Hydrocarbons in Contractional Belts

Hydrocarbons in Contractional Belts The Geological Society of London Books Editorial Committee Chief Editor Bob Pank...

Author: G. P. Goffey | J. Craig | T. Needham | R. Scott

215 downloads 1127 Views 10MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Hydrocarbons in Contractional Belts

The Geological Society of London Books Editorial Committee Chief Editor

Bob Pankhurst (UK) Society Books Editors

John Gregory (UK) Jim Griffiths (UK) John Howe (UK) Rick Law (USA) Phil Leat (UK) Nick Robins (UK) Randell Stephenson (UK) Society Books Advisors

Mike Brown (USA) Eric Buffetaut (France) Jonathan Craig (Italy) Reto Giere´ (Germany) Tom McCann (Germany) Doug Stead (Canada) Maarten de Wit (South Africa)

Geological Society books refereeing procedures The Society makes every effort to ensure that the scientific and production quality of its books matches that of its journals. Since 1997, all book proposals have been refereed by specialist reviewers as well as by the Society’s Books Editorial Committee. If the referees identify weaknesses in the proposal, these must be addressed before the proposal is accepted. Once the book is accepted, the Society Book Editors ensure that the volume editors follow strict guidelines on refereeing and quality control. We insist that individual papers can only be accepted after satisfactory review by two independent referees. The questions on the review forms are similar to those for Journal of the Geological Society. The referees’ forms and comments must be available to the Society’s Book Editors on request. Although many of the books result from meetings, the editors are expected to commission papers that were not presented at the meeting to ensure that the book provides a balanced coverage of the subject. Being accepted for presentation at the meeting does not guarantee inclusion in the book. More information about submitting a proposal and producing a book for the Society can be found on its web site: www.geolsoc.org.uk.

It is recommended that reference to all or part of this book should be made in one of the following ways: Goffey, G. P., Craig, J., Needham, T. & Scott, R. (eds) 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348. Cook, B. S. & Thomas, W. A. 2010. Ductile duplexes as potential natural gas plays: an example from the Appalachian thrust belt in Georgia, USA. In: Goffey, G. P., Craig, J., Needham, T. & Scott, R. (eds) 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 57–70.

GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 348

Hydrocarbons in Contractional Belts

EDITED BY

G. P. GOFFEY PA Resources UK Ltd, UK

J. CRAIG Eni Exploration and Production Division, Italy

T. NEEDHAM Needham Geoscience Ltd, UK

and R. SCOTT CASP, University of Cambridge, UK

2010 Published by The Geological Society London

THE GEOLOGICAL SOCIETY The Geological Society of London (GSL) was founded in 1807. It is the oldest national geological society in the world and the largest in Europe. It was incorporated under Royal Charter in 1825 and is Registered Charity 210161. The Society is the UK national learned and professional society for geology with a worldwide Fellowship (FGS) of over 9000. The Society has the power to confer Chartered status on suitably qualified Fellows, and about 2000 of the Fellowship carry the title (CGeol). Chartered Geologists may also obtain the equivalent European title, European Geologist (EurGeol). One fifth of the Society’s fellowship resides outside the UK. To find out more about the Society, log on to www.geolsoc.org.uk. The Geological Society Publishing House (Bath, UK) produces the Society’s international journals and books, and acts as European distributor for selected publications of the American Association of Petroleum Geologists (AAPG), the Indonesian Petroleum Association (IPA), the Geological Society of America (GSA), the Society for Sedimentary Geology (SEPM) and the Geologists’ Association (GA). Joint marketing agreements ensure that GSL Fellows may purchase these societies’ publications at a discount. The Society’s online bookshop (accessible from www.geolsoc. org.uk) offers secure book purchasing with your credit or debit card. To find out about joining the Society and benefiting from substantial discounts on publications of GSL and other societies worldwide, consult www.geolsoc.org.uk, or contact the Fellowship Department at: The Geological Society, Burlington House, Piccadilly, London W1J 0BG: Tel. þ44 (0)20 7434 9944; Fax þ44 (0)20 7439 8975; E-mail: [email protected]. For information about the Society’s meetings, consult Events on www.geolsoc.org.uk. To find out more about the Society’s Corporate Affiliates Scheme, write to [email protected]. Published by The Geological Society from: The Geological Society Publishing House, Unit 7, Brassmill Enterprise Centre, Brassmill Lane, Bath BA1 3JN, UK (Orders: Tel. þ44 (0)1225 445046, Fax þ44 (0)1225 442836) Online bookshop: www.geolsoc.org.uk/bookshop 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. # The Geological Society of London 2010. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of The Copyright Licensing Agency Ltd, Saffron House, 6 –10 Kirby Street, London EC1N 8TS, UK. Users registered with the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, USA: the item-fee code for this publication is 0305-8719/10/$15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 978-1-86239-317-2 Typeset by Techset Composition Ltd, Salisbury, UK Printed by MPG Books Ltd, Bodmin, UK Distributors North America For trade and institutional orders: The Geological Society, c/o AIDC, 82 Winter Sport Lane, Williston, VT 05495, USA Orders: Tel. þ1 800-972-9892 Fax þ1 802-864-7626 E-mail: [email protected] For individual and corporate orders: AAPG Bookstore, PO Box 979, Tulsa, OK 74101-0979, USA Orders: Tel. þ1 918-584-2555 Fax þ1 918-560-2652 E-mail: [email protected] Website: http://bookstore.aapg.org India Affiliated East-West Press Private Ltd, Marketing Division, G-1/16 Ansari Road, Darya Ganj, New Delhi 110 002, India Orders: Tel. þ91 11 2327-9113/2326-4180 Fax þ 91 11 2326-0538 E-mail: [email protected]

Contents GOFFEY, G. P., CRAIG, J., NEEDHAM, T. & SCOTT, R. Fold –thrust belts: overlooked provinces or justifiably avoided?

1

ROEDER, D. Fold –thrust belts at Peak Oil

7

HILL, K. C., LUCAS, K. & BRADEY, K. Structural styles in the Papuan Fold Belt, Papua New Guinea: constraints from analogue modelling

33

COOK, B. S. & THOMAS, W. A. Ductile duplexes as potential natural gas plays: an example from the Appalachian thrust belt in Georgia, USA

57

TURNER, S. A., COSGROVE, J. W. & LIU, J. G. Controls on lateral structural variability along the Keping Shan Thrust Belt, SW Tien Shan Foreland, China

71

ROURE, F., ANDRIESSEN, P., CALLOT, J. P., FAURE, J. L., FERKET, H., GONZALES, E., GUILHAUMOU, N., LACOMBE, O., MALANDAIN, J., SASSI, W., SCHNEIDER, F., SWENNEN, R. & VILASI, N. The use of palaeo-thermo-barometers and coupled thermal, fluid flow and pore-fluid pressure modelling for hydrocarbon and reservoir prediction in fold and thrust belts

87

CAPOZZI, R. & PICOTTI, V. Spontaneous fluid emissions in the Northern Apennines: geochemistry, structures and implications for the petroleum system

115

RODRIGUEZ-ROA, F. A. & WILTSCHKO, D. V. Thrust belt architecture of the central and southern Western Foothills of Taiwan

137

HESSE, S., BACK, S. & FRANKE, D. Deepwater folding and thrusting offshore NW Borneo, SE Asia

169

Index

187

Fold –thrust belts: overlooked provinces or justifiably avoided? GRAHAM P. GOFFEY1*, JONATHAN CRAIG2, TIM NEEDHAM3 & ROBERT SCOTT4 1

PA Resources UK Limited, Waterfront, Winslow Road, Hammersmith, London W6 9SF, UK

2

Eni Exploration & Production Division, via Emilia 1, 20079 San Donato Milanese, Milan, Italy 3

Needham Geoscience Limited, Lynnthorpe, 4 Easby Drive, Ilkley LS29 9BE, UK 4

CASP, West Building, 181A Huntingdon Road, Cambridge, CB3 0DH, UK *Corresponding author (e-mail: [email protected])

Abstract: This volume results from a conference intended to assess the exploration and exploitation primarily of onshore fold–thrust belts. These are commonly perceived as ‘difficult’ places to explore and therefore are often avoided by companies. However, fold– thrust belts host large oil and gas fields and barriers to effective exploration mean that substantial resources may remain. This volume shows how evaluation techniques have developed over time. It is possible in certain circumstances to achieve good 3D seismic data. Structural restoration techniques have moved into the 3D domain and simple thermal constraints can be enhanced by using more sophisticated palaeothermal indicators to more accurately model burial and uplift evolution of source and reservoirs. Awareness of the influence of pre-thrust structure and stratigraphy and of hybrid thick and thinskinned deformation styles is supplementing the simplistic thin-skinned fault-bend and fault propagation models employed in earlier exploration. The ‘learning curve’ in fold –thrust belt exploration has not been steep and further improvement seems likely to be a slow, expensive and iterative process with information from outcrop, well penetration and slowly improving seismic data. Industry and academia need together to develop and continually improve the necessary understanding of subsurface geometries, reservoir and charge evolution and timing.

This Geological Society Special Publication contains a selection of the papers presented at a conference on ‘Fold–Thrust Belt Exploration’, held in London from 14 to 16 May 2008. The conference was conceived by Enzo Zappaterra and organized by the Petroleum Group of the Geological Society of London, in collaboration with the Geological Society’s Tectonic Studies Group. The conference was intended to assess the current state of the scientific evaluation, exploration and exploitation of onshore contractional ‘fold –thrust’ belts in both academia and industry. A primary objective of the conference was to consider whether fold–thrust belts could still host substantial undiscovered hydrocarbon resources which might be located through commercial exploration efforts. Put another way, given what we know, are fold– thrust belt provinces overlooked or justifiably avoided? Onshore fold–thrust belts are typically perceived as ‘difficult’ places to explore for several reasons (which have been easy for exploration management to perceive as prohibitive). Typically, rugged terrain makes access difficult and substantially increases the cost of exploration. Seismic data quality is often insufficient to allow unambiguous interpretation or to differentiate between alternative models. Reservoir development and

charge history are frequently difficult to unravel because of uncertainty regarding the burial history, while the tectonic regime tends to lead to a perceived high risk of trap breach. On the other hand, some fold– thrust belts do host very large oil and gas fields. Do the technical and logistical barriers to effective exploration mean that there are substantial hydrocarbon resources still to be found in the world’s fold–thrust belts? The Fold –Thrust Belt Exploration conference was intended to address this question by drawing on experience from both industry and academia. Exploration and exploitation techniques used to evaluate the hydrocarbon potential of fold–thrust belts have in many respects evolved substantially over the last few decades. Seismic imaging and interpretation tools have improved significantly, section restoration and balancing software has evolved into the 3D domain, and computer modelling of burial history and fluid flow in the context of tectonic loading/unloading and shortening has been developed. Simultaneously, the oil and gas industry has moved into deepwater provinces where contractional tectonic styles are common and where high quality 3D seismic imaging has allowed richly detailed analysis of structural development. Have these improvements produced sufficient new

From: Goffey, G. P., Craig, J., Needham, T. & Scott, R. (eds) Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 1–6. DOI: 10.1144/SP348.1 0305-8719/10/$15.00 # The Geological Society of London 2010.

2

G. P. GOFFEY ET AL.

insights to reinvigorate exploration under higher oil prices or are the negative perceptions of the challenges of onshore exploration in contractional regimes still justified?

Do fold – thrust belts host substantial undiscovered hydrocarbon resources? Figure 1 shows the worldwide distribution of foreland fold–thrust belts. Roeder (2010) notes that fold–thrust belts contain more than 700 billion barrels of oil equivalent (BOE) of known hydrocarbon reserves. The Zagros fold– thrust belt contains the vast majority of these reserves (c. 517 billion BOE), with the remaining 183 billion BOE distributed across some 30 other fold–thrust belts. Even with the Zagros fold–thrust belt excluded, much of the reserve is concentrated in just a few provinces and the reserves distribution has a long ‘tail’, with some 23 fold– thrust provinces each having established reserves of between 2 and 5 billion BOE. In this context Roeder asserts that much of the yet-to-find reserves in the world’s fold–thrust belts will inevitably be in relatively modest sized fields of, perhaps, less than 100 million BOE, in fold– thrust belts which host a few billion barrels of oil equivalent each. The Zagros fold– thrust belt is clearly a key area for hydrocarbon exploration in fold–thrust belts. In an unpublished presentation at the conference, Sepehr et al. and colleagues from the National Iranian Oil Company showed profiles from 3D seismic surveys over several fields. These illustrate a level of structural complexity not portrayed in the limited published literature, including relatively recent discoveries in sub-thrust traps. A careful study of structural timing in the Zagros belt in southeastern Lurestan, Iran, by Blanc of StatoilHydro and co-workers from National Iranian Oil Company and CSIC Barcelona, indicates that the initial growth of many of the large folds pre-dates deposition of Eocene – Early Miocene strata. A previously unrecognized phase of Early Palaeogene compression is interpreted to account for up to half of the measured shortening. Early structural growth has also been demonstrated in the Dezful Embayment and foreland basin area of the Zagros (Abdollahie Fard et al. 2006) as well as the Lurestan. This can be dated as latest Cretaceous and results from the interaction of basement structures on ‘Arabian’ and ‘Zagros’ trends. The structures extend into the NW Persian Gulf (Soleimany & Saˆbat 2010). The major implication of this work is that some areas of the Zagros previously considered unprospective, because structural development was thought to post-date hydrocarbon charge, may in fact be locally prospective. In

contrast to Roeder’s conclusions of only modest yet-to-find resources in the globally prevalent lower ranked fold–thrust belts, these interpretations hint that a substantial yet-to-find resource may still exist in the world’s most richly endowed fold– thrust belt, the Zagros. In another unpublished conference presentation, Cooper observed that the six largest fold–thrust belts in the world have different structural characteristics, implying that deformation style is not a critical factor in the resource density of these belts. The volume of hydrocarbon resources in any given fold–thrust belt seems to be more closely linked to the presence of an effective source rock. Fold – thrust belts tend to have a high density of structural traps and this was seen by Cooper as the differentiating factor from other structural provinces. Roseway, in another unpublished conference paper, presented a comparative study of 33 foreland basins and fold– thrust belts. This showed that the overwhelming majority of source rocks in these systems are Cretaceous –Tertiary oil-prone marine shales. The dominant top seal is mudstone, some 40% of reservoirs are shelf carbonates and 40% are marine shelf and continental sandstones. Some 60% of fold– thrust belt reservoirs examined in this study are fractured, with unfractured carbonate reservoirs being uncommon.

