Baixe o app para aproveitar ainda mais
Prévia do material em texto
Fractures, Fluid Flow and Mineralization Geological Society Special Publications Series Editors A. J. FLEET R. E. HOLDSWORTH A. C. MORTON M. S. STOKER It is recommended that reference to all or part of this book should be made in one of the following ways: MCCAFFREY, K. J. W., LONERGAN, L. & WILKINSON, J. J. (eds) 1999. Fractures, Fluid Flow and Mineralization. Geological Society, London, Special Publications, 155. CLOKE, I. R., CRAIG, J. & BLUNDELL, D. J. 1999. Structural controls on hydrocarbon and mineral deposits within the Kutai Basin, East Kalimantan. In: MCCAFFREY, K. J. W., LONERGAN. L. & WILKINSON, J. J. (eds) Fractures, Fluid Flow and Mineralization. Geological Society, London, Special Publications, 155, 213-232. GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 155 Fractures, Fluid Flow and Mineralization EDITED BY KEN MCCAFFREY School of Geological Sciences, Kingston University, Surrey, UK LIDIA LONERGAN & JAMIE WILKINSON T. H. Huxley School of the Environment, Earth Sciences and Engineering Imperial College of Science, Technology and Medicine, London, UK 1999 Published by The Geological Society London THE GEOLOGICAL SOCIETY The Society was founded in 1807 as The Geological Society of London and is the oldest geological society in the world. It received its Royal Charter in 1825 for the purpose of 'investigating the mineral structure of the Earth'. The Society is Britain's national society for geology with a membership of around 8500. It has countrywide coverage and approximately 1500 members reside overseas. The Society is responsible for all aspects of the geological sciences including professional matters. The Society has its own publishing house, which produces the Society's international journals, books and maps, and which acts as the European distributor for publications of the American Association of Petroleum Geologists. SEPM and the Geological Society of America. Fellowship is open to those holding a recognized honours degree in geology or cognate subject and who have at least two years' relevant postgraduate experience, or who have not less than six years' relevant experience in geology or a cognate subject. A Fellow who has not less than five years' relevant postgraduate experience in the practice of geology may apply for validation and, subject to approval, may be able to use the designatory letters C Geol (Chartered Geologist). Further information about the Society is available from the Membership Manager, The Geological Society. Burlington House. Piccadilly, London W1V OJU, UK. The Society is a Registered Chanty. No. 210161. Published by The Geological Society from: The Geological Society Publishing House Unit 7, Brassmill Enterprise Centre Brassmill Lane Bath BA13JN UK (Orders: Tel. 01225445046 Fax 01225 442836) First published 1999 Reprinted 2001 The publishers make no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. c The Geological Society of London 1999. All rights reserved. No reproduction, copy or transmission of this publication may be made without written per- mission. No paragraph of this publication may be reproduced, copied or transmitted save with the pro- visions of the Copyright Licensing Agency, 90 Totten- ham Court Road, London W1P 9HE. Users registered with the Copyright Clearance Center. 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 0305-8719 99 $15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 1-86239-034-7 ISSN 0305-8719 Distributors USA AAPG Bookstore PO Box 979 Tulsa OK 74101-0979 USA (Orders: Tel. (918) 584-2555 Fax (918) 560-2652) Australia Australian Mineral Foundation 63 Conyngham Street Glenside South Austral ia 5065 Australia (Orders: Tel. (08) 379-0444 Fax (08) 379-4634) Affil iated East-West Press PVT Ltd G-l 16 Ansari Road New Delhi 110 002 India (Orders: Tel. ( 1 1 ) 327-9113 F a x ( 1 1 ) 326-0538) Japan Kanda Book Trading Co. Cityhouse Tama 204 Tsurumaki 1-3-10 Tama-Shi Tokyo 0206-0034 Japan (Orders: Tel. (0423) 57-7650 Fax (0423) 57-7651) Typeset by Aarontype Ltd, Unit 47. Easton Busines Centre, Felix Road', Bristol BS50HE. UK Printed by Cambridge University Press. Cambridge. UK Contents Dave Johnston: an appreciation and bibliography vii LONERGAN, L., WILKINSON, J. J. & MCCAFFREY, K. J. W. Fractures, fluid flow and 1 mineralization: an introduction Fracture populations ROBERTS, S., SANDERSON, D. J. & GUMIEL, P. Fractal analysis and percolation properties 7 of veins STOWELL, J. F. W., WATSON, A. P. & HUDSON, N. F. C. Geometry and population 17 systematics of a quartz vein set, Holy Island Anglesey, North Wales GILLESPIE, P. A., JOHNSTON, D. J., LORIGA, M. A., MCCAFFREY, K. J. W., WALSH, J. J. & 35 WATTERSON, J. Influence of layering on vein systematics in line samples LORIGA, M. A. Scaling systematics of vein size: an example from the Guanajuato mining 57 district (Central Mexico) Fluid flow and fracture systems SANDERSON, D. J. & ZHANG, X. Critical stress localization of flow associated with 69 deformation of well-fractured rock masses, with implications for mineral deposits JONES, M. A., PRINGLE, A. B., FULTON, I. M. & O'NEILL, S. Discrete fracture network 83 modelling applied to groundwater resource exploitation in southwest Ireland CONNOLLY, P. T. & COSGROVE, J. W. Prediction of static and dynamic fluid pathways within 105 and around dilational jogs Structural controls on mineralization Cox, S. F. Deformational controls on the dynamics of fluid flow in mesothermal gold 123 systems BLENKINSOP, T. G. & SANDERSON, D. J. Are gold deposits in the crust fractals? A study of 141 gold mines in the Zimbabwean craton JOLLEY, S. J., HENDERSON, I. H. C., BARNICOAT, A. C. & Fox, N. P. C. Thrust-fracture 153 network and hydrothermal gold mineralization: Witwatersrand basin, South Africa ROSSETTI, P. & COLOMBO, F. Adularia-sericite gold deposits of Marmato (Caldas, 167 Colombia): field and petrographical data BRANQUET, Y., CHEILLETZ, A., GIULIANI, G., LAUMONIER, B. & BLANCO, O. Fluidized 183 hydrothermal breccia in dilatant faults during thrusting: the Colombian emerald deposits BROWN, M. A. N., JOLLY, R. J. H., STONE, W. & COWARD, M. P. Nickel ore troughs in 197 Archaean volcanic rocks, Kambalda, Western Australia: indicators of early extension CLOKE, I. R., CRAIG, J. & BLUNDELL, D. J. Structural controls on hydrocarbon and mineral 213 deposits within the Kutai Basin, East Kalimantan Irish Zn/Pb deposits: structure and fluid flow HITZMAN, M. W. Extensional faults that localize Irish syndiagenetic Zn-Pb deposits and 233 their reactivation during Variscan compression CONTENTS vi EVERETT, C. E., WILKINSON, J. J. & RYE, D. M. Fracture-controlled fluid flow in the Lower 247 Palaeozoic basement rocks of Ireland: implications for the genesis of Irish-type Zn-Pb deposits LEWIS, H. & COUPLES, G. D. Carboniferous basin evolution of central Ireland - simulation 277 of structural controls on mineralization SEVASTOPULO, G. D. & REDMOND, P. Age of mineralization of carbonate-hosted, base metal 303 deposits in the Rathdowney Trend, Ireland O'REILLY, B. M., READMAN, P. W. & MURPHY, T. Gravity lineaments and Carboniferous- 313 hosted base metal deposits of the Irish Midlands Index 323 Dave Johnston: an appreciation and bibliography On 2 October 1995, Dave Johnston disappeared while working in the field at Annagh Head, Belmullet, County Mayo, Ireland. He is believed to have been washed from the shoreline by a freak wave. Dave graduated in 1980 with a first- class honours degree in Geology from Trinity College Dublin and he returnedthere in 1984 having been appointed as a lecturer in the Department of Geology. In the intervening years, Dave carried out postgraduate research at Monash University, Melbourne and gained a PhD for his work on the structural controls of Uranium deposits in the Rum Jungle area, Northern territories. His subsequent research focused on the structural controls of the precious and base-metal deposits and the underlying tectonics and structure of Ireland. In recent years he published a number of significant papers on these topics and, in particular, he was at the forefront in the application of fractal and chaos theory to geological phenomena and the quanti- fication of mineral deposits in particular. Dave was an enthusiastic teacher and inspired many of the students and researchers who he taught or worked with. Indeed, many former students now work in a diverse range of geology- related industries. He firmly believed that aca- demia could interact and collaborate more closely with industry to their mutual benefit. Thus, he was elected as a council member for dthe Irish Association for Economic Geology. In his other role as President of the Irish Geological Association he was keen to bring geology to a wider public audience. Dave talked freely to students and colleagues and liked nothing better than discussing geology over a pint after a day in the field. He was a deeply caring individual who made time to help anyone who needed it. Throughout, he remained closely attached to his home and family in North Dublin and considered himself to be very for- dtunate to be able to do what he loved best, geological research based in the magnificent surroundings of Trinity College. Dave will be remembered as a lively and colourful character who enjoyed life to the full, whether it be playing rugby, scuba diving, doing fieldwork or travelling. For many years he led the singing at the late-night 'Friends of the Irish' party at the Tectonics Studies Group annual meeting. He always had a bad joke or a strange story to tell, often it was the many slightly bizarre incidents that always seemed to happen