J
I-
z
50 i -
40
Average~O=
uJ r UJ r~
%
o
10
a
m
EXPLANATION
9
Down dtp extension of ore
Fig. 8-15. Cross section along Hinkle and Kantor's traverse B-B' at Johnson Camp, Arizona, showing concentrations of H2S, COS and CO2 in soil air (from Hinkle and Kantor, 1978).
and 15 m, respectively. Soil samples were collected from three depths: the surface microlayer; 0-5 cm; and approximately 35-40 cm. The samples were analysed for total volatile sulphur gases and the results for the surface microlayer are presented in Figs. 8-18 to 8-20. The data are presented as peak areas from the strip-chart recorder of the analytical unit, as calibration of the response of a pulse of gas containing varying proportions of different species is not possible. On traverse A-A' it is clear that there is a distinct pattem of three zones of higher sulphur values. There is a discontinuous series of anomalous values at the southwestem end of the traverse, which begin over the suboutcrop of Zone I and extend into the area underlain by the downdip extension of the mineralisation to a point where the depth to mineralisation is of the order of 80-90 m. This is followed by a series of generally low values until the samples are
275
Sulphur gases C
400
"7
:ff
~
Z
200
160
80
Average C O S , , .
if)
o U
50 F-
z
r er
.
~.
,
C
C'
I
I
EXPLANATION
~ i
~
ORE i D O W N DIP EXTENSION OF ORE '
Fig. 8-16. Cross section along Hinkle and Kantor's traverse C-C' at Johnson Camp, Arizona, showing concentrations of H2S, COS and CO2 in soil air (from Hinkle and Kantor, 1978).
underlain by the suboutcrop of Zone II. These samples are very strongly anomalous with a high contrast relative to background. The anomalous section is confined to the suboutcrop and does not extend over the downdip extension of the mmeralisation. At the northeastern end of the traverse there are a number of anomalous samples which again may indicate an extension of Zone III. The other two traverses showed similar, but not as well developed, expressions of the mineralisation, although Zone III was not detected on traverse B-B', perhaps due to its considerable depth of burial. The samples collected from deeper within the soil profile gave only a comparatively poor expression of the mmeralisation (Fig. 8-21). Subsequently, Oakes and Hale (1987) resampled the surface microlayer and speciated the desorbed sulphur gases. Anomalous
276
M E . Hinkle and J.S. Lovell
. . . . . O~176
A'
0
04
0
0
%
B
o/
0
~',,, 0 0 \
0
o
o o ', o o ~ 'l
a
.~=~-o'-~
o ZONE
f
o I
ZONE
II
I
f,.,,.."
4, o
EXPLANATION
0
300 M
~
Ore zone sub-outcrop - - - - - - - Limit of down-dip extension o
Drill hole
Fig. 8-17. Ore zones, drill hole locations, surface microlayer and soil sampling traverse locations at Johnson Camp, Arizona, for survey performed by Loveli (1979).
concentrations of COS were detected over mineralisation, indicating that COS may be a useful guide to concealed mineral deposits. Hinkle (1986) collected soil samples on the same three traverses along which molecular sieves had been buried in the study by Hinkle and Kantor (1978). On each of the three traverses shown in Fig. 8-13, samples were collected at 60 sites at intervals of 30 m. A soil sample was collected at each site by scraping away surficial materials to a depth of 2-5 cm, collecting the soil beneath, and storing this in a plastic bag. Volatile compounds were removed from soil samples by placing the soil samples in closed vials with screw caps that had Teflon septa in their centres, and heating the vials for one hour at 80~ Headspace gas in the vial over the soil was removed by inserting the needle of a syringe through the Teflon septum in the cap of the vial. The headspace gas was analysed for sulphur species using gas chromatography. As part of this study, CO2 in soil air and soil and helium in soil air were also determined (Hinkle and Kantor, 1978).
Sulphur gases
277
TABLE 8-IV Sulphur compounds detected in drill holes at Johnson Camp, Arizona Mineralisation
Zone II
Ore
Drill hole no. 83 96 125 109
III Ore extension downdip
I
III
74 76 82 57 101 108 116 126 68
II
114
II
No ore nd
-
H2S (ppb) 370 800
COS (ppb) 400 90 >300 >300 >300
350 310 <30 nd
nd 500 330
50 65 <30
nd
nd
no sulphur gases detected 295
z m 120~
100-
z~
Organics
~-
,,~ ,.,. 10 ~a.
Suffur
A
A'
Pediment gravels and callche
0 50 r~ r 100 l,iJ
Peleozoic udimentz ZONE I
Iso ~E2oo 25o 3oo
Zo/ve /t
100
200
300
400
500
600
700
800
900 1000 11(]0 1200 1300 1400 16430 1600 1700 1800 1900 METERS
Fig. 8-18. Cross section along Lovell's traverse A-A' at Johnson Camp, Arizona, showing concentrations of total sulphur and organics in the surface microlayer (from Lovell, 1979).
278
M.E. Hinkle and J.S. LoveIl
~< 901
0~.
Ix:
70
=:z
50
~
I
._j
O1_.~
Organics B,
O-
50 100 n
. , , , , ,
T" i" , , I .
i
, =-,
, , .~,
7r , ,
, , ', i " , ' - ~
~ , , i
, , w ' ,
, ,
~--,
9Sulfur 8'
-
Pediment
gravels
and
caliche
15o
'9 200 11
~ 25o
~
-
300 " 3so -
.
~
.
EXPLANATION Psleozoic
'~o
' 2~o
'-
300 "
,~o
'
s~o
"
6~o .
.
.
sediments
~o.
METERS
~
ORE ZONES
80o ' ' ~o ' 1;oo"~ 1'oo' 1~oo' 1 ~ "
~
Fig. 8-19. Cross section along Loveli's traverse B-B' at Johnson Camp, Arizona, showing concentrations of total sulphur and organics in the surface microlayer (from Lovell, 1979).
z':Jl
~< z (3 R-
Oe"
70
z
2 m,.- n .JT 1,- ~.s
Q.
l'!
.
.
=
9 .
.
i
l
!
I
!
i
0 60
~ and
C caliche
I
'
100 rR . MJ
150
2o0 3ool
25O
35O
~-
0
~ 9 100
i 200
I 300
I 400
I 500
I 800
I 700
I 800
I g00
I 1000
~ O R E i 1100
i 1200
t 1300
i 1400
ZONES i 1500
I 1600
METERS
Fig. 8-20. Cross section along Loveli's traverse C-C' at Johnson Camp, Arizona, showing concentrations of total sulphur and organics in the surface microlayer (from Lovell, 1979).
Sulphur gases
279
N
l,,,,
x
5-40cm
0
~K)O
1000
1S00
1800m
Fig. 8-21. Cross section along Lovell's traverse A-A' at Johnson Camp, Arizona, showing concentrations of total sulphur in subsurface soils (from Lovell, 1979).
The results are shown in Figs. 8-22 to 8-24. Above-average concentrations of CO 2 (for this area) were seen both directly over some of the ore bodies, and on the sides of others in a rabbit ears pattern. The CO 2 was probably formed by sulphuric acid, derived from the weathering of sulphide ores, reacting with limestone host rock beneath the ground surface (see Chapter 14). Helium in soil gases was highest over the central (150 m deep) ore body on all three traverses, and over the northeastern ore body on traverse C-C'. This is probably helium generated at depth that is rising through faults and fractures that controlled the mineralisation in these locations. Of the sulphur species, CS 2, COS, H2S and CH3SH were detected. Carbon disulphide and COS were the primary sulphur species that could be related to the ore bodies. On traverse A-A', CS 2 occurred on the sides of Zone II, and COS occurred both at the southwestern end of the traverse and over Zone II. On traverse B-B', CS 2 occurred over Zone II but not over Zone I, whereas COS occurred from the southwest end of the traverse to over Zone II. On traverse C-C', CS 2 occurred from between Zone I and Zone II to the middle of Zone II, while COS occurred only over Zone II. Neither CS 2 nor COS occurred northeast of Zone II. Because CS 2 and COS occurred in the vicinity of Zone I and Zone II, these gases are believed to be related to the ore; at a depth of 225 m, the ore body may be too deeply buried to be detected by CS 2 and COS. Hydrogen sulphide and CH3SH could not be related to the ore bodies. Although HzS occurred in several samples throughout traverse A-A', the occurrences were scattered and did not form any pattern; CH3SH occurred in essentially the same concentration in most of the samples throughout traverse A-A'. On traverse B-B', H2S occurred in several isolated samples from
280
M.E. Hinkle and J.S. Lovell 2000 O
|
0
0
1500
g
0
o lOO~ m o
-
%
500
2.0 3
MEAN
1.0
0
~ O
0.2
(~ L~
0.1
/
MEAN
o
tn tu0c
300
• 250
T D J@ z
MEAN
;>00 A o I ,=
I-
I('}0 E
,
-~"~-~...._...._
A' .... ~ 2 K M
. I'----"-'-wnn !
i
Alhnvlun!
Fig. 8-22. Cross section along Hinkle's traverse A-A' at Johnson Camp, Arizona, showing conccntrations of COS, CS2, CO2 in soil and COz, He in soil air; means are average conccntrations of gases in the area (reproduced with permission from Hinkle, 1986, J. Geophys. Res., 91:12,359-12,365, copyright by the American Geophysical Union).
the southwestern end of the traverse to Zone II, whereas CH3SH occurred in several scattered samples throughout the traverse. Hydrogen sulphide did not occur in any sample on traverse C-C' and CH3SH occurred mostly east of Zone III. The lack of patterns to the occurrences of H2S and CH3SH suggests that these gases may be related to bacterial activity in the soil.
Sulphur gases '~
281
1500 O
O
o~ 9 1000
--
rj)
0
0
0
0
0
0 0 (1300 0
0
0 0
0
0 0 (3 n
500
cl
Q
A
--
9 pna
0
m
~
9
9
2.0 MEAN 1.o
o 1.9%
II II II
0.3
I
~
co~
1
.2
o.1
MEAN
3O0 Wnr
o~ x
w Z ,..,,250 --ILl
I ~
MEAN
200
~0
"
~
uJ 1 0 0 LU 200
B
~
~ ,
B' ,2 KM
1...... 1 I
I
"~.///2),,'
vium
- ~ _
Fig. 8-23. Cross section along Hinkle's traverse B-B' at Johnson Camp, Arizona, showing concentrations of COS, CS2, CO2 in soil and CO2, He in soil air; means are average concentrations of gases in the area (reproduced with permission from Hinkle, 1986, J. Geophys. Res., 91: 12,359-12,365, copyright by the American Geophysical Union).
282
M.E. Hinkle and J.S. Lovell
.._1tSoo O o) ,..,, 09B 1 0 0 0
0 O0 500 O (..) mm
0
li
im i l
3.0 A_l 0
co
2.0
MEAN
!
O 0
a~
1.0
0 0.3
o~m 0.2
ol
MEAN
o
tO
300
wx
uJz
MEAN
z . ~ "' 2so ~m .o_0 a . , J i~. ~-r
200
2)
5"
w~
C m
o
|
u2J~0s 0100 L3;
C" ~
1
- , . ~
I
~
(I
I
1
1
t
(
)
I
Alluvium
Fig. 8-24. Cross section along Hinkle's traverse C-C' at Johnson Camp, Arizona, showing concentrations of COS, CS2, CO2 in soil and CO2, He in soil air, means are average concentrations of gases in the area (reproduced with permission from Hinkle, 1986, J. Geophys. Res., 9 l: 12,359-12,365, copyright by the American Geophysical Union).
Sulphur gases
283
North Silver Bell Arizona Hinkle and Dilbert (1984) camed out a soil-gas survey at the North Silver Bell porphyry copper deposit, located about 60 km northwest of Tucson, Arizona. The unmined deposit is a northwesterly extension of the Silver Bell porphyry copper system, which occurs along the southwestern flank of the Silver Bell Mountains and is considered to be Laramide in age. Mineralised rocks at North Silver Bell include dacite porphyry and quartz monzonite porphyry that have intruded Palaeozoic and Cretaceous rocks. Potassic, phyllic and propylitic alteration zones are visible on the east side of the ore body. Primary sulphide minerals are pyrite and chalcopyrite. Several cycles of oxidation and leaching have resulted in a supergene zone of chalcocite ore presently lying beneath 10-40 m of leached capping. The alteration on the west side of the deposit is covered by alluvium, which deepens rapidly into the valley to the west. Soil samples were collected by scraping away surficial debris and collecting the soil at 0-5 cm depth at 30 sites on hills and hillsides over a 3 km 2 area covering the northern part of the deposit. Soils were sieved to <30 mesh (<600 mm) at each sample site, placed in plastic bags and stored prior to analysis in a freezer to reduce loss of volatiles. A 10 g soil sample was placed in a 40 ml glass vial with an open-top screw cap lined with a Teflon-covered septum. The soil samples in vials were equilibrated for three days in an incubator at 47~ For analysis of headspace gas surrounding the soil in the vials, 25 ml of ambient air was injected by syringe through the septum into the vial and 25 ml of mixed air and headspace gas was removed for injection into a Tracor 560 gas chromatograph with a Hall detector. The most abundant sulphur gases determined at North Silver Bell were CS2 and COS: 18 of the 30 samples contained CS2, with the highest concentrations located over the ore body (Fig. 8-25); 15 samples contained COS, with the highest concentrations located over the alteration zones around the ore body (Fig. 8-26). Only four of the 30 samples contained measurable HzS, one over the ore body and three over the alteration zones. Although SO2 was not expected, because of its high solubility in water, nine of the 30 samples, located mostly over the ore body, contained 300-1000 ppb SO2 (Fig. 8-27). The occurrence of SO2 may be attributed to the hot, dry environment at North Silver Bell, and to the shallowness of the deposit, which permitted some of the SO2 from oxidising ore to reach the surface before it dissolved in moisture within the leached capping and overlying alluvium. Organic sulphur gases occurred in 16 of the 30 samples, both over the ore body and over the alteration zones, in concentrations from 50-200 ppb. Of the 16 organic sulphur gases, eight were identified a s (CH3)2S, seven as CH3SH and one as C2HsSH. Possibly these organic sulphur compounds were produced by bacterial action in the alluvium over the sulphide mineralisation, but their distribution showed no correlation either with the ore body or with the alteration zones; they formed only random patterns whether plotted by type of compound or by concentrations of a single compound.
284
M.E. Hinkle and J.S. Lovell
32~ 26'30 '
11 l~
''
111~
~
....'" ,'-'~,": " "'~'; ~;-~,;".'.'::.,~."~,~25o0-.~,-,;'-,-":.'?-,"
~ 4
%
q,~
:,r.,-.,','.,.:,:;',,.:,~ -
,~
:,,'_,?:.-,~,-
.,
-
'' ]
"'":"'""" ~ ""' ;:':'_'!~:. . . . ",,". "." . .,,;,:..:,. ,!::..,~':,,~ ,..:.~..;:,:,.,,..,:, !,,..~,,~:
,_, , ..... .,.-.~.,.:..,..-:.::_~
.,.::
.
'
'~
I/(DJ
":,
:
,
,
2g',,~%'~ 32~
.
, .....
"
"'
?:::') :~';':
k
L to
Pt
" B e l l ore body ~ 0
0 I
,
I 0
9'
'
,
i
'
' 0.5
Phyllic alteration
0.5
,
I
I MILE I I KILOMETI:R
9 Sample site
l
Carbon disulfide in head-space gas Range' not detected at lower limit of detection (50 ppb)-1000 ppb ~=~
100 ppb
Fig. 8-25. Geology, mineralisation, soil sample sites and CS2 concentrations at North Silver Bell, Arizona (from Hinkle and Dilbert, 1984).
32 ~
111~
111~
'00"
~.2
//15
..,,..~.:.,,-,'",., ",.i .,.. ..,.,, ,:.. . ,, ":
.,.
. ~
, , oi 4 .'"
"",i~
25'
..........
.0 ~
~'%
' EXPLANATION
V /
~. \ of North Silver
/jIV/./-'~:7~~
U f~"
~
Propyl~icaheration
l. ~
Phyllic alteration
i~//////] Potassic alteration 0 I
I 0
'
,
'
'
,
'
~' 0.5
I
1 MILE
0.5 I
I
1 KILOMETER
I
9 Sample site Carbonyl sulfide in head-space gas Range: not detected at lower limit of detection (0.050 ppm)-O560 ppm
~
> 0.300 ppm
Fig. 8-26. Geology, mineralisation, soil sample sites and COS concentrations at North Silver Bell, Arizona (from Hinkle and Dilbert, 1984).
Sulphur gases
32~
285
11le32"30"
111'31'00"'
"'.~.'~'-
M/~/ff /'~t~L
r
L,, 0
,,, ;
0.5
t,
0.5 I
/
~
'
~
)~
' - ~ 1 7 6- t
I MILE I KILOMETER
, ,.,;,.~"
...,
EXPLANATION
~ ' / / ~ Potonie alteration 9 Samplesite Sulfur dioxide in head-space gas Range: not detected at lower limit of detection (50 pl~)-lOO0 pDb ~ z
300 ppb
Fig. 8-27. Geology, mineralisation, soil sample sites and SO2 concentrations at North Silver Bell, Arizona (from Hinkle and Dilbert, 1984).
Crandon, Wisconsin The Crandon massive sulphide deposit is found in Precambrian volcanic rocks of the Rhinelander-Ladysmith greenstone belt, which trends east-west across the northern part of Wisconsin. Mineralisation occurs in a sequence of volcanic and associated sedimentary rocks which strike about N85~ dip 70-90~ and consists of laminae of sphalerite in pyrite with minor amounts of galena and chalcopyrite. Stringer sulphide mineralisation underlies the massive sulphide ore and consists of quartz-chalcopyrite-pyrite and pyritechalcopyrite-sphalerite veins. The deposit is covered by up to 65 m of glacial drift. McCarthy et al. (1986) made nine soil-gas sampling traverses at Crandon: seven traverses across the deposit; and two traverses away from the deposit for background determination. Sample spacing along the traverses was 15 m over and near the deposit and 30 m away from it. Soil gas was sampled by driving a hollow probe 0.5 m into the ground. Soil gas was withdrawn via small holes near the tip of the probe by inserting a hypodermic needle attached to a syringe through a rubber septum sealed into the top of the probe. The first gas sample withdrawn from the probe was discarded, and the second sample was injected into the inlet system of a quadrapole mass spectrometer mounted in a four-wheel drive vehicle. The mass spectrometer has a range of 1-100 atomic mass units, which
286
M.E. Hinkle and J.S. Lovell
includes the range for sulphur gases. A record of the intensity (relative concentration) of each mass unit in the gas sample was printed out on a paper tape for each analysis. From the range of gases determined at Crandon it was evident that the concealed massive sulphide deposit is indicated by positive CO2 anomalies coincident with negative 02 anomalies. However, the sulphur gases H2S, COS, CS2 and SO2 were found to be absent.
Kazakhstan
Although there has been considerable development of gas geochemistry methods in the former USSR since Glebovskaya and Glebovskii (1960) first suggested the use of sulphur gases, adequate descriptions of successful programmes are not readily available in the literature. The best-documented surveys are those carried out in central Kazakhstan by Elinson et al. (1970). The dry climate and low water table in this region were thought to limit the loss of sulphur gases due to solution in the groundwater. Traverses were made across four copper-molybdenum deposits and soil gas was collected from holes drilled to a depth of 1.0-1.8 m. The soil air was analysed titrimetrically after absorption in an iodine solution. Control analyses were performed by mass spectrometry. The ore deposits, with widths of between 10-400 m, are associated with skarns and granite porphyries. The thickness of the overburden is between 0.6-3 m. The ore minerals are principally chalcopyrite, molybdenite, chalcocite, bornite, galena, sphalerite, magnetite and pyrite. Weathering and secondary enrichment have affected the ores to a maximum depth of 95 m. Sulphur gases, which were not speciated, are clearly anomalous over the ore bodies, some of which were previously undiscovered (Fig. 8-28). As a follow-up to the surface soil-gas determinations the gases dissolved in the mud of drill holes were determined. These revealed higher concentrations of sulphur gases to be associated with the betterdeveloped mineralisation.
Ireland
Sub-economic massive sphalerite and galena occupy high-angle faults at Keel, County Longford, Ireland. In this area, the bedrock and mineralised faults are concealed beneath by 2-7 m of Pleistocene glacial till. Hale and Moon (1982) used gas chromatography to determine COS desorbed from surface soils samples collected along a traverse crossing the mineralised faults. Background concentrations of COS desorbed from soils were less than 750 pg g-~, whereas concentrations of COS up to 1500 pg g-~ occurred directly above the mineralised faults (Fig. 8-29).
Sulphur gases
287
~s%
zs% 1
PROFILE
VI
1.1
g
0.9
r
0.7
COS s
~
Pg/g 3
c~~,o_
!
ZS%
1.1
PROFILE
0.9
III
o.s
IS%
1.1
PROFILE
IX
0.9
0.7
0.7
0.5
0.5
0.3 0.1
o.3
0.1
0.3 _
0.1
Fig. 8-28. Cross sections along copper-molybdenum skarns in Kazakhstan showing total sulphur in soil gas (reproduced with permission from Elinson et al., 1970).
DISCUSSION In geochemical exploration, soil samples, either surficial or (in arid regions) microlayer are the recommended sample medium for the collection of sulphur compounds in soil gases. Soils should be placed in plastic bags or glass containers that are as air-tight as possible. An entire study area should be sampled at one time, over a period of as few days as possible, to avoid weather-related or seasonal fluctuations in soil-gas concentrations. The soil samples should be stored in a cool place or in a freezer. All of the soil samples from a particular area should be analysed as quickly as possible after collection. They should be sieved to the same mesh size and should be of the same weight for analysis. No generalised guidelines for data interpretation can be made. The main problem in interpreting the data is to distinguish sulphur compounds associated with mineralisation from sulphur compounds arising from other sources. The problem of distinguishing between natural and artificial sources of sulphur compounds becomes acute near populated areas. Anthropogenic sources such as air pollution, agricultural activities (dust, fertilisers, pesticides and herbicides) and mining activities (dust, tailings and smelters) produce various sulphur compounds. Natural sources of sulphur compounds such as decaying vegetation in forested or swampy regions must also be differentiated. Each study area must be interpreted individually. The sampling and analysis techniques and interpretations change depending on the environment in which they are used (wet or dry, cold or hot, polluted or non-polluted, different kinds of soil and vegetation cover). A
288 a z
M.E. Hinkle and J.S. Lovell
15oo
oo
0 10oo (,'3~ 0
~
500-
I-. ~
0
=
1000
.....
,,
. . . . . .
I
,,I
1200
1400
\
~Sulftdes
16oo METERS
/
...~
Fig. 8-29. Cross section along the sphalerite deposit at Keel, Ireland, showing COS in soils (reproduced with permission of the Canadian Institute of Mining, Metallurgy and Petroleum from Hale and Moon, 1982).
certain technique and interpretation may be valid for one or more of these environments, but not for all of them. Whilst the general pattern of anomalous sulphur-gas concentrations in a study area remains fairly constant from season-to-season, the concentrations themselves may vary. Therefore, analytical results should be interpreted on samples that were collected during the same time period. Interpolation of results obtained from different field seasons will probably lead to confusion. Sulphur-gas geochemistry is not, ideally, a "stand alone" technique. It should be used in conjunction with other geochemical, geophysical and geologic studies to understand the processes that cause and influence sulphur-gas anomalies around and over mineral deposits. When the processes become more fully understood, predictive models may be developed for the occurrence of volatile sulphur compounds over different types of mineralisation and over buried mineralisation. Future work will then expand the use of sulphur-gas anomalies from studies of specific mineral deposits to regional and reconnaissance studies.
CONCLUSIONS Although considerable interest exists in using sulphur-gas anomalies as indicators of mineralisation, and especially of buried mineralisation, it is apparent from the foregoing data that the routine use of sulphur-gas geochemistry is still some distance in the future. The extremely encouraging results from the Johnson Camp and North Silver Bell areas of Arizona serve to demonstrate that gaseous dispersion of sulphur gases through exotic overburden can occur and may extend a diagnostic signature to the surface. Surveys
Sulphur gases
289
employing direct detection of free sulphur gases within soil air have met with only limited Success.
To summarise, therefore, there is sufficient promise to encourage further research in this field, and it should be aimed at improving the analytical system and detection limits and at defining the causes of background variability unrelated to mineralisation. Furthermore, it is not yet clear how an inorganic non-volatile sulphide precursor can be the source of gaseous sulphur species, although it is probable that there may be a microbiological intermediary. It is to be hoped that research in this field will continue.
This Page Intentionally Left Blank
Geochemical Remote Sensing of the Subsurface Edited by M.Hale
Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
291
Chapter 9
SULPHIDE ANIONS AND COMPOUNDS
X. SUN
INTRODUCTION During the formation of a sulphide mineral deposit there is usually the development of a primary sulphide halo in the host rock. Subsequent to ore formation, sulphide anions and compounds may be dispersed in the aqueous phase in groundwater and the gas phase in air-filled pore spaces to form secondary dispersion patterns. In exploration for sulphide mineral deposits, the simple determination of mobile sulphur species originating from sulphide ore deposits has a clear attraction. In order to understand the way in which such sulphur species occur in soils, a number of in-vitro experiments were performed. On the basis of the results, suitable methods of sampling and analysis have been chosen and tested in soils overlying 30 different mineral deposits.
EXPERIMENTAL INVESTIGATIONS Although H2S is considered to be a gas commonly generated during the oxidation of sulphide mineral deposits (see Chapter 8), no more than 0.01 mg 1-1 could be found in the pore spaces of soils above mineral deposits. Therefore simulation experiments were performed to investigate: (a) the affinity of soil for HzS adsorption; (b) the method of HzS transport through soil; and (c) the redox conditions favourable to HzS persistence in soil. These experiments are very simplistic and there are some difficulties in relating them to the natural environment. They do, however, contribute to understanding the behaviour of H2S in soils and how it can be used in mineral exploration. The limit of detection of the method used to measure H2S in these experiments is about 0.01 mg H2S.
Soil adsorption of hydrogen sulphide Three glass tubes (4 mm and 5 mm in diameter) were partially filled with <40 mesh, HzS-free, dry soil and plugged with cotton wool to form soil columns (70 mm and 45 mm in length). These soil columns were the basis of experiments comprising four steps. In the
292
X. Sun
Fig. 9-1. Schematic diagram of experiment for soil adsorption of H2S.
first step, the lower end of each tube was connected to a H2S source and the upper end led via a detector containing lead acetate paper to a pump (Fig. 9-1). The H2S source was a mixture of Na2S, HCI and SnCI2, the amount of Na2S being varied to generate 10, 5 and 2 mg of H2S. The gas generated in each case was drawn through one of the soil columns to the detector by the pump. After 3 min. the H2S supplies were assumed to be exhausted and disconnected. The lead acetate papers were removed and the amount of HzS each had trapped was determined by comparison with reference charts. In the next step, air was pumped through each tube for a further 3 min. and passed through fresh lead acetate papers. In the third step, each soil column was dismantled and the soils placed into reaction tubes, to which were added 5 ml of 50% HC1 containing 5% SnC12; the tubes were connected via fresh lead acetate papers to the pumps for 2 min. to determine the amounts of acid-released HzS in the soils. Finally, 0.5 g of iron powder was added to the remnants of each of the reaction mixtures and, with fresh lead acetate papers in the lines, the pumps were reconnected for a further 2 min. to determine the amounts of HzS in the soils released only by reduction. The results of this experiment are shown in Table 9-I. The small quantities of H2S eluted directly to the lead acetate detectors in the first step show that HzS gas is strongly adsorbed by the soil in the columns, from which it is deduced that it is difficult for HzS to migrate in the gas phase through dry soil. As the second step shows, passing air through the soil does not lead to desorption of HzS, thUS there is little possibility of detecting HzS in the soil by an active pumping method. But in the third step, most of the adsorbed HzS is released by acid addition, indicating that much of the HzS is held in the soil by chemisorption. Converting to reducing conditions in the fourth step releases most of the remaining HzS, which shows that even in a short period of time, some of the H2S adsorbed on the soil is oxidised. Overall this experiment reveals that soil adsorption affects the migration and persistence of HzS in soil,
Sulphide anions and compounds
293
and further this sorption is chemical and cannot be reversed by a simple physical method; chemical desorption, however, can extract most of the H2S adsorbed on soil.
TABLE 9-I Experimental results of adsorption of H2S by soil Design and results of experiments
Experiment no. 1
Experimental conditions
Weight H2S recovered (mg)
2
3
Weight H2S at source (mg)
I0
5
2
Diameter soil columns (mm) Length soil column (mm)
4
5
4
70
45
70
Pumping rate (ml min ~)
200
300
200
Step 1, direct elution Step 2, air desorption Step 3, acid desorption Step 4, reducing desorption Balance not recovered
0.20 0.0 7.00 2.00 0.80
0.60 0.0 3.00 0.80 0.60
0.01 0.01 1.00 0.30 0.68
Hydrogen sulphide transport through soil Whilst H2S gas may be a product of simulation experiments, its occurrence and detection in natural soils depends upon its transport and persistence m the gas phase. Another experiment was therefore designed to investigate the method of HzS migration m soil above a simulated oxidising sulphide deposit. A mixture of <60 mesh pyrite, chalcopyrite, galena, sphalerite, calcite and silica was placed in a 30 ml reaction tube, then covered by calcite and silica gangue of 1 to 3 mm gram size to a total volume of 22 ml. Water at pH 6 was added to bring the volume to 25 ml, and finally 1 ml of 30% H202 was added to ensure oxidismg conditions. Through the seal at the mouth of the reaction tube, a glass tube, 980 mm long, 4 mm wide and containing <60 mesh HzS-free dry soil, was fitted so that the lowermost 15 mm was immersed in the solution. The soil was supported m the tube by a purified cotton wool plug and a second plug impregnated with lead acetate occupied the top 30 mm of the tube, above the soil. After 45 days the experiment was concluded by noting the conditions in the soil column, then dismantling it and analysing the soil at various points along it for sulphur species that could be determined as acid-released H2S.
294
X. Sun
Fig. 9-2. Summary of results of experiment to investigate the method of H2S migration in soil above a simulated oxidising sulphide deposit.
The results are summarised in Fig. 9-2. The solution rose up the column to a height of about 715 mm, of which the lowermost 550 mm was stained yellow-green by what was subsequently found to be FeSO4. There was not only enrichment of sulphate but also, as the analytical results show, of other sulphur species, found as far as the uppermost margin of soil saturation at 715 mm. Still higher in the column, m dry soil, there was no detectable sulphur, and hence no evidence of gas-phase transport.
Redox conditions for hydrogen sulphide persistence The persistence of H2S in soils was examined by immersing 80 g of sulphides and gangue of mixed gram size m an aqueous solution of pH 5 m two 500 ml narrow-neck bottles. Into one bottle only, 0.5 ml of 30% H202 was added to make conditions more oxidising. Both bottles were sealed m a way which allowed headspace gas to be sampled via a tap. After 8 hours an air sample from each bottle was analysed for H2S using lead acetate paper; samples were subsequently analysed at random intervals ranging from 2 to 8 hours. After three days, two lead acetate papers were introduced into each bottle, one suspended m the headspace and the other just touching the solution. The bottles were warmed to 60~ allowed to cool, then the effect on the lead acetate paper was recorded.
Sulphide anions and compounds
295
During the first three days of the experiment no H2S was evolved from the more oxidising solution and only trace amounts, unrelated to the experiment duration, were found above the more reducing solution. After the bottles were warmed the suspended lead acetate paper above the more oxidising solution was more intensely coloured (30 x 8 mm) than that above the more reducing solution (15 x 8 mm). This result, however, is attributed to COS in the headspace reacting to form H2S on the lead acetate papers after they were moistened by evaporating water: COS + HzO ~
H2S + CO2
The slow (rather than rapid) coloration of the suspended paper is consistent with the slow rate for the above reaction. For the lead acetate paper in contact with the warmed solutions, that in the more oxidising solution was coloured only slightly while that in the more reducing solution was extensively coloured. These results indicate the presence of sulphide anions and/or compounds in both bottles, but in amounts inversely proportional to the Eh of the solutions. From this experiment it can be inferred that, in the vicinity of oxidising sulphides, sulphide anions and/or compounds occur mainly in reducing solutions. Only very small amounts of H2S occur in the gas phase, even over solutions of low Eh. Near to the point of oxidation, COS is formed. These findings conf'trm that, in moist soils, any HzS will be in solution and thus gas-phase migration 9f H2S will not be possible.
Conclusions of experimental investigations The experimental investigations show that H2S does not migrate in the gas phase through soil. However, it migrates through water-saturated soil, probably in solution as sulphide anions and compounds. These are subsequently adsorbed onto soil particles, from which they can be released as HES. This can be determined in concentrations as low as 0.01 mg using lead acetate paper.
FIELD INVESTIGATIONS Two methods have been devised for the detection of the dispersion pattems of sulphide anions and compounds in soil: (a) the shallow hole method; and (b) the container method. In the first method, a 18 mm diameter steel rod is driven into the soil to make a hole that acts as the sample container. After the rod has punched a hole in the soil, it is removed and a device comprising an acid sprayer and gas extraction tube is inserted. (Fig. 9-3). The spray head must be at least 40 cm below the surface and the mouth of the hole must be thoroughly sealed by the spray device. Then the outlet tap of the spray reservoir is opened to allow 8 ml of 30% HC1 to reach the spray head and to spray onto the walls of the hole below 40 cm
296
X. Sun
soil
II[,;"!': ~~ ,[
L.Ill (21
(1) acid reservoir (2) acid spray head (3) detector (4) air pump
Fig. 9-3. Schematic diagram to illustrate the shallow-hole method for determining acid-released HzS.
depth. The tap is closed when the acid has flowed out, and the pump is switched on to extract the air from the hole for 1 min. and pass it to the detector (lead acetate paper). This method, used in the case histories described below, has the advantages of high anomaly contrast and good reproducibility. It is possible to determine quickly the areal extent of an anomaly in the field. Its disadvantages are longer time and more work in the field than are required with the container method, which can be used for determination of acid-released H2S in a (field) laboratory. In the container method, a few grams of soil are put into a 20 ml glass container, 4 ml of 50% HC1 containing 8% SnC12 is added, then the tube is sealed with a rubber bung through which run two air-conducting tubes. The equipment is attached to a pump, and the H2S released in pumped through lead acetate paper. The air inlet tube must be narrower than the outlet tube and reach almost to the bottom of the glass container (Fig. 9-4). In both methods
Air
re action tube- .
soil sam ple
Fig.9-4. Schematic diagram to illustrate the container method for determining acid-released H2S.
Sulphide anions and compounds
297
the limit of detection of 0.01 mg H2S is effectively independent of sample weight. Surveys for dispersion patterns of sulphide anions and compounds determined as acidreleased HzS in soil have been camed out over 30 mineral deposits in China, including skams, porphyry copper deposits and porphyritic iron deposits (sulphur contents higher at the margin of the ore body), altered-brecciated gold deposits and lead-zinc deposits in volcanic breccia. Three successful case histories are summarised here. Of the remainder, only four failed to yield anomalies over mineralisation.
Mineralisation beneath thick transported overburden One test of the method was conducted over a medium-sized iron-copper deposit near the southeastern coast of China. The ore body occurs in a skam near to the contact of granodiorite with interlayered marbles and homfels. The attitude of the ore body is controlled by the contact, which generally is steep. The strike of the ore zones is approximately east-west, plunging west, and their western contact is westerly convex. The top of the ore body, which may have been eroded, is now covered by 140 to 160 m of transported sediments. There is a thick soil cover; the ground surface is quite fiat and suitable for arable farming. Four traverses were laid out at intervals of 100 to 200 m, with a distance between sample sites of 20 m. Three of these traverses (a-c) crossed a known ore body, whereas the most westerly traverse (d) was over an unexplored area.. Acid-released H2S anomalies occur above the known ore body (Figs. 9-5 and 9-6). The most westerly traverse (d) has anomalies
0 d
l O O m ~
~
350" ,
(9'
(9
(9
'
(9
d' '
(9 '
r~5:
',
Hf~
-
'"
0
0
"
0
b
o
l
o', 6. . . . . ' SK ' Hf
.[-O-~._.._drlllh o l e s prior
to experiment
~O--~dri|! h o l e s after
experiment
Fig. 9-5. Acid released H2S content of soils along four traverses over an iron-copper deposit in southeastern China; concealed solid geology, Hf = brecciated schist, r8 = granodiorite, SK skarn.
298
X. Sun
far away from the projected position of the contact (Fig. 9-5). These were thought to reflect a separate ore body, as was proved a year later when a cupriferous pyrite ore body was discovered (Fig. 9-7). Beneath the other three traverses this second ore body is too deeply buried to yield anomalies.
Fig.9-6. Section along traverse 'a' showing acid-released H2S content of soil over an iron-copper deposit in southeastern China; Hf = brecciated schist, r8 = granodiorite, SK -- skarn, Q = Quaternary sediments.
Fig. 9-7. Mineralisation discovered by drilling to investigate acid-released H2S anomalies along traverse 'd', adjacent to a known iron-copper deposit in southeastern China; r8 = granodiorite, SK = skarn, J~g = Jurassic sandy shale, Q = Quaternary sediments.
Sulphide anions and compounds
299
Fig.9-8. Section along traverse 'b' (Fig.9-9) showing acid-released H2S content of soil over a lead-zinc ore body beneath thick lithic cover in China; Klcon = Cretaceous conglomerate; Kid = Cretaceous red sediments, Jlgn = Jurassic Lingkou Group, P~m = Permian Maokou formation, P~q -- Permian Qixia formation, P2L = Permian Leping formation.
Mineralisation beneath thick lithic cover Further investigations were carried out over a large lead-zinc ore body in China, in which ore zones occur in both flanks of a reverse anticline. The ore is richer in the eastern flank, where it is controlled by an interlayer breccia zone. Smaller amounts of ore occur in fractures and fissures in siliceous limestone. The ore body is buried beneath 350 m of barren rock in the south and warps up northward, where it is buried beneath 250 m of barren rock.. The ore is richer in the north than in the south. The hangingwall rocks are conglomerates and red-bed sediments, in which interstitial water is abundant. The topography of the area is rugged. The soil is brick red and 0.8 to 10 m thick. Four traverses were laid out across the area, with sample sites at 20 m intervals along each. There are H2S anomalies above the ore body (Fig. 9-8) and the H2S anomalies are higher in the north than in the south (Fig. 9-9), corresponding to ore grade and depth of burial.
X. Sun
300
N ~
.
9
/9.
,--Y. ......> W
,/~
ii, 9 0!.~\
I,/ 82 ,
/
I
/
/
Kid
/ o
idrlll hole
o
,
\
J,gn
F~
il'~\ / ~; 1/
........ "i/ i unconformity / P=L\\ ._------.drill hole not
9 /Intersecting ! ~rnlnerallzstlon "-
150 m
d
"\\ /
p
o Jlntere.ctlng
" mineralization
\ J~gn
\\
\\ \
I~]
baselIne end HaScontent
Fig. 9-9. Acid released H2S content of soils along four traverses over a lead-zinc ore body beneath thick lithic cover in China; K~d = Cretaceous red sediments, J~gn = Jurassic Lingkou Group, PzL = Permian Leping formation.
Mineralisation beneath mixed eluvium and transported sediment A third investigation was performed where a large lead-zinc-silver sulphide ore body occurs in the mountains of eastern China. Ore zones strike east-west, controlled by a boundary fault between more competent quartzite and less competent dolomite and limestone. The ore body mainly occurs in the limbs of a reversed anticlinorium, is steeply dipping and replaces carbonaceous shales and sandstones interlayered with silt. Sandstones of the basal Jurassic overlie the reversed anticlinorium with steep angular unconformity. The thickness of the sandstones is 200 to 300 m. Some smaller lead-zinc-pyrite-uranium deposits occur in the unconformity and at lithological boundaries. The mineralisation was affected by the Yanshan Orogeny, which produced deep faults cutting the ore bodies and unconformity. The surface of the area is covered by up to 30 m of eluvium and transported sediments. The vegetation is mainly pine forests with arable land at the foot of slopes.
Sulphide anions and compounds
301 0 i
125
0
15
5
r
25
300 , i
600 i
900m J
35
45
55
35
45
55
~ .........
i
Q~D
.. s
........r 0
15
25
~_ ~ . 0 ~ H:S content contour line
~Pb-Zn
5
surface projection of P b - Z n deposit
9 drill hole intersecting mineralization
Fig.9-10. Acid released H2S contours in soil over a lead-zinc-silver ore body in eastern China; outer contour = 0.1 mg H2S, inner contour = 0.2 mg HzS.
A survey was carried out on a grid covering an area of 13.6 k i n 2. Lines were 100 m apart and sample sites 20 m apart along each. There is a series of elongate HzS anomalies above the known ore zone, including where this is covered by transported sediments, and other anomalies indicate mineralisations within fractures and fissures (Fig. 9-10).
DISCUSSION The immediate source of acid-released H2S in soil is sulphide anions, for example, S 2-, HS-, Me(HS),-, and/or mobile sulphur compounds, such as HzS, COS, CS2. At depth, circulating groundwater provides a medium through which sulphide anions are readily transported in aqueous solution. The results of in-vitro experiments indicate that, in the open pore spaces of overburden and soil, it is unlikely that HzS exists in the gas phase. This and the other sulphur gases are discussed further in Chapter 8.
Undetected mineralisation In the four soil surveys above ore deposits that failed to exhibit acid-released H2S anomalies, it was noted that the overburden was dry. This implies a highly oxidising environment not conducive to the preservation of reduced sulphur species. In one case the
302
X. Sun
lead acetate paper was whitened, probably due to the presence of S O 2 formed after soil acidification. Such bleaching interferes with the detection of any acid-released H2S.
F a l s e anomalies
Sources other than ore deposits can give rise to acid-released H2S anomalies and so obscure expressions related to mineralisation. The main sources appear to be: 9 9 * * * 9
sediments at the bottom of lakes, ponds and streams; marshes, swamps and rice-fields; soil mixed with building material (limestone, cement, coal ash, etc.); ore waste, tailings, smelters and refineries; urban areas and roads; livestock farms.
The above sources can generally be observed in the field. In the case of low level or subtle interference where a shallow source is difficult to identify by eye, discrimination may be achieved by acquiring additional geochemical data. An anomaly may be due to the decomposition of organic sulphur compounds. Acid-released H2S of this origin has a strong positive correlation with organic carbon, with the result that the ratio of acid-released H2S to organic carbon is more-or-less uniform. Higher ratios pick out sources of inorganic sulphur such as sulphide anions and compounds. Alternatively, increasing sample density radially around an anomaly and calculation of the deviation from the mean of each measurement helps to characterise the source: an uneven pattern of deviations indicates a biogenic source; a more regular pattern is indicative of mineralisation or some form of soluble pollution. Finally samples from depths in excess of 0.4 m usually avoid biogenic sources and soluble pollution, and hence yield results that can unambiguously reflect mineralisation. Sampling below 0.4 m can be achieved with the container method.
CONCLUSIONS A method for detecting sulphide anions and compounds as acid-released H2S has been shown to yield good-contrast anomalies in soils overlying mineral deposits. Satisfactory results have been obtained in different overburden materials of varying thickness, under which mineral deposits are concealed at considerable depths. The equipment and procedure for measuring acid-released HzS in soil are simple, rapid and efficient, can be widely used and are easily adapted.
Geochemical Remote Sensing of the Subsurface Edited by M.Hale
Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
303
Chapter 10
HELIUM C.R.M. BUTT, M.J. GOLE and W. DYCK
INTRODUCTION The principal isotope of helium, 4He, is generated during the radioactive decay of isotopes of uranium, thorium and a few other elements that may either be components of some mineral deposits or be situated in the basement or country rocks. Being a highly diffusive and inert gas, it has been considered as a possible pathfinder for deposits containing these elements, particularly those that are blind or buried. Exploration targets that might be sought include: 9 deposits containing uranium and/or thorium in sufficient quantities to act as an adequate helium source, for example, uranium deposits, some coals, mineral sands, carbonatites; liquid and gaseous hydrocarbons, which can be enriched in helium derived from the basement or thorium- or uranium-bearing source rocks; 9 deposits associated with, and emplaced along, faults and fissures that act as pathways for the escape of basement-derived helium, and bodies such as kimberlites and carbonatites, which are themselves brecciated or faulted, and could act as pathways for the escape of either basement-derived helium or helium contained or generated within the rocks themselves; 9 geothermal areas, which are the focus for the release of helium contained in juvenile or circulating meteoric water. Basement-derived helium may also leak along unmineralised faults, including active ones, and in consequence its detection may have application in locating faults in areas of poor outcrop. Short-term fluctuations of the helium flux have been suggested as possible predictors of earthquake activity. In addition, dissolved helium contents have been used in attempts to estimate the ages of groundwater, and determinations of helium isotope ratios have been used to indicate whether some gaseous emissions originate from the crust or the mantle.
Revised manuscript of this chapter prepared 1986; further minor revision 1995.
304
C.R.M. Butt, M.J. Gole and W. Dyck
Investigations into these possible applications of helium surveys have been conducted in North America, the former USSR, Australia and northern Europe over the last three decades. Initial results were commonly quite encouraging, but further studies have demonstrated that their potential in mineral exploration is limited and there has been little research and few publications since about 1987. Nevertheless, applications in hydrocarbon exploration and earthquake prediction remain possible. Total He analysis is ineffective for dating groundwaters but He isotope ratios are routinely applied to distinguishing mantlederived gases. In this chapter, the occurrence and properties of helium are briefly outlined, followed by a description of appropriate sampling and analytical techniques and reviews and assessments of the possible uses of helium surveys.
OCCURRENCE
Discovery Helium was recognised as an element in 1895, but the history of its discovery dates back a further 30 years. The bright yellow He lines in the solar spectrum were first noted, but not identified, in 1868. In 1888, helium liberated when uraninite is treated with acid was mistaken for another inert gas, nitrogen.
Abundance and origin Helium is the second most abundant element in the Universe after hydrogen; both elements are concentrated in the stars. The He abundance in the Sun is 3 x 103 that of Si, whereas on Earth it is only 5 x 10-~2 that of Si. There are two naturally-occurring isotopes, ~He and 4He, which may be either primordial or radiogenic in origin. Primordial 3He is formed by the collision of hydrogen and deuterium atoms: ~H + tH = 2H + e + n
2H + ~H = 3He Primordial 4He is then formed by collision of two 3He atoms: ~He + 3He = 4He + 2H The two He isotopes have approximately the same abundance throughout the universe. Some primordial helium was trapped during the accretion of the Earth, with an estimated 3He/4He ratio of 10-4. This is still retained in the mantle and, to a lesser extent, the core, and is now outgassing along deep faults and fissures and during volcanic eruptions. The
Helium
305
3He/aHe ratio of these gases is about 10-5, the difference being due to the generation of radiogenic 4He since accretion. Both isotopes of He can also be radiogenic in origin. Most terrestrial and atmospheric 4He is radiogenic and has formed since the accretion of the Earth as a product of the alpha-decay of some naturally-occurring radioactive isotopes. On ejection from its parent isotope, an alpha particle (which is a He nucleus comprising two neutrons and two protons) attracts two electrons to form a 4He atom. By far the most important sources are 238U, 235U and 232Th, which yield 8, 7 and 6 4He atoms, respectively, during decay half-lives of up to 10 ~~years. The other alpha-emitting radioelements, principally isotopes of Sm, Nd and Pt, have far longer half-lives, in the range 10~t-10 ~5 years, and undergo only single decays, with the result that their contribution to the total 4He flux is negligible. Although most terrestrial 3He is primordial, there is a significant radiogenic component in both the crust and the atmosphere. The main source is the decay of tritium: 3H - 3He + 13(tu = 12.35 years). In crustal rocks, the tritium is generated by (an) reactions: OLi n
= 3H +
4He
7Li c~ = 3H + 24He The neutrons are considered to have been derived from the spontaneous fission of 238U (Morrison and Pine, 1955). Tritium is also generated by bombardment by cosmic radiation of the atmosphere and the Earth's surface. Since U deposits do not contain much Li, production of 3He by this process is small compared to the amount of radiogenic 4He generated by the decay of U itself. A significant proportion of atmospheric 3He is cosmogenic, generated by spallation reactions from cosmic radiation. The testing of hydrogen bombs has added some tritium and hence some 3He to natural levels. On balance, production of terrestrial radiogenic He is dominated by 4He, hence the 3He:He ratio in the crust is now reduced to 10-6.
Abundance in the Earth and atmosphere There have been relatively few systematic descriptions of noble gas concentrations in rocks and minerals. In general, the 4He content is dependent upon the abundances of Th and U, the crystal density of host minerals and the permeability of the rock. Mantle-derived rocks and minerals seem to be relatively enriched in He, with high 3He/4He ratios. Thus, minerals and nodules in kimberlites can have He contents up to 2755 cm3/g x 10.8 (phlogopite) and 350 cm3/g x 10.8 (diamond); in contrast, many igneous rocks may contain <0.1 cm3/g x 10.8 (Hoppe and Alexander, 1978). Very high concentrations of He can occur
306
C.R.M. Butt, M.J. Gole and W. Dyck
in pores in sedimentary rocks, associated with oil and gas accumulations. Concentrations of a few hundred ppm are normal, but concentrations may reach several percent by volume in some instances, these being the commercial sources of He (Boone, 1958). Concentrations of He in some natural gases and waters are given in relevant sections below. The He content and isotope ratio of the atmosphere is a dynamic equilibrium maintained by degassing of the Earth and thermal escape into space. The 4He flux is largely accounted for by the degassing of radiogenic 4He from the crest, and the 3He flux by degassing of primordial 3He from the mantle. The high diffusiveness of He and mixing by wind ensure that the atmospheric He content is effectively uniform worldwide. Determinations by Olivier et al. (1984) and by Holland and Emerson (1987, 1990) have shown the value to be 5.22 ppm He. Analyses reported in this chapter and the earlier literature, however, use 5.24 ppm as a standard (Gluechauf, 1946). The He content of water in equilibrium with the atmosphere at STP is 0.0440 ~tL He/L H20 (Weiss, 1971). The 3He/4He ratio in the atmosphere is 1.4 x 10-6 and waters in equilibrium with the atmosphere have the same ratio. The constant and uniform abundance of He in the atmosphere and hence of the equilibrium concentration in water is extremely important in establishing background values for He concentrations in soil and overburden gases, surface water and groundwater, which are used as sample media in He surveys. Because of the radiogenic origin of 4He, there have been a number of attempts to calculate the ages of groundwaters from their He contents. Given an isolated system, with rocks of known U and Th contents, and assuming constant porosities and release of He from rock to water, the concentration of radiogenic He in the groundwater should be proportional to its age. However, these ages tend to be higher than those from t4C dating, often excessively so, given reasonable values for aquifer rock parameters (Heaton and Vogel, 1979; Andrews and Lee, 1979; Datta et al., 1980; Andrews et al., 1982). There may, nevertheless, be a linear relation between He contents and 14C dates. The excess He is considered to have been derived from leakage from other rock units (Heaton, 1984; Torgerson and Ivey, 1985), probably along faults, the linear relation implying this to have been at a fairly constant rate. Consequently, dates reliant entirely on comparing measured and calculated He abundances (e.g., Marine, 1979) may be erroneously high, since they assume the groundwater system to be isolated.
Isotope ratios The variation of the isotopic abundance of He in nature is by far the largest of any stable isotope ratio known. It spans a range of 101~with a minimum 4He/3He ratio of 1 for iron meteorites and maximum of 101~for certain U minerals (Kamenskiy et al., 1971). It has been established that the 3He/4He ratio has values characteristic of the source of the gas. Deep oceanic waters over rift zones and sea-floor spreading centres are found not only to be enriched in total He, relative to the solubility equilibrium with the atmosphere,
Helium
307
but also to have high 3He/4He ratios (Clarke et al., 1969; Craig et al., 1975; Lomonosov et al., 1976; Lupton et al., 1980). For example, enrichments reported from the Red Sea brines are factors of 3200 for 3He, 370 for 4He and a total abundance of about 14 ~tL/L H20 (Lupton et al., 1977). Calculations show that the ratio of the added He flux is 1.6 x 10.5 (Craig et al., 1975) and it is suggested that this is of mantle origin. Similarly, rocks of mantle origin (e.g., chilled glassy margins of oceanic pillow basalts) have a ratio of 10-5. It is concluded that the He in these deep waters is derived by degassing of freshly erupted basalts by circulating sea water. Crustal rocks exhibit a range of 3He/4He ratios (10-L10-8). The lower values are the result of degassing during eruption and relatively slow cooling, degassing during metamorphism and contamination by radiogenic 4He. Wakita and Sano (1983) have reported that high 3He/4He ratios, up to 8.65 x 106, occur in methane-rich natural gases from oil and gas wells in northeast Japan. These authors conclude that at least one-third of the gas could therefore be of magmatic rather than biological origin, although it is possible that only the He is magmatic. The extremely low 3He/4He ratios in meteoric waters, the oceans and atmosphere means that for most exploration purposes it is unnecessary to determine 3He contents. Unless otherwise specified, the properties and concentrations of He reported herein should be regarded as those of4He.
Properties" and migration Some physical properties of 4He are listed in Table 10-I. For exploration, diffusivity in sample media (air, water, soil, rock) and materials that could be used as sample containers, both of which are largely functions of the small atomic radius, are of prime importance, along with solubility in water. Water in equilibrium with the atmosphere at STP contains 0.044 ~tL He/L H20 (Weiss, 1971). The solubility of He in water decreases with rising temperature; thus, over the temperature range 0-40~ it declines from 0.049 to 0.042 ~tL He/L H20. Solubility is also depressed by salinity (Smith and Kennedy, 1982); for example, at 25~ the solubility decreases from 0.044 to 0.037 ~tL He/L H20 as salinity increases from 0 to 3.5% (seawater). The high diffusivity of He has resulted in the degassing of both primordial and radiogenic He from the lithosphere over time. For radiogenic He, ejection during radioactive decay of the precursor alpha particle with its associated energy of 6 Mev causes initial transport over a distance of 0.01-0.03 mm. This alpha irradiation also disrupts the mineral lattice and prepares pathways for the subsequent migration and loss of He. In general, only minerals with very tight lattice packing (e.g., corundum, rutile, spinel) can retain He effectively. Most rock-forming minerals have looser structures, with diffusion coefficients for He of the order of D = 10t~ cm2/sec. From the expression S2 = pDt, which relates distance travelled (S) to time (t), the diffusion within a mineral lattice is typically
C.R.M. Butt, M.J. Gole and W. Dyck
308
a r o u n d 1 m m per annum. W i t h increasing alpha irradiation and the d e v e l o p m e n t o f m e t a m i c t structures, minerals lose He readily (Faul, 1954), with retention rates o f b e l o w 10% o f the H e generated b e i n g reported (Gerling, 1939).
TABLE 10-I Physical properties of helium Property Atomic number Atomic mass Atomic diameter, 10-8cm Boiling point, ~
Value 2 4.0026 2.55 -269
Melting point, ~ (25 atm.) Density gas, g/L (0~ 1 atm.) - liquid, g/cm 3 (270~ - solid, g/cm 3 (250~ First ionisation potential, volts First ionisation energy kcal/g-mole Hc at vaporisation, cal/g-mole Solubility at I atm., mL (STP)/L - in water at 0~ in water at 20~ in water at 40~ Solubility from moist air at 1 atm., nL (STP)/L in water at 0~ in water at 20~ in water at 40~ Bunsen coefficients x 103, mL (STP)/mL in water at 0~ in water at 20~ in water at 40~ Diffusion coefficients, cm2/sec ~n air at 20~ in soil In water at 10~ in ice at-10~ In solid Si02at 20~ in solid Si02 at 500~ - ~n pyrex at 20~ - ~n pyrex at 500"C n granitoids m argillaceous schists
-272
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.178 0.125 0.361 24.6 567 19.4 9.78 8.61 8.38 49.0 44.8 42.6 9.41 8.73 8.69 0.7 No data (2-5) x 10.5 (0.6-3) x 10.5 (2-6) x 10 1~ (2-14) x 10.8 0.5 x 10-l~ 2 x 10.8 10-8_ 10-!o 10-6
Helium
309
Assuming crustal averages of 4 ppm U and 12 ppm Th, Dyck (1976) has calculated that the present annual production of 4He is 8.2 x 10-13 cm3/g. This compares to production of 31 x 10 -13 cm3/g/year at the time of formation of the Earth. If the mean annual production rate since the formation of the Earth at 4.5 x 109 years has been approximately 31 x 10 -13 cm3/g, the He content of a gram of "average crustal rock" should be 9 x 10.3 cm 3. However, only 10-30% of this amount is found, due to loss by diffusion into fractures and voids. For example 70-90% is lost from granites (Tugarinov and Osipov, 1974) and over 80% is lost from sedimentary rocks (Voronov et al., 1969). Helium diffusion rates are far greater through waters in fissures than through rocks; furthermore, any flow of these waters leads to rapid migration and, ultimately, to loss from the lithosphere. Experimental determination of He diffusion was attempted by Duddridge et al. (1991), who injected He-rich gas at a depth of 35 m into permeable limestones cut by a fault. They recorded a pulse of He in shallow soil gas 5-20 hours later within 10 m of the fault suboutcrop and up to 53 hours later 20 m from the fault suboutcrop. However, the concentration increase recorded (0.032 ppm) is well within the error of the analytical system (mass spectrometer with constant pressure inlet, as discussed below, and analytical sensitivity of 0.030 ppm), the data are patchy with many samples showing no pulse, and there is no estimate of background variation or the effect of changing environmental conditions. Conclusions about diffusion rates based on these data may not be reliable. Mathematical models attempting to predict the diffusion of He away from concealed U deposits have been presented by Novikov and Kapkov (1965) and Jeter (1980). The applicability of such models is uncertain, owing to the lack of diffusion coefficients for He in water-saturated rocks and soils and to the assumptions inherent in such models. Field data gathered by Jeter suggest a diffusion coefficient of D = 10 -4 t o 10 -s c m 2 / s , i.e., about that for water; applying the latter value to an idealised planar body, 1 m thick, at a depth of 100 m and containing 0.6% U308, the model suggests that overburden gases 20 m above will have an excess He content of 3.5 ppm higher than the atmospheric concentration of 5.24 ppm. This excess would decline with increasing distance from the source. From the equations presented by Novikov and Kapkov (1965), over a 0.1% U source 100m below surface, soil gas at 1 m depth would have an excess of 0. 15 ppm He, with D - 10 .4 cm2/s. Given that in air D = 0. 7 cm2/s, a more realistic diffusion coefficient would be much higher; thus with D = 10-2 cm2/s, the excess He calculated by these models would be only 0.0005 and 0.0002 ppm respectively. Such excess concentrations are, of course, negligible and well within sampling and analytical errors, so that if the models have any validity, they predict that both overburden and soil gas sampling will have little application in exploration. This prediction has apparently been borne out by much subsequent experience (Butt and Gole, 1985). Diffusion is not the only mechanism by which He (or indeed any gas) migrates. The evaluation by Newton and Round (1961) suggested that, as a significant proportion of He in sedimentary rocks is dissolved in water, the migration of He is dominated by flow, even if the flow rate is estimated at only 30 cm per annum. Therefore, once He enters an aquifer, it is transported rapidly and essentially unidirectionally by flow, compared to the random
310
C.R.M. Butt, M.J. Gole and W. Dyck
migration due to diffusion. Several million He determinations on groundwaters from the former USSR have convincingly corroborated the hydrologically-dominated flow of He (Golubev et al., 1974; Eremeev et al., 1973; Ovchinnikov et al., 1973; Yanitskii, 1979; Ivanov et al., 1979). Fault and fracture zones in the basement and their extensions into overlying sedimentary rocks constitute major pathways for upward He flow, particularly in tectonically-active regions, and these are the principal source of the observed He distribution pattems. The driving forces behind the upward flow include pressure gradients, regional and local hydrostatic heads and lithostatic formation pressures.
SAMPLING Samples used for He surveys have included interstitial free gases in soils and overburden, gases extracted from soils (including those in regions of permafrost) and lake sediments, gases and waters issuing from springs, lake waters and groundwaters from various depths. Appropriate procedures have been developed for collecting and storing these samples with minimal loss of He or contamination by other gases. The high diffusivity of He has a great bearing on the selection of sample containers. In general, only glasses and metals are sufficiently impermeable to retain He for periods longer than a few hours without appreciable loss. Plastics are highly permeable to He and are quite unsuited for sample storage. For glass, the rate of permeation decreases rapidly with increasing density, with diffusion being the rate-controlling step in the permeation. For example, lead borate glass (permeability constant of 10 -17 for He) provides a superior container compared to soda lime glass ( 10-14) or pyrex ( 10~ !).
Soil and overburden gases In the majority of soil gas surveys, samples have been collected using hollow steel probes hammered 0.5-1.0 m into the soil (Reimer et al., 1979b; Pogorski and Quirt, 1979; Gregory and Durrance, 1985). After being driven to depth, and perhaps allowed to equilibrate for one to five minutes, the probe is purged by withdrawing gas through the septum with a syringe and hypodermic needle. Further gas is then withdrawn for analysis. Probe sampling may be unsatisfactory, however, if there are difficulties of insertion to a suitable depth, or contamination, or equilibration with the atmosphere via existing fissures or cracks formed during insertion of the probe. Even where these general problems do not appear to arise, the shallow depth of sampling attained with a probe may fail to provide an appropriate sample, due to the high diffusibility of He, which promotes equilibration of He in near-surface soil gas with the atmosphere. Thus, Butt and Gole (1985) found that He concentrations increased with depth and that samples from less than 3 m had lost most, if not all, of any excess He. Similarly Hinkle (1994) found no significant variations in samples collected at various depths in the
Helium
31 1
top 2 m. Butt and Gole (1985) chose to collect samples by drilling to a predetermined depth (usually 6 m), placing clear vinyl tubing (3 mm i.d.) fitted with nylon taps at the top into the holes, and backfilling the holes to the surface. The sampling tubes were purged by withdrawing 50 ml of gas with a syringe and left to equilibrate for a minimum of four hours before sampling, again with a syringe. Deep sampling was also recommended by Jones and Drozd (1983), who collected soil gases from holes drilled to 4 m. There are, however, obviously many circumstances in which deep sampling is not possible, such as areas of shallow water table. Whilst it may be argued that shallow sampling must then suffice (Gregory and Durrance, 1985), it may be that in such circumstances gas sampling is not an appropriate procedure. Samples may be retained in syringes if analysed within 24 hours, apparently without loss of He. For storage over longer periods, glass or metal containers should be used. Metal cylinders, fitted with septa for injection and sealed by brass nuts with lead washers, can retain He for several weeks. Prior to use, the cylinders are evacuated and then filled with sample to about two atmospheres pressure. Possible errors due to preferential gain of He when evacuated, or loss when filled, can be detected by duplicate sampling.
Soils
Soil samples, collected from depths of 5 cm to 1 m, have also been utilised in He surveys (Pogorski and Quirt, 1979; Hinkle and Dilbert, 1984). These authors consider that they can thereby effectively sample gas trapped in soil micropores by films of water. Such gas supposedly is less prone to equilibration with the atmosphere than that in the macropores (which is the gas sampled by hammered probes). Soil samples are placed in containers, which are sealed and left to equilibrate for days or weeks. The headspace gas is then analysed for He and the concentration in micropore gases calculated. Such a procedure has been described by Hinkle and Kilbum (1979) and Hinkle and Dilbert (1984), but may be subject to considerable error (Butt et al., 1985). Firstly, during equilibration in the sample container, biological activity changes the major gas composition of the headspace, having residual effects on the He concentration and affecting the performance of constant-pressure inlets (see below). Secondly, the methods by which pore-gas volumes are calculated are unproven and inaccurate. A similar procedure, using much deeper (15 to 25 m) overburden samples, was employed by Van den Boom (1987) to investigate He distribution over an oil field in Germany. However, the calculation for He content does not account for the possible effects of changes in bulk gas composition, although numerous other factors are considered. Mud and permafrost have been used as samples in hydrocarbon exploration, with the assumption that He is retained in the water or ice (Roberts, 1981).
312
C.R.M. Butt, M.J. Gole and W. Dyck
Waters A variety of sampling techniques has been used in He surveys of surface waters (oceans, lakes, springs) and groundwaters. The most important features of such devices are that mixing and turbulence with air are minimised and that samples can be collected from specific depths. Sampling of groundwaters requires pumps or samplers that can operate within the confines of a drill hole; Dyck et al. (1976c) and Butt and Gole (1984) illustrate two such samplers, from which water is removed immediately the sampler is brought to the surface. Pumps are effective in open waters (Stephenson et al., 1992) and in cased and slotted drillholes if degassing is minimised. For the more rigorous requirements of isotope analysis, soft copper tubes crimped with strong clamps are recommended (Clarke and Kugler, 1973). Many studies have shown that, when waters are enriched in He, the He content varies with sampling depth below the water table or water surface (e.g., Eremeev et al., 1973, Dyck, 1976; Dyck and Jonasson, 1977; Butt and Gole, 1984, 1986). The general trend is for the He concentration to decline upwards due to degassing in attaining dynamic equilibration with the atmosphere. The effect in uncased drill holes or wells is greatest in the top 5-10 m of the water column, but is evident to 100 m or more below the water table. In cased drill holes, where the He enters the hole from a specific opening (hole bottom or screen), the He concentration gradient is steepest near the opening. The degassing gradient at a given site appears to vary with time and, in a given area, from site to site, so that no satisfactory correction or normalising factor can be developed. However, as total He concentration variations in the top 50 m or more of the water column are commonly less than 10, whereas regional variations are 10z-105, samples collected from a constant depth below the water table (b.w.t.), selecting the deepest level consistent with an appropriate sample density, are adequate for survey purposes. In surface waters, such as lakes, equilibration with the atmosphere occurs to considerable depths due to mixing by wave action and convective circulation. In Canada, Torgerson and Clarke (1978) found that anomalous He concentrations were retained only close to or below the thermocline (130-150 m). In shallow lakes, in which the thermocline is destroyed in summer, such sampling is only possible in the winter (Dyck and Tan, 1978), so that lake sediment samples have been employed as an alternative (Dyck and Da Silva, 1981). Gregory and Durrance (1987) used stream and spring sampling for a He and Rn survey in southwest England, but variable degassing predictably gave interpretational problems. If water samples are to be analysed within, say, a month after collection, then storage in thick-walled glass bottles (such as the types used for soft-drink or beer) is effective and inexpensive. The bottles can be sealed with metal caps (Dyck et al., 1976c) or rubber bungs, and stored upside down so that any headspace gas or air subsequently degassed will be against the glass and not the stopper (Butt and Gole, 1984). Loss of He by diffusion from glass bottles is usually minimal but losses of up to 15% per month have been reported
Helium
313
depending on the type of glass or other factors (Dyck, 1976; Heaton and Vogel, 1979; Datta et al., 1980; Butt and Gole, 1984). Samples are analysed either by degassing the water under vacuum in the inlet system of the mass spectrometer (Clark and Kugler, 1973; Dyck et al., 1976c) or, more easily, by equilibration with a known volume of atmospheric air, analysing this for He and calculating the He content of the water (Dyck et al., 1976c; Reimer and Otton, 1976; Butt and Gole, 1986; see below). As an alternative, Dyck and Da Silva (1981) equilibrated air in ping-pong balls and latex tubing with lake sediments and analysed the equilibrated air.
ANALYSIS
Mass spectrometry Helium analysis is most conveniently carried out by mass spectrometry and even very simple instruments are capable of sensitivities better than 0.01 ppm at concentrations of 56 ppm 4He, i.e., those commonly found in soil and overburden gases and waters. A schematic diagram of apparatus suitable for He analysis is shown in Fig. 10-1. It consists of an inlet system, a pre-concentration trap, a mass spectrometer with associated vacuum system and a read-out. The gas inlet system, designed for rapid yet reproducible analysis, may be of either a constant-pressure (CP) or a constant-volume (CV) type (Hart et al., 1985). Gas from the inlet system is pre-concentrated by passing it through a cold trap containing either activated carbon or molecular sieve, cooled by liquid nitrogen. These materials pump most gases except H2, He and Ne (Dennis and Heppell, 1968) and hence strongly pre-concentrate the gas entering the spectrometer. A trap consisting of a U-tube of 12 mm o.d. copper pipe containing about 25 g of 8-12 mesh beads of 5A molecular sieve can adsorb over one litre of air (Butt and Gole, 1984). It was found to be effective at flow rates of up to 15 ml min 1 (STP) without deterioration of the vacuum in the mass spectrometer. The sieve can be regenerated overnight, or as required, by allowing it to warm to ambient temperature while pumping with a rotary backing pump. Full regeneration is achieved by baking at 150~ Magnetic sector mass spectrometers are used for He analyses, since they are inherently simpler in design than quadrupole instruments and, importantly, give a response in the form of flat-topped peaks that permit easier and more accurate peak-height measurements. Many studies have utilised modified vacuum leak-detectors (Friedman and Denton, 1975; Reimer, 1976; Dyck and Pelchat, 1977; Martin and Berquist, 1977). These are inexpensive, relatively robust and can be adapted for use in the field. They are small-radius instruments with permanent magnets, normally focused on mass 4. However, it is essential that such instruments have some scanning capability, achieved by varying the ion-acceleration potential, at least to optimise the focus and preferably to resolve adjacent mass peaks. Resolving power should be sufficient to preclude contributions from mass 3 derived from the dissociation of water (Zaikowski and Roberts, 1981).
C.R.M. Butt, M.J. Gole and W. Dyck
314 A. ANALYTICALSYSTEM J Ou
Variable leak
jr
,v.:
o~
~lUmJ tern I
i0 ; il:)~'U/; n~train JJ.__..~__.~
B. CONSTANTVOLUME INLET Inlet
C. CONSTANTPRESSUREINLE1~
Maln valve
(
Luer-hub ~ inlet port
J #4 porous '~ glass frlt Glass syrlnge
].
Mercury reservoir
J Graduated
Luer-hub inlef port
way valve
Fig. 10-1. Schematic diagram of analytical apparatus used for mass spectrometric determinations of helium (from Hart et al., 1985).
Mass spectrometers with electromagnets can scan a wide mass range and hence, with suitable inlet systems, can analyse for a number of gases. They are thus more versatile than modified leak detectors, but are more costly and unsuited for field use. With a CP-inlet, the instrumental response is proportional to the He concentration and may be recorded as a plateau height on a chart recorder, whereas with a CV-inlet, the response is integrated over the period of sample passage. Generally, analyses of unknowns are bracketed between those of standards (normally water-saturated air); eight to ten samples can be analysed per hour. Analytical precisions of 0.01-0.02 ppm in total concentrations in the range 5.0-6.0 ppm He can be achieved (Reimer et al., 1979b; Butt and Gole, 1985). Instrumental response is linear and better precisions can be obtained at the higher concentrations (102-103 ppm He) found in natural gases and headspace gas over groundwaters, provided suitable standards are used. Dilution may be necessary if the pumping capacity of the vacuum system or other instrumental characteristics are exceeded.
Helium
315
Despite such precision, however, the accuracy of some published data is uncertain. Hart et al. (1985) have noted that with a CP-inlet, the concentration is in effect a measure of the flow rate of He through the instrument. This is directly related to concentration if the pressure and major gas compositions of the samples and unknowns are the same. However, at the same pressure, different gases have different flow rates through the inlet system, relative rates being: 02 < dry air < water-saturated air < N2 < CO2 << CH4. Consequently, the net flow rate of a sample, and hence the apparent He content, will vary according to the major gas composition. Such compositions can vary quite markedly in soil gases, due to biological activity, and hence the resultant He determinations can be in error. For example, a water-saturated soil gas containing 10% biogenic CO2 and only 10% 02 will have an apparent He content of 5.33 ppm relative to 5.24 ppm in a dry atmospheric air standard; such a value would be considered significantly anomalous by many workers. Any He survey using soil gases or gas extracted from soils is subject to such errors if the analysis has been by an instrument fitted with a CP-inlet. Unless another inlet has been used or the possibility of error has been recognised and discounted, the results must be treated with caution. Nevertheless, Duddridge et al. (1991) and Hinkle (1994) have ignored the problems with CP-inlets and treat variations within analytical error as significant, so that some of the results of these studies must be questioned. For example, data manipulation by Duddridge et al. (1991) makes the plausible assumption that "He, Rn and CO2 (are) ostensibly independent gases", but this is incorrect given the effects of CO2 absorption in CP-inlets. Similarly, in the important study of environmental effects on soil gas concentrations presented by Hinkle (1994), the value of the He data is diminished since the environmental and analytical effects are not separable. The data do give an estimate of the net variation but it is not possible, for example, to determine whether similarities in behaviour of He and CO2 are real or are artifacts of the analytical method. If relative values are suitable, then He analyses via a CP-inlet will be satisfactory provided they are normalised with respect to 2~ determined by using exactly the same inlet and instrumentation. This will not be possible, however, with simple leak detectors having a restricted mass range. For normalisation using gases that can only be analysed by different inlets or instruments (e.g., 36Ar), absolute He contents must be determined by using a CV-inlet. This procedure was used by Corazza et al. (1993) to determine He enrichments derived from geothermal sources in soil gases affected by atmospheric contamination, dilution by CO2 and, in addition, effects related to the use of CP-inlets.
Gas chromatography Gas chromatographic procedures have been developed that permit the determination of He, H2, N2, Ar and CH4 (Sugisaki et al., 1982). The analysis is slow (about one sample per hour) and the sensitivity of about 1 ppm He at 5 ppm total He concentration is inadequate for soil gases and some waters, but the procedure is useful for studies in earthquake prediction and of thermal waters.
316
C.R.M. Butt, M.J. Gole and W. Dyck
Portable helium analysers A relatively simple portable battery-operated ion pump with limited resolving power and sensitivity, which is suitable for the detection of 50 ppm or more He, has been described by Eremeev et al. (1973) and Yanitskii (1979). This instrument is capable of analysing up to 30 samples per hour but lacks the sensitivity required for near-surface samples. A portable He analyser capable of measuring 1% or more has been described by Sonnek et al. (1965). It performs pre-separation of He from other gases in the sample by means of a chromatographic column, with identification of the He constituent by the response of a thermistor at the end of the column.
Determination of helium isotope ratios Since the abundance of 3He is about 10 -6 lower than that of 4He, the determination of 3He/nile ratios is only possible with mass spectrometers having high sensitivity and high resolution. Although ratio determinations are important in some groundwater studies, including earthquake prediction, there is no evidence that they have any advantage over total 4He analysis in exploration. Consequently, the complex instruments and time-consuming procedures necessary for such determinations (which also make them very costly) are not discussed here. Descriptions of the procedures are given by Mamyrin et al. (1969) and Kugler and Clarke (1972).
Analysis of waters Analysis of He in waters relies on equilibrating the water-containing sample with a known volume of atmospheric air (or other gas) which is then analysed. This is most easily achieved by analysing the headspace gas in partly-filled water sample containers. After analysis, the volumes of the headspace gas and water must be measured. If atmospheric air is assumed to contain 5.240 ~tL/L He and fresh water in equilibrium at STP with atmospheric air contains 0.0441 ~tL/L He (Weiss, 1971), then: CA = 1 18.8 Cw
(10.1)
C A -- measured concentration of He in headspace gas, Cw = concentration of He in water in equilibrium with air sample. Since total He is the same before and after extraction,
where,
CAVA + CwVw = CwoVw + CAoVA
(10.2)
Helium
317
concentration of He in headspace gas before extraction, Cwo = concentration of He in water before extraction, VA = volume of headspace gas, Vw = volume of water. Substituting equation (10.1) into (10.2)" where, CAO -'-
CWO = [CA(V Aqt-vW/118.8) - CAoVA] / Vw
(10.3)
The presence of salts in water alters the solubility of gases, such that equation 10.3 becomes: Cwo = {[CA(VA+Vw/1 18.8)- CAoVA] / Vw} + [3 * CA
(10.4)
where 13 is the Bunsen solubility coefficient in salt solution. The effect is generally small and the 13 value for pure water (0.00870) can be used. For high salt concentrations (>1% total dissolved solids) a correction should be applied using the relationship" ln[13o]/[3 = MK
(10.5)
where, [3o = Bunsen solubility coefficient in pure water (0.00870), M = salt concentration, K = salting coefficient for NaC1 solution, 0.2303 K(mole-~). The analytical sensitivity needed for He analyses of groundwater for He is generally far less than that required for soil gases. Absolute concentrations in dissolved air or headspace gas are generally higher than those in soil gas and, more importantly, the range of groundwater He concentrations found within a given survey area will generally be far greater (up to 104) than that in soil gas (up to 5 x 10-2). Errors arising from use of CP-inlet systems are thus generally not significant. However, the calculation is such that if minor variations are sought, as, for example, in surface waters, normalisation or the use of CV-inlets are essential.
VARIATIONS OF HELIUM CONCENTRATIONS The concentrations of He and other gaseous components of soil and overburden gases and waters varies over periods of hours, days or months. This variation has importance in interpreting survey data and, by recognising significance in the variations themselves, in earthquake prediction.