What techniques are being employed in fold – thrust belt exploration and development and how are these developing? Section modelling and restoration of fold–thrust belts have moved emphatically into the 3D domain. At the conference, Gibbs and co-workers from Midland Valley Exploration, discussed how traditional section restoration and balancing algorithms using flexural slip, simple shear, fault-parallel flow and mixed-mode combinations introduce bias into kinematic prediction and interpretation style, and that such algorithms are often selected on the basis of limited observations. Dominantly planestrain assumptions allow these approaches to be applied in 3D restoration but further limit their utility. In the view of Gibbs, 3D approaches are still powerful, notwithstanding these limitations, and techniques are evolving towards kinematic predictions that are not limited by plane-strain assumptions or by user choice of tectonic transport direction. Analogue modelling is also providing valuable insights into the evolution of fold–thrust belts. At the conference, Sassi and co-workers from the Institut Français du Pe´trole (IFP) presented the results of numerical digitization of a sandbox experiment that modelled the evolution of force folding within a

180°W

150°W

120°W

90°W

60°W

30°W

0°

30°E

60°E

90°E

120°E

150°E

180°E

Conventional Oceanic Plate Boundaries & Motions: Mid Ocean Ridge & Transform Faults

80°N

80°N

Oceanic Subduction Plate Velocities: a. Diverging b. Converging

ICE

a

b

Oceanic Crust (undifferentiated) Oceanic Plates

Brooks Range Parry Islands Timan-Pechora

70°N

70°N

Mostly Oceanic Plates Associated with Backarcs

PA - Pacific Plate NZ - Nazca Plate CO - Cocos Plate JF - Juan De Fuca Plate CAL - Caroline Plate

PH - Philippine Plate CA - Carribbean Plate SC - Scotia Plate

Continental Boundaries & Structures:

Baikal-Predpatom

Passive Ocean / Continent Boundary

60°N

60°N

Alps-Carpathians

B.C. Rockies 50°N

N. Apennines

Alberta Rockies

Basement - Involved Uplifts and/ or Rift Inversions

50°N

Diffuse Plate Boundaries:

NC

Salt Range-Potwar T

Sevier & U.S. Rockies

On continents and close to Subduction Zones (mostly subaerial) Deformation inferred from seismicity, topography, faulting, etc.

40°N

Circum-Sichuan

Appalachians

20°N

Zongolica-Veracruz

PA

Atlas Mts

E. Venezuela CA

Chiapas-Reforma

AR

IN

AF/NU

Trinidad

30°N

Y

S. Apennines

Sierra Madre Oriental Circum-Maracaibo

10°N

OK

Zagros Mts

40°N

30°N

Tertiary Rifts

Caucasus

Pyrenees

NA

JF

EU

Taiwan

0°

New Guinea CAR W-C. Venezuela

MDD-Alto Beni

20°S

20°N

associated with breakup of Gondwana parts of Pangea SA - South America

AR - Arabia

AF/NU - Africa / Nubia Plate

IN - India

AF/SOM - Africa / Somalia Plate LW - Lwandle

CAP - Capricorn AU - Australia

Continent/Ocean Plates

0°

NA - North America

associated with breakup of Gondwana parts of Pangea

EU - Eurasia*

*Including Ill-defined sub-plates within diffuse plate boundaries as follows: O - Okhotsk T - Tarim NC - North China

Y - Yangtse I - Indonesia B - Borneo

Folded (Orogenic) Belts

Santa Cruz-Tarija

NZ

AU

30°S

Submarine Regions where non-closure of plate circuits indicates measurable deformation; deformation mostly inferred from seismicity.

Continent/Ocean Plates

10°N

AF/SOM

SA

10°S

PA

PH

Assam-Arakan

CO Llanos Foothills

Continental (A) Subduction Boundary

Cuyo

LW

Neuquen Foothills 40°S

Oceanic (B) Subduction Boundary

Oceanic Basins

Continental (A) Subduction Boundary

Oceans

Basement Involved Uplifts and /or Inversions

Oceanic Back-Arc Basins

Foreland Folded Belts 50°S

50°S

SC 60°S

60°S

AN

PLATE TECTONIC SETTING Continent-continent

Continent-arc

Non-collisional

ICE 180°W

150°W

120°W

90°W

60°W

30°W

0°

30°E

60°E

90°E

120°E

150°E

Ages of Deformation/ Basement

Sedimentary Basins on Continental and Transitional Crust

Cenozoic / Mesozoic

Foreland Basins

Vendian to Paleozoic

Deep Sea Trenches

Precambrian Basement

Episutural Basins

180°E

Fig. 1. Index map showing orogenic belts worldwide, age of continental crust and distribution of foreland fold– thrust belts. Open red boxes denote areas covered by specific papers in this volume. Map courtesy of Albert Bally and C & C Reservoirs.

FOLD– THRUST BELTS

transpressive regime. The experiment was used to create an evolving 3D model of interfaces and faults and to describe the 3D geometric evolution over time. Hill et al. (2010) used centrifuge analogue modelling to examine structural styles in the Papua New Guinea Fold Belt, by varying the mechanical stratigraphy until the structural styles generated match those interpreted in the subsurface. Analogue models with pre-cut faults, albeit of appreciably lower angle than would be expected for inherited extensional faults, produced early inversion structures similar to that observed at the Kutubu Oilfield, suggesting that pre-thrust weaknesses may play a significant role in the development of the Papua New Guinea Fold Belt. Such techniques provide geometric and kinematic insights which can assist in the interpretation of seismic data. As always, information from outcrop studies is extremely important, especially where integrated with other data. Wilson and colleagues from Oil Search Limited presented a paper on the Moran Field in Papua New Guinea at the conference. Geological complexity here is the result of shortening being influenced by pre-existing structures and partial detachment, within Early Cretaceous mudstones, between the oil-bearing Upper Jurassic reservoirs and the Miocene limestone carapace at surface. Seismic data quality is poor over the field due to a combination of steep and variable dips and the presence of a rugged surface topography of karstified limestone. Careful integration of multiple datasets and techniques, including 87Sr/86Sr isotopic dating of the limestones and the use of produced oil volume mass balance techniques, has helped define cross-cutting faults at reservoir level which play a significant role in oil entrapment and production. It was conspicuous in the conference that it is still relatively rare for subsurface models and geometries to be adequately constrained by seismic data. In a paper on the Kutubu Oilfield in Papua New Guinea, Bradey and colleagues from Oil Search Limited showed that five 2D seismic lines, of variable quality, had been acquired across the Kutubu field, some years after the commencement of production, at a cost of US$ 80 000/km. Seismic acquisition in Papua New Guinea is notoriously expensive, even by fold–thrust belt standards, due to the rugged karstified terrain, dense rain forest and absence of access routes. Kutubu exemplifies the perceptions of the utility of seismic data in these terrains: expensive, difficult and generally of poor quality. However, in an unpublished paper, Duque and colleagues from BP gave some reason for optimism that high quality seismic imaging can eventually be achieved in fold–thrust belts. Based on BP’s

3

experience in the Llanos foothills of Colombia, the authors showed that there have been improvements in acquisition and processing techniques over time. On the acquisition side, improvements in data quality have resulted from increased reliability of the recording equipment and the ability to handle substantially more channels. On the processing side, the key issues are statics, velocities, noise attenuation, migration and imaging. Specific techniques which have been helpful in improving seismic data quality include refraction tomography to address statics issues and pre-stack imaging in 3D datasets. Derivation of the appropriate velocity field requires a good understanding of the geology and is an iterative process that evolves as well data are acquired and seismic data quality progressively improves. Duque et al. showed that an iterative approach to processing, using several different processing contractors working in parallel on the same dataset, enabled BP to achieve outstanding success in imaging to depth in the Cusiana– Cupiagua area of the Llanos foothills. Comparison of the latest seismic dataset with earlier processed versions of the same dataset in this area show that it is possible to acquire high quality seismic data in onshore fold–thrust belts, but it takes significant time and very substantial effort. Several speakers at the conference and authors of papers in this volume presented case studies in which the understanding of structural geometries evolved with both increasing data and the use of techniques that allowed model development to move beyond well-established fault-bend and fault propagation fold geometries. For example, Newson demonstrated the value of continued re-evaluation of existing models in the Moose Mountain Field in the Canadian fold– thrust belt. The Moose Mountain Field had previously been interpreted using a faultbend fold model, but re-evaluation using balanced sections, dip-domain analysis and down-plunge projection indicated a detachment fold origin with several thrust sheets forming an antiformal stack. The re-interpretation resulted in new pool gas discoveries that have doubled the original field size. Similarly, in a paper on the implications of a recently drilled deep geothermal well in the Variscan Aachen fold and thrust belt in the Rhenish Massif in Germany, Becker from RWTH Aachen University, showed how integration of core and wireline log data allowed the revision of interpreted structural geometry at the front of the thrust belt. The implications of this re-interpretation include the recognition that the displacement of individual thrust sheets is probably much less than previously estimated. The evolution in structural understanding of fold– thrust belt fields with expanding datasets appears to be common. Hill et al. (2010) reviewed

4

G. P. GOFFEY ET AL.

how the structural model of the Kutubu field in Papua New Guinea has evolved from a simple thinskinned fault-bend fold, through an overturned fault propagation fold, with break-thrust and small-scale duplexes at reservoir and seal level respectively, to a hybrid model of apparently inversion-related basement uplift overlain by thin-skinned folds. This evolution reflects the progressive incorporation of data from 50þ wells and eight seismic lines acquired over the field during its development. Cook & Thomas (2010), in the context of an interpreted ductile duplex developed within a recess in the Appalachian thrust belt, draw attention to the apparent role of pre-thrust extensional structures in the thickening of basal weak shales within a deep and subsequently inverted thick-skinned graben. This extensional mechanism is considered the most likely mechanism for the observed substantial thickening of ductile shales below the main roof thrust and its recognition has provided a possible solution for volume balance problems encountered in the palinspastic restoration of sections across the recess. The role of pre-existing extensional structures was also assessed by Turner et al. (2010) in the Keping Shan thrust belt of NW China. Here preexisting extensional structures give rise to partitioning of the belt and provide marked stratigraphic breaks which strongly influence the structural architecture of the later thrust belt. Bump et al. from BP in another unpublished conference paper drew attention to the variation in structural complexity in the Llanos foothills of Colombia, depending on whether or not the interpretation is constrained by well data. Sections drawn without well control tend to show simple structures with long, continuous horses, whereas those drawn with well control tend to show tighter folds, more faulting and ramps and flats within horses. A significant limitation of 2D seismic data in this area is the tendency for apparent frontal structures to be either seismic artefacts or to be too shallow to form traps at reservoir level. Roeder (2010) draws attention to examples in the East Venezuelan fold belt, where the interpretation of structural geometries has been compromised by unrecognized limitations in the application of fault-bend fold models. It appears there has been a historical reliance on thin-skinned geometric models in the interpretation of fold–thrust belts. These models were originally developed in predominantly thin-skinned belts such as the Rocky Mountains, where the pre-deformation stratigraphy was treated as having been deposited in a broadly consistent layer cake on a relatively undeformed ramp margin. The Rocky Mountains are now sufficiently well described to show that the effects of lateral facies changes can in fact be incorporated into sections. For example, dominated by folds to the north and with ‘classic’ large thrust