to him. These usually involved the series of clapped-out vehicles that he always seemed to own. There was the time his car rolled down a cliff in Donegal, the time his car's engine went on fire, but because he was late for a meeting at the Lisheen deposit, he threw some water on it and drove on! Anyone who knew or met Dave has a similar story to recount of his various exploits. This volume on 'Fractures, Fluid Flow and Mineralization' is a fitting tribute to his memory. As George Sevastopulo so appropriately said at the conference held in Dublin in November 1997 to commemorate Dave's life, 'Dave would have loved to have been here'. He is greatly missed by his former students, many colleagues and friends around the world. K. J. W. McCaffrey School of Geological Sciences, Kingston University, Kingston-upon-Thames, Surrey KT1 2EE, UK BIBLIOGRAPHY Bibliography WALL, V. J., ETHERIDGE, M. A., Cox, S. F. & JOHNSTON, J. D. 1984. Regional metamorphic ore deposits - structural and chemical controls. In: MARSHALL, B. & GILLIGAN, L. B. (eds) Mechanical and Chemical Mobilization of Metal- liferous Mineralization. Geological Society of Australia Short Course, 33-40. JOHNSTON, J. D. 1985. Interpretation of refolding and asymmetric folds using vergence concepts in drill- core. Journal of Structural Geology, 7, 317-326. JOHNSTON, J. D. 1988. Structural controls of gold mineralization in Western Australia. Irish Associa- tion for Economic Geology Annual Review 1988, 74. SANDERS, I. S. & JOHNSTON, J. D. 1989. The Tor- ridonian Stac Fada Member; an extrusion of fluidised peperite? Transactions of the Royal Society of Edinburgh: Earth Sciences, 80, 1-4. PALMER, D., JOHNSTON, J. D., DOOLEY, T. & MAGUIRE, K. 1989. The Silurian of Clew Bay, Ireland: part of the Midland Valley of Scotland? Journal of the Geological Society, London, 146, 385-388. McELDUFF, B. & JOHNSTON, J. D. 1989. Bedrock occurrence of Platinum Group Metals (PGMS) in Ireland. Irish Association for Economic Geology Annual Review 1989, 111. JOHNSTON, J. D. 1990. The Untamed Earth. Technol- ogy Ireland, 55-58. SANDERS, I. S. & JOHNSTON, J. D. 1990. Reply to The Torridonian Stac Fada Member: a discussion'. Transactions of the Royal Society of Edinburgh: Earth Sciences, 81, 249-250. MURPHY, F. C, ANDERSON, T. B. DALY, J. S., GALLAGHER, V., GRAHAM, J. R., HARPER, D. A. T., JOHNSTON, J. D., KENNAN, P. S., KEN- NEDY, M. J., LONG, C. B., MORRIS, J. H., O'KEEFE, W. G , PARKES, M., RYAN, P. D., SLOAN, R. J., STILLMAN, C. J., TIETZSCH-TYLER, D., TODD, S. P. & RAFTER, J. P. 1991. An appraisal of Caledo- nian suspect terranes in Ireland. Irish Journal of Earth Sciences, 11, 11-41. VAUGHAN, A. M. P. & JOHNSTON, J. D. 1992. Structural constraints on closure geometry across the lapetus Suture in eastern Ireland. Journal of the Geological Societv, London, 149, 65-74. JOHNSTON, J. D. 1992. The fractal geometry of vein systems: the potential for ore reserve calculations. In: BOWDEN, A. A., EARLS, G., O'CONNOR, P. G. & PYNE, J. F. (eds) The Irish Minerals Industry 1980-1990. Irish Association for Economic Geology, Dublin, 105-117. JOHNSTON, J. D. 1993. Ice wedge casts in the Dalradian of south Donegal: evidence for subaerial expo- sure of the Boulder Bed. Irish Journal of Earth Sciences, 12, 13-26. JOHNSTON, J. D. 1993. Three-dimensional geometries of veins and their relationship to folds; examples from the Carboniferous of eastern Ireland. Irish Journal of Earth Sciences, 12, 47-63. MCCAFFREY, K. J. W., JOHNSTON, J. D. & FEELY. M. 1993. Use of fractal statistics in the analysis of Mo-Cu mineralisation at Mace Head. County Galway. Irish Journal of Earth Sciences. 12, 139-148. JOHNSTON, J. D., TAIT, J. A., OLIVER. G. J. H. & MURPHY, F. C. 1993. Evidence for a Caledonian orogeny in Poland. Transactions of the Royal Society of Edinburgh: Earth Sciences, 85, 131-142. JOHNSTON, J. D., MCCAFFREY, K. J. W., LORIGA, M. A.. WATTERSON, J., WALSH, J. J. & GILLESPIE, P. A. 1994. A Manual Describing Recording, Analysis and Prediction of Vein and Related Fracture Distribution. MIRO, Lichfield. FITZGERALD, G., FEELY, M., JOHNSTON, J. D.. CLAY- TON, G., FITZGERALD, L. J. & SEVASTOPULO, G. D. 1994. The Variscan thermal history of west Clare. Ireland. Geological Magazine, 131. 545-558. JOHNSTON, J. D. 1995. Major northwest-directed Caledonian thrusting and folding in Precambrian rocks, northwest Mayo, Ireland. Geological Mag- azine, 132. 91-112. JOHNSTON, J. D. 1995. Pseudomorphs after ikaite in a glaciomarine sequence in the Dalradian of Done- gal, Ireland. Scottish Journal of Geology. 31. 3-9. JOHNSTON, J. D. & PHILLIPS, W. E. A. 1995. Terrane amalgamation in the Clew Bay region, west of Ireland. Geological Magazine. 132. 797-806. JOHNSTON, J. D. 1995. A review of the pre-Devonian rocks of Ireland. ///: ANDERSON. K., ASHTON. J.. EARLS. G., HITZMAN, M. & TEAR, S. (eds) Irish Carbonate-hosted Zn-Pb Deposits. Society of Eco- nomic Geologists Guidebook Series 21. 209-217. JOHNSTON, J. D. 1995. Variscan deformation in Ire- land. In: ANDERSON, K.. ASHTON. J.. EARLS. G.. HITZMAN, M. & TEAR, S. (eds) Irish Carbonate- hosted Zn-Pb Deposits. Society of Economic Geologists Guidebook Series, 21, 111-113. MCCAFFREY, K. J. W. & JOHNSTON. J. D. 1996. Fractal analysis of a mineralised vein deposit: Curraghinalt gold deposit. County Tyrone. Mineralium Deposita. 31. 52-58. JOHNSTON, J. D. & MCCAFFREY, K. J. W. 1996. Fractal geometries of vein systems and the variation of scaling relationships with mechan- isms. Journal of Structural Geology. 18. 349-358. WILKINSON, J. J. & JOHNSTON. J. D. 1996. Pressure fluctuations, phase separation, and gold precipi- tation during seismic fracture propagation.Geol- ogy, 24, 395-398. JOHNSTON, J. D., COLLER, D., MILLAR, G. & CRITCH- LEY, M. F. 1996. Basement structural controls on Carboniferous-hosted base metal mineral depos- its in Ireland. ///: STROGEN, P.. SOMERVILLE. I. D. & JONES, G. LL. (eds) Recent Advances in Lower Carboniferous Geology. Geological Society. Lon- don, Special Publications, 107. 1-21. JOHNSTON, J. D. 1997. Localization of mid-crustal thrust ramps by metadolerite sheets in the Dalradian of northwest Ireland. Geological Magazine. 134, 199 212. Compiled by P. N. Wyse Jackson, Department of Geology, Trinity College, Dublin 2. Fractures, fluid flow and mineralization: an introduction LIDIA LONERGAN,1 JAMIE WILKINSON1 & KEN McCAFFREY2 1 T. H. Huxley School of Environment, Earth Sciences & Engineering, Imperial College of Science, Technology and Medicine, London SW7 2BP, UK School of Geological Sciences, Kingston University, Kingston-upon-Thames, Surrey KT1 2EE, UK The historical context Prior to Hubbert and Rubey's classic paper (Hubbert & Rubey 1959), which specifically recognized the role of high fluid pressures in lowering the shear stress required to move and emplace large thrust sheets, geologists had tended to ignore the importance of fluids in crustal deformation (Fyfe et al. 1978). Much has changed in the intervening 40 years. It is now accepted that not only do fluids enable deforma- tion but that the converse is also true, i.e. that faulting can cause fluid migration. The grow- ing interest in the relationship between fluid flow and deforming rock in a range of geological environments has led to knowledge on the topic becoming increasingly specialized. This is reflected by the number of major books pub- lished in the last decade spanning subject areas more traditionally considered the realm of struc- tural geologists, metamorphic or sedimentary petrologists and hydrogeologists. Notable exam- ples include: Fluids in Subduct ion Zones (Tarney et al. 1991); The Role of Fluids in Crustal Pro- cesses (National Research Council, US Geophy- sics Study Committee 1990); Geofluids (Parnell 1994), focusing on the flow of fluid in sedi- mentary basins; Rock Fractures and Fluid Flow (National Research Council, US Committee on Fracture Characterization and Fluid Flow 1996); and Fluid Flow and Transport in Rocks (Jamtveit & Yardley 1997) which addresses the integration of the physics and chemistry of fluid flow through crustal rocks. Data collated from the major geoscience reference database, GeoRef, shows an exponen- tial growth in the number of papers published since 1975 that include 'faults/fracturing' and 'fluid' in their title or among their keywords (Fig. la). This growth, culminating in a publication rate in excess of 400 papers per annum, reflects the explosion in research interest in the subject area in the last 10-15 years. In the period from 1975 to the early 1980s, most of these publica- tions, typically of the order of 40 per year, were concerned with ore deposits or mineral- ization (Fig 1b). This is not surprising given that many modern ideas concerning the flow of fluids in fractures stem from observations made as a result of human exploitation of mineral resources. The origin of many modern concepts may be traced to works such as Waldemar Lindgren's classic text Geology of Mineral Deposits (Lind- gren 1913) in which he presented the first classifi- cation and interpretation of mineral deposits within a genetic context. It was recognized, even then, that 'the great majority of mineral deposits were formed where the path of [solutions] was prescribed by openings in the rocks other than those of ordinary pore space... These openings of super-capillary size... are chiefly found in the zone of fracture' (considered to be generally <6km in the crust). By the late 1980s and early 1990s papers were no longer limited to mineralization studies but had spread across a number of disciplines, princi- pally rock mechanics, structural geology, meta- morphic petrology, hydrology, water