Soil and overburden gases In tests for diumal variations at fixed sample sites at 60 cm depth, Reimer et al. (1976) noted maximum variations of 0.03 ppm He about a mean of 5.24 ppm. In deeper samples (6 m), a range of 5.38-5.43 ppm was noted at Manyingee, Westem Australia (Butt and
318
C.R.M. Butt, M.J. Gole and W. Dyck
Gole, 1985), and 5.24-5.27 ppm and 5.53-5.95 ppm in background and anomalous sites respectively at Gingin, Westem Australia (Gole and Butt, 1985). The variations tend to be site specific and are probably related to changes in biological activity and moisture content, though barometric pumping, wind speed and climatic factors may be implicated. In a detailed study of variations in the Re, Rn and Hg flux at a single site over a 22month period, Klusman and Jaacks (1987) found the total range to be 5.20 to 5.30 ppm He, 460 to 22000 counts/day Rn and <0.1-3.57 ng/L Hg. The periodicity of variations for He and Rn were quite similar, with both being higher in winter. Most of the variations for Rn and Hg could be explained in terms of changes in meteorological conditions, particularly temperature and moisture; for He, however, these explained less than 33% of the variation, the remainder being "noise". It is hypothesised that the flux of Rn and He from deep sources is influenced by the development of an inversion layer below the sampling depth (3 m), which slows the rate of convection in the summer. Similarly, a minor difference of 0.02 ppm in the mean He content of samples collected five months apart over the Aurora uranium deposit, Oregon, USA, was ascribed to a change in moisture content in the period between surveys (Reimer, 1986). However, in both examples, CP-inlets were used and concomitant changes in other components may be responsible for much of the variation. Detailed monitoring of eight unmineralised sites over 10-14 months by Hinkle (1994) showed that there were both seasonal and diurnal changes in gas compositions at sampling depths between 0.3-2.0 m, related to variations in precipitation, temperature and barometric pressure. Considerable variations in He abundances were recorded, the greatest being from 4.111-8.545 ppm at 0.6 m depth at Golden, Colorado, and the least being from 4.836 to 5.872 ppm at 1.2 m at Arvada, Colorado (samples both above and below this depth showed greater variations). These are net abundances since, again, analysis using CP-inlets means that analytical and environmental effects cannot be separated. Such variations are far greater than those recorded in most surveys and in other monitoring studies. The reason is unclear but may be due, in part, to the use of fixed (rather than hammered) probes, reducing atmospheric contamination. Although there are cases, such as Gingin, where a real anomaly is not masked by periodic variation, the data reported by Hinkle (1994) clearly demonstrate why re-surveys tend to give quite different patterns as a result of periodic change. The ranges of diurnal variation were not reported but are stated to be much less, so that surveys completed in a short period with minimal meteorological change should be intemally-consistent with respect to these factors. Nevertheless, the magnitude of even short-term variations, coupled with background variations between close adjacent sites, implies that the results of single surveys must be viewed with extreme caution, notwithstanding other problems associated with procedures such as shallow sampling by hammered probes. Because variations in the flux tend to be site specific in both magnitude and sign, use of a monitoring site is unlikely to be an effective control. However, normalisation with either 2~ or 3~Ar eliminates many environmental and analytical effects and is essential if real variations of He flux are to be determined (see below).
Helium
319
Groundwaters Helium concentrations of groundwaters sampled from drill holes and wells generally remain fairly constant with time. Russian workers (e.g., Tikhomirov and Tikhomirova, 1971; Golubev et al., 1974) have shown that at depths below 30 to 50 m, He concentrations are relatively stable over time. At Koongarra, Northern Territory, Australia, Gole et al. (1986) conducted two surveys seven months apart in contrasting seasons (end of dry season and end of wet season) using samples from 10 m b.w.t. Very similar results were obtained, in terms of both abundances and distribution patterns, which gives confidence in the survey techniques used. At Mt Weld, Western Australia, there are variations in dissolved He contents, even over a few hours (Fig. 10-2), the long and short term fluctuations observed in samples from 5-10 m b.w.t, are rarely greater than two-fold (Butt and Gole, 1986). Greater variations (up to four-fold), however, occur in wells containing less than 2 m of water (Butt and Gole, 1984). The cause of these variations is not explained, but they do not seem to arise from sampling or analytical errors, nor from changes in pressure, temperature or wind speed. Since variations in He content do occur, there are implications for the interpretation of survey data. The changes that may be observed for individual sites (two- to four-fold) are such that subtle interpretations should not be placed on the data. However, the ranges in abundance observed in most surveys (two to four orders of magnitude) are so great that such variations can be tolerated. Most data points will remain within the same interval and distribution patterns will be similar, despite this uncertainty.
Biological activity and soil gas composition The interaction between soil gas and biological activity is both highly complex and very significant for the interpretation of soil-gas geochemical data. The major gas composition of soil gas varies markedly both spatially and temporally as a result of this interaction. There is extensive literature discussing compositional changes in the soil atmosphere due to the effect of soil type, degree of aeration, drainage status, soil biology and seasonal and diurnal variations on mineralogical and biological reactions unrelated to the presence of mineralisation (De Jong and Paul, 1979, and contained references). These influences have rarely been considered in studies of gas geochemistry in exploration but, unless they can be evaluated, they place severe constraints on the use of these methods. The interaction between biological activity soil gas may also affect inert gases such as He, even though they are not involved in the reactions. For example, if 02 is consumed and not replaced with an equal volume of CO2 (due to solution of CO2 in soil moisture, or the formation of carbohydrates) then the concentration of the residual gases will be enhanced. A 2% enhancement will increase the He concentration from a nominal 5.24 ppm to 5.34 ppm, a level considered to have exploration significance by some workers.
320 b
C.R.M. Butt, M.J. Gole and W. Dyck
9~9
E
970 o.
Day 1
I
Day 2
I
Day 3
I
I
Day 4
P9 5m
r
..........
\ tV
Lilly Pond W. R C 4 5 5m
\
8
.F. .a _,1
I
Day 5
I %1
V A , u,/: 2
6
Om
i
4
,,,...,:'...
Day 1
~-:..'~..':.. i
Day 2
I
"...,'-..
Day 3
_.,~ 5m
=,.o,,m,~'e'o.q
_J i
Day 4
I
Day 5
Fig. 10-2. Variations with time in atmospheric pressure and in the helium content of groundwater from exploration drill holes over the Mt. Weld carbonatite, Western Australia (P9 and RC45) and an adjacent stock well (Lilly Pond Well); depths shown are below water table (from Butt and Gole, 1984).
In order to determine whether residual enhancement has occurred, He must be normalised, most usefully to the non-radiogenic atmospheric components 2~ or 3~'Ar; residual enhancement will change the absolute contents of He, Ne and Ar but not the ratio of their concentrations. Gole and Butt (1985) document several sample sites at which soilgas He concentrations are residually enhanced, the maximum being 6.4 ppm He which is normalised to 5.25 ppm when ratioed with Ne. This particular He enhancement occurred over a natural-gas microseep, but presumably residual enhancement of He occurs in many other environments which do not have any exploration significance.
HELIUM SURVEYS IN MINERAL EXPLORATION
Rationale Helium, like 222Rn(see Chapter 11), should be an ideal pathfinder for U mineralisation. Both are daughter products of U and hence must be present; and being gases, they have different geochemical mobilities to U. Calculations show that 1 g of U and its decay products in equilibrium produce 1.17 x 10.7 ml (STP) He. In a U deposit having a grade of 1%, a porosity of 20% and a bulk density of 2.2, the He concentration of the pore space
Helium
3 21
increases at a rate of 0.013 ppm per annum, assuming a closed system (Pacer and Czarnecki, 1980). Any permeability or leakage will contribute to a He flux. The potential of the technique lies in the possibility that this contribution will result in a detectable increase in the He content of waters and gases that can be spatially related to the mineralisation. Furthermore, because of its radioactive stability, He may, over time, become highly concentrated in suitable traps or migrate considerable distances. These are potential advantages over 222Rn which, because of its short half-life (3.825 days), cannot become highly concentrated nor migrate far. Tanner (1964a, reported by Smith et al., 1976) has calculated that under optimum conditions, 222Rn may migrate 160 cm by diffusion and, assuming a groundwater flow rate of 1-5 m per day, 60-70 m by mass flow transport. At these distances its concentration is 0.1% of that at the source. Greater distances are quoted in the literature (e.g., 120 m by Gingrich and Fisher, 1976) but it is not established whether this is due entirely to migration by Rn or, in part at least, to migration by Ra. Under similar hydrological conditions, He concentrations only decrease by dilution by He-free water or by losses by diffusion to adjacent aquifers. Helium would be expected, therefore, to yield a more widespread and persistent anomaly than Rn. Soil and overburden gases are the most accessible of possible sample media for He surveys and their use has been widely tested. Because of the high diffusibility of He, however, sampling of the atmosphere itself has rarely been attempted. With few exceptions, most blind U deposits and other potential sources of He are situated below the water table, so that He must be released into the groundwater prior to its escape to the soil gas and the atmosphere. In consequence, groundwater, or at least groundwater as represented by water in drill holes and wells, has been sampled extensively as a means of determining the dispersion of He. Surface waters, however, tend to be equilibrated with the atmosphere and are unsuitable as sample media, except when collected from springs and at depth from lakes.
Helium in gases and soils Soil and overburden gas surveys
Most published data on He in soil gases are based on shallow sampling by a hammered probe. As discussed above, such samples are close to equilibrium with the atmosphere and perhaps contaminated (diluted) during sampling. Certainly, He concentrations are uniformly low, with maximum concentrations mostly in the range 5.27-5.40 ppm, with few higher values. Nevertheless, some workers place significance on such values as indicating proximity to mineralisation. One of the most convincing examples is from a roll-type uranium deposit at Weld County, Colorado, USA, (Reimer et al., 1979b), where low grade mineralisation occurs at a depth of 70-80 m in a gently dipping sequence of mudstones, sandstones and conglomerates. The water table is at 3-12 m and groundwater flow is to the
322
C.R.M. Butt, M.J. Gole and W. Dyck
; :- ;
5.28
.........~ii~!ii.i
9 .... 9
I
Fig. 10-3. Contour map of averaged soil-gas (probe) helium concentrations over a roll-type uranium deposit, Weld County, Colorado (from Reimer et al., 1979b).
south-southeast. Soil-gas samples were taken at grid intervals of 320 m and at a depth of 0.60 m over an area of 8 km 2. A smoothed contour map (Fig. 10-3) shows a subtle anomaly peaking at 5.29 ppm He about 1.5 km south of the mineralisation. This displacement is interpreted as being due to dispersion by groundwater movement. Other surveys have tended to show only small areas with elevated He concentrations. Thus, over the Lamprecht ore body in Texas, USA, where mineralisation is at 70-90 m beneath clays and cemented sandstones, soil gas He contents have a total range of 5.205.34 ppm He (Reimer et al., 1979a). The few values over 5.27 ppm He occur over part of one mineralised zone and sporadically elsewhere, but another zone has background to negative values. In many instances, data typically show little or no coherent pattern and may vary with time. This is illustrated by some well-documented studies by Pogorski and Quirt (1979). In two surveys, conducted six months apart in the Red Desert, Wyoming, USA, the concentration ranges and distribution patterns differ (Fig. 10-4). Mineralisation is in flat-lying sediments, overlain by 125-140 m of barren rocks. Extremes of the concentration ranges were found at adjacent sites, less than 50 m apart, during each survey, and at individual sites when the surveys are compared. Higher values tend to be clustered close to the mineralised zone, but this is partly due to a greater sample density. Other high values were interpreted as possibly being due to deeper but unproven ore bodies. When the results of these and other surveys are considered, it is apparent that the observed concentration ranges may be due to background variations plus the sampling and analytical errors discussed previously. Pogorski and Quirt (1979) suggest that the errors
323
Helium A
+,
9 x
9 9
It9
A
9
9
I
i
9
" ~ ' ~~+ ' + ' ~ l e l ~
O*
9 "
I
It
113'
x
1
I
9
9
9
I
9
9
x
0..:.~ ; .'.-.
!
9 < 523
Grsde thickness ~0.25%feet
200 m i
9 It
x
It
9
.
9
x
9
9 x
I
It
I
It
9
x
*
9
9
It
9
It
x
*
I
9
It
9
9
thickness 0 , 2 5 %feel
200
9
x 9
x
It
9
9
9
II
st
9
It
m o
9
it
9
It
It
9
it
1 x
9 It 9
Fig. 10-4. Uncorrected helium data from soil-gas (probe) survey at the Red Desert research site: (A) collected during summer; and (B) collected during winter (from Pogorski and Quirt, 1979).
alone may be 0.05 ppm. The data of Gregory and Durrance (1985) illustrate these effects, with a total range of 5.19-5.29 ppm He being observed in 45 samples from an area of only 90 m 2. Most reported surveys tend to be of very restricted areas and hence give no indication of these background variations. However, in a semi-regional survey of 200 km 2 in Weld County, including the area illustrated in Fig. 10-3, it was noted that the range of He concentrations was greater than on the localised grid around the U deposit, and that there was no obvious pattern to the distribution (Reimer et al., 1979b). This was interpreted as suggesting that the sample density (1 per 2 km 2) was too great for the target size. However, this interpretation also implies that these higher values have significance in terms of a He source, whereas it would probably be more correct to consider them as part of the background population. The significance of the original anomaly described above then becomes harder to assess. Similarly, there is probably little significance in the difference of 0.02 ppm He noted for soil gases over two lithological units at the Aurora uranium deposit, Oregon, USA (Reimer, 1986). Although a higher He flux over the prospective units is a possible explanation, so too are background variations related to the influence of soil type on interaction with the atmosphere, major gas composition and analytical performance. The highest He concentrations reported for probe sampling were found at the Ambrosia Lake district, New Mexico, USA, in a restricted survey of 46 samples from a 1.2 x 3.2 km area over mineralisation 300-400 m deep (Reimer et al., 1979b). The mean soil-gas concentration was 5.30 ppm He, with three samples containing 5.47-5.65 ppm He. The distribution pattern has some coherence but is not readily related to the location of the mines and has probably been affected by the mining operations.
C.R.M. Butt, MoJ. Gole and Hr. Dyck
324 3600N I]I~BD 9 [ ~
|
55
Li .
2850 N. .
r "~"
54
.
A
.
.
t,,
57 F Line 600 W
I~
54
t / I !\1! / ~
/ll
~"
tI
| 200
m.~
56
1000wl~
.
iT ....
0'
"~
, 0
J 200E
Location of blind mineralization
E 5.31 5.34 5.35
December 1980 April 1981 September 1981
"it
J2OOE 2850N
/~
5.2 I I metres 1000W 800 600 400 ,,., .........................................................
55
.
Line
600W
~,~
s 0.05 0.11 0.05
,
5,3 metres
~~3600N
34 o 0 mr
3 ;o 0
3000 . . 2800 . .......................
. 2600.
2400
2200N
Fig. 10-5. Helium concentrations in overburden gas collected at 6 m depth, Angela Deposit, Northern Territory (from Butt and Gole, 1985).
Because of the perceived problems with shallow probe sampling, Butt and Gole (1985) conducted a number of surveys over blind mineralisation in Australia using buried, fixed sampling tubes. It was found that in samples from less than 3 m depth, He contents were low and close to the atmospheric background, but that deeper samples, usually from 6 m, gave higher and presumably more reliable He values. Traverses over the Angela deposit, Northern Territory, gave the most promising results. Mineralisation is in Palaeozoic sedimentary rocks at depths of 25-110 m and the water table is at about 60 m. Samples were collected on several occasions from sites 50 m apart on two intersecting traverses (Fig. 10-5) and although they showed considerable variation in the distribution and abundance of He, higher values (5.30-5.64 ppm He) were generally found over mineralised areas. However, not all samples from such areas had high He contents, nor were sites consistently high on re-sampling; conversely, high values were also found in supposed background areas. Whilst it appeared probable that a detectable He anomaly was present, this was restricted in extent, variable in intensity and modified by local conditions (e.g., the anomaly was reduced or eliminated where there was a shallow, perched water table). Similar variations were shown by a later survey in which the mean He concentration over mineralisation (5.44 ppm) was greater than that over the background area (5.37 ppm; Table 10-II). However, normalisation to constant Ne, to eliminate the effects of variations in bulk
325
Helium _
5"5 1 " I I
Radon
Im probe
May t983
Radon
6m hole,
December 1983
6m holes
December 1983
Helium ~
1765
ibtlOlUte caNncentrlltlon
Plan. Down hole raOlometrlc anomaly (cp~) 1ooo 13
14
16
10
2 3 2 5
Fig. 10-6. Concentrations in overburden gas of helium (at 6 m depth) and radon (1 m and 6 m depth), Mulga Rock, Officer Basin, Western Australia (from Butt and Gole, 1985).
composition, also eliminated this difference, suggesting that the apparent He anomaly was largely fortuitous. A similar study was made at Mulga Rock, Western Australia, where mineralisation is at the water table at a depth of 30 m, and overlain by Tertiary clays and sandstones. Sample tubes were set at 5-8 m beneath aeolian sands, at 50-100 m intervals on a traverse across the mineralised zone. Samples were collected on several occasions but no He anomaly was detected. Typical results (Fig. 10-6) showed there is no significant difference between mineralised (mean 5.27 ppm He) and unmineralised (mean 5.26 ppm He) areas. A few higher values (5.29-5.31 ppm) are found over part of the mineralised zone but these cannot be unequivocally considered to be indicating mineralisation. Normalisation to constant Ne reduced the He concentrations (Table 10-II) and although higher values at the western end of the traverse again suggested an increased flux over part of the mineralised zone, they are still within the range of background variation plus sampling and analytical error. In contrast, Rn determinations of gases from these sites showed a very marked anomaly, which was duplicated, with much reduced contrast and intensity, in probe samples from 1 m depth. Butt and Gole (1985) also found no He anomaly over shallow (15-20 m) zircon- and monazite-rich mineral sands at Eneabba, Western Australia. These contain 300-500 ppm U + Th and thus represent a moderate source of He. However, the He contents of overburden gases were the same over the mineral sands (5.24-5.31 ppm) and a background site 5 km distant (5.21-5.30 ppm). Any He must either be retained within the minerals or lost to the
326
C.R.M. Butt, M.J. Gole and W. Dyck
atmosphere without accumulating, due to the porous nature of the overburden. The data do, however, give an indication of background variations over relatively small areas that have essentially unchanging aspect, slope, overburden and drainage. Even where mineralisation is below the water table and groundwaters are enriched in He, gas anomalies may not be present. Thus, over the Mount Weld carbonatite, Western Australia, where groundwaters contain up to 13.5 ~tL/L He at 5 m b.w.t., overburden gases contain less He than over country rocks with groundwaters having only 0.04-0.10 ~tL/L He (Table 10-II). No anomaly was found at Manyingee, Western Australia, although the highest concentration (5.48 ppm He, and 5.43 ppm after Ne-normalisation) was from a supposedly background site where groundwaters were highly anomalous (80 ~tL/L).
Soil surveys
Pogorski and Quirt (1979) and Pogorski and Pogorski (1982) document studies at three sites in which He has been extracted from soil, correcting their analytical data by an unspecified procedure. They claimed that the results indicate the presence of mmeralisation and that soil sampling is preferable to direct sampling of soil gas. The most detailed study was at the Red Desert site, Wyoming, USA, where in the initial survey, corrected He contents of 5.56-9.14 ppm appeared to indicate the U mmeralisation. In a subsequent sub-regional survey, however, two samples over mineralisation that were anomalous (6.95 and 9.00 ppm He) relative to the local background (5.20-6.37 ppm) proved non-significant relative to the regional background (mean = 6.76 ppm He), which increases strongly to the south. About 25% of the surveyed area had He contents exceeding 7.82 ppm, with many samples having 10-12 ppm, none of which related to U mineralisation. The location of the deposit itself can be revealed by contouring, or by extracting residuals after trend-surface analysis, but it is not a strong feature and is overshadowed by other variations in the data. Hinkle and Dilbert (1984) used the minus 30-mesh (<600 mm) fraction of soils from only 0.8 cm depth for He surveys. After equilibrating the samples in sealed containers for up to five weeks, the headspace gases were analysed for He and other gases. At the North Silver Bell porphyry copper deposit, Arizona, USA, they claimed that pore gases contained the remarkably high concentrations of 7.84-23.84 ppm He and related the distribution to the occurrence of faults and fractures. However, Butt et al. (1985) have questioned whether the assumption that He is retained during sieving can be correct and have pointed out that, during equilibration of samples in sealed containers, biological activity causes gross changes in major gas composition, affecting both the residual concentration of He and analytical performance; also, assumptions and methods of measuring pore volumes are erroneous. By duplicating the procedures, using a desert soil that had been ground to <150-mesh (<105 mm) and stored dry for seven years, Butt et al. (1985) obtained apparent He contents of over 15 ppm for "pore gas"; such results demonstrate the inadequacies of the methods. Hinkle and Dilbert (1987) subsequently reported a negative correlation between He content and both moisture and CO2 contents of soils from North Silver Bell
327
Helium
TABLE 10-II Range and (in italics) mean of total helium concentrations (He) and neon-normalised helium concentrations (HEN) in overburden gas over mineralised zones and background areas in Australia (from Butt and Gole 1984, 1985).
No. Angela (uranium)
20
Mulga Rock (uranium)
20
Eneabba (mineral sand)
16
Mt. Weld (carbonatite)
14
Manyingee (uranium)
22
Mineralised zone He HeN No. (ppm) (ppm) 5.26-5.77 5.18-5.36 25 5.44 5.28 5.25-5.31 5.22-5.30 10 5.27 5.25 5.24-5.31 15 5.28 5.25-5.31 6 5.32 5.35-5.44 5.29-5.34 21 5.39 5.32
Background area He HeN (ppm) (ppm) 5.24-5.72 5.14-5.35 5.37 5.28 5.24-5.28 5.23-5.26 5.26 5.24 5.21-5.30 5.27 5.24-5.47 5.33 5.35-5.48 5.29-5.43 5.39 5.33
and from Crandon, Wisconsin, USA. These data are dominated by very high He values determined from analyses of equilibrated headspace gas; as discussed above, these are subject to inaccuracies, including analytical enhancement due to biogenic C02. The soil C02 contents reported, however, are of gases derived by heating other sub-samples, not of the equilibrated headspace gas. Accordingly, the validity of the He data remains doubtful and any conclusions related either to relationships between moisture and CO2 contents or to the location of mineralisation are speculative.
Discussion o f soil-gas and soil survey techniques
The available data from He surveys over U mineralisation, summarised above and in other publications (Reimer et al., 1979b; Pogorski and Quirt, 1979; Butt and Gole, 1985; Butt et al., 1987) offer no convincing evidence that detectable anomalies are present in either soil gas or gas trapped in soil. Many surveys have been very localised, with insufficient samples collected to establish background variations, so that the often optimistic interpretations were misplaced. Although in some instances, patterns of marginally higher concentrations may indicate an increased flux due to mineralisation, such patterns and concentrations may also be due to background variations. These may result from changes in environmental conditions affecting, for example, sampling efficiency (i.e., the degree of atmospheric contamination) and bulk composition, with its concomitant effects on residual He concentration and analytical performance. Deeper sampling and normalisation using ZONe or 36Ar can eliminate such effects; but in the few
328
C.R.M. Butt, M.J. Gole and W. Dyck
instances when this has been done, many apparent anomalies have been eliminated, suggesting them to be spurious. Consequently, assigning significance to increased He contents may be quite misleading, as too is the assertion (e.g., by Raines et al., 1985) that concentrations exceeding 5.33 ppm He in shallow soil gases are due to anomalous U concentrations in underlying rocks or groundwaters. In particular, isolated high He values should not be interpreted as indicating a He-rich source, but rather to represent local environmental conditions (Roberts, 1981). Even if such false anomalies are present, the technique would be of value if it were certain that U mineralisation did yield He anomalies in overburden gases. Since this is not so, however carefully samples are collected and analysed, the technique must be considered to have very little application. This conclusion is supported by the mathematical modelling by Novikov and Kapkov (1965) and Jeter (1980), described above, which suggest that excess He contents in soil and overburden gas due to underlying, concealed U mineralisation would be far too low to be detected reliably. Soil sampling has been advocated because of the assumption that soil micropores either retain or concentrate He, thus essentially integrating the contributions of the flux emanating from mineralisation (Hinkle and Dilbert, 1984). However, apart from the considerable difficulties in measuring the pore concentrations, the mechanism by which He may accumulate is not explained. The high diffusibility of He would suggest that migration into and out of the pores in response to partial pressure differences would be so rapid that at most the He content of the micropore gas may represent a mean of the fluctuations shown by the inter-crumb voids in the soil. The high concentrations reported for micropore gas (6-24 ppm He) are not supported by independent measurements and suggest that false assumptions have been made in their calculation.
Helium in waters Groundwater surveys in uranium-mineralised areas
That groundwaters in contact with or immediately surrounding U deposits are indeed enriched in He has been documented in numerous studies (Table 10-III). These studies have used different sampling and analytical methods and the results have been reported using a variety of units. Thus the comparisons between the He data listed in Table 10-III, which has attempted to standardise the reported results, must be undertaken with some caution. Nevertheless, it is evident that groundwaters closely associated with U mineralisation are commonly markedly enriched in He compared to the atmospheric equilibrium value of 0.044 ~tL/L, the concentration at the time of groundwater recharge. Exceptions are Yeelirrie, Western Australia (<0.13 ~tL/L He), where mineralisation is very shallow and is mostly above the sample point (Butt and Gole, 1984, 1986) and at the Red Desert and Spokane Mountain sites described by Pogorski and Quirt (1979), where
329
Helium
groundwaters with <0.10 gL/L He were suspected to be subject to dilution by surface runoff.
TABLE 10-III Helium contents of groundwaters in mineralised locations and (in italics) in neighbouring unmineralised areas Country Australia
Location (reference) Bennett Well, W.A. (I) Manyingee, W.A. (1) Honeymoon, S.A. (1) Koongarra, N.T. (2) Yeelirrie, W.A. ( 1)
Olympic Dam, S.A. (1) Mt Gunson, S.A. (I) Mt Weld, W.A. (1)
Canada
USA
Malcolm, W.A. (1) Inda Lake, Labrador (3) Key Lake, Saskatchewan (4) Athabasca, Saskatchewan (5) Lamprecht, Texas (6)
Depth, m 10-20
Sample Type a
No. 6
5-15
d
35
0.08-78.9
14.27
15 56-76 50-60 10
a b b a
92 5 15 11
0.08-1.23 0.14-1.68 6.9-44.4 0.07-11.5
0.56 0.70 27.5 1.04
10 5 O. I-5 5
d a d, e f
29 15 58 51
0.05-0.16 0.00-0.13 O.00-4.11 0.05-110
0.9 0.5 O.45 13.26
50
a
11
910-2495
1590
0.5-10
d
128
0.03-450
6.99
2.5-5.0
a
33
0.18-13.6
2.13
0.5-5.0 0.5-5.0
d,e d.e
36 154
0.05-0.98 0.03-4.10
0.17 0.32
25-100 50, 90
b b
10 2
0.05-28.2 to 20.0
4.62
7.5 50-500
d c,d
2 28
0.3 7, O.43 0.05-23.6
90
c
4
0.63-1.72
70, 160
d
2
O. 71, 1.05
c
1
0.48
d d'
21 7
0.42-4.26 0.22-130
1.92 >41
0.17
Edgemont, South Dakota (7)
Red Desert, Wyoming (8) Copper Mt, Wyoming (8) Spokane Mt, Washington (8)
He (gL/L) Range Mean 9.92-29.5 15.55
19.0
60-113
c
14
0.07-0.91
94-109
d
7
0.08-0.62
0.31
17-125 18-108
c c
13 5
0.06-4.20 0.08-0.12
0.95 0.10
17-117
d
5
0.09-0.10
0.10
References: (1) Butt and Gole, 1986; (2) Gole et al., 1986; (3) Clarke and Kugler, 1973; (4) Dyck and Tan, 1978; Dyck et al., 1978; (5) Earl and Drever, 1983; (6) Reimer et al., 1979a; (7) Bowles et al., 1980; (8) Pogorski and Quirt, 1979. Sample types: (a) holes intersecting mineralisation; (b) mineralised interval; (c) probably mineralised interval; (d) samples equivalent to those associated with mineralisation (d', from structural zones); (e) stock wells and bores; (f) deep bores.
330
C.R.M. Butt, M.J. Gole and W. Dyck
Clarke and Kugler (1973) reported excesses of He in the one hole sampled near the Nordic mine at Elliott Lake, Ontario (1.56-8.69 ~L/L He), and in several holes in a potential ore body in Labrador (0.05-28 ~tL/L He). Similar results were reported by Dyck et al. (1977) from Key Lake, Saskatchewan, where up to 20 ~tL/L He occurred in deep holes (60-100 m) intersecting mineralisation. Both studies showed that, in many holes, He contents increased with depth, presumably due to upward degassing and equilibrium with the atmosphere. Groundwaters away from mineralisation at Key Lake contained less than 0.043 ~tL/L He, suggesting that high values in the U deposits were due to the mineralisation itself, but the data are too few to be conclusive. Limited groundwater surveys have also been conducted by researchers from the US Geological Survey. Rice and Reimer (1977) reported that waters from stock-bores had He contents 30-50 times equilibrium (i.e., about 1.3-2.2 ~tL/L He) down-drainage from a roll-front U deposit in Weld County, Colorado, USA; they also noted a variation in concentration according to the depth of samples. Similar concentrations (0.71 to 1.75 ~tL/L He) were reported from four holes into the Lamprecht ore body in Texas and four holes into the host formation 1 km and 4.5 km distant (Reimer et al., 1979a). Shallow holes not penetrating this formation had background He contents (0.05 pL/L He). Only the four holes into mineralisation had high Rn concentrations, so the He-enrichment of the distant holes implies that either the host formation is He-rich throughout, or the mineralisation, though nearby, is still sufficiently distant for Rn to have decayed. In a groundwater survey in the Edgemont district, South Dakota, USA, Bowles et al. (1980) report 0.48 ~tL/L He from one hole within a U deposit, although far higher concentrations than this were encountered elsewhere in the survey area (see below). Earle and Drever (1983) found He concentrations of 0.05-236 pL/L in groundwaters from various depths (up to 500 m) in the Athabasca Basin, Saskatchewan. They claimed there to be a moderate correlation between the occurrence of He anomalies and known U mineralisation, provided the He concentrations were divided by sample salinity in a rough correction for groundwater residence time. Groundwater He surveys of six U deposits in Australia were reported by Butt and Gole (1984, 1986) and Gole et al. (1986). Groundwaters (5 m b.w.t.) from within the Yeelirrie calcrete deposit, Western Australia, contain 0.001-0.13 pL/L He. These essentially background values are attributed to the shallow depth of mineralisation (mostly above the sampling depth and even above the water table) and the porous nature of the overburden. The below-ambient He values are a result of constant degassmg of the water by biogenic methane from decaying organic matter in old drill holes. Sandstone-hosted deposits in palaeochannels at Bennett Well and Manyingee, Western Australia, and Honeymoon, South Australia, contain 9.9-29.5 ~L/L, 0.008-1.7 ~tL/L and 6.9-44.4 ~tL/L He, respectively. The relatively low He concentrations at Manyingee are thought to be due to inflow of groundwater with near-background He contents displacing the waters in contact with the deposit. At Koongarra, Northern Territory, a vein/unconformity-type deposit in an area of active groundwater flow, the highest
Helium
331
concentration in samples from 20 m b.w.t, at the end of the dry season was 11 ~L/L He; seven months later, at the end of the wet season, a sample from the same site contained 14 ~L/L He, still the highest value in the area. Repeat samples from other holes also showed very similar He concentrations, with only one changing from the background (< 0.2 I.tL/L He) to the anomalous sample population (Gole et al., 1986). Groundwaters (50 m b.w.t.) from prospects in the area to the south of the Olympic Dam Cu-Au-U deposit, on the Stuart Shelf, South Australia, contained 0.58 to 4800 ~tL/L He, the maximum value being an enrichment of 105 relative to the atmospheric equilibrium value. The water samples were collected several hundred metres above the mineralisation. Of the samples from holes drilled into the mineralised basement, 11 were He-rich (mean 1590 ~tL/L He), saline (12.4% TDS) and most contained Rn, whereas eight had lower He contents (22.9 ~tL/L He), lower salinity (4.9% TDS) and no detectable Rn. It is inferred that the saline, He-rich samples probably represent groundwaters from the basement and the remainder are partly or wholly from the cover rocks, either because the holes are blocked or the shallow aquifers are more active. The available data thus show that groundwater in and near most U deposits is enriched in He above the atmospheric equilibrium value of 0.044 ~tL/L He. Exceptions may be found where the mineralisation is largely above the water table or where the groundwater is freshly recharged. It is apparent that the concentration of He around U deposits is generally within the range 1-100 ~tL/L He, i.e., an enrichment of up to l03 above the ambient value. Very much higher concentrations occur near the Olympic Dam deposit on the Stuart Shelf, but the extremely high salinities imply very long residence times for the groundwaters; He accumulation over time probably accounts, at least in part, for the exceptional enrichments found. Indeed, around Mt. Gunson in the southern part of the Stuart Shelf, an area of highly-saline groundwater with no known U concentrations, groundwaters sampled at <10 m b.w.t, contain 0.03-450 ~tL/L He, with deep samples (down to 50 m b.w.t.) yielding a maximum value of 1585 ~tL/L He (Butt and Gole, 1986).
Groundwater surveys in non-mineralised areas
There are, however, geological environments where groundwater may be enriched in He to a similar or even far greater extent than that shown to occur in or around U deposits (Table 10-IV). Concentrations of up to 100 ILtL/L He have been recorded from thermal springs (e.g., Mazor and Verhagen, 1976; Mazor et al., 1983) and He enrichments are recorded from groundwaters of meteoric origin associated with various lithologies. The highest concentrations (136-130,000 ~tL/L) are associated with granitoids (Andrews et al., 1982; Bottomley et al., 1984) and with sedimentary rocks overlying granitoids (14-200 ~tL/L, Marine, 1979). Tikhomirov and Tikhomirova (1971) and Golubev et al. (1974) have shown that the He content of formation waters is directly proportional to the depth to basement, confmnmg that basement rocks are a major source of He. Groundwater from faults and structurally-disturbed zones may also contain high He concentrations. Eremeev
C.R.M. Butt, M.J. Gole and W. Dyck
332 TABLE I 0-IV
Helium contents of groundwaters not associated with uranium mineralisation Country Australia Canada India S. Africa
Sweden UK USA
Location (reference) Great Artesian Basin (1) Chalk River granite (2) Lac du Bonnet granite (2) Gujarat and Rajasthan (3) Beaufort Group (4) Table Mountain (5) Stampiet (5) Stripa granite (6) Bunter Sandstone (7) South Carolina (8)
Sample Depth m n 1000 27 86-576 7 18-800 10 <20 (?) 60 Artesian 4 20-50 25 0-270 8 0-? 14 40-881 10 to 500 27 to 544 7
Water (He ~tL/L) Range x 2.9-340 0.15-870 136-130000 0..5-1.61 0.24 0.12-14.18 6.13 0.05-99 11 0.04-16" 6.3 1-93" 27 0.08-1820 0.04-0.34 0.1 ! 14-200 89 m
Rock (ppm) U Th 1.7-8 6-14 3 12 10 30 2 (?) 2 2 (?) 2 2 7 4 2.5 2.2 6* 17-39 30* 1.9 8.7 1.5 8
* Approximate References: (I) Torgersen and Clarke, 1985; (2) Bottomley et al., 1984; (3) Datta et al., 1980; (4) Heaton and Vogel, 1979; (5) Heaton, 1981; (6) Andrews et al., 1982; (7) Andrews and Lee, 1979; (8) Marine, 1979.
et al. (1973), Ovchinnikov et al. (1973) and Rozen and Yonitskii (1974) describe regional surveys in Russia in which concentrations up to 500 ~tL/L He are attributed to fracture systems in basement rocks. Bulashevich et al. (1973) found concentrations up to 19 ~tL/L He associated with faults, although where faults intersect, concentrations up to 180 IaL/L He occur (Bulashevich and Bashorin, 1971). Bowles et al. (1980) report groundwaters containing >130 ~tL/L He from the Lone Mountain Structural Zone in South Dakota, USA, although in the nearby structural zone related to the Dewey Fault the contents were <2.5 p.tL/L He. Dyck (1980) attributed high He concentrations encountered in a survey of 22,000 km 2 in the Ottawa area of Ontario to the possible presence of buried structural features such as an unconformity, a palaeochannel, or basement faults. Many studies have shown that confined aquifers may accumulate He above concentrations expected from the groundwater age (as determined by ~4C dating) and the contained U and Th (e.g., Heaton and Vogel, 1979; Andrews and Lee, 1979; Datta et al., 1980; Torgerson and Clarke, 1985). Indeed "He ages may be up to 1000 times higher than ~4C-derived ages (Heaton, 1984). Concentrations of 0.05-99 ~tL/L He have been reported from the Beaufort group sediments, South Africa, which contain an average of about 2 ppm U and 7 ppm Th (Heaton and Vogel, 1979). Concentrations of 2.9-340 ~tL/L He have been reported from the Great Artesian Basin, Australia, which contains 1.7-8 ppm U and 614 ppm Th (Torgersen and Clarke, 1985). Further studies (Heaton, 1984; Torgersen and
Helium
333
Fig. 10-7. Distribution of dissolved helium in groundwaters, Edgemont uranium district, South Dakota (from Bowles et ai., 1980).
Clarke, 1985; Torgerson and Ivey, 1985) have shown that this excess He is likely to have been derived from the underlying rocks and to have migrated, probably via fractures, into the aquifer where it accumulates. Thus, it is apparent that there are many geological environments where very significant enrichments of He occur in the groundwater. Indeed, when the He enrichments associated with U deposits are considered in their regional context, they lose much, if not all, of their exploration significance (Table 10-III). In the survey of the Edgemont district, South Dakota, USA (Bowles et al., 1980; McCarthy and Reimer, 1986), groundwater from within mineralisation located near Burdock was above ambient (one sample only, 0.48 ~tL/L He), but higher values occurred in 26 of the other 28 samples collected (Fig. 10-7). Indeed, the Burdock mineralisation is within a zone of low He, contrary to the relationship expected. It was concluded that in this area of artesian groundwater, hydrologic characteristics, such as high flow rates, which cause dissipation of He, have a far greater influence on the distribution of He in the groundwater than does the distribution of the U deposits.
C.R.M. Butt, M.$ Gole and HI. Dyck
334
.5:~' ~o
'\ \
ONSLOW
,woo
%
/
"o,~ --
I
.22"22.5'
\\
\
/ W ' ~ 174
_
\I \ \
\
\
",o
_ \ +'-"~" I
,
\\ ~
~
/
2
WO 75
/ ~36
-,
/
\ , , \ .., w O 5S\
\
-.
A
ii
wo 1
t2 0
~
~.
I
""
163
)woo
\,,!
-)
~
_ ~Ls o o 56
\ 4220 14
/
Ii:
/~.,,06e ///'\ ,~,
'l
]~
2 I
0
"~
W -4llk 29 / "~53' 28 "~'~ W L . ~ t .~"~'.411~ ~1~'~193022 ' (~Manyingee Corner W - . ~ - ~ - - ~ / "'10 ~" .42~',- " " ' ~ 1 ~ 6 2 ~ deposits
/
I
[
I
/
/
'-.
', ,
/
J
.o-'~
,,."
U
"
I
g20
.... o
/-
...'~.~'o~3.,UYx
/ f., w(5-
"~
'~
I
I
\ \ \\\
"-
,
\ \ X
\\
Uranium deposll
~
~ '~
% \
.
.
.
.
B
,/ \/I
Basin-basement boundary
~ ,
Well
W
~59\ ,wO \
~i,
51-80
4,- i t - s o 9 5-10
0 4
'i
200 ~ / ~" ~ Basement Isobath (m) "~" "~
I .~
0 .
',.
-)/
\
~ I
,/
I/
'. +~, K
~
."
(~. ~
~=_. '~\
' s s , Q ~ w O,ge
'~,~
\~\ \\ \ \ )o
/~
"~'1~
r'\
m,+,-
"o .
IOkm
4~ 0
+
I
70 0
2-s
1--2 0-1 Radon (counts/rain)
Fig. 10-8. Helium and radon contents of groundwaters from exploration drill holes, stock wells and bores in the Manyingee-Bennett Well area, Western Australia (from Butt and Gole, 1986).
Similarly, the He concentrations at the Manyingee and Bennett Well deposits, Western Australia, should be seen in their regional context (Fig. 10-8, Table 10-III). The highest values (at equivalent depths b.w.t.) in the neighbouring unmineralised areas were up to 40 times those at the Manyingee deposit and almost four times those at the Bennett Well deposit (Butt and Gole, 1986). The relatively low He content associated with the Manyingee deposit may be due to leakage of He from the deposit as a result of exploration activity and the high values are remnants of a dispersion halo down drainage. However, other interpretations are possible, such as leakage of He up faults in the basement or long groundwater residence times allowing accumulation of He generated in basement rocks. These latter factors appear to be those controlling the He distribution in the Yeelirrie area, a deeply weathered, largely granitoid terrain in which Cretaceous palaeochannels are choked with unconsolidated detritus. Water flow is now generally subsurface, mostly
Helium
335
within the calcrete aquifer that hosts U mineralisation. Groundwater samples were collected, where possible, at 5 m b.w.t, from: (a) 13 shallow holes into the Yeelirrie deposit; (b) 70 stock wells, bores and exploration drill holes that sample the shallowest aquifer; and (c) 50 deep, cased and slotted bores which penetrate fresh granitoid at depths of 20-80 m and thus sample, or have access to, relatively deep aquifers. Although the groundwaters from within the Yeelirrie deposit are not enriched in He (Table 10-III), equivalent samples from the shallowest aquifer in the area surrounding the deposit may contain 0.04-5 ~tL/L He (Fig. 10-9). A similar range in He concentration was obtained from shallow aquifers in the Mount Weld and Malcolm areas in the Yilgarn Block of Western Australia (Table 10-III; Butt and Gole, 1986). Samples from 5 m b.w.t, from the holes that penetrate to the deep aquifers in the Yeelirrie area, however, generally contain more He (Fig. 10-9), with exceptional values of 90-100 ~tL/L He; a maximum value of 215 pL/L He was obtained at 50 m b.w.t. The He distribution appears unrelated to any known geological feature, even when samples from the shallow and deep holes are considered separately. The deep holes show high variability over short distances and indeed some are not markedly enriched. The He distribution pattem is thought to reflect the presence of fractures which allow He generated in the granitoids to migrate upwards, and/or regolith features (e.g., low permeability, basement depressions) which enable He to accumulate to a greater extent. The shallow aquifers probably have a higher flow rate and hence shorter residence times, and also have a greater opportunity to equilibrate with the atmosphere. Palaeochannels incised into shallow-buried granitoid terrains represent attractive exploration targets for U. However, the Yeelirrie data demonstrate that in such terrain marked He enrichments may be unrelated to the presence of mineralisation. Also the He enrichments at Yeelirrie are considerably higher than those encountered in other U deposits within palaeochannels (e.g., Manyingee, Bennett Well, Honeymoon; Table 10-III). From results of a detailed and extensive study of groundwaters in the Monticello area, New York, USA, Rose et al. (1983) claim that He is a good guide to the mineralised horizon in this area. Factor analysis and the He distribution pattern shown on a crosssection suggest, however, that the anomalous He concentrations may be related to sampling of older, deeper groundwaters rather than the U deposits themselves, a possibility acknowledged by the authors. One locality where the He distribution may indicate the location of mineralisation is at Koongarra (Gole et al., 1986), where concentrations of up to 14 pL/L occur at 20 m b.w.t. in holes intersecting mineralisation, and anomalous values >0.6 ~tL/L extend for over 100 m downslope to the southeast (Fig. 10-10). Elsewhere near Koongarra, most groundwaters have only background He contents (<0.18 ~tL/L He), so that there is little doubt that mineralisation and/or the faulted contact in which it occurs are the He source. However, the available regional data are not extensive and the possibility of enrichments unrelated to mineralisation occurring elsewhere in the area cannot be discounted, particularly as springs with He-rich waters (up to 36 pL/L He) are known.
336
CR.M. Butt, M.J. Gole and W. Dyck
Llmeston
Iw., I'~_/c
/ ~
l~'-
......... ............
~ar,
o
9 . . . . . oo
~
9 08,-50
~
, . . . . o,o
.............
ooo
o~
(-
.....
\
\,
\
A
30-
" t Ow
0~0"
I
-
I
B
20
o
0.5
.. 1 . n . . . ~ J T . . - r -, 1o 5
He Air ii.
o
o 5
1o
5
3o 5011o
I.to ,~L/L
Fig. 10-9. Frequency distribution of helium in groundwater, Yeerlirrie area, Western Australia: (A) data from stock wells, bores and shallow drill holes (n=70); and (B) data from slotted, cased stratigraphic bore holes (n=50) (from Butt and Gole, 1986).
Su@lce water surveys
Minor excesses of He have been reported to occur mostly in shallow lake waters in the vicinity of mineralisation in Canada (Clarke et al., 1977; Dyck and Tan, 1978). Maximum concentrations were in the range 0.075-0.3 gL/L He, i.e., excesses of 1.5-6.5 times the relevant ambient solubility value. Similarly, Torgerson and Clarke (1978) reported He excesses in lake waters from north Ontario which, by analogy, they suggest may be due to the presence of hitherto unknown mineralisation. In contrast, Stephenson et al. (1992) relate high He abundances (0.060->2.8 gL/L) in lake waters in Manitoba, Canada, to discharge of deep groundwaters into the lake along faults. Such surveys are affected by the degree of equilibration between the water and the atmosphere. Thus, Dyck and Tan (1978) found that better results were obtained by sampling in winter, presumably because the ice cover prevented mixing and equilibration, a consideration adopted by Stephenson et al.
Helium
337
Fig. 10-10. Distribution of groundwater helium contents, Koongarra deposits, Northern Territory: (A) detailed scale; and (B) semi-regional scale (from Gole et al., 1986).
(1992). They also noted that, in general, the rather higher concentrations were associated with deeper waters, a result confirmed by a profile study reported by Torgerson and Clarke (1978). In a deep lake, these authors found the He content to rise from 0.047 ~L/L He at surface to 0.13 ~tL/L He at 130 m, with a sharp increase below the thermocline to 0.249 ~tL/L He at 150 m and 0.30 He p.L/L at 167 m. Stephenson et al. (1992) noted a similar depth variation in a shallow (<8 m) lake, with values declining sharply above the thermocline at 4-6 m depth. A conclusion is that deep lake waters can act as an integrator of He released over several months and might thus provide a suitable sample medium for regional surveys. Clarke et al. (1977) and Torgerson and Clarke (1978) also showed that the excess 4He concentrations were accompanied by decreases in 3He/aHe ratios, confirming the radiogenic origin of the He. However, the determination of this ratio does not seem to provide much additional information for the considerable extra difficulty involved. These authors attempted calculations as to the source of He in terms of the mass and probable distribution of the U parent. However, even where 3He measurements provide accurate time constraints, such calculations are inconclusive: the mechanisms by which He enters the lakes are unknown; and the relative importance of unmmeralised faults in granitic terrain and pegmatite mineralisation as He sources are undetermined.
338
C.R.M. Butt, M.J. Gole and W. Dyck
Discussion of water survey techniques Data outlined above show that features other than U mineralisation can give rise to similar or higher He concentrations in groundwater than might U mineralisation itself. Even in the vicinity of some U deposits, equivalent or higher He concentrations are encountered away from the deposit compared to those that occur within the deposit itself. The anomalous concentrations unrelated to U deposits are attributed to: (1) leakage, probably along faults and fractures, of He or He-charged waters from underlying rocks, particularly granitoid basement; (2) long residence time of groundwater, allowing accumulation of He; and (3) aquiclude impermeability to He. The influence of these factors, or even the possibility of their influence, effectively makes interpretation of He data impossible in many circumstances and thus nullifies the exploration potential of the method. The contributions of leakage, residence time and aquifer characteristics to He concentration are not only very significant but also largely unquantifiable. The problem is compounded in regional groundwater surveys where samples may be collected from several aquifers. Even where waters from a well-characterised groundwater system are sampled and analysed for numerous dissolved species, the origin of the He may not be readily determined (e.g., Bottomley et al., 1984). The determination of the age of groundwater samples may be an aid to interpretation in some circumstances. However, as He is commonly enriched in aquifers relative to the concentration expected from the groundwater age, the determination of whether such excess He is due to a nearby U deposit, due to addition via leakage, or due to some aquifer characteristic, is generally not possible. Even at Koongarra, where there is a close spatial association between anomalous He and the U deposits, there is a strong possibility that much of the He is derived from granitoids of the underlying Nanambu Complex by leakage up faults, the main one of which postdates the ore. Only where granitoid basement is very deep and there are no interconnecting faults are He contents unaffected by leakage. The properties of He that make it potentially useful in exploration, particularly for blind mineralisation (i.e., stability, inertness and radiogenic origin) are, it appears, the very properties that nullify its potential in exploration practice. The distribution and concentration of He cannot be interpreted unambiguously in terms of occurrence and location of mineralisation, and such data do not add usefully to information available from other sources.
HELIUM SURVEYS IN PETROLEUM EXPLORATION Most oil and gas reservoirs concentrate He to some degree and the most He-rich are sources of commercially-available He (Cook, 1979; Voronov et al., 1981). In some oil and gas fields, concentrations of several percent have been recorded (e.g., Boone, 1958; Dyck, 1976), but data summarised by Voronov et al. (1974) show that the average concentration is much lower.
Helium
339
Fig. 10-11. Helium soil-gas (probe) survey, Cement oil field, Oklahoma (from Roberts, 1981).
By virtue of this association, He, as a component of microseeps, has potential as a pathfinder for hydrocarbon accumulations. A number of He surveys have been carried out with a view to evaluating this potential. Most surveys have utilised soil and soil gases as the sample medium with the expectation that they might give patterns similar to those described for light hydrocarbon gases, such as: anomalies directly above the field, due to seepage through caprocks; halos due to leakage around the perimeter of the seal; and anomalies aligned along faults and fractures intersecting the trap structures. Surveys have been carried out using techniques similar to those used for U exploration. For example, Roberts (1981) described a survey over the Cement field, Oklahoma, USA, a large (20 x 6 km) anticlinal structure with multizone reservoirs at 600-2200 m depth. Soil gases on a 1.67 km 2 grid showed a He pattern with concentrations of 5.28-5.30 ppm immediately over the structure (Fig. 10-11), corresponding to the occurrence of bleaching, carbonate precipitation and other phenomena interpreted as being due to hydrocarbon microseepage (Donovan, 1974). The He pattern may be due to either seepage or environmental effects associated with these phenomena. Much higher concentrations were found over the Harley Dome, Utah, where a gas reservoir at only 250 m depth contains 85% N2 and 7% He. Soil gases collected on a grid with a 100 m sample interval showed an anomaly >5.74 ppm He, approximately 300 m in diameter, overlying the reservoir, with a maximum value of 15.3 ppm (Roberts, 1981). The anomaly is attributed to seepage through the caprock of the shallow reservoir.
C.R.M. Butt, M.J. Gole and W. Dyck
340
Bush Dome oil-water contact
( ~ ~ Q
i
Soil gas survey 1. Sampling traverses with He 5.29-5.48 ppm Soil gas survey 2. 0.8 km grid. He 5.30-5.42 ppm
Groundwater survey 9 <0.5 ~ L / L He 9 0.5-1.0 ~uL/L He 9 >1.0 ~ L / L He
0 L
2km
Fig. 10-12. Soil-gas (probe) and groundwater helium surveys, Bush Dome helium storage reservoir, Texas (from Holland and Emerson, 1979).
Helium surveys over the Bush Dome in the Cliffside oil field, Texas, USA (Holland and Emerson, 1979), are of considerable significance: although once producing natural gas containing 1.8% He, the field has subsequently been used to store crude (70%) He. The storage formation is predominantly dolomite at a depth of 975-1065 m, sealed by a 120 m thick caprock of anhydrite. Two soil gas and groundwater surveys were carried out. Groundwaters were sampled mostly from wells pumping from deeper than 25 m (water table, 5-40 m). The more reliable of the two groundwater surveys found a range of 0.092.56 laL/L He, with five of the six values >1.0 laL/L He occurring in the south and southwest margin of the dome (Fig. 10-12). The high values may represent leakage from the reservoir but, as discussed previously, many other sources are possible. In one soil-gas survey, samples from 0.6 m were collected at 320 m intervals on four traverses. These showed patterns of He contents of 5.28-5.49 ppm He over the dome with smaller areas
Helium
341
peripheral to it possibly forming a halo (Fig. 10-12). However, the contrast between the dome and its surroundings may in part arise from a bias in sampling density due to the radial arrangement of the sample traverses. A second survey, using a grid with sample sites at intervals of 800 m, showed a completely different pattern, with higher values (5.30 to 5.42 ppm He) displaced to the east of the dome, which in the earlier survey showed as a background area. Although the results of these investigations were positive in that anomalies were detected, their lack of spatial correlation and low abundances suggest that they do not represent leakage of the stored He gas. Jones and Drozd (1983) sampled gases from 4 m depth to reduce the effect of atmospheric equilibration. Over the shallow (60-300 m) Lost Hills oil field in California, USA, they noted that He (and H2) anomalies were present near faults, particularly those penetrating into the basement. They concluded that whereas H2, together with the light alkanes, appears to be a petroleum indicator, He may be a tectonic indicator, independent of petroleum and gas. Using samples from 6 m, Gole and Butt (1985) reported a similar study over the Gingin gas field, Western Australia, where the gas-bearing structure is a faulted anticline at a depth of 3860-4160 m. Soil gases with high concentrations of He extend for 1.1 km west of the now-abandoned well, with peak values exceeding 6.0 ppm He accompanied by anomalous methane, C2-C5 alkanes, H2 and ethylene over the suboutcrop of the fault closing the structure. Normalisation using 2~ reduced the extent and magnitude of the He anomaly, although it was retained over the fault suboutcrop. Gole and Butt (1985) postulated that the anomalies were due to microseepage from the reservoir, with microbial oxidation of the hydrocarbons resulting in a net consumption of oxygen and hence a residual enhancement of He. This suggestion was supported by the presence of ethylene and H2, both products of microbial activity. Only close to the fault itself was the He leakage sufficient to yield a detectable anomaly. Butt and Gole (1984) also reported the presence of He and methane anomalies in deep soil gases over faults associated with the Woodada gas field, Western Australia. Roberts (1981) reported that soil samples analysed by the methods of Pogorski and Quirt (1979, 1981) gave similar results to free soil gases sampled by probe, provided the soil was damp enough to retain its integrity during sampling, and noted that it was the only possible sample in water-logged or permanently-frozen soil. For example, over the Rubelsanto oil field in Guatemala, which produces from depths of 1500-2300 m, mud samples collected from 0.5 m contained background He concentrations of 0.14-0.4 laL He/L mud and all higher values (0.85-3.00 laL He/L mud) occurred close to the producing wells. Roberts (1981) also reported three surveys from permafrost terrain in Alaska, USA, but found no clear relation between He distributions and hydrocarbon reservoirs. Most promising was a survey of the Cape Simpson area, in which samples from 0.7 m depth on a 1.5 km grid yielded a broad halo of higher values (>0.7 ~tL He/L H20). A number of seepages are known in the vicinity and the anomaly is in an area where magnetic anomalies have been attributed to diagenetic magnetite formed by the reduction of iron oxides by seeping hydrocarbons.
C.R.M. Butt, M.J. Gole and W. Dyck
342
9~
~
He >6.5 ppm
~
Oil fields Approximate
9 .9 ~
sample
locations
9
.
- - - - - Faults
9
,--..(p-- Anticlinal axis 9
"
> 9
.
9
".
.~..
f.-~
9
t
2 km
9
i
Fig. 10-13. Helium in soils over the Eldingen oil field, Germany (reproduced with permission from Van den Boom, 1987, J. Geophys. Res., 92" 12,547-12,555, copyright by the American Geophysical Union).
Deep (15-25 m) overburden samples were used by Van den Boom (1987) for a survey of the Eldingen oil field, Germany. Samples were collected at a density of 3/km 2 over a 70 km 2 area and equilibrated for two to three weeks in sealed containers. The data appear to have three populations, the highest of which (range 6.2-9.0 ppm He, mean 7.4 ppm He) gives anomalies overlying the reservoirs and associated faults (Fig. 10-13). Although results from soil and overburden sampling appear encouraging, the possibility remains that differences in major gas composition, either original or due to biological activity during equilibration, may have affected He concentrations, either by analytical enhancement (using a CP-inlet) or by residual concentration/dilution. It is unclear whether or not calculation of the results included consideration of meteorological and other variables, and hence the validity of the data and interpretations cannot be readily assessed. The He content of waters has rarely been investigated for hydrocarbon exploration. However, Dyck and Dunn (1987) report that broad, coincident anomalies of >2.4 gL/L He and >0.4 gL/L methane in wells and springs in the Cypass Hills district, Saskatchewan, Canada, correlate with commercial oil and gas fields and/or tectonic features. High He concentrations can distinguish thermogenic methane from biogenic methane (marsh gas).
Helium
343
In summary, high concentration of He in petroleum and natural gas reservoirs represents a richer source of He than do U deposits. Hence, He surveys for petroleum and natural gas may be expected to be more successful than He surveys for U deposits. High soil-gas He contents have been reported over the shallow and unusually He-rich Harley Dome, but most of the available data are not encouraging, possibly because of the sampling procedures used. Even for one of the more convincing surveys, over the Cement field, the anomaly contrast is very low and within the range of sampling error, analytical error and background variation. This range is probably demonstrated by the two surveys over the Bush Dome storage reservoir, in which the distribution patterns differed very significantly though the spread of values was the same. The absence of a significant anomaly over this very rich He source implies that either there is no leakage or that the sampling technique is inadequate. Where deeper soil gases have been sampled, however, more consistent results have been observed, with higher values and greater contrasts than for shallow samples. These results suggest that there is potential for He surveys in hydrocarbon exploration, preferably as an adjunct to hydrocarbon gas geochemistry itself, but further detailed studies are necessary before that potential can be adequately assessed.
HELIUM SURVEYS IN GEOTHERMAL RESOURCE EXPLORATION Helium in thermal waters and gases
Waters and gases issuing from thermal springs are commonly enriched in He and this association has prompted the suggestion that He surveys might have potential in exploration for geothermal reservoirs (Roberts et al., 1975; Hinkle et al., 1978; Corazza et al., 1993; Kahler, undated). Roberts et al. (1975) report that gases dissolved in waters from several hot springs in the westem USA contain over 100 ppm He. However, published data for both gases and waters suggest that rather lower concentrations are usual. A representative, but not exhaustive, list of He, Ne and Ar contents for thermal gases and waters around the world is given in Table 10-V. Gases bubbling from springs generally have He contents of less than 50 ppm He, with a range from 4.0 ppm (i.e., below the atmospheric content) to several percent. The highest values are from springs in granitic terrain in South Africa (Kent, 1969) and France (Batard et al., 1982), and from carbonatites in Swaziland (Hunter, 1969). Waters generally contain less than 5 ~tL/L He, in the range 0.044 (the atmospheric equilibrium value) to 104 ~L/L He. The He in these waters may come from three sources: (1) from the atmosphere, during reservoir recharge; (2) from flushing of radiogenic He from rocks through which circulating waters pass; and (3) by addition of He from deeper in the crust or from the mantle. The mantle component has been recognised by high 3He/4He ratios in a number of areas, for example, in Yellowstone, USA (Craig et al., 1978a), in Iceland (Kononov et al., 1974), in the former USSR along the Baikal Rift zone (Lomonosov et al., 1976) and in Kurile-Kamchatka (Gutsalo, 1976), and in Japan and the Pacific (Nagao et al., 1979; Craig et al., 1978b). The other rare gases (Ne,
344
C.R.M. Butt, M.J. Gole and W. Dyck
Ar; Table 10-V) are generally depleted relative to their concentrations in atmospheric air (18 ppm Ne, 0.93% Ar) and to their atmospheric equilibrium values in water (0.18 ~tL/L Ne, 440 ~tL/L Ar). These gases are generally considered to be derived from the atmosphere, although Ar may also have radiogenic and mantle components. Data for Rn are fewer but it too is frequently enriched in thermal waters and gases. Cox (1980) reports the use of Rn determination for geothermal resources in Hawaii; hence, by analogy with U exploration, He might have advantages where the thermal activity has no obvious surface expression. However, He concentrations in thermal waters are exceeded by those in many long-residence groundwaters. Where such groundwaters degas, as in artesian bores or when pumped, gases may also be He-rich. These waters may or may not be warm, hence the He content itself is not evidence for a source of geothermal energy. Nevertheless, since it may be released from upwardflowing geothermal waters, He has potential as a pathfinder.
Helium surveys for geothermal resources The possibility of using He as an exploration guide for geothermal reservoirs was investigated by Roberts et al. (1975), who conducted a soil-gas survey around the Indian Hot Springs at Idaho Springs, Colorado, USA. Soil gases at 0.5 m sampled by hammered probe were found to have some exceptionally high He contents in comparison with those reported over U and hydrocarbon deposits, but the areal extent of the anomalies was remarkably small. Thus, at a hot spring (40"C), a maximum value of 1000 ppm He was found in an anomalous area of 15 x 20 m as defined by the 5.4 ppm He contour. Similarly, close to a warm spring (26~ the maximum value was 100 ppm He in an area of 50 x 125 m delineated by the 5.4 ppm He contour. These values contrast with the background of 5.2 ppm in the surrounding area. Other surveys (e.g., Hinkle et al., 1978) at the Roosevelt Hot Springs, Utah, have utilised as samples headspace gases over equilibrated soils; hence claims of elevated He contents in the vicinity of faults are doubtful (see above). Gregory and Durrance (1987) used stream water sampling to determine regions of high heat flow over granitic plutons in southwest England. They suggest that elevated concentrations of He (>0.044 ~tL/L, the atmospheric equilibrium value) and Rn (>122 counts/min.) indicated discharge of groundwaters from the upwelling limbs of hydrothermal convection cells. Conversely, they also suggest that minima in the distribution patterns are regions of descending groundwater. How descending groundwaters can influence the He concentration of stream waters is not explained, nor is a geological explanation apparent. It is more reasonable to suggest that most of the variations in the data are artifacts of sampling and analytical techniques. Subtle changes in major gas composition may give analytical problems at low He abundances determined with a CP-inlet. Even if concentrations greater than 0.044 ~tL/L He are real, differences in sampling conditions (e.g., distance from source spring, turbulence, relative flow rates of spring and stream) make inter-sample comparisons hazardous.
345
Helium
TABLE 10-V Range and (in italics) mean of helium, neon and argon contents of thermal spring gases and waters (all data v/v at STP; volume relationships between gases and waters unknown, hence measured values for waters are only minimum estimates) Type Gas
Location (reference) Yellowstone, USA (1)
No. 8
Yellowstone, USA (2)
4
Kerch-Taman, Russia (3)
4
Iceland (4) Baykal Rift, Russia (5)
1 2 1 9
Baykal Rift, Russia (6)
10 Kurile, Russia (7) Nigorikawa, Japan (8) South Africa (9) Swaziland (!0) Tanzania (11) Fairmont, Canada (12) Massif Central, France (16)
Water
5 I 2 i
- CO2 > 93%
7
- CO2 1-40%
10
Yellowstone, USA (2)
4
Zimbabwe (13)
6
Tiberias, Israel (14)
11
Dead Sea, Israel (14)
16
Baykal Rift, Russia (6)
9 12
Marianas Islands (15) Rocky Mts, Canada (12)
2 9
He (ppm) 5.5-34.0 12.9 127 (1 sample) <10-1630 150 43 30, 50 2420 38.5-62.7 49.1 9.5-187.0 72.9 20-100 4.0 to 9.95% 4.70%, 5.54% 0.1-13.2% > 1200 20-200 0.5-10.9% 0.09-6.42 2.04 0.08-104 34.3 0.33-26.6 9.51 0.04-1.79 0.46 3.25-7.65 4.67 2.4-16.4 8.2 0.04, 0.33 0.45-17.9 6.1
Ne (ppm) 0.65-1.70 1.0 0.12-0.72 0.31 -
Ar (ppm) 246-1640 820 153-620 314 10-3100 250 3950, 9850 2230 -
0.014 -
To 1.53% 1 . 6 2 % , 1.66%
_
>2
>2640
0.01-0.19 0.07 0.04-0.19 0.12 0.1 I-0.23 0.19 0.15-0.23 0.18 -
0.5-1.3% 0.89% 9.3-276 98 43-300 205 213-322 295 259-381 306 -
-
-
0.04-2.73 0.48
83-1300 403
References: 1, Mazor and Wasserburg, 1965; 2, Mazor and Fournier, 1973" 3, Lagunova, 1974; 4, Mamyrin et al., 1972; 5, Lomonosov et al., 1976; 6, Yasko, 1981" 7, Gutsalo, 1976; 8, Nagao et al., 1979; 9, Kent, 1969; 10, Hunter, 1969; 11, Walker, 1969; 12, Mazor et ai., 1983" 13, Mazor and Vergagen, 1976; 14, Mazor, 1972; 15, Craig et al., 1978" 16, Batard et al., 1982.
346
C.R.M. Butt, M.J. Gole and W. Dyck
A more thorough and successful soil-gas survey was conducted by Corazza et al. (1993) in the Vulsini Mountains, central Italy. The surveyed area has a high geothermal gradient and is characterised by anomalous concentrations (0.5-> 10%) of CO2 in soil gases. Helium and H2 were determined as the most diffusive gases that might indicate the most geothermally-active areas. The authors countered problems of atmospheric contamination and dilution by CO2 (and, incidentally, the effects of the latter on He analysis using CPinlets) by normalising the He data to Ne. The He and H2 data identified three sectors within the CO2-enriched area. The sector enriched in H2 (4-18 ppm) and He (up to 5.8 ppm) and with high He/Ne, He/CO2 and H2/CO2 ratios was interpreted as having buffed faults and fractures through which deep, hot fluids were escaping, and thus as having potential for geothermal energy. In conclusion, it is evident that geothermal waters do, in general, have elevated He contents and hence He surveys may have some potential for their location. The Indian Hot Springs example (Roberts et al., 1975), however, suggests that the He excess in soil gases, though very intense, is conf'med to a very restricted area. The implication is that lateral dispersion of He from the conduit carrying the thermal waters is negligible. However, alternative sampling and analytical procedures may give better results (e.g., Corazza et al., 1993). Surveys for regions of high heat flow would probably be more reliable if the springs themselves are sampled, rather than soil, soil gas or surface water.
HELIUM ASSOCIATED WITH FAULTS
Faults as secondary sources o f helium Faults and fractures can act as pathways for the migration of He derived from the crust and mantle, and hence are possible secondary sources for anomalies located during exploration surveys. The potential use of He surveys to locate deep faults, and any associated mineralisation, has been discussed in the Russian literature in particular. Faults in seismically-active zones similarly act as channelways for the escape of gases, and He is one of a number of components of soil gas and groundwater that are monitored to determine whether variations in concentration can be used in the prediction of earthquakes.
Groundwater surveys Much of the Russian literature on He geochemistry has discussed the association of high He concentrations in groundwaters with deep faults. In general, samples have been taken from pumped bores at depths of 60-100 m. Where shallower waters have been sampled, results are less certain because of upward-degassing gradients and dilution near the surface (Eremeev et al., 1973; Bashorin, 1980). The conclusions of much of this work are that the He content of the groundwaters is less dependent on the radioactivity of the
Helium
347
rocks than it is on the existence of permeable faults. Bulashevich et al. (1973) sampled along a deep-sounding seismic traverse in the Transurals and found that the peak He concentrations occurred over faults and fractures, some of which extended for more than 30 km from the Precambrian basement into overlying Phanerozoic sediments. Of 26 such faults, 17 were shown by peaks of 3-5 ~tL/L He (maximum 19 ~tL/L He) against a background of 0.05-0.2 pL/L He (concentrations recalculated from Eremeev et al., 1973). Five other faults gave lower peaks and four generated no anomaly, indicating that they were sealed. Higher concentrations (up to 180 ~tL/L) are noted where such faults intersect, since these represent the most permeable zones (Bulashevich and Bashorm, 1971). Eremeev et al. (1973) and Rozen and Yanitskii (1974) report the results of regional He surveys in areas of outcropping or shallow-buffed basement in the former USSR, using sample densities of 1 sample per 100 km 2 closing to 1 sample per 10-25 km 2. These surveys reveal contrasts of 104 in concentration, from 0.05 jaL/L He (background) to 500 ~L/L He. The results are interpreted as reflecting permeable fracture systems extending to depths of 3-50 km, identified from geophysical surveys. These fractures form a northwestand northeast-trending network, which the authors state to be 300-400 m.y. old. They conclude that the Earth's crustal structure is one of impervious solid blocks, such as granitic masses, with surrounding permeable zones. Leakage along faults in the basaltic basement, possibly coupled with some geothermal activity, was considered to be the cause of high He concentrations (up to 14.18 ~L/L) in groundwaters in Gujarat and Rajasthan, India (Datta et al., 1980). The distribution of some high values seemed to be related to known faults and the concentration appeared to have an inverse relationship to the thickness of the cover rocks. Helium surveys could be considered, therefore, to be effective in locating faults and permeable zones, which might be the sites of certain types of mineral deposits. Examples are given of He anomalies associated with carbonatites, Au-quartz veins, Pb-Zn-Ba veins and Cu-pyrite veins (Eremeev et al., 1973), stockwork Mo-deposits (Ovchinnikov et al., 1973), volcanogenic sulphides emplaced in faults (Bulashevich and Kartashov, 1976) and kimberlites (Kravtsov et al. 1976, 1979). However, except in the case of U- and Th-bearing deposits, the He and other rare gases in faults originate deep in the crest, or perhaps the mantle, and are "indifferent" to ore deposition (Ovchinnikov et al., 1973). Thus they indicate only a potential site, not the mineralisation itself. When the fractures contain radioactive mineralisation, the He flux can increase markedly (Tugarmov and Osipov, 1974) and should be reflected in the He content of groundwaters. However, comparison of absolute values for the purposes of mineral exploration can be confounded by the unknown effects of factors such as the depth and permeability of the fracture and its proximity to the sampling site. Stephenson et al. (1992) used a He survey of water in the shallow (<10 m) Boggy Lake, Manitoba, Canada, to detect zones of groundwater discharge from faults. Granitic basement rocks in the region have mean contents of 7 ppm U and 33 ppm Th, 3-4 times the average for the Canadian Shield, and associated groundwaters have elevated He contents of 10 pL/L to a maximum of 56 pL/L (Gascoyne and Wuschke, 1990, as reported by
348
C.R.M. Butt, M.J. Gole and W. Dyck
Stephenson et al., 1992). Lake water sampling at the end of winter, when a maximum He content in lake waters was expected, showed zones of He enrichment of 0.060->2.8 IuL/L corresponding to the interpreted outcrop of the fault. Helium anomalies were also noted in flowing stream waters, which have extremely high baseline concentrations of 0.30-0.40 ~tL/L compared with the atmospheric equilibrium value of 0.044 ~tL/L. Maxima exceed 5 ~tL/L (130 and 272 times equilibrium value) and decline to the baseline value over about 500 m. The He is interpreted as being derived from enriched groundwaters discharging into the stream through fractures, with the peak values corresponding to a major fault.
Soil-gas surveys There have been few attempts to use soil-gas sampling for locating faults. Some examples given by Rosier et al. (1977) show low-contrast anomalies (1.5-2.0 times) for He, Ne and Ar, but no details are given of the sampling procedure. Denton (1977) claimed that a fault at Roosevelt Hot Springs, USA, is indicated by elevated He concentration in shallow soil gases, but it is not obvious from the data. Helium anomalies in deeper (4-6 m) soil-gas samples over hydrocarbon reservoirs have been attributed to leakage along faults, as discussed above. Indeed, Jones and Drozd (1983) consider He anomalies only as indicators of deep fault systems rather than hydrocarbons, citing as examples very specific enrichments of 20-430 ppm He in soil gases over the San Andreas fault (see Chapter 5). They attribute the difference between these values and the maximum of 5.33 ppm He found by Reimer (1981 ) as due to the latter's use of shallow sampling, although differing location may also be partly responsible. However, it cannot be assumed that all faults would yield He anomalies, certainly not of the magnitude of those over the San Andreas fault. Butt and Gole (1984) found no He or methane anomalies over the Warradarge fault, Eneabba, Western Australia, using 6 m deep samples on 25 m spacings. The fault has a throw of at least 1000 m and cuts coal-bearing rocks and sedimentary units that are the reservoir rocks for the Woodada gas field, 20 km to the north. Duddridge et al. (1992) report anomalies of high He (5.257 ppm), Rn and CO2 and low 02 in shallow soil gases over faults in England and Italy, but other random anomalies of equivalent magnitude remain unexplained. The variation in He abundance is within the range of analytical error using a CP-inlet, as discussed above, and the correspondence between He and CO2 distributions may reflect this problem; in addition the He maxima are less than twice the analytical sensitivity of 0.030 ppm. As with all such surveys, the results are equivocal and vary with time, even when data for all gases are integrated. This variability may reflect fluctuating gas emission along the fault, but it is equally possible that it reflects background variation due to a variety of environmental, meteorological, sampling and analytical effects, and has no significance.
Helium
349
Helium monitoring in earthquake prediction Although most emphasis in earthquake prediction has been placed on geophysical procedures, it has been demonstrated that geochemical parameters may give corroborative evidence and, in some cases, be almost the only premonitory phenomenon of some earthquakes. Radon, He, H2 and their isotopes are one group of components whose variations (usually increases) in concentrations might be used to predict the occurrence of seismic activity. For the Tashkent earthquakes of 1966 and 1975, for example, the He and Rn contents rose by factors of 12 and 3 in 1966 (Ovchinnikov et al., 1973), and in 1975 the He content increased fivefold (Sardarov, 1981); a detailed review is presented by King (1986). In order to evaluate these procedures, a number of long-term monitoring projects have been established. Their aim is to determine: firstly, the background variations of the measured parameters due to factors such as fluctuations in atmospheric pressure and temperature, climate and Earth tides; and secondly, whether deviations from the background pattern can be related to seismic activity. Waters and gases from thermal and artesian bores, deep groundwaters and soil gases have been tested as sample media. Numerous instances have now been recorded where concentrations of certain gases do appear to change a few days prior to earthquakes, even where the epicentres are tens or even hundreds of kilometres distant. Chalov et al. (1976) reported that the He and Rn contents of thermal waters fluctuated markedly and periodically reached very high concentrations prior to earthquakes with epicentres 35-265 km distant, although no shocks were recorded at the sampling sites themselves. For He, background variations were about 25-30% of the mean but, before the earthquakes, the amplitude and frequency of fluctuations increased, with maxima from twice the mean to "off-scale". Radon variations were of lower amplitude and returned to normal earlier. Broadly similar results are reported by other Russian workers (e.g., Borodzich et al., 1979; Barsukov et al., 1979) using waters from springs and artesian bores. Sugisaki (1978, 1981, 1987) analysed gas bubbles from a thermal spring in Japan (mean approximately 840 ppm He, 30 x 10-~~Ci/L Rn) and noted that increased He/Ar and Nz/Ar ratios occurred a few days before earthquakes with epicentres over 200 km distant. The length of the period that the ratios were high was also possibly related to the magnitude of the shock. It was noted that the most useful results were obtained from gas bubbles issuing from springs along faults, rather than from gases occluded within essentially static groundwaters. In the USA, investigations have concentrated on Rn, with increases in Rn emission noted in soil gases before volcanic activity in Hawaii (Cox, 1980) and along the San Andreas and associated faults in California (King, 1978, 1980a). Continuous monitoring of Rn and He at the Arrowhead Hot Springs on the San Andreas fault has shown that increases of over 60% above baseline levels preceded seismic activity in 1979 and 1983 (Chung, 1985). Monitoring of He in shallow soil gases has also been carried out in the vicinity of the San Andreas fault and Reimer (1981) reports a decrease in the contents up to three weeks prior to some, but not all, earthquakes. On two occasions, these
350
C.R.M. Butt, M.J. Gole and IV. Dyck
observations were used for forecasts. Reimer reported a maximum of only 5.33 ppm He at 2 m, whereas elsewhere on the fault, Jones and Drozd (1983) reported 20-430 ppm He at 4 m, within 60 m of mapped suboutcrop. Although this disparity might in part be due to atmospheric dilution of Reimer's samples, the results imply differing permeability and sources along the fault. There is, therefore, evidence that the determination of He, Rn and other gases can have value in earthquake prediction. It is generally considered that stresses prior to the event increase the permeability of fractures (e.g., Barsukov et al., 1979), which not only permits the escape of gas from sites of formation or accumulation but also causes increased groundwater flow. This flow is perhaps supplemented by deep-seated waters themselves charged with He, Rn and other rare gases. This hypothesis is supported by increases in 3He/aHe ratios in springs near some seismically-active centres (Sano et al., 1986; Wakita et al., 1987). An alternative explanation is that increased compression before the shock "squeezes" gas from the rock (King, 1978; Sugisaki, 1981). Conversely, Reimer (1981) suggested that the decreases he observed were due to He escape being prevented either by compression, or by dilation allowing the influx of water. Groundwaters appear to be the most suitable sampling medium, particularly for He. Nevertheless, adjacent wells may give different responses (Chalov et al., 1976) and, in some cases, apparently significant fluctuations may not relate to seismic activity, although they may still be related to stress effects. The evidence for the usefulness of soil gas as a sampling medium for He is less convincing. That Reimer (1981) observed decreases appears unusual and it is assumed that this is not due to dilution or effects unconnected with seismic activity. Although very short term effects were apparently eliminated by sampling at 2 m, longer term, possibly seasonal, variations (6-12 months) were present, with an averaged range of 5.24-5.33 ppm He; this is similar to the range reported by Reimer and co-workers for soil gases near U mineralisation. Any decreases due to seismic activity superimposed on this seasonal range can therefore be only very subtle, as the atmospheric content (5.24 ppm He) limits the amplitude. On this evidence, soil-gas sampling may have limited application. However, since Jones and Drozd (1983), using deeper sampling, reported much higher He contents elsewhere on the same fault, it is evident that further experimentation is warranted.