sheets of carbonate ‘beams’ in the south (McMechan 1985; Spratt et al. 2004). Basement control is also now recognized here so that there is a move away from a pure thin-skinned approach; this is particularly the case for structures lying at a high angle to the thrust belt (Fermor 1999). Thin-skinned models have, however, historically been applied widely to fold–thrust belts even though thin-skinned shortening may not be the primary shortening mechanism. In areas where pre-existing extensional and early low strain inversion structures provide strong control on structural development, thin-skinned geometric models may not be applicable or only of value with considerable adaptation. Recognition of the limitations of applying a narrow suite of models and awareness that structural geometries are often strongly influenced by pre-existing weaknesses and variations in the mechanical stratigraphy and reflect thick-skinned inversion, or hybrid thick-skinned inversion with thin-skinned deformation in the carapace, may only come after significant, and perhaps unsuccessful, exploration efforts. The implications in terms of magnitude of shortening and the nature, timing and location of potential traps are profound, yet are often overlooked. Would the progressive improvement in subsurface understanding in Papua New Guinea described by Hill et al. (2010) have taken place if early drilling had not found several giant fields and so given a commercial imperative to develop a thorough understanding of the subsurface and to the drilling of more exploration, appraisal and development wells that helped constrain the structural models? Roeder (2010) highlights the existence of several fold belts where exploration has been based on very limited seismic datasets and with limited geometric models. He suggests that these areas may offer opportunities to reassess the remaining prospectivity. It is not difficult to envisage that some prospective fold–thrust belts may remain underexplored because early drilling was unsuccessful, understanding was limited by sparse data and, perhaps, by the use of inappropriate models, so that there has been little commercial imperative to develop the level of understanding achieved elsewhere. The timing of hydrocarbon charge and the palaeo-burial and palaeo-thermal history of source and reservoir rocks tends to be subject to wide uncertainty. Historically, analysis has been dependent on measurements of the maturity of organic matter, such as Ro and Tmax. Roure et al. (2010) observe that apatite fission track data and analyses of hydrocarbon-bearing fluid inclusions can reduce the range of uncertainty and allow better prediction of the timing of hydrocarbon generation and the pressure/temperature conditions in the reservoir during cementation and hydrocarbon trapping. Capozzi & Picotti (2010) show how the analysis

FOLD– THRUST BELTS

of brines and hydrocarbons leaking at the surface can further constrain the maturation –migration using examples from the Northern Apennines in Italy. Wiltschko & Roa (2010), with reference to the southern Taiwan orogen, show how a multimodal approach to constraining cross sections using thermal maturity and thermochronological data in addition to GPS velocities allow the influence of large pre-existing normal faults to be determined and cross sections to be better constrained. While papers and posters on deepwater fold– thrust belts were presented at the conference, the focus of the conference was on onshore fold– thrust belts and this focus is reflected in the content of this volume. However, as a reminder that there is scope to transfer understanding between onshore and deepwater contractional belts, Back et al. (2010) present an analysis of the relative roles of crustal shortening versus gravity-driven shortening in the NW Borneo deepwater fold–thrust belt. Cross sections based on 2D seismic data are used to show that gravity-driven shortening decreases and basement-driven compression increases from south to north along the fold–thrust belt. Several contributions to the conference touched on the wider context and significance of fold– thrust belts for adjacent basins. Scott’s presentation on Novaya Zemlya was specifically on this topic. Although the Novaya Zemlya fold–thrust belt itself is not prospective, studies of the fold–thrust belt are valuable because it developed after subsidence had begun in important hydrocarbon basins on both the foreland and hinterland side. Fold– thrust belt geometry provides valuable information about the location and timing of this subsidence. With the development of a fold–thrust belt having the potential to affect sediment dispersal patterns, reservoir geometry and lithology, hydrodynamics and charging in adjacent, more-prospective basins, this aspect should not be overlooked.

Are fold – thrust belts set to re-emerge as a major new exploration target? In an unpublished conference paper, Graham presented a pessimistic view, observing that fold– thrust belts elevate rocks that were previously deeply buried and that this has a detrimental effect on key elements of their petroleum systems. Notwithstanding the potential for large traps and often limited prior exploration to condemn prospectivity, fold–thrust belts tend to exhibit complex internal geometries that are often inadequately imaged on poor quality seismic data. In essence, paraphrasing Graham, hydrocarbons trapped in fold–thrust belts are ‘Goldilocks’ accumulations, requiring special circumstances to allow structures to develop in time to

5

receive hydrocarbon charge and then to retain trap integrity as the fold–thrust belt continues to evolve. Roeder (2010), revisits several failed exploration ventures in fold–thrust belts to analyse whether they share common characteristics. In an idiosyncratic view, Roeder argues that many fold–thrust belts could be re-evaluated and exploration re-commenced, having been under-explored in the past with low drilling and seismic density and with wells that were often sited on the basis of geological models rather than observed data. However, based on a review of the distribution of reserves in fold–thrust belts worldwide, and notwithstanding the scope to re-explore, Roeder concludes that only the fold–thrust belts in the Middle East offer a globally significant target for finding new hydrocarbon resources.

Conclusions The technologies, techniques and models available to the exploration and production industry today are a significant improvement on those of the past and offer the possibility of more effective exploration in fold–thrust belts in the future. However, the improvement has been incremental, not a stepchange, and the establishment of even a basic, reliable understanding of regional and trap-scale structural geometries remains highly problematic in many fold–thrust belts. Limited or unreliable understanding of large-scale geometries will undermine the use of some of the available techniques. Undoubtedly, fold–thrust belts are still immensely challenging places in which to explore costeffectively. It seems likely that in the near to medium term, any significant success in fold– thrust belt exploration will be in areas where politics, not geology or technology, has been the barrier to effective exploration. This is well illustrated by recent successes in the Zagros belt in the Kurdistan region of northern Iraq. This area has been barely touched by industry for several decades and major undrilled traps are comparatively easy to locate. Elsewhere, despite all the progress in exploration and modelling techniques, and notwithstanding the undoubtedly high density of remaining traps, it seems that progress will continue to be slow. If hydrocarbon entrapment relies on a ‘not too hot, not too cold’ set of ‘Goldilocks’ circumstances, and while effective exploration remains an imprecise iteration of remote-sensed and field observations, loosely constrained by a few widely spaced well penetrations and expensive seismic data of limited quality, it will continue to be difficult to mitigate risk and reduce uncertainty in the exploration of fold–thrust belts. As exploration has moved away from areas with simple and predictable mechanical stratigraphy where the widely used

6

G. P. GOFFEY ET AL.

geometric models were first developed, to fold– thrust belts with more complex mechanical stratigraphy, pre-contractional features and where thick and thin-skinned structures have developed together, the understanding of the subsurface structure will continue to be very uncertain. The identification of the traps most likely to be hydrocarbon charged will, inevitably, remain a major challenge. Opportunities do exist in many fold–thrust belts to re-evaluate their hydrocarbon potential and to re-commence exploration from a hopefully more enlightened perspective, being fully aware of the possible failings of previous exploration. It would appear that, outside the Middle East, the potential yet-to-find resource in fold–thrust belts may be mostly in fields of tens to low hundreds of millions BOE size in provinces which currently host less than 5 billion BOE each. These will offer attractive possibilities for some companies but, given the cost and exploration risk, these areas will be avoided by many others as long as there are less challenging opportunities elsewhere. The learning curve in global fold–thrust belt exploration is not steep and as a result it seems likely that fold–thrust belts will still be capable of providing new discoveries long after the world’s shelf and deepwater basins are thoroughly explored. Progress, however, is likely to be slow and will involve an expensive iterative process of maximizing the lessons learned from all well penetrations, slow improvement in seismic data quality and industry working closely with academia to develop and continually improve the understanding of subsurface geometries, reservoir and charge evolution. Fold–thrust belt exploration is clearly not for the ‘faint-hearted’ and is unlikely to reward individuals and companies that are either unwilling or unable to devote the necessary time, skills and expenditure. Even with an appropriately resourced effort, the entrapment and preservation of hydrocarbons in fold–thrust belts seems to depend on ‘Goldilocks-like’ circumstances which will continue to challenge the combined predictive abilities of both industry and academia. It is, perhaps, still too early to judge whether fold–thrust belts are overlooked provinces or justifiably avoided, but barring the development of some unforeseen new enabling technology, a major shift in the focus of the oil and gas industry towards renewed exploration of the world’s fold–thrust belts does not seem likely.

References Abdollahie Fard, I., Braathen, A., Mokhtari, M. & Alavi, S. A. 2006. Interaction of the Zagros Fold– Thrust Belt and the Arabian-type, deep-seated folds in the Abadan Plain and the Dezful Embayment, SW Iran. Petroleum Geoscience, 12, 347–362.

Capozzi, R. & Picotti, V. 2010. Spontaneous fluid emissions in the Northern Apennines: geochemistry, structures and implications for the petroleum system. In: Goffey, G. P., Craig, J., Needham, T. & Scott, R. 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 115–135. Cook, B. S. & Thomas, W. A. 2010. Ductile duplexes as potential natural gas plays: an example from the Appalachian thrust belt in Georgia, USA. In: Goffey, G. P., Craig, J., Needham, T. & Scott, R. 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 57–70. Fermor, P. 1999. Aspects of the three-dimensional structure of the Alberta Foothills and Front Ranges. Geological Society of America Bulletin, 111, 317– 346. Hesse, S., Back, S. & Franke, D. 2010. Deepwater folding and thrusting offshore NW Borneo, SE Asia. In: Goffey, G. P., Craig, J., Needham, T. & Scott, R. 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 169–185. Hill, K. C., Lucas, K. & Bradey, K. 2010. Structural styles in the Papua New Guinea Fold Belt; constraints from analogue modelling. In: Goffey, G. P., Craig, J., Needham, T. & Scott, R. 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 33–56. McMechan, M. E. 1985. Low-taper triangle-zone geometry: an interpretation for the Rocky Mountains foothills, Pine Pass-Peace river area, British Columbia. Bulletin of Canadian Petroleum Geology, 33, 31– 38. Roeder, D. 2010. Fold–thrust belts at peak oil. In: Goffey, G. P., Craig, J., Needham, T. & Scott, R. 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 7– 31. Roure, F., Andriessen, P. et al. 2010. The use of palaeo-thermo-barometers and coupled thermal, fluid flow and pore-fluid pressure modelling for hydrocarbon and reservoir prediction in fold and thrust belts. In: Goffey, G. P., Craig, J., Needham, T. & Scott, R. 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 87–114. Soleimany, B. & Sa`bat, F. 2010. Style and age of deformation in the NW Persian Gulf. Petroleum Geoscience, 16, 31– 39. Spratt, D. A., Dixon, J. M. & Beattie, E. T. 2004. Changes in structural style controlled by lithofacies contrast across transverse carbonate bank margins— Canadian Rocky Mountains and scaled physical models. In: McClay, K. R. (ed.) Thrust Tectonics and Hydrocarbon Systems. AAPG Memoir, 82, 259–275. Turner, S., Cosgrove, J. W. & Liu, J. G. 2010. Controls on lateral structural variability in along the Keping Shan Thrust Belt, SW Tien Shan Foreland, China. In: Goffey, G. P., Craig, J., Needham, T. & Scott, R. 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 71–85. Wiltschko, D. V. & Rodriquez-Roa, F. A. 2010. Thrust belt architecture of the central and southern western foothills of Taiwan. In: Goffey, G. P., Craig, J., Needham, T. & Scott, R. 2010. Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 137– 168.

Fold –thrust belts at Peak Oil DIETRICH ROEDER Murnau Geodynamics, 9225 West Jewell Place No. 107, Lakewood, CO 80227, USA (e-mail: [email protected]) Abstract: Outside the Middle East, onshore fold–thrust belts (FTB) of Tertiary to Recent age contain a significant part of the globally developed petroleum, but far less of the oil and gas remaining undiscovered. Depending on high quality data and on deep drilling, renewed exploration of former failures is commercially attractive, and it will help in exploring the deepwater belts of compression. In FTB with a defined petroleum system, an under-explored trend may be the informally named ‘deep-updip’ or DUD trend. Shale-gas-prone formations in FTB require new exploration strategies, but in the public domain, this type of prospect has not yet been discussed. FTB discoveries require geological insight, persistence and exponentially rising investment. The paper includes examples from the Northern Alps, from the Llanos foothills of Colombia, from Eastern Venezuela and from the Po Valley basin of Italy.

Fold –thrust belts (FTB) are the shallow and partly petroliferous fringes of the global orogenic belts. Many FTB, their location, geographical extent and key elements of their structural style, as well as their stratigraphy and petrology, can be studied in mountainous exposures. However, their tectonic and economic understanding requires subsurface data. Beginning in earnest about 50 years ago, reflection seismography and deep drilling have been, and still are, steering the global exploration of the geologically youngest FTB. Significant academic merits notwithstanding, the present and future understanding of FTB depends on funding and other incentives offered by the petroleum industry. The geological advantage of petroleum systems in Tertiary and younger FTB settings (Marc Cooper, pers. comm. to DHR) over objectives in shallower basin parts is found in their more complete stratigraphic record, in a basin setting combining source and reservoir, and in the favourable thermal configuration of stacked thrust sheets. Peak Oil is a concept of global economy. It defines an economic state during which the demand for petroleum exceeds the technically possible rate of supply from the Earth’s developed petroleum reservoirs. The demand for petroleum increases through growth of economies and populations. Limits to rates of supply are not set by the amount of remaining reserves. Rather, they are set by the naturally declining rate of production in all oilfields, by faltering rates of new discoveries, and by the geographic and political inequality between oil-producing nations and nations with growth of demand. For each developed field, producers maintain a Maximum Efficiency Rate (MER). MER is a declining and constantly monitored composite of reservoir physics, wellhead crude price, and cost

of investment, operation and product transport. Peak Oil is generally perceived as a state of instability of conventional energy supply.

