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
._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 ( 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 (-~ 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
/ ,//-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 Ca~RB- O~
T
"~
CAM - I~
" Gas Content
i Gas Content
i.9 ~[
T
I' I
I
I
o.
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 -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, -0.1 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 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 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 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) 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 (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
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 (