resources and petroleum geology. Much of this recent research has benefited from stimuli in three main areas: • an increased subsurface fluid database arising from drilling for economic resources (hydro- carbon, geothermal and mineral indust- ries), as well as ODP and deep crust drilling programs; • the increased computing power of the last 20 years allowing more complex numerical simulations to be performed; • the impetus from groundwater resource man- agement and nuclear waste disposal sectors. Whilst endeavour in these fields has led to significant progress along a number of avenues, interdisciplinary papers and research colloquia LONERGAN, L., WILKINSON, J. J. & MCCAFFREY, K. J. W. 1999. Fractures, fluid flow and mineralization: an introduction. In: MCCAFFREY, K. J. W., LONERGAN, L. & WILKINSON, J. J. (eds) Fractures, Fluid Flow and Mineralization. Geological Society, London, Special Publications, 155, 1-6. L. LONERGAN ET AL. Fig. 1. (a) Total number of references with 'fault/fracture' and 'fluid' in the title, abstract or keywords from 1975 to 1997 inclusive (data from GeoRef database). Data have been normalized for total number of publications in the database for each year. Search terms used: '(fracture* or fault*) and fluid*, (b) Data in (a) shown on non- normalized scale (•) and numbers of those publications which were about ore or mineralization ( )- Search term used was '[(fracture* or fault*) and fluid*] and (ore or mineral*)'. crossing the boundaries of the traditional subject disciplines remain relatively rare. It seemed that now, some 15 years on, was a timely moment to reassess the subject of fluid flow and fracturing in the context of mineralization, and to explore how advances within different research commu- nities could be applied to the understanding of the formation of economic resources. It was also felt that this was an appropriate way to celebrate the life and work of Dave Johnston whose interests were very much in this field. With these aims in mind three communities of geoscien- tists were invited to a meeting in Dublin in November 1997: structural geologists, geoscien- tists, with interests in fluid flow, and economic geologists. The papers collected in this volume evolved from those presented at the meeting and are intended to give a state-of-the-art review of current ideas on the interrelationships between fracturing and fluid flow, particularly in relation to the genesis of hydrothermal mineral deposits. Permeability, fracture networks and mineralization An issue that has long been recognized as fundamental in fluid flow research is how rock permeability is defined and how to quantify it adequately for fractured rock systems. Indeed, in the introduction of Fluid Flow and Transport in Rocks (Jamtveit & Yardley 1997) it is empha- sized that rock permeability is the key parameter in any model of fluid flow and transport in 2 FRACTURES, FLUID FLOW AND MINERALIZATION rocks. This is particularly true for hydrothermal mineral deposits that form as a result of focused flow of a large volume of fluid. As many of the papers in this volume illustrate (indicated in bold), such mineralization is usually structurally controlled, and hence the need to understand the role of faulting and fracturing in enhancing rock permeability becomes paramount. The papers included in this volume reflect how knowledge in this area has moved on sig- nificantly since Lindgren's text (Lindgren 1913). One of the most important new concepts, high- lighted by the landmark paper of Sibson et al. (1975), was the idea that faults and fracture systems were not 'passive' fluid conduits but could be active in transporting fluid through rock during deformation. Research in neotec- tonic areas (e.g. Muir Wood & King 1993) confirmed, with observationaldata, that active faults do move large volumes of fluid through the crust. Now much research effort is focused on how faulting alters the hydrogeological prop- erties of the rock itself, by modifying the frac- ture pore volume as well as the permeability of the fault and adjacent areas (e.g. Knipe 1993; Matthai & Roberts 1997). The coupling between deformation processes and rock permeability through linked fracture systems is one of the main themes that recurs in various sections in this volume. Some of this work has its origins in the 'fractal revolution' that has occurred in structural geology in the last ten years or so. This approach has proved effective in quantifying the geometry and scaling characteristics of fracture systems, and has led to an increased understanding of how fault systems evolve and grow. Localization of deformation in fractures and shear zones leads to fluid flow localization. When active faults and shear zones link to form percolation networks, the large scale flow systems essential for miner- alization can be established (e.g. Cox, Roberts et al., Sanderson & Zhang). The integration of advances from 'traditional' structural geology with developments in percolation theory will aid both the understanding of the genesis of hydrothermal mineral deposits and improve the ability to explore for these resources. Vein populations The first section in the volume addresses vein systems. Veins are a key natural data set preserving a visible and permanent record of fluid flow and mineral transport in the rock mass. Evidence from field studies shows that various aspects of veins, especially their thick- nesses, have fractal properties (for example, the paper by Stowell et al. on veins hosted in Precambrian rocks of Anglesey, Wales). Roberts et al. deduce that the measured fractal dimen- sion (or D-value) is a useful descriptive para- meter and can potentially be used to predict the connectivity of vein systems. From field data and models Roberts et al. conclude that vein systems with power-law thickness distributions initiate due to the opening of isolated fractures, but at this stage only locally derived fluids enter the fractures. When the isolated vein systems become connected they retain power-law thick- ness distributions but with characteristically lower D-values. Deformation is localized and fluid transport over increased length scales can occur, leading to mineral precipitation given suitable physical and chemical conditions. Thus, thickness distributions of mineralized vein sys- tems would be predicted to have low D-values. From a study of the vein system of the Guana- juato epithermal Ag-Au deposit in Mexico, Loriga also provides data which show that vein thickness populations can be described by a power-law distribution over four orders of mag- nitude, inferring a common growth process for the veins in the deposit. Like Roberts et al. and Gillespie et al., she concludes that veins with power-law distributions will be more prone to mineralization because they form larger con- nected fracture networks allowing large scale fluid transport. Gillespie et al. identify funda- mental differences between stratabound vein arrays (arrays in layered rocks, where the veins are restricted to individual mechanical units) and non-stratabound vein arrays based on studies of vein arrays at ten localities in the British Isles, France and Mexico. The strata- bound vein arrays have regular spacings con- trolled by the layer thickness and do not show a power-law thickness distribution. In compari- son, non-stratabound veins are clustered and exhibit a power law thickness distribution, with D-values that appear to be independent of lithological variation. Fluid flow and fracture systems In the second section of the book the specifics of fluid flow in fracture systems is addressed from a number of different perspectives. Sanderson & Zhang investigate fluid flow and deformation in fractured rock using numerical models, in which mechanical and hydraulic behaviour are coupled. They show that, at a critical stress state, initially diffuse flow though fracture net- works changes to a highly localized flow. The 3 L. LONERGAN ET AL. overall hydraulic conductivity within and across the fracture networks increases suddenly when the critical stress state is reached. At the critical state, the variation in vertical flow rates through different parts of the network shows a fractal distribution. Jones et al provide an example of a sophisticated 3D model of a fractured ground- water reservoir in Carboniferous limestones in southwest Ireland, which explicitly represents fractures using data from boreholes to con- strain the distribution of the fracture popula- tions. This contribution illustrates the sort of complex modelling and simulation that are more common in the hydrological and nuclear waste sectors, but which have great potential for mineralization or hydrocarbon studies. Connolly & Cosgrove address the importance of stress distribution as a controlling factor in influenc- ing fluid migration in tectonically active upper crustal regimes. They use photoelastic experi- ments and theory of brittle failure to obtain pre- dictions of stresses, fractures and fluid pathways around dilation jogs during stick-slip faulting. They investigate the case of applying a constant load to faults that are not propagating (static) and then go on to consider fluid migration path- ways associated with dynamic faulting events. In their example of a dilational jog that starts out with no overlap and evolves to an overlap- ping fault system, fluid is initially driven away from the jog but at later growth increments is focused into the jog at high flow rates. Structural controls on mineralization The third section of the volume is devoted to mineral deposits. A collection of seven papers covers a variety of topics with the