Impact of future FTB discoveries At a seemingly critical point in time, the question arises: Can we restore the stability of conventional energy supply by developing more and cheaper oil and gas in the planet’s onshore fold–thrust belts? In the present paper, the short answer is ‘No’ for all onshore FTB outside the Middle East. This ‘No’ does not condemn FTB exploration, but it responds to its limited global impact. It is based on published reviews of global petroleum reserves, of typical productive fold–thrust belts and of their undiscovered potential. For the poorly known and understood deepwater belts, this paper’s short answer is ‘Perhaps yes’. Finally, the shale-gas potential of FTB remains as an unknown economic complexity added to the extant FTB complexity. Its part of the short answer would be ‘Not now’. Even if FTB provinces cannot delay the problems of Peak Oil, their continued exploration is vital to humankind. The present contribution of FTB geology is academic, and it will be needed in exploring the deepwater belts. However, reexploring productive FTB preserves and uses sunk costs, live expertise and extant facilities (brains, seismic data and pipelines). Re-exploring FTB makes business sense. Therefore, it is an acceptable international and corporate task.

Reserve estimates Today, the standard reference for global data on petroleum reserves (Klett et al. 1997) must take

From: Goffey, G. P., Craig, J., Needham, T. & Scott, R. (eds) Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 7–31. DOI: 10.1144/SP348.2 0305-8719/10/$15.00 # The Geological Society of London 2010.

8

D. ROEDER

account of 2800 billion barrels oil equivalent (BOE) plus new discoveries and minus about 12 years of production increasing by an annual 4%. This reference has grouped the global oil-prone area into 406 provinces, has ranked them by their known reserves, and displays the sizes of reserves of entire provinces and their log-normal distribution. This reference is ageing but still valid. In Figure 1, the Klett– Ahlbrandt data are displayed in a stack of three Cartesian histograms of globally ranked petroleum provinces. Using these log-normally distributed data and applying subjective estimates, we assume that undiscovered FTB production will comprise about 1% of the global reserves. In the top histogram, the global reserves in all geological settings appear dominated by two super-giant provinces, the Zagros –Mesopotamide foothills (533 billion BOE) and West Siberia (356 billion BOE), closely accompanied by other giant Middle East sites. In the second histogram we exclude these giants, and we count only listed and named basins clearly associated with FTB tectonics. In the third histogram, the Zagros – Mesopotamide FTB system is excluded, and the known FTB reserves still amount to a staggering 177 billion BOE, while at present, the global

consumption is about 12 billion BOE per year and rising (Edwards 2001). The third histogram is repeated at two scales in Figure 2. The top version is detailed and shows that among the non-Middle East FTB provinces, the seven richest provinces contain about 60% of the reserves. Each of the lesser 23 FTB provinces (outlined in white) contains reserves of between 5 and 2 billion BOE. It is within this field of known reserves that most likely any future FTB discoveries will be made. Their field sizes will depend on an undefined proportionality, say perhaps of 1 in 10, between known and undiscovered reserves. The proportionality depends on the present state of exploration. For conducting renewed exploration in any FTB basin ranked in the white-outlined field, industry can hope for field sizes of between 10 and 100 million BOE. However, field sizes of less than 10 million BOE are up to 23 times more likely than sizes of 100 million BOE. This hopeful guess is based on sparse geological evidence, as well as on the assumed but poorly understood proportionality. Exploration failures have not yet been treated probabilistically, but certainly they will improve the odds for success in discovering new small fields.

Fig. 1. Cascade of three log-normal histograms ranking the world’s known petroleum provinces or basins by their quantities of known and producible hydrocarbons, designed after Klett et al. (1997). Vertical scales are linear in billions of barrels oil equivalent. Horizontal scales are ranking numbers of named and measured petroliferous basins. Discussion in text.

FOLD–THRUST BELTS AT PEAK OIL

9

Fig. 2. Cascade of two scaled versions of third (bottom) histogram shown in Figure 1, after data by Klett et al. (1997). Vertical scales are linear in billions of barrels oil equivalent. Horizontal scales are ranking numbers of named and measured petroliferous basins. Discussion in text.

Cost of studying fold – thrust belt geology Beyond surface mapping and applying popular model concepts of geological interpretation, FTB exploration advances only through the use of seismic surveys and deep wells. These and other items must be financed from returns on investment. Also, FTB exploration must compete with easier and less challenging opportunities. Usually, exploration expenditures rise exponentially with increasing detail and better defined objectives. During a South Alpine venture and between 1982 and 1985 (Anschutz and partners, see Roeder & Bello 2003), seismic-controlled surface mapping and a modern thrust interpretation became available for a famously low US$ 0.87 per acre (US$ 2.15 per hectare). This work sufficed to define the potential, to be granted concession areas, and to attract investing partners. However, to define trends, leads and drillable prospects, the consortium needed an improved seismic survey at a cost of US$ 200 per acre (US$ 500 per hectare). Today, this survey is still incomplete. The deep wells needed to explore FTB leads within both Alpine flanks must reach or exceed 6000 m. The recorded expenditures for circumAlpine wells have a log-normal distribution and have cost an average of US$ 2000 per drilled metre between 0 and 5000 m, and of US$ 4000

per metre in wells between 5000 and 8000 m deep (Spoerker in Brix & Schultz 1993). Everywhere, more surface geology, the use of all available seismic data, new seismic acquisition and high wellhead prices are required for resuming failed or simply abandoned FTB plays.

Failed ventures, Cordilleran origin of FTB understanding Well-prepared FTB ventures have failed, and will fail, before their target has been reached, for a variety of common errors, or for technical shortcomings, or for non-technical business decisions. Remembering and re-enacting these ventures during an era of better technology and higher petroleum prices may globally evolve into a trade of secondary or tertiary exploration, here loosely defined as exploration resumed with revised concepts and increased expenditure. Modern understanding of FTB geology evolved in the foothills of the Canadian Rocky Mountains between 1950 and 1970, when the international petroleum industry applied single-fold reflectionseismic profiling and focused on deep drilling for thrust-faulted leading-edge traps. Discoveries and failures occurred simultaneously, and the competing community of explorationists and mappers learned

10

D. ROEDER

jointly (Bally et al. 1966; Roeder 1967; Dahlstrom 1969, 1970 and many others). The Canadian vintage of FTB experience has been exported into every FTB province on Earth and into all modern textbooks of structural geology. Figure 3 combines partial segments of two seriated structure cross sections of the Canadian Rockies in Southern Alberta, showing named and unnamed exploration wells, producing gas fields and undrilled ventures of renewed exploration. A: Imperial Scalp Creek was drilled in 1956. The well encountered the roof thrust of a duplex structure, but the sub-thrust objectives were beyond the capacity of the rig on location. Drilling to sub-thrust objectives would require an improved access road. B: Shell Limestone Mountain did not explore a sub-thrust prospect that had not been visible in the seismic data of 1955 vintage. C: Wells were located on a shallow and un-prospective structure or missed the leading edge of a potential reservoir. D: Chevron Fording Mountain did penetrate the sub-thrust zone of a classical duplex structure, but it encountered unexpected structural complications. Re-exploring this structure for its Mississippian reservoir and early Mesozoic source rock appears promising. E and F: Savanna Creek has been producing wet gas from Mississippian dolomites, and it may have missed a Devonian play nearby. G: the triangle zone of the Southern Canadian FTB may be a not fully explored trend of gas production from Mississippian dolomite.

Deep-Updip Lead (DUD) Figure 4 is a structure cross section explaining conceptually and hypothetically a top-rated FTB trend or lead just updip of the thick-skinned front of any FTB. This lead may be informally named ‘deep-updip’ lead or DUD lead. In many structural configurations of the thick-skinned front, the DUD lead clearly is absent, but it is productive in Shell’s gas fields of southernmost Peru (Roeder & Chamberlain 1995), in the Tesoro block of Southern Bolivia (Dunn et al. 1995), in the Tecate and El Furrial fields of Eastern Venezuela (Hung 2005), and in the Cupiagua and NW trend fields of eastern Colombia (Martinez 2006). The DUD setting has sometimes been missed by the edge of regional seismic coverage, such as on the Alto Beni block of Bolivia (Roeder & Chamberlain 1995) and in the Folded Molasse ventures of Germany and Switzerland, see Figures 5, 6 and 7. The Macal venture of Eastern Venezuela failed because of a misinterpretation of the DUD setting. The Covenant field of Utah is perhaps located in the DUD setting (Sprinkel & Chidsey 2008). In the Canadian and Alaskan parts of the Cordillera,

the DUD setting awaits scrutiny and exploration. The Doonerak window of the Romanzoff Mountains (Mull et al. 1987) condemns the DUD setting for the east part of the Alaskan Brooks Range. Other FTB settings notwithstanding, the search for the DUD lead would be useful in any scrutiny of the remaining FTB potential.

North Alpine ventures (Figs 6 – 9) Since about 1950, oil and gas exploration of the Alpine north front evolved, succeeded, failed and resumed as competitive or national or consortial efforts. In the present paper, this complex history is sampled by the review of four events. An early and largely non-geological effort in Germany failed but was repeated and improved twice. An international consortial effort drilled two rank wildcats of geological significance, but did not achieve commercial production. A special part of a national venture discovered significant FTB-based oil and gas by close cooperation between one drilling engineer, two structural geologists and a funding and encouraging management. Together, these ventures demonstrate the need for reliable and adaptable models, for old and modern models and modern data, for the continued ability to read and use obsolete data, and for enough encouragement and funding. Collectively, the five decades of North Alpine exploration produced much published Earth science, notably Mueller 1978; Bachmann et al. 1982; Lemcke 1988; Wessely 1993; Schwerd et al. 1995 and many others. However, some fundamental aspects of North Alpine petroleum trapping are still conjectural. Preussag’s North Alpine venture on German and Austrian territory (1950–1975) was based on surface geology and on the new electronically processed analogue reflection-seismic data. It also was based on learning from early FTB discoveries in Austria and Southern Alberta (Canada). Preussag’s early explorationists did not clearly define their leads (Figs 6 & 7). The available seismograph technology was marginal at best. There was no tectonic model to put order into the abundant and wellmapped surface data. New fixistic views (Richter & Schoenenberg 1955; Kockel 1956; Jacobshagen 1975) were proposed to replace mobilistic views (Ampferer 1906; Ganss & Schmidt-Thome 1955; Tollmann 1976). Later, the state-supported and consortial Vorderriss well (see Fig. 7; Bachmann & Mueller 1981 and Bachmann et al. 1982) did build a pre-seismic and clearly mobilistic FTB model for Alpine exploration. The well was located on the flank of a tight surface anticline, but its seismic support is plainly

FOLD–THRUST BELTS AT PEAK OIL Fig. 3. Two structure cross sections selected from an inventory of sections from the Southern Canadian Rockies redrawn from D. H. Roeder (2000). The sections display several producing and failed (dry) exploration wells. The letters identify the wells. West is to the left. Bar scales indicate horizontal and vertical scales (VE ¼ 1). Black surfaces: grouped Mississippian and Devonian carbonates. Shaded surfaces: Cambrian sediments. White surfaces: Mesozoic to Palaeogene sediments. Discussion in text.

11

12

D. ROEDER

Fig. 4. Structure cross section of an imaginary FTB illustrating the concept of the DUD trend. Designed after data for an academic test in structural geology at Colorado School of Mines. Black: potential reservoir horizon. Light shading: passive margin sediments. Dark shading: basement. Bar scale indicates horizontal and vertical scale (VE ¼ 1).

illegible (see Dohr 1981). In 1977– 1978, Preussag drilled the well to 6468 m at an unknown cost. The well defined the Austro-Alpine allochthon, its sole, and details of its internal polyphase compressional structure. However, its implied frontal prospect of the DUD type remains unconfirmed and untested, even with far better CDP-type seismic data and a second consortial well (Schwerd et al. 1995). The Austro-Alpine allochthon, a mapped tectonic unit, has High Alpine topography and therefore is an unlikely prospect for sealed hydrocarbon reservoirs. In Austria NE of Vienna and the River Danube, the Austro-Alpine is productive and is covered by the Vienna Basin, a Neogene combined foredeep and extensional successor basin (Fig. 8). Uebertief, a uniquely successful FTB venture by the Austrian state corporation OMV, discovered

commercial hydrocarbons in buried Austro-Alpine rocks. Until 1992, Uebertief had generated a cumulative production of 70 million barrels (MMB) of Alpine light oil and 995 billion cubic feet (BCF) of Alpine methane. Hydrocarbons (Fig. 9) are trapped in tightly folded Triassic carbonates. They are sealed unconformably by an inter-tectonic shale series (Gosau) and by the shaly and extensional Vienna Basin fill. The largely vertical reservoir beds are not visible on conventional seismic data. Exploration focused on drilling, logging, biostratigraphy and structural geology. As part of an overall programme of Austrian national resource development and with persistent and effective funding, the Uebertief venture combined the teamwork of at least three leading people, Hermann Spoerker, drilling engineer, and Godfrid Wessely and Wolfgang Zimmer, geologists. The

Fig. 5. Stratigraphic colour legend and brief survey of the stratigraphies for coloured versions of structure cross sections in the Northern Alps and in Eastern Venezuela, composed for use with Figures 6– 9 and 16.