CONCLUSIONS
Migration of helium in the near-surface environment Helium liberated from U mineralisation and hydrocarbon reservoirs, or issuing from faults and geothermal conduits, migrates by diffusion and groundwater flow below the water table and by diffusion and barometric pumping in the soil atmosphere. Although theoretical calculations can be made conceming the magnitudes and rates of migration, these are difficult to quantify in practice. There are uncertainties regarding the He source,
Helium
351
hydrology and comparability of groundwater samples, and for soil gases there are doubts about the sources of He and the suitability of sampling procedures. Groundwater data are characteristically "spotty" on local and regional scales, with tenfold or greater variations between apparently similar samples 0.2-3 km apart (Butt and Gole, 1986). Assuming these differences to be real, the implication is that He does not disperse to give a widespread, uniform enrichment or dispersion plume, as might be expected in slow-moving groundwaters. It appears that local influences, either hydrological or source-related, are more important. The most defmite information on lateral dispersion is from Koongarra, Northern Territory, Australia. There the He content of groundwater declines from 3.5-8.0 ~tL/L He in contact with mineralisation to 0.6-1.2 ktL/L He at 100 m and then to background (0.09 ~tL/L He) at 220 m (Gole et al., 1986). There are few data that quantify vertical migration, though it must be considerable through open fractures. In unfractured sequences, there is no evidence to suggest that slow leakage through the aquicludes permits He contents of upper-aquifer waters to reflect those of lower-aquifer waters. Quantitative vertical interchange between or within aquifers is essential if He methods are to have value in exploration for blind deposits, but the spotty nature of distribution patterns suggests that interchange is probably localised and that any concentrations measured in the upper aquifers are probably related to the proximity, size and nature of connecting features. The concentration of He in soil gases is very uniform, due to equilibration with the atmosphere. Apparent spottiness in the data (e.g., Fig. 10-4) results from selection of narrow class intervals and probably represents background variation, sampling errors and analytical errors rather than real differences in He flux. Where significant soil gas anomalies exist, they tend to be immediately above the source, with little lateral spread. This is demonstrated by the intense (5.4-1000 ppm He) but spatially-limited (15 x 25 m) anomaly found by Roberts et al. (1975) in shallow soil gases at the Indian Hot Springs, Colorado. Similarly, at the Gingin gas field, Western Australia, a He anomaly is restricted to about 50 m over the suboutcrop of a fault, although hydrocarbon leakage may be more widespread (Gole and Butt, 1985). Migration by groundwater flow and subsequent degassing has been invoked to explain displacement of possible soil anomalies (e.g., in Weld County, Colorado, Fig. 10-3). However, where the He distributions are recorded in both groundwaters and overlying gases, no relationship is evident, with gas He contents being low when groundwaters are enriched, and vice versa. Patterns shown by shallow soil gases and groundwaters at the Bush Dome He storage reservoir, Texas (Fig. 10-12), and the Edgemont U district, South Dakota (Bowles et al., 1980), are quite different. Similarly, Butt and Gole (1985) found that deeper overburden gases, even within 6 m of the water table, did not reflect quite substantial groundwater anomalies of up to 13.6 ~tL/L He. Nevertheless, it must be assumed that He in groundwaters in confined or semi-confined aquifers is in dynamic equilibrium with that in gases in overburden and soil above the water table. The lack of correlation between the two sample media suggests that the rate of equilibration between water and overburden gas must be slow, relative to that between overburden gas and the atmosphere.
352
C.R.M. Butt, M.J. Gole and W. Dyck
Application of helium surveys The principal conclusion of the considerable research that has been devoted to the use of He in exploration is that it has little to offer, mainly because of a multiplicity of sources, unknown or unquantifiable factors affecting accumulation and dispersion, and the absence of cost-effective means of obtaining reliable samples. Thus, although surveys have shown that uranium-rich mineralisation has associated groundwater He anomalies, far higher concentrations can be due to leakage from granitoid basement or accumulations over time, even where U and Th are at normal crustal abundances. Aquifer characteristics, such as porosity, structure, flow rates and the permeability of the aquiclude, also influence the He concentration. Thus, where hydrology is poorly known, as is probable in areas subject to exploration, samples may be from different aquifers that have different characteristics with respect to the acquisition and retention of He. In general, therefore, present experience suggests that neither the distribution nor concentration of He can be unambiguously interpreted in terms of the occurrence of mineralisation, nor can He data provide useful information additional to that obtained by other means. Even the structural information derived from surveys in the former USSR could probably be obtained more readily using satellite imagery or airborne geophysics. Most gas surveys seem to have been unsuccessful, particularly in U exploration. Gas samples are subject to equilibration with the atmosphere, resulting in loss of excess He. It is doubtful, therefore, whether shallow soil gases (from less than 1 m) indicate spatial or temporal variations in He flux. If gas sampling is to be used for He or any other component, samples should be collected from as deep as possible (water table permitting, certainly below 3 m and preferably at 6 m or deeper). However, there is no certainty that variations in He flux will be present or detect able. Gas samples are also subject to variations in bulk composition due to biological activity, including that associated with the degradation of methane leaking from hydrocarbon reservoirs. This has residual effects on He concentrations and also may lead to analytical errors. Consequently, normalisation of He data using ZONe, which is similarly affected, is desirable. A comparison between raw and normalised data may indicate the extent of the biological activity and reveal variations in He flux. The poor correlation in He contents of enriched groundwaters and gases in soil and overburden implies, perhaps surprisingly, that the gradual loss of He from such waters does not give rise to detectable increases in He flux. This is perhaps as well, given the prevalence of He enrichment unrelated to economic targets. Increases in the He flux detectable in overburden gases seem to occur when migration is confined to faults and fractures, from which there is usually little lateral spread. Useful results are most probable when exploration targets are associated with faults, with He surveys used as a follow-up to remote-sensing and geophysical surveys that initially locate the faults. The greatest potential for He surveys probably lies in hydrocarbon exploration, as an adjunct to other gas surveys.
Geochemical Remote Sensing of the Subsurface Edited by M. Hale Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
353
Chapter 11
RADON W. DYCK and I.R. JONASSON
INTRODUCTION Friedrich Emst Born has been credited with the discovery of 222Rn in 1900 (Partington, 1957), but only because he published his findings before Ramsay and Soddy who a year later, after having determined its atomic weight, characteristic spectral lines and chemical inertness, placed Rn in the Periodic System as the element niton (Stantso, 1974). A year earlier Rutherford and Owens had discovered that the erratic electrometer readings made during the course of measurements of thorium-salt and radium-salt activities were due to the emanation of a radioactive substance from the salts. These emanations (from Latin, emanatus, meaning qualities or properties issuing from a source) of finite lifetimes were subsequently proven to belong to the noble gas family. Their gaseous and chemically-inert nature, ease of detection, and presence in all natural materials made them excellent tracers in studies of atmospheric circulation, lithospheric emanations and environmental health, and suggested their application as a pathfinder in exploration for U deposits. Perhaps the earliest reference to the use of Rn in prospecting is found in Le Radium (1904) where the use of a giant ionisation chamber for the measurement of soil emanations is described and collection of Rn from a stream using an inverted cone-andbottle assembly is illustrated. In those days the quest was for radioactive springs for health spas rather than for U, which had few known uses. Some years later, Behounek (1927) concluded from atmospheric and soil-air Rn measurements in St. Joachimstal that the radioactive halo around U mineralisation was measurable within a 300 m radius. Most recently Rn has been used for earthquake monitoring and prediction (Hauksson, 1981 ; Sato et al., 1980). The decline in fossil fuel reserves and the advent of nuclear power eventually created a great deal of interest in Rn as a specific sensitive pathfinder for buried U deposits. In the 1930s it had been tested in Europe as one of the principal tracers for U; in North America it came into wide acceptance only in the last few decades. The excellent and comprehensive work of Russian scientists (Grammakov, 1934; Alekseev et al., 1959; Starik, 1959; Baranov, 1961; Novikov and Kapkov, 1965) on the behaviour of Rn in natural environments went largely unnoticed in the West for several decades. However, comprehensive reviews on the migration of Rn in the ground (Tanner, 1964a,
3 54
W. Dyck and I.R. Jonasson
1978, 1980; Gesell and Lowder 1980; United Nations Scientific Committee, 1977) coupled with several original research endeavours on Rn behaviour (Adams and Lowder, 1964; Adams et al., 1972; Clements, 1974; Soonawala, 1976; Korner and Rose, 1977) and the successful application of Rn methods to the search for U deposits brought recognition to the early work on Rn methods of prospecting into renewed prominence. To some readers the review on radiation in U mines (a manual of Rn gas emission characteristics and techniques for estimating ventilating requirements) by Thompkins (1982) and a number of articles on biological uptake and distribution of Ra in natural environments contaminated by U mine and mill wastes (IAEA, 1982) may also be of interest. The mistrust that has built up around certain applications of the Rn method of prospecting is largely the result of a lack of understanding of the geochemical behaviour of Rn and, perhaps more importantly, a lack of sound experimental data. This chapter reviews the physical and chemical properties of Rn, analytical methods for Rn and the applications of Rn to mineral exploration. It also attempts to answer the two most frequently asked questions: (1) how much Rn emanates from a given source; and (2) how far does that Rn move? Finally it mentions problem areas for future research.
pi IYSICAI, AND CI II:~MICAI_,I)R()I)ERTIF~S()1: RAI)()N The journal Chemical Abstracts lists 61 radioactive isotopes of Rn starting with ~87Rn. Fortunately there are only three that are part of the natural radioactive decay series: 222Rn (radon) is produced by the 238U series; 2-~~ (thoron) is produced by the 232Th series; and 2~4Rn (actinon) is produced by the 235U series. Their noble gas configurations make them chemically inert at all conditions prevailing in the surficial environment. At ordinary temperatures Rn is a colourless gas. When cooled below its freezing point of-7 I"C, Rn exhibits a brilliant phosphorescence, which becomes yellow as the temperature is lowered and orange-red at the temperature of liquid air. It has been reported that, at high temperatures, Rn reacts with F and CI to form halides such as RnF2, RnF4, RnF~, and RnCI4. Radon, along with its analogues Ar, Kr and Xe, tbrms a crystalline hydrate, Rn.6H20, with a dissociation pressure of 1 atm at 0~ and which is held together entirely by Van der Waais forces. The main physical properties of Rn are listed in Table !!-!, and its radioactive properties and place in the 238U decay series are recorded in Table 1 1-II. Thoron (Tn), with a halt-life of 54.5 sec., plays only a minor role in mineral exploration and actinon (An), with a half-life of 3.92 sec., can be considered insignificant in this regard.
Radon
355
T A B L E 11-I Physical properties o f radon Property
Value
A t o m i c no.
86
A t o m i c mass
222.0175 4.5 x 10 .8 cm
A t o m i c diameter (estimated) H a l f life Decay constant
3.825 days 2.097 x 10 -6 sec -~
M o d e o f d e c a y and energy: -
alpha particle
5.486 M e V 9 9 % + 4.98 M e V 0 . 0 8 %
- g a m m a ray
0.510 M e V 0 . 0 8 %
Boiling point
-61.8~
Melting point
.71~
Density (gas)
9.73 g/L
Specific gravity: 4.4
-liquid -
solid First ionisation potential
4
l-leat o f vaporisation, c a l / g - m o l e
4325
10.8
ev
Solubility at 1 arm" -
in water at 0"C
5 I()ml~ (S'I'P/I,)
......
2()"C
23O
'
...... ......
4()"C 60"(7
139
"'
96
""
l)istribution coefficient, Rn (liquid)/Rn (gas)" - water at 0"C
-
0.51
....
20"C
0.27
....
40"C
0.16
....
6()"C
0.13
glycerine at 18"C
!.7
- kerosene, benzene, vaseline at 18~
10
- toluol, xylol, benzol at 18~
13
olive oil at 18"C Diffusion constant (cm2/scc) at room temperature:
28
-
- i n air
I x I 0 -I
- i n soil
n x I 0 3 - n x I 0 -2
- i n water
I.! x 10 -s
- in mud
n x ! 0 -~' n x 10 .2o
- i n Ba(NO3)2 (solid)
W. Dyck and I.R..lonasson
3 56 DEFINITIONS
9 The Becquerel (Bq) is the SI unit o f radioactivity. One Bq corresponds to one disintegration per second and is equal to 27 pCi. 9 One Curie (Ci) is the a m o u n t o f a radioactive substance in which 3.7 x 10 ~~ disintegrations occur per second. One picocurie (pCi) is equal to 10 -~2 Ci. One gram o f 226Ra disintegrates at a rate o f one Ci. 9 One Eman is equal to 10 ~~ Ci/L or 100 pCi/L or 3.7 Bq. 9
Emanation
power,
emanation
efficiency,
coefficient
of
emanation,
percent
emanation, and escape-to-production ratio all mean the same thing, n a m e l y the fraction o f Rn atoms formed in a solid that escape from the solid. 9 The Mache unit (ME) corresponds to the a m o u n t o f Rn per litre o f liquid or gas that produces an ionisation saturation current of l0 3 esu. The ME is encountered in older medical and hydrological reports; I ME = 364 pCi/L. 9 The W o r k i n g Level ( W L ) is equal to any combination o f the short-lived d e c a y products o f Rn in 1 L o f air resulting in the ultimate emission by them o f 1.3 x 105 MeV o f alpha-particle energy (obtained from 9800 Rn atoms or about 0.5 pCi/L). 9
Equivalent uranium (eU) is U concentration estimated from the 2~4Bi concentration (usually determined by g a m m a - r a y spectrometry) assuming secular equilibrium.
"I'ABI,I'~
I1-I!
l'ypc and cncrgy of dccay and rclativc radioactive cquilibrium concentrations of the (from l:ricdlandcr ct al.. 1964) i';lcmcnt
2.~stj 234-i,h
l~rincipal decay mode and energy (McV) 4.19 ().19 2.31
234pa "~34 j
>o.1,h 22~ 1
222Rn 2 8po 2
4pb
2 4po 20pb
2~176
0.59 1.51
0.053 0.61
0.015
0.046
1.16 5.31
4.51 x 10'>y 24.1 d I 18m 2 48 x l()Sy 7 52 x 104y
(irams
Atoms
I
2.53 x I() 2~
1.44 x I() l l
3.71 x I() l~
4.89 x I() re'
1.26 x I()r 1.39 x I() Iv
1.62 x I0 -s
4.23 x I() Ir
1 62 x l03 y
3.42 x I0 -v
9.11 x I() ~4
382d
2.13 1.17 1.02 7.50 1.01 4.32 2.69 7.43
5.80 3.22 2.87 2.11 2.85 1.24 7.71 2.13
26.8 m 19.7 m
1.6 x l0 -4 s 22y 5.01 d 138d Stable
scrics
l';quilibrium concentration relative to 238tj
5.40 x I0 s
3 05 in
7.69
20Bi
20po
0.053 ().()68 ().188
4.77 4.68 4.78 5.49 6.00
2 4Bi
0.045 0.029
1ialf-lifc
238[j
x x x x
I()12 I0 -15 I0 t4 10-15
x 10 "21 x IO "9
x 1012 x 1011
x x x x x x x x
i ()'~ I0 ~' 107 107 10~ 1()13 !0 v IO II
Radon
357
GEOCHEMISTRY OF RADON To treat Rn separately from its parents and daughters, as is being attempted here, is not strictly the optimal approach to a review nor, for that matter, to an exploration programme. Each element in the U decay series is unique in some respects and hence is best suited to certain conditions and to a certain phase in the exploration programme. Also there exists a virtually-inseparable link between Rn and Ra; this makes it mandatory to include the geochemistry of Ra in any discussion of Rn geochemistry. This strong link is, of course, the relatively short half-life of Rn (3.825 days) compared to that of Ra (1622 years). As an inert gas, Rn enters into very few chemical reactions, and those only at high temperatures and under rigorous chemical conditions. Radium, on the other hand, manifests chemical properties similar to the other alkaline-earth elements of Group IlA, to which it belongs. However, its low natural abundance (10 12 g/L in surface waters and 10~2 g/g in rocks) rarely allows Ra salts to reach solubility-product concentrations in natural waters. Therefore the important chemical reactions are those of adsorption to active surfaces of all kinds and coprecipitation with Ca and Ba salts. For example, radiobarite is a common secondary mineral near to some oxidising U deposits and it has recently been tound on the modern sea floor. Although Ra concentrations are so low in rocks, leaching tests have shown that Ra and its daughters are more concentrated in microfractures and along grain boundaries than within the matrix of rocks (Starik et al., 1966). Apparently the radiation damage caused by tile recoil of an atom when it undergoes alpha decay permits increased mobility of its daughters. During the process of weathering, this loose or labile Ra accumulates on the surfaces of clays and ill the matrix of decaying organic matter in soils. Vinogradov (1959) found that sandy soils are the poorest and clayey soils the richest ill Ra, with a range of Ra concentration from 0.1-3.8 pCi/g. Grey forest soils have three times more Ra in the subsoil than in upper horizons. Soils above carbonate rocks are richer ill Ra than the rocks themselves, with a marked relationship with Ba and SO42 but not with Ca, suggesting that the Ra is fixed in BaSO4 or in related colloids. Dciwiche (1958) studied the Rn emanation from a variety of soils from three continents, and tbund that red-brown soils, regardless of parent material, have greater Rn generating power in a subsoil rich in clay and concluded that this increased emanation comes fi'om Ra adsorbed on colloidal clay. Taskayev et al. (1978), studying soil profiles from podzolicgley soils with high natural radiation, tbund decreasing Ra content with depth ill direct proportion to the decreasing humus, Ca, and Mg contents of the soils. Leaching tests with various organic acids showed that only about 6% of the Ra could be removed from A horizon soils and I% from B horizon soils. Titayeva (1967) showed that under static and mildly acid conditions, U adsorbs better than Ra on peat. Under dynamic conditions U reaches saturation sooner than Ra, but Ra reaches higher equivalent values. Dynamic desorption of U and Ra from peat showed that U is washed out but Ra is not. In general the Ra distribution coefficient (moles Ra/g solid / moles Ra/mL solution) is 25-500 for soils, 150-300 for sand and 1500-3000 for peat (Sheppard, 1980). The transfer
358
W. Dyck and I.R. Jonasson
coefficient o f Ra in plants (Ra in plant ash / Ra in soil ash) is usually less than one but can go up to 25. Such a variety of environments and Ra sources leads to a large spectrum of Ra concentrations in soils and sediments and consequently to a great range of Rn concentrations. The main chemical reactions of Ra in the natural environment can be summarised with a few equations: Ra 2+ + Ba 2+ + Ca-(clay)= Ca 2+ + Ra-(clay)-Ba xCa z+ + (l-x)Ra 2+ + MCO3 = CaxRa(~.x)CO3 + M 2+ Ra 2+ + SO4 2"= RaSO4 RaZe+ 2Ci- = RaCI2 etc.
- adsorption
-
coprecipitation
- surface reaction - soluble species
The scavenging effects of clays, organic debris and hydrous oxides outweigh the dissolution of Ra by chloride in the surficial environment. As a result there is a net deficit of Ra in stream waters entering the oceans, even though the water in the oceans contains an excess of Ra (Cochran, 1979).
('oncentrutions o[rudon and radium in natural environments With very few exceptions, surface and near surface waters contain an excess of Rn (Table I !-I!I) compared to Ra (Table I I-IV) and U (Table I I-VI). This Rn must come from Ra in solids such as rocks, soils and sediments. The solubility product of Ra salts is seldom reached in natural waters, because invariably it is adsorbed onto sulphates and carbonates at the surfaces of rocks and minerals, in the zone of oxidation it is also coprecipitated by hydrous oxides of Fe and Mn. Only in the vicinity of strong sources of very saline waters do Ra concentrations rise to 10 ~~ or even 10 8 g/L. The link between Rn in water and Ra in sediments must be firmly implanted in the mind of the prospector, who could easily miss a deposit at the bottom of a large deep lake if the surface water is analysed for Rn only. Radon will not travel in water beyond 8 m by true diffusion, although mechanical agitation by stream turbulence or wind action on lakes can increase the Ra-Rn separation to 50-100 m. A second factor is the emanation efficiency of Rn from the solid materials through which the water moves; this seldom reaches 20% and is commonly a few percent. Even though Ra is highly immobile in the surficial environment, the law of dynamic equilibrium demands that some of it passes into solution. Since it has a relatively long half-life (1622 years), it can migrate considerable distances, perhaps several kilometres in well established aquifers, at concentrations below the detection limit of most analytical techniques, eventually accumulating by adsorption-desorption mechanisms on rock surfaces. Whilst most o f
359
Radon
TABLE 11-III Radon content of natural waters (pCi/L) Reference Andrews and Wood, 1972 Asikainen and Kahlos, 1979 Cherdyntsev, 1969 Dyck, 1978 Dyck et al., 1969 Dyck et al., 1971 Dyck et al., 1976a King et al., 1976 King et al., 1972 Korner and Rose, 1977 Lester, 1918 Polanski, 1965 Satterly and Elworthy, 1917 Smith et al., 1961 Smith and Dyck, 1969 Stoker and Kruger, 1975 Tokarev and Shcerbakov, 1 9 5 6
Groundwater 1400-2400 2600-55000 46-20500 355 (a)
Stream
Lake
510-38600 4 (a) 6-27 (g)
0.8-1.2 (g)
875 (a) 60-230 (a) 5-59000 1370 (a) Trace-30500 40-164000 11-640 0-260000 Ca. 1 (a) 38-340000 5000-5000000
(a) arithmetic mean or range of arithmetic means; (g) geometric mean or range of geometric means this Ra is adsorbed on surfaces at any one time, the Rn emitted by it enters the water phase easily. Thus the tap water in the town of Bancroft, Ontario, contains easily detectable amounts of Rn, due to surface accumulation of Ra, even though the lake water that is its source has no detectable Rn levels. Similarly, old domestic well casings, when logged with a gamma-ray probe, are found to have much higher activities than fresh holes drilled adjacent to them. Although the rate of weathering of U ores is low under reducing conditions, the solubility of Ra and Fe 2+ is comparatively high. Thus accumulations of Ra are particularly prominent in groundwaters from depth, where reducing conditions prevail. As these groundwaters enter the zone of oxidation, Fe and Mn oxides precipitate and adsorbed Ra is coprecipitated. Also deep waters usually contain large amounts of CO2 (and bicarbonate) which escape when the waters reach atmospheric pressure. This CO2 loss causes Ca and Mg carbonates to precipitate, again coprecipitating Ra with them, a phenomenon particularly evident in mineral springs (Felmlee and Cadigan, 1978). The Ra content of rocks, soils and stream sediments is summarised in Table l l-V. Idealised hydrogeochemical zonalities of Ra and other parameters in groundwaters free of organic matter are illustrated in Fig. 11-1. As groundwaters change from oxidising to reducing conditions the rate of weathering of U ores decreases but the production of Ra and Fe z+ increases.
W. Dyck and I.R. ,]onasson
360
I|YDROCIi[MICAL
lONf.$
COLUMN; tEVfl Of UNt)[RGROUND WA[
] G[OCH MICAL ] Z( I[S
[1| 1t01~ WAiIRS . W!ltl pH 6 1 8 . 5)
IN A SOLUilON
RA1 [ OF WLAIII[RING Of u OR[
RADIUM COttl[NI IN UNDI.RGIIOUNO WAIfRS .........
f [ ,2 CONI[NI iN UNUs WATfRS
N
Ul'Pf R lONt O[; WtAKLY MINIRALIIID Wkl'fR$
h h .
._
.
...........
M/AN /ON~
I
Of MOil[ MIN[RA! I / [ I ) WAI[H$
( t .15 gll) 4 ......................
tOW{~ /ONf
0~" fill;ill Y MINLRALIZ[O WAILR~ (
935 =till
"" .
"'
.
.
,il II II
.
.
1 'l
tl It II II
.
. . . .
4
II . . . . . .
v 1 .........
....
tl
II 11 . . . . . . . . .
..
Fig. I ! - I . Hydrogeochemical zonality of rocks devoid ol'organic matter.
"I'AI3LI'] I I-IV
l~,adium content of natural waters (pCi/l,) R c l'crc n cc Alcksccv ct al., ! 959 Andrews and Wood, 1972 Asikaincn and Kahlos, 1979 Baranov, 196 I Chcrdyntscv, 1969 [)yck, ! 974 l,~vans ct ai., 1938 l:clmlce and Cadigan, 1978 Khristianov and Korchuganov, i 971 King ctal., 1982 l~cstcr, 1918 Novikov and Kapov, 1965 l'olanski, 1965 Satterly and Elworthy, ! 9 i 7 Schutteikopf and Keifer, 1982 Scott and Barker, 1962 Smith et al., 1961 White et al., 1963
(; rou nd water
Stream
l,akc
( )ccan ().()1-().2
()-13.2 0. 1-8. I (a) 3-7480 7-3000
0.07 0.2-7.3
().()3-().47
(). 1-().20 0.04-0.14 Trace-46 <().()2-! .6 <0. ! -0.6 ()-709800 0-1.4 (a)
(a) arithmetic mean or range of arithmetic means
<0.03-1550 0.01-3.4
5
361
Radon
TABLE I I-V Radium content of rocks, soils and sediments (pCi/g) Reference
Rocks
Soils
Alekseev et ai., 1959 Baranov, 196 ! Dyck, 1974 Kuznetsov et ai., 1968 Levinson et al., 1978 Naumov et al., 1963 NCRPM, 1950 * Novikov and Kapov, 1965 Polanski, 1965 Rankama and Sahama, 1950 Schuttelkopf and Keifer, ! 982 Vinogradov, 1959
0.01-2 0.2-1.5 0.5-2.8 (g)
0.1-1 0.1 - 1
.......... Sediments .......... Stream Lake ()cean
0.7-0.8 (g)
0.9 (g) 2.7 (a) 3.8 (a)
0.8-1.1 0.5-2 0.1-1.2 0.8-12 0.009-1.4 I
0.2-560 0. !-2.9
* National Council on Radiation Protection and Measurements (a) arithmetic mean or range ot'arithnletic means; (g) geometric mean or range of geometric means
Novikov and Kapkov (1965) provide the following criteria which can help in deciding the significance of radioactive anomalies in groundwater. 9 A threetbid or greater increase in the content of Rn or Ra compared to the background of a region. 9 Occurrence of anomalous amounts of all fbur direct-indicator elements (Rn, Ra, U and He). 9 Increased content of trace indicators such as Mo, Pb, Cu, Zn, As, P, V, Ni and F. 9 A sharp rise in the Rn concentration after a rain or thaw period of up to ten times normal levels in the presence of U ore; not more than a tburfold rise above natural levels in the absence of U deposits. Waters from acidic rocks are usually more radioactive than waters circulating in basic rocks. Waters with extensive circulation and high flow rates are weakly radioactive. Groundwaters with a limited circulation tend to become mineralised and may become strongly radioactive in acidic rocks enriched in U. in mountainous areas, waters near the peaks are c o m m o n l y weakly radioactive but springs at the foot of the mountains may be highly radioactive, even in the absence of U ore. Usually, some mineralisation is necessary to produce highly-radioactive waters, such as secondary mineralisation in the fractures through which the water moves.
362
W. Dyck and I.R. .lonasson
TABLE I I-VI Uranium content of natural waters (ppb) Reference Alekseev et al., 1959 Baranov, 1961 Germanov, 1963 Naumov et al., 1963 Novikov and Kapov, 1965 Polanski, 1965 Starik et al., 1963
Groundwater
0.03-700 5-50
Stream 0.3 1 0.03-50 0. !-5.2
0.02-120
0.01-47
Lake
0.03-1000 0.2-5 8
Ocean 0.3-6.5 1.3-2.2 0.4-3.7 0.1-5.2 0.6-2 !-3 !.3-3
The widespread measurement of the radioactivity of waters in prospecting as practised in Eastern Europe (Baranov, 1961; Novikov and Kapkov, 1965) only slowly found similar acceptance in North America. No doubt the complexity of the method and the large variations in Rn levels, largely due to its gaseous nature and short half-life, have contributed to this reluctance to use the method (Rogers, 1958; Makkaveev, 1960; Smith et al., 1961). Nevertheless some North American workers were in the forefront of attempts to apply it to U prospecting. Lester (1918) could perhaps be credited for defining the position of U roll-fronts in Colorado, without knowing it. During a survey of the radioactivity of some 200 springs, he concluded that, "generally speaking, the most (radio-)active springs are found on both slopes of the Continental Divide and not far from it". More recent maps (Fisher, 1974) of the known U roll-fronts in Colorado and Wyoming suggest a similar conclusion with respect to these fronts. Senftle (1946), while engaged in exploration work near Great Bear Lake, Canada, found a clear positive relation between conductivity, total dissolved solids, Ra content of lake waters and degree of mineralisation. Perhaps the main reasons for the seemingly confusing pictures of the Rn-Ra method is the failure to recognise and make allowance for disequilibrium in the U series and weigh anomalies against the typical background levels of an area. Radon backgrounds vary considerably and depend not only on the amount of Ra in the vicinity but also on the emanation efficiency of the solid. However, the following rough guide may be useful: water and soil from granitic terrane usually contain more Rn than water and soil from sediments such as sandstone; and subsurface water contains more Rn than does surface water. As a very rough guide, background Rn levels are as follows: 9 in lakes, 0-5 pCi/L; 9 in streams, 5-100 pCi/L; 9 in wells and springs, 100-1000 pCi/L.
363
Radon
Any deviations by a factor of ten or more from the above may indicate uranium mineralisation or accumulation of Ra due to earlier chemical leaching, transport and redeposition. Springs and drill holes near uraniferous pegmatite in the Ottawa and Bancroft areas of Ontario contain from several thousand to several hundred thousand pCi/L Rn (Dyck and Jonasson, 1977). Such high Rn levels point to high Ra concentrations, or high Rn emanation efficiencies, or both. By contrast, Rn values of more than 10 pCi/L in lake-bottom water above the Athabasca sandstone in Saskatchewan are significant in terms of uranium mineralisation. Typical soil-air Rn concentrations at 25 cm are about 100 pCi/L and air just above the ground contains about 0.1 pCi/L. However, the type of uranium mineralisation, the concentration of U and Ra, and the porosity of the soil also affect the Rn content of soil air. Meteorological factors such as frequency and amounts of rain, barometric pressure and wind also influence Rn. Zverev et al. (1980) give average contents of Rn, Ra and U in natural waters (Table 1 I-VII).
"FABLE 1I-Vii Comparison of radon, radium and uranium contents of natural waters (Zverev et al., 1980) Waters
"i'ype
Surface water
l,akcs Strcams Stagnant Circulating
Sedimcntary rocks Basic magmatic rocks Acid magmatic rocks Uranium deposits
Stagnant Circulating Stagnant Circulating
Rn (pCi/l,)
R.a (pCi/l,)
I I (ppb)
< ! 00 < 100 340 890 890 16 ! 0 2 ! 20 50000 200?
0.5 (). 1 500 ! 0.5 4 I I O0 50
5 0.8 0.2 5.0 4 4 8 6 200
The above results are a useful guide and reflect the geochemistry of the three elements very well. For example, in the oxidising or surface environment lakes have higher Ra and U values than streams. The lake values are higher than those commonly encountered in Canada; most likely they reflect a large proportion of lakes from arid regions. Values in groundwaters in rocks exhibit similar patterns. Stagnant or very slowly-moving groundwaters contain relatively more Ra and less U and Rn than actively-circulating waters. This is so because circulating waters will carry dissolved Ra and U into aquifers where Ra eventually is adsorbed on the walls, thus increasing the Ra content on solids and Rn emanating into the water. Uranium, on the other hand, remains in solution provided the environment is not strongly reducing. Usually stronglycirculating waters are oxidising, enhancing the oxidation and dissolution of U. The
364
W. Dyck and I.R. Jonasson
400
Murho
23~
300
g
r-o C)
........ 'IIIN)!\.<.., 2 .x.\... ' Fla\" ~ ....- ...."-.. ru
~~
I00
~ - ~
~
o
I I_L[I . . . . . . . . I. . . . . I
..~~.
j
-~
"'.....-z...:~. - ~ " . ........
"'--~_33 ~ u 0
"-----./-"--- - /
J__L_LI_LII_I
.~...~....I
I L__LA_.J.L LLI....... L....1....II I_11I.
Io
10o
1o o o
Size ( / ~ m )
"~..........~
.,,,
,-a
210 p ~
"\
\-,
."~'~,. 9 .. \ -
\'~.
._
.~ 5 ~-
"-.,,. ~
t"lokko ',.226 -." ~a . .\
' . \"'.\
238b~- . . , \
Q)
i "',..
.~
. . . . -,. \
t
9. . ~
i-
z f-'"
.
(-)
N
....--...~~ o - - L - l - i l l .... o l
1
I
I ! IJ_.lll
I
10
_l.._l.l
L ~ _ _ _ L 1 O0
S i z e (/~ m )
l:ig. 11-2. Relationship between concentration of 23s(1 (234Th), 2~o.1.h and 22('Ra in soils from Murho and Rokko, Japan (reproduced with permission from Mcgumi and Mamuro, 1977, .I. (icophys. Rcs., 82" 3 53-356, copyright by the American (;cophysical (Jnion).
quoted Rn value of 200 pCi/L for circulating waters in U deposits appears to be in error; 200,000 pCi/L would be a more likely value .judging from the experience of the authors. The intensity of groundwater circulation plays a most important role in the concentration of these elements and must be taken into consideration when evaluating the significance of an anomaly.
Radon
365
Disequilibrium in the uranium decay series Disequilibrium in the U decay series has been studied for many years and is observed in almost all surficial environments. Rosholt (1959) measured U, 234pa, 23~ 226Ra, 222Rn and 21~ in a large number of U ores and offered explanations for six types of disequilibrium. Davy (1974) outlines the geological significance of radioactive disequilibrium situations with respect to exploration. Levinson and Coetzee (1978) discuss possible mechanisms by which separations of parent and daughter elements are effected and suggest pitfalls that may be encountered in the use of radiometric techniques when disequilibrium has occurred. Generally, it is the oxyphile and easily complexed parent, 238U (with its associated isotopes 235U and 234U), that is leached from rocks and soils at a faster rate than are the daughters, 23~ 226Ra, 2~~ and 2~ This mobilised U is adsorbed on sediments in streams and lakes or enters groundwaters, where it is eventually precipitated under reducing conditions, occasionally in ore grade concentrations. Surface leaching and adsorption has been observed by numerous workers including Richardson (1964) and Ostrihansky (1976). Richardson (1964), Nash (1979) and Charbonneau et al. (1981) show that U and 214Bi (eU) in surface rocks are essentially in equilibrium. Charbonneau et al. (1981) also found evidence for leaching of surface rocks, depleting both U and 214Bi, without disturbing radioactive equilibrium significantly. Morse (1970) found that weathered rocks and A-horizon soils from Bancroft, Ontario, contained considerably more Ra than that needed to satisfy radioactive equilibrium consistent with the U content. Megumi and Mamuro (1977) give excellent illustrations of disequilibrium in the U series in soils from Murho and Rokko, Japan (Fig. 11-2). The plot of particle size versus concentrations clearly shows that the decay products remain near the site of weathering and adsorb on surfaces much more readily than does the parent U. Dyck and Boyle (1980) found that the eU/U ratio in plants rooted in mineralised ground in the Goldfields districts of northern Saskatchewan shifts from "greater than one to less than one" for those rooted in background soils. A similar shift was observed in the transition from rocks to stream sediments and lake sediments in the Beaverlodge area of Saskatchewan. Results of a Rn-Ra-U survey of lake bottoms in the Key Lake district of Saskatchewan showed that anomalous hydromorphic U from U mineralisation is paralleled by anomalous Rn and Ra, even though there was much less Ra in the lake bottom sediments than one would expect from the amount of U present (Dyck and Boyle, 1980; Fig. 11-3). Levinson et al. (1978) observed that organic-rich lake sediments from the vicinity of the Key Lake U deposit contained 18 times as much U as one would infer from the Ra content. Similarly, borehole samples from southern Africa (Levinson and Bland, 1978) had excess chemical U, whereas soils from Thailand contained excess radiometric U. Syromyatnikov et al. (1967) describe a gamma-ray anomaly over decomposed rhyolites which contained excess 23~ and 22~Ra. Nikiforova and Yufa (1970) give a mean eU/U ratio of 1.16 for 520 different soil samples, and 1.03 for 73 rocks, with the following
W. Dyck and I.R. Jonasson
366 2000 1000 QS
SO0
QD
iw Q|
IOO_-
_ -
~-
/
QM
-
-| |
-/ -
/
/
/
/QZ
@Z*
/
oK
-
-
/
/
/
/
/
e M zb e z / / / / / -
C)Z QZ
QS QZ
E)KE
200
M*
@s
/
/
/
/
/
/
/
eU, pom 2
5 C)K,,
I0
20
i
i
Li~,l
50
1O0
l
l
200
l
I
I 11,11 SO0 1000
Fig. 1-3. Relationship between U and eU in lake-bottom sediments from the Key Lake district, Saskatchewan: D- Dieter Lake, Z- Zimmer Lake, S- Sea Horse Lake, M- McDonald Lake, K- Key Lake, KE- Karl-Ernst Lake, *Samples with less than 19% loss-on-ignition (reproduced with permission of the Canadian Institute of Mining, Metallurgy and Petroleum from Dyck and Boyle, 1980).
breakdown by type: steppes = 1.0-1.75; forest-steppes - 1.1-3.0; taiga soil and subsoil = 2.29-2.93; superaquatic soils and peats -- 0.2-0.4. They also found that loose soils have higher eU/U ratios and that the largest variations, from 0.2-7, occur in taiga landscapes. Asikainen (1981) found that 310 drilled well-water samples from bedrock in Finland showed a high state of radioactive disequilibrium. Activity ratios of 238U/226Ra usually fell in the range 1-20 and the activity ratios of 238U/222Rn in the range (1-20) x 10-4. In the prospecting history of the Blind River- Elliot Lake discoveries in pyritiferous quartz-pebble conglomerates, one finds remarks about the lack of U in assays of surface samples, despite a strong radioactive response (Lang et al., 1962). Phair and Gottfried (1964) relate that "hot spots" on mine dumps in Colorado contained 150 times as much Ra as that required to satisfy the secular equilibrium corresponding to the amount of U present. Hansink (1976) found that oxidised U mineralisation in a roll-front deposit in Wyoming is depleted and that the mineralisation in the zone of reduction is enriched in chemical U; based on this data he claims that there has been migration of daughter products as well as parents. Using U-Pb isotope systematics, Stuckless and Nkomo (1978) have shown that extensive U loss from the granites of the Granite Mountains, Wyoming, took place during several erosional cycles. These authors also refer to eight
Radon
367
other publications that document loss of U from igneous rocks under near-surface processes of weathering and erosion. We may conclude from the above that U is the most mobile element in the U decay series in the surficial environment. Its removal from one site and deposition in another leads to radioactive disequilibrium over time spans of several half-lives of the respective daughter products. Concomitantly, decay and growth of daughter products, depending on the abundance of the parent, tend to re-establish equilibrium, given sufficient time.
Emanation and mobility of radon The answers to the questions "how much Rn emanates from a given source" and "how far does that Rn move", which are crucial in the evaluation of the significance of Rn survey results, are generally to be found in studies of the emanation efficiencies of particular media and the rates of diffusion of Rn in particular media. For the great majority of cases the approximate answers are simple: on the average about i% of the total Rn escapes into the voids of rocks and about 20% from loose soil; and from diffusion experiments and field tests, the range of Rn, i.e., the average distance a Rn atom moves from its source before it decays is approximately 8 m in static media. However, nature provides tbr an infinite variety of natural settings. Lateral and vertical groundwater movements displace Rn anomalies accordingly, whilst the density, coarseness and moisture content of soils, degree of fracturing of rocks, temperature, wind and barometric pressure influence the emanation rate and range of Rn. During a tour of the U deposits of Wyoming and Colorado, one of the authors (Dyck) was astonished to hear that no Rn or other geochemical-logging techniques were carried out during systematic drilling tbr U. As will be shown elsewhere, the Rn technique gives greater sensitivity, but knowledge of the theoretical and practical capabilities of the method is essential. Most importantly, a knowledge of when, where and how to sample for Rn improves the chances of detection markedly. To help in this regard, some knowledge of the emanation power and mobility of Rn in a medium is essential. The escape of Rn from solids has been studied by numerous workers, whose results are summarised in Tables 1 I-VIII to I I-XII. Typical values and ranges are illustrated in Fig. 11-4. Various terms, such as emanation power, emanation coefficient, emanation efficiency, Rn leakage, etc. (see definitions) have been applied to this phenomenon; all refer to the fraction or percentage of total generated Rn that escapes from a solid. Given a crustal activity of Ra of about 10~2 pCi/g, the crustal activity of Rn is fractionally lower because of the escape of some Rn from the upper 8 m into the atmosphere. This escaping fraction (4.25 x 20 -s pCi/cm2.sec, according to Wiikening et al., 1972) is indeed a minute amount, being 133 g or 43 Ci of Rn mixed in approxiately 57 x 10 ~s tonnes of air. Yet because of its specific radioactivity and half-life it can be measured with appropriate instruments. Early investigators were surprised not only to find radioactivity in the air but also to observe large variations in the intensity (Delwiche, 1958).
368
W. Dyck and I.R. ,lonasson
Fig. !1-4. Radon escape from different types of rocks.
Quantitative determination of Rn in atmospheric air near ground level (ca. l m above the ground) in various parts of the world by a number of scientists gives values ranging from 0.01-0.6 pCi/L in natural settings with mean values averaged over several weeks of 0.10.3 pCi/L. While Tit gives similar concentrations, it usually does not rise above 10 m because of its short half-life of 54 sec. Typical exponential attenuation of Rn with elevation is illustrated in Fig. !1-5. However, depending on atmospheric conditions and moisture content of soils many locations near ground level can reach several pCi/L. Such accumulations occur during quiet, cool, clear summer nights alter warm dry days. According to some studies, enclosures such as buildings and basements can accumulate relatively high Rn levels (occasionally several 100 pCi/L). However, the great majority of studies show conclusively that near-ground air contains, on average, less than i pCi/L, but somewhat more titan 0. ! pCi/L. Virtually all generated Rn decays within the crystal lattice of the solid or the pore spaces within the solid. The escape of Rn from solids occurs mainly through direct recoil, indirect recoil and diffusion. Tanner (1978, 1980) treats these mechanisms in some detail but finds none can account tbr emanation powers much greater than a few percent. He comes to the conclusion that "any appreciable emanation of Rn atoms comes from Ra distributed in secondary crusts or films or in the shallow surface layers of intact crystals of the host minerals". Inside a mineral grain a recoiling Rn atom has a range of 20-70 nm, or roughly 200-700 layers of Si atoms. Leaching experiments convinced Starik (1959) that successive recoils and natural weathering have caused Ra to coat surfaces of capillaries and inter-granular boundaries from where Rn can escape much more efficiently into the soils.
369
Radon
4O 36
-/f•16•Metefs t -.~~..M.eters
32
2s 24
til l
"5 20 E
~ t2 8
~
/ -/
,~
\
1Cm.
Surfaceto
//
/
4
0
9
! 0.2
0
'.
Radon-222
I 0.4
I
J 0.6
I
~---~ 0.8
, 1.0
Concentrarion in tO .~2 Curies Liter
~
I
1.2
1.3
.t
28 24
Meters
f ~ 6
20
f '"X\
ou
/,
c,,)
"1,"
\4
/I
,,,
,,I M ~ , :"
t//// ILV /
X.%L
\'..,-
_.. ....... ..........
0 ~ - - - - [ - -
0
Meters
0.2
-
~-~ I
..L___I_
0.4
......
_J.~_
0,6
..........................
~- .... . I
Radol~-222 Conce~tratJonln
-............. I ........
0.8
L
I
l,O
I ....
I
1.2
.__
1.3
10 -12 ( ' u r i e s l i t e r l
l:ig. 11-5. l:rcqucncy distribution of Rn in air at two locations as a function of elevation above ground.
It is obvious from the data of Tables I I-Vlll to I I-XII that the emanation power varies widely. As a rule, the finer tile particle size the greater the emanation power. However, when samples are severely weathered or damaged, as in the case of radioactive ores, further comminution of a specimen has little effect on the emanation power. In both cases it is believed that Ra has migrated from the original site of its parent in the crystal lattice and has adsorbed on the surfaces of microffactures facilitating Rn escape. When certain samples are pulverised in the laboratory they may even emit less Rn than the whole lump. This can happen when U and its decay products are associated with minerals that do not pulverise as easily, i.e., are less friable. Starik and Melikova (1957) found that a whole lump of ore had an emanation power of 46%
370
W. Dyck and I.R. Jonasson
and the powdered lump 41%; they think that in this case the Ra had leached out o f the mineral and adsorbed on the surfaces of microfractures and grains of the rock. Crushing increased the surface area available for retentive adsorption.
TABLE 1 I-VIII Emanation power (%) of uranium ores in air at ambicnt temperature Reference
Type
Austin, i 975
Sandstone, Colorado (i) Sandstone, New Mexico (!) Sandstone, Utah (1) Sandstone, Wyoming (l) Primary and secondary ore (p) Sandstone with kasolite (1) Sandstone with kasolite (p) Silicified coaly shales (I) Silicified coaly shales (p) Silicified coaly shales (p) Albite with iron (p) Alumino-silicates (p) Alumino-silicates, carbonates (p) (:lay-sandstone-organics (p) Clays with sandstones (p) l,imestones (p) Quartz-sulphide (p) Sandstones with coal (p) Shale-clay-coal-carbonate-sulphide (p) ()rcs with secondary minerals (1) Carnotitic sandstone l)ictyonemic shale (;ranitc with autunite Granite with torbemite (;ranitc with uranium, black l~imestone with carnotitc Limestone with uranium, black Siderite with nasturan Uranium-bearing coal l,Jranium-bcaring limonite
Novikov and Kapkov, 1965
l'rutkina and Shashkin, ! 967
Starik and Melikova, 1957 "l'okarev and Sllcherbakov, 1956
(1) lump or bulk solid; (p) powder or material passing fine sieve
Emanation Min Max <1 47 2 56
47
30 2
65 59
0.2 4 22 2 2 I 37 2
20 92 7() 74 63 32 91 70
Av 15 27 10 37 17 18 41 5 18 32 9 31 13 37 32 14 20 26 16 34 24 14 44 44 34 26 8
76 16
371
Radon
TABLE 1 l-IX Emanation power (%) of rocks in air at ambient temperature Reference Andrews and Wood, 1972
Barretto, 1975
Tokarev and Shcherbakov, 1956
Type Limestones (l-p) Sand (p) Sandstones (l-p) Basaits (p) Conglomerates (p) Gneiss (p) Granites (p) Granodiorites (p) Laterites (p) Limestones (p) Quartz-diorites (p) Quartzites (p) Sandstones (p) Shales (p) Argillites Gneiss and gneissic granites Granites and granodiorites Granitic pegmatites Limestones Marls Quartz porphyries Quartzites Sandstones Shales Trachyliparites
Emanation Max Av 13 4 26 16 14 11 9 6 26 13 14 8 15 7 40 20 3 3 2 2 8 6 11 6 12 7 8 5 22 20 27 22 5 24 17 28 1 25 I1 13 4 5 5 13 30 21 1 12 !1 18 6 15
Min 1 6 3 3 2 1 3 4 3 1 5 2 3 3
(1) lump or bulk solid; (p) powder or material passing fine sieve Andrews and Wood (1972) found that Rn release from rock into water was inversely proportional to the square root of the rock particle diameter for Midford Sand and Carboniferous Limestone and independent of particle diameter for Old Red Sandstone; Giletti and Kulp (1955) had similar results. For example, one single 35 g piece of pitchblende gave 3.0% Rn leakage and the same piece powdered to <200 mesh gave 3.8% Rn leakage. From the great variation in observed emanations regardless of particle size and the general findings of many studies, we may conclude that the degree of weathering, nature of weathering products and microfracturing of a rock or soil are the controlling factors of emanation power, although recoil and diffusion are the mechanisms of emanation.
372
W. Dyck and I.R. Jonasson
TABLE 11-X Emanation power (%) of minerals in air at ambient temperature Reference
Type
Emanation Max Av 0.3 0.8 2.8 6.6 3.0 4.0 0.2 0.2 4.7 0.1 0.2 4.8 2.6 8 4.4 2.3 12 4.1 (I. I 2. () ().4 0.6 3.4 1.6 0.4 46 1.7 10 37 21
Min Barretto, 1975
Cherdyntsev, 1961
(iiletti and Kulp. 1955
! laissinsky, 1964 l'rutkina and Shashkin, 1 9 6 7 Starik and Mclikova, 1957
Allanite (p) Apatite (p) Biotite (p) Glauconite (p) Magnetite (p) Monazite (p) Sphene (p) Xenotime (p) Zircon (p) Cassiterite llmcnite Monazite Sphene "l'antalitc-colunlbite 'l'ungstales (Jranitc Autunite (I) l~,ranncritc (1) (_?arnotitc (I) Pitchblende (I-p) %amarskitc (1) [Jraninite (!) Zircon (!) Calcite Aragonitc Monazite l~itchblcndc Autunitc (I-p) Carnotitc (l-p) Khlopinite (l-p) 'l'orbcrnitc (i)
(i) lump or bulk solid: (p) powder or material passing tine sieve
().1)4
O.6
17 ().I 0.03
27 16.6 ().()3
0.6 !.6
().8 92
5.9 6.0 22 3.5 ().()3
0.7 2 3
0.2
().3
6 32 0.07 12
7 33 ().2 17
().3 1.3 6 32 ().I 15
Radon
373
The concentration of Rn in soil air at infinite depth is given by: (Rn) - e.d.(Ra). 1000 / p where, (Rn) -- concentration of Rn in soil air in pCi/L, (Ra) = concentration of Ra in soil in pCi/g, e = emanation coefficient, d = density of soil in g/cm 3 and p - porosity of soil = (volume of air + volume of water) / total volume Infinite depth is the depth at which escape of Rn into the atmosphere is negligible (>8 m). A complex mathematical treatment of Rn emanation as a function of porosity, Ra concentration, specific gravity and diffusion is given by Baranov (1961). For a typical soil, (Ra) = I pCi/g, e = 20% = 0.2, d = 1.4 g/cm ~ and p = 0.4; hence (Rn) -- 700 pCi/L. Nearly all measured soil Rn values tall below this value because measurements are usually carried out at depths much less than infinite depth. The pronounced effect that depth has on the measured Rn concentration and the precision of measurements in routine Rn surveys is described below. Moisture content of soils also affects the Rn content of soil air. The above expression for soil-air Rn content needs modification. As tile voids fill with water, thctors such as changes of density, solubility of Rn in water and changes ill emanation efficiency come into play. Prutkina and Shashkin (1967) tbund that about twice as much Rn escapes from 150-mesh U ore samples when they are submerged in water than when they are in air. Monazites emanate 10-20% more Rn into water than into air. These authors suggest that it is mainly a surface effect, water molecules displacing and dissolving adsorbed Rn atoms. Similar tests with kerosene and alcohol gave emanation coefficients between those for air and water. The average emanation coefficients fbr nine different ores are given in Table I I-XI. Delwiche (1958) reports 2.4 times as much Rn release from soil in boiling water than the same soil in dry air. Austin (1975) found that moisture content of 10-80% saturation in U ore had little effect on the emanation power of U ore. Age, stratigraphy and U mineral species have no clear effect on emanation power but emanations from ore are almost always greater than those from the pure minerals. Porosity and permeability do have a control on the emanation power. Similarly increase in temperature from -80~ to 300"C increases emanation of Rn (Giletti and Kuip, 1955; Barretto, 1975). However, specimens annealed at higher temperatures (300-1200"C) show a progressive decrease in Rn loss probably because of a reduction of effective surface area. it is somewhat surprising that so few Rn emanation power studies have been carried out on soils. The study by Delwiche (1958) on Rn release from "Great World Soil Groups" gives relative emanation powers for all but one soil. His results show clearly that the more highly weathered soils emanate more Rn and that the clay intervals emanate more than the silty and sandy intervals. Red podzolic, lateritic podzolic and solodic soils emanate the most and desert soils and western brown tbrest soils the least. Quantified emanation powers tbr soils and stream sediments are shown in Table I I-XII.
W. Dyck and I.R. Jonasson
374
Barretto et al. (1972) found that the emanation powers of 19 samples of various soil types were in the range 10-56%. The results of Delwiche (1958) and Bonotto and Andrews (1999) on other soil types lie at the lower end of this range. Korner and Rose (1977) report an average emanation efficiency of 42% for five stream-sediment samples.
TABLE l l-XI Average radon emanation coefficients in different media for nine different ores Med ium Mean percentage Standard deviation
Air 22 10
Water 39 13
K erosen e 28 9
A !coh o I 28 9
The following general conclusions about the Rn emanation power of natural materials can be drawn. 9 Alpha particle recoil and diffusion are the processes responsible for Rn emanation from solids. 9 Radium adsorbed on surfaces of clay-silt particles and capillaries and microfractures of rocks is a significant source of Rn in soil gases, U mines and groundwaters. 9 Increasing temperature and moisture content increases the emanation power of soils and rocks but may decrease the Rn content of pore gas under certain conditions. 9
A clear direct dependence of Rn emanation rate, but not of emanation power, on U concentration is evident. 9 The most highly weathered, the finest grained, or most porous samples emanate the most Rn. 9 Typical emanation coefficients range from 1-42%. 9 The measurable range of emanation power for any one type of material is great.
Tanner's reviews on the migration of Rn in the ground (Tanner, 1964a, 1978, 1980) contain a detailed description of the processes responsible for Rn emanation and migration and an excellent summary of diffusion studies and coefficients; therefore this topic will not be discussed here in detail. The emanation processes of recoil and diffusion release Rn into pore spaces of minerals and rocks, permitting diffusion and mechanical transport to move Rn away from its source. Typical diffusion constants of Rn in various media are listed in Table 11-I. From the many diffusion studies carried out it is evident that diffusion alone cannot move Rn beyond about 8 m within the lifetime of a Rn atom. A simple approximation of the distance-time relationship is given by S 2 = pDt, where S - distance, D = diffusion coefficient and t = time. Thus, assuming a typical diffusion constant of 0.05 cmZ/sec., after six half-lives or 23 days, only 1.5% of a Rn
375
Radon
TABLE 1I-XII Emanation power (%) of soils and sediments in air at ambient temperature Reference Barretto, et al., 1 9 7 2
Bonotto and Andrews, 1 9 9 9 Delwiche, 1958 Korner and Rose, 1977
Type Alluvium-elluvium soils (p) Calcareous soils (p) Clayey soils (p) Granitic soils (p) Lignitic soils (p) Sandy soils (p) Volcanic soils (p) Silty-clay to stony loam soils (p) Brown tbrest soil Stream sediments
Min 18 18 19 37 36 10 47 I0 29
Emanation Max Av 48 56 36 46 20 20 56
10 42
sample has survived and has moved about 5.6 m from its source. The 8-metre range quoted above is obviously an optimistic figure considering the large variations in Rn content that are encountered at a typical field site. A persistent Rn signal 10% above background is a more realistic figure given a time of about 13 days and a distance of 4 m for a detection range in a typical soil. A number of equations, tables and graphs have been prepared by various authors relating the Rn content of soil, the diffusion coefficient and depth of source (Baranov, 1961; Alekseev, et al., 1959; Novikov and Kapkov, 1965; Clements, 1974; Soonawala, 1976). All assume diffusion to be the mode of transport of Rn and provide solutions to Fick's law of diffusion assuming various boundary conditions. The transport equation developed by Clements (1974) incorporates vertical flow of soil gas and was verified by him experimentally. Three solutions to the equation are illustrated in Fig. 11-6. The values chosen for the soil-gas flow are maxima that can be expected from atmospheric pressure variations. When the barometric pressure drops, soil gas flows upward resulting in a rise in the Rn concentration at depth less than "infinite" relative to zero flow-rate conditions. Similarly Rn concentrations fall when atmospheric air flows into the soil, i.e., when barometric pressure rises. There is a rapid change in the Rn concentration at shallow depth, particularly during periods when the barometric pressure is falling. Field survey crews should therefore pay particular attention to depth of sample sites and barometric pressure changes in order to improve precision. Soonawala (1976) used a computer-adapted finite difference method to solve the diffusion equation for Rn in two and three dimensions. Using theoretical and laboratory studies he was able to explain satisfactorily field emanation data from the Eldorado area of Saskatchewan and the Kaipokok Bay area of Labrador with the diffusion model of Rn transport. The results shown in Table I I-XIII are taken from Novikov and Kapkov (1965). To interpret them the reader must visualise an inactive soil layer of thickness h
3 76
W. Dyck and I.R. Jonasson
on top o f an active or R n - e m a n a t i n g layer o f infinite thickness and a s a m p l i n g point x metres above the interface o f the two layers. T w o cases are p r e s e n t e d in T a b l e 11-XIII, in one h is infinitely thick and in the other 3 m thick. For each case the distances, x, are given for six different bulk diffusion constants and for different ratios o f the R n concentration N / N 0, where N = Rn concentration at x and N O = Rn concentration in the active layer. For example, if the bulk diffusion constant is 0.05 cmVsec., 10% o f the original Rn content can be detected 3.54 m or 2.48 m above a radioactive layer, depending on whether the inactive layer is infinitely thick or 3 m thick, respectively.
TABLE 1 I-XIII Upward penetration distance (in metres) of radon from a buried source in terms of bulk diffusion coefficient (D*), fraction of radon detectable (N/No) and thickness of inactive cover above the source (from Novikov and Kapkov, 1965) Thickness of cover
3 metres
D* (= D/n) 5x 1x 5x 2x Ix 5x 2x 5x ! x 5x 2x !x 5x 2x
10-z 10.2 10-~ 10-~ 10-3 10.4 10.4 10-2 10-2 10.3 10-~ 10.3 10.4 10-4
N/N0 0.5 1.06 0.48 0.34 0.21 0.15 0.11 0.07 0.99 0.48 0.32 0.21 0.15 0.11 0.07
0.1 3.54 !.59 1.12 0.71 0.50 0.35 0.22 2.48 1.60 1. ! 2 0.71 0.50 0.35 0.22
0.01 7.08 3.18 2.24 1.42 1.00 0.71 0.44 2.95 2.74 2.22 1.42 1.00 0.71 0.44
0.001 10.62 4.77 3.36 2.13 !.51 1.06 0.66 2.99 2.97 2.89 2.13 1.50 1.06 0.66
n -- porosity = 0.4
O n e - d i m e n s i o n a l and multiple-layer models have been d e v e l o p e d to simulate the transportation o f gaseous 222Rn away from a u r a n i u m ore source (Jeter et al., 1977; Jeter, 1980). Calculations suggest that a 1 m thick planar ore deposit (0.6% U308) located 100 m b e l o w surface and with 20% emanation efficiency should be detectable within 15 m with a signal to b a c k g r o u n d ratio o f 5:1. Experimental results, using 2t~ confirm m o d e l calculations, which can be improved further if migration o f 226Ra to the top o f the water table is taken into account.
377
Radon
V T =3"3X10"4 Cm.S-1
2
V=O
V~,=3.3X10 "4 cm S"1 4--
6-
O =0.030 c m 2 s" I 9 =0.40
A -2.1X10 "6 s "I
0
0.5
1.0
C/C~=
l:ig. 1 I-6. Concentration of Rn as a function of depth from surface (from Clcmcnts, 1974).
There exists a handful of documented cases where the Rn flux in the soil could not be explained by diffusionai transport. Perhaps the most striking of these is the one reported by Gabelman (1974) from a detailed Rn survey across the Starks Salt Dome, Louisiana. During a period of 19 months and when 24 hour, 6 hour and 2.5 hour cycles were being observed, on the morning of 17 December 1971 the Rn content at one station unaccountably rose from about 200 pCi/L to 2140 pCi/L in three hours and then fell as rapidly to 50 pCi/L. One possible explanation may be the influence of an unnoticed seismic event. Alekseev et al. (1959) were able to explain most of the observed Rn concentrations using the diffusionai transport model but in a few cases Rn penetration was found to be larger than expected and attributed to convective transfer of Rn.
378
W. Dyck and I.R. Jonasson
Seismic events can and do alter the Rn flux, particularly along faults and fracture systems. Mogro-Campero et al. (1980) recorded a tenfold increase in the Rn flux at a site in the Blue Mountain Lake area of New York State following a pronounced increase in earthquake activity. Other investigators have made similar, though less dramatic observations (King, 1978, 1980a; Talwani et al., 1980; Teng, 1980; Birchard and Libby, 1980) during earthquake activity. To explain such extraordinary variations in the Rn content of soil gases, Mogro-Campero and Fleisher (1976) have, on the basis of current theories, postulated fluid convection in the Earth driven by local thermal gradients in areas of relatively high permeability as the means of moving Rn over distances greater than 8 m. Kristiansson and Malmqvist (1981) postulate an upward flow of geogas carrying Rn with it in order to explain their field data. However attractive and sensible these convective theories are, data in support of them usually lack one vital analytical component, namely the Ra content of the soils in which the Rn content was measured. Alekseev et al. (1959) report that "thanks to Ra aureoles, it has become possible in individual cases to reveal mineralisation located at considerable depth by means of emanation surveys". The theory of seismic pumping proposed by Panov and co-workers at the Donetzk Polytechnical Institute (pets. commun., 1981; Panov et al., 1980) attributes anomalous Rn occurrences over fault zones to micropulsations in active fault zones that increase the emanation efficiency of soils. They base their theory partly on the fact that Tn activities rise simultaneously with Rn activities over active faults and partly on the fact that there are no unusual Ra accumulations in the soils where the anomalous Rn values are observed. The theory is attractive and warrants further investigation. Zverev et al. (1980) have also observed increased Rn and Tn emanation over faults and in laboratory experiments; soil samples under treatment with ultrasound do indeed emanate more Rn than samples not so treated. Wilkening (1980) has reviewed the processes by which 222Rn is transported from the soil to the Earth's surface and concludes that "222Rn transport by ordinary molecular diffusion appears to be the dominant process". We conclude consideration of the mobility and range of Rn by affirming that diffusion is the main mechanism whereby Rn is moving through soils and in the great majority of cases a Rn anomaly will reflect a source less than 8 m from the anomalous site. However, sometimes movement of Ra with groundwaters generates a larger range for Rn. Over active fault and fracture zones convective forces may move Rn over larger distances and seismic micropulsations may result in higher emanation efficiencies. Anyone who has carried out Rn measurements in natural environments knows that differences of a factor of two in concentration are not uncommon at a site; occasionally tenfold or even greater differences are observed. Baranov (1961), for example, reports fluctuations of Rn at one soil site as a result of changes in meteorological conditions and Kramer et al. (1964) report fluctuations of Rn with depth in weathered tuff on the Rainier Mesa over a period of four months (Table XIV).
379
Radon
TABLE 11-XIV Fluctuations in radon concentrations at different depths due to meteorological changes and time Changing meteorological conditions (Baranov, 1961 ) Depth (cm) (Rn)max / (Rn)min
March - July (Kramer et al., 1964) Depth (cm) Fractional SD Rn conc.
25 50
320 80
20 40
0.34 0.12
! 00
!0
60
0.08
200
I
80
0.06
100
0.04
Evidently the ideal sampling depth from the viewpoint of a constant Rn signal is 2 m or more. In practice such a depth is seldom attained because of practical difficulties. Quite often overburden thicknesses are less than 2 m, and as was shown above, variation in depth from site-to-site in near-surface surveys can have a pronounced effect on the soil-gas Rn concentration, apart from any meteorological variations. The meteorological variables responsible for these large changes in the soil-gas Rn concentration are barometric pressure, moisture content of soil, wind, temperature, convection and Earth tremors. Barometric pressure effects are the best documented. Bogoslovskaya et al. (1932) measured the Rn flux in a plot of ground before and alter burial of uraniferous rock at a depth of 5 m, and related changes in soil-gas Rn to barometric pressure. Their results are summarised in Table ! I-XV. Note the large variation even at a depth of 4 m between rising and falling barometric pressures when U ore is present. Unfortunately these authors did not report the fluctuations of Rn in the overburden prior to the burial of ore, but did report the mean Rn content stabilised at a depth of 2 m. The results by
TABLE I I-XV Variations in soil-gas radon concentration with dcpth and barometric pressure (from Bogsiovkaya et al., 1932) Barometric pressure condition After being stable for 5 days Alter rising for 2 days After falling for 2 days Overall mean Mean for overburden only Mean for ore only
0.5 2.8 2.0 3.2 3.4 -
Measurement depth (metres) I 2 3 4 4.1 7.6 21.7 74.8 3.4 5.1 10.5 38.0 5.8 9.7 28.2 126.5 4.5 8.3 22.9 82.0 2.6 5.0 5.6 6.2 2.3 3.3 17.3 75.8
5 436 436 447 437 6.6 431
3 80
W. Dyck and I.R. Jonasson 1979
pc~
SOIL GAS R A D O N - 2 2 2 AND W E A T H E R MIDWEST LAKE, SASK.
IIOO
~t,v RADON
900
I
-,,.
700
,'
_kPo
"c 50 20 IO
/'~
-,"':-!W I, ......... I
0 -*iO
t/
....
3
t .......
b
"/"
I 9
Ill..,
II
13
Ib
~
17
JULY
19
21
.....I,
275 2b
,. I
2-1
29
.....~
31
2
~.,I
4
6
....... 1 1 8
IO
12.
14
...........
16
18
20
AUGUST
l:ig. I i-7. Soil-gas Rn and mctcorological variablcs at Midwcst l,akc, northcrn Saskatchcwan.
Bogoslovskaya et al. (1932) therefore suggest that variations in Rn concentrations by a factor of two or three at depths of 2-4 m during barometric pressure swings confirm U mineralisation. Other workers have lriade similar observations. Kovach (1945) measured an increase in Rn concentration in soil gas within the top 2 in during a decrease in barometric pressure. Kramer et al. (1964) measured a 20% drop in the Rn concentration at a depth of 60 cm during a 0.38 cm Hg rise in barometric pressure. Clements (1974) reported that a 1-2% change in presstlre produces a 20-60% change in the Rn flux at the soil-air interlace over a period of 1-2 days, with falling pressures causing a rise in near-surface Rn and rising pressures a corresponding decrease. The total effect of meteorological variables on soil-gas Rn in glacial till laced with weathered radioactive sandstone boulders is illustrated in Fig. 11-7 (data source, Dyck). A persistent inverse correlation of Rn and barometric pressure is evident. However, heavier rainfalls also tend to coincide with falling pressures. Soil-moisture effects may reduce or enhance presstlre effects. For example, a short strong rain may clog the soil surface allowing a build up of Rn. Prolonged rain, on the other hand, will tend to wash out soil Rn. In other words, the combined short-term meteorological effects of barometric pressure, wind, rain and temperature (and perhaps other factors) cause a
Radon
381
maximum short term (daily) fluctuation of about +20% in the Rn content at about 40 cm. In August, however, a long-term rising trend is evident which seems to be unrelated to any of the measured variables; a steady warming and drying of the soil could be responsible for this rise. The effect of temperature, particularly freezing temperatures, can be pronounced. Kovach's (1945) seasonal soil-Rn study showed that a frozen wet surface layer resulted in a fourfold increase in the Rn content at the 25 cm level whereas the 200 cm readings remained virtually unchanged through the year. Obviously the frozen layer acted as an effective lid on the overburden allowing Rn to build up. A study of the annual amplitude of temperature versus depth by Baranov (1961) shows that it is virtually zero at 8 m but reaches 30~ at about 25 cm. Such large swings in temperature at the depth at which most soil-Rn measurements are made is certain to affect the Rn concentration and particularly its rate of diffusion. The reader who wishes to study in more detail the factors that affect Rn exhalation rates from soils may consult Wilkening et al. (1972) and Tanner (1964a, 1964b) and references quoted in these papers. The build up of Rn at the frost-no frost interface was held to be the cause of a gamma-ray anomaly that showed up only in winter (Wennervirta and Kauranen, 1960). The authors have come across several accounts of gamma-ray anomalies over a caked layer of ice and snow which disappeared when the ice and snow layer was removed. A study of the nature and build-up of dissolved gases, including Rn, in snow-melt waters was made by Jonasson and Dyck (1978). The development of reducing conditions, evidenced by the presence of CH4 and H2S, effectively restricted dispersion of U from the source area. Radium was found to be closely bound to organic trash in the soils and to soluble and suspended humic material, and was dispersed from the source area. Radon dissolved early into ice and melt-water and was freely mobilised. As the thaw progressed and waters became oxygenated, U became mobile as UO22+ while Rn virtually disappeared. Pradel et al. (1963) observed a significant drop in Rn in soil gas during a period of frost but did not attempt to explain this unusual phenomenon. Novikov and Kapkov (1965) illustrate how the Rn content of thawed soil rises with depth until permafrost is encountered, at which point the Rn concentration drops sharply, simply because a water-saturated frozen layer has no voids into which Rn can diffuse or through which it can be transported by convection. In lakes the build-up of Rn under the ice is a common occurrence (Dyck et al., 1976b; Dyck and Tan, 1978) and can enhance an anomaly significantly. Similarly, springs emptying into lakes or streams inject varying amounts of Rn into the lake or stream, depending on the amount and frequency of rainfall, groundwater level fluctuations and whether or not these fluctations encounter U mineralisation. Novikov and Kapkov (1965) claim that an increase in Rn concentrations in waters by a factor of 5-10 times after a rainfall can occur in the presence of a U deposit and up to 3-4 times in the absence of a U deposit. Tanner (1964b) found that the Rn content of groundwater near the Great Salt Lake was closely related to the Ra content of the nearby sediments. Similarly Komer and Rose (1977) found that Rn in groundwaters in Pennsylvania reflected U mineralisation nearby, but Rn in streams did
3 82
W. Dyck and L R. Jonasson
not show such a relationship because of rapid degassing of turbulent streams. Such large fluctuations of Rn in stream waters are comparable to the large fluctuations of Rn in near-surface soil air. In both cases the Rn signal stabilises when samples are taken from below the surface. In summary, the causes of fluctuations of Rn in soil gases are: meteorological variables (barometric pressure, wind, temperature and moisture), changes in fluid flow (thermal convection, fault-controlled pressure release and water movement) and Earth tremors. Changes due to meteorological conditions are seldom more than +20% at shallow depth and can be reduced by deeper sampling stations or, in the case of soil gases, by integrating Rn detection techniques. Integrating techniques will also reduce Rn fluctuations due to the other causes noted above. Lakes and groundwaters already provide an integrating medium, making such techniques superfluous. As stressed earlier, the heterogeneous distribution of Ra and the sampling-site depth variations in routine soil-gas Rn surveys are the two main causes of poor precision. Faulty equipment and poor sampling techniques can also cause considerable variations in Rn measurements. Pradel et al. (1963) demonstrated that by discarding the first 20 L of soil gas, fluctuations of the Rn content in hole-to-hole and day-to-day samples could be reduced from a factor of I0 to 1.5 using their particular sampling and counting system.
ANAI.Y'I'ICAI, MlTI'I I()I)S All current Rn detection methods depend on the radioactive properties of Rn and its decay products. There are no conventional chemical analytical methods sensitive enough to detect natural levels of Rn. As shown in Table I !-!!, at equilibrium one gram of U is equivalent to 2.13 x ]0 "12 g o f Rn. This means that 1 kg of rock containing 4 ppm U has at most 8.5 x 10-~5 g of Rn; yet by recording individual atomic disintegrations, Rn concentrations are measured with relative ease in the natural environment. Lord Ruthertbrd did his famous experiments on the nature of the atom by visually observing light flashes produced by alpha particles on an activated ZnS screen. But most of the early work was carried out with ionisation chambers using a gold-leaf electroscope, and later with electronic ion current meters to measure radiation intensities. With the invention of the photomultiplier in 1940, activated ZnS screens became the dominant Rn detectors. The introduction of integrated circuits and microprocessors led to the development of compact sophisticated instruments now on the market. These together with the refinement of signal integrating techniques raised the popularity of Rn methods of prospecting. Several articles describing one or more methods are in print; theretbre this discussion will deal only briefly with the principle of each method and refer the reader to original publications for more details. For convenience the methods are subdivided into instantaneous, semi-integrating and fully-integrating categories. Each has advantages and disadvantages depending on the mode of application. By
Radon
383
"instantaneous" is meant that an answer is obtained within minutes and that is its main advantage. Semi-integrating methods give an answer in several hours or days. Their advantage lies in the fact that they integrate or smooth out diurnal fluctuations but the response is slower. Fully-integrating techniques can average Rn signals for several weeks or months, thus further reducing diurnal variations and longer-term environmental fluctuations, but with the unavoidable delay in the result.
Principles of methods In the 238U radiodecay chain there are, between 226Ra and 21~ three alpha decays and two beta decays taking place within a reasonably short time span compared to the half-lives of 226Ra and 2~~ (Table I 1-I1). When a radioactive daughter has a short halflife compared to the parent, so-called secular equilibrium is achieved when parent and daughter remain together. Mathematically this relationship is expressed" X i N i = X2 N2 : X3 N3 etc. where X is the decay constant. Since X equation (! I. I )" N~/t~
=
N2/t 2 -- N3/t3 etc.
( ! i . I) In2/t, where t is the half life, we may rewrite
(l 1.2)
Thus in secular equilibrium the ratio ot'the number ot'atoms to the half-life is a constant, i.e., the shorter the half life the taster the decay rate, since the rate of decay, d N / d t - XN
(11.3)
in practical terms, this means that the longer the half-life, the larger the amount of substance needed to make a quantitative determination by means of its radioactivity. The alpha particle ejected from a disintegrating atom dissipates its energy by colliding with matter and ionising atoms and molecules in its path until it reaches thermal velocities. Then it acquires two electrons and becomes a He atom. In dry air at 15"C and 760 mm Hg pressure, the range of the alpha particle from Rn is 37 ram, i.e., it will take 37 mm of air to stop the alpha particle emitted from Rn. In photographic emulsions the range is about 40 x 10-3 mm and in rocks about 20 x l0 -3 mm. Each ion pair produced by the alpha particle requires about 35 eV, thus the alpha fiom Rn produces about 16 x 104 ion pairs during its lifetime of less than a microsecond. When this packet of ion pairs is confined in an enclosed space, collected on electrodes under the influence of a potential difference, integrated and measured as an ion current with an electrometer, the detector is called an ionisation chamber. When the ionisation takes place on activated ZnS, flashes of light in the visible region are produced and the
384
W. Dyck and I.R. .]onasson
detector becomes an alpha scintillometer. The activation is achieved by introducing an impurity, usually Ag, into the lattice of ZnS, giving rise to electron transition bands which emit energy or photons when excited. These can be detected by the human eye and the photomultiplier and can be recorded by a pulse counter. When the alpha particle expends its energy in a p-n junction of a silicon diode, a bundle of electrons is lifted into the conducting band. When these electrons are recorded we talk of a solid state detector or surface barrier detector. The thermoluminescent dosimeter is a special phosphor, such as CaSO4 doped with Dy, with the ability to trap electrons between the conducting and the valence bands. These trapped electrons will return quickly to the valence band when the phosphor is heated, producing a light pulse that can be recorded with a photomultiplier tube. When the alpha particle strikes matter such as a cellulose nitrate film, it damages the surface by leaving a trail of broken molecular bonds in its wake. This damaged microtrack is attacked more readily than the undamaged surface by etching fluids, such as NaOH, and eventually becomes visible under the microscope as a track or dot. These are called etched track detectors. If instead of the alpha particles we wish to record the gamma rays emitted by the short-lived decay products of Rn, we use a Tl-activated Nal crystal; then the technique is known as gamma-ray scintillometry. The beta particle counter detects and records beta particles. Because they are much less penetrating than gamma rays and of much lower energy than alpha particles they require much more sophisticated detection equipment than the gamma rays or the alpha particles. The alpha-particle detectors are in general the simplest, most rugged and most sensitive devices for the detection ot" Rn. 'Fheir high sensitivity stems partly from the tact that they are able to discriminate these alpha particles from other sources ot" natural radiation such as cosmic rays, natural rock K and most debris from nuclear reactors and bomb testing. Each of the above principles can be used in an instantaneous, serni-integrating or fully-integrating mode, but cost, sensitivity and convenience dictate preferences. A summary of favoured applications of methods and principles is given in Table I I-XVI. Several reviews of methods, instruments and applications have been published (Morrison et al., 1969, IAEA, 1974; Smith et al., 1976; Morse 1976; Dyck, 1979a: Dyck 1979b). A more detailed description of each mode is presented below.