general theme of structural controls of mineral deposit for- mation, including some new thoughts on old themes. A provocative paper by Cox re-examines existing ideas on the genesis of mesothermal gold deposits, applying new concepts in percolation theory to their formation. Arguments are pre- sented for the critical percolation state of these systems, providing a natural analogue for the models described by Sanderson & Zhang. Gold mineralization of this type is considered on a regional scale by Blenkinsop & Sanderson who demonstrate that deposits in the Zimbabwean craton have a fractal distribution with a D-value of c. 1. If it is assumed that this D-value can be extended to other areas, then the authors predict that companies can improve their exploration strategies by comparing observed deposit dis- tributions with fractal dusts. An important new data set is presented by Jolley et al., who discuss the major control of lithology on the geometry of structures formed during thrusting in the extra- ordinarily gold-rich Witwatersrand Basin in South Africa. Critical evidence is presented which shows that deformation was coincident with, and controlled the distribution of, hydro- thermal fluid flow and gold mineralization. This is key information in relation to the ongoing controversy surrounding the genesis of Witwa- tersrand gold and strongly supports a hydro- thermal origin for these deposits, long regarded by many workers to be alluvial placers. New thoughts on an old theme are presented by Brown et al., who provide a novel interpretation of the famous Kambalda nickel deposits. Structural data are used to infer that the komatiitic flows hosting mineralization were constrained within early extensional graben structures, similar to the fissure swarms observed in Iceland at the present day. This interpretation has considerable sig- nificance with regard to exploitation of these deposits. Descriptions of vein textures areuseful in interpreting the competing processes of fracture opening and mineral precipitation, a competition which ultimately governs the temporal evolution of permeability in dynamic systems. Vein data and fluid inclusions - the only direct record of palaeofluid properties - are used in two other contributions. Rossetti & Colombo present field and petrographic data from the little known adularia-sericite-type epithermal gold deposits of Marmato, Colombia. Mineralization, hosted within a Neogene volcanic centre, is controlled by fractures associated with a regional fault system. Variations in fluid properties with depth are observed, indicating that gold precipitation occurred in response to competing processes of boiling and dilution in the vein system. Branquet et al. consider the development of fluidized hydrothermal breccias in dilatant zones during thrusting to be the major control of mineralization in the unusual emerald deposits in the Eastern Cordillera of the same country. The final paper in this section, by Cloke et al.. integrates seismic, gravity and structural data to develop a model for the Kutai Basin in Indonesia that shows how basement structure controlled basin evolution and the location of volcanism and hydrothermal activity responsible for major gold deposits. It is shown how a regional understanding of basin evolution and fluid pathways is important for mineral and hydrocarbon exploration. This contribution illustrates how economic geologists, whether their quest be for ore or hydrocarbons, can benefit from modern concepts of basin analysis coupled with effective use of regional geophysi- cal data sets. 4 FRACTURES, FLUID FLOW AND MINERALIZATION Irish base metal deposits At the November 1997 conference in Dublin we were fortunate to be able to bring together a unique group of researchers working on the Carboniferous base metal deposits of Ireland for a lively discussion covering many aspects of the genesis of Europe's premier Zn-Pb deposits. A selection of these papers is presented in the final section of this volume as a unified case study where readers can see aspects of the differ- ent approaches documented in previous sections applied. There has been exciting recent progress in understanding the genesis of the Irish base metal deposits with Hitzman, Everett et al., Lewis & Couples and O'Reilly et al. discussing various aspects of the importance of structural reactiva- tion in basement rocks for controlling fluid flow and mineralization. Hitzman discusses the nature of the extensional fault systems which localize the deposits and considers the evi- dence for, and implications of, their inversion during subsequent Variscan compression. This represents a development of some of the ideas presented by Dave Johnston (Johnston et al. 1996) in one of his last papers. One of the currently popular models for mineralization in Ireland proposes that the deposits formed as a result of long distance flow of fluids through a sandstone aquifer, driven by topographic head from the rising Variscan mountains to the south of the Irish Midlands Basin. This interpretation is challenged by Everett et al., who present new evidence that indicates that fluid flow in fractured basement rocks was an important process in the genesis of the Zn-Pb deposits. This is a theme also addressed by O'Reilly et al., who use reprocessed gravity data to show how the deposits are strongly localized by Caledo- nian basement faults and that intersections with north-northeast and northwest trending struc- tures appear to play an important control on localizing fluid flow and mineralization. Lewis & Couples also contribute to the debate on fluid flow within and beneath the sedimentary basins of the Irish Midlands by constructing 2D coupled fluid and heat flow simulations of the two competing models (topo-graphically driven regional flow and deposit specific density driven convection). Their paper also shows the power of sequential 3D basin reconstructions in aiding the understanding of the tectonic evolution of the Irish Midlands during the Carboniferous. Finally, a controversial interpretation of the timing of mineralization with respect to the age of host rocks is presented by Sevastopulo & Redmond. Their work implies that younger host rocks than previously considered may be pro- spective for Zn mineralization in Ireland. A multidisciplinary problem The integration of all available techniques is the hallmark of current research in the field of fluid flow and mineralization, and many of the papers presented in this volume exemplify the use of multidisciplinary approaches. The geochemistry of fluids trapped within veins (e.g. Everett et al.), geophysical remote sensing such as gravity or seismic methods (O'Reilly et al. and Cloke et al.) and numerical modelling (e. g. Jones et al., Lewis & Couples and Sanderson & Zhang) all have their part to play alongside more classical field geology in helping to solve out- standing complex problems. One of the main aims of this volume is to promote the cross- fertilization of ideas between geoscientists inter- ested in either the passage of fluids through the Earth's crust or the genesis of mineral deposits. Such interdisciplinary collaboration is essential if further advances are to be made in this research area. This volume arose from a joint meeting of the Tectonic Studies Group, Mineral Deposits Studies Group, Irish Association for Economic Geology and Irish Geolo- gical Association on 'Structural Controls and Genesis of Economic Resources: Mineral and Hydrocarbon Deposits' which was held in November 1997 in Trinity College, Dublin. We thank Rio Tinto Mining & Exploration, BHP Minerals, Billiton International Development pic, CSA Ltd, ERA-Maptec and the IAEG for their generous sponsorship of the meeting which allowed us to bring together nearly 200 delegates from all around the world, to celebrate and commemorate the life of Dave Johnston with a stimulating scientific meeting and field trip to the Irish base metal deposits. Behind the scenes, the tireless logistical support given by Patrick Wyse Jackson, Neil Kearney, Frank Hendron, Declan Burke, Stephen Donnolly, JefT Lord, Mags Duncan, Mary Foody, Joann Layng, Lynn Pink and Dick Campain ensured a smoothly run, successful con- ference. Garth Earls, Mike Boland and the staff at Tara Mines (Navan), Minorco Services Ireland (Lish- een) and Arcon Mines (Galmoy) are acknowledged for their assistance in running the field trip. We thank all contributors to the meeting for presenting the science that is contained in this volume and for creating an occasion that Dave himself would have truly enjoyed. The following referees donated their time and expertise for reviewing the manuscripts, often under tight time constrains, and helped us bring this volume to fruition: David Alderton, John Ashton, Chris Bean, Tom Blenkinsop, Adrian Boyce, Olivier Bour, Joe Cartwright, Mike Coward, Patience Cowie, Stephen Cox, Ian Davidson, Garth Earls, Stuart Egan, Chris 5 L. LONERGAN ET AL. Elders, Terry Engelder, Martin Feely, Kerry Galla- gher, Paul Gillespie, Alan Herbert, Richard Herring- ton, Murray Hitzman, Bob Holdsworth, Rob Hunsdale, David James, Gawen Jenkin, Steve Jolley, Richard Jolly, Richard Lisle, Iain Main, Ken McClay, John McCloskey, Eric Nelson, Noelle Odling, Carl Renshaw, Stephen Roberts, Dave Sanderson, George Sevastopulo, Robin Shail, Rick Sibson, Ian Somer- ville, Cees Swager, John Walsh, Nicky White, James Wood, Robert Zimmerman and three other anon- ymous referees. References FYFE, W. S., PRICE, N. J. & THOMPSON, A. B. 1978. Fluids in the Earth's Crust. Elsevier. GEOREF (CD-ROM) 1975-6/1998. American Geolo- gical Institute, SilverPlatter. HUBBERT, M. K. & RUBEY, W. W. 1959. Role of fluid pressure in mechanics of overthrustfaulting. Geo- logical Society of America Bulletin, 70, 115-166. JAMTVEIT, B. & YARDLEY, B. 1997. Fluid Flow and Transport in Rocks. Chapman & Hall. JOHNSTON, J. D., COLLER, D., MILLAR, G. & CRITCH- LEY, M. F. 1996. Basement structural controls on Carboniferous-hosted base metal mineral deposits in Ireland. In: STROGEN, P., SOMERVILLE, I. D. & JONES, G. LI. (eds) Recent Advances in Lower Car- boniferous Geology. Geological Society, London, Special Publications, 107, 1-21. KNIPE, R. J. 1993. The influence of fault zone pro- cesses on fluid flow and diagenesis. In: HORBURY, E. D. & ROBINSON, A. G. (eds) Diagenesis and Basin Development. AAPG Studies in Geology. 