FOLD–THRUST BELTS AT PEAK OIL Fig. 6. Structure cross section No. MW-2-SB-17 of the Alpine north front in Bavaria (Germany) showing three identified pre-seismic foothills wells that have missed the deep belt of potentially prospective imbricates. There is minor seismic control (not shown). The frontal and triangle part of the foothills or Folded Molasse belt is based on seismic data (Lohr 1969). North is to the left. Scale bar indicates horizontal and vertical scale (VE ¼ 1). More discussion in text.

13

14 D. ROEDER Fig. 7. Montage or composite of two seriated structure cross sections of the Northern Alps in Bavaria, redrawn from D. H. Roeder (2001). Well names are 1 Eberfing, 2, 4, 5, Murnau, 3 Staffelsee and 6 Vorderriss. The more westerly and more northerly segment (drawn after Mueller 1978) shows the structural setting of the old Murnau wells. The montage does not equalize the flexural curvature of the basement top taken from Sommaruga (1997). Scale bar indicates horizontal and vertical scale (VE ¼ 1). More discussion in text.


Fig. 8. Regional structure cross section through the middle part of the Vienna Basin, strongly simplified after Zimmer & Wessely (1996). Of the numerous wildcat wells and producers, only the well symbols are shown. Inset square shows location of more detailed Figure 9. Bar scale indicates horizontal and vertical scale (VE ¼ 1). A: The dotted line is the Sommaruga North Alpine standard flexural basement top located radially 29 km above the seismic-controlled Moho. B: Implied but not documented external massif required to move the North Alpine thrust wedge during the Neogene. C: Rheno-Danubian Flysch and suture-associated late Cretaceous sediments. D and E: Basement parts (‘Altkristallin’) of the allochthonous imbricate complex, outcropping in Leitha hills, redrawn after Zimmer & Wessely (1996).

15

16 D. ROEDER Fig. 9. Structure cross section of the frontal Austro-Alpine allochthon with two stacked and up-righted thrust sheets unconformably covered by inter-tectonic upper Cretaceous clastics and by post-compressional Vienna Basin clastics. Wells of the Uebertief venture encountered and missed commercial oil and gas accumulations. At this detail and limited control, the standard tectonic models lose their applicability, not their validity. Redrawn after Zimmer & Wessely (1996). Red ¼ gaseous petroleum. Green ¼ oil.


venture includes the hitherto deepest European well for hydrocarbon exploration, Zistersdorf UT-2A, with total depth (TD) at 8553 m (1992). The well data developed during the Uebertief venture contain numerous geological details of locally vital importance, but of unlikely regional or textbook significance. Nevertheless, OMV’s Alpine exploration is a global milestone in applying and modifying the FTB style as recognized in the Canadian Rockies. The Gosau seal in the Uebertief area may or may not encourage Austro-Alpine exploration outside the Vienna Basin. Even during the years of academic campaigning for an Alpine fixistic reinterpretation, the OMV team maintained an inventory of seriated structure cross sections, in part with seismic control and built with modern structural geology. The inventory was used to plan and drill a major series of sub-thrust wildcat wells with a mixture of production and dry tests (Zimmer & Wessely 1996). Between 1994 and 2002, two American independents, Anschutz and Forest Oil, resumed exploration in the old Preussag concession with a new inventory of field-checked seriated cross sections. The attraction of risk-sharing partners failed because of German legal restrictions to the re-use of Preussag’s state-supported consortial and early Common-Depth Point (CDP) seismic data. The most recent and unlikely last North Alpine venture, again by the Austrian corporation of OMV and in the German part of the Molasse basin, is

17

continuing the work resumed by Anschutz and Forest. OMV’s exploration is focusing on deep thrust imbrications near the south edge of the Molasse basin, by using 3D seismic surveys and extant surface geology. Results have not yet reached the public domain. However, one older well, Au, near the Austrian– Swiss border, may have reached a shallow analogue of the DUD trend but was dry and abandoned (Wessely 1993).

More Cordilleran ventures An FTB setting is evident in three of the four giant Cordilleran petroleum provinces of South America, and two of them are reviewed here: the Oriente foothills of Colombia and the Serrania foothills of Venezuela (Fig. 10). Both clearly show that it is the abundance and maturity of source rock that makes a petroleum province. They also show that it takes the development drilling after a commercial discovery to reveal and confirm the structural style. At least one of them shows significant limits to the guiding or misguiding concepts of structural styles. The South American Cordillera is an assembly of thick-skinned and locally pluton-invaded cratonic basement blocks. They are separated by crustal strike-slip faults that strain-partition (Fitch 1972) the Peru –Chile and Caribbean subduction (Figs 10 & 14). The north part of the Cordillera affects an Atlantic– Tethyan passive margin series

Fig. 10. Sketch map of northwestern South America after GSA Tectonic Map of South America and Bellizzia & Dengo (1990) showing in three shades of grey from west to east, the Andean tectonic belts, high blocks with outcropping basement, many strike-slip faults, and the Subandean foredeep east of the Cordilleran frontal fold belt. The map also shows two rectangular areas of ventures and structural inventories reviewed in the present paper.

18

D. ROEDER

of post-Jurassic age with unevenly distributed petroleum source rock. A Cordilleran foredeep series of Tertiary to Recent age covers the passive margin series and migrates diachronously northward and eastward. The thick-skinned Cordilleran east front is accompanied by a discontinuous or vaguely coherent FTB. In Colombia, the FTB style is more clearly imbricate. In East Venezuela, the frontal FTB is a belt of detached and internally coherent folds dissected by several thrust faults. Cordilleran petroleum is sourced in Cretaceous to Palaeogene shales. It is trapped in intercalated sandstones within compressional intra-Cordilleran belts, such as the Middle Magdalena valley, in extensional intra-Cordilleran belts (Maracaibo), and in foreland FTB (Oriente, Serrania, Trinidad). The Orinoco tar belt suggests long-range updip migration of petroleum. There are no public-domain data on the geographic distribution of the associated and free gases, of their origin, and of their geochemistry.

Oriente of Colombia Klett et al. (1997) rank this petroleum province as No. 53, with 5.4 billion barrels (BB) of liquid, 10.3 TCF of gas, and 7.4 billion barrels oil equivalent (BOE). Geologically, the Oriente is well described (Colletta et al. 1990; Cooper et al. 1995; Martinez 2006; unpublished talk by R. Graham from Fold-and-Thrust belt conference). It is still under active exploration after world-class discoveries at Cusiana and Cupiagua. The Oriente belt could serve as a global type locality for interfering thrust progradation and triangle back-thrusting, as well as for a successful history of inter-related commercial and public FTB ventures. Figures 11– 13 are cross sections sampling the geology of the Oriente FTB. Its significance started with the Farmout by Triton company (see Triton Colombia, Inc. 1982) showing an abundance of source rock, a somewhat discontinuous layer of fluviatile sands, and classical seismic-documented foothills architecture. Detrimental to industry was its remote location relative to tidewater, and the absence of any local infrastructure. Today as well as initially, security problems and environmental problems are costly and common in the Llanos province. Ecopetrol, the Colombian state petroleum company, persisted in its regional seismic exploration, but had only limited economic success in the Medina segment of the Oriente FTB.

East Venezuela basin (EVB) Three points distinguish this world-class FTB. First, its production rate and its potential are blessed by an abundance of mature source rock. Secondly,

fatal errors in tectonic analysis were caused by unrecognized or acknowledged but ignored limits to the Bally–Dahlstrom –Suppe model of faultbend folding. The errors suggest that this enormously successful concept of structural geology needs restraint and control by data. Thirdly, East Venezuela exploration shows that even with a nearperfect digital system of data retrieval, key data can get lost among hundreds of carefully logged wells, thousands of miles of high quality seismic data, and dozens of well-designed seriated cross sections. Along its cratonic south flank, the thick-skinned and strike-slip-faulted Cordillera of Eastern Venezuela (Fig. 14) is lined by a thin-skinned and petroliferous FTB. Its details are described by varieties of the Bally–Dahlstrom –Suppe structural style (Roure et al. 1994; Passalacqua et al. 1995; Audemard & Serrano 2001; Hung 2005). Its petroleum reserves, as an East Venezuela Province, are ranked (Klett et al. 1997) as No. 13, with 30.2 BB of liquid to solid petroleum, 129.7 TCF of biogenic and thermal gas and 52.6 billion BOE. East Venezuela contains dozens of producing or shut-in anticlinal and thrust-faulted oil and gas fields, hundreds of new field wildcats, and thousands of kilometres of CDP 2D seismic data, all managed and funded by PDVSA, the Venezuelan state petroleum corporation. Geological maps play a key role in defining the northern limits of the productive basin. Also, geological maps are key data for a valid tectonic model. Surface-controlled photogeological maps of the Cordillera and its foothills date back to PDVSA’s precursors such as Esso-Creole Petroleum, renamed Lagoven. Between about 1950 and 1965, and using the same personnel, Geophoto of Denver and Calgary together with Esso– Imperial– Exxon, simultaneously mapped East Venezuela and the productive foothills of Southern Alberta (Canada). In contrast to the setting and style of Alberta, the East Venezuela maps show no thrust faults but internally coherent open and steep-flanked folds. Given the common history of these maps, this contrast of style probably is well observed. There is no record of Exxon –Lagoven’s original structure cross sections. Between 1990 and 2000, exploration by a new generation of geologists and their modern CDP seismic data applied the key elements of the modern Suppean FTB style (Suppe 1983). New structure cross sections (Figs 15–17) show problems with the updated style and their solutions. Two versions of a structure cross section appear in Figure 15. Originally (top image), it proposed a shallow and large main thrust covering a wide field of footwall imbrications (Roeder 2001). A revision to comply with well data soon became needed.

FOLD–THRUST BELTS AT PEAK OIL Fig. 11. Two cross sections of productive foothills structures. West is to the left. Two identical scale bars, both implying equal horizontal and vertical scales (VE ¼ 1). Top: line tracing of one of the pre-discovery 2D seismic lines across the Cusiana field, with shading applied to the Palaeogene stratigraphic interval containing the petroleum system. Groups of letters identify details of the local stratigraphy, after Valderrama (1982), not explained in the present paper. Redrawn after Roeder & Chamberlain 1995. Bottom: Foothills of the Southern Canadian Rocky Mountains in Southern Alberta. Black: Mississippian and Devonian passive margin carbonates. The Mississippian is involved in two productive thrust sheets, Turner Valley and Highwood–Quirk Creek. Redrawn after Bally et al. (1966). 19

20 D. ROEDER Fig. 12. Two parallel and partly seismic-controlled structure cross sections of the Llanos foothills, east front of the Eastern Cordillera, Colombia. West is to the left. Scale bars show equal horizontal and vertical scales (VE ¼ 1). Black: Palaeogene interval of the passive margin or rift sequence, containing the petroleum system. Dark grey: Palaeozoic and older rocks forming the basement of the Eastern Cordillera. Lighter grey, Jurassic and Cretaceous sediments. Prepared during Ecopetrol’s bid round study for JNOC in 1998.

FOLD–THRUST BELTS AT PEAK OIL 21

Fig. 13. Composite structure cross section of the productive Llanos foothills (top) and its partial SNIP restoration (bottom). West is to the left. Bar scale is valid for horizontal and vertical scales (VE ¼ 1). Black: Palaeogene interval of passive margin or rift sequence containing the petroleum system. Grey: unidentified older and younger intervals illustrating the structural setting. This section and its restoration display a gradual or step-wise dissection of a triangle zone by late and out-of-sequence thrusts. Simplified and redrawn from material by Martinez (2006). More discussion in text.

22 D. ROEDER Fig. 14. Crustal structure cross section of Cordillera and Caribe Borderland of Eastern Venezuela, sketched after published data, unpublished geological maps courtesy PDVSA, and the position of the subducting Caribbean slab (Van der Hilst & Mann 1994). North is to the left. Bar scale is valid for horizontal and vertical scales (VE ¼ 1). Black: Cretaceous and Palaeogene sediments. Light grey: continental crust. Dark grey: oceanic crust and mobilized continental crust.