Instantaneous mode The ionisation chamber and the alpha scintillometer lend themselves best to the instantaneous mode of Rn determination because of their higher sensitivity compared to other methods. In this mode, the time of sample collection and analysis are of the order of minutes. Both instruments are also known as emanometers and both depend on a system of taking a soil-gas sample, usually by pounding a hollow rod into the ground,
385
Radon
TABLE 11-XVI Favoured applications of different methods of radon measurement Method Instantaneous 0.5L ionisation chamber Alpha scintillometer
Semi integratin,~ Collector (daughters) Collector (Rn and daughters) Fully integratin~ Etched track Solid state alphameter Thermoluminescence
Application(s)
Turn-around
Advantages
Problems
Soil gas Groundwater Soil gas Surface water Groundwater
Minutes Minutes
Rugged Quick turnaround High sensitivity Simplicity Inexpensive
Humidity Dust Background Light
Soil gas in situ
Hours
Slow turnaround
Soil gas in situ
Days
Simple Inexpensive Simple Inexpensive
Soil gas in situ
Weeks
Simple
Soil gas in situ
Hours/weeks
Replicationeasy
Soil gas in situ
Hours/weeks
Simple
Expensive Slow turnaround Expensive Slow turnaround Expensive Slow turnaround
withdrawing a soil-gas sample from the desired depth with a piston pump and bringing the gas into the detector chamber. The i o n i s a t i o n c h a m b e r was used almost exclusively prior to the invention of the photomultiplier tube. A 0.5 L chamber is a convenient size for a portable field instrument. It has a detection limit of about 30 pCi/L and hence is sensitive enough for soil-air Rn determinations since samples usually contain 100 pCi/L or more. The instrument is fairly rugged but requires extremely good air filters because fine dust can easily introduce a signal that is greater than the Rn signal. Extremes of temperature and humidity can also be problems. Detailed descriptions of ionisation chamber theory and applications for prospecting are given by Baranov (1961) and Novikov and Kapkov (1965). A field instrument incorporating solid state electronics is described by Scintrex Ltd. (1975). With a vacuum-tight chamber and a suitable degassing system, a 0.5 L model could be used for Rn determinations in groundwaters where concentrations are usually greater than 100 pCi/L. Larger chambers, including some over 300 L, have been used for the determination of Rn fluxes from soils and for the determination of atmospheric levels of Rn (Kramer et a1.,1964; Israel and Horbert 1970; Dyck, 1973).
3 86
W. Dyck and I.R. Jonasson
However, the ionisation chamber is of little use in detecting background values in surface waters. The alpha scintillometer is probably available in more versions than any other Rn detection device. When Rn gas is brought into the ZnS(Ag) cell it is commonly referred to as the Rn sniffer. The reason for its popularity is its relative simplicity, ruggedness, versatility, sensitivity and low cost. Radon concentrations of a few pCi/L in gases or waters can be measured in ten minutes without appreciable interference from humidity, dust or temperature changes. The two main drawbacks of the sniffer are: (1) the build-up of a background count in a ZnS(Ag) cell when very active samples are encountered; and (2) stray light striking the ZnS(Ag) produces an avalanche of pulses which takes up to ten minutes to dissipate. The solution to the first problem is a large number of cells; and to the second problem, a dark environment during cell changes. At least one manufacturer has been successful in discriminating electronically against the light pulses. There are many accounts of alpha particle detection instruments for the measurement of Rn (Sedlet, 1966). Lucas (1957) was one of the first to develop a very low background counting system. Higgins et al. ( 1961 ) adapted the method for well-water Rn and Ra surveys. Peacock and Williamson (1962) developed a shallow-borehole probe using ZnS(Ag) that could make in-situ Rn determinations without the use of a pump, but required a 5 cm diameter light-proof hole into which the ZnS(Ag)-coated lucite rod and photomultiplier assembly was inserted. Rushing et al. (I 964) used a similar technique tbr the determination of Rn in effluents and environmental samples. For U prospecting, Dyck (1969) and Allen (1976) applied a Lucas-type cell to determine Rn in soil, lake waters and stream waters.
Semi-integrating mode" Two methods lend themselves to this mode of Rn determination: the "Collector" and the "ROAC" (Rn on activated charcoal) methods. Semi-integration implies that Rn and its daughters are allowed to build up until their rate of production is equal to their rate of decay (a steady state concentration) and then their activities are measured. The Collector method requires a means of collection and storage of Rn decay products and the ROAC method a means of collecting and storing Rn and daughters. The Collector method is a relatively recent innovation in U exploration instrumentation (Card and Bell, 1979) even though the principle has been employed in Rn studies for decades. It is based on the collection and detection of the decay products of Rn. When a clean surface, such as a metallic or plastic disc, is exposed to an atmosphere containing Rn, such as a hole in the ground, the decay products of Rn will collect on it if they strike it. The activity of these daughters builds up with a composite half-life of about 35 rain. and reaches a maximum in about 3.25 h. Hence any fluctuations in the Rn concentration which are of much shorter duration than 4 h will be
Radon
387
integrated. Since the decay products emit alpha and beta particles and gamma rays, there is a wide range of detectors to choose from. One of the simplest and most sensitive is the alpha scintillometer described by Card and Bell (1979). The collector is suspended from an inverted cup, placed in a hole and the cup is then covered with a plastic blanket and soil and left for four hours or more. Then the collector is transferred to the ZnS(Ag)coated chamber of the alpha-scintillometer and alpha activity determined. To minimise the effects of thoron-daughter alphas, a five minute waiting time can be used between removal and counting. However, counting has to be done soon after removal and the counting times have to be sequenced accurately for high precision. A well established method for the detection of airborne Rn in mines is based on the collection of dust on filters and counting the radioactivity. In this case the collector comprises millions of small dust particles. A novel commercial innovation of the collector method is the alphacard of Alpha Nuclear Ltd., Toronto, Canada. It uses a plastic film in a slide for the collector and two solid state detectors facing each other inside a portable computerised alpha counter. The collector method has also been applied to soil samples in the laboratory (Card and Bell, 1979; McCorkell et al., 1981). By sealing a collector in a jar with a dried, screened and weighed soil sample, semi-quantitative Rn determinations are possible. Also frozen, swampy or water- logged soils can be tested during the winter months. The ROAC method is based on the tact (which has been known and utilised tbr a long time) that Rn adsorbs onto activated charcoal. Wennervirta and Kauranen (1960) demonstrated its usefulness in prospecting by pumping soil air through an ampoule containing activated charcoal and counting the ampoule with a shielded gamma-ray counter. Kapitanov et al. (1970) studied the use of charcoal for the removal of Rn from mine air and tbund that when working with air that is at 100% relative humidity the equilibrium distribution coefficient of activated charcoal for Rn was decreased by a factor of two compared to dry air. The South African Atomic Energy Board (1980) has developed and streamlined the ROAC system for U prospecting. By burying opened charcoal cartridges contained in cups the system can be used in the semi-integrating mode. Because it takes time for Rn to diffuse from soil into the cartridge, 8-10 days are needed to saturate the charcoal. Hence signal integration is in the order of days and diurnal variations are smoothed out. Since the parent Rn remains trapped with its daughters when the cartridge is removed from the ground and sealed, there is ample time for counting providing Rn escape is negligible. By inserting a diffusion filter such as a thin layer of wool or foam, Tn can be prevented from absorbing on the charcoal and hence will not contribute to the activity. Counting is done with a good-quality scintillometer inside a 5 cm Pb shield. The attractive features of the system are very simple field procedures with purchased cartridges, integration of diurnal Rn fluctuations and higher sensitivity than in-situ gamma-ray measurements. A disadvantage is that two trips have to be made to the field.
388
W. Dyck and 1.R. Jonasson
Fully-integrated mode Photographic and etched cellulose nitrate films and solid state detectors with built-in pulse registers fit into the fully-integrating mode methods. They keep accumulating and integrating alpha pulses as long as the operator chooses to expose them to alpha activity. Three proprietary integrating techniques, Track-Etch, the alphameter, and the Thermoluminescent Dosimeter came onto the market in the 1970s. The Track Etch method is probably the most widely publicised Rn method of prospecting. Although radiography is an old and reliable method of detecting radioactivity (Hatuda applied it to prospecting in 1954), the sensitivity of photographic emulsions to light made it an impractical field method. However, the use of cellulose nitrate film (Morrison et al., 1969; Fleischer et al., 1975; Gingrich and Fisher, 1976) has made it a rugged and simple technique which truly integrates the Rn flux from soils. The film, placed in a plastic cup, is buried in the soil for three or four weeks, recovered and etched in NaOH solution to make visible the tracks left by the impinging alpha particles. The tracks per unit area are counted under a microscope. The only real drawbacks of the method are its high cost when the film cups and the counting are purchased commercially, the long turn-around time and the inconvenience of two trips to the field. Tracks from Tn can be prevented by placing a thin plastic film over the cups (Fleischer and Mogro-Campero, 1978, 1979); with a 54 sec. half-life, Tn decays before it can diffuse through the film. The alphameter method became practicable when inexpensive silicon diodes and integrated circuits were developed. Each alphameter is in fact a miniature detectorcounter assembly that weighs less than 0.5 kg, and which thirty years ago would have been very bulky and weighed 20 kg. Alpha particles striking the active surface of the diode are counted and stored until a reader is connected and the accumulated count recorded. Thoron and humidity effects are minimised by covering the detector with a thin plastic film. Its sensitivity is such that 24 h exposure gives sufficient counts in most soils (Gaucher, 1976; Warren, 1977). Its main advantages are that, once the detector is in place, many readings can be taken consecutively without disturbing the site, and moreover the detector lasts almost indefinitely. However, since each unit is a sophisticated piece of equipment, capital costs are high and it needs considerable care in handling. Thermoluminescent dosimeters are a relatively recent addition to Rn methods of prospecting (Smith, 1978), although the principle of thermoluminescence has been applied in other fields for decades. As with the other integrating devices, the thermoluminescent device (TLD), usually a protected thin wafer of Dy-doped CaSO4, is placed inside a cup and buried in the soil for a predetermined time. By keeping the TLD very thin (76 mm), gamma radiation and beta particles have little effect on it, but alpha particles deposit all their energy on it, causing a proportional number of electrons to be trapped in intermediate energy states or bonds. After the desired exposure time the phosphor or TLD is placed on a hotplate coupled to a lightmeter. The heat from the
Radon
389
hotplate excites the trapped electrons to the conducting band. When they return to the ground state (or valence band), the energy release at a luminescence centre of the phosphor causes visible light to be emitted, which is proportional to the initial alpha particle energy. Thus the total amount of light given off by the phosphor is a measure of the length of exposure and Rn concentration of the environment in which it was placed.
FIEI~D METHODS
Determination o f radon in natural waters Methods of determining Rn in natural waters usually involve the removal of Rn from an aliquot of sample by passing a fine stream of air bubbles through it, filling an evacuated ZnS(Ag) cell with this air-Rn mixture and counting the alpha particle activity in the cell with an alpha scintillometer. Water-tight cups coupled to integrating-type detectors can be used for in-situ measurements of Rn fluxes from lake or stream beds; but since the water itself integrates the Rn flux from the bottom sediments there seems little ,justification for going to the trouble of planting and retrieving cups when it is much simpler and faster to take bottom water samples. The ZnS(Ag)-coating in the cell is prepared by mixing a powder, such as Du Pont luminescent chemical type No. I101, with acrylic cement (dichloroethylene with a little acrylic plastic dissolved in it) until a freely-flowing paste is obtained. Tile coating is applied to a clean cell wall making sure that an infinitely-thick layer of scintillator is obtained, i.e., 20 mg/cm 2. Cells prepared in this way last almost indefinitely when treated with care. A high-viscosity grease coating, dusted with ZnS(Ag) powder, also give a satisfactory scintillating surface, but efficiency and precision may be poorer. The amount of light received by a photomuitiplier from an alpha decay in the cell depends upon the amount of energy the alpha particle expends on the sulphide layer, the angle and distance between points of photon generation and detection, and the light transmission efficiency of the window and filling gas. Such considerations, plus tile fact that the range of tile alpha particles from Rn in air is about 37 ram, indicate that the hemispherical or coneshaped cell design is ideal. However, the cylindrical shape is usually chosen because it is easier to manufacture. The parts can be interconnected with metal and plastic tubing. Use of rubber or plastic tubing should be minimal, as Rn has an affinity for adsorption on rubber. Consequently, it may take several hours to clean a rubber tube which has been exposed to a very active sample in order to minimise the risk of cross-contamination of more normal samples. Tile fritted disc at tile bottom of tile sample holder disperses air into fine bubbles, resulting in more efficient scrubbing of the water sample. A simple vacuum pump is most convenient and efficient for the transfer of gases. However, where electric power is not available, a good quality hand pump is satisfactory, although with a 10% reduction in efficiency. The major electronic components are very similar to those of a gamma-ray scintillometer. The photomultiplier is enclosed in a light-tight tube with a removable top
390
W. Dyck and I.R. Jonasson
TABLE 11-XVII Radioactive decay series from radon to lead Isotope ZZZRn 218po 214pb
214Bi 214po 2~~
Decay c~ or. B B ct B
Half-life 3.82 days 3.0 min. 26.8 min. 19.7 min. 10-4 sec. 22 years
or cap connected to a safety switch, which disconnects the high voltage from the tube to protect it from being overloaded when exposed to light during sample cell changes. A 1000 pCi/L 226Ra solution accumulates sufficient Rn in one day to permit a daily check on the performance of the instrument and Rn extraction line. By treating the standard in the same way as the unknown, faults in procedure and equipment are detected quickly. An active sample produces a rise in cell background counts fbr the fbllowing reason. When a clean cell is filled with Rn at time zero and evacuated after three hours, a constantly-increasing alpha activity during this time is observed as a result of the growth of Rn daughters which are also radioactive. After about three hours, equilibrium between growth and decay of the daughter products and Rn is reached and the whole series from Rn to 2X'lPo decays within the half-life of Rn. Because of its very long half-life of 22 years, 2~~ does not contribute measurably to the total alpha activity. When the Rn is removed from the cell, the activity decreases; rapidly at first due to the 3-minute :~SPo, then more slowly with a composite half-life of about 35 minutes due to the 27-minute 2~4pb and the 20-minute 2~4Bi. The decays in this sequence are shown in Table 1 I-XVII. Radon loss from water stored in plastic bottles, taking natural decay into account, can be appreciable. Even greater losses occur when bottles are only partly filled: loss is exponential with time, and up to 90% of the Rn content can escape from the water in a half-filled bottle in three days. in relating Rn concentrations of water samples to source concentration it is important to know the detection limit as well as the reproducibility or errors of the method. The use of larger cells and samples, sample enrichment and longer counting times lowers the detection limit but decreases the output. The errors associated with the method can be divided into four types: instrumental, experimental, sampling and statistical. Instrumental errors include detection efficiency changes of cells and photomultipliers, changes in the gain of amplifiers, high-voltage drifts and faulty operation of other components such as pressure gauges, timers and valves. The main experimental errors are pressure, volume and time measurements, and the degassing procedure. Sampling errors result mainly from the escape tendency of Rn. To minimise these, glass bottles
Radon
3 91
with well-sealing caps should be used and filled to the top except for one air bubble, necessary to prevent the tightly sealed bottle from cracking during temperature changes. The largest uncertainty in the Rn values results from the inhomogeneous distribution of Rn in natural water bodies. Experience has shown that the act of sampling can disturb the water and change the Rn concentration at the site by as much as a factor of five. In view of such large variations, the other errors, including the errors associated with the randomness of radioactive decay, are negligible. With constant Rn sources, the statistical counting error is dominant in all but very active samples (several hundred pCi or more). The same method can be used for the determination of Ra in natural waters, simply by degassing a sample completely and letting the Rn grow back in for a known length of time in a sealed bottle. Time permitting, a sample may simply be stored for a long time (40 days or more) to allow all unsupported Rn to decay. This long waiting period is necessary in case there are samples with high initial Rn activity but little Ra. The same technique may also be used for the determination of Ra in rocks and soils. The solid samples are dissolved and the solutions treated in the same manner as the water samples. Equivalent U can be calculated from the Ra concentration using the expression: eU ( p p m ) - 2.962 x Ra (pCi/g) The constant embodies the half-lives of allows for 0.7% 2~5U in natural U.
238U (4.468 x 10'~ y) and 226Ra (1600 y) and
Determination o f radon in soil emanations Semi-quantitative determinations of Rn in soil air are carried out relatively quickly with the alpha scintillometer. In sandy well-drained soils a portable auger is the preferred method of making a hole. In the dynamic mode, soil gas is pumped from the hole through the detector to the atmosphere and a Rn reading taken when the signal stabilises. In the static mode soil gas is circulated through hole and detector in a closed loop. A great variety of soil-gas probes have been developed for different soil types and applications. Some are designed for difficult overburden such as glacial till containing many boulders, others are relatively light and work well in light fine overburden. It is important to make certain that only soil air is sampled and not some mixture of soil and atmospheric air, and for this purpose some probes are equipped with inflatable balloons which provide a seal against atmospheric air when used in boreholes in very porous overburden. A vacuum gauge in soil-gas extraction lines alerts the sampling team to leaks in lines or channelling in soils and thereby improves sampling precision greatly. However, all probes have one common deficiency, they are difficult to hammer into the ground to a depth of 50 cm or deeper; only at depths below 100 cm does the Rn flux become sufficiently stable to allow quantitative determinations to be made.
392
W. Dyck and I.R. Jonasson
All soil gases contain a mixture of radon, thoron and actinon. Whilst the half-life of the last (5.4 sec.) is too short to interfere with Rn determinations, Tn is always present in in-situ determinations. To relate Rn and Tn emanations quantitatively to U and Th concentrations is difficult, if not impossible. Too many factors, such as radioactive disequilibrium of the series in the soil as a result of weathering, changes in the emanation efficiency due to soil density and grain size, moisture content, and distribution of radioactive particles, cause erroneous extrapolations. However, precise, quantitative results are not of prime importance in prospecting. The important point is that there be correspondence between Rn in soil emanations and U in the area nearby. That this correspondence exists has been demonstrated numerous times and is discussed in more detail below. To avoid the Tn effect, the authors favour the following modes of operation. Evacuate a vacuum tight container (or cell, if sufficient cells are available) with a highquality hand pump, allow the pure soil air from a probe in the ground flow into the cell and take the filled cells to a field laboratory for counting and cleaning. In the case of integrating-type devices, a thin plastic film is placed over the cups in which the device is held. Thoron decays before it can diffuse through the film but Rn does not.
C()MI'ARISON S'I'UI)II:~S AND CASE I IIS'I'ORII'S There is such a large number of comparison studies and case histories in the literature that it is virtually impossible to cite them all. Many of them have already been referred to in the foregoing sections. Tanner (1964b, 1978) described a number of case histories and comparison studies involving soil-gas radon. Adkinson and Reimer (1976) prepared a bibliography of references of geological interest to uranium exploration. Ill their comparison of soils from Red Desert and Copper Mountain, Wyoming, and Spokane Mountain, Washington, Pacer and Czarnecki(1980), employed tile following Rn detection techniques: zinc sulphide detectors, charcoal cannisters, an ionisation chamber, alpha track detectors, thermolurninescence detectors and partial extraction of 2~~ Their results showed that integrating techniques exhibited lower variability and lower peak to background ratios than instantaneous techniques. All techniques outlined the Iocalised U mineralisation, gave similar anomaly patterns, and correlated positively with equivalent U in the soil. Stevens et al. ( 197 I) obtained very diverse results for three areas in the western United States using the soil-gas Rn prospecting technique. In the Denver Basin, Colorado, anomalous Rn (eight times background) was found above a uranium bearing sandstone at a depth of 3-9 m. In the Lisbon Valley area, Utah, Rn measurements along traverses over U ore bodies 6-800 m below interbedded shales, mudstones and sandstones gave no clear indication of the ore bodies. Soil-gas Rn in the Laramie Basin outlined anomalous bands which coincided with anomalous U in the soil but no anomalous radioactivity was encountered in drill holes over these anomalies. Severne (1978) evaluated the alpha scintillation counter and alpha-sensitive films at
Radon
393
Yeelirrie, Australia, where significant near-surface carnotite mineralisation occurs in a semi-arid environment. Both Rn techniques provided essentially the same response which was keyed to local U distribution rather than to the overall disposition of the mineral zones several metres distant. Wennervirta and Kauranen (1960), Dyck et al. (1976b) and others have demonstrated the Rn build-up under frozen soil or water near U mineralisation. Gingrich and Fisher (1976) describe five successful soil-gas Rn surveys, two from the USA, one from Canada and two from Australia. They claim to have obtained Rn anomalies from U ore located up to 120 m below the surface. In some cases the anomalies were displaced horizontally relative to the location of the ore for some unknown reason. For Rn to traverse such distances in highly anomalous amounts requires strong upward gas flows or secondary U mineralisation, too weak to be detected by conventional means, near the surface. Neither mechanism was proven to be operative at these test sites. From studies of U deposits in general, it is clear that secondary mineralisation is a common occurrence and therefore a likely explanation for the Rn anomalies over or near the U deposits. Gas flow over 50-100 m cannot be ruled out but has not as yet been proven independently, other than in such natural phenomena as geysers and volcanos. Nonetheless the fact remains that soil-gas Rn surveys detect U mineralisation at depth and are a most useful tool in U exploration, particularly in arid regions. In temperate environments, surface waters (lakes and streams) become plentiful and groundwater tables rise to the surface in many places, absorbing and integrating the Rn emanating from soils and rocks. A summation of extensive studies and case histories of Rn and other constituents in water by Russian scientists is given by Novikov and Kapkov (1965) and reproduced in part in Table l l-Ill. Groundwater regimes may be quite complex, particularly in mountainous terrain, and the localisation of the source of radioactive water is a formidable task in such regimes. However, Lester (1918) was able to conclude from his study of mineral springs in Colorado that 75% of the more active springs are near metamorphic rock, that the most active springs are found on both slopes not far from the continental divide, and that the radioactivity of the waters is picked up from radioactive matter underground. In the Port Radium region, northern Canada, Senftle (1946) was able to show a direct relationship between Ra, conductivity of lake waters and degree of mineralisation of the rocks nearby. Several regional surface and groundwater surveys in Canada have demonstrated quite convincingly the effectiveness of the Rn technique in outlining U mineralisation (Dyck et al., 1971; Dyck et al., 1976a; Dyck, 1980; Dyck, 1982). Korner and Rose (1977) and Rose and Korner (1978) found that groundwater Rn surveys in Pennsylvania were effective for outlining U deposits, but that stream water Rn surveys were not. Using the multi-method attack, King et al. (1976) were able to locate several drilling targets and subsequently intercepted mineralised ground in several holes.
394
W. Dyck and L R. .lonasson
FUTURE NEEDS 9 In the realm of basic research, thorough studies are required to determine to what extent soil-gas Rn anomalies reflect local flow from depths beyond the diffusion range of Rn. To do this properly will require extensive soil Ra analyses and the simultaneous analysis of other soil gases such as Tn, 02, CO2 and He. 9 A comprehensive study of the geochemistry and migration of Ra in the ground is required to help explain Rn anomalies encountered in waters. 9 There is a need for standardised and calibrated Rn detection equipment and procedures, so that absolute concentration levels may be obtained for world comparisons.
Geochemical Remote Sensing of the Subsurface Edited by M.Hale Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
395
Chapter 12
MERCURY G.R. CARR and J.R. WILMSHURST
INTRODUCTION The rationale for using Hg as a pathfinder element in mineral exploration is attractive. Because of its volatility, Hg is presumed to form broader halos in the hypogene environment than most elements. It is envisaged that both vapour-phase and solution transport are responsible for a wide dispersion of the element. In the secondary environment it is well known that Hg exerts a measurable vapour pressure at ambient temperatures and possesses redox properties that allow the metal to exist in the elemental state under a range of natural conditions. Therefore it has been claimed that, as a host sulphide-body weathers, it can be expected that Hg will be converted partly to the vapour state, thereby overcoming the constraint of hydromorphic or solution dispersion that applies to other target and indicator elements. Vapour-phase dispersion through permeable rock or cover would allow Hg to be detected in soil or soil gas, and perhaps as an atmospheric anomaly. The evidence for such claims comes principally from experience in the former USSR, where the pioneering work on the theory and application of Hg techniques took place (e.g., Saukov, 1946; Fursov, 1958; Ozerova, 1959, 1962, 1971; Dvornikov and Petrov, 1961). In the Western literature, the development of ideas on the use of Hg in mineral exploration can be traced through such publications as Hawkes and Williston (1962), James (1962, 1964), McCarthy (1972), McNerney and Buseck (1973), Wu and Mahaffey (1978), Ryall (1979b), Zonghua and Yangfen (1981) and Carr et al. (1986). Despite the considerable literature on the subject, these claims can rarely be validated, due to principally incomplete case-history documentation. The concept of a broader halo in both the primary and secondary environments assumes a comparison with other elements. In case-history studies, however, such comparisons are often not presented and we are left in some doubt as to the significance or extent of Hg dispersion. Over the past two decades, many field geologists in North America and Australia have been disappointed by unsuccessful attempts to use Hg pathfinder techniques in their exploration programmes. This lack of success can be attributed both to falsely-high expectations and to technical problems.
396
G.R. Carr and J.R. Wilmshurst
TABLE 12-I Typical mercury contents of common silicate, sulphide and gangue minerals (from Kleinevoss, 1971) Group
Mineral
Silicates
Feldspar Biotite Pyroxene Quartz Hornblende
Sulphides
Molybdenite Arsenopyrite Pyrite Pyrrhotite Bornite Antimonite Galena Sphalerite - scdimentary - hypothermai - mesothermal - epithermai Tetrahedrite Reaigar
Gangue
Barite Carbonate Fluorite
Range (ppm) 0.1- 0.8 0.3 - 0.4 0.5 - 0.6 0.8 - 10 1 - I0 1 - 10 1 - 10 1 - 10 1 - I0 1 - 10 1 - 100 1 - 200 I - 100 1 - 100 I 0 - 1000 1 0 - 50000 I - 50000 100 - 10000 I - 200 0.1 - 100 0.01 - 100
It is the a i m o f this chapter to help the e x p l o r a t i o n g e o l o g i s t and g e o c h e m i s t to d i s c r i m i n a t e those e x p l o r a t i o n situations w h e r e the use o f Hg t e c h n i q u e s is valid and cost effective f r o m those w h e r e other t e c h n i q u e s are p r o b a b l y m o r e appropriate.
GEOCHEMISTRY OF MERCURY A l t h o u g h m i n e a b l e deposits o f H g occur only as cinnabar ( H g S ) and native Hg, there are at least 38 m e r c u r y m i n e r a l s (Tunell, 1978) and trace c o n c e n t r a t i o n s o f m e r c u r y are found in a wide range o f m i n e r a l s (Table 12-I). L e v e l s in m o s t i g n e o u s rocks not associated with m i n e r a l i s a t i o n are g e n e r a l l y less than 100 ppb, but no definitive statistics exist and there are wide regional variations. S e d i m e n t a r y rocks in g e n e r a l h a v e H g
Mercury
397
TABLE 12-II Typical mercury content of common rocks (after Jonasson and Boyle, 1972) Group
Rock type
Igneous
Metamorphic
Sedimentary
Range (ppb)
Mean (ppb)
Ultrabasites Basic plutonites Basic volcanites Intermediate plutonites Intermediate volcanites Acid plutonites Acid volcanites Alkali rocks
7 - 250 5 - 84 5 - 40 13 - 64 20- 200 7- 200 2 - 200 4 0 - 1400
168 28 20 38 68 62 62 450
Quartzite Amphibolite Hornfeis Mica schist Gneiss Marble, recta-dolomite
10 - 100 30- 90 35 - 400 i 0 - 1000 25 - 100 ! 0 - 100
50 225 100 50 50
Recent fluvial Recent lacustrine Recent marine Sandstone, arkose Clay, shale Carbonaceous shale Limestone, dolomite Gypsum, anhydrite Halite, sylvanite
! 0 - 700 I0- 700 10 - 2000 10- 300 5 - 300 100 - 3250 10- 220 10 - 60 20- 200
53
73 73 100
55 67 437 40 25 30
contents similar to those of igneous rocks, but shales and clays contain higher amounts. Carbonaceous and bituminous shales may be enriched in Hg, containing up to several thousand ppm. The Hg content of various rock types is shown in Table 12-II. It is considered that in the primary (hypogene) environment, Hg is transported principally as complexes in hydrothermal solutions, although the relative importance of sulphide, bisulphide, chloride, hydroxy and organic complexes remains controversial (Barnes et al., 1967; Barnes, 1979). At temperatures of over about 200~ vapour-phase transport may also be important, but very little evidence exists to indicate the possible scale and practical significance of any such transport (Krauskopf, 1964; Barnes et al., 1967; Khodakovsky et al., 1975). The main natural source of Hg in the secondary environment is believed to be the oxidation of sulphide minerals (Saukov, 1946; Williston, 1964; Ryall, 1979b), although
398
G.R. Cart and J.R. Wilmshurst
McNemey and Buseck (1973) suggest that some Hg in the secondary environment may originate from deposits located below the zone of oxidation. The vapour pressure of Hg at ambient temperatures is significant (13 and 22 ng/mL at 20 and 25~ respectively) and this factor alone has led to the concept of the vapour-generated anomaly in the secondary environment. The solubility of the element in water is also high, exceeding the level in the vapour phase by roughly a factor of three. The partition coefficient therefore favours the aqueous phase and this has practical consequences that are discussed below. Mercury is also significantly soluble in a number of organic solvents and this is of possible relevance in organic-rich soils. In its chemical properties Hg differs distinctly from Zn, with which it is commonly associated in the primary environment. Whereas Hg has fields of stability for the zero, one and two valence states, Zn in the natural regime exists only as Zn (II). In addition to Eh and pH, Hg phase stability is dependent on the concentration of the halide and sulphur species. Within soils Hg can exist in the ionic or associated forms, HgC1, Hg(OH)2 and HgC12, as elemental Hg and within various solid phases. The presence of gas-phase Hg in soil is therefore understandable, although it is evident that under conditions of equilibrium the level of free Hg in aerated chloride-containing soil will be low. It can be expected that similar equilibrium conditions apply to the aqueous solutions derived from sulphide oxidation, since the sulphate ion does not interact strongly with the several Hg species. In practice, however, it seems that much of the Hg contained within soils and weathered rock is more or less firmly associated with solid phases. This can be shown to extend from simple sorption to structural incorporation within oxidate minerals. The manner of the association between Hg and organic matter has been discussed by Jonasson (1970). The question of adsorption is a rather grey area and, whilst it is clear that it plays a very significant role in soil experiments (e.g., Andersson, 1979), it seems that in mature soils the Hg is held rather more firmly. (This is apart from discrete mineral hosts deriving from sulphide oxidation.) It is possible that, as a simple adsorbent ages, the associated Hg may enter the structure, becoming a "residual" element effectively removed from the dynamic weathering cycle. In simple adsorption from aqueous solution, Hg has features in contrast and in common with the base metals. The hydroxy-cation is the active species in the model for heavy-metal adsorption and this also appears to be true for Hg. However, in contrast with Cu, Pb and Zn, the adsorption is less efficient and is strongly inhibited by the formation of halide complexes, as has been shown by Forbes et al. (1974) (Fig. 12-1). These authors also demonstrate that the adsorption of Hg to goethite is effective at pH as low as 4, allowing it to be trapped subsequent to sulphide oxidation. Whilst many minerals in weathered rocks and soils may each adsorb Hg, the relative efficiency of the hydrous iron oxides (Andersson, 1979) implies that these phases will be the dominant host in most exploration samples. However, the soil organic matter is also of importance and, although the association with Hg has been described as adsorption, it seems more
Mercury
399
100 #
80 -
Cu
/Pb
Co Zn
I:}
.~ so
~.o
cd
-
20 0
t"'
-
z,
. . . .
-
!
5
~ Hg (NO 3 ion)
Hg i 6 pH
i 7
j 8
Fig. 12-1. The adsorption of mercuric ion, Cd. Co, Cu, Pb and Zn on goethite in the presence of chloride ion (reproduced with permission of Academic Press from Forbes et al., 1974, J. Colloid Interface Sci., v.49).
likely that direct chemical bonding is involved, whether by chelation or direct bonding of Hg to S (Jonasson, 1970). Mercury accumulated at low pH by humic substances in soils is important (Andersson, 1979), as it seems likely that the strong bonding involved will promote the oxidation of the elemental form and the resulting stronger fixation will lead to the high background values observed in certain terrains. Thus, the major factors contributing to the distribution of Hg in soils and weathered rock are the soil pH, the organic component, the halide component, the sesquioxide component and the clay fraction.
BEHAVIOUR OF MERCURY IN THE PRIMARY ENVIRONMENT
Low temperature epithermal base-metal deposits As has been previously noted, the high volatility of Hg over a wide range of geological environments has led many authors to speculate on the possibility of broad primary dispersion halos surrounding sulphide deposits. Until about thirty years ago the only published case-history studies of the primary distribution of Hg around ore deposits were to be found in the Soviet literature. The most often-quoted example is that of Ozerova (1962, 1971), who studied the dispersion of Hg around Hg deposits as well as those containing Hg as a trace component. Of these the most pertinent to our considerations are the deposits of South Fergana, Uzbekistan. The deposits of South Fergana, described originally by Ozerova (1962), occur in Devonian to Permian carbonate and terrigenous sediments with minor intercalated volcanics. The Pb-Zn ores are described as being structurally controlled: those deposits
G.R. Carrand J.R. Wilmshurst
400
Pb
1000 p p m I-
I0
'
ppm ~
.
t
w..,j!
'0ooot A l p p m ~-
Metres 0"!~....~
!
~
~
~
BEZYMYAN NOY E DEPOSIT
KARAOT EK
DEPOSIT
.~:: : : : : : : ~ ~ ~ ~ . . : 1000:"
9" ' ' . - ' ' ' " '
9' , ' , '
:..~-- ~ ! . : , ' , ~
Lower Cretaceous sandstones a n d m a r l s ~
_
Hg
Upper C a r b o n i f e r o u s
conglomerates and sandstones
~-~
:.-:: ~;:: :- : :-- :: :~.:.:..:-::'.:..:.: :..:..
.'~--~'.:'.',:L'-C~'-
9~
O ~ ~km
- =-
"
=- "
"
"
=-
"~." :" : .'-" : " .' ."
Lower Carboniferous limestones and $hol~
~ Upper D e v o n i a n " ~:~ dolomite~ and l i m e s t o n e s
Fig. 12-2. The distribution of Hg and Pb in rocks from a traverse across Bezymyannoye and Karaotek deposits, South Fergana (reproduced with permission from Ozerova, 1962).
in carbonates are lens-shaped, bedded or flat pipe-like bodies with forms determined by the intersection of fissures and bedding; those in terrigenous rocks occur in fault planes, in the hinges of anticlines and at specific intersections of structures and host rocks. The deposits are considered to be epigenetic and relatively low temperature in origin; they apparently outcrop. A traverse some 7 km long across two major deposits (Karaotek in carbonate rocks and Bezymyannoe in terrigenous rocks) defined two significant Hg soil anomalies up to 2 km wide and containing up to 30 ppm Hg (Fig. 12-2). In contrast Pb and Cu have far more restricted distributions. Unfortunately, insufficient data on Zn distribution are available for comparison. Although soils were the medium analysed, Ozerova considered that the contained Hg reflected the primary dispersion of the element rather than its secondary dispersion.
Volcanogenic massive sulphide deposits The Woodlawn volcanogenic Zn-Pb-Cu deposit is situated on the flank of a folded Middle to Upper Silurian volcaniclastic pile within the Lachlan Fold belt of eastem Australia (Malone et al., 1975). Primary ore comprises some 6 million tonnes of finelybanded complex massive sulphide ore (average grade 14% Zn, 5% Pb, 1.6% Cu and 88 g/t Ag). Ryall (1979b) compared the distribution of Hg with that of Cu, Pb, Zn, Ag, As and Sb in massive sulphide ore and wall rocks. The Hg content of massive sulphide ore ranges from <100-18,000 ppb and thermal release curves indicate Hg resides principally
Mercury
401
~
.
9150 N
\
ORILL HOLE (in section)
J
ORILL HOLE ( projected ) MERCURY ( PPB; ~ - ~10 PPB]]
J
[
\.
9280 N
Fig. 12-3. Geological cross sections at Woodlawn showing the location of diamond drill holes through ore and country rock with the Hg content of rocks and stringer sulphide minerals; S =sphalerite, C = chalcopyrite, P = pyrite, MS = massive sulphide (reproduced with permission of Economic Geology, v. 74:6, p. 1473, Fig. 2, Ryall, 1979b).
in the sphalerite and tetrahedrite crystal structures. The enclosing country rocks generally contain less than 100 ppb Hg, except where sulphide stringers or disseminations occur and values up to 850 ppb have been registered (Fig. 12-3). Ryall (1979b) concluded that there is no widespread Hg halo developed about the ore independent of the disseminated or stringer mineralisation. Subsequently Van den Boom and Poppelbaum (1980) defined, in rock about 500 m to the west of the Woodlawn ore body, an extensive area of elevated Hg contents, which they considered to result from primary dispersion associated with the Woodlawn mineralising event. However, subsequent work by company geologists (McKay, pers. commun.) would suggest that weak disseminated mineralisation found in this area is associated with rocks lithologically-distinct from the Woodlawn sequence, and probably associated with a different alteration event. Sakrison (1971) briefly described a halo of anomalous Hg in drill core and surface rocks associated with a blind massive sulphide deposit near Noranda, Quebec. The halo
402
G.R. Cart"and J.R. Wilmshurst
apparently extends vertically for several hundred metres and laterally for 100-200 m. No other geological or geochemical information was presented nor was there any discussion of the analytical techniques used. Sharp (1976) analysed chert around the Hidden Creek massive sulphide deposit near Anyox, British Columbia. He showed that, although S dispersed more than 1 km beyond the recognised vent area, Ba, F and Hg show weak dispersion away from the sulphide occurrences. In general, then, the extent of primary Hg dispersion associated with massive sulphide deposits remains essentially untested. Halos almost certainly do exist but the size of these halos relative to those of target and indicator elements remains equivocal.
Sediment-hosted massive sulphide deposits The Howard's Pass (XY) Zn-Pb deposit in the Selwyn Basin of western Canada consists of concordant laminae rich in sphalerite and galena within Lower to Upper Silurian carbonaceous cherty mudstone and cherty limestones. Goodfellow et al. (1983) studied the zonation of chalcophile elements, including Hg, about the deposit. In particular they examined the distribution of these elements through vertical profiles up to one km thick and extending across the Howard's Pass Basin into laterally-equivalent unmineralised rocks. The average Hg content of sulphide ore from the Active Member (140 ppb) is low compared to other sediment-hosted massive sulphide deposits. Within the ore sequences, and in the footwall and hangingwall, the distribution of Hg and Cd correlate strongly with Zn (Fig. 12-4). This pattern is evident in both vertical and lateral directions and indicates that there is no Hg halo that is independent of Zn. The authors concluded that the transport and deposition mechanisms were similar for these elements.
Fig. 12-4. Stratigraphic zonation of ore-forming and ore-associated elements in DDH 42, Howard's Pass (XY) deposit (from Goodfellow et al., 1983).
Mercury
403
Proterozoic sedimentary basins in northem Australia are host to a number of large stratiform massive sulphide deposits, including McArthur River HYC, Mount Isa and Lady Loretta. Each of these deposits has been studied to determine the distribution of Hg relative to Zn and Pb. Ryall (1980) reported that at McArthur River the ores contain about 700-2000 ppb Hg, decreasing to background levels of about 25 ppb in laterallyequivalent unmineralised rocks. Although there is a recognisable primary dispersion halo of Hg, the distribution of Hg correlates strongly with Zn, which has a halo of a similar size. At Mount Isa, where sulphide ores contain significantly higher levels of Hg (150070,000 ppb) than at McArthur River, Ryall (1979c) described an essentially identical relationship between Hg and Zn. Again, anomalous levels of Hg in country rock are coincident with anomalous levels of Zn. The Lady Loretta deposit also contains high levels of Hg (3500-33,000 ppb) and here again a strong correlation exists between Hg and Zn both in the ore and host rocks (Carr, 1981, 1984). Broad, coincident primary dispersion halos of Zn, Pb and Hg occur to the southwest of the major ore accumulation, indicating similar modes of transport and deposition for each of the elements. In summary, there is no evidence that the higher volatility of Hg relative to the target elements, Pb and Zn, has resulted in its wider dispersion during the ore-forming processes of the sediment-hosted massive sulphide deposits described here.
Gold deposits The Drake Ag-Au deposits in northem New South Wales occur in PermoCarboniferous volcanic rocks; mineralisation is considered to result from hydrothermal activity associated with the acid volcanism (Bottomer, 1986). At one of these deposits, Lady Hampton, stratabound mineralisation occurs as disseminations in a volcaniclastic unit exhibiting sericite-clay-quartz alteration. Pyrite is the major sulphide present with minor sphalerite, galena, chalcopyrite, Ag sulphosalts, tennantite and native alloys. In a study at the Commonwealth Scientific and Industrial Research Organisation (CSIRO), Australia, the primary vertical zonation of Hg at Lady Hampton was compared with that of Cu, Pb, Zn, As, Au and Ag. In one section studied, the ore unit is overlain by 100 m of a second volcaniclastic unit (Fig. 12-5). Within this unit Au and Ag were not detected, As, Cu, Pb and Zn occurred at low concentrations and Hg varied up to 400 ppb with a mean of 100 ppb. No trends are apparent in these data and they are considered to represent background populations. Immediately above the hangingwall contact of the ore zone (108 m), Hg, Cu and As show significant enrichment over a distance of about 10 m. The highest Hg values occur in this narrow zone. Towards the footwall, Hg contents progressively decrease while As remains high. Thus while Hg shows strong enrichment in hangingwall rocks, its halo is narrow and coincident with those of As and Cu.
G.R. Carr and J.R. Wilmshurst
404
!
50 r
r't
100
L
150
,oo
200
Cu (ppm)
;.'
~'2'~b'
~ } ~ ;.
Pb Zn (ppm/1000) (ppm/1000)
iogo~koi~ Ag [ppm)
Au { ppm)
~
200 ,-oo
As I ppm)
Hg (log ppb)
Fig. 12-5. Distribution of Hg, Cu, Pb, Zn, As, Ag and Au in percussion drill hole, Lady Hampton ore body, Drake mineral field.
Aftabi and Azzaria (1983) studied the distribution of Hg in ore and host rocks from the Sigma Au Mine, Val D'Or, Quebec. The deposit occurs in deformed Archaean metavolcanic rocks of ultrabasic and tholeiitic composition which are intruded by diorite porphyries and undeformed porphyry dykes. Gold mineralisation occurs in sub-vertical and sub-horizontal veins associated with chloritisation, silicification, albitisation, sericitisation and minor carbonatisation. Background Hg contents in the metavolcanic host rocks averaged about 5 ppb compared with 80 ppb in ore veins and rocks associated with ore. However, the distribution of Hg was found not to be wider than that of Au, Ag, As, Zn or Cu in the various host rocks (Fig. 12-6).
The role o f metamorphism The effect of metamorphism on the distribution of Hg in and about sulphide deposits remains controversial. Ozerova et al. (1975) described the Hg content of massive sulphide deposits in the Pacific and Ural provinces of Russia in relation to the type and environment of deposition of the ores and the post-ore metamorphism. In the Urals, deposits in rocks of prehnite-pumpellyite facies contain between 1000 and 100,000 ppb Hg, those in rocks of greenschist facies contain about 1000 ppb and those in amphibolite-grade rocks contain about 100 ppb Hg. They concluded that during metamorphism, sulphides lose their Hg as a vapour phase, which may travel some distance from the deposit.
Mercury 20 10
405 Ni ppnn Zr pprn
100
208 . I00
~
0 200 100' 0
Cu
ppm
_-
Znppm
20(
.
.
.
.
ASppm
! O(
Agppm
I
38
-
~ppb
A
9
A
Auppm
l/
Im
Jl
z
w
~
~
I
9
9
m
ml
m
9
I
o 9
m
9
9
9
m
I
e
al
,
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: :::::::::::::::::::::
:::':'**~"*:"t*".~
......................... **. .* ..:* ::,..:)..):... .......... . ....... : .... ::: ::::::::::::::::::::::: :" :1::' .:'.:,:..::.,;... * ' ' ~ . . . . . . . . ==================================================..**.*:::: .'-:':'"..'"~i
::::::::::::::::::::::::::::::::::::::
'~:. . . . . . . . . . . . . . . . . . . . . .
:::!:!.:~i!:!!E:J":iHHH~!~!!~iHL:i~x!.:*::.:!..:~.i!!:.!!~!!!.:~!H!iiHHiW.i~i::i!iii::::::::::. (1
10 meters
;:'.:::::
!~:i:i;~.;.>.>
211 ORE
VEIN
C-PORPHYRY G - P O R P HYR'~
Fig. 12-6. Distribution of Hg, Cu, Zn, Ag, As, Au, Ni and Zr along traverse through 38th level, Sigma Au mine (from Aftabi and Azzaria, 1983).
Ryall (1981a) discussed the problems of comparing Hg contents of similar style deposits from different metallogenic provinces. Variations in Hg content can occur between deposits in the same province and with the same metamorphic grade. Thus it is difficult to distinguish differences resulting from primary geochemistry or later metamorphism. Ryall (1981a) therefore examined Hg distribution in present-day sulphide mud and in four Australian Proterozoic stratiform massive sulphide deposits of varying metamorphic grade and compared this with their overall Hg content. He found that in these deposits there is no relationship between metamorphic grade and Hg content or the size of the Hg halo. It appears that some redistribution of Hg may have taken place between sulphide phases during diagenesis and metamorphism, but no Hg halos formed that are independent of primary geochemical halos of target and indicator elements.
406
G.R. Carr and J.R. Wilmshurst
BEHAVIOUR OF MERCURY IN THE SECONDARY ENVIRONMENT The dispersion of Hg in the secondary environment can be considered to result from both solution and vapour transport. However, the relative importance of these two mechanisms for different geological, geomorphological and climatic environments cannot be estimated based on our theoretical knowledge of the behaviour of Hg, but must be deduced from observations of the distribution of Hg, relative to other metals, around weathering sulphide bodies. Whereas studies of soils and weathered rocks surrounding outcropping deposits are likely to indicate the nature and extent of the solution transport of Hg, possible vapourgenerated halos can only be confidently described in relation to buried or geochemically blind deposits. In a major project at CSIRO, Australia, on the behaviour of Hg in the secondary environment, over 40 case-history studies of base metal, Au and U deposits and prospects were carried out (Carr et al., 1986). The project was funded by 15 mining companies operating in Australia through the Australian Minerals Industries Research Association (AMIRA). Case histories were chosen on the basis of those features of the geology and environment that are likely to highlight any distinctive behaviour of Hg relative to the more commonly-determined elements, Cu, Pb, Zn, Ag, As, Fe and Au. The examples used in this section draw heavily on these case-history studies.
Outcropping mineral deposits in dry climates In the CSIRO study, 14 case histories of outcropping deposits in low-rainfall terrains were described. Of these, ten were associated with high-contrast Hg soil anomalies. An example is the Dugald River stratiform Zn-Pb-Cu deposit in northern Queensland. The main mineralisation comprises a major stratiform lode (the Eastern Lode) which has a strike length of about 2.5 km (Fig. 12-7). Sulphide mineralisation consists of pyrrhotite, pyrite and sphalerite, minor galena and traces of chalcopyrite, arsenopyrite and sulphosalts. A smaller, less continuous lode (the Western Lode) occurs to the west of the major lode and chalcopyrite mineralisation has been delineated in the hangingwall. The area has moderate relief. The Eastern Lode outcrops as a continuous ridge up to 20 m above the surrounding areas, and forms a well-defined gossan with an average width of 5 m. In the south of the prospect, Cu mineralisation also outcrops as a gossan (Fig. 12-7). Sulphides are completely oxidised to a depth of approximately 20 m. The weathering products are Fe oxides, plumbogummite, kaolinite, montmorillonite, barite and jarosite (Taylor and Appleyard, 1983). The soils are almost everywhere lithosols. The initial sampling delineated three high-contrast Hg soil anomalies directly over the Eastern and Western Lodes and over rocks midway between the lodes on Line 14.36 N (Fig. 12-8). Later sampling duplicated the central and eastern anomalies, but produced only low-contrast anomalies over the Western Lode. Background Hg contents of soils over footwall limestone and hangingwall slate are in the range 10-30 ppb. Copper is
Mercury
407
u.i o o
o
i.-
ILl
WESTERN
EASTERN
~,,.,. --
l& ~,00N 1 Line I/.,.36N-
I I
]
Slate
Lime I stone
I
1~000 N
DR 46
COPPER GOSSAN ,~
~r
13600N 0
200
~.00m
Fig. 12-7. Location of sample traverses in relation to the distribution of gossans at Dugald River (from Carr et al., 1986).
enriched in soils over and to the west of the Western Lode, but occurs at background levels over the Eastern Lode. Lead forms two very high-contrast anomalies directly over the lodes and in the footwall carbonates to the east. Zinc is anomalous over the Western Lode and in the footwall rocks of the Eastern Lode. The Hg anomalies in soils over the footwall (to the east) are related to supergene enrichment in the carbonate host rocks and anomalies over the rocks between the lodes are related to thin gossanous outcrops in these areas. The variability in intensity of Hg soil anomalies between repeat samplings is considered to result from the highly specific relationship between the Hg content of the soil and the Hg content of the rock directly below, and correlates with variations in Cu, Pb and Zn content. As these rocks show strong variability over very small distances due to the stratiform nature of the mineralisation, so too do the soil metal contents. Consecutive surveys cannot hope to sample exactly the same locations. The relationship of the Hg soil contents to those of Cu, Pb and Zn is also readily explained by the geology. Taylor and Scott (1983) showed that Zn was mobilised early in the oxidation of the ores and redeposited in the footwall carbonates, thus explaining the high Zn soil contents over these footwall rocks. Lead, on the other hand, is fixed within phases such as plumbogummite and jarosite, which remain stable even in the surface gossan. It is apparent that in this environment Hg has also been stabilised within the gossan at the
408
G.R. Carr and J.R. Wilmshurst LRDY
LORETTR
1950N
200 Q.
ff
/
lee 0 6
o.
q~
"
300
o.
200
O.
1oo o 30
o.
20
,,
&
,
i
I
.
.
.
.
.
15
~
~Jo n GL v
5
200
100
o~
3:
0 0
TOPO
SLOPE
200
400
600 (mE)
~ -:-t-:--:
800
1000
--
Fig. 12-9. Distribution of Hg, Cu, Pb, Zn, As and Ag in soils along line 1950N, Lady Loretta (from Cart et al., 1986).
time of oxidation of the ore, with the result that very little lateral dispersion has occurred. A similar relationship between Hg and target elements has been defined at the Lady Loretta deposit, also in northern Queensland (Loudon et al., 1975; Carr, 1981, 1984). Here the ore body consists of a stratiform lens, up to 40 m thick, hosted within fine-grained sedimentary rocks occurring in a steep-sided basinal structure. The ore outcrops on the western limb of this structure; however, to the east, Zn and Pb sulphides pinch out rapidly, giving way to massive pyrite with interbedded dark, fine silt-shales. The ore-bearing rocks and their pyritic equivalents stand out as ridges, remnants of a
Mercury
409 DUGRLD
RIVER
14.36N
15eeO
1978---
618800 c
N
5800 8
. . . . .~
_
500 Q.
Q"
I
I
I
I
i
i
250
I1.
0
J 25~r
.'",
,,,
,,AA
01
le4se
TOP0 SLOPE
_
lesse
GOgSRN'~ esse ~
~
ERSTINGS (m) =
t
_ _~.--~"~'"
,,. _
--
9
m
l lese ---
Fig. 12-8. Distribution of rig, Cu, Pb and Zn in soils along line 14.36N, Dugald River (from Carr et al., 1986).
Mesozoic peneplain which underwent deep weathering and lateritisation during the Tertiary period (Cox and Curtis, 1977). Total relief in the area is 70 m. There is more than 50% outcrop of the ore-bearing rocks along the ridges. Total leaching of carbonates and oxidation of sulphides extends to depths of 50-300 m. The deepest weathering is associated with permeable zones near the synclinal axis and along faults. Soils on and adjacent to the synclinal ridges are lithosols. Gossans collected over the outcrop of the ore have a wide range of Hg contents (20-40,00 ppb) and an arithmetic mean of 730 ppb Hg, which represents about 95% depletion from fresh ore (range 3500- 33,000 ppb, mean 21,000 ppb). There are significant differences in the distribution of the elements Hg, Cu, Pb, Zn and Ag in soils (Fig. 12-9). At 400 m E only Hg and Ag occur as residual concentrations over the ores. The results for Cu, Pb and Zn are consistent with those of Cox and Curtis (1977), who showed that although Zn was strongly dispersed in all situations, Pb formed both residual and hydromorphic anomalies. It is considered that anomalous levels of Hg, Pb, Ag and Zn within the synclinal axis (550 m E) and east of the eastern gossan (weathered pyritic rocks) result from hydromorphic dispersion either along faults which intersect the ores or through surficial processes. In the above examples it is apparent that the secondary dispersion of Hg is no greater than, and commonly less than, that of the target elements. This was the case in most of
410
G.R. Carr and J.R. Wilmshurst
the 14 case-history studies of outcropping deposits in dry terrain conducted in the CSIRO project. Some examples of broader Hg anomalies do exist, but these can be related to primary lithological control rather than dispersion. One such deposit is at Kookynie, in the Eastern Goldfields of Western Australia, where Archaean tholeiitic basalts overlain by acid volcanics consisting of rhyolitic flows, breccias and tufts outcrop as a series of steep-sided domes. A black shale unit containing minor cherts occurs along the contact with the basalts and contains disseminated-to-massive sulphides. Diamond drilling has indicated the presence of chalcopyrite mineralisation in both basalt flow-breccias and the shale. In addition, a massive pyrrhotite lens has been intersected in the shale. No economic mineralisation has been discovered. The area is one of low relief and is covered by residual lithosols to red-brown loam soils. Outcrop of all rock types in the area is good. A large gossan outcrop (Fig. 12-10) is considered to be the surface representation of the massive barren sulphides encountered in drilling. The depth of weathering is unknown to the authors, although it is probably at least 50 m. The distribution of Hg in the soils is significantly different from the distributions of Cu and Zn (Fig. 12-11). Whereas anomalous Hg levels occur in soils both at the contact of the basalt and the acid volcanics and over the basalt itself, Cu and Zn anomalies are restricted to the contact zone. This difference is interpreted as resulting essentially from primary lithological control. Whereas Cu and Zn
w 55 000 N
0
I~
2~
:
3~ -
....
~ , i
5~m i
';" : " ",.
: ~,VOLCANICS", : , , ,,,?',,, .... ;
~
Line I I'. : ,,,?, ,
~r.~,~,-.,~,:~: " " , . , , < ; .;,.:_'- ; . , , . ,
5;~!;:j,!~:1:.:.:.'i),b' . s'sN. ~?-~.A~_ , ,.",, : ~:)LCANICS: "" :l
Fig. 12-10. Local geology and location of soil sample lines at Kookynie, Western Australia (from Carr et al., 1986).
Mercury
411 KOOKYNIE LINE
I
2oe s o. o. .,'9
log ..
0
.
. . . .
,. . . . .
|
9. . . . . . .
,
20
.Q n
200
Ix
leo
d IO
"
t flCID VOLCflNICS"~
0
2gg
^
4 O0
A |VOLCRNICS
/\
flog
BBO
(m) TOPO
SLOPE
-:
~
I OOO
=
Fig. 12-11. Distribution of Hg, Cu, Pb, Zn and As in soils over Cu mineralisation and host rocks at Kookynie, Western Australia (from Carr et al., 1986).
are strongly enriched relative to the acid volcanics in the black shales and basalt near the contact, Hg is enriched throughout the basalt as well as in the black shales. In fresh basalt, such Hg occurs within sulphides and upon oxidation is fixed by the abundant Fe oxides. These CSIRO examples do not stand alone in indicating a common highly-specific relationship between outcrop of weathering sulphides and the extent of anomalous levels of Hg in soils. Warren et al. (1966) compared the Hg, Pb and Zn contents in soils associated with base-metal mineralisation in British Columbia, Canada, and concluded that because anomalous levels of Hg were associated with anomalous Zn and/or Pb, there was no advantage in using Hg as a pathfinder. Friedrich et al. (1984) determined the distribution of Hg, Ba, Cu and Zn associated with Cu deposits of the Troodos Complex, Cyprus, and, on a number of profiles across faults that intersect blind mineralisation, found anomalous Hg soil contents coincident with high-contrast Zn, Cu and Ba anomalies (Fig. 12-12). Broad halos of Hg in soil and surface rock have been described elsewhere in the literature, but unfortunately no comparisons are made with other elements. Some
412
G.R. Carr and J.R. Wilmshurst
Fig. ! 2- ! 2. Distribution of Hg, Cu and Ba along a traverse over the 40 m deep Sha mineralisation, Cyprus: (a) rocks; and (b) soils (from Friedrich et al., 1984).
examples are presented by Williston (1964), Friedrich and Hawkes (1966) and Wu and Mahaffey (1978). A case-history report published by the Jerome Instrument Corporation detailed a broad Hg anomaly associated with the gossanous outcrop of a steeply-dipping massive pyrite deposit at Lake Yindarlgooda, near Kalgoorlie in Western Australia. The distribution of anomalous Hg soil contents over the gossan and hangingwall rocks was interpreted as resulting from a combination of a residual anomaly and Hg vapour transport to the soil from the downdip weathering of the sulphides (Fig. 12-13). This deposit was also studied by Carr et al. (1986), who found the distribution of Hg correlated closely with that of other elements and that narrow, high-contrast anomalous metal contents in soils over the hangingwall of the major gossan could be related to outcropping, narrow gossanous layers (Fig. 12-14).
Outcropping mineral deposits in wet climates In the CSIRO study, six deposits in areas of moderate-to-high rainfall were studied. Five of these deposits occur in the Mt Read volcanics of western Tasmania, Australia. These volcanics are of acid to intermediate composition and consist of a central belt of massive rhyolites and dacites (together with minor andesites) flanked to the west by a
Mercury
413
Fig. 12-13. Distribution of Hg in skeletal soils at the Lake Yindarlgooda gossan, Western Australia (from Jerome Instruments Corporation technical sheet).
volcano-sedimentary sequence and to the southeast by mixed volcanics and volcaniclastic conglomerates of the Tyndall Group (Corbett, 1981). They are host to a number of important massive sulphide deposits, including Mt Lyell, Rosebery, Que River and Hellyer, and have been the subject of intense exploration. The combination of high rainfall, dense vegetation and steep topography makes access and logistics for exploration in the region difficult. The climate is typical of west-coast cool-temperate regions with year-round rainfall averaging over 2500 mm in the lower regions and about 3000 mm along the West Coast Range. This range rises to an elevation of 1275 m and is characterised by steep gorges and valleys, which dissect a glacially-modified plateau surface (Reid and Meares, 1981). Recent fluvioglacial sediments cover parts of the area to a depth of several metres. At each deposit sulphides are actively oxidising near the surface and in some cases can be recognised in outcrop. High-contrast Hg anomalies with no obvious lateral extent occur at each deposit directly over the sulphide-rich rocks, for example, at Que River (Fig. 12-15). The distributions of Pb, As and Cu are also closely related to the outcrop of these rocks, but Zn may form broad hydromorphicallydisplaced anomalies. The Currawang massive sulphide deposit in eastern New South Wales occurs in Siluro-Devonian pillowed andesitic lavas to the north of the Woodlawn deposit. The geomorphic regime is more mature than in western Tasmania and the rainfall is lower. Lower rates of erosion result in deeper penetration of sulphide oxidation, although fresh sulphides have been observed in outcrop as armoured residues within highly silicic rocks. Small gossanous outcrops represent minor apophyses of the main lode, which is blind. Lead forms a narrow, high-contrast residual anomaly; Cu, Zn and Hg form equally high-contrast anomalies, but show evidence of hydromorphic dispersion down slope (Fig. 12-16).
414
G.R. Carr and J.R. Wilmshurst ROCKY
DAH
LINE
I
200 el
! 00
,5 9
i
#
40 m
.11
,.,
I o0
Ok
""
50 0 300
-~
100
t,,l
Q.
.
2?eB
.
.
3600
.
.
.
.
.
i
15~
i
0 TOPO
SLOPE~'~-
e~e ~
~
~ (m )
\\\ 400
GOSSRN
s00
"--4"--
Fig. 12-I4. Distribution of Hg, Cu, Pb, Zn, Fe and As in skeletal soils at the lake Yindarlgooda gossan, Western Australia; CSIRO data are compared with the data presented in Fig. 12-13 (from Cart et ai., 1986).
In an exploration area in northern Peru, igneous rocks cover an area of several square km within a metamorphosed series of quartzites and limestones. The younger sequence consists of ignimbrites, tufts and tuffites of approximately 1000 m thickness. The metamorphism produced skarns, which are connected with the ore body. A porphyry Cu deposit occurs in the metamorphosed rocks and is characterised by intense hydrothermal alteration (quartz-kaolinite-sericite) which makes determination and classification of magmatic and sedimentary rocks at the surface very difficult. A zone of propylitic alteration can locally be followed up to 500 m into the andesitic-dioritic wall rocks.
Mercury
415 OUE RIVER LINE 7488N
--
4@0r lI
"
J
o-',
p
HORIZON-----
C HORIZON - - -
,
i';
:'
B
~
A
l
,"/~iA'
~
.
,,
3e
~- 20
2000[ [
"" 1500
,;', ,' .
1000|
:~
000 468o
TOPO SLOPE
"
',
SULFIDE O / C
"~
5~ee
"'"
54~9 EAST
I NGS
- - ~
soee
(m )
Fig. 12-15. Distribution of Hg, C, As, Cu, Pb and Zn in the A horizon and C horizon of soils along Line 7400N, Que River, western Tasmania (from Carr et al., 1986).
Indications of ore are rare at the surface, where Cu-oxide minerals occur sporadically. The ore body consists of a primary ore and a supergene enrichment zone, which are separated by a transition zone. An investigation of the area by Van den Boom (written commun., 1986) included a Hg survey with more than 800 samples of altered rocks collected from the surface. The Hg concentrations in the samples showed maximum values of more than 2000 ppb with a regional background of 150 ppb. The frequency distribution of Hg is characterised by two populations; the second, consisting of 250 samples is considered to be anomalous (Fig. 12-17). The ore body, at approximately 50 m depth, generated a distinct doughnut-shaped anomaly of Hg at the surface which corresponds with the projection at surface of the 0.8% Cu isoline (Fig. 12-18). The continuation of the anomaly to the west might be in response to skarn mineralisation adjacent to the porphyry.
416
a.
G.R. Carr and JR. Wilmshurst
-f
AA
CURRRklRNG L I N E
2 HN03 HCI
-----
---
1000 E
500
0
2oor
5 3
1oI E00
.D g. Q.
100 A 3000 ERSTINGS
2800 TOPO
SLOPE
3200
(m)
---I---
,---l---
Fig. 12-16. Distribution ofllg, Cu, Pb and Zn in soils sampled at a depth of 150 mm, Currawang, New South Wales (from Carr et al., 1986).
POPULATION I N = 563 MEAN: 110 PPB MEAN LOG.: 2.04 STAND. DEV. LOG.:
POPULATION
FREOUENCT
0.34
DISTRIBUTION
HG
2
N = 250 MEAN:
420
PPB
MEAN L O G . : STAND.
LO
!
.
2.62
DEV.
LOG.:
INTERV.:
0
.
1|
l|ll
le||
tPPO)
Fig. 12-17. Frequency distribution of log-transformed Hg concentrations in rock samples from Peru and separation into two populations.
Mercury
417
3O00
/
-, I," , ~ ,
x'xk
Illitezone
~.~.
~rlclt~
~j
I~E~dlrdte~ ,'~,.~ Hg-~'~rnah/(concer~ra'aor~ >42op#:)) -=,=-- 0.8%Cu Isollne
0 1000
I 2ooo
I 3o00
I
4000
Fig. 12-18. Distribution of rig in rock samples over an altered zone and ore body in Peru.
Blind and buried mineral deposits in dry climates In the CSIRO project, four deposits classified as blind by company geologists and located in areas with dry climates were studied. Significant Hg and base-metal soil anomalies were detected at three of these deposits, indicating that the deposits or their primary halos outcrop. The fourth locality, Elura, in central New South Wales, is a large massive sulphide deposit (27 million tonnes grading 5.8% Pb, 8.4% Zn and 130 g/t Ag) which occurs as a steeply-plunging elliptical body (Taylor et al., 1984; Wilmshurst, 1980). The top of the body divides into two apophyses, only one of which (the southern apophysis) reaches the surface (Fig. 12-19). The Hg content of the sulphide ore is high. In two drill-hole intersections, the Hg contents ranged from 1000-200,000 ppb with an average of 53,000 ppb in one and 11,000 ppb in the other. The results of a soil survey over the southern aphophysis indicate anomalous Hg levels to the west of the outcrop of the ore body, but not over it (Line 50 740N, Fig. 12-20). The Hg contents of soils along the northern traverse over the blind apophysis of the ore body are at or below the threshold except at one sample site from 300 mm depth (Line 50 880N, Fig. 12-20). These patterns were matched by the distributions of Cu, Pb, Zn, Ag and As.
418
G.R. Carr and J.R. Wilmshurst 'l]ossan trench'
SOUTH
E25
i
NORTH
E28
Water table Base of complete" ?~ weathering ?~
~..Im i" ? r !
q
Base of partial weathering
"?~ No2 ~
sub-level
Massive sulphide ore
Fresh country rock
Pyritic
No. 2
level
Pyrrhotitic
0
i
50 m V&H
I
Fig. 12-19. Geological cross section through the Elura ore body showing the weathering profile (from Taylor et al., 1984).
Out of a total of six case histories of buried deposits in dry climates in the Northern Territory, Queensland, New South Wales and Westem Australia, in only one instance was a possible vapour-generated halo detected. At Woodcutters a carbonate-hosted Pb-Zn deposit is represented at surface by an iron-rich gossan which, in the south of the prospect, has a shallow cover of transported soil. Soil and soil gas measurements for Hg along two lines did not reveal the presence of the underlying gossan. The Pegmont Zn-Pb deposit has also weathered to an iron-rich gossan, which is buried locally by up to 10 m of sediments. A soil survey over these sediments (Line 3400N, Fig. 12-21) revealed only background levels of Hg, in contrast to the strong residual anomalies further to the north (Line 4920N, Fig. 12-21). The failure to detect anomalous Hg above these buried deposits can at least in part be related to features of the weathering environment. In some of the case histories it was determined that the present water table was above the depth of oxidation of sulphides with the result that active sulphide oxidation had ceased. In others it was apparent that the low levels of Hg restricted the development of a halo.
Mercury
419
ELURR LINE 58 748N
ELURR LINE 50 890N 388mm
~[ O, 15BB
,
2000
-.
,
25~
ERSTINGS
(m)
3~NND
.
350e
ELURR LINE 58 748N
L
Z S ~ I -N D o L i SULFIDES 2000
2500
ERSTINGS (m)
3808
ELURR LINE 50 850N
i
20
. . . . . . . . . . .
2e
d
OI
~3~
~3'~e
4B
.
.
.
.
.
.
.
Z
! BE
235B
.
61I
z~ 9
9
24B0
....
2458
ERSTINGS
(m)
2s'ee
BLIND SULFIDES \ e4~ e4~e ~s~ ERSTINGS (m)
z~'~e
Fig. 12-20. Distribution of Hg. Cu, Pb, Zn, Ag and As in soils along traverses across the Elura gossan and the blind apophysis (from Carr et al., 1986).
The successful case history was the Jabiluka U-Au deposit in the Northern Territory. The Jabiluka One and Two deposits occur in Lower Proterozoic metasedimentary rocks of the Cahill Formation, which are partially exposed in a "window" bounded to the northeast by the overlying Kombolgie Formation and to the south and west by faults (Hegge, 1977). The Jabiluka Two deposit is covered by sandstone of the Kombolgie Formation which is itself covered by sandy colluvium. The soils developed over Jabiluka One and the western part of Jabiluka Two are predominantly sands, which are coarse grained and unconsolidated away from the floodplain. X-ray diffraction analysis of the sands indicates that they consist of quartz with minor clays, mainly kaolinite. Weak profile development is present in soils adjacent to the floodplains. The soils are generally above flood height and, although they may become waterlogged during the rainy season, their sandy nature ensures that they drain rapidly. The Jabiluka Two ore body contains
420
G.R. Carr and J.R. Wilmshurst
PEGMONT
LINE
4920N
1~176176 t PEGHONT L I N E
34OON
n 4
~
20
3
a.
10
N
L.J
4700
4800
4900
5000
EFISTINGS (m)
5
100
A
5
d
so
2
5
.
5
|
..,
i 2O0
o. '1
100
GOSSRN
1" I~ 4250
4450
4G50
ERSTINGS TOPO SLOPE
-- ; ~
4050
(m) -- f - f - - -
;
....
-
Fig. 12-21. Distribution of Hg, Cu, Pb. Zn, As and Ag in lithosol soils over outcropping gossans and distribution of Hg in transported soils over buried gossan at Pegmont, northwest Queensland (from Carr et al., 1986).
appreciable Hg, with uraninite ores averaging up to 1950 ppb (Ryall, 1981b; Ryall and Binns, 1980). Higher Hg concentrations are present in the Au-rich western part of the ore body where native Au contains up to 450 ppm Hg and uraninite concentrates up to 11 ppm. The zone of oxidation of the Jabiluka One deposit extends down to about 15 m. No information is available on the oxidation of the Jabiluka Two deposit. Soil and soil-gas surveys were carried out at Jabiluka late in the dry season of two different years. Soils were sampled at 300 mm depth and soil gases were collected onto SIROSORB (Wilmshurst and Ryall, 1980) using techniques described elsewhere in this chapter. Mercury in soils over the blind Jabiluka deposit generally ranges from 10-20 ppb (Fig. 12-22). Peripherally, concentrations up to about 40 ppb are evident, producing a weak "rabbit-ears" effect, probably as a result of variations in soil mineralogy. The Hg content
Mercury
421 JRBILUKR LINE 4
2 i \ o~
1981
-r
8
5 3 -18 58 JD O. n.
25
m "I" g
8
IBB
28B
3B8
(mN)
Fig. 12-22. Distribution of Hg in soil and soil gas over faulted Au mineralisation at Jabiluka Two, Northern Territory; soil gas was sampled at the end of the dry season (from Carr et ai., 1986).
of soils does not reflect the presence of a fault that intersects the Jabiluka Two deposit at depth. However, over the Au-rich mineralisation, Hg in soil gas measured in the first survey rose to a sharp peak of 8 ng/L in the vicinity of the fault, with less than 2 ng/L away from the fault. In the second survey the soil-gas anomaly was reproduced even though absolute Hg levels were lower. This may have been due to variations associated with the barometric cycle, the response being lowest at the peak of the pressure increase and/or when gas generation (through sulphide oxidation) is at a minimum. In addition, improved gas sampling techniques used in the second survey considerably reduced the detection level and revealed lower background levels, perhaps indicating background contamination in the earlier sampling. The Hg vapour anomaly detected over the Au mineralisation at Jabiluka is significant in that it is the expression of a halo detected through transported cover which is essentially sand derived from the overlying Kombolgie Sandstone. There is no soil anomaly, a feature which we interpret as due to the relatively low surface area of the sandy soil, its low organic content and to the absence of minerals capable of entrapping and hosting Hg. Studies at CSIRO have shown that quartz sand has a very low sorptive capacity for Hg. In contrast to the very restricted size of the soil-gas anomaly at Jabiluka, Zonghua and Yangfen (1981) describe broad anomalies associated with a buried skarn Cu deposit near Shanghai, China. The ore body is confined to the contact zone of granodiorite and Palaeozoic and Mesozoic sedimentary rocks. The major sulphide minerals in the ore body are chalcopyrite, pyrite and molybdenite. Part of the deposit has been oxidised to limonite. The area is covered by 140-180 m of alluvium. A soil-gas survey of the area was complemented by a multielement study of soils including the determination of Hg. Significant anomalies of Hg in soil and soil gas occurred above the buried sulphide
422
G.R. Carr and J.R. Wilmshurst
~(a) ~'m1 5
~
~1~ 5 "1"
[
~
0
|
I
l
10
20
3O
0
10
9
2
20 Metres
3'0
I~ Alluvium I
~ e -body zn F Fig. 12-23. Distribution of Hg over the Naking Pb-Zn mine: (a) near-surface atmosphere; (b) soil gas (from a CNNC study).
mineralisation, with peak values of 0.4 ng/L in the soil gas compared to a threshold of 0.015 ng/L. The multielement study gave no indication of the underlying sulphides. The results are surprising considering the sampling method involved passing the soil gas through water before the Hg was concentrated on Au wire. Similar anomalies over buried ore deposits have been reported by other Chinese workers (see Chapter 13). Using a gold-film detector, workers at the Chinese National Non-ferrous Metals Corporation (CNNC) analysed Hg in soil gas and near-surface atmosphere at the Naking Pb-Zn mine, where stratabound ores occur in fractures in Permo-Carboniferous strata. The region is covered by about 30m of allochthonous overburden. Mercury anomalies of up to three times background occur in both soil gas and near-surface atmosphere above a fault that intersects the ore (Fig. 12-23). These are coincident with higher-contrast Hg soil anomalies, but corresponding Pb and Zn data are not available.
Blind and buried mineral deposits in wet climates Four case studies in areas of high rainfall were carried out in the CSIRO project. In Tasmania, three small stratiform massive sulphide occurrences covered by fluvioglacial
Mercury
423 HENTY
FRULT
ZONE
IN
49.
1000 l ,-"
--.
o
..
.
.
. . . . . . . . . . . . . . . . . . . . . . . . . . .
,
.-~_- ~-,
d _
4ooi
1200
1300
BURIED
A
1400 EAST INGS
( ~t)
1500
! 600
Fig. 12-24. Distribution of Hg, Cu, Pb, Zn and Fe in the A horizon and the B/C horizon of soils above buried massive sulphide mineralisation at the Henty Fault Zone, western Tasmania (from Carr et al., 1986).
sediments were sampled. Levels of Hg in soils at each of these deposits fail to indicate the presence of the underlying oxidising sulphides. At the Henty Fault Zone the Hg contents of fresh sulphides are relatively low, averaging about 600 ppb. Soil was sampled from both the A and B/C horizons, with Hg being concentrated in the A horizon (Fig. 12-24). Possibly-anomalous levels of rig in the A horizon are coincident with Zn and Cu anomalies, which strongly suggests a mechanism of enrichment other than vapour transport. Also, analysis of samples from overburden profiles (fluvioglacial deposits and overlying soils) above the sulphide lens do not indicate the formation of vapour-generated Hg halos (Fig. 12-25). Similar patterns are associated with sulphide veining in a road-cutting near the Bastyan dam. Mineralisation is restricted to several narrow sphalerite-rich veins over a stratigraphic interval of about 0.5 m. High levels of Hg (10,000 ppb) associated with the veining are not reflected in the overlying tills and soils (Figs. 12-26 and 12-27). The failure to detect anomalous Hg levels in permeable fluvioglacial sediments directly overlying weathering sulphides in these Tasmanian case histories can be directly attributed to the high rainfall. It would seem that the soluble species generated during oxidation, including Hg, remain in solution and are removed by aqueous flow through the till. A small proportion is retained near the base of the profile, presumably largely
424
G.R. Carr and J.R. Wilmshurst
'"I 4~
o i.
a.
3"f/
wj 'if 9
9
~
-
il
.-
9
•
H~ 0
~
0 m. tiC" 13- :} u
!
o ,
,
3
"ia. a.
18
.....
9
/
2
8
DEPTH
(
Fig. 12-25. Element distributions in two profiles from bedrock to the soil A horizon in a costcan across the stratiform sulphide lens at the Hcnty Fault Zone (in Profile 1 Ag is below the detection limit, in Profile 2 As is below the detection limit); S = soil, T = fluvioglocial sediment, B = bedrock (from Carr et al., 1986).
within Fe oxides. Some Hg may well be transported in the vapour phase during dry intervals, but any anomalies so produced are too subtle to be detected. The fourth case history in this group is the Wet Lagoon prospect near Goulburn, New South Wales, where the distribution of Hg differs significantly from the distribution of other trace elements. A small Zn-rich stratiform sulphide deposit occurs within an acid volcanic sequence 9 The lodes, which in drilling have been intersected at a depth of 100 m, project to the subsurface beneath 10 m of transported clay and organic-rich sediments of the Wet Lagoon. Surrounding outcropping acid volcanics contain disseminated oxidising sulphides with significant residual soil anomalies of Hg, Cu, Pb, Zn and As (Fig. 12-28). The nearby lagoonal sediments, flooded with hydromorphically- and mechanically-transported trace elements derived from the volcanics, show very high background levels. No significant contrasts in Cu, Pb, Zn or As occur across lagoonal sediments. Mercury has only a slightly elevated background level in the lagoon, but superimposed on this are very high-contrast, narrow anomalies over the projected mineralised intervals (Fig. 12-29). Although active oxidation of sulphides occurs in the
Mercury BOO
425
~,00 r
*~*
Hg
x ~ x
ZN
Ioo
oo o '-.,.,-.....~..o -
2oo
..,.......,.
o ~ o
--
o
x\\
_ _ L _3
.~ . ~
~.oo
. _ ~ . . ~ L--=-:~'~"~
0
Hg (ppb)
~00
a
I
o. zoo
/I
TILL 2
/
'~-- -- "\ \
',\x/ i
/
""
o
TILL 1
x ....... " " -K
,~176 t
0
,
t.,.
"
4"
9 TILL 1
~o
*
*---20m SOlt. ~)
800 Zn
~-.--~.~-"""
0 ~ , .~ ' ~
0
4 ;
,
'
....