36, 135-154. LINDGREN, W. 1913. Geology of Mineral Deposits. McGraw-Hill. MATTHAI , S. K. & ROBERTS, S. G. 1997. Trans- ient versus continuous fluid flow in seismically active faults: an investigation by electric analogue and numerical methods. In: JAMTVEIT. B. & YARDLEY, B. (eds) Fluid Flow and Transport in Rocks. Chapman & Hall, 263-295. MUIR WOOD, R. & KING, G. C. P. 1993. Hydrological signatures of earthquake strain. Journal of Geo- physical Research, 98, 22 035-22 068. NATIONAL RESEARCH COUNCIL. US COMMITTEE ON FRACTURE CHARACTERIZATION AND FLUID FLOW. 1996. Rock Fractures and Fluid Flow. Contemporary Understanding and Applications. National Academy Press. NATIONAL RESEARCH COUNCIL. US GEOPHYSICS STUDY COMMITTEE. 1990. The Role of Fluids in Crustal Processes. National Academy Press. PARNELL, J. (ed.) 1994. Geofluids: Origin, Migration and Evolution of Fluids in Sedimetary Basins. Geological Society, London. Special Publications. 78. SIBSON, R. H., MOORE. J. M. & RANKIN, A. 1975. Seismic pumping - a hydrothermal fluid trans- port mechanism. Journal of the Geological Societv, London, 131, 653-659. TARNEY, J., PICKERING, K. T., KNIPE, R. J. & DEWEY. J. F. 1991. The Behaviour and Influence of Fluids in Suhduction Zones. Royal Societv. London. 6 Fractal analysis and percolation properties of veins S. ROBERTS,1 D. J. SANDERSON13 & P. GUMIEL2 1 School of Ocean and Earth Science, Southampton University, Southampton Oceanography Centre, Southampton SO 14 3ZH, UK Instituto Tecnologico Geominera de Espana, Rios Rosas 23, 28003 Madrid, Spain 3Present address: T. H. Huxley School of Environment, Earth Sciences of Engineering, Imperial College of Science, Technology and Medicine, Prince Consort Road, London SW7 2BP, UK Abstract: Systematic sampling of vein systems often reveals a power-law distribution of the number of veins (N) to the vein thickness (t) of the form N = Ct-D . A comparison of vein thickness distributions from different geological settings establishes systematic variations between vein populations, with lower D-values characterizing well-connected vein systems. Low D-value vein systems are reported from mineral deposits and from veins developed within and around fault zones, often at constant vein frequency. Simple models are described to explain the field observations which suggest that the observed variations in vein thickness distributions are directly related to the connectivity of the sampled vein network. Mineralized vein systems are an economically important manifestation of the flow of hydro- thermal fluids through the Earth's crust. Major ore deposits of Au, Cu, Sn, W, Pb and Zn are all found within the confines of vein systems. This paper describes physical methods of investigat- ing the development and connectivity of hydro- thermal vein systems and examines the increased range of information which can be acquired. Field data are reported from mineralized and barren vein systems, and theoretical models designed to describe their development. It is argued that the methods described provide new and useful information in the quest to under- stand, explore and ultimately exploit mineralized fracture networks. Vein thickness distributions Measurements of vein thickness along linear traverses, ideally normal to the trend of the veins, can often be made alongside more typical vein descriptions of orientation, architecture, age relationships, mineral content and fibre orientations. In such instances, the thickness, position and orientation of all veins exceeding some minimum thickness, e.g. >l mm, are recorded along the traverses. The data can be compared with normal, log-normal, power-law and negative exponential distributions based on the shapes of the cumulative frequency curves. Typically, cumulative frequency plots of vein thickness conform to power-law distributions, of the form N= Ct-D (1) where N is the number of veins with thickness >t\ C represents the frequency of veins > unit size; the exponent D is often termed the fractal dimension of the distribution. A plot of log N v. log t will be a straight line for a power-law distribution, with slope —D (Fig. 1). Some of the departure from the power-law distribution can be attributed to two types of truncation: under- sampling of the thinnest veins (usually <2 mm); the absence of larger veins, either due to some finite limit of vein opening or the low probability of intersects on a sample traverse of short finite length (Barton & Zoback 1990; Pickering et al 1996); these effects can be noted to varying degrees in many of the plots given in this paper. Vein thickness distributions have been widely studied during the past few years (Sanderson el al 1994; Clark et al 1995; Johnston & McCaffrey 1996; McCaffrey & Johnston 1996; Roberts et al 1998) and have generally been found to conform to power-law distributions. Data from two regions of Central Iberia are outlined below to illustrate this approach. Quartz veins, Central Portugal Metamorphic quartz veins were investigated from Precambrian Complejo Esquisto Grauva- quico of the Castello Branco region of Central ROBERTS, S., SANDERSON, D. J. & GUMIEL, P. 1999. Fractal analysis and percolation properties of veins. In: MCCAFFREY, K. J. W., LONERGAN, L. & WILKINSON, J. J. (eds) Fractures, Fluid Flow and Mineralization. Geological Society, London, Special Publications, 155, 7-16. S. ROBERTS, D. J. SANDERSON & P. GUMIEL Fig. 1. Cumulative frequency plot of vein thickness (t) from two transects in the Castello Branco area, Portugal. The data conform to power-law distributions and Locality 3 shows a lower exponent (D-value) reflecting the development of thicker veins. Both data sets show evidence of truncation of data in the 1-3 mm region. Portugal (Murphy & Roberts 1997) (Fig. 1). The Complejo Esquisto Grauvaquico is a thick sequence of metagreywackes, comprising inter- bedded shales and silts, all metamorphosed to lower greenschist facies. Well-exposed road sections, typically >50m in length, contain syn- and post-kinematic quartz veining of variable character and intensity, typically orien- tated N140°E, and ranging in thickness from 1 to 420mm. A series of line traverses were completed, CB1-CB12, and a summary of the data is given in Table 1. Two contrasting examples of the data sets, CB3 and CB12, are plotted on Fig. 1. Both data sets have similar frequencies of veining > l m m (0.5-lm"1) but contrast in their power-law exponents (D = 0.15 for CB3 and D= 1.3 for CB12). This reflects the presence of thicker veins within the CB3 section, with the majority of data falling between 10 and 80mm thickness, whereas for CB12 the majority of the data fall between 2 and 10mm thickness. Thus, the line samples plotted as log-log cumulative frequency plots provide information about the vein density for a given thickness at each site, and the extent to which the system is dominated by large or small veins. The latter property is reflected in the exponent (D-value) of the distribution, with low D-values reflecting an increased presence of thick veins. Furthermore, such an approach readily allows different sites tobe compared. Ore-Mineralized system, La Codosera, Western Spain Analysis of ore-mineralized vein systems have also revealed power-law vein thickness distribu- tions (Sanderson et al. 1994; McCaffrey & Johnston 1996; Roberts et al. 1998). In parti- cular, a study involving the systematic measure- ment of thickness data was introduced as part of the routine drill core logging procedure in an exploration programme of the La Codosera area of western Spain (Roberts et al. 1991; Sanderson et al. 1994). Measured vein thickness distribu- tions generally conformed to power-law distri- butions (Fig. 2). A comparison of the borehole data shows a decrease in the exponent through boreholes #25, #28 and #27 (Fig. 2), which reflects an increase in vein thickness and vein density, particularly at 10mm thickness, evident within each hole. Notably, the lower D-value is Table 1. Summary of line transect data from the Castillo Branco area of Portugal Sample CB1 CB2 CB3 CB4 CB5 CB6 CB7 CB8 CB9 CB10 CB11 CB12 n 94 72 46 45 71 49 12 38 72 36 94 79 Traverse length (m) 90 80 90 60 60 60 60 60 60 60 60 90 Thickness Min. 2 2 3 2 2 2 2 2 2 2 2 1 (mm) Max. 420 500 500 480 220 170 60 110 460 85 195 50 Thickness (mm/m) 30.00 26.30 24.70 34.40 45.40 21.40 3.05 13.40 33.30 13.30 36.20 4.00 Density (m-1) 1.04 0.90 0.51 0.75 1.18 0.82 0.20 0.63 1.20 0.60 1.57 0.88 D 0.80 0.60 0.75 0.60 0.70 0.77 0.73 0.75 0.75 0.70 0.95 1.30 8 FRACTAL ANALYSIS AND PERCOLATION PROPERTIES OF VEINS Fig. 2. Vein thickness plots from the La Codosera area, southwest Spain. Number of veins per metre of core (N) with thickness >t plotted v. thickness (t) from three different drill holes. Drill hole #21 shows the highest Au assays and the lowest power-law exponent [after Sanderson et al. (1994)]. also broadly correlated with increasing amounts of gold, reported