Fig. 15. Structure cross section of the Serrania del Interior in Anzoategui (East Venezuela), two dated versions. Top: A pre-Macal section (Roeder 2001) assuming a style of thin-skinned thrust stacking at and below the major and shallow Urica–Pirital thrust or Cordilleran main thrust. Bottom: A revised version after the Macal evidence of moderate and thick-skinned displacement within a field of internally coherent detached folds. In both versions, the Urica–Pirital thrust, lettered X, cuts up-section and blind-soles beneath the Neogene of the East Venezuela basin. The bar scale is valid for horizontal and vertical scales (VE ¼ 1). The colouring does not conform to the colour legend of Figure 5. Grey shading: Barranquin and older Cretaceous sediments. Thick black line: El Cantil Cretaceous limestone, a seismic marker. Colours are as labelled in Figure 5. Well symbols: Extant well and proposed location of New Field Wildcat Capiricual (2001). The bottom of the section is the gravimetrically assumed and very uncertain basement top.

24 D. ROEDER Fig. 16. Manual line tracing of seismic data with colours showing a stratigraphic interpretation. Seismic data are two combined and overlapping versions of reprocessed seismic line ET88-17A, courtesy PDVSA, near the south front of the Serrania del Interior and near the state line of Anzoategui and Monagas (East Venezuela). The patchwork shows two wells and two geological interpretations. North is to the left. Colours are as labelled in Figure 5. A and D: Lower Cretaceous El Cantil limestone, a seismic marker. B: Shallow high-strain Cordilleran main thrust assumed to define duplex target for MAC-1X. C: interpreted trace of Cordilleran main thrust or Pirital or Prepirital thrust. E: Assumed allochthonous duplex or horse unit disproved by MAC-1X. This combination of an older seismic line with a newer reprocessed patch illustrates a failed attempt to reinterpret a thick-skinned FTB front in terms of a high-strain version of the Bally– Dahlstrom–Suppe style.


Fig. 17. Two versions of a regional structure cross section illustrating the impact of the Macal failure. Top: pre-Macal version composed of the work by Saul Osuna and Leroy Hernandez. This version shows a large and shallow Urica–Pirital thrust. It detaches the Cordilleran field of Biotian folds and covers a field of prospective Jurassic to Neogene imbricates. Bottom: As suggested by two Macal wells and by the seismic line ET88-17A, the Urica– Pirital thrust, moderately dipping, is the sole of the thick-skinned main part of the Serrania del Interior. There is major established production only south of this thrust. Light grey: pre-Cretaceous sediments and basement, undivided. Black: Lower Cretaceous El Cantil limestone, a seismic marker. Dark grey: Palaeogene foredeep fill containing the petroleum system. The bar scales are valid for horizontal and vertical distances (VE ¼ 1). Colours are as labelled in Figure 5.

26 D. ROEDER Fig. 18. Top: Structure section across the Pirital front and the foothills field of El Furrial. Bottom: Line tracing of the seismic support for this cross section. Production of this field comes from several sand levels within the (light-brown) interval of Palaeogene foredeep fill above the (dark brown) Vidono source rock (Hung 2005). The Pirital thrust is strongly discordant to its hanging wall and less discordant relative to its footwall. Its displacement is 25–30 km. The Pirital hanging wall shows strong mountainous topographic relief infilled by the Morichito-Las Piedras successor basin, essentially the Orinoco delta. To the right (south) of El Furrial, foredeep fill may be present below the Morichito–Las Piedras unit. Colours are as labelled in Figure 5.


The large thrust fault had to blind-sole southward, and the poorly controlled north-dipping basement top presented space problems. Surface data do not support a frontal thrust, even though it has been inferred and mapped tentatively as the Urica fault (Ostos 1990), a local branch of the more regional Pirital thrust. South-dipping panels of foredeep fill suggest a triangle zone and the blind south front of the assumed main thrust. The Macal project (2001–2002) involved two deep wells and was targeted for Duplex-type thrust imbrications (Dahlstrom 1970; Suppe 1983) or horses enclosed between the shallower Pirital thrust and the deeper Prepirital thrust. This interpretation assumed a Pirital thrust overlap of 40– 50 km. In a graphically combined form, Figure 16 illustrates the Macal pre-drilling and post-mortem play concepts. The well MAC-1X bottomed in Cretaceous or in older tight and sterile sediments, well below the assumed Pirital or Quiriquire thrust. It never encountered the anticipated Tertiary sourcereservoir system. Later, a paper copy of dip line ET88-17, hand-interpreted by Dr Peter Varrell

27

(about 1992) was rediscovered. This line shows the Pirital fault as a much deeper thick-skinned thrust fault with a setting of constant cut-off angle, with a ramp-on-ramp geometry, and with an estimated throw of 30 km. This setting is drastically different from the anticipated setting of a shallow Pirital –Prepirital system. The Macal failure appears in Figure 17 with two versions of a regional structure cross section. Two wells (only one of them shown) and a forgotten seismic line have strategically limited the Cordilleran mountain front area prone to contain unexplored leads. They also have limited the preseismic assumptions to a moderate-strain version of the Bally–Dahlstrom –Suppe style of FTB tectonics. On the positive side, they have opened the field for exploring detached Biotian folds. A new and critically revised systematic account of East Venezuela leads and prospects is needed to support PDVSA’s petroleum strategy. The El Furrial field (Fig. 18) appears as a duplex in sub-thrust position beneath the Pirital thrust. This thrust fault shows about 25 km of separation, and it discordantly dissects folds of the detached Serrania

Fig. 19. Three stratigraphic well profiles or formation-top logs of two wells in Lombardia of Northern Italy. The actual Cascina Riviero well (formerly Zandobbio) contrasts strongly with its prognosis and with the incompletely drilled Franciacorta well (labelled F-Corta) 5 km to the SE. Vertical scale in metres. Black: Upper Triassic Dolomia Principale, an important reservoir in the Po Valley basin. Dark grey: shaly and in part organic-rich Rhetic or uppermost Triassic. Light grey: Other Mesozoic and Tertiary formations. Black lines in Franciacorta log: High-energy seismic events.

28 D. ROEDER Fig. 20. Two seriated and parallel structure cross sections, Alpine south front in Lombardia, Italy, showing productive or shut-in fields of Malossa and Adda and dry consortial well Cascina Riviero, formerly Zandobbio. Sections are based on surface geology, on partial seismic control and on an assumed Bally– Dahlstrom–Suppe model of thinskinned FTB tectonics just updip of the thick-skinned (Orobic) front. Black: Lower Cretaceous Majolica and upper Triassic Dolomia Principale. Light grey with parallel lines of equal depth below basement top: Pre-Triassic crystalline rocks or basement. Dark grey: Upper Triassic and Jurassic passive margin sediment, also (in foreland part of bottom section) Gonfolite Flysch or Alpine –Apennine foredeep fill of Neogene age, Shallow black unit above Malossa field: Gasiferous Messinian boulder beds. North is to the east, bar scales are valid for horizontal and vertical distances. Both sections display a significant and not entirely controlled difference in depth to foreland basement. The deepest slice in the Alpine frontal stack of thrust sheets is visible on some but not all seismic lines. If present as designed in these sections, the Main Dolomite reservoir of the productive Foothills unit could be reached in wells 7.6 km and 6.8 km deep.


type. A second, more internal footwall imbrication may or may not form a drillable prospect, shown in the present interpretation.

South Alpine venture This failed and unfinished venture could be re-opened given recent attempts to lessen Europe’s dependence on West Siberian gas. In Northern Italy, Anschutz and consorts (1982–1998) tried to discover wet gas on acreage down-dip from producing fields and outside of the exclusive ENI Reserved Area. The global rank of the Po Valley basin (Klett et al. 1997) is 84, with 0.4 BB of liquid, 18.9 TCF of gas, and 3.2 billion BOE. Geologically, the venture aimed at a DUD trend, but it failed with the first well Agip–Chevron Cascina Riviero-1 dry and abandoned. The well neither confirmed nor condemned the objective, that is, a deep imbricate directly down-dip from extant production. A seismic image of the objective appears visible, deep and tectonically covered by units of the thick-skinned Alpine south front. Success in this play would require not only the definition and drilling of the trapping tectonic unit but also firm knowledge of the burial history at the Alpine front, and from the start, a full consortial control over all relevant data both inside and outside the ENI Reserved Area. Feasible locations for the first well were extremely difficult to obtain. The Alpine foothills terrain is composed of densely populated valleys and steeply sloped and forested ridges. At the site finally chosen, the well did reach the tight and dry top of the reservoir in a flank position and not at the crest. The short reservoir interval opened was dry. Finally, the well failed conceptually (Fig. 19) because of a major error in extrapolating the northward thickening of the passive margin wedge containing both the source rock and the reservoir. This error dates back to inaccurate consortial cross sections designed by Dietrich Roeder (see Roeder 1992). Two structure cross sections of the South Alpine front have been in part revised but not re-balanced (2008, Fig. 20) to account for the Cascina Riviero well. In both sections, the potential reservoir horizons, productive in the Po Valley basin, include the (lower Cretaceous) Majolica limestone and the (upper Triassic) Dolomia Principale. Of several known source rocks, the (upper Triassic) Riva di Solto is the most likely local candidate.

Revisiting the South Alpine target Encouraged by the data available from Cascina Riviero, by the political need to replace West

29

Siberian gas, and by the economic need to replace European atomic energy, a second search for a productive DUD trend at the Alpine south front would be based on a restudy of all relevant 2D seismic data. This search would focus on defining a deep slice conterminous with the buried and productive foothills belt. Based on the cross sections shown in Figure 20, the undrilled but seismically visible deep slice perhaps involves the upper Triassic reservoir at an estimated depth of 7.6 km near Cascina, and 6.8 km near the Adda River. The thermal state of a deep foothills slice depends on the kinematic sequence and on the tectonic ages. In the common progressive sequence, the deep slice forms by piggybacking the thick-skinned front, and therefore it is over-mature. In an anomalous setting generated by the Messinian sea-level lowstand (see Roeder 2004), emplacement of the thick-skinned Alpine front may be late and analogous to the Friuli seismicity (Slejko et al. 1987). In this case, the present blind front may have overrun an extant and still cold foothills imbricate.

Conclusions Globally, there are few onshore FTB areas not controlled by state petroleum interest. However, the challenges for states and private industry are the same. Few FTB areas are still in the frontier state, and renewed exploration will be required in most future FTB ventures. Modern exploration is vastly superior to the technical state that was valid during the era of FTB failures. Modern visualization easily replaces the old hand-designed structure cross sections, but it depends on the availability of 3D seismic data. The typical FTB failures occurred in domains of low-fold and obsolete 2D data not feasible for visualization. Investment in modern 3D data must be based on a technical estimate of potential return. Therefore, a typical restudy of FTB failures, preparatory to any decision on investment, requires an intelligent use of obsolete data. Discovery will be possible if the old data are organized and reinterpreted, if high quality seismic data of any vintage can be obtained, if complex new dynamic models are used, and if the new exploration can be integrated with earlier attempts. Local and global shortages of marketable natural gas may or may not help to maintain the high gas prices that are needed to account for the high risk of very deep drilling. Enzo Zappaterra invited the author to attend the Geological Society of London conference on petroleum exploration LGS conference on petroleum exploration of fold –thrust belts and to write the present paper. Enzo

30

D. ROEDER

had been in charge of the Zandobbio-Cascina Riviero well and its geological preparation based on incomplete and uncertain data (1991– 1996). After the LGS conference (2008) he freely shared his revised and better understanding of the drilled prospect. The present paper profited from suggestions by Marc Cooper, Rodney Graham, Rob Butler, Ken McClay and Francois Roure. Finally, reviews by Graham Goffey and an anonymous referee substantially improved this paper.