]
BID O(.K
SCALE L . . . . --.a
~ ;I 9
V,/~ Ot I
Fig. 12-26. Mercury and Zn contents of soils, fluvioglacial cover and bedrock at the Bastyan damsite road cutting (from Cart et al., 1986).
nearby acid volcanics, the lagoon soils are waterlogged below a depth of one metre, even when the surface of the lagoon is dry. Thus it is unlikely that there is a present-day Hg flux through to the soils from the oxidation of underlying sulphides. It is, however, probable that over the period of formation of the soils the lagoon dried out completely, allowing the movement of Hg through the cracked, clay-rich horizons. Any such Hg that adsorbed onto or complexed with the highly-adsorptive soil components, especially in the more organic-rich upper layers, would remain despite later flooding of the lagoon. An attempt was made by Van den Boom (written commun., 1986) to test the application of Hg geochemistry for the detection of shale-hosted pyritic ore bodies near the village of Lohrheim, Germany, where an unsuccessful shallow-depth soil sampling survey had previously been conducted. The exact location of the mineralisation was known from drilling completed several decades earlier. The main ore body is situated at a depth of about 70 m in the north of the area and a second ore body occurs at a depth of more than 100 m in the south. To test Hg geochemistry, soil samples were taken from 3 m depth using a powered drill and the <63 mm fraction of samples was analysed using a
426
Iml O-
1 -
G.R. Carr and J.R. Wilmshurst
1.0
80
1001ppml
A_~
2
/,
P_bb
1000
2000
i
T
9
w
6 Ippml ],a3 3000 (ppm) ~ l-
,
9
--
T
~n
SOIL _ (Aol
//// /
TILL 3
i.//
x 2-
TILL Z
&-
-
2000 ~00 9
3009 lppm ) 6r./J lop b I
l
r-
1
I/ 9
t/
7' )~ 0.7'S .--~x
Cu
B( OROCK
9
,
/
/
TILL I
200
9
/'t
3-
1000
1250 As
"---~.~./o
Pb I
x
___
x .~--~ 923S0
.x2.2% -"Pc I0 000
Fig. 12-27. Element distributions in a vertical profile from bedrock containing sphalerite veins through to the soil layer at the Bastyan dam-site road cutting (from Carr et al., 1986).
portable HGG3 atomic absorption spectrophotometer. Mercury in the soil samples occurs in two populations with a threshold for anomalous values of 70 ppb (Fig. 12-30). The main ore body, situated in the northwest comer of the area and intersected at a
o
I
9
~/.~~ ~
Thm
~ Pb
I ~ 100- 500 ppm SO0 - 1SO0 ppm m isoo- so00 ppm
sulfide
project
/-
within
OON
~.
~/-
lenses
to surface these
zones
" Line 128900
-- Line 12e,600 9
Fig. 12-28. Location of sample traverses at the Wet Lagoon also showing Pb geochemistry in surrounding residual soils as determined by North Broken Hill Ltd; to the south of the lagoon, Pb contents were determined on soils from the base of shallow auger holes to bedrock (from Carr et al., 1986).
Mercury
427 HET
LRGOON
LINE
128
988N
leo
o I Bo
5O
5O
o
3O
____
20
LIMIT OF LFK~O0~ SEDIMENTS
IO g
^
~A
IN
245
7411
rl
,
P2S
5o
B ~"v'~ B
~.,,,,
,IN
....
6OO
12N
(m)
Fig. 12-29. Distribution of Hg, Cu, Pb, Zn, As and Fe in lagoonal sediments and surrounding residual soils along a traverse across the Wet Lagoon prospect (from Carr et al., 1986).
depth of 70 m, is not shown by the distribution of Cu, Pb, and Zn in the soil samples, whereas the distribution of Hg shows three anomalous areas, that in the northwest being the projection of the main pyrite ore body (Fig. 12-31). The southern Hg anomaly may correspond with mineralisation encountered at more than 100 m depth.
SAMPLING MEDIA
Atmospheric air With the introduction in the late 1960s of very high sensitivity instrumentation for the analysis of air and soil gas for Hg, it became possible to determine Hg levels in the field. Air surveys were commonly carried out from vehicles mounted with a preciousmetal collector. Mercury within measured volumes of air passing over the collector amalgamated with the precious metal, which was subsequently heated to release the Hg for analysis. A less sensitive alternative method was to pass air directly through a Hg analyser without pre-concentration.
428
G.R. Carr and JR. Wilmshurst
m
.... i ........
~--
i
'
......
I ......
lO
9 -1-
-
.
.
.
.
.
.
.
l
.........
r
lO0
Fig. 12-30. Bimodal frequency distribution of log-transformed Hg concentration in soils, Lohrheim, Germany (horizontal axis in ppb Hg).
McCarthy (1972) reviewed publications on the use of Hg vapour (and other volatile components) as a guide to ore deposits. He quotes Soviet workers who detected high concentrations of gaseous Hg in air over Hg deposits. Results from both the former USSR and North America of gas surveys over base-metal deposits (including porphyry Cu deposits) did not indicate significant anomalies; it was suggested that in these instances primary Hg concentrations are too low to produce observable anomalies. McCarthy (1972), however, quotes a written communication from R. Gedde, who stated that Hg-in-air anomalies were detected over Cu, Ni and Zn sulphide deposits in Australia using a truck-mounted Scintrex Hg spectrometer. Robbins (1973) quotes a personal communication from J.G. Baird, who had detected anomalous levels of Hg in air over the Mons Cupri Cu-Zn sulphide deposit in the West Pilbara of Western Australia, again using a truck-mounted spectrometer, lJnfortunately the deposit has been intersected by shafts and underground workings and the site is contaminated with excavated material. Rowntree and Mosher (1976) reported that a vehicle-mounted "Hg sniffer" survey over the Jabiluka One deposit failed to produce Hg anomalies related to mineralisation. There are a number of well-documented examples of the use of Hg vapour related to geothermal studies (e.g., Phelps and Buseck, 1978; Shiikawa, 1983; Openshaw, 1983; Zhu et al., 1986). In many of these Hg, together with the volatile elements As and B, are shown to be useful pathfinder elements in geochemical exploration for geothermal resources. In our opinion there is insufficient evidence to indicate that atmospheric air sampling for Hg is a successful exploration technique, except perhaps for Hg deposits and
Mercury
429
geothermal areas. In other geological situations, discrimination between background and anomaly is required close to the level of sensitivity of the analytical equipment, and there are thus many possible sources of error.
Fig. 12-31. Distribution of Hg, Cu, Pb and Zn in soils at Lohrheim, Germany.
430
G.R. Carr and J.R. Wilmshurst
M
I
I
I 6
I
I I I NOON
I
I 6
f
r
I
I
M
I
I 0
I
I ~
\
-
I
I
I
I 6
I
"- \
, 24.2
"J'(P "~
li
[_l
i_ _ ~
I , ~I 0 t
5m~sm'nO~
I
t
I
i
t
1
I
k
l
l
l
I,
,
,
28.97
.-,,
9
~
',,~,,,,~ 9
i O
l
% ..i 24.0 24.1
BarcermbtC pre~um
-~'~"
50"
M
"I
-
i
I
A
\
| 12
i
A,
/ ~,
J
28.8"/
n m e ~
Fig. 12-32. Influence of meteorological parameters on Hg values measured in soil air (from McCarthy et al., 1969" Kromer et al., 1981).
Soil gas Determination of Hg in soil air can be carried out either statically, by using collectors in plexiglass tents (McCarthy et al., 1969), or dynamically, either by pumping the soil air into a collector for later analysis or by pumping it directly into the analyser. A Hg collector consists of Au or Ag with a large surface area. Gold wire or shavings have commonly been used; at CSIRO a diatomaceous earth collector (Chromosorb) coated with Au was developed (SIROSORB, Wilmshurst and Ryall, 1980). This collector has a very large surface area and very good adsorption characteristics. A number of meteorological factors affect the results of soil gas sampling for Hg. Table 12-III shows the effect of heavy rain on the Hg content of soil gas. In addition, barometric pressure and air temperature affect Hg emission from soil (McCarthy et al., 1969; Kromer et al., 1981; McNerney and Buseck, 1973) (Fig. 12-32).
Mercury
431
TABLE 12-III Mercury content of soil air before and after heavy rainfall (ng Hg / 250 ml air) Station no.
Soil dry, before rain
After heavy rain
18
0.15
0
20 22 24 26 27 29 31 33 35 36 38 40 41 43
0.25 0.05 0.10 0.20 0.30 0.20 0.15 0.20 0.15 0.15 0.30 0.10 0.10 0.30
0.25 0.05 0.05 0.05 0.05 0 0.05 0 0.10 0 0 0 0 0.05
Soil In the CSIRO study, ten case histories were selected for comparisons of Hg concentration in different soil horizons. The soils varied from strongly-differentiated acid peats and peaty podsols, through duplex soils, red earths, lateritic soils and lithosols. In those soils with a strongly-developed Ao horizon (see Figs. 12-15 and 12-24), background Hg levels increase towards the surface due to the scavenging effect of the organic material (Table 12-IV). The results from Currawang, Elura and Dugald River are compared in Fig. 12-33. The duplex soils at Currawang were sampled towards the base of the A horizon (here the A0 horizon is poorly developed), within the B horizon and where possible within the C horizon. Except for an enigmatic single site anomaly towards the base of the hill on Line 1, no significant difference in Hg contents is observed between the various soil horizons. At Elura, red earths containing ferruginous gravel show very little evidence of textural or gross chemical differentiation down to the maximum depth sampled (300 mm). The undifferentiated lithosols at Dugald River were sampled at depths of 50 and 150 mm and results indicate that the strong anomaly over the outcrop of the Western Lode is independent of sampling depth.In the limited number of case histories presented here, soils have been sampled down to a maximum depth of 500 mm. Where there is no obvious textural or colour differentiation, the sample depth within this interval is not critical. Sampling below the top 100 mm is, however, advisable. Soils with a well-developed A0 horizon should be sampled towards the top of
432
G.R. Carr and J.R. Wilmshurst ELURA LINE 58 748N
i
""
751
158ram
o
CURRAWFING LINE
1
-
18~mm
~
~
.
.
ENSTINGS (m)
.
1588
~,,
5~
.
.
.
.
G48
Z79B
z099
TOPO SLOPE
2989
DUGALD RIVER LINE
38B9
3198
ERSTINGS (m) - ~-
3298
OL
t ~65e
.
t ~s9
.
19e5~ ~ FIq Y I IqF.":, ' m "
14.3GN
.
t ~95o
Fig. 12-33. Comparison of Hg content of different soil horizons along traverses across the Currawang, Elura, and Dugald River deposits (from Carr et al., 1986).
the underlying horizon. It may well be desirable to sample at even greater depth within deeply-weathered profiles. Although the volatility of Hg requires that care be taken with soils to prevent contamination of samples or, rarely, Hg loss from samples, the sampling techniques
TABLE 12-1V Background thresholds and anomaly peaks of mercury content of soils over massive sulphide deposits in western Tasmania Deposit
Beatrice Howard's anomaly Lake Salina Que River Pinnacles
Background threshold in soil (ppb) A horizon B/C horizon 70 (?) 100 (?) 200 290 120
40 (?) 30 I00 210 80
Anomaly peak in soil (ppb) A horizon B/C horizon 280 260 400 900 300
400 240 450 2000 520
Mercury
433
differ little from those commonly used for soils (Ryall, 1979a). After sampling, soils should be sealed in plastic bags for storage and transport; if left unprotected, they readily adsorb above-background levels of Hg from contaminated laboratory atmospheres. Drying of soil samples can be achieved by opening the sample bags to the atmosphere at ambient temperature (20-30~ in an environment determined to be free of significant Hg contamination. The dry samples should be lightly crashed, if necessary, and sieved through nylon mesh after removing obvious organic fragments. A mesh size of 180 or 230 is recommended to maximise the clay component of the sample.
Comparison of soil and soil gas In four case histories from the CSIRO study, soil gas was collected in addition to soil. In the example of the Currawang deposit, a soil-gas Hg anomaly of about twice background occurs in association with a soil anomaly of about 100 times background (Fig. 12-34). The detection limit of Hg in soil gas was 0.07 ng/L and the background levels were in the range 0.1-0.2 ng/L. In two other examples, the Woodcutters Zn-Pb deposit (Roberts, 1973) and the Ranger IV U deposit, both in the Northem Territory, Hg soil anomalies, but not soil-gas anomalies, were detected. In the fourth example, the Jabiluka U-Au deposit, a soil-gas anomaly was present but not a soil anomaly (Fig. 12-
22). McNerney and Buseck (1973) analysed Hg in soil and soil gas over the Vekol Ag-Pb mine in Arizona. The deposit was defined by both a soil-gas anomaly of 6 ng/L over a background of about 1 ng/L and by a soil anomaly of 300 ppb over a background of about 30 ppb. Due to the problems inherent in analysing for the low concentrations of Hg in soil gas, especially when compared with the relative ease of soil analysis, it is important to define those situations where soil gas is likely to give more geochemical information than soil. There are several factors that influence the concentration of Hg in soil gas relative to soil: (i) the Hg source; (ii) the sorptive capacity of the soil; and (iii) the moisture content of the soil. The source of Hg may be either a flux resulting from the active oxidation of underlying sulphides or gas-solid equilibration of residual Hg adsorbed onto solid soil phases due to inorganic or biological processes. The sorptive capacity of the soil determines the partition coefficient between soil and soil gas in dry locations. This capacity is essentially due to the sesquioxides, the clays and organic matter. Significant amounts of any of these components will result in the moderate to strong partitioning of Hg into the solid phase. A high moisture content in a soil results in the partitioning of Hg from the gas phase into the liquid phase. Based on these soil parameters and the experience of case-history studies, Carr et al. (1986) considered soil gas to be an appropriate sampling medium only in soils of low sorptive capacity (such as transported quartz sands) in dry locations. In addition, the soil
434
I-,,.
\ o~ C" v
\ O] C
v
G.R. Carr and J.R. Wilmshurst
8.4 3hr8 0
20
0.4
Imln
0.2
T
0 -.~ 2000[ SOIL
1000[ "1OL 2930
2948
2950 ERSTINGS (m)
2960
Fig. 12-34. Comparison of Hg content of soil and soil gas (sampled at different times) at Currawang, New South Wales (from Carr et al., 1986).
profile must be deep enough and sufficiently free of large rocks to allow penetration of the soil probe.
Rocks
Rock samples, especially those containing sulphides, should be crushed in a percussion mortar and then ground to about 20 mm in an agate mortar. High-energy crushing and grinding techniques are to be avoided and particular care should be taken to minimise contamination of sulphides by atmospheric Hg and cross-contamination of unmineralised samples by sulphides. The factors to be considered when determining the relative merits of sampling rocks or soils for Hg analysis in a particular area differ little from those factors considered in any geochemical survey. In general, soils are the preferred medium because they tend to "smooth" noisy bedrock data by effectively sampling the weathering products from a large area of rock suboutcrop. Also, vapour-generated anomalies are less likely to be detected in rock samples. Soil samples are more readily collected in a routine manner and subsequent preparation procedures are simpler.
Mercury
435
RECOMMENDED ANALYTICAL PROCEDURES Numerous analytical methods are available for Hg analysis (Jonasson, 1970), but in geochemical applications only a limited number are practical. The commonly-used techniques incorporate two important stages: release of Hg from the sample; and measurement of the amount of Hg released. Of the two stages of analysis, the measurement of Hg presents few problems. It can be carried out conveniently using a type of atomic absorption spectrophotometer (AAS) or a resistivity-measurement instrument. The Scintrex HGG3 Hg analyser is a non-dispersive flameless AAS unit designed for field use. It employs the Zeeman principle for interference compensation (Robbins, 1973). Another instrument, SIROMAN, developed by J.R. Wilmshurst for field use in geochemical exploration, is a flameless, non-dispersive AAS with solid-state detectors; background interference (spectral absorbance) is detected and compensated by use of a continuum source. The Jerome Instruments Corporation Hg analyser uses the change in resistivity of a gold film to measure the Hg content of a gas stream. The unit itself is readily adapted for field use. It has been claimed that the instrument does not suffer from interferences to the same extent as do the AAS systems. The Chinese report a model JM-3 gold-film Hg detector with considerably enhanced sensitivity. The release of Hg from the sample can be accomplished by either reducing Hg in solution with stannous chloride (Hatch and Ott, 1968) or heating the sample. The thermal release method is the most reliable and easiest, provided sufficient precautions are taken. The heating of geological samples (soil, gossan, fresh or weathered rock) commonly involves the release of not only Hg but also other volatile components, some of which cause severe interference in AAS and gold-film analysis. The principal interferants are SO2, H20 and incompletely oxidised organic matter, although other volatile organic components can cause major problems. Such interferences are minimised by the admixture of an appropriate buffer with the sample: freshly calcined CaO for sulphides; and ignited K2CO3 for soils and other samples. Oxidation of organic matter is promoted by the use of oxygen as the carrier gas. Residual concentrations of these interferants are further reduced by the preferential sorption of Hg in the gas stream onto a gold collector such as SIROSORB. The SIROSORB is then heated in another gas stream and the desorbed Hg passed into the analyser. However, if the interferants have not been suppressed by the addition of buffers, significant contamination of the gold collector can still occur due to its lower temperature and large surface area. These interferants can then be passed on to the analyser. Significant loss of Hg can also occur during thermal release when working with certain classes of samples, principally those with high levels of halides and/or oxidate minerals. The initial sample generally gives an acceptable recovery, but that for subsequent samples tends to be low. The problem appears to be due to the volatilisation of components of the sample that condense on the cooler part of the release tube and
436
G.R. Carr and J.R. Wilmshurst FURNACE - Temperature Zone I
_
02/AIR
F-"
_!.
FURNACE 21 -I-Ternp'erature Zone \
---,
SodQ-Lime trnp Somple boot
SilicQ tube 25 and 12 mm diQ.
Sirosorb
Fig. 12-35. Schematic of two-stage furnace used for Hg determination.
subsequently "react" with Hg, removing it from the gas stream. In these cases it is essential to use the K 2 C O 3 buffer and periodically to heat the cooler parts of the release tube. Since the effect intensifies with higher release temperatures, it is best to work at the lowest temperature that will give a quantitative recovery. This is determined empirically for any given set of samples, and reference materials are analysed frequently to monitor performance. A commonly-used reference material is a volume of Hg-saturated air, the Hg content of which is determined from temperature-vapour pressure data. A glass syringe with a Teflon plunger is used for transfer. Mercury can also be lost after volatilisation, by adsorption and chemical reaction, as above. Adsorption is essentially a surface-area and temperature effect and so surface areas should be minimised. Absorption can be noticeable when plastic tubing is used for interconnections. Mercury is also lost if the release tube is heated excessively; with vitreous silica this occurs at temperatures of around 900"C. A recommended analytical procedure, and the one used at CSIRO, employs a twostage furnace (Fig. 12-35): the first stage determines the release temperature (between 300"C and 800~ the second promotes the oxidation of organic smoke or vapours. The volatilisation tube is of vitreous silica. Oxygen is used as the carrier gas at a flow rate of 60 mL/min. The sample, typically 100 mg, is contained in a porcelain boat and the size of sulphidic samples is adapted to the Hg content. The gas carrying the Hg from the volatilisation tube is passed through a soda-lime trap and then through a tube containing SIROSORB, the two traps being at room temperature. Typical furnace temperatures for soil samples of low organic content are 450-500~ for both temperature zones whilst for samples of higher organic content the oxidation stage is held at 800~ For sulphidic samples a temperature of 500~ is adequate for the second stage but the optimum release temperature depends on the nature of the sulphides and their ignition temperature. With air as the carrier, a temperature of some 800~ may be necessary for galena, which is the most "stable" of the sulphides. Mercury is released from the SIROSORB when the trap is heated in an aluminium slot furnace at 400~ with a constant nitrogen or air flow of between 50 and 60 mL/min, the gas passing directly into the analyser. System calibration is carried out at frequent intervals with a reference material introduced into the release tube at the relevant operating temperature.
Mercury
437
CONCLUSIONS Primary dispersion halos of Hg have been well documented in association with Hg deposits and geothermal fields and the use of Hg is a viable method of exploration for such targets. However, it has not been consistently nor convincingly demonstrated that Hg halos around precious-metal and base-metal deposits are any larger than those of the more commonly-determined indicator and target elements. The development of metamorphogenic Hg halos about Hg-containing sulphide deposits does not appear to be a common phenomenon. In residual overburden, Hg anomalies tend to be local to the surface expression of the mineralisation. Although broad halos have been described they are commonly lithologically controlled and as such are likely to be of similar dimensions to those of other metals. In regional scale exploration of terrains without transported cover, Hg is useful as a pathfinder element, especially in situations where information is needed rapidly and in-field analysis is possible. However, because anomalies are likely to be restricted in size, there is generally no advantage in using Hg in preference to target or other pathfinder elements. In prospect level exploration, the common specific relationship between the distribution of Hg soil anomalies and the suboutcrop pattern of units containing oxidising sulphides can be used to advantage. Vapour-generated Hg anomalies can occur above buried or blind sulphide deposits. However, they tend to be subtle and restricted in area and are thus easily missed, especially if the sample interval is large. The formation of a vapour-generated halo is critically dependent on aspects of climate and geomorphology. In wet terrains, Hg may be lost by hydromorphic dispersion with no vapour-generated expression at surface. Only in dry or periodically-dry terrains can a significant vapour halo form. The position of the water table relative to the depth of oxidation of sulphides dictates whether a flux of Hg occurs through to the soil layer. Whereas soils with a high sorptive capacity preserve a vapour-generated halo (as a "fossil" halo if the Hg flux ceases due to changing weathering conditions), soils with a low sorptive capacity, such as quartz sands, yield a measurable soil gas anomaly only during active oxidation of sulphides. Thus, in all but the situation of transported quartz-rich alluvium, soil is preferred relative to soil gas as a sample medium. In soil sampling, sample depth is not critical with non-differentiated soils, provided the sample is taken from beneath the A horizon. However, in differentiated and highly organic soils which accumulate Hg in the A horizon and may thus show a high background, the best contrast is obtained by sampling the B or C horizon.
This Page Intentionally Left Blank
Geochemical Remote Sensing of the Subsurface Edited by M.Hale Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 02000 Elsevier Science B.V. All rights reserved
439
Chapter 13
DISCRIMINATION OF MERCURY ANOMALIES Z. HU
INTRODUCTION Mercury, by virtue of its coexistence with many types of base metal and precious metal deposits and its high vapour pressure, often forms secondary halos above ore deposits, including in soils. When soil is heated, Hg is released at different temperatures according to the form in which it occurs in the soil (Fig. 13-1). The choice of a particular temperature for thermal release of Hg in a certain form can produce geochemical data that are useful for locating deep blind ore deposits. Comparison of data obtained by thermal release at different temperatures allows the discrimination of significant anomalies related to such mineralisation from false anomalies related to various forms of pollution.
ppb
$0
.-,;,
3000310 ~
HgS----.."
,,o.
40"
HgS04
\ ....
0"~=:" x H g a C l z - - - - J , - - , x 2 o ..'-..Jr 11r ~ ~ ~
:
,. :
.,
:
:
Y.
0
-~
....
!
100
---
-
!
200
.....
s40 o SSo~
~ i
!
/ ~ -,...,~ I \ ~ !
20
O ~
i
HgO
i
,
300
\ ,z~'/
'7
~
400
Fig. 13-1.Thermo-emission spectra of different Hg compounds.
~
500
/
,I
!
I
I,
i
I
! 600
440
Z. Hu
Fig. 13-2. Scheme for measuring Hg content of soil using thermo-emission method: 1 -temperature control, 2 = thermo-electric sensor, 3 - sample, 4 = furnace, 5 = N a z C O 3 filter, 6 -Au wire amalgamation trap, 7 = vacuum pump, 8 = Hg detector.
METHOD Soil samples are collected at depths of at least 40-50 cm and below any stone debris, tree roots and the cultivated layer. Samples are air dried, and then sieved and bagged immediately. Mechanical grinding must be avoided or Hg may be thermally released during the resulting heating. Prolonged storage of samples may result in loss of Hg by vaporisation or in sample contamination. Thermally-released Hg is determined by means of a simple flameless atomic absorption detection system (Fig. 13-2). This has a sensitivity of 0.1 ng Hg and a limit of detection of 0.01 ng Hg. In the temperature range 120-150~ adsorbed Hg (subsequently termed Hg x), comprising Hg, HgC12, Hg2C12 and any Hg", is released and measured. When the temperature is raised to 800"C, all forms of Hg (Hg,) are released. The determination of Hg as these two groups of compounds aids the understanding of Hg dispersion and hence exploration.
CASE HISTORIES Experimental studies have been carried out in field locations in China. These studies include soil traverses over iron, copper, molybdenum and lead-zinc deposits, as well as a 20 km regional traverse and a regional grid. The results show significant anomalies above mineralisation, with low background values and good contrast. Both Hgx and Hg, contribute to indicating the location of buried mineral deposits.
Traverses over known mineral deposits
The Meishan iron ore deposit, a large porphyritic iron deposit that is located in the northern flank of the faulted Ning-Wu basin, occurs in the crushed boundary between
441
Discrimination of mercury anomalies ppb
Hgx Hg~
I
700 600
.~~
500 4OO
300
I
II~1~
--Hg~ I
I
Hgz
Fig. 13-3. Profile of Hg content of soil over the Meishan iron deposit.
porphyritic diabase-diorite and andesite. The ore minerals are mainly magnetite and martite, with accessory siderite and specularite. Most of the ore minerals have a high sulphur content. The ore body is buried by 100-200 m of non-mineralised rock. Along two traverses about 565 m apart, soils were collected and their thermally-released Hg content determined. Anomalies of both occur over the ore body, many of them coincident or almost so. Beyond the limits of the mineralisation, background levels of Hgx and Hg, are mainly low. The results along one traverse are shown in Fig. 13-3. The Funiushan copper deposit is a skam deposit on the inverted northem limb of the western Tang-Lun anticline. It occurs within the contact zone of a quartz diorite and comprises chalcopyrite, subordinate pyrite and sphalerite, as well as minor amounts of magnetite and martite. The ore body lies beneath sedimentary formations and 20-30 m of eluvium and alluvium. Thermally-released Hg was determined along three soil traverses. As the traverse illustrated in Fig. 13-4 shows, the Hgx pattem picks out the mineralisation better than the Hgt pattern. The shape of the Hgx anomaly is characterized by twin peaks over the margins of the ore body. A large thrust fault seems to have a role in anomaly development. The Tongshan lindgrenite skarn deposit lies on the northern margin of the NingZheng Dome folded zone. The ore bodies are fault-controlled along the contact between
442
Z. Hu ppb
Hgx 5
~,3JI
-100
i
w
1
|
!
Fig. ! 3-4. Profile of Hg content of soil over the Funiushan copper deposit.
quartz-diorite and limestone. The ore minerals are chalcopyrite and molybdenite, with subordinate magnetite and pyrite. The ore bodies are, for the most part, deeply buried, but outcrop locally. Three soil traverses were sampled. Along one traverse which crosses steeply dipping ore concealed beneath eluvium, good contrast single peaks of Hgx and Hg, are well developed (Fig. 13-5). On another traverse, anomalies are present over a worked-out ore body, but their intensities are only about one third of those on the first traverse, and to the north of these anomalies are much higher peaks caused by waste water from the refinery in that locality (Fig. 13-6). Comprising mainly magnetite, sphalerite, galena and a little chalcopyrite, the Xiaomaoshan polymetallic skarn deposit occurs at the contact of porphyritic granite and limestone, at a depth of 110 m. The surface of the area is covered by exotic overburden, on which rice is cultivated. Soil samples taken at 20 m intervals along a traverse yielded twin peaks of Hgx above the hangingwall and footwall contacts of the mineralisation (Fig. 13-7).
443
Discrimination of mercury anomalies 400 ppb
Pl~b Hgx Hgt 10 250
/ ~,,~1 ~t~o ~ . _ ,,,,/~ju 0/ :'" ,'" , ,
I
l. ,, , ~ _ . , , , ,~ ~ . . , 7"
,
Fig. 13-5. Mercury content of soil over the Tongshan copper-molybdenum deposit, Profile 1.
Hosted in the contact between granodiorite, hornfels and dolomitic marble, the Zhangyan copper-molybdenum deposit is a skam-type deposit, containing chalcopyrite, magnetite and molybdenite, buried beneath 160-200 of sediments. Along three traverses 100 m apart, soil samples were collected at 20 m intervals and at depths of 40 and 60 cm. These samples were analysed for Cu, Pb, Zn, As, Sb and thermally-released Hg. The first five elements yielded no anomalies. As illustrated in the traverse shown in Fig. 138, samples from 60 cm yielded twin peaks for both Hgx and Hgt located approximately above the hangingwall and footwall contacts of the mineralisation; the same pattern is present but less clear in the samples from 40 cm. This difference is attributed to interference from the use of various inorganic and organic fertilisers in rice cultivation. The Shanbaidu polymetallic deposit comprises genetically related but spatially separate iron and lead-zinc skarn deposits in the contact zone of quartz-diorite and limestone. The ore minerals are mainly magnetite, hematite, galena and sphalerite. The ore bodies are found at depths of 5-40 m, and the surface of the area is covered by 5-15 m of transported exotic overburden. Thermally-released Hgx data from samples collected at depths of 30, 60 and 80 cm in the soil along a traverse crossing both ore deposits are shown in Fig. 13-9. In the samples from 60 and 80 cm, clear anomalies mark the positions of the mineralisation. In the samples from 30 cm, however, the background is
Z Hu
444
ppb
Hgx 8
.g, 200
Hgt
150
Hg=
I '
|| I/~,,
,
,
,
,
,
,
Fig. 13-6. Mercury content of soil over the Tongshan copper-molybdenum deposit, Profile 2.
very disturbed, especially in the northern part of the traverse where values of Hg~ exceed those over mineralisation. Field observations suggest that these high values are caused by pollution from Hg-containing pesticides sprayed over cotton fields.
Regional traverse A soil traverse of 20 km was made in the Ning-Zeng region in order to investigate the applicability of thermally-released Hg at the regional scale. The sample interval was 50 m, closing to 25 m across the Qixiashan lead-zinc deposit and the Mierenshan, Anjishan and Funiushan copper deposits. These sample intervals were strictly maintained so that any effects of pollution were included. Samples were analysed for Cu, Pb, Zn and thermally-released Hgx and Hgt. Above the Qixiashan and Funiushan deposits there are prominent peaks for both Hgx and Hgt (Fig. 13-10). There is a weak Hg~ response over the Mierenshan deposit. The lack of response over the Anjishan deposit is attributed to its low grade, small size and deep burial. The low, broad peak of Hgt extending to the north of the Funiushan reflects the gossan cap. The remaining unmatched Hgt peak in the north-centre of the traverse
Discrimination of mercury anomalies
A ,a Q, Q. ...-
445
5
4
.!.
3 2 1
Fig. 13-7. Profile of Hgx content of soil over the Xiaomaoshan polymetallic deposit.
coincides with a residential area. Except at this village, there is a correlation between one or both of the forms of thermally-released Hg and one or more of the base metals in soils. The traverse demonstrates that thermally-released Hg in soils successfully detects a number of mineralisations against a generally low and stable background, but anomalies are also produced by other sources, including villages.
Regional grid Using a grid of 200 x 50 m, an area of 180 km 2 over the Qixiashan lead-zinc deposit and its surroundings was surveyed for thermally-released Hg in soils. The results are satisfactory, as illustrated here by those from a 14 km 2 sub-area. The mineralisation occurs in folded and faulted Palaeozoic rocks, upon which Jurassic strata lie unconformably (Fig. 13-11). The ore, which occupies faults in carbonates and a faulted unconformity, takes the form of layers, lenses and veins. The ore minerals include sphalerite, galena, pyrite and a little chalcopyrite in a gangue of
Z. Hu
446 P ~b Hgx Hgt 5
I
too
.o
I
II
rl
I
,A,r
40 cm soil sampling depth 0
ppb
Hgx
9
9
....
9
9
"
Hgt
80 cm soil sampling depth
Fig. 13-8. Profile of Hg content of soil over the Zhangyan copper-molybdenum deposit.
quartz and calcite. The area is hilly and partly covered by soil, 0.4 - 40 m thick. The dominant components of the soil are clay, sandy clay, eluvium and alluvium. The pattern of thermally-released Hgx is shown in Fig. 13-12. The background is generally stable at <2 ppb. Regional anomalies are in the range 2-4 ppb and pick out the metallogenetic fracture area. Local anomalies are usually in the range 8-16 ppb, but can reach 25 ppb over shallow mineralisation, and tend to lie parallel to the strike of known ore zones. Anomaly I overlies the northeastern extension of the ore-controlling fault (F2) and coincides with a manganese cap. This anomaly is considered prospective, with further deep drilling expected to reveal ore at depth. Anomaly II overlies the main mineralised fault and its areal extent is consistent with that of the ore body, despite a depth of burial which reaches 300-350 m in the west. Anomaly III is developed on thick overburden, and deep drilling would be necessary to determine its significance. Anomaly IV is considered more prospective, since it coincides with Cu, Pb, Ag and electrical anomalies, and extends beyond the area shown in Fig. 13-12.
447
Discrimination of mercury anomalies I 'false' anomalyl
i.lgnlllcantlanomaly ' A, A^A' I
3
~2 30 cm soil sampling depth
0 ~3 iX
~2
&
z
1 .
.
.
.
.
.
ing depth
Fig. 13-9. Profile of Hgx content of soil over the Shanbaidu polymetallic deposit.
DISCUSSION Under most circumstances, the vertical distribution of thermally-released Hg in the soil profile is isotropic. In background areas there is a slight enrichment of Hg in the A horizon of the soil compared to the B horizon (Fig. 13-13a) and there is a similar pattern, but at higher concentrations, over mineralisation (Fig. 13-13c). Where pollution affects otherwise background levels of Hg, elevated concentrations of Hg occur in the A horizon but drop quickly to reach normal background levels at a depth of 50 cm (Fig. 13-13b).
448
Z. Hu
E 1 500~ o. 1000-1
s ool ___._._e~_~
,
i
o.E 1500-~
1oo01 !
v
E ,x 1500~ -"
rl
100o1 SO0o~
1000-
: ::
5000 ,', 1000] = 500J
-r
O:
,
,
' .....
N
O-r---|
-500~!
--
/ Qlxlashan
,
, ivlllaile
~'
+
I I
(m) Pb-Zn deposit
,
'
-
,
, S
I lp
~-
I
l OpO m
I
ImelrenBhlnl
mineral occurrence
I I
South
I
I
I
I
Anjishan Cu deposit
~ Funluahan
Cu deposit
I I
Fig. 13-10. Regional geochemical profiles showing base metal and Hg contents of soil over known mineral deposits. The groundwater level in the soil has no detectable effect on Hg patterns. At Shanbaidu, anomalies over mineralisation occur in low ground near to a river and are often covered by flood waters; samples from depths of 60 and 80 cm actually come from below the water table at this locality, but show good anomalies (Fig. 13-9). Similarly, rainwater run-off has no detectable effect; anomalies on a slope overlying mineralisation at Tongshan show no downslope displacement (Fig. 13-5). In general, data for both Hgx and Hgt reveal similar patterns: statistical analysis using 1146 soil samples from several areas show that Hgx and Hgt are closely correlated. Background levels are typically 0.5-1.0 ppb Hgx and 20-50 ppb Hgt, so the ratio Hgx : Hgt is usually within the range 0.01-0.05. However, when pollution is present the ratio moves outside this range. In this way the ratio provides a means of distinguishing anomalies caused by pollution from those over mineral deposits.
CONCLUSIONS Sampling and analysis for thermally-released Hg in soils can be carried out easily and economically. The resulting data are effective in exploration for concealed mineral deposits.
Discrimination of mercury anomalies
Fig. 13-1 i. Geologic map of the Qixiashan lead-zinc deposit.
Fig. 13-12 Distribution of Hgx content of soil over the Qixiashan lead-zinc deposit.
449
450
Z. Hu
.y
. .._ ppb
/
5A_
9 ,.~ ppb
. /
> ppb
50
'ID
o
e~
Hg
Hg
-Hg
r
Fig. 13-13. Vertical profile of Hgt content of soil in (a) background area, (b) polluted area, and (c) mineralised area; A = soil A horizon, B = soil B horizon.
Various studies have shown that background values are low and anomalies over iron, copper, lead, zinc and molybdenum deposits are of good contrast and often characterised by several peaks. Twin peaks occur above gently-dipping ore bodies (<70 ~ whereas a single peak appears above more steeply-dipping ore bodies. Anomaly size and shape are also influenced by the genesis of the ore body, intensity of weathering, type of ore minerals, depth of burial and soil properties. False anomalies unrelated to mineralisation arise due to the use of fertilisers, pesticides and even simply proximity to villages. These false anomalies can be recognised by the departure of their Hgx : Hgt ratios from the norm for the area, which is otherwise stable over both background and mineralised locations. Better still, false anomalies can be avoided by sampling below 50 cm, which has been found to be the lower limit of penetration of surface pollution in the areas studied.
Geochemical Remote Sensing of the Subsurface Edited by M.Hale
Handbook of Exploration Geochemistry, Vol. 7 (G.J.S. Govett, Editor) 9 Elsevier Science B.V. All rights reserved
451
Chapter 14
OXYGEN AND CARBON DIOXIDE IN SOIL AIR J.S. LOVELL
INTRODUCTION It is self-evident that the oxidation of sulphide minerals entails the consumption of oxygen. The initial source is molecular oxygen from the atmosphere but this must pass into solution in groundwater or soil solutions before any reaction with sulphides is possible. Interstitial air in soils, overburden or porous rocks forms an intermediate reservoir of oxygen between buried sulphides and the free atmosphere. The oxidation may be entirely chemical or may be enhanced by the microbial action of bacteria such as Thiobacillus thiooxidans. The oxidation of sulphides leads to the production of sulphuric acid, which will be neutralised by any available carbonates with the release of gaseous carbon dioxide into the subsurface surroundings and ultimately into the atmosphere. Thus the oxidation of a sulphide ore body will lead to consumption of molecular oxygen and probable production of gaseous carbon dioxide. A porous overburden will form a buffer in which restricted diffusion, dispersion and replenishment will accentuate and retain anomalous gaseous activity. It is rewarding and surprising to calculate the quantities of oxygen consumed by sulphides and the amount of carbon dioxide that may be generated. If the oxidation of pyrite proceeds to goethite thus, 4FeS2 + 1502 + 10H2O
=
2Fe203(H20)
+ 8H2804
then one tonne of pyrite will consume one tonne of oxygen and produce 1.6 t of sulphuric acid. If sufficient carbonate is available then the reaction, CaCO3
+ H2804
=
CaSO4 +
H20 + CO2
will produce 0.72 t of carbon dioxide. The oxygen consumed has a volume of 700 m 3 and the carbon dioxide generated will occupy 370 m 3. With an atmospheric concentration of 20.9% oxygen, this would totally deplete 3350 m 3 of air. If the gas is derived within a rock with a porosity of 20%, then the oxygen will be entirely removed from 16,750 m 3 of cover. As diffusion and mass flow will tend to replace the lost oxygen, an overall fall of 1% oxygen could be produced in 350,000 m 3 of overburden.
452
J.S. Lovell
The carbon dioxide produced would give a concentration of 0.53% in this same volume of rock or soil. The oxidation of sulphides may proceed with extreme rapidity and spontaneous sulphide fires in mines have not been an uncommon occurrence. Bateman (1950) notes that a sulphide vein in a blind and warm stope at the Leonard Mine (Butte, Montana) was oxidised to a depth of 1 m within two years. At the Ely Mine, Nevada, chalcocite ore in a bench in an open pit was oxidised so quickly that, at a depth of 10-15 m, about 15% copper within the ore was removed in solution. Thus, within the overburden above a weathering sulphide deposit, there is a potential for the development of gaseous CO2 and 02 concentrations that are anomalous with respect to the local soil-air regime. This was first considered in the former Soviet Union (Glebovskaya and Glebovskii, 1960), and there are numerous reports in the Russian literature describing the use of these gases in soil air as an exploration method (Kulikova, 1960; Khayretdinov et al., 1965; Kravtsov and Fridman, 1965; Glebovskaya, 1969; Elinson et al., 1970; Dadashev et al., 1971; Fridman and Petrov, 1976). Given that much of this literature describes apparent success in locating buried ore using CO2 and 02 in soil air, it is perhaps difficult to understand why this particular exploration technique was largely ignored in the West prior to the research of Lovell (1979) and Lovell et al. (1980).
OXYGEN AND CARBON DIOXIDE IN THE SUBSURFACE The extent to which oxidising sulphides affect the composition of the subsurface atmosphere will depend upon the rate of oxidation and the intensity of other activities that remove oxygen and generate carbon dioxide. The oxidation of pyrite was reviewed in the introduction to this chapter, and other sulphides more commonly of economic interest would be expected to behave in a similar manner. However, the stabilities of different metallic sulphides vary greatly in the secondary environment and consequently their oxidation rates differ. The rate of oxidation of pyrite has been studied in the greatest detail and may be summarised thus: 9 oxidation proceeds most rapidly below a pH of 3.5, at which point the activity of sulphide-oxidising bacteria becomes the dominant effect, whereas above this pH the oxidation is purely chemical; 9 factors reducing the solubility of iron (high pH, high concentrations of phosphate ions) retard oxidation; 9 a high static water table restricts the circulation of atmospheric oxygen and limits the sulphide oxidation to zones of freshly-oxygenated groundwater; and 9 low temperature reduces the rate of oxidation.
Oxygen and carbon dioxide in soil air
453
It is probable that the conditions that most favour a high rate of oxidation of pyrite are those which are most likely to produce a gaseous expression of sulphide oxidation in the subsurface, i.e. high ambient temperatures, fluctuating water table and emergence of sulphides above the water table from time-to-time. The presence in a sulphide mineral assemblage of minerals relatively unstable under oxidising conditions (pyrite, marcasite, chalcopyrite) is likely to lead to a better gaseous expression in oxygen and carbon dioxide than when only more stable sulphides (such as galena) are present. There are, of course, several processes that can frustrate the detection of such expressions. It is important to consider the rate at which gas enters and leaves the soil air. Baver (1972) quotes several authors and estimates that there would have to be a complete renewal of soil air every hour to a depth of 20 cm in a normal cropped soil in order to maintain its usual average composition and microbiological activity. If mineral deposits are to have adequate expression in the soil air, sulphide oxidation must clearly influence its composition at a rate commensurate with such rapid aeration. Furthermore, the oxidation of sulphides is among the least significant of activities that deplete the oxygen in the soil air and which contribute carbon dioxide. By far the most important is the mineralisation of organic carbon by microbial activity, followed by other biological processes in the soil. Any gaseous expressions of mmeralisation have to be seen against this background, which therefore deserves a brief review here. The rates of oxygen consumption and carbon dioxide production in normal agricultural soils depend upon the activity of plant roots and on microbiological activity, which are in tum dependent upon the soil moisture content and temperature, and the ease of decomposition of the organic matter in the soil. The respiratory quotient of a wellaerated soil (i.e., the ratio of the volume of carbon dioxide produced to the volume of oxygen consumed) is close to unity (Dixon and Bridge, 1968). It should only rise above unity where water or clay minerals restrict free circulation of gases in soils and produce anaerobic pockets. There is a very wide range of published figures for oxygen consumption and carbon dioxide production in soils, because these depend upon a great many factors. For example, Monteith et al. (1964) found that the carbon dioxide flux (i.e., CO2 lost to the free atmosphere) in a non-vegetated Rothamsted clay-loam soil was 1.5 g m -2 day -! in winter and 6.7 g m -2 day -~ in summer and, for the dependence of flux upon temperature, derived the equation, R = IL,Q x/,o where Ro is the flux at 0~ and R is the flux at T~ They found that Q had a value of about 3 and the results of their investigations are given in Fig. 14-1. These figures may be rather low because Currie (1970), using a more direct method, obtained much higher summer figures than Monteith et al. (1964). The figures in Table 14-1 show a much greater rate of gaseous exchange by soil under vegetation.
454
J.S. Lovell June Aua,~
a
"O
6
uly
~N
E a X
M a r c9h
4
F:b
_= O o
A p r 9i l .J M a y J 9 S e p t
2t ~Nov
Dec I 5
I 10
I 15
M e a n s o i l t e m p . ( o C ) at 10 cm
Fig. 14-1, Relation between the daily soil respiration and mean soil temperature of a bare Rothamsted soil" the two curves are the plot of RoQwl~ for Q = 3 and R0 = 1.2 (upper curve) and R0 = 0.9 (lower curve) (from Russell, 1973).
Ross and Roberts (1970), among others, demonstrated that 0 2 uptake and C O 2 production were significantly positively correlated with the numbers of viable bacteria in soils and negatively correlated, although not usually significantly, with mean annual temperature. The gaseous activity of the soils did not appear to be influenced by differences in the nature of the vegetative cover, the clay or organic carbon content of the soil, or the mean annual rainfall. In a classic early study, Wollny (1881) reported that the CO2 content of the soil at 30"C increases about ten times as the soil moisture content changes from 6.8% to 26.8%. The CO2 content of soil air also increases with depth and the 02 content falls (Table 14II). The gradient is influenced by the nature of the soil: granular soils contain less than
TABLE 14-1 Oxygen consumption and carbon dioxide production from a bare soil and a soil under kale at Rothamsted, UK (g m -2 day -I) Gas exchange Oxygen consumption Carbon dioxide production
Summer (17~ Cropped Bare 24 12 35 16
Winter (3 ~ Cropped Bare 2.0
0.7
3.0
1.2
Oxygen and carbon dioxide in soil air
45 5
2O 16 10
8 0
A t 3 0 cm
20
"'.. . . . . . . . ... . . . . . . "....
............................ e
15
%
Vol.
Gas
02
....... S a n d y loam
811ty c l a y
10
20
~o
At 90 cm
,.
Jan
.............................
Mar
May
July At
8apt
2
Nov
1 6 0 cm
Fig. 14-2. Oxygen and carbon dioxide content of the soil air at three depths in a sandy-loam and a silty-clay apple orchard.
one half as much CO2 as powdery soils; loams contain more CO2 than sands; and clays contain more CO2 than loams. There are also considerable seasonal fluctuations in the 02 and CO, contents of soil air (Fig. 14-2). The diversity of the phenomena that can affect the soil-air composition and the mutual inter-dependence of these phenomena suggest that the production of a suitable equation to predict the O, and CO2 concentrations in a given soil may be difficult if not
TABLE 14-11 Oxygen and carbon dioxide content of soil under cacao, Rivers Estate, Trinidad (Russell, 1973) Depth (cm) 10 25 45 90 120 Rate*
02 content Wet (I) Dry (2) 13.7 20.6 12.7 19.8 12.2 18.8 7.6 17.3 7.8 16.4
content . . . . . . . . . . Early dry (3) Late dry (4) 1.0 0.5 2.1 !.2 4.3 2.1 6.7 3.7 8.5 5.1 14.8 35.0
CO 2
Wet (1) 6.5 9.7 10.0 9.6 13.4
(I) Oct-Jan; (2) Feb-May; (3) Feb; (4) May; (5) Apr-May. * Observed CO2 diffusion rate from the soil (g m 2 day ~)
CO 2gradient Wet (I) Dry (5) 0.65 0.05 0.13 0.06 0.04 0.07 0.01 0.06 -0.01 0.06
456
.LS. Lovell
impossible. Furthermore, the collection of sufficient data to enable this to be even attempted would be impossible within the context of mineral exploration. Moisture content affects the activity of the soil population in two ways: through the thickness of the water films in the soil and consequently its aeration; and through the reduction in free energy of the water as the films become thinner and the soil drier. Bacteria can move only in water films and, although many are smaller than 1 ~tm in size, they only appear to be readily mobile in films appreciably thicker than this. The rate of movement becomes slower as the soil becomes drier, probably because the water-film pathway between two points in the soil becomes more tortuous (Hamdi, 1971). A point that does not appear to have received attention (Russell, 1973) is that microbial activity always involves the excretion of by-products that are toxic to the organism and, as the soil becomes drier and the water films thinner, these will diffuse into the surroundings increasingly slowly. Microbial activity tends to increase with increasing moisture content, with a corresponding increase in CO2 production. The composition of the soil air upon the microbiological population is not important until the 02 content falls below 1%. Until this level, the soil behaves as a fully aerobic environment (Greenwood, 1967). Vegetated soils have higher CO2 contents, due to root respiration, than bare soils, and the presence of decomposable organic matter increases 02 uptake and hence CO2 production (Table 14-III). The very process of preparing soils for cultivation increases their aeration and rates of both gaseous exchange and organic matter mineralisation.
TABLE 14-I11 Composition of the air in soils, percent by volume (Russell, 1973) Soil type Arable, no dung for 12 months Pasture land Arable, uncropped, no manure: - sandy soil - loam soil - moor soil Sandy soil dunged and cropped: potatoes, 15 cm - serradella, 15 cm Arable land: - fallow unmanured - dunged Grassland -
-
Usual composition 02 CO2 19-20 0.9 18-20 0.5-1.5
Extreme limits 02 CO2 10-20
0.5-11.5
20.6 20.6 20.0
0.16 0.23 0.65
20.4-20.8 20.0-20.9 19.2-20.5
0.05-0.30 0.07-0.55 0.28-1.40
20.3 20.7
0.61 0.18
19.8-21.0 20.4-20.9
0.09-0.94 0.12-0.38
20.7 20.4 20.3
0.1 0.2 0.4
20.4-21.1 18.0-22.3 15.7-21.2 16.7-20.5
0.02-0.38 0.01-1.4 0.03-3.2 0.3-3.3
Oxygen and carbon dioxide in soil air
457
SAMPLING AND ANALYTICAL METHODS A great deal of the research that has been carried out into the concentrations of 0 2 and CO2 in soil air has been for the purposes of agriculture. There is a large body of literature describing different sampling methods (e.g., Yamaguchi et al., 1962; Tackett, 1968; Dowdell et al., 1972; Burford and Stefanson, 1973; Bunting and Campbell, 1975). The Russian literature is unfortunately deficient in adequate descriptions of the analytical systems first used in mineral exploration. However, amongst the methods that have been mentioned are interferometry (Khayretdinov et al., 1965), thermal conductivity (Glebovskaya, 1969; Kulikova, 1960), gas chromatography (Dadashev et al., 1971) and mass spectrometry (Glebovskaya, 1969). The chief characteristic that distinguishes O2 and CO2 from other gases that have been used in mineral exploration is that the levels and changes in concentrations encountered in soil air are in the percentage range. The analysis is thus much simpler and can, where appropriate, rely upon the use of portable instruments.
Oxygen and carbon dioxide analysers In a number of surveys, Lovell (1979), Lovell and Hale (1983) and Lovell et al. (1979, 1983) used a Taylor Servomex DA272 oxygen analyser and a Lab-Line 2245 carbon dioxide analyser. These instruments were connected in series and samples of soil air from a probe driven about 1 m into the ground were introduced by means of a hand pump. The Taylor Servomex DA272 oxygen analyser exploits a property that distinguishes 02 from most common gases" oxygen is paramagnetic and hence tends to move to the strongest part of a non-uniform magnetic field. This movement can be measured in a Pauling cell, in which two nitrogen-filled diamagnetic spheres of glass are mounted at the end of a bar to form a dumb-bell. This dumb-bell is mounted horizontally on a vertical torsion suspension. The whole measuring cell operates within a strong, non-uniform magnetic field. The spheres are repelled from the strongest part of the field and so rotate until the force produced by the twist of the suspension is equal to the force acting on the spheres (Fig. 14-3). The strength of the magnetic field varies with the magnetic properties of the gas with which the cell is filled and this will thus govern equilibrium positions attained by the dumb-bell. The analyser uses a platinum torsion suspension, which imparts physical strength and is able to conduct a current to an electromagnetic coil used to maintain the dumb-cell at a zero position. The current required to maintain this zero position is a function of the 02 content of the gas present in the cell. The instrument can be calibrated easily in the field by the use of pure nitrogen for the zero and dry air for the span (up to 21% O2). Table 14-IV gives the magnetic susceptibilities of some common gases compared with 02. It can be seen that none of those with a susceptibility likely to interfere with the measurement would be expected to
458
J.S. Lovell
\
Force
on S p h e r e Pole Pie
\
(a)
~..
/:.
,
..~-_"_--7 T F,o, d
..-7---- ~'"...
Restoring F o r c e of Suspension
(b)
Fig. 14-3. Determination of oxygen concentration using its paramagnetic properties: (a) principle of the Pauling cell; (b) analytical cell of Taylor Servomex OA272 oxygen analyser.
occur in significant quantities in soil air. The instrument is fitted with a scale expansion system that permits a full-scale response in the range 16-21% 02. With this system it is possible to detect changes in the 02 content of the soil air of 0.05%. The carbon dioxide analyser exploits the fact that CO2 is about 40% less efficient at conducting heat than air. This difference in thermal conductivity is measured by means of a pair of thermistors in cells which are mounted in a massive aluminium block and which form the two arms of a sensitive bridge circuit. These thermistors respond to temperature changes with comparatively large changes in resistance. The thermistors are held at an elevated temperature relative to the aluminium by a fixed current. Heat is lost to the block via the gas surrounding the thermistors. The equilibrium temperature of each thermistor, and hence its resistance, is therefore dependent upon the thermal conductivity of the surrounding gas. Air fills one cell, which forms the reference arm of the bridge, and the sample fills the other. The sample and reference air are dried prior to analysis by means of a silica gel absorption tube. For use in the field, it was found necessary to fit the instrument in an insulated polished metal box, to minimise drift due to the effects of wind and direct sunlight. The instrument is calibrated in two ranges, 0-10% and 0-20% CO,. With multiple measurements it is possible to read the meter to within 0.05% CO2 with a detection limit of 0.05% CO2. However, extreme care is needed to obtain reliable measurements at this level. A disadvantage of both instruments is that, in the field, repeated and tedious zeroing is necessary before and after each measurement in order to ensure the requisite precision.
Oxygen and carbon dioxide in soil air
459
TABLE 14-IV Magnetic susceptibilities of common gases at 100% concentration in sample, expressed in terms of oxygen equivalent Gas Acetylene Ammonia Argon n-butane iso-butane l-butane Carbon dioxide Carbon monoxide Ethane Ethylene Helium Hydrogen
Percent 02 equivalent -0.24 -0.26 -0.22 - 1.3 - 1.3 -0.85 -0.27 +0.01 +0.46 -0.26 +0.30 +0.24
Gas Hydrogen sulphide Methane Neon Nitric oxide Nitrogen Nitrogen dioxide Oxygen n-pentane iso-pentane Propane Propylene Water
Percent 0 2 equivalent -0.39 -0.2 +0.13 +43 0 +28 + 100 - 1.45 - 1.49 -0.86 -0.54 -0.02
Orstat gas analyser Ball et al. (1983a, 1990) pumped soil air from a hollow probe into a modified Orstat gas analysis apparatus. A known volume of soil air is first pumped into the gas burette of the apparatus, and subsequently transferred to integral absorption vessels (Fig. 14-4). The absorbent for CO2 is 40% aqueous KOH. The volume of gas absorbed is recorded from the gas burette. The remaining gas is exposed to a mixture of saturated aqueous ammonium chloride and ammonia in contact with copper coils for absorption of 02. Again the gas volume reduction is recorded from the gas burette. Practical limits of detection under field conditions are 0.01-0.1% for each gas.
Draeger tubes For the determination of CO2 only, Romer and Finlay (1984) pumped 100 cm 3 of soil air from a probe through a calibrated glass tube containing reactants which respond to CO2 with colour change (Draeger type CH31401). Each tube is used once and then discarded. Its pre-sealed ends are snapped off before it is attached to the pump, and the CO2 that passes through the tube changes the colour of the reactant from white to blue over a part of the length of the tube. The length of the colour change is proportional to CO2 concentrations over the range 0.5-10% by volume.
460
J.S. Lovell
I1,, ,, II ,,',
,
/\
It
II II
II
II
Absn.
II
vo.,e!
/\ i
t
~
I ve.,.l
OB8 burette
,,/ Absn.
t
m
i , I
Levelling i bottle
J
Flexible tubing
,
,
[
!
I
Fig. 14-4. Schematic of the Orstat gas analyser (from Ball et al., 1990).
Gas chromatography and mass spectrometry The determination of 0 2 and C O 2 c a n be performed with considerable sensitivity by gas chromatography (Tackett, 1968; Bailey and Beauchamp, 1973; Bunting and Campbell, 1975; Blackmer and Bremner, 1977). Lovell and Reid (1989), using a hollow sampling probe driven into the ground, drew 30 c m 3 soil-air samples into a syringe and injected these into air-evacuated sealed cylinders for later gas chromatographic determination of 02 and CO 2 in the laboratory. At concentrations above about 0.1% v/v, 02 and CO 2 can be measured by mass spectrometry (Anderson et al., 1972; Nerken, 1972; Ball et al., 1973; McFadden, 1973; Newton et al., 1975; Robertson and Bracewell, 1979). McCarthy and Bigelow (1990) developed a truck-mounted mass spectrometer which McCarthy and McGuire (1998) used for on-site determination of CO 2 and other soil gases in the Carlin trend of Nevada. Soil-air samples were extracted from a hollow sampling probe by means of a syringe, from which they were injected directly into the mobile mass spectrometer.
461
Oxygen and carbon dioxide in soil air
b-surface 0 (3
air
1 14
13
12
1 1 10
9
8
7
6
5
4
3
2
1
Sample points 8-8E
N-NW
~!~:..:.:l
3 :;F__~,~~,~.~i'~,~ ~~,~.V'~~i ! . - ~ . 6 ~ r 9t~)~:--~
,,
Ey~i-
"
-~i,!
t
./.
Lower Cretaceous (Balel suite)
.
. .........
sandstone deposits
Lower Cretaceous sandstone-conglomerate deposits (Tergen suite) Varlacsn
.-9 ""
(Unda) granltolda
Ore zones
Fig. 14-5. Relation of carbon dioxide in soil air and geology at Balei gold veins, Baykal region, Russia (reproduced with permission from Kulikova, 1960).
CASE HISTORIES
Russia
Kulikova (1960) describes the results of a gas survey over the Balei deposits in the Baykal region of Russia. The mineralisation is gold in quartz veins and the source of the gas is believed to be emissions of CO2 from depth along mineralised fractures within Lower Cretaceous arenaceous sediments. The area is covered by alluvial Quatemary deposits that are up to 10 m thick on the fluvial plains and may reach 30 m on the terraces. The samples were collected in bottles with water seals of saturated NaCI. Despite the fact that the area of investigation is underlain by continuous permafrost at depths of 3-40 m, there is a good expression of the mineralised fractures in the soil-gas data from a depth of 1.5-2 m, with values over the ore in the range 1.0-2.0% CO2 against a background of 0.2-0.8% CO2 (Fig. 14-5). Soil samples were collected and analysed for sorbed subsurface gas, which also showed evidence of an enhanced CO2 flux.
J.S. Lovell
462
5.0 r 0
4.0 3.0 2.0 1.0 0
~ ~ ,,
,,
,,,,,
, *,,,
,
, ,
,
,
, 1966 ~""~,
,
~ ...... ,
, ,
, ,
,
June, ,
Teat
1966 points
eluvl81 and colluvl81 sediments aleurolltlc
clays
z o n e of f o l d i n g a n d I n t e n s i v e dlcklte mineralization
fracture
of r o c k s
with
z o n e of f o l d i n g a n d i n t e n s i v e calcite mineralization
fracture
of r o c k s
with
orebodles faults
Fig. 14-6. Relation of carbon dioxide in soil air and geology at Sakhalinsk mercury deposit, northwest Caucasus, Russia (reproduced with permission from Ovchinnikov et al., 1972).
Ovchinnikov et al. (1972) give only a few details of a survey over the Sakhalinsk mercury deposit, northwest Caucasus, but its inclusion here is of interest as it illustrates the influence of a nearby earthquake. The initial survey showed a weU-arked anomaly over the mineralisation with a maximum value of 4.8% CO2 against a background of 0.5-1.5% CO2 (Fig. 14-6). The second survey, some two months later, showed values elevated by approximately a factor of three. The influence of seasonal factors cannot be ruled out, but there was a very marked change in the ~3C ratio and gaseous anomalies associated with suboutcropping faults became more marked. The source of additional CO2 is thought to be abyssal gas preferentially moving from depth along the mineralised fractures.
Azerbaijan Dadashev et al. (1971) describe a soil-gas survey over the Filizchai pyrite deposit in Azerbaijan. They report a definite relation between the position of the ore body and the composition of the soil air. As the survey line enters the ore field in the southwest, there
Oxygen and carbon dioxide in soil air
463
Fig. 14-7. Oxygen and carbon dioxide in soil air over the Filizchai pyrite deposit, Azerbaijan (reproduced with permission from Dadashev et al., 1971).
is a steady increase in the C O 2 content of soil air and a decrease in the 0 2 content. There is a marked anomaly in the composition of the soil gas northeastwards, with the O2 content falling and the CO2 content rising until the ore body plunges to greater depth (Fig. 14-7). The changes in gas concentrations are ascribed to the effects of the oxidation of sulphide minerals; the patterns provoke interest in whether the anomaly maxima coincide with the position of maximum sulphide oxidation.
Kyrgyzstan
A survey carried out over a polymetallic ore deposit in Kyrgyzstan is described by Glebovskaya and Glebovskii (1960). The traverse shows an increase over the background CO2 concentration of 1-1.5% to a diffuse anomaly of 2-3.5% over the suboutcrop of the ore zone. There is an intense 02 anomaly, with the 02 concentration falling sharply from a consistent background of 20-20.5% 02 immediately over the mineralisation (Fig. 14-8).
Namibia
At Witvlei, 160 km west of Windhoek in central Namibia, copper mineralisation occurs within a sedimentary sequence ranging from conglomerates through grits, sandstones and arkoses to siltstones and claystones. The area is underlain by a sandstone-siltstone sequence dipping at about 40 ~ to the southeast. Trenching revealed that surficial geochemical copper anomalies are due to chalcopyrite mineralisation within certain horizons. Limited drilling showed that the sulphides have been oxidised to malachite and chrysocolla to a depth of about 30 m, and that the mineralised units are
J.S. Lovell
464 % CO 2
% 02
8.0 7.0
21.0
6.0
20.0
5.0
19.0
4.0
18.0
3.0
17.0
2.0 i
16.0
1.0
15.0
0
02
CO 2 v 15
!
u
u
v
10
0
-5
-10
, -15
| -20
i -26
9
.-,'of,.'.; ""
9
. . . . : . ~: .... : .p... 9;. :.;: .-.~:
9, . . , , .,., : . . . . : ' ~ . . :
9 4
...'-~ ., .. .- , ", .
gneisaoid granite leucocratlc granite ~
tectonic dislocation
quartz porphyry
~
ore body
daclte
~
un 9
deposits
Fig. 14-8. Oxygen and carbon dioxide in soil air (from !.5 m) over polymetallic ore deposits in Kyrgyzstan (reproduced with permission from Glebovskaya and Glebovskii, 1960).
interbedded with calcareous arkose. There is a cover of up to 2 m of bright-red Kalahari sand with a basal rubble horizon and occasional calcrete; outcrop is rare. Within this area two soil-gas traverses were conducted; one of these is shown in Fig. 14-9. Using a convention in which 02 concentration is reported as AO2, the reduction in the 02 concentration from the background atmospheric concentration of 21%, the results show a single, low-contrast anomaly at the site of the oxide/sulphide interface with a AO2 value of 0.3% and CO2 value of 0.4%, set against a quiet background of 0.1-0.2% AO2 and 0-0.1% CO2. The other traverse carried out during this survey revealed a similar pattern, and gave a good expression of the mineralisation.
Johnson Camp, Arizona The mineralisation at Johnson Camp (see also Chapter 8) consists of chalcopyrite, sphalerite, bomite and pyrite. It was pyrometasomatically-introduced into limestones of the Naco Group of Upper Pennsylvanian to Lower Permian age by the nearby Tertiary Texas Canyon quartz monzonite intrusive (Cooper and Silver, 1964). The mineralisation has invaded the host rock along minor fractures and as disseminations. The mineralised zones have been partly oxidised near their suboutcrops and, adjacent to the innumerable mineralised minor fractures, along much of their down-dip extensions. The overall
465
Oxygen and carbon dioxide in soil air
0.4
." ~ 16 CO 2 ~ , - - - - , ~ 11, ~O 2
~
.f~
l~
,
, ,oo . ,,o.
OO t ~O2. 0., 2400
. 2:!O0
2200
/,.a. 2100
2000
,~
"
t 800
1800
Cu
-~.....,,,~... 1700
! 000
1800
$400
1800
.~ 1200
zone
................................................................. D~
L o w e r limit o f o o p p e r
oxide mlnorele
SECTION FROM TRENCH. LINE 2, DAHEIM
Fig. 14-9. Oxygen and carbon dioxide in soil air over sedimentary copper mineralisation at Witvlei, Namibia (from Lovell et al., 1983).
sulphide concentration is approximately 3%. Three zones of mineralisation have been outlined by drilling. These are concealed by pediment gravels and alluvium that increase in thickness to the northeast from 10 m over the most southerly ore zone to some 225 m over the most northerly. A soil-gas survey was carried out across this area during a dry summer (Lovell et al., 1983). Three traverses were completed, with soil air being collected from a depth of 50-100 cm; one of these traverses is shown in Fig. 14-10. The suboutcrop and down-dip extension of the shallowest body of mineralisation (Zone I) is marked by a discontinuous series of anomalous values, with maximum values of 0.75% AO2 and 0.9% CO2. These maximum values do not occur over the actual suboutcrop, but at a position where the sulphide zone is some 60 m below the surface. This is then followed by a series of background values in the range 0.15-0.3% AO2 and 0-0.15% CO2, although the traverse is still partially underlain by the down-dip extension. It is interesting to note here that this particular mineralised zone is believed to be an extension, along strike, of the old Peabody Mine (J. Kantor, pers. commun.). Scott (1916) reports that in the mine the ore was oxidised to 60 m. If the same conditions extend into the survey area, then the highest AO2 and CO2 soil-air values are to be found over the projected interface of oxide and sulphide ore. The suboutcrop of the central zone of mineralisation (Zone II), lying to the northeast and covered by 90-150 m of overburden, has a good expression in the soil-gas data, but the down-dip extension is not indicated. This is perhaps a result of oxidation being confined to the upper sections of the mineralisation by the greater depth of cover. The two other traverses conducted during this survey showed that Zones I and II, the shallowest zones, have a consistent expression in the soil air, but that Zone III, the deepest zone, covered by 225 m overburden, had no expression. A subsequent soil-air survey, following a very wet summer, revealed similar patterns in the data, but the values were greatly reduced. The background values for both AO2
J.S. Lovell
466 % CO 2 end ,~02 1.0
o.
t
0.8
2 /
9
0.7
o., ~ , . 0.4
N
Ore -,ones
/\
I !
o.s
~
o.,
~ ~.~,,-~ , , / ~ / ~
~t"~,
~
'~T'~
v
I/",,/ ~,,~ ,~',..,,.~
~
....~" .
.
~,,,/.
'~'--
~
.
,
,~..
.,
,,02
;i
o
P e d i m e n t g r a v e l s end c s l i c h e
1O0
P a l a e o z o i c ,,ediments
,oo
\\~o.,, i
0
|
100
i
200
|
300
1
400
i
600
,, i
800
700
1
800
coo
,0;0
,1;0
1200
,300
14;0
1800
1coo
, 700 '
1800
1.00
Metere
Fig. 14-10. Oxygen and carbon dioxide in soil air over pyrometasomatic mineralisation at Johnson Camp, Arizona (from Lovell et al., 1983).
and CO2 were below the limits of detection, with maximum values of 0.2-0.4% over the mineralisation (B.W. Oakes, pers. commun.).
Colorado Plateau, Arizona Mineralised breccia pipes occur in Palaeozoic sedimentary rocks of the Colorado Plateau in northern Arizona (Weinrich, 1985). These pipes are usually circular or oval at suboutcrop, have horizontal dimensions that are typically a few tens of metres and vertical dimensions that may extend to 1000 m. They appear to have developed over solution-collapse structures resulting from karstic weathering of Lower Mississippian carbonate sediments between the Upper Mississippian and Triassic. The mineralisation, comprising pitchblende and sulphides of Fe, Cu, Mo, Pb and Zn, is found beneath a massive pyrite cap, several hundreds of metres below surface. Some of the ore bodies are, or have been, mined for uranium. Gangue minerals include calcite and dolomite. The surface features of the collapse structures include: concentric, inward-dipping beds; circular areas of brecciated rock; circular bleached or ferruginous tonal anomalies; circular vegetation anomalies and circular concave topography. However, not all of these structures are mineralised, and the concentration of CO: in the overlying soils has proved to be a valuable guide to the extent of the concealed mineralisation, which may lie as much as 500 m below the surface. The development of a gaseous expression of mineralisation at such considerable depths is facilitated by a number of features. The sulphide content of the ore is very high (tens of percent) and its oxidation far below the
Oxygen and carbon dioxide in soil air
467
Fig. 14-1 I. Carbon dioxide in soil air (sample sites and contours in %) over mineralised breccia pipe, northern Arizona (from Lovell and Reid, 1989).
ground surface is promoted by a low water table and continuous supply of oxygenated groundwater, as a result of the deep canyons that developed during the uplift of the Colorado Plateau. In addition, the isolated and transgressive nature of the breccia structures appears to favour vertical gaseous movement and to limit the gaseous anomalies to the confines of the pipe. The low organic content and the poorly-developed nature of the surface soils yields a low CO2 background and permits subtle expressions of the mineralisation to be recognised. Figure 14-11 shows contoured soil-air CO2 data collected from a depth of 1 m, overlying a breccia pipe which is heavily mineralised at depths of 150-250 m. The anomaly peak is only 0.3% CO2 (only ten times the atmospheric background) but the circular gas halo is centred over the mineralisation (Lovell and Reid, 1989; Reid and Rasmussen, 1990). The fact that these low concentrations accurately reflect mineralisation at these depths is a testament to the sensitivity and precision of the gaschromatographic method of analysis that was used, and to the low background. This technique was subsequently routinely adopted as an exploration tool in northern Arizona. However, prior to its adoption, a series of trials were carried out to determine its applicability and confidence levels in this environment. Of 14 prospects that were tested, 12 were subsequently drilled. From the soil-air CO2 data it was possible to predict correctly both the presence and tenor of the underlying mineralisation in 11 out of the 12
468
J.S. Lovell
Fig. 14-12 Variation of CO2 patterns in soil air over mineralised breccia pipe, northern Arizona: a) early summer; b) 18 days later; c) a further 10 days later (from Lovell and Reid, 1989).
cases (A.R. Reid, pers. commun.), an excellent success rate for any exploration technique. Further surveys at different times of the year proved disappointing: a series of exploration targets failed to return any detectable concentrations of CO2 in the soil air; and re-sampling of areas that had previously been shown to contain CO2 related to mineralisation also failed to yield a response. As these further surveys had been carried out during the dry summer months, it was surmised that the dry soils had no capacity to retain CO2 in concentrations significantly out of equilibrium with the atmosphere. This phenomenon was studied by monitoring, for several weeks, a grid of soil sites over mineralisation (Fig. 14-12). As the soil dried, the anomaly over mineralisation faded and shifted. A later survey after heavy rainfall showed that the anomaly was re-established, and demonstrates the continuous evolution of CO2 from its source. Other surveys after winter rain and snow also revealed anomalies, including an anomaly over the one mineralised breccia pipe that was not anomalous in the initial CO2 survey of prospects.
DISCUSSION Since the techniques described here measure transient fluxes of 02 and CO2 in soil air, it is hardly surprising that measurements are not always reproducible over long
Oxygen and carbon dioxide in soil air
469
periods of time. At Johnson Camp, Arizona, surveys in different seasons yielded comparable patterns but at different concentration levels. Over the breccia pipes of the Colorado Plateau the distinctive but subtle anomalies found in one season were completely lost in another. Ball et al. (1990) found a similar seasonal pattern over auriferous sulphides in West Africa, with CO2 anomalies in soil air detected during the wet season disappearing by the end of the dry season. Even in the temperate climate of the UK, Ball et al. (1983b) found the best 02 and CO2 responses over mineralisation in the wetter winter months. The inference that dryness of the soil plays a significant role in soil-air CO2 anomaly persistence gains support from Hinkle (1990), who found that, amongst a considerable number of soil-air environment factors investigated, soil moisture content influences the CO2 level in soil air the most, although its effects are rather unpredictable. It is likely that isotopic analysis to determine the ~3C and ~80 content of various soilair samples offers a means of distinguishing those containing some CO2 generated during sulphide oxidation at depth from those whose CO2 derives solely from the atmosphere and plant respiration. This has been attempted by Alpers et al. (1990) at Crandon, Wisconsin, where McCarthy et al. (1986) found a CO2 anomaly in soil air over massive sulphides hosted by Proterozoic limestones and concealed beneath glacial till containing clasts of Palaeozoic limestone. In fact the CO2 in soil air over the mineralisation was isotopically indistinguishable from that in background samples 1-2 km away. This finding is, however, far less conclusive than it first seems to be: at the time when the samples where taken for isotopic analysis, the CO2 anomaly over the mineralisation was not particularly well developed due to recent precipitation.
CONCLUSIONS There is a substantial body of evidence to support the belief that the concentrations of 02 and CO2 in soil air can reveal the position of concealed mineralisation, even through considerable thicknesses of overburden. Surveys can be conducted using any one of several proven field and laboratory methods. The equipment required is, for the most part, inexpensive and readily available. However, there is a very wide range of concentrations of 02 and CO2 in the soil air of different environments. Also, measurements can vary markedly with the seasons. Thus it is extremely important when applying the technique in a given area to carry out local orientation tests in order to establish the true background and the contrast that may be expected.
This Page Intentionally Left Blank
471
REFERENCES
Abaturova, M.A. and Shkarupa, P.N., 1981. Application of the MPF method for prospecting quartz poor-sulphide mineralisation zones. In: Investigation and Application of the Electrochemical Processes for Prospecting and Exploration of Mineral Resources. Rotaprint VITR, Leningrad, pp. 56-60 (in Russian). Acharya, C.L. and Prihar, S.S., 1969. Vapour losses through soil mulch at different wind velocities. Agronomy J., 61: 666-668. Adams, D.F., Farweli, S.O., Pack, M.R. and Robinson, E., 1981. Biogenic sulphur gas emissions from soils in eastern and southeastern United States. J. Air Pollution Control Assoc., 31 : 1083-1089. Adams, J.A.S. and Lowder, W.M. (eds.), 1964. The Natural Radiation Environment. Univ. Chicago, 1069 pp. Adams, J.A.S., Lowder, W.M. and Gesell, T.F. (eds.), 1972. The Natural Radiation Environment II. National Technical Information Service, US Dept. of Commerce, Springfield, Va., 959 pp. Adkinson, C.W. and Reimer, G.M., 1976. Helium and radon emanation bibliography - selected references of geologic interest to uranium exploration. US Geol. Survey, Open File Report, 76860, 44 pp. Aftabi, A. and A~aria, L.M., 1983. Distribution of mercury compounds in ore and host rocks at Sigma gold mine Val d'Or, Quebec, Canada. J. Geochem. Explor., 19: 447-464. Alckseev, S.G., Dukhanin, A.S., Veshev, S.A. and Voroshilov, N.A., 1996. Some aspects of practical use of geoelectrochemical methods of exploration for deep-seated mineralisation. J. Geochem. Explor., 56: 79-86. Alekseev, S.G, Goldberg, I.S, Veshev, S.A, Gracheva, T.R., Dukhanin, A.S. and Ivanova, A.V., 1981. Regularities of the electrochemical extraction of elements in the simplest physico-chemical systems. In: Investigation and Application of the Electrochemical Processes for Prospecting and Exploration of Mineral Resources. Rotaprint VITR, Leningrad, pp. 38-48 (in Russian). Alekseev, V.V., Grammakov, A.G., Nikonov, A.I. and Tafeev, G.P. (eds.), 1959. Radiometric methods in the prospecting and exploration of uranium ores. Translation Series, AEC-TR-3738 (Book 1) pp. 1-228, (Book 2) pp. 339-640. Allen, C.R., Bonilla, M.G., Brace, W.F., Bullock, R.W., Hamilton, R.M., Hofheinz, R., Kisslinger, C., Knopoff, L., Park, N., Press, F., Raleigh, C.B. and Sykes, L.R., 1975. Earthquake research in China. Eos Trans. AGU, 56: 838-881. Allen, J.W., 1976. Development of a portable radon detection system. US Dept of Energy, Open File Report, GJBX-50, pp. 1-32. Alpers, C.N., Dettrnan, D.L., Lohman, K.C. and Brabec, D., 1990. Stable isotopes of carbon dioxide in soil gas over massive sulphide mineralisation at Crandon, Wisconsin. J. Geochem. Explor., 38: 69-86.