from the assays returned from each hole, which is not simply related to the total vein thickness (Sanderson et al. 1994, fig. 2). Vein thickness distributions and the development of fracture networks The characteristics of the power-law vein thick- ness distributions observed, in particular the D-values, must reflect the mechanisms respon- sible for the formation and evolution of the vein system. Thus, models produced to explain the development of the vein systems should ideally result in the generation of power-law distribu- tions of vein thickness as opposed to normal, log-normal, etc. This section considers a theore- tical model and a detailed analysis of veins developed within the damage zones of faults. The results provide insights into how these vein thickness distributions arise and their potential significance. A theoretical model A simple conceptual model for the develop- ment of a vein system is shown in Fig. 3 in which an initial set of small fractures grows to pro- duce a linked fracture network (Roberts et al. 1998). Zhang & Sanderson (1994) have discussed network evolution in terms of the random addi- tion of fractures with a power-law length dis- tribution. They recognize this as a percolation phenomenon (e.g. Stauffer & Aharony 1992), suggesting that a system of fractures will event- ually connect to form an infinite cluster at a certain fracture density, which can be character- ized by its fractal dimension. The set of fractures within the infinite cluster, which provide path- ways across the network, is termed the backbone and this provides a path for fluid flow and localized deformation. An & Sammis (1996) developed a cellular automaton model to simulate the growth of a network of faults, which employed a set of simple rules based on fracture mechanics, to Fig. 3. Conceptual model of growth of a vein system: (a) initial, small, isolated fractures; (b) growth proportional to length produces local linkage; (c) linkage to form an infinite cluster. Note that the opening (vein thickness) is concentrated on the backbone of the infinite cluster. After Roberts et al. (1998). 9 10 S. ROBERTS, D. J. SANDERSON & P. GUMIEL control initiation, growth and linkage of frac- tures. Based on these studies a conceptual model has been developed (Fig. 3) in which the following rules are incorporated schematically: • an initial power-law distribution of fracture length with no further nucleation (Nur 1982; Segall & Pollard 1980) (Fig. 3a); fractures are located randomly and with random devia- tion from a single mean orientation; • thickness (opening) is proportional to frac- ture length (e.g. Vermilye & Scholz 1995); • change in length (AL) occurs where ALocZ/, where k is a constant dependent on the mechanisms controlling growth: fol- lowing An & Sammis (1996), a value of k = 1 is used which develops geometries similar to those observed in nature; • linkage occurs between two fractures of length LI and L2 when the distance between their tips is <0.1 (L1 +L2). Thus, it is possible to model the development of an infinite cluster by growth rather than addition of fractures. At the percolation threshold, a cluster of infinite length suddenly develops, which localizes both deformation and fluid flow (Zhang & Sanderson 1994) (Fig. 3c). As thick- ness is proportional to length, this will also produce a sudden opening of fractures located on the backbone of this cluster, probably due to the reduced elastic bending stresses required to open the long arrays of linked fractures. Thus, veins which form part of the infinite cluster will have a wider range of thicknesses than those forming isolated veins or clusters, and this will be reflected in a lower fractal dimension (D). For line samples through such vein systems, the isolated veins will be dominated by small length and thickness (D > 1), whereas the connected systems will be dominated by larger veins (D < 1). Using stochastic models, Clark et al. (1995) suggest that power-law distributions of vein thicknesses can arise if the growth of the veins and the associated network involves limited initi- ation of new fractures and incremental growth in proportion to the existing vein length. Adapting these ideas and incorporating the fracture linkage model (Fig. 3) allows vein thickness distributions to be interpreted in the following simple model. Pairs of cracks are randomly sampled from an initial power-law distribution of lengths and thicknesses (D = 1). Each pair is linked by summing the lengths and thicknesses: this assumes tip to tip linkage but the model could be modified to incorporate other types of Fig. 4. Results of simple model for vein thickness distribution. An initial power-law distribution of lengths and thickness (D = 1) is randomly sampled and pairs of cracks linked by summing their lengths and thicknesses. The resulting distribution for linking of 50 and 75% of cracks, with growths of two and four times, respectively, shows lower D-values and a departure from ideal power-laws (after Roberts et al. 1998). linkage. The process is repeated until a given proportion of cracks are linked and all cracks have grown in proportion to their length. The resulting distribution is shown in Fig. 4 for linking of 0, 50 and 75% of cracks. The resulting distributions have lower D-values (D < 1) and show departure from ideal power-laws, particu- larly for the larger cracks. The models also produce patterns of vein thickness distribution similar to those seen in field studies, cf. Fig. 4 with Figs 1 and 2. Vein development in damage zones of faults The above models predict the localized open- ing of fractures, promoting enhanced fluid flow, when connected fracture systems develop. One situation where the transition from uncon- nected to connected vein systems can be examined is in the vicinity of fault zones where regionalveins (usually unconnected) often intensify and become connected in the damage zones. In this section the relationship between vein thickness distribution and localization of deformation is examined from three traverses across damage zones associated with exten- sional faults. FRACTAL ANALYSIS AND PERCOLATION PROPERTIES OF VEINS 11 Kilve, Somerset. An 8m long traverse across three east-west trending normal fault zones (FZ1-FZ3) of displacement 0.4, 0.1 and <0.1 m, respectively, was examined within a single lime- stone unit in the Lower Lias limestones and shales to the east of Kilve Pill (Fig. 5). A total of 3.8m of the traverse was within damage zones of faults, with the remaining 4.2 m lying outside these zones. The limestone layer contains many thin calcite veins with displacements <1 mm. All veins that were observable (probably of thick- ness >0.2 mm) were recorded, but the sample is only considered to be complete for veins c. >0.5 mm in thickness, the average density at this thickness being c. 10 m"1. The veins in the fault zones penetrate the 0.25m thick limestone bed and connect with veins in the underlying and overlying shales, the remainder being confined to the limestone. The frequency of veins does not change within the fault zones (Fig. 5a) but their opening, as shown by steeper cumulative thickness slopes, does. Log-log plots (Fig. 5b) suggest an approx- imate power-law distribution, but the power- law exponent (D-value) within the damage zones is 0.5, compared with a value of c. 1.5 out- side (Table 2). The significant result from this study is that the thickness distributions are markedly different for veins inside and outside the fault damage zones. These vein distribu- tions are interpreted as resulting from the local- ized opening of fractures within the limestone layers due to displacement and strain within the fault zones. Fig. 5. (a) Cumulative vein thickness and vein frequency developed over three small fault zones within an 8 m traverse line at Kilve, Somerset. Note the increase in vein thickness but not vein frequency through the fault zones, (b) Cumulative frequency thickness distributions. O, Data collected outside the recognized damage zones of the faults (D = 1.55); • , data collected from within the damage zones of the faults (D = 0.5). S. ROBERTS, D. J. SANDERSON & P. GUMIEL Table 2. Summary of data from the damage zones of faults Sample Kilve FZ E Millook FZ E Wick FZ E n 64 95 66 102 113 52 Traverse length (m) 3.8 4.2 1.6 6.9 45 60 Thickness Min 0.2 0.2 0.2 0.2 1 1 (mm) Max 100.0 5.5 210 8 163 11 Veins (%) 16.37 1.10 40.01 1.68 2.45 0.15 D 0.5 1.5 0.7 1.1 0.8 1.5 Millook, Cornwall. A low displacement, normal-dextral fault (trending northeast-south- west) is exposed in a sandstone-shale sequence (Crackington Formation, Namurian) on the foreshore at Millook. Two traverses of 4 and 4.5 m (total 8.5 m) were made in a sandstone unit across the fault zone, which has a damage zone (c. 1m wide). A total of 1.6m of traverse was measured within the damage zone, which com- prises a network of connected quartz veins up to 210mm in thickness. Outside this damage zone the veins are <8mm thick and occur as isolated veins, or in short arrays, of overlapping vein segments. The sandstone layer contains many thin veins (displacements <l mm). Although attempts were made to measure all observable veins (thickness >0.2mm), the sample is only considered to be complete for veins >1 mm. The average density at this thickness is c. 7m"1 out- side the damage zone, increasing to c. 30m"1 within the zone. Log-log plots (Fig. 6a) sug- gest an approximate power-law distribution, but the power-law exponent (D-value) within the damage zones is 0.7, cf. c. 1 outside (Table 2). South Head, Wick, north Scotland. A section through two small normal faults (trending north-northwest-south-southweast), with down- throws of c. 1 m and 0.5m to the west, occurs in Middle Devonian rocks (Caithness Flagstones) at South Head, Wick. Extension is associated with a series of N160°E calcite and Fe-carbonate veins. Three traverses, taken approximately normal to the faults, were made; a 45 m traverse through the two faults and the intervening rocks, and two 30m traverses in the same units away from the faults (Table 2). All veins with thickness >1 mm were measured. In the faulted traverse the veins are up to 100mm thick and form a well-connected network with thicker carbonate areas forming pull-aparts or jogs between more planar veins. Away from faults. Fig. 6. (a) , Data from outside the damage zones (D = 1.1); • . data from within the damage zones of faults (D = 0.7); both observed at Millook, Cornwall, (b) Similar data from South Head, Wick, north Scotland. , Data from outside the damage zones of the faults (D = 1.5); • . data from within the damage zones of the faults (D = 0.8). 