References ¨ ber das bewegungsbild von FalAmpferer, O. 1906. U tengebirgen: Jb. Geologischen Reichs-Anstalt, 56, 539– 622. Audemard, F. E. & Serrano, I. C. 2001. Future Petroliferous Provinces of Venezuela. In: Downey, M. W., Three, I. C. & Morgan, W. A. (eds) Petroleum Provinces of the Twenty-First Century. AAPG Memoir, 74, 353–372. Bachmann, G. H. & Mu¨ller, M. 1981. Geologie der Tiefbohrung Vorderriss 1. Geologica Bavarica, 81, 17–53, Mu¨nchen. Bachmann, G. H., Dohr, G. & Mu¨ller, M. 1982. Exploration in a classic Thrust Belt and its Foreland. Bavarian Alps, Germany. Bulletin AAPG, 66, 2529–2542. Bally, A. W., Gordy, P. L. & Stewart, G. A. 1966. Structure, seismic data and orogenic evolution of southern Canadian rocky mountains. Bulletin of Canadian Petroleum Geology, 14, 337– 381. Bellizzia, A. & Dengo, G. 1990. Chapter 6, The Caribbean mountain system, northern South America; A summary. In: Dengo, G. & Case, J. E. (eds) The Geology of North America, vol. H, The Caribbean Region. Geological Society of America, Boulder, 167– 175. Colletta, B., Hebrard, F., Letouzey, J., Werner, P. & Rudkiewicsm, J. L. 1990. Tectonic Style and Crustal Structure of the Eastern Cordillera (Colombia) from a balanced cross section. In: Letouzey, J. (ed.) Petroleum and Tectonics in Mobile Belts. Editions Technip, Paris, 81–100. Cooper, M. A. et al. 1995. Basin Development and Tectonic History of the Llanos Basin, Colombia. In: Tankard, A. J., Suarez Soruco, R. & Welsink, H. J. (eds) Petroleum Basins of South America. AAPG Memoir, 62, 659– 665. Dahlstrom, C. D. A. 1969. Balanced cross-sections. Canadian Journal of Earth Science, 6, 743–757. Dahlstrom, C. D. A. 1970. Structural geology in the eastern margin of the Canadian Rocky Mountains. Bulletin of Canadian Petroleum Geology, 18, 332– 402. Dohr, G. 1981. Geophysikalische Untersuchungen im Gebiet der Tiefbohrung Vorderriss 1. Geologica Bavarica, 81, 55– 64. Dunn, J. F., Hartshorn, K. G. & Hartshorn, P. W. 1995. Structural Styles and Hydrocarbon Potential of the Sub-Andean Thrust Belt of Southern Bolivia. In: Tankard, J., Suarez Soruco, R. & Welsink, H. J. (eds) Petroleum Basins of South America. AAPG Memoir, 62, 523–543.

Edwards, J. D. 2001. Twenty-First Century Energy: Decline of fossil fuel, Increase of renewable nonpolluting energy sources. In: Downey, M. W., Three, I. C. & Morgan, W. A. (eds) Petroleum Provinces of the Twenty-First Century. AAPG Memoir, 74, 21–34. Fitch, T. J. 1972. Plate convergence, transcurrent faults, and internal deformation adjacent to SE Asia and the Western Pacific. Journal of Geophysical Research, 77, 4432–4460. Ganss, O. & Schmidt-Thome, P. 1955. Die gefaltete Molasse am Alpenrand zwischen Bodensee und Salzach. Zeitschrift der Deutschen Geologischen Gesellschaft, 105, 402–495, Hannover. Hung, E. J. 2005. Thrust belt interpretation of the Serrania del Interior and Maturin subbasin, eastern Venezuela. In: Venezuela, H. G., AveLallemant, V. B. & Sisson, B. (eds) Caribbean– South American Plate Interactions, Venezuela. Geological Society of America, Special Paper, 394, 1– 346. Jacobshagen, V. 1975. Zur Struktur der su¨dlichen Allgaüer Alpen. Gebundene Tektonik oder Deckenbau? Neues Jahrbuch fur Geologie and Paleontologie Abhandlungen, 148, 185– 214, Stuttgart. Klett, T. R., Ahlbrandt, T. S., Schmoker, J. W. & Dolton, G. L. 1997. Ranking of the World’s Oil and Gas Provinces by Known Petroleum Volumes. US Geological Survey Open File Report, 97–463. Kockel, C. W. 1956. Der Umbau der no¨rdlichen Kalkalpen und seine Schwierigkeiten. Verhandlungen der Geologischen Bundesanstalt, 205–214, Vienna. Lemcke, K. 1988. Geologie von Bayern 1: Das bayerische Alpenvorland vor der Eiszeit. Erdgeschichte, Bau, Bodenscha¨tze, E. Schweizerbart’sche Verlagsbuchhandlung Stuttgart, 175p. Lohr, J. 1969. Die seismischen Geschwindigkeiten der ju¨ngeren Molasse im ostschweizerischen und deutschen Alpenvorland. Geophysical Prospecting, 17, 111– 125, Den Haag. Martinez, J. A. 2006. Structural evolution of the Llanos foothills, Eastern Cordillera, Colombia. Journal of South American Earth Sciences, 21, 510–520. Mueller, M. 1978. Miesbach 1 und Staffelsee-1, two basement tests below the Folded Molasse. In: Closs, H., Roeder, D. & Schmidt, K. (eds) Alps Apennines Hellenides. Inter-Union Commission on Geodynamics, Science. Report 38, 64–68. Mull, C. G., Roeder, D. H., Tailleur, I. L., Pessel, G. H., Grantz, A. & May, S. D. 1987. Geologic sections and maps across the Brooks Range and Arctic Slope to Beaufort Sea, Alaska. Geological Society of America Maps and Charts Series, MC-28S, 1:500 000. Ostos, Marino 1990. Evolucion tectonica del margen Sur-Central del Caribe basado en datos geoquimicos, GEOS No. 30, 1– 294. Passalacqua, H., Fernandez, F., Gou, Y. & Roure, F. 1995. Crustal architecture and strain partitioning in the eastern Venezuelan ranges. In: Tankard, A., Suarez Soruco, R. & Welsink, H. (eds) Petroleum Basins of South America. AAPG Memoir, 62, 667–679. ¨ ber der Bau der Richter, M. & Scho¨nenberg, R. 1955. U Lechtaler Alpen. Zeitschrift der Deutschen Geologischen Gesellschaft, 105, 57–79, Hannover.

FOLD–THRUST BELTS AT PEAK OIL Roeder, D. 1967. Rocky Mountains – Der Geologische Aufbau des Kanadischen Felsengebirges. Beitr. Regionale Geologie der Erde, 5, Borntraeger, 318p. Roeder, D. 1992. Thrusting and Wedge growth, Southern Alps of Lombardia (Italy): final reports. European Geotraverse (EGT). Tectonophysics, 207, 199–243. Roeder, D. 2000. Canadian Rocky Mountains – Inventory of Cross Sections. BERTFEST, Albert W. Bally Spring 2000 Symposium & Fest, Rice University, Houston, TX. Roeder, D. 2001. Eastern Venezuela and Cutofito Fold– Thrust Belts: Constructing a Series of Structure Cross Sections: A report of a regional structural review for 2001: PDVSA internal report. Roeder, D. 2004. Messinian Salinity crisis and SouthAlpine tectonics. Part 1 – thin thrust wedges: – NAFTA, Year 55, No. 3, (March), Zagreb, R. Croatia, 107–116. Roeder, D. & Bello, H. 2003. Global Inventory of Fold– Thrust Belts: Petroleum Geology, Geodynamics (2 volumes). A short course for AAPG International Conference, Barcelona, Espana. Roeder, D. & Chamberlain, R. L. 1995. Eastern Cordillera of Colombia: Jurassic– Neogene Crustal Evolution. In: Tankard, A., Suarez Soruco, R. & Welsink, H. (eds) Petroleum Basins of South America, AAPG Memoir 62, 633 –646. Roure, F., Carnevali, J., Gou, Y. & Subieta, T. 1994. Geometry and kinematics of the North Monagas thrust belt (Venezuela). Marine and Petroleum Geology, 11, 347– 362. Schwerd, K., Huber, K. & Mueller, M. 1995. Tektonik und regionale Geologie der Gesteine der Tiefbohrung Hindelang 1 (Allgaüer Alpen). Geologica Bavarica, 100, 75–115, Mu¨nchen. Slejko, D., Carulli, G.-B. et al. 1987. Modello Sismotettonico dell’Italia Nord-Orientale: Consiglio Nazionale delle Ricerche. Grupo Nazionale per la difesa dai terremoti, Rendiconto No. 1, 1 –82.

31

Sommaruga, A. 1997. Geology of the Central Jura and the Molasse Basin: New Insight into an EvaporiteBased Foreland Fold and Thrust Belt. Memoire de la Societe Neuchateloise des Sciences Naturelles, Tome XII, 176p. Spo¨rker, H. 1993. Technik des Bohrens. In: Brix, F. & ¨ sterreich. Schultz, O. (eds) Erdo¨l und Erdgas in O Verlag Naturhistorisches Museum, Wien, 100– 114. Sprinkel, D. A. & Chidsey, Th. C. 2008. A Review of Petroleum Exploration in the Central Utah Thrust Belt with an Emphasis on the Failures and How they guide Current Research. In: Emme, J. & Carr, M. (eds) Rocky Mountain “Dusters” Lessons Learned and Opportunities Created. RMAG Seminar, Denver September 22, 2008. Suppe, J. 1983. Geometry and kinematics of fault-bend folding. American Journal of Science, 283, 648– 721. Tollmann, A. 1976. Der Bau der No¨rdlichen Kalkalpen. Franz Deuticke, Wien, 449p. TRITON COLOMBIA, INC. 1982. Farming Possibilities in Colombia. Unpublished public-domain brochure. Valderrama, R. 1982. Desarollo de facies en la cuenca de los Llanos Orientales Colombianos. In: Roberto, L. (ed.) Exploracion Petrolera de las Cuencas Subandinas. Volume 3, Bogota´ 1982. Van Der Hilst, R. D. & Mann, P. 1994. Tectonic implication of tomographic images of subducted lithosphere beneath northwestern South America. Geology, 22, 451– 454. Wessely, G. 1993. Beilage 9, Geologischer Tiefbau Flysch-Kalkalpenzone. In: Brix, F. & Schultz, O. ¨ sterreich. Verlag Naturhis(eds) Erdo¨l und Erdgas in O torisches Museum, Wien. Zimmer, W. & Wessely, G. 1996. Exploration results in thrust and subthrust complexes in the Alps and below the Vienna Basin in Austria. In: Wessely, G. & Liebl, W. (eds) Oil and Gas in Alpidic Thrustbelts and Basins of Central and Eastern Europe. EAGE Special Publication No. 5, The Geological Society, London, 81–108.

Structural styles in the Papuan Fold Belt, Papua New Guinea: constraints from analogue modelling KEVIN C. HILL1*, KATIE LUCAS2,3 & KEITH BRADEY1 1

Oil Search Limited, Angel Place, 123 Pitt Street, Sydney, NSW Australia 2000

2

Geological Sciences and Geological Engineering, Miller Hall, Queen’s University, Kingston, Ontario, K7L 3N6, Canada 3

Present address: Premier Gold Mines, 401-1113 Jade Court, Thunder Bay, Ontario, P7B 6M7, Canada *Corresponding author (e-mail: [email protected])

Abstract: Cross sections, seismic data and centrifuge analogue modelling reveal the structural styles in the oil-producing areas of the Papuan Fold Belt. They include inverted basement faults, detachment faults in the Jurassic section 1 –2 km beneath the Neocomian Toro Sandstone reservoir, and tight, overturned folds in the reservoir sequence with stretched and boudinaged forelimbs, cut by break-thrusts. Additional features include highly variable thicknesses in the Cretaceous Ieru Formation, the regional seal sequence, including through-going detachments that isolate the overlying thick Miocene Darai Limestone. Centrifuge analogue modelling of intact, plane-layered strata determined that the mechanical stratigraphy and the thickness of weak beds above the lower dećollement horizon exert the greatest control on the structural style. Large-offset thrust faults were only produced in models with pre-cut faults, generating early inversion and then large ramp anticlines, similar to those in the Kutubu Oilfield, which has reserves of .350 million barrels. It is suggested that the Kutubu Oilfield trend was underlain by a large normal fault and that, by analogy with the Vulcan Sub-basin, oil-rich source rocks may be confined to the hanging wall or north side of this fault. Oil would have been generated and expelled during thin-skinned deformation.

The aim of this paper is to describe the structural style of hydrocarbon traps in the Papuan Fold Belt (Fig. 1), particularly the thin-skinned structures, and to relate the structures to mechanical stratigraphy through physical, centrifuge analogue modelling. Structural interpretation has incorporated data from the many wells that have been drilled, from surface mapping and from 2D seismic and other geophysical surveys. However, the jungle-covered mountains in Papua New Guinea (PNG) severely hamper geological mapping and result in poor quality seismic data, so there remains much ambiguity in the structural interpretations. The ambiguity is greatest in the deeper, undrilled, parts of the structure and in the sheared forelimbs, both areas that are being investigated for additional hydrocarbon potential. In order to constrain the structural interpretation and hence assess this potential, physical scaled-modelling was carried out in a centrifuge using horizontal layers of plasticine and silicone putty to reflect the mechanical stratigraphy of the main beds. The models are generic rather than designed to replicate specific structures, but by varying the thickness and competence of the individual beds it was possible to obtain a very good fit to the known structures.

Oil and gas exploration commenced in PNG in the late 1920s, drilling shallow wells on seeps near major rivers on the Fly Platform. Exploration in the fold belt commenced in the 1950s on the accessible mountains, resulting in significant gas discoveries such as Barikewa in 1958 and Juha in 1983. Commercial oil was discovered in the Iagifu – Hedinia (Kutubu) anticlines in 1985 (Bradey et al. 2008) followed by nearby discoveries at Agogo, SE Mananda, Moran and Gobe, which have collectively been on production since 1992. Initial recoverable reserves in the known fields were well over 500 million barrels of low viscosity oil. The giant Hides gas field (Johnstone & Emmett 2000), with over 5 trillion cubic feet (TCF) reserves, was discovered in 1986 and will be the core of the gas development project planned in the near future. This project, combined with new technologies, has opened the door to a renewed phase of exploration for deeper, more cryptic oil and gas plays (Hill et al. 2008). Currently, over 250 wells and sidetracks have been drilled in the Papuan Fold Belt, and more than 3000 km of 2D seismic data have been acquired, generally of poor to moderate quality. There has been considerable surface geological

From: Goffey, G. P., Craig, J., Needham, T. & Scott, R. (eds) Hydrocarbons in Contractional Belts. Geological Society, London, Special Publications, 348, 33– 56. DOI: 10.1144/SP348.3 0305-8719/10/$15.00 # The Geological Society of London 2010.