472
References
Anderson, D.M., Orgel, L.E., Oro, J., Owen, T., Shulman, G.P., Toulmin, P.and Urey, H.C., 1972. Mass spectrometric analysis of organic compounds, water and volatile constituents in the atmosphere and surface of Mars. Icarus, 16:111-38. Anderson, R.J., 1977. Environmental effects on an emanometric survey. Unpubl. BSc thesis, Brock Univ., St. Catherines, Ontario, 33 pp. Anderson, R.K., Scalen, R.S., Parker, P.L. and Behrens, E. W., 1983. Seep oil and gas in Gulf of Mexico slope sediment. Science, 222:619-621. Anderson, T.F. and Arthur, M.A., 1983. Stable isotopes of oxygen and carbon and their application to sedimentologic and palaeoenviromnental problems. In: M.A. Arthur (ed.), Stable Isotopes in Sedimentary Geology. Society for Sedimentary Geology, Short Course Manual 10, pp. 1.1 - 1.151. Andersson, A., 1979. Mercury in soils. In: J.O. Nriagu (ed.), Biogeochemistry of mercury. Elsevier, Amsterdam, pp. 79-106. Andrews, J.N. and Lee, D.J., 1979. Inert gases in groundwaters from the Bunter Sandstone of England as indicators of age and palaeoclimatic trends. J. Hydrol., 41: 233-252. Andrews, J.N. and Wood, D.F., 1972. Mechanism of radon release in rock matrices and entry into groundwaters. Trans. Instn. Min. Metall., 87: B 198-209. Andrews, J.N., Giles, I.S., Kay, R.L.F., Lee, D.J., Osmond, J.K., Cowart, J.B., Fritz, P., Barker, J.F. and Gale, J., 1982. Radioelements, radiogenic helium and age relationships for groundwaters from the granites at Stripa, Sweden. Geochim. Cosmochim. Acta, 46: 1533-1543. Antonov, P.L., Bartashervich, O.V., Zorkin, L.M., Mogilevskii, G.I. Mysichenko, N.I., Proshnievo, N.V. and Stroganov, V.A., 1971. The theory and methodology of geochemical prospecting for oil and gas. In: The results of research studies - application of direct geochemical methods in prospccting for oil and gas. Nicodva, Moscow. Antropov, P.Y., 1981. Laser gas analysis in solving geological and production problems. Intl. Geol. Rcv., 23: 314-318. Antropova, L.V., 1975. Forms of Element Occurrence in Dispersion Halos of Ore Deposits. Nedra, Lcningrad, 145 pp. (in Russian). Antropova, L.V., Goidberg, I.S., Voroshilov, N.A. and Ryss, Yu.S., 1992. New methods of regional cxploration for blind mincralisations: application in the USSR. J. Geochem. Expior., 43: 157-166. Arp, G.K., 1992. An integrated interpretation for the origin of the Patrick Draw oil-field sage anomaly, Amcr.Assoc. Pet. Geol. Bull., 76: 301-306. Asikaincn, M., 1981. State of disequilibrium between 238U, 234U,226Ra and 222Rn in groundwater from bedrock. Geochim. Cosmochim. Acta, 45: 201-206. Asikainen, M. and Kahlos, H., 1979. Anomalously high concentrations of uranium, radium and radon in watcr from drilled wells in the Heisinki region. Geochim. Cosmochim. Acta, 43: 1681-1686. Asimov, 1., 1966. The noble gases. Basic Books, Inc., New York, 171 pp. Austin, S.R., 1975. A laboratory study of radon emanation from domestic uranium ores. Radon in uranium mining, IAEA Vienna, PL-565/8, pp. 151-160. Bailey, L.D. and Beauchamp, E.G., 1973. Gas chromatography of gases emanating from a saturated soil system. Can. J. Soil Sci., 53: 122-24. Bailey, R.V. and Childers, M.O., 1977. Applied mineral exploration with special reference to uranium. Westview Press, Boulder, Colorado, 542 pp. Bajc, A.F., 1998. A comparative analysis of enzyme leach and Mobile Metal Ion selective extractions; case studies from glaciated terrain, northern Ontario. J. Geochem. Explor., 61 : 114-138. Ball, G.W., Lawson, G. and Todd, J.G.J., 1973. Application of dynamic mass spectrometers to problems in gas analysis. Dynamic Mass Spectrometry, 3: 99-182.
References
473
Ball, T.K., Crow, M.J., Laffoley, N., Piper, D. and Ridgway, J., 1990. Application of soil-gas geochemistry to mineral exploration in Africa. J. Geochem. Explor., 38:103-115. Ball, T.K., Hextall, J.P., Nicholson, R.A. and Peachey, D., 1983a. Application of Orstat stack-gas analyser to geochemical exploration research. Trans. Instn. Min. Metall., 92:B49-51. Ball, T.K., Nicholson, R.A. and Peachey, D., 1983b. Effects of meteorological variables on certain soil gases used to detect buried ore deposits. Trans. Instn. Min. Metall., 92: B 183-190. Bamesberger, W.L. and Adams, D.F., 1969. Improvements in the collection of hydrogen sulphide in cadmium hydroxide suspension. Environ. Sci. and Tech., 3:258-261. Bammel, B.H. and Bimic, R.W., 1994. Spectral response of big sagebrush to hydrocarbon-induced stress in the Bighom Basin Wyoming. Photogrammetric Engineering and Remote Sensing, 60: 8796. Banwart, W.L. and Bremner, J.M., 1975. Formation of volatile sulphur compounds by microbial decomposition of sulphur-containing amino acids in soils. Soil Biol. Biochem., 7: 359-364. Baranov, V.I., 1961. Radiometry. AEC-TR-4432, 381 pp. (translation from Izdatel'stvo Akademii Nauk SSSR, 1956). Baranov, V.I. and Khristianova, L.A., 1966. Radioactivity of ocean deposits. Chemistry of the Earth's Crust, pp. 425-432 (translated from Russian). Baranov, V.I., Morozova, N.G., Akimova, T.G. and Orlova, A.V., 1968. Naturally radioactive elements in surface and soil groundwaters. Geokhimiya, 3:334-340 (in Russain). Barker, C., 1982. Geochemical considerations of deep earth gas including the thermodynamic stability and migration of hydrocarbons in the mantle. In: Deep Source Gas Workshop, METC, Morgantown, WV, May 3-4. Bames, H.L., 1979. Solubilities of ore minerals. In: H.L. Barnes (ed.), The Geochemistry of Hydrothermal Ore Deposits (2nd edn.). John Wiley and Sons, New York, pp. 404-454. Barnes, H.L., Romberger, S.B. and Stemprok, M., 1967. Ore solution chemistry II: Solubility of HgS in sulphide solutions. Econ. Geol., 62: 957-982. Barnes, 1., O'Neil, J.R., and Trescases, J.J., 1978. Present day serpentinisation in New Caledonia, Oman and Yugoslavia. Gcochim. Cosmochim. Acta, 42: 144-145. Barnes, I., Platfker, G., White, D. and Armstrong, A.K., 1980. Potential natural gas in the Gulf of Alaska indicated by calcite depleted in carbon-13. In: W.L. Coonrad (ed.), The United States Geological Survey in Alaska - accomplishments during 1980. US Geol. Survey, Circular 844, pp. 143-146.Barretto, P.M.C., 1975. Radon-222 emanation characteristics of rocks and minerals. Reprint from "Radon in Uranium Mining", IAEA, Vienna, pp. 129-150. Barretto, P.M.C., Clark, R.B. and Adams, J.A.S., 1972. Physical characteristics of radon-222 emanation from rocks, soils and minerals: its relation to temperature and alpha dose. In: J.A.S. Adams, W.M. Lowder and T.F. Gesell (eds.), The Natural Radiation Environment II. National Technical Information Service, US Dept. of Commerce, Springfield, Va., pp. 731-740. Barringer, A.R., 1981. Airbome geochemical reconnaissance methods for oil and gas exploration. In: B.M. Gottlieb (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas II. Southern Methodist University Press, Dallas, pp. 219-239. Barsukov, V.L., Serebrennikov, V.S., Varshal, G.M. and Garanin, A.V., 1979. Geochemical methods of predicting earthquakes. Geochem. Intemat., 16:1 - 13. Bartlett, J.B. and James, R.J., 1995. System for categorising soil redox status by chemical field testing. Geoderma, 68:21 I-218. Bashorin, V.N., 1980. The helium survey as a method of studying long-lived faults. Dokl. Akad. Nauk. SSSR, 255:91-93.
474
References
Bass Becking, L.G.M., Kaplan, J.R. and Moore, D., 1960. Limits of natural environment in terms of pH and oxidation- reduction potentials. J. Geol., 68: 243-284. Batard, F., Baubron, J.C., Busch, B., Marce, A. and Risler, J.l., 1982. Isotopic identification of gases of a deep origin in French thermomineral waters. J. Hydrol., 56: 1-21. Bateman, A.M., 1950. Economic mineral deposits (2nd edn.). Wiley, New York. Baver, L.D., Gardner, W.H. and Gardner, W.R., 1972. Soil Physics (4th edn.). John Wiley and Sons, New York, 498 pp. Behounek, F., 1927. Uber die Verhaltnisse der Radioactivitat im Uranpecherz-bergbaurevier von St. Joachimstal in Bohmen. (Radioactivitat der Quellen, Boden-und Grabenluft und der Atomosphere.) Physik. Zeitchrift XXVIII, pp. 333-342. Bensman, V.R., Goldberg, I.S., Grigorovich, V.V., Krivodubsky, R.I. and Timchuk, A.V., 1982. Application of the CHIM method for deep-seated beryllium deposit prospecting. In: Investigation and Application of Physico-Chemical and Geoelectrochemical Processes for Prospecting and Exploration of Mineral Resources. NPO Rudgeofizika, Leningrad, pp. 82-88 (in Russian). Bernard, B.B., 1982. Geochemical prospecting for petroleum and natural gases. In: Workshop Report on Basic Research in Organic Geochemistry Applied to National Energy Needs, 1980. National Technical Information Service Report. CONF-801249, pp. B7-B33. Bernard, B.B., Brooks, J.M. and Sackett, W.M., 1976. Natural gas seepage in the Gulf of Mexico. Earth and Planet. Sci. Let., 31: 48-54. Bernard, B.B., Brooks, J.M. and Sackett, W.M., 1978. Light hydrocarbons in recent Texas continental shelf and slope sediments. J. Geophys. Res., 83: 4053-4061. Birchard, G.F. and Libby, W.F., 1980. Soil radon concentration changes preceding and following four magnitude 4.2 -4.7 earthquakes on the San Jacinto fault in Southern California. J. Geophys. Res., 85:3100-3106. Blackmer, A.M. and Bremner, J.M., 1977. Gas chromatographic analysis of soil atmosphere. Soil Sci. Soc. Amer.J., 41: 908-12. Bockris, J.O'M. and Reddy, A.K.N., 1970. Modern Electrochemistry, Vol. 1, Plenum Press, New York, 622 pp. Bogoslovskaya, T.N., Grammakov, A.G., Kirikov, A.P. and Tverskoy, P.N., 1932. Account of the work of the experimental - methodical group of Kavgolovsk in 1931 (Otchet Krabote Kavgolovskoy opytno-metodichcskoy parti v 193 ! g). Izv. Vses. Geol. Razved. Ob'yedineniye, 51: 1283-1293 (in Russian). Bolviken, B. and Logn, O., 1975. An electrochemical model for element distribution around sulphide bodies. In: I. Elliot and K. Fletcher (eds.), Geochemical Exploration 1974, Elsevier, Amsterdam, pp. 631-648. Bonotto, D.M. and Andrews, J.N., 1999. Transfer of radon and parent nuclides 238U and 234U from soils of the Mendip Hills area, England, to the water phase. J. Geochem. Explor., 66: 255-268. Boone, W.J., 1958. Helium-bearing natural gases of the United States: analyses and analytical methods. US Bureau of Mines, Bull. 376, 115pp. Borodzich, E.V., Yeremeyev, A.N. and Yanitskii, I.N., 1979. Preliminary results of helium pattern variations in seismically active zones. Geochem. lnternat., 16: 37-41. Bottomer, L.R., 1986. Precious metal mineralisation in the Drake area, northeastern NSW. Aust. J. Earth Sci., 33:457 -473. Bottomley, D.J., Ross, J.D. and Clarke, W.B., 1984. Helium and neon isotope geochemistry of some groundwaters from the Canadian Precambrian Shield. Geochim. Cosmochim. Acta, 48: 19731985.
References
475
Bowles, C.G., Reimer, G.M., Been, J.M. and Murray, D.G., 1980. Helium investigations in the Edgemont Uranium District, Southern Black Hills, South Dakota and Wyoming. US Geol. Survey, Open File Report 80-1077, 30 pp. Boynton, D. and Reuther, W., 1938. A way of sampling soil gases in dense subsoils and some of its advantages and limitations. Soil Sci. Soc. Amer.Proc., 3: 37-42. Breck, D.W., 1974. Zeolite Molecular Sieves: Structure, Chemistry and Use. John Wiley and Sons, New York, 771 pp. Bremner, J.M. and Banwart, W.L., 1974. Identifying volatile S compounds by gas chromatography. Sulphur Inst. J., 10: 6-9. Brock, J.S., 1972. The use of dogs as an aid to exploration for sulphides. Western Miner, 45, 12: 28-32. Brody, S.S. and Chaney, J.E., 1966. Flame photometric detector. J. Gas Chromatog., 4: 42. Brooks, J.M., Kennicott, M. C. and Carey, B. D., 1986. Offshore surface geochemical exploration. Oil and Gas J., 84: 66-72. Brooks, R.R., 1972, Geobotany and Biogeochemistry in Mineral Exploration. Harper and Row, New York, 290pp. Bukova, E.N., 1959. The formation of heavy gaseous hydrocarbons in the anaerobic decomposition of organic substances. Geol. Nefti Gaza, V. 3, No. 8, pp. 44-47. Bulashevich, Y.P. and Bashorin, V.N., 1971. Confinement of high helium concentrations in groundwater to intersections of faults. Doki. Akad. Nauk SSSR, 201: 36-37. Bulashevich, Y.P., and Kartashov, N.P., 1976. Application of helium to structural studies in mineralised areas. Doki. Akad. Nauk SSR, 229:137-139. Bulashevich, Y.P., Bashorin, V.N., Druzhinin, V.S. and Rybalka, V.M., 1973. Helium in groundwater along the Sverdlovsk deep seismic sounding traverse. Doki. Akad. Nauk SSSR, 208:15-18. Bunting, B.T. and Campbell, J.A., 1975. Improvements in sampling, storing and analysis of soil air for gas chromatography. Can. J. Soil Sci., 55: 68-71. Burford, J.R. and Stefanson, R.C., 1973. Measurement of gaseous losses of nitrogen from soils. Soil Biol. Biochcm., 5: 133-141. Burr, S.V., 1982. A guide to prospecting by the self-potential method. Ontario Geol. Survey, Misc. Paper 99: I - 15. Burson, K.R. and Thompson, C.K., 1985. The successful utilisation of geochemical reconnaissance in petroleum exploration (abs.). Presented at ACS 190th National Meeting, Chicago, IL, September 8-13. Burtell, S.G., 1988. Geochemical investigations at Arrowhead Hot Springs, San Bernardino, and along the San Andreas Fault in southern California. Unpubl. MS thesis, Univ. Pittsburgh. Burtell, S.G., Jones, V.T., S.G., Hodgson, R.A., Okada, K., Ohhashi, T., Kuniyasu, M. Ando, T. and Komai, J., 1986. Remote sensing and surface geochemical study of Railroad Valley, Nye County, Nevada - detailed grid study. Presented at Fifth Thematic Conference, Remote Sensing for Exploration Geology, Reno, Nevada, September 29 - October 2. Busigin, A., Van der Vooren, A.W., Lin Pai, H. and Phillips, C.R., 1979. Development of instrumentation for the measurement of radon, thoron and their daughters in uranium mines and their environment. Minerals Sci. Eng., 11: 193-206. Butt, C.R.M. and Gole, M.J., 1984. Helium determinations in mineral exploration. Aust. Dept. Resources and Energy, Report no. NERDDP/EG/84/271,351 pp. Butt, C.R.M. and Gole, M.J., 1985. Helium in soil and overburden gas as an exploration pathfinder - an assessment. J. Geochem. Explor., 24: 141-173. Butt, C.R.M. and Gole, M.J., 1986. Groundwater helium surveys in mineral exploration in Australia. J. Geochem. Explor., 25: 309-344.
476
References
Butt, C.R.M., Gole, M.J. and Hart, M.K.W., 1987. A review of Australian experience in the evaluation of helium as a pathfinder in exploration. J. Geophys. Res. 92: 12595-12603. Butt, C.R.M., Hart, M.K.W. and Gole, M.J., 1985. Gases and trace elements in soils at the North Silver Bell Deposit, Pima County, Arizona: Discussion. J. Geochem. Explor., 24: 129-134. Byman, J.E., 1977. Investigation of sulphur vapour techniques in exploration geochemistry. Unpubl. M. Phil. thesis, Univ. London. Caneer, W.T. and Saum, N.M., 1974. Radon emanometry in uranium exploration. Mining Eng., 26: 26-28. Card, J.W. and Bell, K., 1979. Radon decay products and their application to uranium exploration. Can. Inst. Min. Metall. Bull., 72:81-87. Carr, G.R., 1981. The mineralogy, petrology and geochemistry of the zinc-lead-silver ores and host sedimentary rocks at Lady Loretta, northeast Queensland. Unpubl. Ph.D. thesis, Univ. Wollongong, NSW. Carr, G.R., 1984. Primary mineralogical and geochemical dispersion in the vicinity of the Lady Loretta Zn-Pb-Ag deposit, Northwest Queensland. J. Geochem. Explor., 22: 217-238. Carr, G.R., Wilmshurst, R.R. and Ryall, W.R., 1986. Evaluation of mercury pathfinder techniques: base metal and uranium deposits. J. Geochem. Expior., 26: 1-117. Carter, J.S. and Koger, D.G., 1988. Successful applications of remotely-sensed data for oil and gas exploration. Proc. Sixth Thematic Conference on Remote Sensing for Exploration Geology, Houston, vol. !, pp. 19-25. Chalov, P.I., Tuzova, T.V. and Alekhina, V.M., 1976. Fluctuations of helium and radon anomalies in some springs located along faults of the north Tien-Shan seismic zone during the period preceding the Kochkor earthquake of 1974. Doki. Akad. Nauk SSSR, 23 I: 26-28. Charbonneau, B.W., Ford, K.L. and Cameron, G.W., 1981. Equilibrium between U and cU (214Bi) in surface rocks of Canada. Geol. Survey Canada, Paper 81-IC, pp. 45-50. Chcng, K.C. and Porritt, J.W.M., 1981. The measurement of radon emanation rates in a Canadian cutand-fill uranium mine. Can. Inst. Min. Metall. Bull., 74: i 10-118. Chcrdyntscv, V.V., 1961. Abundance of Chemical Elements (translated by Walter Nichiporuk). Univ. Chicago Press, 304 pp. Chcrdyntsev, V.V., 1969. Uranium-234. IPST, Jerusalem, 234 pp. Chung, Y., 1985. Radon variations at Arrowhead and Murrieta Springs; continuous and discrete measurements. Pure Appl. Geophys., 122: 294-358. Clark, I.D., 1987. Groundwater resources in the Sultanate of Oman: origin, circulation times, recharge processes and palaeoclimatology. Doctoral thesis, Universite de Paris Sud, France, 264 pp. Clark, J.R., 1992. Detection of bedrock-related geochemical anomalies at the surface of transported overburden. Explore, 76:4-11. Clark, J.R., 1993. Enzyme-induced leaching of B-horizon soils for mineral exploration in areas of glacial overburden. Trans. Instn. Min. Metall., 102: B19-29. Clark, J.R., 1996. Unique significant enzyme leach anomaly patterns in areas of tropical/subtropical weathering. Proc. Conference Queen's University, Kingston, Onatrio, pp. 241-256. Clark, J.R., 1997. Concepts and models for interpretation of enzyme leach data for mineral and petroleum exploration. Internal report, Enzyme- Actlabs, 62 pp. Clarke, W.B. and Kugler, G., 1973. Dissolved helium in groundwater: a possible method for uranium and thorium prospecting. Econ. Geol., 68: 243-251. Clarke, W.B., Beg, M.A. and Craig, H., 1969. Excess 3He in the sea: evidence for terrestrial primordial helium. Earth Planet. Sci. Lett., 6: 213-220.
References
477
Clarke, W.B., Top, Z., Beavan, A.P. and Gandhi, S.S., 1977. Dissolved helium in lakes - uranium prospecting in the Precambrian terrain of Central Labrador. Econ. Geol., 72: 233-242. Clements, W.E., 1974. The effect of atmospheric pressure variation on the transport of 222Rn from the soil to the atmosphere. Unpubl. PhD thesis, New Mexico Institute of Mining and Technology, 111 PP. Cochran, J.K., 1979. The geochemistry of 226Ra and 228Ra in marine deposits. Unpubl. PhD thesis, Yale Univ., 259 pp. Coleman, D.D., 1979. the origin of drift-gas deposits as determined by radiocarbon dating of methane. In: R. Berger and H.E. Seuss (eds.), Radiocarbon Dating, Proc. of the Ninth International Radiocarbon Dating Conference, 1976. University of California Press, Berkley, pp. 365-387. Coleman, D.D., Meents, W.F., Liu, C. and Keogh, R.A., 1977. Isotopic identification of leakage gas from underground storage reservoirs. Illinois Petroleum, 111, Illinois State Geol. Survey. Collins, W., Chang, S.H., Raines, G., Canney, F. and Ashley, R., 1983. Airborne biogeochemical mapping of hidden mineral deposits. Econ. Geol., 78: 737-743. Conel, J.E. and Alley, R.E., 1985. Lisbon Valley, Utah, uranium test site report. In: M.J. Abrams, J.E. Conei, H.R. Lang and H.N. Paley (eds.), Joint NASA/Geosat Test Case Project: final report. Amer.Assoc. Pet. Geol., Spec. Publ., part 2, vol. 1, pp. 8.1-8.158. Cook, E., 1979. The Helium Question. Science, 206:1141-1147. Cook, G.A., 1961. Argon, Helium and the Rare Gases. Wiley Interscience, New York. Cooper, J.R. and Silver, L.T., 1964. Geology and ore deposits of Dragoon Quadrangle, Arizona. US Geol. Survey, Prof. Paper 416. Corazza, E., Magro, G., Pieri, S. and Rossi, U., 1993. Soil gas survey in the geothermal area of Bolscna Lake (Vulsini Mountains, central Italy). Geothermics, 22: 201-214. Corbett, K.D., 1981. Stratigraphy and mineralisation in the Mt. Read Volcanics, western Tasmania. Econ. Geol., 76: 209-230. Cox, M.E., 1980. Ground radon survey of a geothermal area in Hawaii. Geophys. Res. Lett., 7: 283286. Cox, R. and Curtis, R., 1977. The discovery of the Lady Loretta zinc-lead-silver deposit, northwest Queensland, Australia- a geochemical exploration case history. J. Geochem. Explor., 8: 189-202. Craig, H., Clarke, W.B. and Beg, M.A., 1975. Excess 3He in deep water on the East Pacific Rise. Earth Planet. Sci. Lett., 26:125-132. Craig, H., Lupton, J.E. and Horibe, Y., 1978a. A mantle component in Circum-Pacific volcanic gases: Hakone, the Mananas and Mt. Lassen. In: E.C. Alexander and M. Ozima (eds.), Advances in Earth and Planetary Sciences 3, Terrestrial Rare Gases. Center for Academic Publications, Japan, pp. 316. Craig, H., Lupton, J.E., Welhan, J.A. and Poreda, R., 1978b. Helium isotope ratios in Yellowstone and Lassen Park volcanic gases. Geophys. Res. Lett., 5: 897-900. Crawford, M.F., 1986. Preliminary evaluation of remote-sensing data for detection of vegetation stress related to hydrocarbon microseepage: Mist gas field, Oregon. Proc. Fifth Thematic Conference on Remote Sensing for Exploration Geology, Reno, pp. 161-177. Currie, J.A., 1960. Gaseous diffusion in porous media. Brit. J. Appl. Physics, 11:318. Currie, J.A., 1970. Movement of gases in soil preparation. Soc. Chem. Ind., Monograph 37: 152-169. Cwick, G.J., Bishop, M.P., Howe, R.C., Mausel, P.W., Everitt, J.H. and Escobar, D.E., 1995. Muitispectral video data for detecting biogeochemicai conditions at an Alabama oil field site. Geocarto International, 10: 59-66.
478
References
Dadashev, F.G., 1982. The background of methane homologs in near surface sediments as a regional indicator of source rock potential. Ne~oegozovaya Geol., Geofiz, No. 1, pp. 19-20. Dadashev, F.G., Guseinov, R.A. and Zykov, Yu., 1971. Gas surveying of sulphide ore deposits. Geology of Ore Deposits, Nauka Publ., Moscow, pp. 111-112 (in Russian). Dalziel, M.C. and Donovan, T.J., 1980. Biogeochemical evidence for subsurface hydrocarbon occurrence, Recluse oil field, Wyoming: preliminary results: US Geol. Survey, Circular 837, 11 pp. Dando, P.P., O'Hara, S.C.M., Schuster, U., Yaylor, L., Clayton, C.J., Bayliss, S. and Laier, T., 1994. Gas seepage from a carbonate-cemented sandstone reef on the Kattegat coast of Demnark. Marine Pet. Geol., 11 : 182-189. Datta, P.S., Gupta, S.K., Jayasurya, A., Nuampurkar, V.N., Sharma, P. and Plusnin, M.I., 1980. A survey of helium in groundwater in parts of the Sabarmarti basin in Gujarat State and in the Jaisalmer district, Rajasthan. Hydrol. Sci. Bull., 26: 183-193. Davidson, M.J., 1963. Geochemistry can help find oil if properly used. World Oil, July, pp. 94-106. Davis, J.B., 1967. Petroleum Microbiology. Elsevier, New York, 604 pp. Davis, J.B. and Squires, R.M., 1954. Detection of microbially produced gaseous hydrocarbons other than methane. Science, 119: 381-382. Davy, R., 1974. The geochemical and radiometric behaviour of uranium and other radioactive elements in relation to exploration for secondary uranium. Amdel Bull., 17: 17-32. Davy, R. and Stokes, M., 1977. A preliminary evaluation of sulphur dioxide as a prospecting tool in WA. Annual. Report, Western Australia Geol. Survey, pp.72-76. De Camargo, O.A., Grohmann, F., Salati, E. and Matsui, E., 1974. A technique for sampling the soil atmosphere. Soil Sci., 117: 173-174. De Jong, E., and Paul, E.A., 1979. Aeration, respiration and atmosphere. In: R.W. Fairbridge and C.W. Finkl (eds.), The Encyclopaedia of Soil Science, Part 1. Dowden, Hutchinson and Ross Inc., Stroudsberg, pp. 10-21. De Souza, T.L.C., Lane, D.E. and Bhatia, S.P., 1975. Analysis of sulphur-containing gases by gassolid chromatography on a specially treated Porapak QS column packing. Anal. Chem., 47: 543544. Debnam, A.H., 1969. Geochemical prospecting for petroleum and natural gas in Canada. Geol. Survey of Canada Bull. 177, 26 pp. Delwiche, C.C., 1958. Weathering of great world soil groups as related to general atmospheric radioactivity. In: Waste Treatment and Environmental Aspects of Atomic Energy, 18:551-556. Dennis, N.T.M. and Heppell, T.A., 1968. Vacuum System Design. Chapman and Hall, London. Denton, E.H., 1977. Helium sniffer field test: Roosevelt Hot Springs, Utah. US Geol. Survey, Open File Report, 77-606, 10 pp. Derjaguin, B.V., Zorin, Z.M. and Sobolev, V.D., 1980. Properties of the thin water layer close to a solid surface. The Attached Water in the Dispersion Systems, 5:4-13 (in Russian). Devine, S.B., 1977. Soil hydrocarbon geochemistry, a potential petroleum exploration tool in the Cooper Basin, Australia. J. Geochem. Explor., 8:397-414. Devine, S.B. and Sears, H.W., 1977. Soil hydrocarbon geochemistry, a potential petroleum exploration tool in the Cooper Basin, Australia. J. Geochem. Explor., 8: 397-414. Devine, S.B. and Sears, H.W., 1985. An experiment in soil geochemical prospecting for petroleum, Della Gas Field, Cooper Basin. Australian Pet. Explor. Assoc. J., 15:103-110. Dickinson, W.R., 1979. Cenozoic plate tectonic setting of the Cordilleran region in the United States. In: J.M. Armentrout, M.R. Cole and H. TerBest (eds.), Cenozoic Palaeogeography of the Western
References
479
United States. Pacific Section, Soc. Econ. Palaeo. Min., Pacific Coast Palaeogeography Symposium 3, pp. 1 - 13. Dickinson, R.G. and Matthews, M.D., 1993. Regional microseep survey of part of the productive Wyoming-Utah thrust belt. Amer. Assoc. Pet. Geol. Bull., 77: 1710-1722. Dixon, A.J. and Bridge, B.J., 1968. Respiratory quotient arising from microbial activity in relation to matrix suction and air-filled pore space of soil. Nature, 218: 961. Dobrin, M.B. and Savit, C.H., 1988. Introduction to Geophysical Prospecting (4th edn.). McGraw-Hill, New York, 867 pp. Dolly, E.D., 1979. Geologic techniques used in the discovery of Trap Springs Field. In: Rocky Mountain Assoc. Geol., Basin and Range Symposium 1979, pp. 455-467. Donovan, T.J., 1974, Petroleum microseepage at Cement, Oklahoma: evidence and mechanism, Amer. Assoc. Pet. Geol. Bull., 58: 429-446. Donovan, T.J. and Dalziel, M.C., 1977. Late diagenetic indicators of buried oil and gas. US Geol Survey, Open File Rpt. 77-817, 44 pp. Donovan, T.J., Friedman, I. and Gleason, J.D., 1974. Recognition of petroleum-bearing traps by unusual isotopic composition of carbonate-cemented surface rocks: Geology, 2:351-354. Donovan, T.J., Roberts, A.A. and Dalziel, M.C., 1981. Epigenetic zoning in surface and near-surface rocks resulting from seepage-induced redox gradients, Velma oil field, Oklahoma: a synopsis. Oklahoma City Geological Society Shale Shaker, 32(3): 1-7. Dortman, N.B. (ed.), 1992. Petrophysics- Reference Book, the First Book. Rocks and Mineral Resources. Nedra, Moscow, 391 pp. (in Russian). Dowdell, R.J., Smith, R.A., Crees, R., and Restall, S.W.F., 1972. Field studies of ethylene in the soil atmosphere - equipment and preliminary results. Soil Biol. Biochem., 4:325-31. Drever, J.l., 1982. The Geochemistry of Natural Waters. Prentice-Hall, Englewood Cliffs, 437 pp. Driscoll, J.N., Ford, J., Jaramillo, L., Becket, J.H., Hewitt, G., Marshall, J.K. and Onishuk, F., 1978. Developments and applications of the photoionisation detector in gas chromatography. Amer.Lab., 10: 137-147. Drozd, R.J., P~dersky, G.J., Jones, V.T. and Weismann, T. J., 1981. Use of compositional indicators in prediction of petroleum production potential (abs.). Presented at1981 American Chemical Society Meeting, Atlanta, GA, March 29-April 3. Duchscherer W., 1980. Geochemical methods of prospecting for hydrocarbons. Oil and Gas J., 78: 194-208. Duchscherer, W., 1982. Geochemical exploration for hydrocarbons, no new tricks - but an old dog. Oil And Gas J., July, pp. 163-176. Duchscherer, W., 1984. Gcochemical Hydrocarbon Prospecting, with Case Histories. PennWell, Tulsa, 196 pp. Duddridge, G.A., Grainger, P. and Durrance, E.M., 1991. Fault detection using soil gas geochemistry. Quart. J. Eng. Geol., 24: 427-435. Dukhanin, A.S., 1989. Experience of using of the diffusion extraction method for prospecting of deepseated mineral deposits. In: Methods of the Exploration Geophysics, Application of Geoelectrochemical Methods for Ore Deposit Prospecting and Exploration. NPO Rudgeofizika, Leningrad, pp. 99-109 (in Russian). Dukhanin, A.S., 1990. Distribution of heavy metals in gas samples from mineral deposits. Abstracts of the Second All-union Conference for Inorganic Gas Analysis. Leningrad State Univ. pp. 255-256 (in Russian). Dvomikow, A.G. and Petrov, V. Ya., 1961. Some data on the content of mercury in the soils of the Nagollnyi Range. Geochemistry, 10:1021 - 1028.
480
References
Dyck, W., 1968. Radon-222 emanations from a uranium deposit. Econ. Geol., 63: 228-291. Dyck, W., 1969. Radon determination apparatus for geochemical prospecting for uranium. Geol. Survey Canada, Paper 68-21, 13 pp. Dyck, W., 1973. An ionisation chamber for continuous monitoring of atmospheric radon: Radon-222 levels in Ottawa and Gatineau Hills, Canada. Geol. Survey Canada, Paper 73-28, 89 pp. Dyck, W., 1974. Geochemical studies in the surficial environment of the Beaverlodge area, Saskatchewan. Geol. Survey Canada, Paper 74-32, 30 pp. Dyck, W., 1975. Geochemistry applied to uranium exploration. Geol. Survey Canada, Paper 75-26, pp. 33-47. Dyck, W., 1976. The use of helium in mineral exploration. J. Geochem. Explor., 5: 3-20. Dyck, W., 1978. The mobility and concentration of uranium and its decay products in temperate surficial environments. In: Short Course in Uranium Deposits, Min. Assoc. Canada, pp. 57-100. Dyck, W., 1979a. Uranium, radon and helium in waters of the Key Lake area, Saskatchewan. Preprint of talk presented at the Annual Meeting of Can. Inst. Min. Metall., Montreal. Dyck, W., 1979b. Application of hydrogeochemistry of the search for uranium. Geol. Survey Canada, Econ. Geol. Report, 31: 489-510. Dyck, W., 1980. Uranium, radon, helium and other trace elements and gases in well waters of parts of the St.Lawrence Lowlands (Ottawa region), Canada. J. Geochem. Explor., 13: 27-39. Dyck, W., 1982. Hydrogeochemical investigations of the Cypress Hills area, Saskatchewan (unpubl. data). Dyck, W. and Boyle, R.W., 1980. Radioactive disequilibrium in surficial materials from uraniferous environments in northern Saskatchewan. Can. Inst. Min. Metall. Bull., 73: 77-83. Dyck, W. and Da Silva, F.G., 1981. The use of ping-pong balls and latex tubing for sampling the helium content of lake sediments. J. Geochem. Explor., 14: 41-48. Dyck, W. and Dunn, C.E., 1986. Helium and methane anomalies in domestic well waters in southwestern Saskatchewan, Canada, and their relationship to other dissolved constituents, oil and gas fields and tectonic patterns. J. Geophys. Res., 91: 12343-12353. Dyck, W. and Jonasson, I.R., 1977. The nature and behaviour of gases in natural waters. Water Res., 11: 705-711. Dyck, W. and Meilleur, G.A., 1972. A soil gas sampler for difficult overburden. J. Geochem. Explor., 1:199-202. Dyck, W. and Pelchat, J.C., 1977. A semi-portable helium analysis facility. Geol. Survey Canada, Paper 77-1 C, pp. 85-87. Dyck, W. and Tan, B., 1978. Seasonal variations of helium, radon and uranium in lake waters near the Key Lake uranium deposit, Saskatchewan. J. Geochem. Explor., 10: 153-167. Dyck, W., Campbell, R.A. and Pelchat, J.C., 1978. Evaluation of He and Rn geochemical uranium exploration techniques in the Key Lake area, Saskatchewan. Geol. Survey Canada, Paper 78-1B, pp. 39-44. Dyck, W., Chatterjee, A.K., Gemmell, D.E. and Murricane, K., 1976a. Well water trace element reconnaissance, Eastern Maritime Canada. J. Geochem. Explor., 6:140-147. Dyck, W., Dass, A.S., Durham, C.C., Hobbs, J.D., Pelchat, J.C. and Galbraith, J.M., 1971. Comparison of regional geochemical uranium exploration methods in the Beaverlodge area, Saskatchewan. Can. Inst. Min. Metall., Spec. Vol. 11, pp. 132-150. Dyck, W., Jonasson, I.R. and Liard, R.F., 1976b. Uranium prospecting with 222Rnin frozen terrain. J. Geochem. Explor., 5:115-127. Dyck, W., Pelchat, J.C. and Meilleur, G.A., 1976c. Equipment and procedures for the collection and determination of dissolved gases in natural waters. Geol. Survey Canada, Paper 75-34, 12 pp.
References
481
Earle, S.A.M. and Drever, G.L., 1983. Hydrogeochemical exploration for uranium within the Athabasca Basin, northern Saskatchewan. J. Geochem. Explor., 19: 57-73. Elinson, M.M., Pashkov, Y.N., Agababov, G.M. and Ignatiev, I.B., 1970. Research on gas aureoles around copper-molybdenum ore bodies. In: Outline of the Geochemistry of Mercury, Molybdenum and Sulphur in the Hydrothermal Process. Nauka Publ., Moscow, pp. 217-236 (in Russian). Elworthy, R.T., 1918. Mineral springs of Canada, the chemical character of some Canadian mineral springs, Part II. Ottawa, Canada, 173 pp. Enhalt, D.H., 1974. The atmospheric cycle of methane. Tellus, 26: 58-70. Eremeev, A.N., Sokolov, V.A., Solovov, A.P. and Yanitskii, I.N., 1973. Application of helium surveying to structural mapping and ore deposit forecasting. In: M.J. Jones (ed.), Geochemical Exploration 1972. lnstn. Min. Metall., London, pp. 183-192. Evans, R.D., 1969. Engineers' guide to the elementary behaviour of radon daughters. Health Physics, 17: 229-252. Evans, R.D., Kip, A.F. and Moberg, E.G., 1938. The radium and radon content of Pacific Ocean water, life and sediments. Amer. J. Sci., 36: 241-259. Everett, J.R. and Petzel, G., 1973. An evaluation of the suitability of ERTS data for the purposes of petroleum exploration. Proc. Third Earth Resources Technology Satellite Symposium. NASA, SP356, pp. 50-61. Fairbridge, R.W., 1972. Encyclopedia of Geochemistry and Environmental Sciences. Dowden, Hutchinson and Ross, Stroudsburg, Pa., 134 pp. Farwell, S.O. and Gluck, S.J., 1980. Glass surface deactivants for sulphur-containing gases. Anal. Chem., 52: 1968-1971. Farwcll, S.O., Gage, D.R. and Kagel, R.A., 1981. Current status of prominent selective gas chromatographic detectors: a critical assessment. J. Chromatog. Sci., 19: 358-376. Farweil, S.O., Gluck, S.J., Bamesberger, W.L., Schutte, T.M. and Adams, D.F., 1979. Determination of sulphur-containing gases by a deactivated cryogenic enrichment and capillary gas chromatographic system. Anal. Chem., 5 I: 609-6 i 5. Faul, H., 1954. Rare gases and fission in nature. In: Nuclear Geology, a Symposium on Nuclear Phenomena in the Earth Sciences, pp. 133-144. Favorskaya, M.A.and Tomson, I.N., 1989. About the nature of Transrockal ore-concentration structures. In: The Transrockal Ore-Concentration Structures. Nayka, Moscow, pp. 5-9 (in Russian). Feder, A.M. and Pcnfield, G.T., 1985. Remote sensing detection of microseepage in hydrocarbon exploration. Proc. Fourth Thematic Conference on Remote Sensing for Exploration Geology, San Francisco, pp. 79-91. Fei, Q. and Ruan, T.J., 1991. Surface geochemical patterns over oil and gas traps. J. China Univ. Geosciences, 2: 29-35. Felmlee, J.K. and Cadigan, R.A., 1978. Radium and uranium data for mineral springs in eight western states. US Geol. Survey, Open File Report 78-561, pp. 1-45. Fenn, l.J., 1940. A hypothesis for the origin of certain geochemical anomalies associated with petroleum accumulation. In: E.E. Rosaire (ed.), Symposium on Geochemical Prospecting for Petroleum. Amer. Assoc. Pet. Geol. Bull., 24: 1400-1433. Ferguson, J.D., 1975. Surface alteration and mineralisation of Permian red beds overlaying several oil fields in southern Oklahoma. Unpubl. MS thesis, Oklahoma State Univ., 95 pp.
482
References
Ferguson, T.D., 1979a. The subsurface alteration and mineralisation of Permina red beds overlying several oil fields in southern Oklahoma, Part I. Oklahoma City Geological Society Shale Shaker, 29:172-178. Ferguson, T.D., 1979b. The subsurface alteration and mineralisation of Permina red beds overlying several oil fields in southern Oklahoma, Part II. Oklahoma City Geological Society Shale Shaker, 29: 200-208. Fleischer, R.L. and Likes, R.S., 1979. Integrated radon monitoring by the diffusional barrier technique. Geophysics, 44:1963-1973. Fleischer, R.L. and Mogro-Campero, A., 1970. Radon enhancements in the earth: evidence for intermittent upflows? Geophys. Res. Lea., Paper 9LO399, pp. 361-364. Fieischer, R.L. and Mogro-Campero, A., 1978. Mapping of integrated radon emanation for detection of long-distance migration of gases within the Earth: techniques and principles. J. Geophys. Res., 83: 3539-3549. Fieischer, R.L. and Mogro-Campero, A., 1979. Integrated radon mapping in the earth - assessment of the Rn-220 signal and its exclusion. Geophys., 44: 1541-1548. Fleischer, R.L. and Mogro-Campero, A., 1980. Techniques and principles for mapping of the integrated radon emanation within the earth. In: T.F. Gesell and W.M. Lowder (eds.), Natural Radiation Environment III. National Technical Information Service, US Dept. of Commerce, Springfield, Va., pp. 57-71. Fleischer, R.L., Hart, H.R. and Mogro-Campero, A., 1980. Radon emanation over an ore body: search for long-distance transport of radon. Nuclear Instruments and Methods, 173:169-181. Fleischer, R.L., Price, P.B. and Walker, R.M., 1975. Nuclear tracks in solids: principles and applications. Univ. California Press, 605 pp. Forbes, E.A., Posner, A.M. and Quirk, J.P., 1974. The specific adsorption of inorganic Hg(ll) species and Co(Ill) complex ions on goethite. J. Colloid Interface Sci., 49: 403-409. Forrest, J. and Newman, L., 1973. Ambient air monitoring for sulphur compounds: a critical review. J. Air Pollution Control Assoc., 23: 761-768. Foster, N.H., 1979. Geomorphic exploration utilised in the discovery of Trap Springs Field, Nye County, Nevada. In: Rocky Mountain Assoc. Geol., Basin and Range Symposium 1979, pp. 477486. Frank, D.S., Sackett, W.M., Hall, R. and Fredericks, A.D., 1970. Methane, ethane, and propane concentrations in the Gulf of Mexico. Amer. Assoc. Pet. Geol. Bull., 54: 1933-1938. Freeze, R.A. and Cherry, J.A., 1979. Groundwater. Prentice-Hail, Englewood Cliffs, 604 pp. Fridman, A.I., 1970. The Natural Gases of Ore Deposits. Nedra, Moscow, 192 pp. (in Russian). Fridman, A.I. and Pctrov, V.A., 1976. Principal results of gas surveying in ore deposits under permafrost conditions (Noril'sk region). Intemat. Geol. Rev., 18:545-550 Friedlander, G., Kennedy, J.W. and Miller, J.M., 1964. Nuclear and Radiochemistry. John Wiley and Sons Inc., New York, 585 pp. Friedman, I. and Denton, E.H., 1975. A portable helium sniffer. US Geol. Survey, Open File Report, 75-532, 6 pp. Friedrich, G.H. and Hawkes, H.E., 1966. Mercury dispersion haloes as ore guides for massive sulphide deposits, West Shasta District, California. Mineral. Deposita, 2: 77-88. Friedrich, G.H., Herzig, P., Keyssner, S. and Maliotis, G., 1984. The distribution of Hg, Ba, Cu and Zn in the vicinity of cupriferous sulphide deposits, Troodos Complex, Cyprus. J. Geochem. Explor., 21: 167-174.
References
483
Froidevaux, C.M., 1980. Radar, an optimum remote sensing tool for detailed plate tectonic analysis and its application to hydrocarbon exploration (an example in Irian Jaya, Indonesia). In: Radar Geology: an Assessment (Publ. 80-61). JPL, Pasadena, pp. 457-501. Fursov, V.Z., 1958. Halos of dispersed mercury as prospecting guides at the Achisai lead-zinc deposits. Geochemistry, 3: 338-344. Fursov, V.Z., Vol'fson and Khvalorski, A.G. 1968. results of a study of mercury vapour in the Tashkent earthquake zone. Dokl. Akad. Nauk. USSR, 179:1213-1215. Gabelman, J.W., 1972. Radon emanometry of Starks salt dome, Calcasiew Parish, Louisiana. US Atomic Energy Commission, Report RME-4114 UC-51, 75 pp. Gabelman, J.W., 1974. Frontiers of uranium exploration. Paper presented at the 103rd Annual Meeting of the AIME Society of Mining Engineers, Dallas, 51 pp. Galimov, E.M., 1968. Isotopic composition of carbon in gases of the crust. Int. Geol. Rev., 11 : 10921104.
Garrels, R.M, and Christ, C.L., 1965. Solutions, Minerals and Equilibria. Harper and Row, New York, 450 pp. Gascoyne, M. and Wuschke, D.M., 1990. Fracture detection and groundwater flow characterisation in poorly-exposed ground using helium and radon on soil gases. United Kingdom Dept. Energy, Report no. DOE/RW/90/079. Gaucher, E., 1976. Alphameters: uranium prospecting leakage from radioactive minerals. Amer. Mineral., 40:481-496. Gerling, E.K., 1939. Part taken in the close packing of crystals in the diffusion of helium. Compt. Rend. Acad. Sci. USSR., 24: 274-277. Gerling, E.K., 1957. The migration of helium from minerals and rocks. Trudy Radievogo Inst. Im. V.G. Khiopina 6:64-87 (AEC-tr-4208). Gerling, E.K., Mamyrin, B.A., Tolstykhin, I.N. and Yakovleva, S.S., 1971. Helium isotope composition in some rocks. Geokhimiya, I0:1205-1217. Germanov, A.I., 1963. Uranium in natural waters. In: A.P. Vinogradov (ed.), Fundamental Features of the Geochemistry of Uranium, pp. 290-336 (in Russian). Germanov, A.I., Batulin, S.G., Volkov, G.A., Lisitsin, A.K. and Serebrennikov, V.S., 1958. Some regularities of uranium distribution in underground waters. Proc. 2nd UN Internat. Conference on Peaceful Uses of Atomic Energy, Geneva: 2, Survey of Raw Material Resources, pp. 161-177. Gesell, T.F. and Lowder W.M. (eds.), 1980. Natural Radiation Environment Ill. National Technical Information Service, US Dept. of Commerce, Springfield, Va., 1736 pp. Giletti, B. and Kulp, J.L., 1955. Radon leakage from radioactive minerals. Amer. Mineral., 40: 481496. Gingrich, J.E., 1974. Results from a new uranium exploration method. Terradex Corp., Ca., 4 pp. Gingrich, J.E. and Fisher, J.C., 1976. Uranium exploration using the Track-Etch method. In: Exploration for Uranium Ore Deposits. IAEA, Vienna, pp. 213-227. Glasstone, S. and Lewis, D., 1960. Elements of Physical Chemistry (2nd edn.). Van Nostrand, Princeton, 758 pp. Giebovskaya, U.S. and Glebovskii, S.S., 1960. The possibility of the application of gas surveys in prospecting for sulphide ore deposits. Vopr. Rudnoy Geofiziki, 1:48-56 (in Russian). Gieezen, C.B., 1985. New Criteria for distinguishing active microseeps from background (abs.). Presented at 190th ACS National Meeting, Chicago, IL, September 8-13. Gluechauf, E., 1946. A micro-analysis of the helium and neon contents of air. Proc. Roy. Sot., 185: 98-119.
484
References
Go, Z. and Wang X., 1994. Study on abiogenic gases in Songliao Basin. Sciences of China (B), 24(3): 34-39 (in Chinese). Goetz, A.F.H. and Rowan, L.C., 1981. Geologic remote sensing. Science, 211: 781-791. Gole, M.J. and Butt, C.R.M., 1985. Biogenic-thermogenic near-surface gas anomaly over Gingin and Bootine gas fields, Western Australia. Bull. Amer.Assoc. Pet. Geol., 69:2110-2119. Gole, M.J., Butt, C.R.M. and Snelling, A.A., 1986. A groundwater helium survey of the Koongarra uranium deposits, Pine Creek Geosyncline, Northern Territory. Uranium, 2: 343-360. Goleva, G.A., 1977. Hydrogeochemistry of the Ore Elements. Nedra, Moscow, 216 pp. (in Russian). Golubev, V.S., Eremeev, A.N. and Yanitskii, I.N., 1974. Analyses of several models of migration of He in the lithosphere. Geokhimiya, 7: 1067-1076 (in Russian). Goodfellow, W.D., Jonasson, I.R. and Morganti, J.M., 1983. Zonation of chalcophile elements about the Howards Pass (XY) Zn-Pb deposit, Selwyn Basin, Yukon. J. Geochem. Explor., 19: 503-542. Goriainov, I.N., Gramberg, I.S., Smekalova, A.S, and Tikhomirov, Yu.V., 1989. To question about the distribution occurrence laws of the deep-water sulphides in the world ocean (hypothesis of the mantle spicules). Reports of Academy of Sciences of the USSR, 306:1185-1188 (in Russian). Gottschalk, V.H. and Buehler, H.A., 1912. Oxidation of sulphides. Econ Geol., 7:15-34. Goumay, L.S., Harreli, J.W. and Dennis, C.L., 1979. Remote sensing of hydrocarbon gas seeps utilising microwave energy. US Patent No. 4, 132, 943. Govett, G.J.S., 1973. Differential secondary dispersion in transported soils and post-mineralisation rocks: an electrochemical interpretation. In: M.J. Jones (ed.), Geochemical Exploration 1972. Insm. Min. Metall., London, pp. 81-91. Govett, G.J.S., 1976. Detection of deeply-buried and blind sulphide deposits by measurement of H§ and conductivity of closely-spaced surface soil samples. J. Gcochem. Explor., 6: 359-382. Govett, G.J.S. and Atherden, P.R., 1987. Electrochemical patterns in surface soils- detection of blind mineralisation beneath exotic cover, Thalanga, Queensland, Australia. J. Geochem. Explor., 28: 201-221. Govctt, G.J.S. and Chork, C.Y., 1977. Detection of deeply buried sulphide deposits by measurement of organic carbon, hydrogen ion and conductance in surface soils. In: Prospecting in Areas of Glaciated Tcrrain 1977. Insm. Min. Metall., London, pp. 49-55. Govett, G.J.S., Dunlop, A.C. and Atherden. P.R., 1984. Electrical techniques in deeply weathered terrain in Australia. J. Geochem. Explor. 21:311-33 I. Govett, G.J.S., Goodfellow, W.D. and Whitehead, R.F.S., 1976. Experimental aqueous dispersion of elements around sulphides. Econ. Geol., 71:925-940. Grammakov, A.G., 1934. Emanation (radon) method of search, investigation and reconnaissance of radioactive objects. Tr. TSNI GRI 7 (in Russian). Granger, H.C. and Warren, G.C., 1969. Unstable sulphur compounds and the origin of roll-type uranium deposits. Econ. Geol., 64: 160-171. Grant, W.B. and Menzies, R.T., 1983. A survey of laser and selected optical systems for remote measurement of pollutant gas concentrations. J. Air Pollution Control Assoc., 33:187-194. Greenland, D.J., 1981. The Chemistry of Soil Processes. University Press, Belfast, 222 pp. Greenwood, D.J., 1961. The effect of oxygen concentration on the decomposition of organic materials in soils. Plant Soil, 14: 360-376. Gregory, R.G. and Durrance, E.M., 1985. Helium, carbon dioxide and oxygen soil gases: small scale variations over fractured ground. J. Geochem. Explor., 24: 29-49. Gregory, R.G. and Durrance, E.M., 1987. Helium, radon and hydrothermal circulation associated with the Carnmeneilis radiothermal granite of southwest England. J. Geophys. Res., 92: 12,567-12,586.
References
485
Gutsalo, C.K., 1976. On the sources and distribution of helium and argon isotopes in the thermal waters of the Kurile islands and Kamchatka. Geochem. Intemat., 13:167-175. Haissinsky, M., 1964. Nuclear Chemistry and its Applications. Addison-Welsley, Paris, pp. 623-627. Hale, M. and Moon, C.J., 1982. Geochemical expressions at surface of mineralisation concealed beneath glacial till at Keel, Eire. In: P.H. Davenport (ed.), Prospecting in Areas of Glaciated Terrain 1982. Can. Inst. Min. Metall., pp. 228-236. Hamdi, Y.A., 1971. Soil water tension and the movement of rhyzobia. Soil Biol. Biochem., 3: 121. Hamilton, S.M., 1998. Electrochemical mass-transport in overburden: a new model to account for the formation of selective-leach geochemical anomalies in glacial terrain. J. Geochem. Explor., 63: 155-172. Hamilton, S.M. and McClenaghan, M.B., 1998. Field data in support of electrochemical transport of elements through thick glacial overburden. Ontario Geol. Survey, Misc. Paper 169: 267-274. Hamilton, S.M., Bajc, A.F., Farrow, C.E.G. and McClenaghan, M.B., 1995. Multi-media geochemical sampling in areas of thick overburden: a case study. Ontario Geol. Survey, Misc. Paper 164: 258262. Hamilton, S.M., Michel, F.A. and Jefferson, C.W., 1990. CO2-rich waters of the Flat River Valley, N.W.T. In: T.D. Prowse and C.S.L. Ommanney (eds.), Northern Hydrology, Selected Perspectives. National Hydrology Research Institute, Saskatoon, Saskatchewan, 532 pp. Hansink, J.D., 1976. Equilibrium analysis of sandstone roll-front uranium deposits. In: Exploration for Uranium Ore Deposits. IAEA, Vienna, pp. 683-693. Harper, C.T., 1968. Track etch, emanometer, spectrometer and soil geochemistry surveys across the Carswell Structure, northwestern Saskatchewan. Saskatchewan Geol. Survey, Spec Publ., 1-24. Hart, M.K.W., Golc, M.J. and Butt, C.R.M., 1985. Possible errors in helium analyses of soil gases by mass spectrometers having constant-pressure inlet systems. J. Geochem. Explor., 24: 121-128. Hatch, W.R. and Ott, W.L., 1968. Determination of sub-microgram quantities of mercury by atomic absorption spectrometry. Anal. Chem., 40: 2085-2086. Hatuda, Z., 1954. Radioactive method for geological exploration. Memoirs College Sci., Univ. Kyoto, Ser. B, 21, Art. 9:231-271. Hauksson, E., 1981. Radon content of groundwater as an earthquake precursor: evaluation of worldwide data and physical basis. J. Geophys. Res., 86:9397-9410. Hawkes, H.E. and Webb, J.S., 1962. Geochemistry in Mineral Exploration. Harper and Row, New York, 415 pp. Hawkes, H.E. and Wiiliston, S.H., 1962. Mercury vapour as a guide to lead-zinc-silver deposits. Min. Cong. J., 48: 30-32. Heaton, T.H.E., 1981. Dissolved gases: some applications to groundwater research. Trans. Geol. Soc. S.A., 84: 91-97. Heaton, T.H.E., 1984. Rates and sources of 4He accumulation in groundwater. Hydrol. Sci. J., 29: 2947. Heaton, T.H.E. and Vogel, J.C., 1979. Gas concentrations and ages of groundwaters in Beaufort Group Sediments, South Africa. Water S.A., 5: 160-170. Heaton, T.H.E. and Vogel., J.C., 1980. "Excess air" in groundwater. J. Hydrol., 50:201-216. Heemstra, R.J., Ray, R.M., Wesson, T.C., Abrams, J.R. and Moore, G.A., 1979. A critical laboratory and field evaluation of selected surface prospecting techniques for locating oil and natural gas. Dept. Energy, Bartlesviile Energy Technology Center, BETC/RI-78/I 8, 84 pp. Hegge, M.R., 1977. Geologic setting and relevant exploration features of the Jabiluka uranium deposits. Proc. Austr. Inst. Min. Metali., 264: 19-32.
486
References
Henderson, P., 1986. Inorganic Geochemistry. Pergamon, Oxford, 353 pp. Heyrovsky, Ya. and Kuta, Ya., 1965. The basis of polarography, lnostrannaja Literatura, Moscow, 559 pp. (in Russian). Hickey, J.C., 1983. Fault leakage characterisation by integrative gas geochemistry/mass spectrometry/pattern recognition procedures. Amer. Assoc. Pet. Geol. Bull., 67: 1342. Hickey, J.C., Klusman, R.W. and Voorhees, K.J., 1983. An integrative gas geochemical technique for surficial petroleum exploration. Amer. Assoc. Pet. Geol. Bull., 67: 483. Higgins, F.B., Jr., Grune, W.N., Smith, B.M. and Terrill, J.G., Jr., 1961. Methods for determining radon-222 and radium-226. J. Amer. Water Works Assoc., 53: 63-64. Hinkle, M.E., 1978. Helium, mercury, sulphur compounds and carbon dioxide in soil gases of the Puhimau thermal area, Hawaii Volcanoes National Park, Hawaii. US Geol. Survey, Open File Report, 78-246, 14 pp. Hinkle, M.E., 1986. Using volatile constituents of soils and soil gases to determine the presence of copper-zinc ore bodies at Johnson Camp, Arizona. J. Geophys. Res., 91: 12,359-12,365. Hinkle, M.E., 1990. Factors affecting concentrations of helium and carbon dioxide in soil gases. In: Geochemistry of Gaseous Elements and Compounds. Theophrastus Publ., pp. 421-448. Hinkle, M.E., 1994. Environmental conditions affecting concentrations of He, CO2, 02 and N2 in soil gases. App. Geochem., 9: 53-63. Hinkle, M.E. and Ayres, K.P., 1985. Gas chromatographic procedure for analysis of drill core samples for carbon dioxide and volatile sulphur species. US Geol. Survey, Open File Report, 85-245, 16 PP. Hinkle, M.E. and Dilbert, C.A., 1984. Gases and trace elements in soils at the North Silver Bell Deposit, Pima County, Arizona. J. Geochem. Explor., 20: 323-336. Hinkle, M.E. and Harms, T.F., 1978. CS2 and COS in soil gases of the Roosevelt Hot Springs Known Geothermal Resource Area, Beaver County, Utah. US Geol. Survey J. Res., 6: 571-578. Hinkle, M.E. and Kantor, J.A., 1978. Collection and analysis of soil gases emanating from buried sulphide mineralisation, Johnson Camp area, Cochise County, Arizona. ,i. Geochem. Explor., 9: 209-216. Hinkle, M.E. and Kilburn, .I.E., 1979. The use of vacutainer tubes for collection of soil samples for helium analysis. US Geol. Survey, Open File Report, 79-1441,23 pp. Hinkle, M.E. and Ryder, J.L., 1987. The effect of moisture and CO2 on concentrations of helium in soils and soil gases. J. Geophys. Res. 92: 12,587-12,594. Hinkle, M.E., Denton, E.H., Bigelow, R.C. and Turner, R.L., 1978. Helium in soil gases of the Roosevelt Hot Springs Known Geothermal Resource Area, Beaver County, Utah. US Geol. Survey ,i. Res., 6: 563-570. Hitarov, N.I., Kravtsov, A.N., Voitov, G.I., Fridman, A.I., Ortenberg, N.A. and Pavlov, A.S., 1979. Gases of the free extrusions of the Hibin Massif. Soviet Geology, N2:62-73 (in Russian). Holland, P.W. and Emerson, D.E., 1979. Helium in groundwater and soil gas in the vicinity of Bush Dome Reservoir, Cliffside Field, Potter County, Texas. US Bureau Mines Inf. Circ. No. 8807, 22 PP. Holland, P.W. and Emerson, D.E., 1987. A determination of the helium-4 content of near-surface atmospheric air within the continental United States. J. Geophys. Res. 92: 12,557-12,566. Holland, P.W. and Emerson, D.E., 1990. The global helium-4 content of near-surface atmospheric air. In: Geochemistry of Gaseous Elements and Compounds. Theophrastus Publications, pp. 97-109. Hoppe, R. and Alexander, E.C., 1978. Helium and the other noble gases. In: K.H. Wedepohi (ed.), Handbook of Geochemistry. Springer Verlag, Berlin. Horvitz, L., 1939. On geochemical prospecting. Geophysics, 4: 210-228.
References
487
Horvitz, L., 1945. Recent developments in geochemical prospecting for petroleum. Geophysics, 10: 487-493. Horvitz, L., 1950. Chemical methods. In: J.J. Jakosky (ed.), Exploration Geophysics (2nd edn.). Trija Publishing, Los Angeles, pp. 938-965. Horvitz, L., 1954. Near-surface hydrocarbons and petroleum accumulation at depth. Mining Eng., December, pp. 1205-1209. Horvitz, L., 1957. How geochemical analysis helps the geologist find oil. Oil and Gas J., 45: 234-242. Horvitz, L., 1965. Case histories show results of drilling geochemical prospects. World Oil, November, pp. 161-165. Horvitz, L., 1969. Hydrocarbon geochemical prospecting after thirty years. In: W.B. Heroy (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas. Southern Methodist University Press, Dallas, pp. 205-218. Horvitz, L., 1972. Vegetation and geochemical prospecting for petroleum. Amer. Assoc. Pet. Geol. Bull., 56: 925-940. Horvitz, L., 1980. Near-surface evidence of hydrocarbon movement from depth. In: W.H. Roberts and R.J. Cordell (eds.), Problems of Petroleum Migration. Amer. Assoc. Pet. Geol. Studies in Geology, 10:241-263. Horvitz, L., 1981. Hydrocarbon geochemical exploration after forty years. In: B.M. Gottlieb (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas II. Southern Methodist University Press, Dallas, pp. 83-95. Horvitz, L., 1985. Near-surface hydrocarbons and non-hydrocarbon gases in petroleum exploration. Presentation at Amer. Assoc. Pet. Geol., Rocky Mountain Section, Denver. Horvitz, L., 1986. Hydrocarbon geochemical exploration after fifty years. In: M.J. Davidson (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas IV. Southern Methodist University Press, Dallas, Texas, pp. 147-162. Hunt, G.R., 1977. Spectral signatures of particulate minerals in the visible and near-infrared. Geophysics, 42:501-513. Hunt, G.R., 1979. Near-infrared (1.3-2.4 ~tm) spectra of alteration minerals -potential for use in remote sensing. Geophysics, 44:1974-1986. Hunt, G.R. and Salisbury, J.W., 1970. Visible and near-infrared spectra of minerals and rocks: I, silicate minerals. Modem Geol., 1: 283-300. Hunt, G.R. and Salisbury, J.W., 1971. Visible and near-infrared spectra of minerals and rocks: II, carbonates. Modem Geol., 2: 23-30. Hunt, G.R., Salisbury, J.W. and Lenhoff, C.J., 1973. Visible and near-infrared spectra of minerals and rocks: VI, additional silicates. Modem Geol., 4: 85-106. Hunt, J.M., 1974. Hydrocarbon and kerogen studies. Initial Reports of Deep Sea Drilling Project, 22: 673-675. Hunt, J.M., 1975. Origin of gasoline range alkanes in the deep sea. Nature, 254:41 I-413. Hunt, J.M., 1979. Petroleum Geochemistry and Geology. W.H. Freeman, San Francisco, 617 pp. Hunt, J.M., 1981. Surface gcochemical prospecting, pro and con. Amer. Assoc. Pet. Geol. Bull., 65: 939. Hunt, J.M. and Whelan, J.K., 1979. Volatile organic compounds in Quaternary sediments. Organic Geochem., 1:219-224. Hunter, D.R., 1969. Thermal waters of Swaziland. In: G. Kacura (ed.), Mineral and thermal waters of the World, B - Overseas Countries. Proc. XXIII Int. Geol. Cong., 19:165-170.
488
References
IAEA, 1974. Recommended instrumentation for uranium and thorium exploration. Technical Report Series, No. 158. International Atomic Energy Agency, Vienna, 93 pp. IAEA, 1982. Environmental migration of low-lived radio-nuclides. International Atomic Energy Agency, Vienna, 834 pp. Israel, H. and Bjornasson, S., 1966. Radon (Rn 222) and thoron (Rn 22~ in soil air over faults. Forschung. fur Geophys. 33: 49-64. Israel, H. and Horbert, M., 1970. Tracing atmospheric eddy mass transfer by means of natural radioactivity. J. Geophys. Res., 75:2291-2297. lvanov, V.U., Medovii, V.I. and Dobrolovskaya, V.I., 1979. Field of helium concentrations in sedimentary sequences, lnternat. Geol. Rev., 21:891-903. Jackson, R.G., 1995. The application of water and soil geochemistry to detect blind mineralisation in areas of thick overburden. Ontario Geol. Survey, Open File Report 5927, 151 pp. James, C.H., 1962. A review of the geochemistry of mercury and its application to geochemical prospecting. Tech. Comm. No. 41, Geochemical Prospecting Research Centre, Imp. Coll. Sci. Tech., London, 42 pp. James, C.H., 1964. The potential role of mercury in modem geochemical prospecting. Min. Mag., July, pp. 23-32. Janezic, G.G., 1979. Biogenic light hydrocarbon production related to near-surface geochemical prospecting for petroleum (abs.). Amer. Assoc. Pet. Geol. Bull., 63: 403. Jeter, H.W., 1980. A modelling study of gaseous Rn-222, Xe-133 and He-4 for uranium exploration. US Dept. of Energy, Open File Report, GJBX-140(80), 117 pp. Jctcr, H.W., Martin, J. and Schutz, D.F., 1977. The migration of gaseous radionuclides through soil overlying a uranium deposit: a modelling study. US Dept. of Energy, Open File Report, GJBX-67 (77), 61 pp. Jonasson, I.R., 1970. Mercury in the natural environment: a review of recent work. Geol. Survey Canada, Paper 70-57. Jonasson, I.R. and Dyck, W., 1978. Dissolved gases in snow-melt waters. Chem. Geol., 21:15-24. Jones, V.T., 1983. Surface monitoring of retort gases from an underground coal gasification reactor: time dynamics (abs.). Proc. Annual Meeting Amer. Chem. Soc., Washington, D.C. Jones, V.T., 1984. Overview of hydrocarbon extractions in soils and waters: research needs and problems. Presented at American Petroleum Institute Workshop for Sampling and Analytical Methods for Determining Petroleum Hydrocarbons in Groundwater and Soil, Denver. Jones, V.T. and Drozd, R.J., 1979. Predictions of oil or gas potential by new surface geochemistry. Amer. Assoc. Pet. Geol. Bull., 63: 699. Jones, V.T. and Drozd, R.J., 1983. Prediction of oil or gas potential by near-surface geochemistry, Amer. Assoc. Pet. Geol. Bull., 67: 932-952. Jones, V.T. and Thune, H.W., 1982. Surface detection of retort gases from an underground coal gasification reactor in steeply dipping beds near Rawlins, Wyoming. Soc. Petroleum Engineers, SPE 11050, 24 pp. Jones, V.T., Burteli, S.G., Hodgson, R.A., Whelan, T., Milan, C. Ando, T., Okada, K., Agtsuma, T. and Takono, O., 1985. Remote sensing and surface geochemical study of Railroad Valley, Nye County, Nevada. Presented at Fourth Thematic Mapper Conference, Remote Sensing for Exploration Geology, San Francisco, California, April 1-4. Kahler, D.E., undated. Helium, a guide to geothermal energy resources. Western Systems Inc., Evergreen, Colorado, Paper WSI-1811,2 pp.
References
489
Kahma, A., 1965. Trained dog as tracer of sulphide-bearing glacial boulders. Sedimentol., 5, Atlas 1, no. 4. Kahma, A., Nurmi, A. and Mattsson, P., 1975. On the composition of the gases generated by sulphidebearing boulders during weathering and on the ability of prospecting dogs to detect samples treated with these gases in the terrain. Geol. Survey Finland, Report No. 6, 4 pp. Kalinko, M.K., 1987. Geology and Geochemistry of Naphthides. Nedra, Moscow, 242 pp. (in Russian). Kamenskiy, L.L., Yakutseni, V.P., Mamyrin, B.A., Anufriyev, S.G. and Tolstykhin, I.N., 1971. Helium isotopes in nature. Geokhimiya, 8: 914-931. Kapitanov, Yv.T., Pavlov, I.V., Semikin, N.P. and Serclyukova, A.S., 1970. Adsorption of radon on activated carbon. Intemat. Geol. Rev., 12: 873-878. Karar, V.P.and Sakovich, B.B., 1989. Application of the CHIM method for prospecting placer gold deposits at the Enisey mountain-ridge. Application of Geoelectrochemical methods for Prospecting and Exploration of Ore Deposits. NPO Rudgeofizika, Leningrad, pp. 189-112 (in Russian). Karim, M.F., 1964. Some geochemical methods of prospecting and exploration for oil and gas. Unpubi. Ph.D. thesis, Univ. Southern California, 227 pp. Kartsev, A.A., Tabarsaranskii, Z.A., Subbota, M.I. and Mogilevskii, G.A., 1959. Geochemical methods of prospecting and exploration for petroleum and natural gas. University of California Press, Berkely, 349 pp. Katz, D.L. and Williams, B., 1952. Reservoir fluids and their behaviour. Amer. Assoc. Pet. Geol. Bull., 36: 342-358. Keller, G.V. and Frischknecht, F.C., 1966. Electrical Methods in Geophysical Prospecting. Pergamon Press, Toronto, 519 pp. Kent, L.E., 1969. The thermal waters in the Republic of South Africa. In: G. Kacura (ed.), Mineral and Thermal Waters of the World, .B- Overseas Countries. Proc. XXIII Int. Geol. Cong., 19: 143-164. Khayretdinov, I.A., Dokokin, G.P., Prokhorov, J.G. and Sverlova, V.N., 1965. The use of gas sampling for surveys in fault dislocations of East Sayan. Geology and Geophysics, Nauka Publ. Moscow, 10:135-137 (in Russian). Khodakovsky, I.L., Popova, M.Ya. and Ozerova, N.A., 1975. On the role of sulphide complexes in the transport of mercury by hydrothermal solutions. Geochim. Int., 12:207-2 ! 5. Khristianov, V.K. and Korchuganov, B.N., 1971. On the radon content of the waters of the upper Volga. Geokhimiya, 4: 492-496. Kim, A.G. and Douglas, L.J., 1972. Hydrocarbon gases produced in a simulated swamp environment. US Dept. of Interior, Bureau of Mines, Report of Investigations, 7690. King, C.Y., 1978. Radon emanation on San Andreas fault. Nature, 271: 516-519. King, C.Y., 1980a. Episodic radon changes in subsurface soil gas along active faults and possible relation to earthquakes. J. Geophys. Res., 85: 3065-3078. King, C.Y., 1980b. Geochemical measurements pertinent to earthquake prediction. J. Geophys. Res., 85:3051. King, C.Y., 1986. Gas geochemistry applied to earthquake prediction: an overview. J. Geophys. Res., 91: 12,269-12,281. King, J., Tauchid, M., Frey, O., Basset, M., Cetinturk, I., Aydinoz, F. and Keceli, B., 1976. Exploration for uranium in southwestern Anatolia: a case history. IAEA/NEA International Symposium on Exploration of Uranium Ore Deposits, Mexico, 25 March - 2 April. King, P.T., Michel, J. and Moore, W.S., 1982. Groundwater geochemistry of 228Ra, 226Ra and 222Rn. Geochim. Cosmochim. Acta, 46:1173-1182.
490
References
Kiriukhin, A.V., Korotkov, A.I. and Pavlov, A.H.,. General Hydrogeology. Nedra, Leningrad, 358 pp. (in Russian). Kleinevoss, A., 1971. Zur geochemischen Charakteristik des Quecksilbers unter besonderer Ber~icksichtigung der Hg-Verteilung in den Erzlagem des Rammelsberges und ihrer Umgebung. Unpubl. Ph.D. thesis, Univ. Clausthal, 191 pp. Klochkov, V.G., Bugorkov, M.S. and Delivron, V.A., 1989. Application of polarographic logging in ore regions. Application of Geoelectrochemical Methods for Prospecting and Exploration of Ore Deposits. NPO Rudgeofyzika, Leningrad, pp. 61-63 (in Russian). Klusman, R.W. and Jaacks, J.A., 1987. Environmental influences upon mercury, radon and helium concentrations in soil gases at a site near Denver, Colorado. J. Geochem. Expior. 27: 259-280. Klusman, R.W. and Voorhees, K.J., 1983. A new development in petroleum exploration technology. Col. Sch. Mines Mag., 73: 6-10. Klusman, R.W. and Webster, J.D., 1981. Meteorological noise in crustal gas emission and relevance to geochemical exploration. J. Geochem. Explor., 15: 63-76. Klusman, R.W., Saeed, M.A. and Abu-Ali, M, 1992. The potential use of biogeochemistry in the detection of petroleum microseepage. Amer.Assoc. Pet. Geol. Bull., 76:851-863. Klusman, R.W., Voorhees, K.J., Hickey, J.C. and Malley, M. J., 1986. Application of the K-V fingerprint technique for petroleum exploration. In: M.J. Davidson (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas IV. Southern Methodist University Press, Dallas, pp. 219-243. Kononov, V.I., Mamyrin, B.A., Polyak, B.G., Khabarin, L.V., 1974. Heliumcisotopes in the thermal waters of Iceland. Dokl. Akad. Nauk SSSR, 217:41-43. Komer, L.A. and Rose, A.W., 1977. Radon in streams and groundwaters of Pennsylvania as a guide to uranium deposits. Mineral Conservation Section, Report GJO-1659-2. Penn. State Univ., 152 pp. Korostin, V.P., i 976. Dependence of the limiting current of electrochemical reactions on velocity of its increasing for objects with different surfaces. In: Principles and Techniques of Exploration, Rotaprint VITR, Leningrad, pp. 15-17 (in Russian). Kovach, E.M., 1945. Meteorological influences upon the radon content of soil gas. Trans. Amer. Geophys. Union, 26:241-248. Kramer, M.W., Schroeder, G.L. and Evans, R.D., 1964. Measurements of the effects of atmospheric variables on radon-222 flux and soil-gas concentrations. In: J.A.S. Adams and W.M. Lowder (eds.), The Natural Radiation Environment. Univ. Chicago Press, pp. 191-215. Krat, V.N., 1983. Modem hydrothermal activity occurrence in the Sasyk-kul hollow (Pamir). Abstracts of the IV All-union symposium. Naukova Dumka, Kiev, pp. 137-138 (in Russian). Krauskopf, K.B., 1964. The possible role of volatile metal compounds in ore genesis. Econ Geol., 59: 22-45. Krauskopf, K.B., 1979. Introduction to Geochemistry (2 "d edn.). McGraw-Hill, New York, 617 pp. Kravtsov, A.I. and Fridman, A.I., 1965. Natural gases of ore deposits. Doki. Akad. Nauk. SSSR, 165: 192-193. Kravtsov, A.I., Voytov, G.I., Bobrov, V.A., Akimov, A.P., Ivanov, A.N. and Serdyukov, L.I., 1979. Composition and isotopes of gases in the Mir kimberlite pipe. Dokl. Akad. Nauk SSSR, 245: 214216. Kravtsov, A.I., Voytov, G.I., lvanov, V.A. and Kroptova, O.J., 1976. Gases and bitumens in rocks of the Vdachnaya pipe. Dokl. Akad. Nauk SSSR, 228:231-234. Krchmar, B., 1988. The field of infiltered-form elements in the atmospheric air and its relation to geological structures. Praha. pp. 191-205.
References
491
Kristiansson, K. and Malmqvist, L., 1981. A new model mechanism for the transportation of radon through the ground (written commun.), 18 pp. Kristiansson, K. and Malmqvist, L., 1982. Evidence for non-diffusive transport of 222Rn in the ground and a new geophysical model for the transport. Geophysics, 47: 1444-1452. Kristiansson, K. and Malmqvist, L., 1986. Trace elements in the geogas and their relation to bedrock composition. Geoexploration, 24:517-534. Kromer, E., Friedrich, G. and Walker, P., 1981. Mercury and mercury compounds in surface air, soil gas, soils and rocks. J. Geochem. Explor., 15:51-61. Kugler, G. and Clarke, W.B., 1972. Mass spectrometric measurements of 3H, 3He and 4He produced in thermal-neutron ternary fission of 235U; evidence for short range 4He. Phys. Rev., Ser. C., 5:551560. Kulikova, N.N., 1960. On the use of a gas-core survey for detecting blind ore zones in the region of the Balei deposits, lzvestiya of Higher Educational Institutions, Geology and Exploration, Moscow, 12: 78-82. Kuznetsov, Y.U.V., Simoniak, S.N., Lisitzyn, A.P. and Frenklikh, M.S., 1968. Uranium and radium in the surface layer of ocean sediments. Geokhimiya, 3: 323-333. Lagunova, I.A., 1974. On the origin of carbon dioxide in the gases of mud volcanoes of the Kerch'Taman' region. Geochem. Intemat., 11: 1209-1214. Lang, A.H., 1956. Prospecting in Canada (3 rd edn.). Econ. Geol. Series, No. 7, Geol. Survey Canada, 401 pp. Lang, A.H., Griffith, J.W. and Steacy, H.R., 1962. Canadian Deposits of Uranium and Thorium. Econ. Geol. Ser., 16, Geol. Survey Canada., 324 pp. Lang, H.R., Aldeman, W.H. and Sabins, F.F., 1985a. Patrick Draw, Wyoming, petroleum test case report. In: M.J. Abrams, J.E., Conei, H.R. Lang and H.N. Paley (eds.), Joint NASA/Geosat Test Case Project: Final Report. Amer.Assoc. Pet. Geol., Spec. Publ., part 2, vol. 2, pp. 11.1-11.28. Lang, H.R., Curtis, J.B. and Kovacs, J.C., 1985b. Lost River, West Virginia, petroleum test site report In: M.J. Abrams, J.E., Conel, H.R. Lang and H.N. Paley (eds.), Joint NASA/Geosat Test Case Project: Final Report. Amer.Assoc. Pet. Geol., Spec. Publ., part 2, vol. 2, pp. 12.1-12.96. Larin, V.N., 1980. Hypothesis of the originally-hydride Earth. Nedra, Moscow, 216 pp. (in Russian). Laubmeyer, G., 1933. New geophysical prospecting method, especially for hydrocarbon strata. Petroleum, 29: 1-4. Le Radium, Journal, 1904. La radioactivate de I'atmosphere et du sol. Publication mensuelle. Lester, O.C., 1918. The radioactive properties of the mineral springs of Colorado. Amer. J. Sci. (4th series), 46: 623. Levinson, A.A. and Bland, C.J., 1978. Examples of the variability of disequilibrium and the emanation factor in some uraniferous materials. Can. J. Earth Sci., 15: 1867-1871. Levinson, A.A., Bland, C.J. and Parslow, G.R., 1978. Possible pitfalls in the research for uranium deposits using lake sediments and lake waters; Can. Inst. Min. Metall. Bull., 71: 59-62. Levinson, A.A. and Coetzee, G.L., 1978. Implications of disequilibrium in exploration for uranium ores in the surficial environment using radiometric techniques- a review. Min. Sci. Eng., 10: 1927. Levorsen, A.I., 1967. Geology of Petroleum. W.H. Freeman, San Francisco. Leythaeuser, D., Schaefer, R.G. and Yukler, A., 1980. Diffusion of light hydrocarbons through nearsurface rocks. Nature, 284: 522-525.