12 FRACTAL ANALYSIS AND PERCOLATION PROPERTIES OF VEINS 13 veins are usually <10mm thick (Table 2) and occur as isolated veins and side-stepping arrays. Locally, veins are arranged in en echelon arrays, some of which extend from the tips of recogni- sable faults. The average density of veining is usually < 1 m"1 away from faults but increases to c. lOm"1 within the fault zones, averaging 2.5m"1 in the traverse through the two faults. Log-log plots (Fig. 6b) suggest an approximate power-law distribution with D c. 1.5 away from the faults, decreasing to 0.8 within the faulted traverse and, locally, lower still in the fault zones themselves. Vein thickness distributions and grade Gold mineralization in quartz veins Analysis of veins from the La Codosera area of southwest Spain shows that boreholes sam- pled which contained elevated gold grades (>300ppb) display low D-values (fitted to the bulk of the data) compared to boreholes which consistently provided assays of <300ppb (San- derson et al 1994) (Fig. 2). By comparison, borehole and adit data from the Curraghinault gold deposit (McCaffrey & Johnston 1996) have consistently lower exponents (Fig. 7). These data from mineral deposits are consis- tent with the models and observations of the damage zones of faults. They predict that natural vein networks with low D-values char- acterize well-connected systems and are more prospective. Connectivity facilitates the ingress of 'externally' derived fluid and permits a significant increase in the fluid flux through such systems. For example, at La Codosera, a nitrogen anomaly was observed for fluid inclu- sions within quartz samples from assay intervals with enhanced gold values (Dee & Roberts 1993). This was attributed to the trapping of fluids derived from metamorphic devolatization reactions at depth within the host shear zone. This observation is consistent with recent work by Pettke & Diamond (1997) who used Sr isotope signatures to trace fluid evolution and suggested that fluids derived from deep within the crust are an important component of mineralized shear zones at Brusson, northwest Alps. At La Codosera, the gold mineralization is associated with an unmixing aqueocarbonic fluid. Recently, Wilkinson & Johnston (1996) demonstrated that unmixing and gold precipita- tion can be closely associated with the devel- opment of vein systems and, in particular, the linking of veins. These data tend to suggest that the mechanical development and the geochem- ical evolution of the vein systems is intimately linked. Sn-W Mineralization Similar data sets have been acquired from a series of Sn-W veins within the Iberian Penin- sula (Roberts et al. 1998). Sn-W mineralization is associated with intrusions of syn- to late kinematic Hercynian granites, typically occur- ring as quartz veins and stockworks which cut the granites and local countryrocks. The veins sampled were oriented northeast-southwest and dip towards the southeast. The vein-fill is predominantly quartz with no apparent fibres or development of crack-seal textures. The deposits are small with economic extraction of ore occurring only at the La Parilla mine [c. 6000 tonnes of Sn (0.2 wt%) and WO3 (0.05wt%)]. Barren vein sets of similar age, but with no visible sulphides, were studied from Fig. 7. (a) Vein thickness distribution from Curraghinalt gold deposit, after McCaffrey & Johnston (1996). Data shown are typical for adit and drill-hole analysis, (b) Comparative data from a borehole at La Codosera; this sample has lower gold grades but an increased vein density compared to Curraghinalt. 14 S. ROBERTS, D. J. SANDERSON & P. GUMIEL Table 3. Summary of data from the Sn-W mineralization Sample La Parrilla, 1 La Parrilla, 2 A. Montanches El Trasquilon Albala Trujillo, 1 Trujillo, 2 Trujillo, 3 Trujillo, 4 Trujillo, 5 Trujillo, 6 Trujillo, 7 Trujillo, combined n 37 112 74 60 40 9 23 16 17 10 27 23 125 Traverse length (m) 110.0 100.0 37.0 34.0 36.0 11.5 39.0 16.0 30.0 6.0 9.0 25.0 136.5 Thickness (mm) Min. 16.0 1.0 2.5 6.0 1.0 4.0 .0 .0 .0 .0 .0 .0 1.0 Max. 250 400 136 300 48 12 15 15 30 17 22 30 30 Thickness - (m mm) 26.30 42.30 36.60 79.80 14.60 5.70 3.10 4.30 3.70 5.50 16.20 5.36 5.00 Density D (m-1) 0.34 0.7 1.12 0.7 2.00 1.0 1.76 0.9 1.11 0.7 0.78 1.3 0.59 .2 1.00 .2 0.57 .4 1.67 3.00 .4 0.92 .5 0.92 .6 the Trujillo granite, which elsewhere hosts Sn-W mineralization. The data are shown in Table 3 and plotted on log-log frequency distribu- tion plots (Fig. 8). Data from the mineralized sites at La Parrilla, Arroyomolinos de Montanches, Albala and El Trasquilon (Table 3) show a similar intensity of vein development (one vein >1 m m m 1 ) , with vein thicknesses ranging from 2.0 to 2.5 orders thickness (mm) Fig. 8. Log-log plot of the number of veins per metre (N) v. thickness (t) for selected samples of veins from the Caceres district: Trujillo, unmineralized veins; Albala, partially mineralized quartz veins; La Parrilla, quartz-scheelite veins from worked ore deposit [after Roberts et al. (1998)]. of magnitude. At all four localities the bulk of the data conform to a power-law distribution, with a tendency for the thicker veins to depart from the power-law distribution (Fig. 8). This may in part be attributed to truncated samples, but probably reflects an approach to some upper limit of vein opening. The former mining locality of La Parrilla shows the lowest D-value of 0.7, with vein thicknesses ranging between 1 and 400mm. Thus, thickness distributions from the variably mineralized sites contrast with those from the unmineralized sites within the Trujillo granite, by showing lower D-values, indicating a greater influence of thicker veins and breaks within the vein thickness distributions. The data obtained for the Sn-W mineralized veins are consistent with the models outlined above and with models for Sn-W mineralization which attempt to explain the development of the mineralization through the tapping of a fluid reservoir (Pollard & Taylor 1986). Moreover, recent geochemical modelling (Heinrich 1995) suggests that for Sn-bearing vein systems to develop, fluid flow must be focused within a restricted part of the fracture network in order to prevent reaction with wall rocks which would otherwise lead to Sn deposition at a much earlier stage. The development of a critical cluster of connected fractures is required to tap the fluid reservoir in an efficient way (Roberts et al. 1998). The focusing of the fluid flow on the backbone of the critical cluster (see Stauffer & Aharony 1992) provides the necessary channelling of the fluid to prevent significant fluid-rock reaction at an early stage. The vein thickness data suggest that fractal analysis of Sn-W mineralized vein systems can provide information to help inter- pret the formation of Sn-W ore bodies. Sn-W veins, Caceres FRACTAL ANALYSIS AND PERCOLATION PROPERTIES OF VEINS 15 Thickness distributions and alteration Models predicting increasing fracture connectiv- ity and enhanced fluid flow suggest that an increase in the alteration may well be observed at sites with well-connected fracture networks; assuming that an impermeable armour of vein material does not form along the walls of the fracture. A geochemical traverse was completed alongside vein thickness measurements for a 60m road section at Castello Branco, Portugal. Using alteration indices, such as relative deple- tion/enrichment of major elements (SiO2, Na2O, A12O3, K2O and TiO2), mass balance calcula- tions (isocon diagrams), XRD analysis and ele- mental analysis (C and N), two distinct zones of enhanced alteration were recognized within the slates. These zones tended to show a loss of Si, Na and Ca and an increase in K with the accompanying development of a two-mica plus chlorite assemblage. The vein thickness data were then divided into two subsamples accord- ing to the presence or absence of local alteration (Fig. 9). Overall, the distributions of vein thicknesses in the Castello Branco section suggest the presence of a well connected system. Closer inspection of the altered zones shows a distribu- tion with an increased vein density and/or thickness compared to the overall section. This observation would appear to suggest that the more pervasive alteration is related to rock volumes with greater fracture permeability, due to increased density and/or fracture opening (aperture) (see Sanderson & Zhang 1999). The Fig. 9. Vein thickness distributions from a line transect in the Castello Branco area, Portugal. Data are subdivided according to recognized alteration of the host rock. The areas of alteration are characterized by a vein population with an increased vein density. data are