492
References
Lilbum, R.A. and Alshaieb, Z., 1984. Geochemistry and isotopic composition of hydrocarbon-induced diagenetic aureole (HIDA), Cement, Oklahoma. Oklahoma City Geological Society Shale Shaker, 34: 57-67. Link, W.K., 1952. Significance of oil and gas seeps in world oil exploration. Amer. Assoc. Pet. Geol. Bull., 36: 1505-1541. Lomonosov, I.S., Mamyrin, B.A., Prasolov, E.M. and Tolstykhin, I.N., 1976. Helium and argon isotopic composition in some thermal waters of the Baykal Ritt Zone. Geochem. Int., 13: 78-81. Loudon, A.G., Lee, M.K., Dowling, J.F. and Bourne, R., 1975. Lady Loretta silver-lead-zinc deposit. In: C.L. Knight (ed.), Economic Geology of Australia and Papua New Guinea, Metals, Vol. 1. Australas. Inst. Min. Metall., Monogr. Series, No. 5, pp. 377-382. Lovell, J.S., 1979. Applications of vapour geochemistry to mineral exploration. Unpubl. Ph.D. thesis, Univ. London, 195 pp. Lovell, J.S. and Hale, M., 1983. Application of soil air carbon dioxide and oxygen measurements to mineral exploration. Trans. lnstn. Min. Metall., 92: B28-32. Lovell, J.S. and Reid, A.R., 1989. Carbon dioxide / oxygen in the exploration for sulphide mineralisation. In: G.G. Garland (ed.), Proc. Exploration '87. Ontario Geol. Survey, Spec. Vol., 3, pp457-470. Lovell, J.S., Hale, M. and Webb, J.S., 1979. Soil air disequilibria as a guide to concealed mineralisation at Keel, Eire. In: Prospecting in Areas of Glaciated Terrain 1979. lnstn. Min. Metall., London, pp. 45-50. Loveli, J.S., Hale, M. and Webb, J.S., 1980. Vapour geochemistry in mineral exploration. Mining Mag., 143: 229-239. Lovcil, J.S., Hale, M. and Webb, J.S., 1983. Soil air carbon dioxide and oxygen measurements as a guide to concealed mineralisation in semi-arid and regions. J. Geochem. Explor., 19:305-317. Lucas, H.F., 1957. Improved low-level alpha-scintillation counter for radon. Rev. Scientific Instruments, 28: 680-683. Lupton, J.E., Klinkhammer, G.P., Normark, W.R., Haymon, R., Macdonald, K.C., Weiss, R.F. and Craig, H., 1980. Helium-3 and manganese at the 20~ East Pacific rise hydrothermal site. Earth Planet. Sci. Lett., 50:115- i 27. Lupton, J.E., Weiss, R.F. and Craig, H., 1977. Mantle helium in Red Sea brines. Nature, 266: 244-246. MacEivain, R.C., 1963. What do near surface signs really mean in oil finding? Oil and Gas J., February, pp. 132-146. MacEIvain, R.C., 1969. Mechanics of gaseous ascension through a sedimentary column. In: W.B. Heroy (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas. Southern Methodist University Press, Dallas, pp. 15-18. Makkaveev, A.A., 1960. Problems and methods of radiohydrogeological investigations. Works of the Institute of Geological Sciences, ANBSSR, pp. 209-215 (in Russian). Malhotra, R.V., Bimie, R.W. and Johnson, G.D., 1989. Detection of surficial changes associated with hydrocarbon seepage, Sheep Mountain anticline, Bighorn basin, Wyoming. Proc. Seventh Thematic Conference on Remote Sensing for Exploration Geology, Calgary, pp. 1097-l 110. Maimqvist, L. and Kristiansson, K., 1984. Experimental evidence for an ascending microflow of geogas in the ground. Earth and Planetary Science Letters, 70: 407-416. Malmqvist, L. and Kristiansson, K., 1985. A physical mechanism for the release of free gases in the lithosphere. Geoexploration, 23: 447-453.
References
493
Malone, E.J., Olgers, F., Cucchi, F.G., Nicholas, T. and McKay, W.J., 1975. Woodlawn copper-leadzinc deposit. In: C.L. Knight (ed.), Economic Geology of Australia and Papua New Guinea, Metals, Vol. 1. Australas. Inst. Min. Metall., Monogr. Series, No. 5, pp. 701-710. Mamyrin, B.A., Tolstykhin, I.N., Anufriyev, G.S. and Kamenskiy, I.L.,1969. Isotopic analysis of terrestrial helium on a magnetic resonance mass spectrometer. Geokhimiya, 5. Mamyrin, B.A., Tolstykhin, I.N., Anufriyev, G.S. and Kamenskiy, I.L., 1972. Helium isotopic composition of thermal springs of Iceland. Geochem. Intemat., 9: 948. Mamyrin, B.S., 1979.3He/aHe ratios in earthquake forecasting. Geochem. lntemat., 16: 42-44. Marine, I.W., 1979. The use of naturally occurring helium to estimate groundwater velocities for studies of geologic storage of radioactive waste. Water Resour. Res., 15:1130-1136. Martin, J.P. and Berquist, L.E., 1977. Study of the applicability of 3He/4He ratio for uranium prospecting. US Dept. of Energy, Open File Report, GJBX-58(77), 95 pp. Matthews, M.D., 1981. Effects of hydrocarbon leakage on earth-surface materials. In: B.M. Gottlieb (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas II. Southern Methodist University Press, Dallas, pp. 29-31. Matthews, M.D., 1986. Effects of hydrocarbon leakage on earth surface materials. In: M.J. Davidson (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas IV, Southern Methodist University, Dallas, pp. 27-44. Mat-thews, M.D., 1996a. Importance of sampling design and density in target recognition. In: D. Schumacher and M.A. Abrams (eds.), Hydrocarbon Migration and its Near-Surface Expression, Amer. Assoc. Pet. Geol.., Memoir 66: 243-253. Matthews, M.D., 1996b. Migration - a view from the top. In: D. Schumacher and M.A. Abrams (eds.), Hydrocarbon Migration and its Near-Surface Expression, Amer. Assoc. Pet. Geol., Memoir 66: 139-155. Matthews, M.D., Jones, V.T. and Richers, D.M., 1984. Remote sensing and hydrocarbon leakage. Presented at 3rd Intl. Symp. Remote Sensing of Environment, Colorado Springs. M~or, E., 1972. Palaeotemperatures and other hydrological parameters deduced from noble gases dissolved in groundwaters, Jordan Rift Valley, Israel. Geochim. Cosmochim. Acta, 36:1321-I 336. Mazor, E. and Foumier, R.O., 1973. More on noble gases in Yellowstone National Park hot waters. Geochim. Cosmochim. Acta, 37:515-525. Mazor, E. and Verhagen, B.T., 1976. Hot springs of Rhodesia: their noble gases, isotopic and chemical composition. J. Hydrol., 28: 29-43. Mazor, E. and Wasserburg, G.J., 1965. Helium, neon, argon, krypton and xenon in gas emanations from Yellowstone and Lassen volcanic National Parks. Geochim. Cosmochim. Acta, 29: 443-454. Mazor, E., Van Everdingen, R.O. and Krouse, H.R., 1983. Noble-gas evidence for geothermal activity in a karstic terrain: Rocky Mountains, Canada. Geochim. Cosmochim. Acta, 47:1111-1115. McAuliffe, C., 1966. Solubility in water of paraffin, cycloparaffin, olefin, acetylene, cycloefin, and aromatic hydrocarbons. J. Phys. Chem., 70: 1267-1275. McCarthy, J.H., 1972. Mercury vapour and other volatile components in the air as guides to ore deposits. J. Geochem. Explor., 1: 143-162. McCarthy, J.H. and Bigelow, R.C., 1990. Multiple gas analyses using a mobile mass spectrometer. J. Geochem. Explor., 38: 233-245. McCarthy, J.H. and McGuire, E., 1998. Soil gas studies along the Carlin Trend, Eureka and Elko Counties, Nevada. In: Contributions to the Metallogeny of Northern Nevada. US Geol. Survey, Open File Report, 98-338, pp. 243-250. McCarthy, J.H. and Reimer, G.M., 1986. Advances in soil gas geochemical exploration for natural resources: some current examples and practices. J. Geophys. Res., 91: 12,327-12,338.
494
References
McCarthy, J.H., Lambe, R.N. and Dietrich, J.A., 1986. A case study of soil gases as an exploration guide in glaciated terrain - Crandon massive sulphide deposit, Wisconsin. Econ. Geol., 81: 408420. McCarthy, J.H., Vaughn, W.W., Learned, R.E. and Mueschke, J.L., 1969. Mercury in soil gas and air; a potential tool in exploration. US Geol. Survey, Circ. 609, 16 pp. McCorkell, R.H., Archambault, G. and Brameld, M.P., 1981. Laboratory apparatus for measuring radon emanation from solids. Can. Inst. Min. Metall. Bull., 74: 90-96. McCoy, R.M. and Wullstein, L.H., 1988. Mapping iron and manganese concentrations in sagebrush over the Blackburn oil field, Nevada. Proc. Sixth Thematic Conference on Remote Sensing for Exploration Geology, Ann Arbor, pp. 523-528. McCoy, R.M., Scott, L.F. and Hardin, P., 1989. The spectral response of sagebrush in areas of hydrocarbon production. Proc. Seventh Thematic Conference on Remote Sensing for Exploration Geology, Calgary, pp. 751-756. McCrossan, R.G., Ball, N.L. and Snowdon, L.R., 1971. An evaluation of surface geochemical prospecting for petroleum, Olds Caroline Area, Alberta. Geol. Survey of Canada. Paper 71-31, pp. 4-9. McDermott, E., 1939. Concentrations of hydrocarbons in the earth. Geophysics, 4: 195-209. McDermott, E., 1940. Geochemical exploration (soil analysis) with speculation about the genesis of oil, gas, and other mineral accumulations. Amer. Assoc. Pet. Geol. Bull., 24:. 859-881. McFadden, W.H., 1973. Techniques of combined gas chromatography/mass spectrometry. Wiley and Sons, New York, 463 pp. McKenna, E.J. and Kallio, R.E., 1965. The biology of hydrocarbons. Annual Reviews of Microbiology, 19:182-208. McNerney, J.J. and Buseck, P.R., 1973. Geochemical exploration using mercury vapour. Econ. Geol., 68: 1313-1320. Megumi, K. and Mamuro, T., 1977. Concentration of uranium series nuclides in soil particles in relation to their size. J. Geophys. Res., 82: 353-356. Melvin, J.D., Shapiro, M.H. and Copping, N.A., 1978. An automated radon-thoron monitor for earthquake prediction research. Nucl. Instrum. and Methods, 153:239-251. Meyer, W.T. and Peters, R.G., 1973. Evaluation of sulphur as a guide to buried sulphide deposits in the Notre Dame Bay area, Newfoundland. In: Prospecting in Areas of Glaciated Terrain. Instn. Min. Metail., London, pp. 55-66. Meyer, W.T., Lovell, J.S. and Hale, M., 1986. Detection of concealed mineral and energy resources by vapour geochemistry. In: I. Thornton and R.J. Howarth (eds.), Applied Geochemistry in the 1980s. Graham and Trotman, London, pp. 86-102. Mills, R. van Auken and Wells, R., 1919. The evaporation and concentration of waters associated with petroleum and natural gas. US Geol. Survey, Bull. 693,104 pp. Mogro-Campero, A. and Fieischer, R.L., 1976. Subterrestrial fluid convection: a hypothesis for longdistance migration of radon within the earth. Earth Planet. Sci. Lett., 34:321-325. Mogro-Campero, A. and Fleischer, R.L., 1980. Search for long-distance migration of subsurface radon. In: T.F. Geseil and W.M. Lowder (eds.), Natural Radiation Environment Ill. National Technical Information Service, US Dept. of Commerce, Springfield, Va., pp. 72-89. Mogro-Campero, A., Fleischer, R.L. and Likes, R.S., 1980. Changes in subsurface radon concentration associated with earthquakes. J. Geophys. Res., 85: 3053-3057. Monteith, J.L., Szecz, G. and Yabuki, K., 1964. Crop photosynthesis and the flux of CO2 below the canopy. J. Appl. Ecol., I: 321. Mooney, J.R., 1982. Sniffing for oil and earthquakes. The Orange Disk, 25: 26-29.
References
495
Moore, H.E. and Poet, S.E., 1976. Background levels of 226Ra in the lower troposphere. Atmos. Environ., 10:381-383. Moore, V.S. and Anderson, R.L., 1985. Application of thematic mapper data for hydrocarbon exploration in the Hardeman Basin of north Texas. Proc. Fourth Thematic Conference on Remote Sensing for Exploration Geology, San Francisco, pp. 645-653. Morrison, P. and Pine, J., 1955. Radiogenic origin of the helium isotopes in rock. Ann. N.Y. Acad. Sci., 62: 69. Morrison, W.A., Pollock, H.C. and Scarth, R.F., 1969. Radon detection as a method for uranium prospecting. General Electric Research and Development Center, 69-C-085, 29 pp. Morse, R.H., 1969. Radium geochemistry applied to prospecting for uranium. Can. Min. J., 90: 75-76. Morse, R.H., 1970. The surficial geochemistry of radium, radon and uranium near Brancroft, Ontario, with applications to prospecting for uranium. Unpubl. Ph.D. thesis, Queen's Univ., Kingston, Ontario, 186 pp. Morse, R.H., 1976. Radon counters in uranium exploration. In: Exploration for Uranium Deposits. IAEA, Vienna, pp. 229-239. Mousseau, R.J. and Williams, J.C., 1979. Dissolved hydrocarbons in coastal waters of North America (abs). Amer. Assoc. Pet. Geol. Bull., 63: 699-700. Nagao, K., Takaoka, N. and Matsubayashi, O., 1979. Isotopic anomalies of rare gases in the Nigorikawa geothermal area, Hokkaido, Japan. Earth. Planet. Sci. Lett., 44: 82-90. Nash, J.T., 1979. Uranium and thorium in granitic rocks of northeastern Washington and northern Idaho with comments on uranium resource potential. US Geol. Survey, Open File Report, 79-233, 39 pp. National Council on Radiation Protection and Measurements, 1976. Environmental Radiation Measurements. NCRP Report, 50, 234 pp. Naumov, G.B., Korenov, A.B., Gerasimovskii, A. and Germanov, A.I., 1963. Uranium in sedimentary rocks. In: A.P. Vinogradov, (ed.), Fundamental Features of the Geochemistry of Uranium, pp. 238289 (in Russian). Ncrken, A., 1972. Procedure for calibrating gas analysers of the mass spectrometer type. J. Vac. Sci. Technol., 9:1260-1274. Ncwton, J.C., Crawford, R.W. and Stump, R.K., 1975. A portable mass spectrometer for field and laboratory use. Lawrence Livermore Lab. Univ. Cal. Report U.C.I.D. 16823, 17 pp. Newton, R., and Round, H.F., 1961. Diffusion of helium in sedimentary rocks. Geochim. Cosmochin. Acta., 22: 106-132. Nikiforova, Y.H. and Yufa, B.Ya., 1970. Features of uranium and radium migration in mountain landscapes and their role in radiometric prospecting. Moscow, Univ., Vestn., Ser. Geogr., 4: 75-81. Nikonov, V.F., 1971. Distribution of methane homologs in gas and oil fields. Akademiya Nauk SSSR Doklady, 206: 234-236. Niisson, G., 1971. The use of dogs in prospecting for sulphide ores. Geol. Foeren. Stockholm, Foerh., 547: 725-728. Niisson, G., 1973. Nickel prospecting and the discovery of the Mjoevattnet mineralisation, northern Sweden: a case history of the use of combined techniques in drift-covered glacial terrain. In: Prospecting in Areas of Glaciated Terrain. Instn. Min. Metall., London, pp. 97-109. Novikov, T.F. and Kapkov, Y.V., 1965. Radioactive Methods of Prospecting. Nedra, Leningrad, 780 pp. (in Russian).
496
References
Nuutilainen, J. and Peuraniemi, V., 1977. Application of humus analysis in geochemical prospecting: some case histories. In. Prospecting in Areas of Glaciated Terrain 1977. Instn. Min. Metall., London, pp. 1-5. Oakes, B.W., 1984. Vapour geochemical pathfinders for oxidising sulphide mineralisation beneath exotic overburden. Unpubl. PhD thesis, Univ. London. Oakes, B.W. and Hale, M., 1987. Dispersion patterns of carbonyl sulphide above mineral deposits. J. Geochem. Explor., 28: 235-249. Oehler, D.Z. and Stemberg, B.K., 1984. Seepage-induced anomalies, false anomalies and implications for electrical prospecting. Amer.Assoc. Pet. Geol. Bull., 68:1121-1145. Okabe, S., 1956. Time variation of the atmospheric radon-content near the ground surface with relation to some geophysical phenomena, Univ. Kyoto, Mem. Coll. Sci., 28: 99-115. O'Keefe, A.E. and Ortman, C., 1966. Primary standards for trace gas analysis. Analyst, 38: 760-763. Oliveria, W.J. de and Crosta, A.P., 1996. Detection of hydrocarbon seepage in the Sao Francisco basin, Brazil, through Landsat TM, soil geochemistry and airborne/field spectrometry data integration. Proc. Eleventh Thematic Conference and Workshops on Applied Geologic Remote Sensing, Las Vegas, pp. 1-155-165. Olivier, B.M., Bradley, J.G. and Farrar, H., 1984. Helium concentration in the Earth's lower atmosphere. Geochim. Cosmochim. Acta, 48: 1759-1767. Openshaw, R.E., 1983. Hg and As soil geochemistry of the Meager Creek geothermal area. J. Geochem. Explor., 19: 339-344. Orlov, A.P., Roboneu, V.I. and Kirilenko, G.M., 1969. Geologicheskie poiski S rudorozysknumi sobakami. Nedra, Moscow, 47 pp. (in Russian). Ostrihansky, L, 1976. Radioactive disequilibrium investigations, Elliot Lake area, Ontario. Geol. Survey Canada, Paper 75-38, pp. 21-48. Ovchinnikov, L.N., Sokolov, V.A., Fridman, A.I. and Yanitskii, I.N., 1972. Gaseous geochemical methods in structural mapping and prospecting for ore deposits. In: M.J. Jones (ed.), Geochemical Exploration 1972. Instn. Min. Metall., London, pp. 177-183. Owen, R. L., 1972. Hydrocarbon Gas Sensor. US Patent No. 3-651-395. Ozerova, N.A., 1959. The use of primary dispersion haloes of Hg in search for Pb-Zn deposits. Geokimya, 7: 793-802. Ozerova, N.A., 1962. Primary dispersion haloes of mercury. Acad. Sci. USSR, Inst. Geol. Ore Deposits, Petrography, Mineralogy and Geochem., Proc. 72, 135 pp. (in Russian). Ozerova, N.A., 1971. Primary dispersion haloes of mercury. IGR Book, Section 13, 109 pp. Ozerova, N.A. and Mashianov, N.R., 1989. Achievements in the mercury measuring survey. In: Geochemical Methods and Scientific-Technical Progress in Geochemical Investigations of the Earth entrails. Nauka, Moscow, pp 69-77 (in Russian). Ozerova, N.A., Rusinov, V.L. and Ozerov, Y.K., 1975. The mercury in sulphide deposits emplaced in volcanic suites. Mineral. Deposita, 10: 228-233.
Pacer, J.C., 1977. Radon and helium studies in areas of known uranium mineralisation. Bendix Field Engineering Corp., Grand Junction, Colorado, 52 pp. Pacer, J.C., 1981. Study of radon released from open drill holes. National Uranium Resource Evaluation, Report GJBX-146, Bendix Field Engineering Corp., Grand Junction, Colorado., pp. 124. Pacer, J.C. and Czamecki, R.F., 1980. Principles and characteristics of surface radon and helium techniques used in uranium exploration. US Dept. of Energy, Report GJBX-177(80), 63 pp.
References
497
Panchenkov, G.M., 1964. Dynamics of unstable physico-chemical processes on the condition of bimolecular reaction speed law. J. Physical Chemistry, 38:770-773 (in Russian). Panov, B.S., Alexin, V.I. and Riaboshtian, Yu.S., 1980. Sedimentary rocks and ores. Naukova Dumka, Kidv., pp. 101 - l I l (in Ukranian). Parasnis, D.S., 1973. Mining Geophysics (2nd edn.). Elsevier, New York, 395 pp. Parasnis, D.S., 1979. Principles of Applied Geophysics (3 ~dedn.). Wiley, New York, 275 pp. Partington, J.R., 1957. The discovery of radon. Nature, 179: 912. Patton, K.H. and Manwaring, M., 1984. Evaluation of a Landsat-derived tonal anomaly for hydrocarbon microseeps, southwest Kansas. Proc. 3rd Int. Symp. Remote Sensing of the Environment. Colorado Springs, pp. 441-450. Paull, C.K., Spiess, F.N., Ussler, W. and Borowski, W.A., 1995. Methane-rich plumes on the Carolina continental rise: association with gas hydrates. Geology, 23: 89-92. Peacock, J.D. and Williamson, R., 1962. Radon determination as a prospecting technique. Trans. Instn. Min. Metall., 71: 75-83; 271-277. Pearson, J.E. and Jones, G.E., 1966. Soil concentrations of " Emanating radium-226" and the emanation of Radon-222 from soils and plants. Tellus, 18: 655-662. Pearson, J.E. and Moses, H., 1963. The effects of atmospheric stability and horizontal shear on the accumulation and diffusion of radon. Division of Radiological Physics, Argonne National Laboratory, Report ANL-6646, pp. 151 - 167. Pearson, J.E. and Moses, H., 1966. Atmospheric radon-222 concentration variation with height and time. J. Appl. Meteorol., 5: 175-181. Peters, R.G., 1973. Secondary dispersion of sulphur from sulphide deposits in the Notre Dame Bay area, Newfoundland. Unpubl. Ph.D. thesis, Univ. London, 274 pp. Pflug, K.A., Killeen, P.G. and Mwenifumbo, C.J., 1996. Borehole geophysical logs in gold deposits in the Kirkland Lake area, Ontario (Macassa gold mine and Victoria Creek gold deposit). Geol. Survey Canada, Open File Report 3153, 24 pp. Phair, G. and Gottfried, D., 1964. The Colorado Front Range, Colorado, USA, as a uranium and thorium province. In: J.A.S. Adams and W.M. Lowder (eds.), The Natural Radiation Environment. Univ. Chicago Press, pp. 7-38. Phelps, D.W. and Buseck, P.R., 1978. Mercury in soils as an indicator of geothermal activity: Yellowstone National Park, Wyoming. In: J.R Watterson and P.K. Theobald (eds.), Geochemical Exploration 1978. Assoc Explor. Geochem., Rexdale, Ontario., pp. 153-160. Phiip, R.P. and Crisp, P.T., 1982. Surface geochemical methods used for oil and gas prospecting - a rcview. J. Geochem. Explor., 17: 1-34. Pirkle, R.J., 1985. Hydrocarbon seeps in the Gulf of Mexico. Presented at 189th Annual Meeting, Amer. Chem. Soc., Miami Beach, Florida. Pirkle, R.J. and Drozd, R.J., 1984. Hydrocarbon contamination in the near soil (abs). Presented at 187th Annual Meeting, Amer. Chem. Soc., St. Louis. Pirkle, R.J. and Jones, V.T., 1983. Helium and hydrogen anomalies associated with deep or active faults. Presented at 181 st Annual Meeting, Amer. Chem. Soc., Atlanta. Pirson, S.J., 1946. Disturbing factors in geochemical prospecting. Geophysics, 1 l: 312-320. Pirson, S.J., 1969. Geological, geophysical and chemical modifications of sediments in the environment of oil fields. In: W. B. Heroy (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas. Southern Methodist University Press, Dallas, pp. 159-186. Pirson, S.J., 1981. Significant advances in magnetoelectrical exploration. In: B.M. Gottlieb (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas Il. Southern Methodist University Press, Dallas, pp. 169-196.
498
References
Pixler, B.O., 1969. Formation evaluation by analysis of hydrocarbon ratios. J. Petroleum Technology, 21: 665-670. Pogorski, L.A. and Pogorski, S.R., 1982. The effect of the physical properties of the overburden on helium emanometry. US Dept. of Energy, Open File Report, GJBX- 114(82), 191 pp. Pogorski, L.A. and Quirt, G.S., 1981. Helium emanometry in exploration for hydrocarbons, Part I. In: B.M. Gottlieb (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas II. Southern Methodist University Press, Dallas, pp. 136-149. Pogorski, L.A. and Quirt, G.S., 1979. Helium emanometry as an indicator of deeply buried uranium deposits. US Dept. of Energy, Open File Report, GJBX-22(80), 362 pp. Polanski, A., 1965. Geochemistry of Isotopes. TT61-31327. Warsaw, Poland, 474 pp. (transl. from Polish). Poll, J.J.K., 1975. Onshore Gippsland geochemical survey: a test case for Australia. J. Australian Petroleum Explor. Assoc. 15: 93-101. Pomarskii, A.A., Sevemyi, S.A. and Trifonova, E.P., 1968. Determination of ultra-microquantities of 226Ra in various substances. Soviet Atomic Energy, pp. 718-721 (transl. from Atomnaya Energiya, 27: 36-39). Pradel, J., Billard, F., Mirabel, J., Gangloff, A., Puybarand, Y. and Tayeb, G., 1963. La Prospection de l'uranium par le radon. Commissariat a l'Energie Atomique, Report CEA-2330 (translated into English as "Uranium Prospecting with Radon"), AEC-TR-6472, 30 pp. Price, A. and Heatherington, A., 1984. The influence of soil/sediment pH of minerals on the adsorbed hydrocarbon technique for geochemical exploration for petroleum (abs). Presented at 187th ACS National Meeting, St. Louis, April 8-13. Price, L.C., 1986. A critical overview and proposed working model of surface geochemical exploration. In: M.J. Davidson (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas IV. Southern Methodist University, Dallas, pp. 245-304. Prutkina, M.I. and Shashkin, B.L., 1967. The emanation of radon from uranium ores and minerals into liquids, USSR Atomic Energy, 22: 140-141. Punanova, S.A., 1974. Trace elements in oils and their use for geochemical investigations. Nedra, Moscow, 216 pp. (in Russian). Purvis, D.B., Albright, J.L., Hart, G., Smith, H.F. and Whelan, T. 11I, 1977. Seismic fluorimetric and chromatographic studies of reservoir leakage. Presented at 9th Annual AIME Offshore Technological Conference, Preprint No. OTC-2937, 3: 456-458. Putikov, O.F., 1977. Polarographic logging for groundwater composition investigation. In: Principles and Techniques of Exploration, Rotaprint VITR, N117:37-45 (in Russian). Putikov, O.F., 1980. Geoelectrochemical Methods of Prospecting and Exploration. Rotaprint LGI, Leningrad, 56 pp. (in Russian). Putikov, O.F., 1987. Theoretical basis of the geoelectrochemical exploration methods. Leningrad State University Press, Leningrad, 176 pp. (in Russian). Putikov, O.F., 1993. The Geoelectrochemical Methods of Prospecting and Exploration. Rotaprint SPSMI, St Petersburg, 118 pp. (in Russian). Putikov, O.F., 1995. Principles of theory and practice of the non-linear geoelectrochemical methods. In: Proc. |ntemational Conference, Nijne-Voljsky Institute Geologii i Geofyziky, Saratov, pp. 2937 (in Russian). Putikov, O.F. and Dukhanin, A.S., 1994. About the possible mechanism of formation of "stream" halos. Reports of the Russian Academy of Sciences, 338:219-221 (in Russian).
References
499
Putikov, O.F. and Wen, B., 1997. Mathematical and physico-chemical modelling of concentration distribution of heavy metals in "stream" halos of dispersion. Abstracts of International Geophysical Conference, Moscow '97, G.2-8, 1 p. Putikov, O.F., Dukhanin, A.S. and Mashianov, N.R., 1994. Toward the foundation of a physicomathematical model of"stream" halos of dispersion. Russian Geophys. J., N2: 5-10 (in Russian). Raines, G.L., Erdman, J.A., McCarthy, J.H., Jr. and Reimer, G.M., 1985. Remotely-sensed limonite anomaly on Lordsburg Mesa, New Mexico: possible implications for uranium deposits. Econ. Geol., 80: 575-590. Rankama, K. and Sahama, Th.G., 1950. Geochemistry. Univ. Chicago Press, 912 pp. Rasmussen, R.A., 1974. Emission ofbiogenic hydrogen sulphide. Tellus, 26: 254-260. Reid, A.R. and Rasmussen, J.D., 1990. The use of soil-gas CO2 in the exploration for sulphide-bearing breccia pipes in northern Arizona. J. Geochem. Explor., 38: 87-101. Reid, K.O. and Meares, R.M.D., 1981. Exploration for volcanic-hosted sulphide deposits in western Tasmania. Econ. Geol., 76: 350-364. Reid, N., lwashita, A., Yamashita, Y. and Thompson, K., 1988. High resolution imaging of geobotanical anomalies associated with subsurface hydrocarbons. Proc. Sixth Thematic Conference on Remote Sensing for Exploration Geology, Houston, Vol. 1, pp. 213-223. Reimer, G.M., 1976. Design and assembly of a portable helium detector for evaluation as a uranium exploration instrument. US Geol. Survey, Open File Report, 76-398, 10 pp. Reimer, G.M., 1980. Use of soil gas helium concentrations for earthquake prediction, limitations imposed by diagonal variation. J. Geophys. Res., 85:3107-3114. Reimer, G.M., 1981. Helium soil-gas variations associated with recent central California earthquakes: precursor or coincidence. Geophys. Res. Lett., 8: 433-435. Rcimer, G.M., 1986. Helium soil-gas survey of the Aurora uranium deposit, McDermitt caldera complex, Oregon. J. Geophys. Res. 91: 12,355-12,358. Reimer, G.M. and Otton, J.K., 1976. Helium in soil gas and well water in the vicinity of a uranium deposit, Weld County, Colorado. US Geol. Survey, Open File Report, 76-699, 10 pp. Rcimer, G.M., Bowles, C.G., Murray, D.G. and Been, J.M., 1979a. Helium in soil-gas and helium/radon in groundwater in the vicinity of a south Texas uranium roll-type deposit. US Geol. Survey, Open File Report, 79-1625, 10pp. Reimer, G.M., Denton, E.H., Friedman, I. and Otton, J.K., 1979b. Recent developments in uranium exploration using the US Geological Survey's mobile helium detector. J. Geochem. Explor., 11: 112. Reimer, G.M., Roberts, A.A. and Denton, E.H., 1976. Diurnal effects on the helium concentration in soil-gas and near surface atmosphere. US Geol. Survey, Open File Report, 76-715, 6 pp. Reitsema, R.H., Lindberg, F.A. and Kaltenback, A.J., 1978. Light hydrocarbons in Gulf of Mexico water: sources and relation to structural highs. J. Geochem. Explor., 10:139-151. Rice, D.D., 1980. Chemical and isotopic evidence of the origins of natural gases in offshore Gulf of Mexico. Trans.Gulf Coast Assoc.Geol. Socs., 30: 203-213. Rice, R.S., and Reimer, G.M., 1977. Helium analysis of subsurface waters as a uranium exploration tool. Symp. on Hydrogeochemical and Stream Sediment Reconnaissance for Uranium in the USA. US Dept. Energy, Oct. 1977. Richardson, K.A., 1964. Thorium, uranium and potassium in the Conway granite, New Hampshire, USA. In: J.A.S. Adams and W.M. Lowder (eds.), The Natural Radiation Environment. Univ. Chicago Press, pp. 39-50.
500
References
Richers, D.M., 1984. Comparison of commonly applied soil-gas techniques used in evaluating hydrocarbon potential. Presented at 187th Annual Meeting, Amer. Chem. Soc., St. Louis. Richers, D.M. and Jones, V.T., 1986. The 1983 Landsat soil-gas geochemical survey of Patrick Draw area, Wyoming. Amer. Assoc. Pet. Geol. Bull., 70: 869-887. Richers, D.M. and Weatherby, C., 1985. A continued study of the Patrick Draw test site, Sweetwater County, Wyoming (abs.). Presented at 4th Int. Symp. Remote Sensing of the Environment, San Francisco, April 1-4. Richers, D.M., Jones, V.T., Matthews, M.D., Maciolek, J., Pirkle, R.J. and Sidle, W.C., 1986. The 1983 Landsat soil gas geochemical survey of the Patrick Draw area, Sweetwater County, Wyoming. Amer. Assoc. Pet. Geol. Bull., 70: 869-887. Richers, D.M., Reed, R.J., Horstman, K.C., Micheals, G.D., Baker, R.N., Lundell, L. and Marrs, R.W., 1982. Landsat and soil-gas geochemical study of Patrick Draw oil field, Sweetwater County, Wyoming. Amer. Assoc. Pet. Geol. Bull., 66: 903-922. Robbins, J.C., 1973. Zeeman spectrometer for measurement of atomic mercury vapour. In: Geochemical Exploration 1972. lnstn. Min. Metall., London, pp. 315-323. Roberts, A.A., 1981. Helium emanometry in exploring for hydrocarbons, Part II. In: B.M. Gottlieb (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas II. Southern Methodist University Press, Dallas, pp. 136-149. Roberts, A.A., Friedman, I., Donovan, T.J. and Denton, E.H., 1975. Helium survey, a possible technique for locating geothermal reservoirs. Geophys. Res. Lett., 2: 209-210. Roberts, W.M.B., 1973. Dolomitisation and the genesis of the Woodcutters lead-zinc prospect, Northern Territory, Australia. Mineral. Deposita, 8: 35-36. Robertson, G.W. and Bracewell, J.M., 1979. Rapid representative sampling of soil gases and their determination by mass spectrometry. J. Soil Sci., 30: 681-689. Robertson, J.B., 1969. Diffusion from a gaseous source in a porous medium. A field and theoretical comparison. US Geol. Survey, Prof. Pap., 650-D: 265-273. Rock, B.N., 1984. Remote detection of geobotanical anomalies associated with hydrocarbon microseepage. Proc. Third Thematic Conference on Remote Sensing for Exploration Geology, Colorado Springs, pp. 183-195. Rock, B.N., 1985. Vegetation response to geochemical stress, an overview, El-El8. In: Surface and Near Surface Geochemical Methods in Petroleum Exploration. Presentation at Amer. Assoc. Pet. Geol., Rocky Mountain Section, Denver. Rock, B.N., Hoshizaki, T. and Miller, J.R., 1988. Comparison of in situ and airborne spectral measurements of the blue shit~ associated with forest damage. Rem. Sens. Env., 24: 109-127. Rogers, A.S., 1958. Physical behaviour and geologic control of radon in mountain streams. US Geol. Survey, Bull. ! 052-E, pp. 187-211. Romer, D.M. and Finlay, S., 1984. Soil gas investigations at Navan ore body, Ireland. In: Prospecting in Areas of Glacial Terrain 1984. Instn. Min. Metall., London, pp. 53-57. Rosaire, E.E., 1938. Shallow stratigraphic variations over Gulf Coast structures. Geophysics, 3: 96115. Rosaire, E.E., 1940. Geochemical prospecting for petroleum. Amer. Assoc. Pet. Geol. Bull., 24: 14001433. Rose, A.W. and Korner, LA., 1978. Radon in natural waters as a guide to uranium deposits in Pennsylvania. In: J.R. Watterson and P.K. Theobald (eds.), Geochemical Exploration 1978. Assoc. of Exploration Geochemists, Rexdale, Ontario, pp. 65-76. Rose, A.W., Hawkes, H.E. and Webb, J.S., 1979. Geochemistry in Mineral Exploration (2"d edn.), A cad. Press, London, 657 pp.
References
501
Rose, A.W., Wesolowski, D. and Smith, A.T., 1983. Groundwater geochemistry of uranium and other elements, Monticello area, New York. J. Geochem. Explor., 19:163-186. Rosholt, J.N., Jr., 1959. Natural radioactive disequilibrium of the uranium series. US Geol. Survey, Bull. 1084-A, 30 pp. Rosier, H.J., Beuge, P., Pilot, J. and Tischendorf, G., 1977. Integrated geochemical exploration for deep-seated solid and gaseous mineral resources. J. Geochem. Explor., 8: 415-429. Ross, D.J. and Roberts, H.S., 1970. Enzyme activities and oxygen uptakes of soils under pasture in temperature and rainfall sequences. J. Soil Sci., 21: 368-381. Rouse, G.E. and Stevens, D.N., 1971. Address to American Institute of Mining and Metallurgical Engineering, March. Rowntree, J.C. and Mosher, D.V., 1976. Case history of the discovery of the Jabiluka uranium deposits, East Alligator River region, Northern Territory of Australia. In: Exploration for Uranium Ore Deposits. IAEA, Vienna, pp. 551-573. Rozen, O.M., and Yanitskii, I.N., 1974. Main structures of the Precambrian Kokchetav Block and subsurface helium flow. Dokl. Akad. Nauk SSSR, 216: 53-56. Ruan, T.J. and Cheng, J.P., 1991. Hydrostripping of absorbed hydrocarbons in soil samples - a new method in geochemical exploration for oil and gas. J. Southeast Asian Earth Sci., 5(4): 5-7. Ruan, T.J., Hale, M. and Howarth, R.J., 1985a. Numerical modelling experiments in vapour geochemistry, II: vapour dispersion patterns and exploration implications. J. Gcochem. Explor., 23: 265-280. Ruan, T.J., Howarth, R.J. and Hale M., 1985b. Numerical modelling experiments in vapour geochemistry, I: method and Fortran program. Computers and Geosciences, 11: 55-67. Rushing, D.R., Garcia, W.J. and Clar, K.A., 1964. Analysis of effluents and environmental samples. In: Proc. symposium on radiological health and safety in mining and milling of nuclear materials. IAEA, Vienna, pp. 187-230. Russell, E.H., 1973. Soil Conditions and Plant Growth (10th edn.). Longman, London. Ryall, W.R., 1979a. Guidelines for the collection, preparation and storage of geochcmical samples for mercury analysis. Tech. Commun., 67, Institute of Earth Resources, CSIRO. Ryall, W.R., 1979b. Mercury distribution in the Woodlawn massive sulphide deposit, New South Wales. Econ Geol., 74: 1471-1484. Ryall, W.R., 1979c. Mercury in and around lead-zinc-silver ore bodies at the Isa mine, Mount lsa. Unpubl. Restricted Inv. Report, 1042R, Institute of Earth Resources, CSIRO. Ryall, W.R., 1980. Mercury distribution in and around zinc-lead silver deposits of the McArthur area, Northern Territory. Unpubl. Restricted Inv. Report, 1101R, Institute of Earth Resources, CSIRO. Ryall, W.R., 1981a. The forms of mercury in some Australian stratiform Pb-Zn-Ag deposits of different metamorphic grades. Mineral. Deposita, 16: 425-435. Ryall, W.R., 1981b. Mercury contents of some Australian uranium deposits. Econ. Geol., 76: 157-60. Ryall, W.R. and Binns, R.A., 1980. Mercury distributions in the Jabiluka Two ore body and its host rocks. In: J. Ferguson and A.B. Goleby (eds.), Proc. Int. Uranium Symposium on the Pine Creek Geosyncline. IAEA, Vienna, pp. 469-476. Ryss, Yu.S., 1969. The contact polarisation curve (CPC) technique for definition of mineral composition and size of ore bodies. In: Principles and Techniques of Exploration. Rotaprint VITR, N65:34-42 (in Russian). Ryss, Yu.S., 1973. Ore body prospecting and exploration by means of the contact polarisation curve (CPC) method. Nedra, Leningrad, 166 pp. (in Russian).
502
References
Ryss, Yu.S., 1979. The contactless polarisation curve (CLPC) technique for geophysical anomaly estimation and prospecting for economic ores in the prospective zones. In: Principles and Techniques of Exploration, Rotaprint VITR, N 127: 5-17 (in Russian). Ryss, Yu.S., 1981. The use of the contact polarisation curve (CPC) technique to evaluate and explore ore deposits. Geoexploration, 18:281-295. Ryss, Yu.S., 1983. Geoelectrochemical Methods of Exploration (Introduction to Geoelectrochemistry). Nedra, Leningrad, 255 pp. (in Russian). Ryss, Yu.S. and Goldberg, I.S., 1973. Method of partial extraction of metals (CHIM) for ore deposit prospecting. In: Principles and Techniques of Exploration. Rotaprint VITR, N84:5-19 (in Russian). Ryss, Yu.S., Antropova, L.V., Goidberg, I.S. and Voroshilov, N.A., 1987a. Geoelectrochemical methods for geological survey and prospecting at the 1:50000 Scale. In: Methods of Exploration Geophysics. NPO Rudgeofyzika, Leningrad, pp. 85-95 (in Russian). Ryss, Yu.S., Bakhtin, Yu.G., Popov, Yu.V. and Korostin, V.P., 1978. Prospecting and exploration of ore deposits by the CPC method: methodical recommendations. Rotaprint VITR, Leningrad, 87 pp. (in Russian). Ryss, Yu.S., Goldberg, I.S., Alekseev, S.G. and Dukhanin, A.S., 1987b. Jet migration of substances in the formation of secondary dispersion halos. Reports of Academy of Sciences of USSR, 297: 956958 (in Russian). Ryss, Yu.S., Goldberg, I.S., Vasilieva, V.I. and Voroshilov, N.A., 1990. Possibility of application of geoelectrochemical methods for oil-gas deposit prospecting. Soviet Geology, N6:28-33 (in Russian). Ryss, Yu.S., Voroshilov, N.A., Voronkov, V.D. and Tiasto, A.S., 1988. Results of a regional geoelectrochemical survey (MPF, TMGM, SP) for metallogenic prognosis. In: Geochemical Prospecting Theory and Practice Under Modern Conditions, Issue 5, p. 72 (in Russian). Sackett, W.M., 1977. Use of hydrocarbon sniffing in offshore exploration. J. Geochem. Explor., 7: 243-254. Sakrison, H.C., 1971. Rock geochemistry: its current usefulness on the Canadian Shield. Can. Inst. Min. Metali. Bull., 64:28-31. Sanderson, R.T., 1940. Some neglected aspects of chemical exploration. Geophysics, 5: 284-294. Sandy, J., 1988. Overview of gas-sensing radar survey theory, instrumentation, surveying and interpretation. Oil and Gas J., no. 2, pp. 1-8. Sano, Y., Nakamura, Y., Wakita, H., Notsu, K. and Kobayashi, Y., 1986. 3He'He ratio anomalies associated with the 1984 western Nagano earthquake; possibly induced by a diapiric magma. J. Geophys. Res., 91:12,291-12,295. Sardarov, S.S., 1981. Empirical relationship between anomalies that are short-term predictors of impending earthquakes and earthquake parameters. Dokl. Akad. Nauk. SSSR, 258: 30-34. Sato, K., Kancoka, I. and Sato, J., 1980. Rare gas releasing experiments and Rn degassing from eruption magma. J. Geochem. Explor., 14: 91-94. Sato, M. and Mooney, H.M., 1960. The electrochemical mechanism of sulphide self-potentials. Geophysics, 25: 226-249. Satterly, J. and Elworthy, R.T., 1917. Mineral springs of Canada, the radioactivity of some Canadian mineral springs, Part I. Ottawa, Canada, 53 pp. Saukov, A.A., 1946. Geochemistry of mercury. Trans. Inst. Geol. Sci., Acad. Sci. USSR, Issue 78, Series Mineralogy and Geochemistry, 17, 128 pp.
References
503
Saunders, D.F., Burson, K.R. and Thompson, C.K., 1991. Observed relation of soil magnetic susceptibility and soil gas hydrocarbon analyses to subsurface petroleum accumulations. Amer.Assoc. Pet. Geol. Bull., 75: 389-404. Scaringelli, F.P., Saltzman, B.E. and Frey, S.A., 1967. Spectrophotometric determination of atmospheric sulphur dioxide. Anal. Chem., 39:1709-1719. Schaefer, R.G, Weiner, B. and Leythaeuser, D., 1978. Determination of sub-nanogram per gram quantities of light hydrocarbons (C2-C8) in rock samples by hydrogen stripping in the flow system of a capillary gas chromatograph. Anal. Chem., 50:1848-1854. Scherbakov, A.V. and Kozlova, N.D., 1986. Hydrogen occurrence in underground fluids and the relation of anomalously high concentrations to deep faults in the USSR. Geotectonics, N2:56-66 (in Russian). Schlegel, H.G., 1974. Production, modification and consumption of atmospheric trace gases by microorganisms. Tellus, 26:12-20. Schoell, M., 1983a. Isotope techniques for tracing migration of gases in sedimentary basins. J. Geol. Soc., 140:415-422. Schoell, M., 1983b. Genetic characterisation of natural gases. Amer. Assoc. Pet. Geol. Bull., 67: 22252238. Schumacher, D., 1996. Hydrocarbon-induced alteration of soils and sediments. In: D. Schumacher and A.A. Abrams (eds.), Hydrocarbon migration and its near-surface expression. Amer.Assoc. Pet. Geol., Memoir 66, pp. 71-89. Schuttelkopf, H. and Kiefer, H., 1982. Radium-222 contamination of the black forest and the radioccological behaviour of radium. Proc. symp. Knoavil, 27-31 July, 1981. IAEA, Vienna, pp. 345-352. Scintrex Ltd., 1975. Scintrex ETR-I Emanometer. Scintrex Ltd., 222 Snidercrot~ Road, Concord, Ontario. Scott, L.F. and McCoy, R.M., 1993. Near surface soil: the interface between deep reservoir hydrocarbons and metals in sagebrush, Lodgepole oil field, Utah. Proc. Ninth Thematic Conference on Geologic Remote Sensing, Pasadena, pp. 861-869. Scott, R.C. and Barker, F.B., 1962. Data on uranium and radium in groundwater in the United States, 1954 to 1957. US Geol. Survey, Prof. Paper 426, 115 pp. Scott, W.A., 1916. Mining operations at Johnson Camp, Arizona. Mining Engineering World, 45:14 I143. Sealey, J.Q., 1974a A geomicrobiological method of prospecting for petroleum, Part 1. Oil and Gas J., vol. 72, April 8, pp. 142-146. Sealey, J.Q., 1974b. A geomicrobiological method of prospecting for petroleum, Part 2. Oil and Gas J., vol. 72, April 15, pp. 98-102. Sedlet, J., 1966. Radon and radium. In: Treatise on Analytical Chemistry, Part 2, Vol. 4. Wiley and Sons, New York, pp. 219-366. Segal, D.B. and Merin, I.S., 1989. Successful use of Landsat Thematic Mapper data for mapping hydrocarbon microseepage-induced mineraiogic alteration, Lisbon Valley, Utah. Photogramm. Eng. Rein. Sens., 55:1137-1145. Segal, D.B., Ruth, M.D. and Merin, I.S., 1986. Remote detection of anomalous mineralogy associated with hydrocarbon production, Lisbon Valley, Utah. Mountain Geologist, 23: 51-62. Segal, D.B., Ruth, M.D., Merin, I.S., Watanabe, H., Soda, K., Takano, O. and Sano, M., 1984. Correlation of remotely detected mineralogy with hydrocarbon production, Lisbon Valley, Utah. Proc. Third Thematic Conference on Remote Sensing for Exploration Geology, Colorado Springs, pp.273-292.
504
References
Senftle, F.E., 1946. Relation of conductivity and total solid content of lake water to its radium content, Great Bear Lake region. Trans. Can. Inst. Min, Metall., 49: 439-446. Seveme, B.C., 1978. Evaluation of radon systems at Yeelirrie, Western Australia. J. Geochem. Explor., 9: 1-22. Shapiro, M.H., Melvin, J.D., Tombrello, T.A., Jiang Fongliang, Li Guiru, Mendenhall, M.H. and Rice, A., 1982. Correlated radon and CO2 variations near the San Andreas Fault. Geophys. Res. Let., 9: 503-506. Shapiro, M.H., Melvin, J.D., Tombrello, T.A., Mendenhall, M. H., Larson, P.B. and Whitcomb, J.H., 1981. Relationship of the 1979 southern California radon anomaly to a possible regional strain event. J. Geophys. Res., 86: 1725. Sharp, R.J., 1976. The Hidden Creek cupriferous massive sulphide deposit, Anyox, B.C. (abs.). Can. Inst. Min. Metall. Bull., 76: 53. Sheppard, M.I., 1980. The environmental behaviour of radium. Atomic Energy of Canada Ltd., Report PAECL-6796, 28 pp. Shigaev, V.Yu., 1997. Geoelectrochemical investigations for oil and gas prospecting. Dissertation Abstracts, Saratov University, Saratov, 22 pp. (in Russian). Shiikawa, M., 1983. The role of mercury, arsenic and boron as pathfinder elements in geochemical exploration for geothermal energy. J.Geochem. Explor., 19: 337-338. Shipulin, F.K., Genkin, A.D., Distler, V.V., Komarov, P.U., Malinovskii, E.P., Ozerova, N.A., Percliman, A.I., Rekharskii, V.I., Rozanov, Yu.A., Filimonova, L.G., Chernyshev, V.F. and Elinson, M.M., 1973. Some aspects of the problem of geochemical methods of prospecting for concealcd mineralisation. J. Geochem. Explor., 2: 193-235. Shturm, L.D., 1952. Morphological characteristics of sulphate-reducing bacteria encountered in oilbearing formations. Trudy Inst. Mikrobiologii Akad. Nauk SSSR, 2: 3-32. Shvcts, V.M., 1973. Organic Substances in Underground Water. Nedra, Moscow, 190 pp. (in Russian). Siegcl, F.R., 1974. Applied Geochemistry. Wiley and Sons, New York. Silvcrman, S.R. and Epstein, S., 1958. Carbon isotopic composition of petroleums and other scdimentary organic materials. Amer. Assoc. Pet. Geol. Bull., 42: 998-1012. Simpson, C.J., Wilford, J.R., Macias, L.F. and Korsch, R.J., 1991. Remote sensing of petroleum gas seepage alteration, Palm Valley gas field, Amadeus basin, central Australia. Proc. Eighth Thematic Conference on Geologic Remote Sensing, Denver, pp. 1463-1472. Sinclair, A.J., 1976. Probability Graphs. Assoc. Explor. Geochemists, Spec. Vol. 4, 95 pp. Singhroy, V.H., Stanton-Gray, R. and Springer, J., 1986. Spectral geobotanical investigation of mineraliscd till sites. Proc. Fifth Thematic Conference on Remote Sensing for Exploration Geology, Rcno, pp. 523-544. Sittig, M., 1980. Geophysical and Geochemical Techniques for Exploration of Hydrocarbons and Minerals. Noyes Data Corp., Newark, NJ. Sivenas, P. and Beales, F.W., 1982. Natural geobatteries associated with sulphide ore deposits, I: theoretical studies. J. Geochem. Explor., 17:123-143. Smee, B.W., 1979. A theoretical estimation of ion mobilities through glaciolacustrine sediments: diffusion down a concentration gradient. Geol. Survey Canada., Paper 79-IA, pp. 367-374. Smee, B.W., 1983. Laboratory and field evidence in support of the electrochemically-enhanced migration of ions through glaciolacustrine sediment. J. Geochem. Explor., 19: 277-304. Smee, B.W., 1998. A new theory to explain the formation of soil geochemical responses over deeplycovered gold mineralisation in arid environments. J. Geochem. Explor., 61: 149-172. Smee, B.W. and Sinha, A.K., 1979. Geological, geophysical and geochemical considerations for exploration in clay-covered areas: a review. Can. Inst. Min. Mctall. Bull., 72: 67-82.
References
505
Smith, A.Y. and Dyck, W., 1969. The application of radon methods to geochemical exploration for uranium (abs.). Can. Inst. Min. Metall. Bull., 62: 215. Smith, A.Y., Barretto, P.M.C., and Poumis, S., 1976. Radon methods in uranium exploration. In: Exploration for Uranium Ore Deposits. IAEA, Vienna, pp. 185-211. Smith, B.M., Grune, W.N., Higgins, F.B. and Terrill, J.G., Jr., 1961. Natural radioactivity in groundwater supplies in Main and New-Hampshire. J. Amer. Water Works Assoc., 53: 75-87. Smith, G.H. and Ellis, M.M., 1963. Chromatographic analysis of gases from soils and vegetation related to geochemical prospecting for petroleum. Amer. Assoc. Pet. Geol. Bull., 47: 1897-1903. Smith, K.A., 1977. Gas chromatographic analysis of the soil atmosphere. Adv. Chromat., 15:197-231. Smith, R.C., 1978. Radon detection using thermoluminescent calorimeters in uranium exploration. In: Uranium Exploration Techniques. Saskatchewan Geol. Survey, Spec. Publ. 4, pp. 85-108. Smith, S.P. and Kennedy, B.M., 1982. The solubility of noble gases in water and NaC! brines. Geochim. Cosmochin. Acta., 47: 503-515. Sokolov, V.A., 1933. The gas survey as a method of prospecting for oil and gas formations. Technika, NGRI, No. 1, February. Sokolov, V.A., 1935. Summary of the experimental work of the gas survey. Neftyanoye Khoziastvo, 27:28-34 (in Russian). Sokolov, V.A., 1965. The processes of generation and migration of oil and gas. Nedra, Moscow. Sokolov, V.A., 1971a. The theoretical foundations of geochemical prospecting for petroleum and natural gas and the tendencies of its development. Can. Inst. Min. Metall., Spec. Vol. 11, pp. 544549. Sokolov, V.A., 1971b. Geochemistry of Naural Gases. Nedra, Moscow. Sokolov, V.A., Alexeyev, F.A., Bars, E.A., Geodekyan, A.A., Mogilevskii, G.A., Yurovsky, Y.M. and Yasenev, B.P., 1959. Investigations into direct oil detection methods. In: Proc. 5th World Petroleum Congress, Section I, Geology and Geophysics, pp. 667-687. Sokolov, V.A., Gcodekyan, A.A., Grigoryev, G.G., Krems, A. YA., Stroganov, V.A., Zorkin, L.M., Zcidelson, M.I. and Vainbaum, S.J.A., 1970. New methods of gas surveys, gas investigations of wells and some practical results. Can. Inst. Min. Metali., Spec. Vol. 11, pp. 538-543. Sokolov, V.A., Gcodekyan, A.A., Stroganov, V.A. and Petzenko, M. G., 1971. Theory and techniques involved in direct geochemical prospecting for oil and gas. Presented at Prague Symposium. Soli, G., 1954. Geomicrobiological prospecting. Amer. Assoc. Pet. Geol. Bull., 38: 2255-2558. Soli, G., 1957. Microorganisms and geochemical methods of oil prospection. Amer. Assoc. Pet. Geol. Bull., 41: 134-140. Solovov, A.P., 1985. Geochemical Methods for Mineral Deposit Prospecting. Nedra, Moscow, 294 pp. (in Russian). Sonnck, A.A., Mceks Jr., G.C., Klugman, C.L. and Seitz, C.A., 1965. A portable helium analyser. Rhode Is. Bur. Min., Invest. Report, 6600, 15 pp. Soonawala, N.M., 1976. Diffusion of radon-222 in overburden and its application to uranium exploration. Unpubl. Ph.D. thesis, McGill Univ., 187 pp. South African Atomic Energy Board, 1980. The ROAC system for uranium exploration. Inter-Science Research and Development Services, Ltd., PO Box 39265, Bramley 2018, South Africa, 8 pp. Stahl, W., 1975. Carbon isotope ratios of natural gases. Erdol und Kohle, 28:188-191 (in German). Stantso, V.V., 1974. Radon. Khimiia i zhizn, 10:17-21 (in Russian). Starik, I.E., 1959. Principles of Radiochemistry. AEC-TR-6314, 349 pp. Starik, I.E. and Melikova, O.S., 1957. Emanating power of minerals. AEC-TR-4498 (translated from Trans. V.G. Khlopin Radium Inst., 5: 206-226).
506
References
Starik, I.E., Ousnetsov, Yu.V., Petryayev, E.P. and Legin, V.K., 1966. Certain aspects of the geochemistry of radioactive isotopes. Chem. Earth's Crust, 1: 369-412. Starobinetz, I.S., 1976. The Role of Sedimentary Rocks in Distribution of Disseminated Hydrocarbons. Nedra, Moscow. Starobinetz, I.S., 1983. Development of scientific bases for geochemical methods in exploration for oil and gas. In: Geochemical Prospecting for Oil and Gas. Nauka, Acad. of Science of USSR, Moscow. Starobinetz, I.S. and Lomeyko, 1977. Extraction and analysis of gases during geochemical prospecting for oil and gas. Ministry of Geology, USSR, Nedra, Moscow. Stellmack, M.L. and Street, K.W., Jr., 1982. Permeation devices as gas standards. Amer. Lab., 14: 2533. Stephenson, M., Schwartz, W.J., Evenden, L.D. and Bird, G.A., 1992. Identification of deep groundwater discharge areas in Boggy Creek catchment, southeastern Manitoba, using excess helium. Can. J. Earth Sci., 29: 2640-2652. Stevens, D.N., Rouse, G.E. and De Voto, R.H., 1971. Radon-222 in soil gas: three uranium exploration case histories in the Western United States. In: Geochemical Exploration. Can. Inst. Min. Metall., Spec. Vol. 11, pp. 258-264. Stevens, R.K., O'Keefe, A.E. and Ortman, G.C., 1969. Absolute calibration of a flame photometric detector to volatile sulphur compounds at sub-parts per million levels. Env. Sci. Tech., 3: 652-655. Stoker, A.K. and Kruger, P., 1975. Radon measurements in geothermal systems. Report SGP-TR-4, Stanford Univ., I16 pp. Stroganov, V.A., 1969. Some recommendations of the employment of geochemical prospecting, Petroleum Gas Geol. Geophys., 7. Stuckless, J.S. and Nkomo, I.T., 1978. Uranium-lead isotope systematics in uraniferous alkali-rich granites from the Granite Mountains, Wyoming: implications for uranium source rocks. Econ. Geol., 73: 427-441. Stumm, W. and Morgan, J.J., 1970. Aquatic Chemistry. Wiley Interscience, New York, 583 pp. Sugisaki, R., 1978. Changing He/Ar and N2/Ar ratios of fault air may be earthquake precursors. Nature, 275:209-211. Sugisaki, R., 1981. Deep seated gas emission induced by the Earth tide: a basic observation for geochemical earthquake prediction. Science, 212:1264-1266. Sugisaki, R., 1987. Behaviour and origin of helium, neon, argon and nitrogen from active faults. J. Geophys. Res., 92: 12,523-12,530. Sugisaki, R., Takeda, H., Kawabe, I. and Miyazaki, H., 1982. Simplified chromatographic analysis of H2, Re, Ne, Ar, N2 and CH4 in subsurface gases for seismo-geochemical studies. Chem. Geol., 36: 217-226. Supelco, 1977. Analysis of Sulphur Gases: Trace Quantities. Supelco, Inc., Bellefonte, Pennsylvania, 4 pp. Swinnerton, J.W. and Lamontagne, R.A., 1974. Oceanic distribution of low molecular weight hydrocarbons - baseline measurements. Environ. Sci. Tech., 8: 657-663. Syromyatnikov, N.G., Ibrayev, R.A. and Mukashev, F.A., 1967. Interpretation of uranium anomalies in arid regions with the aid of 234U/238Uratio. Geokhimiya, 7: 834-841. Tackett, J.L., 1968. Theory application of gas chromatography in soil aeration research. Proc. Soil Sci. Soc. Am., 32: 346-50. Taiwani, P., Moore, W.S. and Chiang, J., 1980. Radon anomalies and micro-earthquakes at Lake Jocassee, South Carolina. J. Geophys. Res., 85: 3079-3088.
References
507
Tanner, A.B., 1964a. Radon migration in the ground: a review. In: J.A.S. Adams and W.M. Lowder (eds.), The Natural Radiation Environment. Univ. Chicago Press, pp. 161-190. Tanner, A.B., 1964b. Physical and chemical controls on distribution of radium-226 and radon-222 in groundwater near Great Salt Lake, Utah. In: J.A.S. Adams and W.M. Lowder (eds.), The Natural Radiation Environment. Univ. Chicago Press, pp. 253-276. Tanner, A.B., 1978. Radon migration in the ground: a supplementary review. US Geol. Survey, Open File Report, 78-1050, pp. 1-62. Tanner, A.B., 1980. Radon migration in the ground: a supplementary review. In: T.F. Gesell and W.M. Lowder (eds.), Natural Radiation Environment III. National Technical Information Service, US Dept. of Commerce, Springfield, Va., pp. 5-56. Taskayev, A.I., Ovchenkov, V.Ya and Aleksakhi'n, R.M., 1978. Forms of 226Ra in the horizons of soils with a high concentration of this isotope. Soviet Soil Sci., 2:45-51. Taylor, C.H., Kesler, S.E. and Cloke, P.L., 1982. Sulphur gases produced by the decomposition of sulphide minerals: application to metallic minerals exploration. J. Geochem. Explor., 17:165-168. Taylor, G.F. and Appleyard, E.C., 1983. Weathering of the zinc-lead lode,Dugald River, northwest Queensland: I, the gossan profile. J. Geochem. Explor., 18:87-110. Taylor, G.F. and Scott, K.M., 1983. Weathering of the zinc-lead lode, Dugald River, northwest Queensland: II, surface mineralogy and geochemistry. J. Geachem. Explor., 18:111-130. Taylor, G.F., Wilmshurst, J.R., Togashi, Y. and Andrew, A.S., 1984. Geochemical and mineralogical haloes about the Elura Zn-Pb-Ag ore body, western New South Wales. J. Geochem. Explor., 22: 265-290. Tedosco, S.A., 1995. Surface Geochemistry in Petroleum Exploration, Chapman and Hall. Telegina, Z.P. and Cherkinskaya, B.S., 1971. Possibility of the formation of unsaturated hydrocarbons during oxidation of saturated hydrocarbons by microorganisms. In: Results of Development and Testing of Direct Geochemical Methods of Finding Deposits of Oil and Gas, No. 10. Nedra, Moscow, pp. 61-64 (in Russian). Teng, T.L., 1980. Some recent studies on groundwater radon content as an earthquake precursor. J. Geophys. Res., 81: 3089-3099. Tcplitz, A.J. and Rodgers, J.K., 1935. Research Project DA-51-3(1). Gulf Research and Development Company, Achives Location 71-B23, Box 14D-34. Testury, M.H., 1996. Results of geoelectrochemical experimental investigations by means of the diffusion extraction of metals (MDE) and their significance in geophysical prospecting for kimberlite explosion pipes. VINITI, St. Petersburg State Mining Institute, Russia, 12 pp. (in Russian). Thompkins, R.W., 1982. Radiation in uranium mines (a manual of radon gas emission characteristics and techniques for estimating ventilating requirements). Can. Inst. Min. Metall. Bull., 75: (845) 149-156, (846) 62-67, (847) 98-104. Thompson, C.K., 1981. The airborne gas sensor and its role in exploration reconnaissance. In: B. M. Gottlieb (ed.), Unconventional Methods in Exploration for Petroleum and Natural Gas 11. Southern Methodist University Press, Dallas, pp. 240-248. Thompson, C.K. and Burson, K.R., 1984. Geochemical reconnaissance in petroleum exploration (abs). Presented at 187th Annual Meeting, Amer. Chem. Soc., St. Louis. Thompson, C.K., Saunders, D.F. and Burson, K.R., 1994. Model advanced for hydrocarbon microseepage, related alterations. Oil and Gas J., November, pp. 95-99. Thompson, R.R., 1971. Extraction of hydrocarbon gases from earth samples. Canadian Patent no. 861464.
508
References
Thompson, R.R., Duschatko, R.W. and Nash, A.J., 1974. Critical carbonate minerals in geochemical prospecting. US Patent no. 3,801,291. Thomber, M.R., 1975a. Supergene alteration of sulphides, I: a chemical model based on massive sulphide deposits at Kambalda, Western Australia. Chem. Geol., 15:1 - 1 4 . Thornber, M.R., 1975b. Supergene alteration of sulphides, II: a chemical study of the Kambalda nickel deposits. Chem. Geol., 15:117-144. Tikhomirov, V.V. and Tikhomirova, V.G., 1971. Factors controlling helium background contents in gas accumulation. Geokhimiya, 12:1503-1509. Tissot, B.P. and Welte, D.H., 1978. Petroleum Formation and Occurrence. Springer-Verlag, New York, 538 pp. Titayeva, N.A., 1967. Association of radium and uranium with peat. Geochem. lntemat., 4:1168-1174 (translated from Geokhimiya, 12:1493-1499). Tokarev, A.N. and Shcherbakov, A.V., 1956. Radiohydrogeology. AEC-TR-4100, pp. 16-164. Tomkins, R., 1990. Direct location technologies: a unified theory. Oil and Gas J., 24 Sept., pp.126134. Torgerson, T. and Clarke, W.B., 1978. Excess 4He in Teggau Lake: possibilities for a uranium ore body. Science, 199:769-781. Torgerson, T. and Clarke, W.B., 1985. Helium accumulation in groundwater: I, an evaluation of sources and the continental flux of crustal 4He in the Great Artesian Basin, Australia. Geochim. Cosmochim. Acta, 49:1211-1218. Torgerson, T. and lvey, G.N., 1985. Helium accumulation in groundwater: 1I, a model of the accumulation of the crustal 4He degassing flux. Geochim. Cosmochim. Acta, 49: 2445-2452. Townsend, T.E., 1984. The significance of iron oxides and clays in petroleum and mineral exploration. Int. Rept., Exxon Prod. Res. Co. Tracor, 1973. Model 270HA Atmospheric Sulphur Analyser, Operation Manual. Tracor Instr., Austin, Texas, 44 pp. Tracor, 1980a. Tracor 560 GC, Operation Manual. Tracor Instr., Austin, Texas, 77 pp. Tracor, 1980b. Tracor 700A Hall Electrolytic Conductivity Detector, Operations Manual. Tracor Instr., Austin, Texas, 70 pp. Tugarinov, A.I. and Osipov, Yu. G., 1974. Flow of helium through granite massifs and zones containing radioactive ores. Geochem. Intemat., 1 I: 981-986. Tunell, G., 1978. Mercury. In: K.H. Wedepohl (ed.), Handbook of Geochemistry, vol. 11. SpringerVerlag, New York. United Nations Scientific Committee, 1977. Natural sources of radiation. In: Sources and Effects of lonising Radiation. Report to the General Assembly, United Nations, New York, pp. 35-114. Uvarov, N.N., 1981. Equipment of polarographic logging for composition investigation of underground water. Transactions of Leningrad State Mining Institute, 89: ! 15-122 (in Russian). Uvarov, N.N., 1984. Peculiarities of the pulse modes in the application of backgroundless polarography to natural water. Transactions of Leningrad State Mining Institute, 101:99-104 (in Russian). Van de Laan, J.E., Rosengreen, A. and Altpeter, T., 1985. Remote and selective detection of natural gas leaks using a frequency mixed CO2 laser. Van den Boom, G., 1987. Helium distribution pattern of measured and corrected data around the Eidingen oil field, NW Germany. J. Geophys. Res., 92: 12,547-12,555.
References
509
Van den Boom, G. and Poppelbaum, M., 1980. Determination of the presence of primary mercury dispersion around the volcanogenic sulphide ore deposit "Woodlawn", New South Wales, Australia. Geol. Jb., D37: 15-27. Van den Boom, G. and Poppelbaum, M., 1982. Volatile compounds in the soil as indicators in geochemical exploration for concealed ore deposits: application of a gas chromatographic headspace method. Geol. Jb., D58: 3-30. Vasilieva, V.I. and Voroshilov, N.A., 1995. Estimation of oil-prospective areas by means of geoelectrochemical methods. Geophysics, N2:29-36 (in Russian). Verbanac, B. and Dunia, P., 1982. Hydrocarbon ratios in gas mixtures as indices of reservoir saturation and producing capacity. Geologija i Geofizika, 33: 185-192. Vilenskiy, V.D. and Anikiyev, V.V., 1974. Lead-210 and radium-226 in the bottom deposits of freshwater bodies. Geochem. lntemat., 11: 874-877. Vinogradov, A.P., 1959. The Geochemistry of Rare and Dispersed Chemical Elements in Soils (2nd edn.). Consultants Bureau Inc., New York, 209 pp. Vogel, J.C., Talma, A.S. and Heaton, T.H.E., 1980. Gaseous nitrogen as evidence for denitrification in groundwater. J. Hydrol., 50:191-200. Vogel, T.M., Oremland, R.S. and Kvenolden, K.A., 1982. Low temperature formation of hydrocarbon gases in San Francisco Bay sediment (California, USA). Chem. Geol., 37: 289-298. Voronov, A.N., Prasolov, E.M. and Tikhomirov, V.V., 1974. The relation between radiogenic argon and helium Isotopes in gas pools. Geochem. lntemat., 11:1311-1322. Voronov, A.N., Tikhomirov, V.V. and Tikhomirova, V.G., 1969. Specific capacity of helium supply into porous spaces of sedimentary strata. Geokhimiya, 4: 505-507. Voronov, A.N., Tikhomirov, V.V. and Yakutseni, V.P., 1981. Patterns of Iocalisation of helium deposits in the sedimentary cover. Dokl. Akad. Nauk SSSR, 257:176-179. Voytov, G.I., Lebedev, V.S., Blokhina, G.G. and Cherevichnaya, L.F., 1975. Chemical and isotopic composition of subsoil and marsh gases of the Kola tundras. Doklady Akad. Nauk. SSSR., 224: 186-188.
Wakita, H., 1978. Helium spots: caused by diapiric magma from the upper mantle. Science, 200: 430432. Wakita, H., ! 980. Hydrogen release: new indicator of fault activity. Science, 210:188-190. Wakita, H. and Sano, Y., 1983. 3He/4He ratios in CH4-rich natural gas suggest magmatic origin. Nature, 305: 792-794. Wakita, H., Sano, Y. and Mizoue, M., 1987. High 3He emanation and seismic swarms observed in a nonvoicanic forears region. J. Geophys. Res., 92: 12,539-12,546. Walker, B.G., 1969. Springs of deep seated origin in Tanzania. In: G. Kacura (ed.), Mineral and Thermal Waters of the World, B - Overseas Countries. Proc. XXIII Geol. Congr., 19:171-180. Wang, X., 1994. Abiogenic natural gases. In: Theory and Application of Gas Genetics. Scientific Press, Beijing, pp. 15-24 (in Chinese). Warner, M.A., 1982. The source and time of generation of hydrocarbons in the Fossil Basin, Western Wyoming Thrust Belt. In: Geologic Studies of the Cordilleran Thrust Belt, pp. 805-815. Warren, H.V., Delavault, R.E. and Barakso, J., 1966. Some observations on the geochemistry of mercury as applied to prospecting. Econ. Geol., 61 : 1010-1028. Warren, R.K., 1977. Recent advances in uranium exploration with electronic alpha cups. Geophysics, 42: 982-989. Webber, G.R., 1975. Efficacy of electrochemical mechanisms for ion transport in the formation of geochemical anomalies. J. Geochem. Explor., 4:231-233.
510
References
Weinrich, K.J., 1985. Mineralisation breccia pipes in northern Arizona. Econ. Geol., 80:1722-1735. Weismann, T.J., 1980. Developments in geochemistry and their contribution to hydrocarbon exploration. Proc. 10th World Petroleum Cong., Bucharest, Romania, Vol. 2, pp. 369-386. Weiss, R.F., 1971. Solubility of helium and neon in water and seawater. J. Chem. Eng. Data, 16: 235241. Wen, B., 1997a. Physico-chemical modelling of the natural ionic flotation of metals by gas bubbles in porous rocks. In: Abstracts of the All-Russian Scientific-Practical Conference of Young Scientists and Specialists, "Geophysics-97". VIRG Rudgeofizika, St.Petersburg, pp. 44-45 (in Russian). Wen, B., 1997b. Solution of the physico-mathematical model of the "stream" halo of dispersion and comparison with results of physico-chemical modelling. In: Abstracts of the Annual Scientific Conference of Young Scientists, "Mineral Resources of Russia and their Exploitation". St.Petersburg State Mining Institute, St. Petersburg, p. 4 (in Russian). Wennervirta, H. and Kauranen, P., 1960. Radon measurement in uranium prospecting. Bull. de la Commission geologique de Finlande, 188: 24-40. West, P.W. and Gaeke, G.C., 1956. Fixation of sulphur dioxide as disulfito-mercurate (II) and subsequent colorimetric estimation. Anal. Chem., 28:1816-1819. Whelan, J.K., 1979. C1-C7 hydrocarbons from IPOD holes 397 and 397A. In: Initial Reports of the Deep Sea Drilling Project. US Govt Printing Office, Washington DC., Vol. 47A, pp. 531-539. Whelan, J.K., Hunt, J.M. and Berman, J., 1980. Volatile C1-C7 organic compounds in surface sediments from Walvis Bay. Geochim. Cosmochim. Acta, 44: 1767. White, D.E., Hem, J.D. and Waring, G.A., 1963. Chemical composition of subsurface waters. In: Data of Geochemistry (6th edn.). US Geol. Survey, Prof. Paper 440-F, 67 pp. Wilkening, M.H., 1980. Radon transport processes below the earth's surface. In: T.F. Gesell and W.M. Lowder (eds.), Natural Radiation Environment III. National Technical Information Service, US Dept. of Commerce, Springfield, Va., pp. 9 I-104. Wilkening, M.H., Clements, W.E. and Stanley, D., 1972. Radon-222 flux in widely separated regions. In: J.A.S. Adams, W.M. Lowder and T.F. Gcscli (cds.)., The Natural Radiation Environment II. National Technical Information Service, US Dept. of Commerce, Springfield, Va., pp. 717-730. Williams, B.G., 1985. Landsat and geochemical investigation of the Permian Basin., Regan and Crockett Counties, Texas. Unpubl. MS thesis, Univ. Texas. Williams, J.C., Mousseau, R.J. and Weismann, T.J., 1981. Correlation of well gas analysis with hydrocarbon seeps (abs). Proc. Meeting American Chemical Society, Atlanta, GA, March 29April 3. Williston, S.H., 1964. The mercury halo method of exploration. Eng. Min. J., 165: 98-101. Wilmshurst, J.R., 1980. The Elura gossan: Part 1, geochemistry and macroscopic detail, DDH 25, DDH 28 and individual samples. Unpubl. Restricted Inv. Report, 1177R, Institute of Earth Resources, CSIRO. Wilmshurst, J.R. and Ryall, W.R., 1980. SIROSORB: a collector for use in the determination of mercury in geochemical samples. J. Geochem. Explor., 13: 1-8. Wollny, E., 1881. Untersuchegen uber den Eifluss der physikalischen Eigenscha~en des Bodens auf dessen Gehalt and Kohlensaure. Forsch. Geb. Agr. Phys., 4:1-24 and 9: 165-194. Wu, l.J. and Mahaffey, E.J., 1978. Mercury in soils geochemistry on massive sulphide deposits in Arizona. In: J.R. Watterson and P.K. Theobald (eds.), Geochemical Exploration 1978. Assoc. Explor. Geochem., Rexdale, Ontario, pp. 201-208. Wyatt, D.E., Richers, D.M. and Pirkle, R.J., 1995. Barometric pumping effects on soil gas studies for geologic and environmental characterization. Environmental Geol., 25: 243-250.