certainly consistent with the more intense alteration related to interaction with an externally derived fluid and a high fluid flux (fluid-rock interaction) through the zone. Summary and conclusions The fractal and percolation properties of veins provide an important insight into the role of fracture connectivity in the formation of vein systems and vein-style ore deposits. Measure- ment of the fractal dimension of vein thick- ness distributions provides a powerful descrip- tive parameter and a potential predictor of connected vein systems. Observation of natural systems and models suggests that vein sys- tems initiate due to opening of isolated frac- tures. At this stage, only locally derived fluid enters fractures, with negligible fluid migra- tion. As veins grow, a percolation threshold is reached at which point a connected vein system develops through the rock mass. Data from mineral deposits and damage zones of faults indicates that connected vein networks are char- acterized by power-law distributions and by lower D-values than unmineralized networks. The development of the connected vein net- work localizes deformation and fluid flow. This allows fluid to be transported over increased lengths and increases the likelihood of external fluids entering the system. The transported fluids produce vein-hosted ore deposits where physi- cal and chemical conditions are suitable for mineral precipitation. Thus, thickness distribu- tions of mineralized vein systems should show low D-values, recognition of which can serve as a guide to prospecting in such systems. This work was carried out as part of a grant (GR9/ 1485) funded by NERC. Joe Cartwright and Paul Gillespie are thanked for detailed and constructive reviews of earlier versions of the manuscript. References AN, L.-J. & SAMMIS, C. G. 1996. A cellular automaton for the development of crustalshear zones. Tectonophysics, 253, 247-270. BARTON, C. A. & ZOBACH, M. D. 1990. Self-similar distribution of macroscopic fractures at depth in crystalline rock in the Cajon Pass scientific drill hole. In: BARTON, N. & STEPHANSSON, O. (eds) Rock Joints. Balkema, Rotterdam, 163-170. CLARK, M. B., BRANTLEY, S. L. & FISHER, D. M. 1995. Power-law vein-thickness distributions and positive feedback in vein growth. Geology, 23, 975-978. 16 S. ROBERTS, D. J. SANDERSON & P. GUMIEL DEE, S. J. & ROBERTS, S. 1993. Late kinematic gold mineralization and the role of anomalous nitro- gen; an example from the La Codosera area, SW Spain. Mineralogical Magazine, 57, 437-450. HEINRICH, C. A. 1995. Geochemical evolution and hydrothermal mineral deposition in Sn (-W-Base Metal) and other granite related ore systems: some conclusions from Australian examples: In: THOMPSON, J. F. H. (ed.) Magmas, Fluids and Ore Deposits. Mineral Association of Canada Short Course Series, 23, 203-220. JOHNSTON, J. D. & MCCAFFREY, K. J. W. 1996. Fractal properties of vein systems and the variation of scaling relationships with mechanism. Journal of Structural Geology, 18, 349-358. MCCAFFREY, K. J. W. & JOHNSTON, J. D. 1996. Fractal analysis of mineralized vein deposit: Curraghinalt gold deposit, County Tyrone. Mineralium Deposita, 31, 52-58. MURPHY, P. J. & ROBERTS, S. 1997. Evolution of a metamorphic fluid and its role in lode gold mineralization in the Central Iberian Zone. Mineralium Deposita, 32, 459-474. NUR, A. 1982. The origin of tensile fracture linea- ments. Journal of Structural Geology, 4, 31-40. PETTKE, T. & DIAMOND, L. W. 1997. Oligocene gold quartz veins at Brusson, NW Alps: Sr isotopes trace the source of ore-bearing fluid to over 10-km depth. Economic Geology, 92, 389-406. PICKERING, G., BULL, J. M. & SANDERSON, D. J. 1996. Scaling of fault displacements and implications for estimation of sub-seismic strain: In: BUCHA- NAN, P. G. & NIEUWLAND, D. A. (eds) Modern Developments in Structural Interpretation, Valida- tion and Modelling. Geological Society, Lodnon, Special Publications, 99, 11-26. POLLARD, P. J. & TAYLOR, R. G. 1986. Progressive evolution of alteration and tin mineralization: controls by interstitial permeability and fracture- related tapping of magmatic fluid reservoirs in tin granites. Economic Geology, 81, 1795-1800. ROBERTS, S.. SANDERSON, D. J. & GUMIEL, P. 1998. Fractal analysis of Sn-W mineralization from Central Iberia: Insights into the role of fracture connectivity in the formation of an ore deposit. Economic Geology, 93, 360-365. DEE, S. J. & GUMIEL. P. 1991. Tectonic setting and fluid evolution of auriferous quartz veins from the La Codosera area. Western Spain. Economic Geology, 86, 1012-1022. SANDERSON, D. J. & ZHANG, Z. 1999. Critical stress localization of flow associated with deformation of well-fractured rock masses, with implications for mineral deposits. This volume. , ROBERTS, S. & GUMIEL, P. 1994. A fractal relationship between vein thickness and gold grade in drill core from La Codosera. Spain. Economic Geologv, 89. 168-173. SEGALL, P. & POLLARD, D. D. 1980. Mechanics of discontinuous faults. Journal of Geophysical Research, 85. 4337-4350. STAUFFER, D. & AHARONY, A. 1992. Introduction to Percolation Theory. Taylor & Francis. Bristol. VERMILYE, J. M. & SCHOLZ, C. H. 1995. Relation between vein length and aperture. Journal of Structural Geology, 17, 423-434. WILKINSON. J. J. & JOHNSTON, J. D. 1996. Pressure fluctuations, phase separation, and gold precipita- tion during seismic fracture propagation. Geology* 24, 395-398. ZHANG, X. & SANDERSON, D. J. 1994. Fractal struc- ture and deformation of fractured rock masses: In: KRUHL, J. H. (ed.) Fractals and Dynamic Sys- tems in Geosciences. Springer, Berlin. Geometry and population systematics of a quartz vein set, Holy Island, Anglesey, North Wales JULIA F. W. STOWELL, ADRIAN P. WATSON & NEIL F. C. HUDSON Division of Earth Sciences, School of Environmental and Applied Sciences, University of Derby, Kedleston Road, Derby DE22 1GB, UK Abstract: A steeply dipping set of quartz-chlorite-muscovite-biotite veins was emplaced, at depths of at least 14km and temperatures >400°C, into Mona Complex metasedimentary rocks of Holy Island, after the D4 event. Vein trends range from northeast-southwest to north-northwest-south-southeast. Vein offshoots, consistently oriented c. 10-35° anticlock- wise relative to the main vein, are common. Quartz fibres within different veins show a range of orientations, from northwest-southeast to east-west, and are interpreted as tracking vein opening directions. Vein and fibre orientations are integrated into a four-stage model for vein emplacement. At each stage, new main veins either open extensionally, with fibres subnormal to their boundaries, or by hybrid extension-shear. Hybrid main veins commonly have extensional offsoots. Extensional main veins are orientated progressively further anticlockwise with time from a dominantly northeast-southwest trend to a dominantly north-south trend, but hybrid main veins have out-of-sequence orientations and probably utilized pre-existing fractures. The dominance of extensional and hybrid extension-shear vein opening indicates that the differential stress was small and pore fluid pressure was high. A study of length-thickness relationships has revealed that this vein set is self-affine with vein widening progressing more quickly than elongation during growth. Linear transect analyses show that the vein spacings have fractal characteristics. Vein arrays are a ubiquitous feature of regional metamorphic terrains and are an important indication that large quantities of fluids are channelled through the crust during tectonic events (Walther & Orville 1982). Recent studies of vein arrays have largely focused on three aspects: the geometry of the array (e.g. Sander- son et al 1994; Johnston & McCaffrey 1996; Smith 1996); the origin of the fluid from which the vein was precipitated (e.g. Boiron et al. 1996; Haggerty & Bottrell 1997); the fluid-rock inter- action at scales ranging from subgrain (Cathe- lineau et al. 1993) to whole terrains (Cox 1993). Ultimately, integration of information from these different fields of investigation will enhance the understanding of the development and growth of veins and mechanisms of fluid flow through the crust. This study concentrates on the geometry and population systematics of a vein array on Holy Island, Anglesey, North Wales. Preliminary observations on the fluid composition and conditions of emplacement are also presented. Geological setting Rocks from the New Harbour and South Stack Groups of the Monian Supergroup (Shackleton 1975) are exposed on the southwest coast of Holy Island (Fig. 1), Anglesey. Both groups comprise a series of greenschist facies metasedi- ments, deposited as turbidite fans (Phillips 199la), with the New Harbour Group being intruded by a suite of serpentinites prior to the onset of deformation (Maltman 1977). Four main phases of deformation have been recog- nized in the New Harbour Group (Hudson & Stowell 1997) and the South Stack Group (Phillips 19916), although correlation between them is problematical (Hudson & Stowell 1997). Veins are common throughout both groups and have been divided into five sets, V1-V5 (Table 1), on the basis of the timing of their emplacement relative to the established struc- tural history. This paper examines the geometry and systematics of the V3 vein set. V3 veins V3 veins are common throughout southwest Holy Island and are ideal for a study of the development and growth of vein systems. The V3 veins developed post-D4 (Table 1), post-dating all the major deformation events, and their observed orientations reflect their emplacement orientations. In addition, V3 veins commonly
Compartilhar