References
51 1
Yamaguchi, M., Howard, F.D., Hughes, D.L. and Flockes, W.H., 1962. An improved technique for sampling and analysis of soil atmospheres. Proc. Soil Sci. Soc. Am., 26:512-13. Yanitskii, I.N., 1979. Surveying for Helium. Nedra, Moscow, 91 pp. Yasko, V.G., 1981. Radon in spontaneously-evolving gases of hot springs of the Baykal ri~ zone. Dokl. Akad. Nauk. SSSR, 245: 124-126. Zaikowski, A. and Roberts, E.H., 1981. Importance of resolution for helium detectors used in uranium exploration. US Dept. of Energy, Open File Report, GJBX-38(81), 19 pp. Zarella, W.M., Mousseau, R.J., Coggeshall, N.D., Norris, M. S. and Schrayer, G.J., 1967. Analysis and significance of hydrocarbons in subsurface brines. Geochim. Cosmochim. Acta, 31: 1155. Zhu Bingqiu, Zhang Jinmao, Zhu Lixin and Zheng Yaxin, 1986. Mercury, arsenic, antimony, bismuth and boron as geochemical indicators for geothermal areas. J. Geochem. Explor., 25: 379-388. Zinger, A.S., Fiedorov, D.U., Talnova, L.D., Sorkokina, T.A., Navrotski, O,K, Timofeev, G.N., Fiedorova, S.V., Sidorov, I.N., 1983. Complex of geological and hydrogeochemical methods in prospecting for subtle hydrocarbon traps. In: Integration of Prospecting Methods in Exploration for Oil and Gas. Academy of Science USSR, Moscow. Zonghua, W. and Yangfen, J., 1981. A mercury vapour survey in an area of thick transported overburden near Shanghai, China. J. Geochem. Explor., 15: 77-92. Zorkin, L.M., 1977a. Geochemical prospecting for oil and gas. Soviet Geol., 11: 127-142. Zorkin, L.M., 1977b. The effects of earthquakes on hydrocarbon gas concentrations in surface rocks (as observed on Sakhalin Island). Doklady Navd. Navk USSR, 235: 476-478. Zorkin, L.M., Dadashev, F.G., Vayramova, G.A., Guliyev, I.S. and Feyzulayev, A.A., 1982. Geochemical prospecting for oil and gas in Azerbaijan. Geol. Nefli i Gasa, 4: 42-45. Zorkin, L.M., Zabairaevi, S.L., Karus, E.V., Kilmetov, K.Kh., 1977. Experience of geochemical prospecting in petroleum and gas deposits in the seismically active zone of Sakhalin Island, IZU. UZSSh Uchebn. Zaved, Geol. Razved, 20:52-62 (in Russian). Zverev, V.L., Tokarev, A.N., Tyminskii, V.G. and Shvets, V.M., 1980. Radioisotopic Geochemistry. Nedra, Moscow, 204 pp. (in Russian).
This Page Intentionally Left Blank
513
AUTHOR INDEX*
Abaturova, 53,471 Abrams, 477, 485, 503 Abu-Ali, 490 Acharya, 12, 471 Adams, 261,267, 354, 471,473, 481, 490, 497, 499, 507, 510
Adkinson, 392, 471 Aftabi, 404, 405,471 Agababov, 481 Agtsuma, 488 Akimov, 490 Akimova, 4 73 Albright, 498 Aldeman, 49 ! Alekhina, 4 76 Aleksakhi'n, 507 Alekseev, 17, 353,361,362, 375,377, 378, 471,502 Alexander, 305,477, 486 Alexeyev, 505 Alexin, 497 Allen, 153, 386, 4 71 Alley, 237, 477 Alpers, 469, 4 7/ Alshaieb, 239, 492 Altpeter, 158, 508 Anderson, 151,233,238, 244, 460, 472, 495
Andersson, 398, 472
Ando, 488 Andrew, 507 Andrews, 306, 331,332, 360, 371,374, 472, 474
Anikiyev, 509 Antonov, 144, 146, 472 Antropov, 155, 157, 158, 472 Antropova, 17, 19, 51, 4 72, 502 Anufriyev, 489, 493 Appleyard, 406, 507 Archambault, 494 Armstrong, 4 73 Arp, 242, 472 Arthur, 238, 472 Ashley, 477 Asikainen, 360, 366, 472 Asimov, 4 72 Atherden, 112, 484 Austin, 370, 373,472, 508 Ayres, 486 Azzaria, 404, 405,471 Bailey, 460, 472 Bajc, 81, 85, 112, 472, 485 Baker, 500 Bakhtin, 502 Ball, 459, 460, 469, 472, 473, 494 Bamesberger, 267, 4 73, 481 Bammel, 240, 4 73
*Pagenumbersin italicsreferto the referencelist,pp. 471-511
514
Banwart, 251,260, 4 73, 4 75 Barakso, 509 Baranov, 353,360, 361,362, 373,375, 378, 381,385,473 Barker, 360, 4 72, 4 73, 503 Barnes, 84, 95, 96, 99, 239, 397, 473 Barretto, 371,372, 373,374, 473, 505 Barringer, 158, 220, 473 Bars, 505 Barsukov, 154, 349, 350, 473 Bartashervich, 4 72 Bartlett, 93, 95,473 Bashorin, 332, 346, 4 73, 4 75 Bass Becking, 91, 96, 98,474 Batard, 343,345, 4 74 Bateman, 452, 4 74 Batulin, 483 Baubron, 4 74 Baver, 11,453, 4 74 Bayliss, 4 78 Beales, 102, 105, 504 Beauchamp, 460, 472 Beavan, 477 Becker, 4 79 Been, 475, 499 Beg, 4 76, 4 77 Behounek, 353, 4 74 Behrens, 4 72 Bell, 386, 476, 486 Bensman, 46, 474 Berman, 510 Bernard, 140, 168, 474 Berquist, 313,493 Bhatia, 4 78 Bigelow, 13,460, 486, 493 Billard, 498 Bimic, 473 Binns, 420, 501 Birchard, 378, 474 Bird, 506 Birnie, 240, 492 Bishop, 4 77
Author index
Bjomasson, 488 Blackmer, 460, 4 74 Bland, 365,491 Blokhina, 509 Bobrov, 490 Bockris, 87, 474 Bogoslovskaya, 379, 474 Bolviken, 82, 85, 98, 99, 100, 102, 103, 105, 108, 112, 114, 474 Bonilla, 4 71 Bonotto, 375, 4 74 Boone, 306, 338, 4 74 Borodzich, 154, 4 74 Borowski, 497 Bottomer, 403,474 Bottomley, 331,332, 338, 474 Bourne, 492 Bowles, 330, 332, 333, 351,475, 499 Boyle, 366, 397, 480 Boynton, 258,475 Brabec, 4 71 Brace, 4 71 Bracewell, 460, 500 Bradley, 496 Brameld, 494 Breck, 258, 475 Bremner, 251,260, 460, 4 73, 4 74, 4 75 Bridge, 453, 4 79 Brock, 13,250, 4 72, 4 75 Brody, 263, 4 75 Brooks, 180, 240, 4 74, 4 75 Buehler, 251,484 Bugorkov, 490 Bukova, 137, 138, 475 Bulashevich, 332,347, 475 Bullock, 4 71 Bunting, 457, 460, 475 Burford, 457, 4 75 Burr, 92, 94, 107, 113, 4 75 Burson, 157, 475, 503, 507 Burtell, 162, 163, 4 75, 488 Busch, 4 74
Author index
Buseck, 395,398,428, 430, 433,494, 497
Busigin, 4 75 Butt, 309, 310, 311,312, 313, 314, 317, 319, 320, 324, 325,326, 327, 328, 329, 330, 331,334, 335,336, 337, 341,348, 351,475, 476, 484, 485 Byman, 250, 251,253, 476 Cadigan, 359, 360, 481 Cameron, 4 76 Campbell, 457, 460, 475, 480 Caneer, 4 76 Canney, 477 Card, 386, 476 Carey, 4 75 Carr, 395,403,406, 407,408,409, 410, 411,412,414, 415,416, 419, 421, 423,424,425,432,433,434, 476 Carter, 4 76 Chalov, 349, 350, 476 Chaney, 263, 4 75 Chang, 477 Charbonneau, 365,476 Chatterjee, 480 Cheng, 220, 223,224, 230, 476, 501 Cherdyntsev, 360, 372, 476 Cherevichnaya, 509 Cherkinskaya, 507 Chernyshev, 504 Cherry, 83,482 Chiang, 506 Childers, 4 72 Chork, 82, 85, 112,484 Christ, 95,483 Chung, 349, 476 Clar, 501 Clark, 83, 85, 96, 99, 102, 105, 110, 112, 117, 118, 313,329, 473, 476 Clarke, 307, 312, 316, 330, 332, 336, 337, 474, 476, 477, 491,508 Clayton, 4 78
515 Clements, 354, 375,377, 380, 477, 510 Cloke, 50 7 Cochran, 358, 477 Coetzee, 365,491 Coggeshall, 511 Coleman, 137, 139, 140, 477 Collins, 162, 477 Conel, 237, 477 Cook, 338, 477 Cooper, 464, 4 77, 4 78 Copping, 494 Corazza, 315,343,346, 477 Corbett, 413, 4 77 Cowart, 4 72 Cox, 344, 349, 409, 477 Craig, 307, 343,345,476, 477, 492 Crawford, 241,477, 495 Crees, 4 79 Crisp, 170, 216, 497 Crosta, 241,496 Crow, 4 73 Cucchi, 493 Currie, 11,453,477 Curtis, 409, 4 77, 49 / Cwick, 244, 477 Czarnecki, 321,392, 496 Da Silva, 312, 313,480 Dadashev, 452, 457,462, 478, 511 Dalziel, 142, 239, 478, 479 Dando, 239, 478 Dass, 480 Datta, 306, 313, 332, 347, 478 Davidson, 233,478, 487, 490, 493, 498 Davis, 138, 140, 233,478 Davy, 250, 365, 478 De Camargo, 258, 478 De Jong, 319, 4 78 De Souza, 260, 4 78 De Voto, 506 Debnam, 141, 176, 4 78 Delavault, 509
516
Delivron, 490 Delwiche, 357, 367, 373,375,478 Dennis, 313, 4 78, 484 Denton, 313,348, 478, 482, 486, 499, 500
Derjaguin, 23,478 Dettman, 4 71 Devine, 178, 220, 222, 478 Dickinson, 191,209, 4 78, 4 79 Dietrich, 494 Dilbert, 250, 283,284, 285, 311,326, 328, 486 Distler, 504 Dixon, 453,479 Dobrin, 94, 4 79 Dobrolovskaya, 488 Dokokin, 489 Dolly, 199, 479 Donovan, 141, 142, 187, 237, 239, 339, 4 78, 4 79, 500
Dortman, 22, 23,479 Douglas, 138, 489 Dowdell, 457, 479 Dowling, 492 Drever, 93, 96, 329, 330, 479, 481 Driscoll, 266, 479 Drozd, 134, 141,144, 149, 154, 155, 156, 161,164, 175, 183, 189, 194, 195,213, 311,341,348,350, 479, 488, 497
Druzhinin, 4 75 Duchscherer, 141,234, 235,238, 239, 479
Duddridge, 309, 315,348, 479 Dukhanin, 18, 30, 31, 52,471,479, 498, 499, 502
Dunia, 167, 509 Dunlop, 484 Dunn, 342, 480 Durham, 480 Durrance, 310, 311, 312, 323,344, 4 79, 484
A uthor index
Duschatko, 508 Dvomikov, 395 Dvomikow, 4 79 Dyck, 258, 309, 312, 313,329, 330, 332, 336, 338, 342, 360, 361,363, 365,366, 367, 380, 381,384, 385, 386, 393,480, 488, 505 Earle, 329, 330, 481 Elinson, 154, 249, 286, 287, 452, 481, 504
Ellis, 138, 188, 505 Elworthy, 360, 481,502 Emerson, 306, 340, 486 Enhalt, 6, 481 Epstein, 504 Erdman, 499 Eremeev, 154, 310, 312, 316, 331,346, 347, 481,484 Escobar, 4 77 Evans, 360, 481,490 Evenden, 506 Everett, 244, 481 Everitt, 477 Fairbridge, 238, 4 78, 481 Farrar, 496 Farrow, 485 Farwell, 261,265, 4 71, 481 Faul, 308, 481 Favorskaya, 23,481 Feder, 240, 481 Fei, 218, 481 Felmlee, 359, 360, 481 Fenn, 186, 481 Ferguson, 141,142, 237, 481,482, 501 Feyzulayev, 511 Fiedorov, 511 Fiedorova, 511 Filimonova, 504 Finlay, 459, 500 Fisher, 321,362, 388, 393,483
Author index
Fleischer, 378, 388, 482, 494 Flockes, 511 Forbes, 398, 399, 482 Ford, 4 76, 4 79 Forrest, 267, 482 Foster, 199, 482 Fournier, 345,493 Frank, 140, 482 Fredericks, 482 Freeze, 83,482 Frenklikh, 491 Frey, 489, 503 Fridman, 22, 23, 24, 249, 452, 482, 486, 490, 496
Friedlander, 375,482 Friedman, 313,479, 482, 499, 500 Friedrich, 411,412, 482, 491 Frischknecht, 101, 102,489 Fritz, 4 72 Froidevaux, 233,483 Fursov, 153, 154, 395,483 Gabelman, 377, 483 Gaeke, 249, 267, 510 Gage, 481 Galbraith, 480 Gale, 4 72 Galimov, 136, 483 Gandhi, 4 77 Gangloff, 498 Garanin, 4 73 Garcia, 501 Gardner, 4 74 Garrels, 95,483 Gascoyne, 347, 483 Gaucher, 388,483 Gemmell, 480 Genkin, 504 Geodekyan, 505 Gerling, 308, 483 Germanov, 362, 483, 495
517
Gesell, 471,473, 482, 483, 494, 507, 510
Giles, 4 72 Giletti, 371,372, 373,483 Gingrich, 321,388, 393,483 Glasstone, 113,483 Gleason, 4 79 Glebovskaya, 249, 286, 452, 457, 463, 464, 483 Glebovskii, 249, 286, 452, 463,464, 483
Gleezen, 189, 483 Gluck, 481 Gluechauf, 306, 483 Go, 23,484 Goetz, 233,242, 484 Goldberg, 17, 38, 42, 44, 471,472, 474, 502
Gole, 309, 310, 312, 313,314, 318, 319, 320, 324, 325,327, 328, 329, 330, 331,333,334, 335,336, 337, 341,348, 351,475, 476, 484, 485 Goleva, 28, 484 Golubev, 310, 319, 331,484 Goodfellow, 402,484 Goriainov, 24, 484 Gottfried, 366, 497 Gottschalk, 251,484 Gournay, 156, 484 Govett, 82, 85, 102, 103, 112, 113,484 Gracheva, 4 71 Grainger, 4 79 Gramberg, 484 Grammakov, 353, 4 71, 4 74, 484 Granger, 251,252,484 Grant, 158, 484 Greenland, 150, 484 Greenwood, 456, 484 Gregory, 310, 311, 312, 323,344, 484 Griffith, 491 Grigorovich, 4 74 Grohmann, 4 78
518
Grune, 486, 505 Guliyev, 511 Gupta, 4 78 Guseinov, 4 78 Gutsalo, 343,345,485 Haissinsky, 372, 485 Hale, 3,250, 275,286, 288,457, 485, 492, 494, 496, 501 Hall, 478, 479, 482, 507, 508 Hamdi, 456, 485
Hamilton, 81, 84, 85, 86, 87, 89, 94, 99, 102, 104, 105, 109, 110, 112, 114, 471,485
Hansink, 366, 485 Hardin, 494 Harms, 250, 258, 486 Harrell, 484 Hart, 314, 476, 482, 485, 498 Hatch, 435,485 Hatuda, 388, 485 Hauksson, 353,485 Hawkes, 395, 412,482, 485, 500 Haymon, 492 Heatherington, 498 Heaton, 306, 313,332,485, 509 Heemstra, 159, 485 Hegge, 419, 485 Hem, 510 Henderson, 4, 486 Heppell, 313, 4 78 Herzig, 482 Hewitt, 4 79 Hextall, 4 73 Heyrovsky, 73,486 Hickey, 159, 486, 490 Higgins, 386, 486, 505 Hinkle, 12,250, 254, 258, 259, 264, 269, 271,272, 273,274, 275,276, 280, 281,283,284, 285, 310, 311, 315,318,326, 328, 343,344, 469, 486
Author index
Hitarov, 24, 486 Hobbs, 480 Hodgson, 475, 488 Hofheinz, 471 Holland, 306, 340, 486 Hoppe, 305,486 Horbert, 385,488 Horibe, 477 Horstman, 500 Horvitz, 134, 140, 142, 159, 176, 183, 185,213,218, 219, 221,486, 487 Hoshizaki, 500 Howard, 511 Howarth, 494, 501 Howe, 477 Hughes, 511 Hunt, 134, 151,173, 175, 184, 216, 233,236, 239, 487, 510 Hunter, 343, 345,488 IAEA, 354, 384, 472, 473, 483, 485, 488, 489, 495, 501,503, 505 Ibrayev, 506 Ignatiev, 481
Israel, 385,488 Ivanov, 310, 488, 490 Ivanova, 4 7/ Ivey, 306, 333,508 Iwashita, 499 Jaacks, 12, 490 Jackson, 81, 85, 112, 488 James, 93, 95,395,473, 488 Janezic, 134, 137, 140, 149, 488 Jaramillo, 4 79 Jayasurya, 4 78 Jefferson, 485 Jeter, 309, 328, 376, 488 Jiang, 504 Johnson, 486, 492, 503 Jonasson, 312, 363,381,397, 398, 399, 435,480, 484, 488
Author index
Jones, 134, 141,144, 146, 148, 149, 152, 154, 155, 161,164, 175, 183, 185, 189, 194, 195, 197, 198, 200, 207, 209213,220, 222, 311, 341, 348, 350, 475, 479, 481,484, 488, 493, 496, 497, 500
Kagel, 481 Kahler, 343, 488 Kahlos, 360, 472 Kahma, 13,250, 267, 489 Kalinko, 23, 489 Kallio, 137, 494 Kaltenback, 499 Kamenskiy, 306, 489, 493 Kancoka, 502 Kantor, 250, 258, 271,272, 273,274, 275,276, 486 Kapitanov, 387, 489 Kapkov, 309, 328, 353,360, 361,362, 370, 375,376, 381,385,393,495 Kaplan, 4 74 Karar, 47, 489 Karim, 159, 489 Kartashov, 347, 4 75 Kartsev, 137, 140, 141,142, 154, 159, 185,218, 219, 489 Karus, 511 Katz, 166, 489 Kauranen, 381,387, 393, 510 Kawabe, 506 Kay, 4 72 Keller, 101, 102, 489 Kennedy, 307, 482, 505 Kennicott, 4 75 Kent, 343,345,489 Keogh, 4 77 Kesler, 507 Keyssner, 482 Khabarin, 490 Khayretdinov, 452,457, 489 Khodakovsky, 397, 489
519
Khristianov, 360, 489 Khristianova, 4 73 Khvalorski, 483 Kiefer, 360, 361,503 Kilbum, 311,486 Killeen, 497 Kilmetov, 511 Kim, 138, 489 King, 155,349, 350, 360, 378, 393,489 Kip, 481 Kirikov, 4 74 Kirilenko, 496 Kiriukhin, 23,490 Kisslinger, 4 71 Kleinevoss, 396, 490 Klinkhammer, 492 Klochkov, 77, 490 Klugman, 505 Klusman, 12, 83, 134, 159, 216, 222, 223,240, 486, 490 Knopoff, 4 71 Kobayashi, 502 Koger, 244, 476 Komarov, 504 Kononov, 343,490 Korchuganov, 360, 489 Korenov, 495 Komer, 354, 374, 375, 381,393, 490, 500
Korostin, 17, 60, 490, 502 Korotkov, 490 Korsch, 504 Kovach, 380, 381,490 Kovacs, 491 Kozlova, 24, 503 Kramer, 378, 379, 380, 385,490 Krat, 18, 30, 490 Krauskopf, 95,397, 490 Kravtsov, 249, 347, 452, 486, 490 Krchmar, 18, 30, 490 Kristiansson, 12, 18, 30, 378, 491,492 Krivodubsky, 4 74
520 Kromer, 430, 491 Kroptova, 490 Krouse, 493 Kruger, 359, 506 Kugler, 312, 313,316, 329, 330, 476, 491
Kulikova, 452,457, 461,491 Kulp, 371,372, 373,483 Kuta, 73,486 Kuznetsov, 361,491 Kvenolden, 509 Laffoley, 4 73 Lagunova, 345,491 Laier, 4 78 Lambe, 494 Lamontagne, 140, 506 Lane, 4 78 Lang, 92, 242,243, 366, 477, 491 Larin, 24, 491 Larson, 504 Laubmeyer, 140, 185, 213,491 Lawson, 4 72 Learned, 494 Lebedev, 509 Lee, 306, 332,472, 492 Legin, 506 Lenhoff, 487 Lester, 360, 362, 393,491 Levinson, 361,365,491 Levorsen, 182, 491 Lewis, 113,484 Leythaeuser, 168, 491,503 Liard, 480 Libby, 378,474 Likes, 482, 494 Lilburn, 239, 492 Lin Pai, 475 Lindberg, 499 Link, 133, 182, 214, 492 Lisitsin, 483 Lisitzyn, 491
Author index
Logn, 82, 85, 98, 99, 100, 403, 102, 105, 108, 112, 114, 474 Lohman, 471 Lomeyko, 506 Lomonosov, 307, 343,345,492 Loudon, 408, 492 Lovell, 11, 83, 250, 258, 259, 260, 262, 269, 273,276, 277, 278,279, 452, 457, 464, 465,466, 467, 492, 494 Lowder, 354, 471,473, 482, 483, 490, 494, 497, 499, 507, 510
Lucas, 386, 492 Lundell, 500 Lupton, 307, 477, 492 Macdonald, 492 MacElvain, 217, 218, 221,238, 492 Macias, 504 Maciolek, 500 Magro, 4 77 Mahaffey, 395, 412, 5 I0 Makkaveev, 362, 492 Malhotra, 237, 492 Malinovskii, 504 Maliotis, 482 Malley, 490 Malmqvist, 12, 18, 30, 378, 491,492 Malone, 400, 493 Mamuro, 364, 365,494 Mamyrin, 155,345,483, 489, 490, 492, 493
Manwaring, 133, 142, 238,239, 497 Marce, 4 74 Marine, 306, 331,478, 493 Marrs, 500 Marshall, 4 79 Martin, 313,488, 493 Mashianov, 18, 496 Mashijanov, 499 Matsubayashi, 495 Matsui, 4 78
Author index
Matthews, 142, 143, 144, 149, 161, 163, 183, 185, 187, 191,209, 479, 493, 500
Mausel, 477 Mazor, 331,345,493 McAuliffe, 149, 493 McCarthy, 13, 83,250, 285,333,395, 428,430, 460, 469, 493, 494, 499 McClenaghan, 85, 112, 114, 485 McCorkell, 387, 494 McCoy, 240, 494, 503 McCrossan, 134, 176, 183,494 McDermott, 177, 183, 185,238, 494 McFadden, 460, 494 McKay, 493 McKenna, 137, 494 McNerney, 395,398,430, 433,494 Meares, 499 Medovii, 488 Meeks, 505 Meents, 477 Megumi, 364, 365,494 Meilleur, 480 Melikova, 369, 370, 372, 505 Melvin, 154, 494, 504 Mendenhall, 504 Menzies, 158, 484 Merin, 237, 238,503 Meyer, 242, 249, 494 Micheals, 500 Michel, 485, 489 Milan, 488 Miller, 482, 500 Mills, 186, 494 Mirabel, 498 Miyazaki, 506 Mizoue, 509 Moberg, 481 Mogilevskii, 472, 489 Mogilevsky, 505 Mogro-Campero, 378, 388, 482, 494 Monteith, 453,494
521
Moon, 250, 286, 288, 485 Mooney, 92, 98, 103, 105, 104, 112, 155,494, 502 Moore, 233,244, 474, 485, 489, 495, 506
Morgan, 91,506 Morganti, 484 Morozova, 4 73 Morrison, 305,384, 388, 495 Morse, 365,384, 495 Moses, 497 Mosher, 428, 501 Mousseau, 134, 149, 170, 495, 51 O, 511 Mueschke, 494 Murray, 475, 499 Murricane, 480 Mwenifumbo, 497 Mysichenko, 4 72 Nagao, 345,495 Nakamura, 502 Nash, 365,495, 508 Naumov, 361,362, 495 Navrotski, 511 Nerken, 460, 495 Newman, 267, 482 Newton, 460, 495 Nicholas, 493 Nicholson, 4 73 Nikiforova, 365,495 Nikonov, 166, 167, 188, 194, 195,471, 495
Nilsson, 13,250, 495 Nkomo, 366, 506 Normark, 492 Norris, 511 Notsu, 502 Novikov, 309, 328, 353,360, 361,362, 370, 375,376, 381,385,393,495 Nuampurkar, 4 78 Nuutilainen, 112,496
522 Oakes, 3,496 Oehler, 235,239, 496 O'Hara, 4 78 Okabe, 153,496 Okada, 488 O'Keefe, 268, 496, 506 Oliveria, 241,496 Olivier, 306, 496 O'Neil, 4 73 Onishuk, 479 Openshaw, 428, 496 Oremland, 509 Orgel, 4 72 Orlov, 250, 496 Orlova, 4 73 Oro, 4 72 Ortenberg, 486 Ortman, 268, 496, 506 Osipov, 309, 347, 508 Osmond, 4 72 Ostrihansky, 365,496 Ott, 435,485 Otton, 313,499 Ousnetsov, 506 Ovchinnikov, 154, 310, 332,347, 349. 462,496, 507 Owen, 156, 472, 496 Ozerov, 496 Ozerova, 18, 395,399, 400, 404, 489, 496, 504
Pacer, 321,392, 496 Pack, 4 71 Panchenkov, 34, 497 Panov, 378,497 Parasnis, 91, 92, 94, 113,497 Park, 471 Parker, 4 72 Parslow, 491 Partington, 353,497 Pashkov, 481 Patton, 133, 142, 238,239, 497
Author index
Paul, 319, 4 78 Paull, 239, 497 Pavlov, 486, 489, 490 Pazdersky, 4 79 Peachey, 4 73 Peacock, 386, 497 Pearson, 497 Pelchat, 313,480 Penfield, 481 Perellman, 504 Peters, 249, 494, 497 Petrov, 395,452, 479, 482 Petryayev, 506 Petzel, 244, 481 Petzenko, 505 Peuraniemi, 112, 496 Pflug, 108, 114, 497 Phair, 366, 497 Phelps, 428, 497 Phillips, 4 75 Philp, 170, 216, 497 Pieri, 4 77 Pilot, 501 Pine, 305,484, 495, 501 Piper, 4 73 Pirkle, 134, 149, 151,155,497, 500, 510
Pirson, 110, 111,112, 141,159, 185, 235,497 Pixler, 149, 168, 498 Platfker, 4 73 Plusnin, 4 78 Poet, 495 Pogorski, 310, 311,322,323,326, 327, 328,329, 341,498 Polanski, 361,362, 370, 498 Poll, 177, 498 Pollock, 495 Polyak, 490 Pomarskii, 498 Popov, 502 Popova, 489
523
Author index
Poppelbaum, 269, 401,509 Poreda, 477 Porritt, 4 76 Posner, 482 Pournis, 505 Pradel, 381,382, 498 Prasolov, 492, 509 Press, 4 71 Price, 149, 152, 215,218, 219, 222, 234, 235,482, 498 Prihar, 12, 4 71 Prokhorov, 489 Proshnievo, 4 72 Prutkina, 370, 372, 373,498 Punanova, 29, 498 Purvis, 179, 498 Putikov, 17, 18, 19, 24, 30, 31, 33, 37, 38, 39, 40, 41, 56, 59, 62, 64, 70, 73, 74, 78,498, 499 Quirk, 482 Quirt, 310, 311,322, 323,326, 327, 328, 329, 341,498 Raines, 328, 477, 499 Raleigh, 4 71 Rankama, 361,499 Rasmussen, 251,253,460, 467, 499 Ray, 485 Reddy, 87, 474 Reid, 241,242,460, 467, 492, 499 Reimer, 155,310, 313,314, 317, 318, 321,322, 323,327, 329, 330, 333, 348,349, 350, 392, 471,475, 493, 499
Reitsema, 140, 499 Rekharskii, 504 Reuther, 258,475 Riaboshtian, 497 Rice, 170, 171,499, 504 Richardson, 365,499
Richers, 134, 144, 149, 150, 151,163, 173, 175, 183, 184, 185, 187, 222, 242, 493, 500, 510 Ridgway, 4 73 Risler, 4 74 Robbins, 428, 435,500 Roberts, 311,313,328, 339, 341,343, 344, 346, 351,433,454, 479, 487, 499, 500, 501,511
Robertson, 9, 460, 500 Robinson, 471 Roboneu, 496 Rock, 240, 241,500, 502 Rodgers, 149, 507 Rogers, 362, 500 Romberger, 4 73 Romer, 459, 500 Rosaire, 140, 183, 185, 186, 219, 234, 481,500
Rose, 250, 253,335,354, 374, 375, 381,393,490, 500, 501 Rosengreen, 508 Rosholt, 365,501 Rosler, 348,501 Ross, 454, 474, 478, 481,501 Rossi, 4 77 Round, 309, 495 Rouse, 501,506 Rowan, 233,484 Rowntree, 428, 501 Rozanov, 504 Rozen, 332, 347, 501 Ruan, 9, 10, 218, 220, 223,224, 230, 481,501
Rushing, 386, 501 Rusinov, 496 Russell, 454, 455,456, 501 Ruth, 503 Ryall, 395,397, 400, 401,403,405, 420, 430, 433,476, 501,510 Rybalka, 4 75 Ryder, 486
524
Ryss, 17, 19, 20, 37, 38, 42, 44, 48, 51, 54, 57, 61, 62, 64, 65, 66, 67, 68, 69, 71, 72, 73, 75, 76, 472, 501,502 Sabins, 491 Sackett, 140, 4 74, 482, 502 Saeed, 490 Sahama, 361,499 Sakovich, 47, 489 Sakrison, 401,502 Salati, 4 78 Salisbury, 487 Saltzman, 503 Sanderson, 177, 502 Sandy, 220, 502 Sano, 307, 350, 502, 503, 509 Sardarov, 349, 502 Sato, 92, 98, 103, 104, 105, 112,353, 502
Satterly, 360, 502 Saukov, 395,397,502 Saum, 4 76 Saunders, 234, 503, 507 Savit, 94, 479 Scalen, 4 72 Scaringelli, 249, 503 Scarth, 495 Schaefer, 223,491,503 Scherbakov, 24, 503 Schlegel, 6, 503 Schoell, 136, 503 Schrayer, 511 Schroeder, 490 Schumacher, 234, 236, 238,239, 240, 244, 503 Schuster, 4 78 Schutte, 481 Schuttelkopf, 360, 361,503 Schutz, 488 Schwartz, 506 Scintrex, 385, 503 Scott, 360, 407,465,494, 503, 507
Author index
Sealey, 141,503 Sears, 178, 220, 222, 478 Sedlet, 386, 503 Segal, 237, 238, 503 Semikin, 489 Senftle, 362, 393,504 Serclyukova, 489 Serdyukov, 490 Serebrennikov, 473, 483 Seveme, 392, 504 Sevemyi, 498 Shapiro, 155,494, 504 Sharma, 4 78 Sharp, 402, 504 Shashkin, 370, 372, 373,498 Shcherbakov, 370, 371,508 Sheppard, 357, 504 Shigaev, 21,504 Shiikawa, 428,504 Shipulin, 249, 504 Shkarupa, 53, 4 71 Shturm, 253,504 Shulman, 472 Shvets, 23, 31,504, 51 ! Sidle, 500 Sidorov, 511 Siegel, 240, 242,504 Silver, 464, 476, 477, 486 Silverman, 136, 504 Simoniak, 491 Simpson, 239, 504 Sinclair, 191,504 Singhroy, 241,504 Sinha, 504 Sittig, 233,504 Sivenas, 102, 105, 504 Smee, 82, 83, 85, 102, 112, 118, 504 Smekalova, 484 Smith, 138, 188, 253,307, 321,360, 362, 384, 388, 479, 486, 498, 501, 5O5
Snelling, 484
525
Author index
Snowdon, 494 Sobolev, 478 Soda, 503 Sokolov, 140, 145, 146, 148, 151, 154, 183, 185, 188, 213,218, 481,496, 505
Soli, 141,505 Solovov, 19, 481,505 Sonnek, 316, 505 Soonawala, 354, 375,505 Sorkokina, 511 Spiess, 497 Springer, 504 Squires, 137, 138, 140, 4 78 Stahl, 137, 505 Stanley, 510 Stanton-Gray, 504 Stantso, 353,505 Starik, 353,357, 362, 368,369, 370, 372, 505, 506 Starobinetz, 146, 188, 190, 506 Steacy, 491 Stefanson, 457, 475 Stellmack, 268, 506 Stemberg, 235,239, 496 Stemprok, 4 73 Stephenson, 312, 336, 337, 347, 506 Stevens, 268, 392, 501,506 Stoker, 359, 506 Stokes, 250, 4 78 Street, 506 Stroganov, 187, 472, 505, 506 Stuckless, 366, 506 Stumm, 91,506 Stump, 495 Subbota, 489 Sugisaki, 315,349, 350, 506 Supelco, 260, 506 Sverlova, 489 Swinnerton, 140, 506 Sykes, 471 Syromyamikov, 365,506
Szecz, 494 Tabarsaranskii, 489 Tackett, 457, 460, 506 Tafeev, 471 Takano, 503 Takaoka, 495 Takeda, 506 Takono, 488 Talma, 509 Talnova, 511 Talwani, 378, 506 Tan, 312, 329, 336, 381,480 Tanner, 9, 353, 368, 374, 381,392, 507 Taskayev, 357, 507 Tayeb, 498 Taylor, 253,254, 255,256, 257, 406, 407,417,418,420, 507 Tedosco, 233,234, 507 Telegina, 188, 189, 507 Teng, 378,507 Teplitz, 140, 149, 507 Terrill, 486, 505 Testury, 49, 507 Thompkins, 354, 507 Thompson, 156, 177, 239, 475, 499, 503, 507, 508
Thornber, 105, 106, 107, 112,508 Thune, 144, 146, 148, 149, 185,488 Tiasto, 502 Tikhomirov, 319, 331,484, 508, 509 Tikhomirova, 319, 331,508, 509 Timchuk, 4 74 Timofeev, 511 Tischendorf, 501 Tissot, 134, 508 Titayeva, 357, 508 Todd, 4 72 Togashi, 507 Tokarev, 370, 371,508, 511 Tolstykhin, 483, 489, 492, 493 Tombrello, 504
526 Tomkins, 110, 111, 112, 508 Tomson, 23,481 Top, 477 Torgerson, 306, 312, 332, 336, 337, 508 Toulmin, 4 72 Townsend, 237, 255,508 Tracor, 262, 263,264, 265, 511 Trescases, 4 73 Trifonova, 498 Tugarinov, 309, 347, 508 Tunell, 396, 508 Turner, 486 Tuzova, 4 76 Tverskoy, 4 74 Tyminskii, 511 Urey, 4 72 Ussler, 497 Uvarov, 17, 73,508 Vainbaum, 505 Van de Laan, 158, 508 Van den Boom, 269, 311,342, 401, 508, 509
Van der Vooren, 475 Van Everdingen, 493 Varshal, 4 73 Vasilieva, 54, 55,502, 509 Vaughn, 494 Vayramova, 511 Verbanac, 167, 168, 509 Verhagen, 331,345,493 Veshev, 4 71 Veyher, 4 71 Vilenskiy, 509 Vinogradov, 357,361,483, 495, 509 Vogel, 306, 313,332, 485, 509 Voitov, 486 Vol'fson, 483 Volkov, 483 Voorhees, 134, 159, 216, 486, 490 Voronkov, 502
Author index
Voronov, 309, 338, 509 Voroshilov, 54, 55,471,472, 502, 509 Voytov, 137, 138, 490, 509 Wakita, 154, 183,307, 350, 502, 509 Walker, 4345, 82, 491,509 Wang, 23, 24, 240, 484, 509 Waring, 510 Warner, 209, 509 Warren, 251,252, 388, 411,484, 509 Wasserburg, 493 Watanabe, 345,503 Weatherby, 175,500 Webb, 485, 492, 500 Webber, 82, 113,509 Webster, 12, 490 Wedepohl, 486 Weiner, 503 Weinrich, 466, 510 Weismann, 134, 141,149, 479, 510 Weiss, 306, 307, 316, 492, 510 Welhan, 477 Wells, 186, 494 Welte, 134, 508 Wen, 17, 24, 31, 36, 499, 510 Wennervirta, 381,387,393, 510 Wesolowski, 501 Wesson, 485 West, 249, 267, 482, 491, 510 Whelan, 151, 173, 174, 175,487, 488, 498, 510
Whitcomb, 504 White, 360, 473, 510 Whitehead, 484 Wilford, 504 Wilkening, 367, 378, 381, 510 Williams, 134, 149, 163, 166, 170, 171, 189, 489, 495, 510 Williamson, 386, 497 Williston, 395,397, 412, 485, 510 Wilmshurst, 417, 420, 435,476, 507, 510
527
Author index
Wollny, 454, 510 Wood, 360, 371,472 Wu, 395,412, 510 Wullstein, 240, 494 Wuschke, 347, 483 Wyatt, 152, 510 Yabuki, 494 Yakovleva, 483 Yakutseni, 489, 509 Yamaguchi, 457, 511 Yamashita, 499 Yangfen, 395, 421, 511 Yanitskii, 310, 316, 347, 474, 481,484, 496, 501, 511
Yasenev, 505 Yasko, 511 Yaylor, 478 Yeremeyev, 4 74
Yufa, 365, 495 Yukler, 491 Yurovsky, 505 Zabairaevi, 511 Zaikowski, 313, 511 Zarella, 145, 51 ! Zeidelson, 505 Zhang, 511 Zheng, 511 Zhu, 240, 428, 511 Zinger, 190, 51 l Zonghua, 395, 421, 511 Zorin, 4 78 Zorkin, 144, 147, 154, 192, 472, 505, 511
Zverev, 363,378, 511 Zykov, 4 78
This Page Intentionally Left Blank
529
GEOGRAPHICAL INDEX
Argentina Neuquen Basin, 192 Australia, 79, 304, 332, 395,400, 428 New South Wales, 403, 413, 417, 418 Goulburn, 424 Northern Territory, 324 Amadeus Basin, 239 Cooper Basin, 178 Koongarra, 319, 330, 335,338, 351 Queensland, 406, 408,418 South Australia Stuart Shelf Mt. Gunson, 331 Tasmania, 412, 422 West Coast Range, 413 Victoria Gippsland Basin, 178 Western Australia Eneabba, 325,348 Gingin, 318, 341, 351 Kalgoorlie, 412 Kookynie, 410 Manyingee, 317, 326, 330, 334, 335 Mt. Weld, 320 Mulga Rock, 325 Yeelirrie, 328, 330, 334, 335,393 Azerbaijan, 79, 462 Apsherouskaya, 192 Nizhniekurinskaya, 192
Kirovobadskaya, 192 Kobystano-Shemahinskaya, 192 Brazil Sao Francisco Basin, 241 Byelorussia, 45, 79 Canada, 79, 92, 113,250, 312, 387, 393 Alberta, 222 British Columbia, 411 Anyox, 402 Manitoba, 336 Boggy Lake, 347 Newfoundland Labrador, 330, 375 Northwest Territories Great Bear Lake, 362 Ontario, 336 Bancroft, 359, 363,365 Elliott Lake, 330 Ottawa, 332, 363 Toronto, 258 Quebec Noranda, 401 Val D'Or, 404 Rocky Mountains, 222, 345 Saskatchewan, 330, 363,365,375 Cypass Hills district, 342 Key Lake district, 366 Midwest Lake, 380 Yukon Selwyn Basin, 402
530 China, 23, 79, 130, 153,213,221,223, 297, 299, 300 Anhui, 123, 126 Gansu, 123 Huaitongshan, 124 Jiangsu, 123, 124, 131 Jingbian, 225,227, 231 Junggar Basin, 239 Lixian Depression, 229-231 Ning-Wu Basin, 440 Ning-Zeng, 444 Ning-Zheng, 441 Ordos Basin, 225, 231 Shandong, 123 Wang-wang, 124 Shanghai, 123, 421 Tang-Lun, 441 Xi'an, 225,229 Xinjiang, 239 Cyprus, 411 Denmark Fredrikshavn, 239 France, 343,345 Germany, 140, 311,342 Lohrheim, 425 Guatemala, 341 Gulf of Alaska, 239 Mexico, 170-172, 177, 181 Cameron, 197 Green Canyon, 151 High Island, 192, 195 Iceland, 343,348 India, 79 Gujarat, 347 Rajasthan, 347 Ireland Longford County, 286
Geographical index
Israel Dead sea, 345 Tiberias, 345 Italy, 348 Vulsini Mountains, 346 Japan, 153,307, 343,349 Murho, 364, 365 Nigorikawa, 345 Rokko, 364, 365 Kazakhstan, 48, 78, 79, 286 Kyrgyzstan, 463 Marianas Islands, 345 Namibia, 463 Peru, 414, 416 Russia, 17, 43, 153,332 Baykal, 345,461 Caucasus, 78,462 Kaliningrad district, 28 Karelia, 75, 76 Kerch-Tamon, 345 Kola peninsula, 23, 44, 48, 67, 68, 75 Kurile-Kamchatka, 343,345 Leningrad district, 79, 190 Orenburg, 78 Rudny Altay, 44, 49, 72, 78, 79 Salchalin, 144 Siberia, 45, 79, 90 Transurals, 347 Urals, 404 South Africa, 332, 343,345 Swaziland, 343,345 Tadjikistan, 75, 79 Tanzania, 345
Geographical index
Trinidad River Estates, 455 UK, 469 England, 312, 344, 348 Rothamsted, 454 USA, 79, 134, 140, 153,343 Alabama, 243 Appalachian Mountains, 242 Arizona, 258, 277, 283,288, 464 Casa Grande, 254, 259 Colorado Plateau, 466, 467, 469 Little Dragoon Mountains, 271 California, 239, 341,349 Cholame Valley, 154 Sacramento, 165, 166, 169 San Bernardino County, 163 San Joaquin Basin, 144, 165, 169 Colorado, 362, 366, 367, 393 Arvada, 318 Denver Basin, 392 Golden, 318 Weld County, 321,323,330, 351 Hawaii, 258,344, 349 Illinois, 139 Kansas, 239 Louisiana, 170, 197,377 Michigan, 241 Montana, 452 Nevada, 192, 240, 452 Railroad Valley, 197-203,205, 207, 209 New Mexico Ambrosia Lake district, 323 New York Monticello, 335 Oklahoma, 237, 244, 339 Arkoma Basin, 239 Oregon, 241, 318, 323
531
Pennsylvania, 381,393 Pittsburgh, 175, 176 Titusville, 133 South Dakota, 332 Edgemont district, 330, 333 Texas, 141, 163, 165,239, 244, 322, 330 Bush Dome, 340 Utah, 144, 192, 209, 339 Great Salt Lake, 381 Lisbon Valley, 237, 238, 392 Overthurst Belt, 210, 211 Roosevelt Hot Springs, 344, 348 Virginia, 242 Hardy County, 163 Lee County, 175 Lost River, 144, 163 Rose Hill, 173, 175, 176 Washington Spokane Mountain, 392 Wisconsin, 285 Wyoming, 192,209, 239, 242, 362, 366, 367 Bighorn Basin, 237, 240 Copper Mountain, 392 Overthurst Belt, 210, 211 Patrick Draw, 144, 163, 185 Rawlins, 146 Red Desert, 322, 323,326, 328, 392 Sweetwater County, 163 Yellowstone, 343 USSR (Former), 78, 79, 134, 140, 144, 153, 166, 185,250, 286, 304, 310, 343,347, 352, 395,404, 428, 452 Uzbekistan, 78, 79 South Fergana, 399 Zimbabwe, 345
This Page Intentionally Left Blank
533
PETROLEUM AND MINERAL DEPOSIT INDEX
Base-metal deposits, 428,439 Epithermal, 396, 399 Massive sulphide, 403-405,422 Elura, Australia, 417, 418, 431, 432 Hidden Creek, Canada, 402 Woodlawn, Australia, 400, 401, 413 Stratiform Dugald River, Australia, 406, 407, 431,432
Mierenshan, China, 444 Copper-gold-uranium deposits Olympic Dam, Australia, 331 Copper-molybdenum deposits Skarn Tongshan, China, 131, 441,443, 444, 448 Zhangyan, China, 443,446 Copper-nickel deposits Pechenga, Russia, 48, 50, 51, 76 Kola Peninsula, Russia, 63, 68
Copper deposits, 440, 450 Massive sulphide Crandon, USA, 285,286 Metasomatic Johnson Camp, USA, 258, 259, 271-282, 288,464, 469 Porphyry Kyzyl-Tu, Kazakhstan, 48, 52 North Silver Bell, USA, 283-285, 288,326 Santa Cruz, USA, 254 Sedimentary Witvlei, Namibia, 465 Skarn, 297 Funuishan, China, 441,442, 444 Huaitongshan, China, 124 Troodos Complex, Cyprus, 411, 412 Unspecified Anjishan, China, 444
Epithermal deposits Base-metal deposits, 396, 399 Zinc-lead deposits Bezymyannoe, Uzbekistan, 400 Karaotek, Uzbekistan, 400 Gas deposits Ashland, USA, 239 Gingin, Australia, 341, 351 Harley Dome, USA, 339, 343 Hugoton, USA, 239 Jingbian, China, 225-227, 228, 231 Lost River, USA, 242 Mist, USA, 241 Palm Valley, Australia, 239 Sacramento, USA, 165 Woodada, Australia, 341,348 Gold deposits, 45, 79, 461 Balei, Russia, 461
534 Kookynie, Australia, 410, 411 Lady Hampton, Australia, 403 Sigma, Canada, 404, 405 Iron deposits, 443,450 Porphyritic, 297 Meishan, China, 440, 441 Wang-wang, China, 124, 125 Lead-zinc deposits. See Zinc-lead deposits Massive sulphide deposits Base-metal deposits, 403-405,422 Elura, Australia, 417, 418, 431, 432 Hidden Creek, Canada, 402 Woodlawn, Australia, 400, 401, 413 Copper deposits Crandon, USA, 285,286 Zinc-lead deposits Keel, Ireland, 286, 288 Currawang, Australia, 416, 432, 434 Henry Fault Zone, Australia, 423, 424 Howard's Pass, Canada, 402 Lady Hampton, Canada, 404 McArthur River HYC, Australia, 403 Mount Isa, Australia, 403 Que River, Australia, 415 Mercury deposits Sakhalinsk, Russia, 462 Molybdenum deposits, 440, 450 See also Copper-molybdenum deposits Nickel deposits. See Copper-nickel deposits
Petroleum and mineral deposit index
Oil deposits Blackburn, USA, 240 Cement, USA, 237, 239, 339, 343 Chickaska, USA, 237 Cliffside, USA, 340 Cholame Valley, USA, 155 Currant, USA, 200-203,205,207, 209 Davenport, USA, 239 Deyminshoe, Russia, 30 Eldingen, Germany, 342 Eola, USA, 237 Filo Morado, Argentina, 192-195 Fox-Graham, USA, 239 Healdton, USA, 237 High Island, USA, 195 Lisbon Valley, USA, 237, 238 Loma de La Lata, Argentina, 192195 Lost Hills, USA, 156, 165, 341 Lost River, USA, 144, 163, 186 Mulchto, Russia, 144 Ocho Juan, USA, 239 Ordos Basin, China, 225 Patrick Draw, USA, 144, 163, 185, 242 Pinqiao, China, 228,229 Pollard, USA, 243 Recluse, USA, 239 Rose Hills, USA, 175 Rubelsanto, Guatemala, 341 Severo-Krasnoborshoe, Russia, 30 Stoney Point, USA, 241 Uinta Basin, USA, 165 Uvalde, USA, 165 Velma, USA, 237, 239 Volga-Ural, Russia, 55 Yuzno-Radovskoe, Russia, 54 Polymetallic deposits, 78, 79, 463 Skarn, 443 Xiaomaoshan,China, 442, 445
Petroleum and mineral deposit index
Unspecified Altyn-Topkan, Tadjikistan, 77 Korbalikhinskoe deposits, Russia, 49,52 Rudny Altay, Russia, 44, 64, 65 Shanbaidu, China, 124, 126, 443, 448 Porhyry(-itic) deposits Copper deposits Kyzyl-Tu, Kazakhstan, 48, 52 North Silver Bell, USA, 283-285, 288,326 Santa Cruz, USA, 254 Iron deposits Meishan, China, 440, 441 Wang-wang, China, 124, 125 Precious-metal deposits, 439. See also Gold deposits Pyrite deposits Filizchai, Azerbaijan, 462, 463 Lake Yindarlgooda, Australia, 412414 Xuanchengshan, China, 126, 129 Rare-metal deposits, 79 Skarn deposits Copper deposits, 297 Funuishan, China, 441,442,444 Huaitongshan, China, 124 Tongshan,China, 448 Troodos Complex, Cyprus, 411, 412 Zhangyan, China, 443 Copper-molybdenum deposits Tongshan, China, 441,443,444 Zhangyan, China, 446 Polymetallic deposits, 443 Xiaomaoshan, China, 442, 445 Stratiform deposits Base-metal deposits
535
Dugald River, Australia, 406, 407, 431,432 Zinc-lead deposits Lady Loretta, Australia, 403, 408 Tar deposits Athabasca, Canada, 143 La Brea, USA, 143 Tin deposits, 79 Primorsky Kray, Russia, 45 Uranium deposits Angela, Australia, 324, 327 Athabasca, Canada, 329 Aurora, USA, 318, 323 Bennett Well, Australia, 329, 330, 334, 335 Copper Mountain, USA, 329 Edgemont, USA, 329 Honeymoon, Australia, 329, 330, 335 lnda Lake, Canada, 329 Key Lake, Canada, 329 Koongarra, Australia, 330 Lamprecht, USA, 329 Malcolm, Australia, 329 Manyingee, Australia, 327, 329, 330, 334, 335 Mt. Gunson, Australia, 329 Mt. Weld, Australia, 329 Mulga Rock, Australia, 325,327 Red Desert, USA, 329 Spokane Mountain, USA, 329 Yeelirrie, Australia, 328-330, 334336 Uranium-gold deposits Jabiluka, Australia, 419-421, 428, 433 Olympic Dam, Australia, 331
536 Volcanogenic massive sulphide deposits. See Massive sulphide deposits Zinc-lead deposits, 443 Epithermal Bezymyannoe, Uzbekistan, 400 Karaotek, Uzbekistan, 400 Hydrothermal Qixiashan, China, 125-128, 444, 445,449 Massive sulphide Currawang, Australia, 416, 432, 434
Petroleum and mineral deposit index
Henry Fault Zone, Australia, 423, 424 Howard's Pass, Canada, 402 Keel, Ireland, 286, 288 Lady Hampton, Canada, 404 McArthur River HYC, Australia, 403 Mount Isa, Australia, 403 Que River, Australia, 415 Stratiform Lady Loretta, Australia, 403,408 Unspecified Naking, China, 422
537
SUBJECT INDEX
Absorption, 150, 216, 393,435,459 Actinon in soil gases, 392 half-life, 354, 392 Adsorbent, 267, 269 activated charcoal, 14, 159, 160, 222, 387 gaseous bubbles, 31 molecular sieve, 159, 160, 250, 258, 259, 271,273,313 silica gel, 159, 160, 458 soil, 14, 177, 273,365,398,433 Adsorption, 18, 32, 146, 147, 150, 155. 157-159, 164, 176, 177, 178, 221, 253,258, 259, 269, 291-293,357, 358,365,370, 389, 398,430, 436 Aerosol, 156, 158,216 Alkali elements in oil, 29 Alkaline earth elements in oil, 29 Alteration clay mineral, 235 diagenetic, 142 hydrocarbon-induced, 235,236, 245 phyllic, 283 potassic, 283 propylitic, 283 Ammonia, 459 in atmosphere, 5 remote monitoring, 158
Analytical calibration, 268, 274 enhancement, 327, 342 error, 309, 315, 319, 322, 325, 343, 348, 351,352 methods acid treatment, 226-228, 230 acid-extraction, 176, 178 alpha scintillometer, 384, 386, 387,389,391 alphameter, 385,388 atomic absorption spectrophotometer (AAS), 435 beta particle counter, 384 collector method, 386 colorimetric method, 249, 250 detector tubes, 267, 268 Draeger tubes, 459 EDTA technique, 177 etched track detectors. See track etch field, 222 FID. See flame ionisation detector flame ionisation detector, 159, 164, 172, 175, 193,222 flame photometric detector, 262, 263,265 flameless atomic absorption detection system, 440 FPD. See flame photometric detector gamma-ray scintillometry, 384
538
gas chromatography, 137, 138, 149, 164, 172, 175, 193,213, 222, 250, 253,255,259-266, 268, 269, 276, 283,457, 460 GC. See gas chromatography Hall electrolytic conductivity detector, 262, 264, 265 HECD. See Hall electrolytic conductivity detector hydrogen stripping, 230 infrared adsorption, 164 interferometry, 457 ionisation chamber, 353,383-385, 386, 392 LIDAR, 158 mass spectrometry, 213,265,285, 313-15,457, 460 Orstat gas analyser, 459 photoionisation detector, 262, 265, 266 PID. See photoionisation detector mass spectrometer portable, 267, 268 quadrupole, 222 Rn on activated charcoal (ROAC), 386, 387 solid state detector, 384 standard mud gas logging, 149 thermal conductivity, 164, 457, 458 thermal desorption, 223,227 thermoluminescent dosimeter, 384, 388 titration, 123, 127 track etch, 384, 388 UV fluorescence spectroscopy, 179 West-Gaeke reaction, 267 wet chemical methods, 267 X-ray diffraction, 419 performance, 323,326, 327 precision, 314, 315
Subject index
recording, 270 sensitivity, 250, 309, 313, 317, 348, 367, 429 Anomaly annular, 218, 219, 227, 229, 231 apical, 83, 85, 112, 118, 218, 219, 227,231 false, 94, 113,302,439, 450 hydrocarbon, 187, 213,227, 234 hydromorphic, 409 electrical conductivity, 112 linear, 218, 231 magnetic, 113 probability map, 209 rabbit-ear, 83-85, 112, 117, 118, 279, 420 residual, 409, 412, 413 selective leach, 81 spontaneous potential, 92 Antimony dispersion, 30 in massive sulphide deposits, 400 in oil, 28, 29 in soil, 443 Archimides forces, 22-24 Argon, 261,220 determination, 315 in atmosphere, 5,344 in thermal springs, 343,345 in water, 343 normalisation of helium, 320 Arsenic in massive sulphide deposits, 403 in oil, 29 in soil, 413,439, 441,443 Arsenopyrite, 102, 396, 406 Atmosphere composition, 5 pressure, 83 Barite, 406
Subject index
Beryllium in oil, 29 Biogeochemical analysis, 243 anomalies, 235,243 effects, 244 Bismuth in oil, 29 half-life, 356 Butane analysis, 138, 148, 193 generation, 134, 137 in microseeps, 216 in soil gases, 134, 137, 138, 161, 166, 169, 194, 216, 341 migration, 168 to methane ratio, 168 Cadmium in massive sulphide deposits, 402 in oil, 29 Calcite, 235,238, 239, 293 Carbon isotopes, 141,239, 462,469 Carbonyl sulphide, 249 abundance, 253 analysis, 250, 255,260, 261,264, 271,295 generation, 254 in drill hole, 277 in soil, 259, 276, 272, 279, 283, 286, 301 Carbon dioxide as carbon source, 253 associated with faults, 348 in fluid inclusions, 131 in groundwater, 84 in soil air, 327, 394, 452-454, 457, 465-469 analysis, 123,457-458 anomalies, 118, 124, 131,132 dispersion pattern, 123, 131
539
primary halo, 126 sampling, 123 sources, 127-128, 130-131 in water, 359 measurement, 451, 453,457-460 remote monitoring, 158 Carbon disulphide, 249 abundance, 253 analysis, 250, 255,260, 261,264, 271,295 generation, 254 in drill hole, 277 in soil, 259, 272,276, 279, 283,286, 301 Carbon monooxide, 265 Carbonatites, 303,343, 347 Chalcocite, 57, 452 Chalcopyrite, 57, 102,253,254, 271, 283,285,286, 293,403,406, 410, 421, 441-443,445,453,463,464 Chloritisation, 126 Chlorosis, 241,244 Clay glaciolacustrine, 83 Clay minerals kaolinite, 238,406, 414, 419 montmorillonite, 406 Cobalt in oil, 29 Copper in deposits epithermal, 400 hydrothermal, 125, 126 massive sulphide, 400, 403,413 polymetallic, 28, 44, 49, 78 porphyry, 48, 49, 254, 283,297, 326 sedimentary, 465 skarn, 124, 297, 421, 441,443, 444 stratiform, 406 in oil, 29
540 in soil, 126, 400, 403,406, 407, 409411, 413, 415, 417, 423,424, 427, 443,444, 446 in voltaic cell, 87 in water, 18, 31 native, 102 migration, 18 oxidation, 93, 94 reduction, 87 Diffusion, 8-10, 21, 33, 36-39, 41, 46, 47, 58, 82-85, 91, 97, 113, 117, 143147, 168, 169, 217, 218,234 coefficient, 8 gradient, 87 rates, 9 Dispersion by groundwater, 82, 322 halo, 3, 17 diffusion, 21 hydrocarbon, 185-187 jet, 19-21, 32 metal anomaly, 110 heavy metal, 19 hydromorphic, 395,413 petroleum, 214 stream. See jet hemisphere, 3-4, 7 Earth cell, 118 Earthquake prediction, 153-155. See also Helium, for earthquake prediction; Radon, for earthquake prediction Effusion, 143, 144, 182, 234 Eh. See Redox Electrical conductivity, 85, 100, 101, 105, 107, 111,115, 118 anomaly, 112 field, 88, 100, 102, 113
Subject index
Electrochemical cell, 105 field, 85 gradient, 82, 85, 99, 114, 117 processes, 81, 85, 112, 116, 119 reaction potential, 57, 62 transport, 81, 82, 85 Electrode potential. See Standard electrode potential Electrogeochemical processes, 81 Electrolytic cell, 88, 102, 105 Equipotential lines. See Redox, equipotential lines Ethane analysis, 138, 143, 148, 158, 159 in microseeps, 216 in soil gases, 134, 137, 138, 140, 147, 158, 159, 166, 168, 169, 193, 194, 195, 197, 208,209, 216, 341 in water, 171 migration, 168 measurement. See analysis Factor analysis, 335 Fluid pressure, 83, 84 Fulvates, 49, 50 Galena, 57, 102, 253,254, 285,286, 293,396 Galvanic corrosion, 85 Galvanodynamic polarisation, 54, 58 Gas permeability of igneous rocks, 22 of sediments, 22 Geobattery, 85, 86, 101 Geobotanical anomalies, 235 effects, 243 indicators, 240 Geoelectrochemical exploration, 17, 19, 43, 53, 67, 78, 79
Subject index
methods, 17, 20, 24, 43, 73, 79 CHIM. See partial extraction of metals DCPL. See direct current polarographic logging diffusion extraction of metals (MDE), 40, 41, 46-49, 51, 79 air mode, 47 ground mode, 47 direct current polarographic logging (DCPL), 73-75 MDE. See diffusion extraction of metals MPF. See organometallic organometallic (MPF), 49-51, 79 partial extraction of metals (CHIM), 36, 40-49, 51, 79 equipment, 43 ground mode, 43, 44 logging mode, 43 polarographic logging, 73, 75-77, 79 PPL. See pulse polarographic logging pulse polarographic logging (PPL), 73-76 thermomagnetic (TMGM), 49-51, 79 TMGM. See thermomagnetic prospecting, 17, 19, 20, 27, 36, 45, 46, 48-51, 69, 73, 76, 79 Geophysical fields, 17 information, 134 measurements, 219 methods, 17, 217 procedures, 349 survey, 347, 352 Goethite, 451 Gold collector, 430, 435 film, 13,435
541 in deposits, 45, 79, 102, 112, 119, 297,403,404, 406, 420, 421, 461 in soil, 118 wire, 422 Graphite, 57, 119, 134 conductance, 54, 92, 101, 102, 105, 112 Gravitational force, 24 Gravity data, 233 specific gravity, 355, 373 Groundwater aquifers, 321, 331,332, 335, 338, 351,352 as an electrolyte, 114 contamination, 75 meteoric, 83, 84, 99, 303, 331 phreatic zone, 83, 117 radioactive dating, 332, 333 redox potential, 106 role in dispersion, 82, 83, 113 surveys, 330, 338,340 vadose zone, 83, 97, 113 water table, 83-85, 97, 106, 108, 109, 110,113,115, 118,311,312,321, 324-326, 330, 331,340, 350-352 Half-life period, 321,367, 374, 383, 391 actinon, 354, 392 bismuth, 356 lead, 356, 390 polonium, 356 palladium, 356 radium, 358, 391 radon, 7, 12, 321,355,357, 362, 367, 386, 390 thoron, 354, 356, 368, 388, 392 uranium, 356 Halogen elements in oil, 29 Heavy metals, 27, 76
542 chemical bonds, 18 concentrations, 18 co-precipitated, 17 dispersion, 19 in capillary moisture, 17 in gaseous and quasi-gaseous states, 17 in groundwater, 17 in mineral lattices, 17 in oil and gas, 28 metallo-organic compounds, 17 migration, 21-22 sorbed, 17 Helium abundance, 304-306 as pathfinder, 303,320, 339, 344 as tectonic indicator, 341 associated with faults, 346-348 diffusion, 307, 309, 310, 312, 321, 350 discovery, 304 for earthquake prediction, 154, 303, 304, 315-317, 349-350 for radioactive dating, 306 geochemical mobility, 320 in gas reservoirs, 343 in gas traps, 216 in geothermal waters, 343-346 in groundwater, 304, 306, 310, 312, 314, 319, 321,326, 328, 336, 340, 344, 346-352 in microseeps, 339 in soil gases, 279, 309, 310, 315, 317, 319-323,326, 327, 339-341, 343,344, 346, 348,350, 351 periodic variation, 317, 318 sampling, 309-311,348, 350 in surface waters, 336, 337 in thermal springs, 331,343 in water, 312, 313 analysis, 316, 317 isotope ratios, 304, 306, 307
Subject index
determination, 316 isotopes, 23,303,304 migration, 309, 310 normalisation, 315,320, 327, 352 origin, 305 portable analysers for, 316 properties, 307-308 spots, 183 Humates, 49, 50 Hydrocarbon gases (unspeciated). See also Butane, Ethane, Methane, Pentane, Propane adsorbed, 141,149-151,175 bound. See adsorbed chimney, 234, 235 chromatographic separation, 146, 188 detection techniques, 134 diffusion coefficients, 144 dissolved, 149 effusion transport, 144 free, 150 in soil gases, 140, 141, 144, 149, 152, 157, 159, 164-166, 168, 183, 185, 187, 189, 192-195,211,212, 234, 241 auger hole surveys, 164, 165, 169, 173, 175, 176 compositional ratios, 141 isotopic composition, 188 mode of occurrence, 219-222 origin, 188 shallow probe sampling, 161-163 halo anomalies, 185 migration, 133, 134, 143-148, 150, 168, 169, 182, 185, 187-189, 194, 217, 234, 244 oxidation, 219, 234, 235 saturated, 188, 189 selective adsorption, 146, 147, 159 spots, 183 thermogenic, 133, 143
543
Subject index
trapping, 134 volatile hydrocarbons, 213 Hydrochemical prospecting, 75 Hydrogen as reducing agent, 99 biogenic, 23 bomb, 315 diffusion, 39, 46 in atmoshere, 5 in electochemical reactions, 38, 56, 74, 88, 91, 92 in water, 24, 99 measurement, 154, 164 stripping, 230 Hydrogen sulphide, 5, 6, 23, 216, 238, 249 abundance, 253 acid-released, 292, 293,296, 297, 301,302 adsorption, 291-293 analysis, 251,255,260, 261,267, 268 as reducing agent, 90, 236, 381 in drill hole, 277 in soil, 249, 272,279, 280, 283,286, 292 dispersion, 7 solubility, 23 Hydrolysis, 88, 94 Industrial contamination, 75, 77 Inert gases, 357 Iodine, 386 in atmosphere, 5 Ion at anode, 86 at cathode, 86 carrier, 58 conductance, 114 exchange, 94, 115 flow, 37, 112
migration, 37, 99, 113-15 mobility, 106, 112, 114 pump, 316 redox behaviour, 117 sorption, 115 transfer, 37 Iron hydroxides, 18, 19, 49, 235 in crystalline lattice, 19 in deposits polymetallic, 443 porphyritic, 297, 440 skarn, 124 in industrial effluent, 77 in soil, 117, 177, 414 dispersion, 30 meteorites, 306 oxides, 18, 19, 49, 235,236, 341, 348, 358, 359, 398,406, 411,424 oxidation, 4, 106, 117 Jarosite, 406, 407 Kimberlites, 96, 119, 303,305,347 Komatiites, 119 Krypton, 5 Lamprophyres, 119 Laplace transformation, 59 Lead half-life, 356, 390 in deposits epithermal, 125, 127 massive sulphide, 286, 288,402404, 415, 416, 423,424, 432, 434, polymetallic, 28, 49, 78 skarn, 124, 443-445,449 stratiform, 403,408 in soil, 400, 403,407, 409, 411, 413, 417,422,424, 427,433,440, 443, 444, 446
544 in voltaic cell, 88 isotopes, 366 migration, 18 Lineament, 198,200-203,205,207 photolineaments, 183, 202, 207 Lithogeochemical survey, 44 Lithostatic pressure, 110 Macroseepage. See Seeps, macroseeps Magnetic data, 233 susceptibility, 50 Magnetite, 57, 102, 11O, 113, 216, 441443 Manganese hydroxide, 18, 19, 49 in industrial effluent, 77 in oil, 29 migration, 21 oxides, 18, 19, 39, 358, 359 Marcasite, 453 Martite, 441 Mass flow, 8, 11, 12, 217, 218, 321, 451 Mercury analytical methods, 435-437 as pathfinder, 395, 411 diffusion, 9 dispersion, 395,398-403,406, 408410, 437,440 halos, 125,395,399, 403,437 geochemistry, 396-398 in atmosphere, 427-429 in hypogene environment, 395 in rocks, 434 in secondary environment, 395, 397, 398,406 in massive sulphide deposits, 400403
Subject index
in soil, 395,398, 399, 400, 406, 407, 411,412, 417, 418,420-427, 430437, 439-450 in soil gases, 226, 395, 418, 420-422, 427, 430, 433,437 survey, 415,417, 418, 421,425, 428, 434 pollution, 439, 444, 447, 448, 450 solubility, 398 thermally-released, 126, 131,439441,443-445,447, 448 vapour pressure, 439 Metallogenic province, 405 Methane adsorbed, 160, 210 analysis, 138, 146, 148, 158, 159, 193 as reducing agent, 119 dating, 139 biogenic, 216, 330, 342 dependent ratios, 189 generation, 134 in atmosphere, 5 in loess, 226 in microseeps, 216, 220 in soil gases, 134, 137-140, 146-148, 155, 158, 159, 166, 168, 169, 175, 188-191, 193, 194, 195, 197, 208, 209, 227, 241,243, 341,348, 352 in water, 171 isotope composition, 136, 137 migration, 168, 194 oxidation, 238 remote monitoring, 158 measurement. See analysis thermogenic, 216, 342 to ethane ratio, 140, 166, 168 Microbiological activity, 153, 160 techniques, 141, 217 Microseepage. See Seeps, microseeps
Subject index
Modelling experimental, 23, 31,103, 105-107, 110-112,375,376 laboratory. See experimental numerical. See mathematical mathematical, 9, 10, 309, 328 predictive, 288 Molybdenite, 57, 102, 396, 421,442, 443 Molybdenum in soil, 126 migration, 18 Neon in atmosphere, 5 in thermal springs, 345 in water, 345 normalisation of helium, 315, 320, 327,352 Nernst equation, 87, 89 Nickel in oil, 29 Nitrogen, 173,223,261,266, 268, 304, 313,315,339, 436, 457 determination, 216 in atmosphere, 5 Nitrogen dioxide, 5, 158 Nitrous oxide, 264, 265 in atmosphere, 5 in soil, 160, 177 Ohm's law, 40, 61,100, 118 Olefin, 188, 189 Olivine, 95, 99 Oxidation potential, 86-89, 103, 104 state, 86, 88, 92, 117 suite of elements, 118 Oxidation cell, 85 Oxygen as oxidising agent, 96, 97 associated with faults, 348
545 concentration cell, 85 diffusion, 82 electrical potential, 95, 98 in atmosphere, 451 in groundwater, 451 in soil air, 394, 451-453,457, 458, 465,468 in soil solutions, 451 isotopes, 142, 469 measurement, 451, 453,457-460 subsurface, 97 Ozone in atmosphere, 5, 158 remote monitoring, 158 Palaeochannels, 330, 334, 335 Palladium half-life, 356 Paraffin deposits, 113 Pearson correlation analysis, 191 Pentane adsorbed, 160 in microseeps, 216 in soil, 166 Pentlandite, 57, 107 Petroleum migration, 133 origin, 214 pH, 91-96, 98, 113, 114, 235,240, 242, 253,254, 293,294, 452 Plumbogummite, 406, 407 Polarisation curve, 54, 55, 62 contact polarisation curve (CPC), 37, 53, 57, 60, 61, 63, 64, 66-71, 73, 78, 79 contactless polarisation curve (CLPC), 53, 57, 69-73, 79 Polarogram, 73-75, 77 Pollution air, 267, 268, 287 industrial, 75, 77, 267 soil, 73,439, 444, 448, 450
546 soluble, 302 Polonium half-life, 356 Porosity of igneous rocks, 22 of sediments, 22 Preferential pathway model, 187 Propane adsorbed, 160 analysis, 138, 148 generation, 134, 137 in microseeps, 216 in soil gases, 199, 216, 341 migration, 168, 193 Pyrite, 57, 102, 105-107, 216, 235-237, 253-255,271,283,285,286, 293, 298, 300, 396, 406, 408, 412, 421, 427, 441,442,445, 451,453,464, 466 oxidation, 451-453 Pyroxenes, 99 Pyrrhotite, 57, 102, 99, 105-107, 396 Quartz, 226 Radium, 353 adsorbed, 374 geochemistry, 357 in groundwaters, 359, 361 in soil, 361 in water, 360, 362, 363 half-life, 358,391 migration, 358,381 Radon as pathfinder, 353 associated with faults, 348 determination, 389-392 detection, 382-383 methods, 384-389 diffusion, 358, 367, 368,371,373376, 378, 381,387, 394 discovery, 353
Subject index
emanation, 353,356-358, 362, 363, 367-369, 371,373-376, 378, 392 for earthquake prediction, 154, 349, 350,353 geochemical mobility, 320 half-life, 7, 12, 321,355,357, 362, 367, 386, 390 in atmosphere, 353,368 in oil, 216 in soil, 357 in soil gases, 325,353,373,378382,385,392 fluctuations, 378,379, 381-383, 386, 387 surveys, 379, 382, 392-394 in surface waters, 358 in water, 359-364 isotopes, 354 leakage. See emanation migration, 321,353,374, 369-378 mobility, 357, 367, 378 properties, 354-356 radioactivity, 367 Redox, 91-99, 104, 108, 114, 115, 117, 216, 235,240, 242, 253,254 anisotropy, 108-110 buffering, 96 cell, 110, 112, 116, 118 differential, 107, 108, 111 equipotential lines, 88, 101, 108, 109, 111,112 field, 98, 101,102, 105, 111, 118 gradient, 98, 105, 106, 108, 111, 114, 118 stratification, 98 Reflectance bleached red beds, 236, 237 carbonate enrichment, 238-240 kaolinite, 238 vegetation, 240
Subject index
Remote sensing aerial photography, 142, 180, 233, 239 Aircraft Thematic Mapper Simulator (NS-001), 239, 242, 243 imaging spectrometers, 233 Landsat Thematic Mapper (TM), 233,237-239, 241,242, 244 multispectral scanning system (MSS), 233 multispectral video data, 244 satellite imagery, 142, 144, 180, 239, 244 SPOT, 233 stress image, 241 Resistivity, 94, 100, 101, 107 Sampling contamination, 123, 162, 170, 189, 307, 310, 315,318, 327, 346, 389, 421,432-435,440 density, 197, 209, 230, 341 depth, 271, 311,312, 318, 327, 330, 348,350, 379, 382,425, 431,437, 450 efficiency, 327 error, 189, 309, 311,319, 322, 325, 343, 351,390 gases soil, 151, 159-169, 198, 256, 258, 259, 262, 267, 273,285,309, 310, 311, 321,323,324, 325, 326, 341,348, 350, 352, 391, 422, 433,457 atmospheric, 153-158, 321,427 dissolved, 169-171 headspace, 172-173 interval, 250 media, 6, 427, 433 methods, 14, 31,131,155-173,215, 231,256, 258, 259, 262, 267, 285, 287, 291,302, 304, 310, 311, 312,
547 321,323,324, 325,328, 344, 351, 376, 379, 382, 390, 421,422, 432, 457 overburden, 392 precision, 391 soil, 131,326, 328, 406, 407, 425, 433,434, 437, 450 strategy, 180 techniques. See methods water, 195, 312, 335,336, 344, 346, 348, 350, 351,391 Seepage. See Seeps Seeps, 149, 150, 163, 168, 170, 192, 239 macroseeps, 133, 151,152, 158, 182, 183,214, 233,234 microseeps, 133, 134, 141-145, 152, 182, 183, 187, 190, 211,214, 216 -219, 228, 231,233,234, 238, 240-242, 244, 245,320, 339, 341 role of faults, 183, 184 Seismic methods, 141 pumping, 378 Seismogeoelectrochemistry, 79 Selective leach anomaly, 81-85, 110, 113, 115, 117 element zonation, 117 procedures, 115 sampling, 81 Selenium in oil, 29 Self potential. See Spontaneous potential Siderite, 216, 441 Silica, 293 Simulation, 254, 291,293,376 Soil aeration, 453,456 agricultural, 453 air, 249, 250, 271,276, 286, 289 atmosphere, 319, 350
548 biological processes, 453 biology, 319 emanations, 353, 391 free gases, 6, 7, 310 horizons, 423, 431,432,437,447 lithosols, 406, 409, 410, 431 micro-biological activity, 453 micropores, 311,328 moisture content, 456 organic matter, 453 profile, 275,357, 447, 448 respiratory quotient, 453 sampling, 328 Specularite, 441 Sphalerite, 57, 102, 119, 253,254, 271, 285,286, 293, 441-443,445 Spontaneous potential, 91-94, 98, 101, 103, 105-107, 109, 111-113, 115, 118 anomaly, 92, 112 cell, 101-103, 106, 107, 112 current fluxes, 110 origin, 112 Standard electrode potential, 88, 90, 91 Standard potential. See Standard electrode potential Standard voltage. See Standard electrode potential Statistical tests, 191 Stoke's law, 24 Sulphide anions, 291,295,297, 301,302 detection, 291,295,297, 302 dispersion, 291,295,297 in groundwater, 291, 301 minerals, 250, 253-255,270, 283. See also Galena; Pyrite; Pyrrhotite; Sphalerite oxidation, 249, 251,291, 451-453, 463,469 deposits, simulated, 254, 293 Sulphur, 250. See also Sulphur gases
Subject index
organic compounds, 254, 255,260262,265,283 Sulphur dioxide, 5, 90, 216, 249,435 abundance, 253 analysis, 251,255,260, 261,264, 267,268, 271,283,285,286 generation, 252 in soil, 259, 271,302 remote monitoring, 158 solubility, 253,283 Sulphur gases (unspeciated). See also Carbonyl sulphide, Carbon disulphide, Hydrogen sulphide, Sulphur dioxide analysis, 253, 255 by gas chromatography, 259-266 by other methods, 267 sampling, 256-258 stability, 253 in soil, 249, 250, 253-256, 260, 261, 267, 268,270, 271,273-275,277, 283,286, 288,289 sampling, 256-259 Super-deep drilling, 23 Supergene enrichment, 112 zone, 283 Thoron emanations, 392 half-life period, 354, 356, 368,388, 392 in atmosphere, 368 in soil gases, 392, 394 Thallium in oil, 29 Thermal springs, 84, 331,343 Thermodynamic data, 91 reactions, 91 Thorium isotopes, 303
549
Subject index
Tin in soil, 126 Trend-surface analysis, 326 Tungsten in soil, 126 Uraninite, 235 Uranium decay series, 357, 365-367, 383 disequilibrium, 362, 365 emanation, 373 half-life, 356 in water, 328,363,365 isotopes, 303,365 radioactive decay, 320 roll-front, 362 Vanadium in oil, 29 Vapour pressure, 83, 84 Vegetation anomaly, 242 density, 241,243 stress, 244, 245 Violarite, 107 Voltaic cell, 85-88, 91, 92, 99, 101, 102, 104, 105, 107 Voltammograms. See Polarisation curve
Water capillary, 18, 23 groundwater. See groundwater juvenile, 303 radioactivity, 361,362 redox stability field, 98 sampling, 342 Weathering profile, 107 Xenon, 5 Zeolites, 110 Zinc in deposits epithermal, 125, 127 massive sulphide, 286, 288,402404, 415, 416, 423,424, 432, 434, polymetallic, 28, 49, 78 skarn, 124, 443-445,449 stratiform, 403,408 in soil, 400, 403,407,409-411, 413, 417,422-424, 427, 440, 443,444 in voltaic cell, 87, 88 oxidation, 87
This Page Intentionally Left Blank