34

K. C. HILL ET AL.

Fig. 1. Sun-shaded, digital elevation model showing the main features of the Papuan Fold Belt. The main structural belts are labelled, after APC (1961). The Darai Plateau is a very large asymmetric anticline overlying an inverted extensional fault that was active from Triassic to Miocene times. The structure is offset by a tear fault, the Bosavi Lineament, to underlie the mountain front in the NW part of the fold belt. All the producing oil and gas fields lie within the Strongly Folded Belt. Key wells are labelled and oil field and gas field outlines are shown in green and red, respectively. Wells within the fields are shown by white dots but, for clarity, are not labelled.

mapping, aided by 87Sr/86Sr isotope dating of the Miocene surface limestones (Hornafius & Denison 1993) and by analyses of synthetic aperture radar images. Structural interpretation has also been improved by the acquisition of regional and high-resolution aeromagnetic surveys, gravity surveys, and of earthquake seismic data. In this paper, a regional cross section is presented and two oil-producing structures and one breached structure are discussed, each of which has been drilled by numerous wells with dipmeter data, has been covered by widely spaced 2D seismic data and has good surface outcrop data. The structures are the Moran, Agogo and Paua anticlines (Fig. 1).

Tectonics, stratigraphy and structure of the Papuan Fold Belt Tectonically, the island of New Guinea comprises the northern margin of the Australian continent

that has undergone Miocene to Pliocene oblique convergence with the Pacific Plate resulting in collision with intervening microplates. The reader is referred to Hill & Hall (2003) for a recent discussion of PNG tectonics. The Papuan Fold Belt straddles the middle of the island and comprises precipitous mountains of heavily karstified Miocene limestone covered with dense equatorial jungle. The Fold Belt is made up of folded and thrust Mesozoic clastic rocks and Tertiary limestones and is bound to the south by the Fly Platform containing similar rocks, but undeformed. Compression in the fold belt occurred mainly in the Late Miocene and Pliocene (Hill & Raza 1999) directed roughly from NE to SW until the Middle Pliocene. Crowhurst et al. (1997) proposed that there was then a change to east –west compression in PNG, continuing to the present and resulting in increased strike-slip deformation. The stratigraphy of the Papuan Fold Belt is summarized in Figure 2a and the simplified stratigraphy

STRUCTURAL STYLES IN THE PAPUAN FOLD BELT

35

Fig. 2. Stratigraphy across the Papuan Fold Belt. (a) Simplified lithostratigraphic section (after Hill et al. 2000) flattened on the top Miocene. The Mesozoic section is dominantly mudstone, but contains the Upper Jurassic to Neocomian Iagifu, Hedinia, Digimu and Toro sandstone reservoirs. These are collectively modelled as Toro Sandstone. The Cretaceous Ieru Mudstone is the regional seal and is unconformably overlain by the thick Miocene Darai Limestone and Orubadi Marls. The Upper Triassic and Lower Jurassic syn-rift sequence is schematic on this section. (b) Stratigraphic column used for mechanical modelling in a centrifuge, showing the mechanical stratigraphy and real versus model thicknesses. P ¼ plasticine, SP ¼ silicone putty. 1 mm in the model is equivalent to 1 km in the prototype (see Table 1). In the later models, a thicker analogue Imburu sequence and pre-cut fault were used as shown.

36

K. C. HILL ET AL.

used in centrifuge analogue modelling (discussed later) is shown in Figure 2b. Beneath the Fly Platform and the fold belt, ‘basement’ comprises Upper Palaeozoic rocks, mainly Permian, that were deformed in the Early Triassic New England Orogeny and intruded by Middle Triassic granites (Van Wyck & Williams 2002; Crowhurst et al. 2004). The sequence was deeply eroded and the granites were exposed at that time. In the Late Triassic and Early to Middle Jurassic, the area was subject to extension and rifting (Home et al. 1990), depositing the syn-rift Kana Volcanics, Magobu Coal Measures and Barikewa Mudstone, the latter two being probable source rocks (Fig. 2a). Regional Late Jurassic subsidence flooded the margin allowing deposition of the Imburu Formation, Toro, Digimu, Iagifu and Hedinia sandstone reservoirs and the Cretaceous Ieru Formation seal. In distal facies of the northeastern Fold Belt, both the Imburu and Ieru mudstones are hydrocarbon source rocks. During the latest Cretaceous to Paleocene, southern PNG was uplifted, probably associated with northern Tasman and Coral Sea rifting. Subsequent erosion stripped some Upper Cretaceous sediments in the fold belt and Fly Platform area and deposition did not resume until Late Oligocene flooding allowed widespread deposition of Miocene shallow marine carbonates, the Darai Limestone (Fig. 2a). Carbonate deposition was halted by the Late Miocene onset of compressional deformation, which was also responsible for generation and migration of most hydrocarbons.

Structural models of the Fold Belt APC (1961) divided the Papuan Fold Belt into three NW–SE-trending belts, illustrated in the regional section shown in Figure 3 (discussed below). In the SW was the ‘Gently Folded Belt’, including the giant but low relief Darai anticline, 40 km wide and 100 km long. Structures in this belt are generally considered to be inverted basement structures (e.g. Hobson 1986; Hill 1991). The second belt, c. 30 km wide, commenced NE of the Darai anticline and comprised the ‘Strongly Folded Belt’. All the hydrocarbons to date have been found in this belt and its structural style is addressed in this paper. The area further to the NE was classified as the ‘Imbricate Zone’ and consists of thrust repeats

of Miocene limestone and uppermost Cretaceous shales, which are recorded in outcrop. This area is not considered here, but was analysed by Hill et al. (2000) and is currently being explored by oil companies. As little was known about the Papuan Fold Belt, early structural models followed those from betterstudied fold belts in North America and Europe, for instance fault-propagation folds (Smith 1965), imbricate thrusts (Findlay 1974; Jenkins 1974), duplexes (Hobson 1986) and fault-bend folds (Hill 1991). It was only with the detailed analysis of individual anticlines following the drilling of several wells that it was found that the structures comprised detached folds with overturned and/or thrusted forelimbs (Lamerson 1990; Eisenberg 1993; Franklin & Livingston 1996). Regional sections continued to show underlying basement thrusts or basement inversion structures (Buchanan & Warburton 1996; Thornton et al. 1996; Cole et al. 2000). Drilling and seismic acquisition over the Moran and Paua anticlines indicated breakthrust structures with strongly sheared forelimbs (Davis et al. 2000; Lingrey 2000). Recently, Hill et al. (2008) and Bradey et al. (2008) used seismic and potential field data to confirm along-strike partition of the fold belt into zones with differing basement involvement. To interpret and model the structure of the Papuan Fold Belt, it is important to know or infer the pre-compression configuration of the margin. For instance, was it a relatively undeformed ramp as beneath parts of the Canadian Rocky Mountains (e.g. Bally et al. 1966) or a highly faulted margin with abrupt thickness changes as recorded in parts of nearby Indonesia (e.g. Chambers et al. 2004)? Cooper et al. (1996) interpreted two regional seismic sections across the Timor Sea, a margin along strike to PNG with a similar Mesozoic history. There they found that the major basinbounding fault abuts the stable Londonderry High, and that two large faults bound the Jurassic Swan Graben, but otherwise the Mesozoic section gradually thickens seaward over a distance of 180 km (Fig. 4). They also noted a significant offset of structures across the Paqualin transfer zone (e.g. Woods 1992). Cooper et al. (1996) proposed that this area is a good structural and stratigraphic analogue for the Papuan Fold Belt prior to thrusting.

Fig. 3. (Continued) Regional cross section over the Papuan Fold Belt based on projected well data, surface geology and, in part, on poor to fair quality seismic data. The seismic data were most useful in defining regional dip and elevation of basement and occasionally the dip and shape of the base of the Darai Limestone. The bulk of the section results from structural interpretation of well and surface data. The lowermost section, at half-scale, shows a restoration of the Gently Folded and Strongly Folded Belts such that the shortening has not yet propagated over the hypothetical graben containing source rocks. See Figure 1 for location and text for detailed discussion.


37

Fig. 3.

38

K. C. HILL ET AL.

Fig. 4. (a) Location map of regional seismic lines across the Vulcan Sub-Basin in the Timor Sea, along strike to PNG. Cooper et al. (1996) inferred that this area was a good analogue to the pre-deformation stratigraphy and structure of the Papuan Fold Belt. Jabiru, Challis and Skua are oilfields, each with recoverable reserves of 50– 150 million barrels. (b) Block diagram of the top Callovian in the Vulcan Sub-Basin showing the Upper Jurassic rift geometry, after Cooper et al. (1996). Note the large growth fault adjacent to the Londonderry High similar to the Darai Fault in PNG (Fig. 3). Oil source rock is largely restricted to the syn-rift fill in the Swan and Paqualin grabens (Kennard et al. 1999) and may be similarly restricted in PNG.

Regional structure of the Papuan Fold Belt A regional cross section was constructed in 2DMove from the foreland across the Darai Plateau, the Strongly Folded Belt and the Imbricate Belt

(Fig. 3). The section was drawn to honour all stratigraphy and dips from outcrop and ten boreholes in addition to synthetic aperture radar images and limited potential field data. It was constructed along or close to seismic lines, particularly the semiregional line PN05-404 (shown in Hill et al. 2008)


across the Fold Belt. In general seismic data quality was moderate, ranging from poor to occasionally good. Migrated seismic lines were used for all interpretations, although regularly checked against those with stack and wave-equation processing. The interpreted horizons were imported into GXII and vertically depth stretched using laterally varying interval velocities for each of the main stratigraphic units. The resulting depth horizons were then imported into 2DMove and were used to guide the form of structures. Due to the moderate quality of the seismic data, they were not relied upon in detail to construct the cross sections. The top and base of the Darai Limestone were usually reasonably imaged so the underlying structure was interpreted by projecting down using known stratigraphic thicknesses. The section was incrementally restored using 2DMove to help validate the interpretation and show the likely structural evolution. Due to the complexity of the structures, several methods were used in restoration. Late-stage basement inversion structures were restored using a tri-shear algorithm that accommodated local changes in thickness of the overlying sediments. The tri-shear algorithm was a good approximation to the deformation for the Mesozoic section, but not for the competent Darai Limestone, that required separate, fault-parallel flow restoration. Relatively simple fault-bend fold structures were restored using fault-parallel flow algorithms. More complex structures, such as the overturned SE Hedinia and Moran anticlines, had late break-thrusts restored by fault-parallel flow, were partially unfolded using a flexural slip unfolding algorithm, and then again restored using fault-parallel flow. Usually a small degree of area balancing was required for the core of the structures. The regional section was restored incrementally in six stages, but only one is shown in Figure 3, illustrating restoration of the Gently Folded and Strongly Folded Belts.

Fly Platform and Gently Folded Belt The thickness of sediments above basement within the Darai Plateau is more than double that of the adjacent Fly Platform (4800 m v. 2200 m) indicating inversion of a previously extensional growth fault (Fig. 3). The extensional fault was probably a major basin-bounding fault, as the sedimentary thickness to the north remains at 4–6 km. The growth appears to be continuous through geological time, suggesting a relatively stable platform to the south, similar to the Londonderry High of Cooper et al. (1996; Fig. 4). Offset of the Darai Limestone and continuing earthquake activity suggest that Darai Plateau inversion was Pleistocene to Recent. However, a component of Late Miocene inversion cannot be ruled out.

39

There is clearly a significant unconformity between the Cenomanian upper Ieru Formation and the Late Oligocene basal Darai Limestone. Apatite fission track analyses (e.g. Hill & Gleadow 1989, 1990) combined with vitrinite reflectance profiles in the Kanau-1 well indicate .2 km erosion of the uppermost Ieru Formation beneath the Darai Plateau, but ,1 km erosion beneath the foreland, prior to Darai Limestone deposition. This suggests that in the Early Tertiary the old normal fault was inverted and that the hanging wall was eroded prior to regional Oligo-Miocene subsidence. Although they appear to have structural closure, neither the Kanau-1 nor the Bosavi-1 wells drilled on the Darai Plateau recovered hydrocarbons. This is thought to be due to lack of charge. It is notable on the cross section that the Toro Sandstone reservoir in both wells is near sea-level. To the SW, the Toro Sandstone almost abuts the basal Darai Limestone across the Darai Fault. It is considered likely that along strike the Toro connects to the basal Darai Limestone and hence is in pressure communication with the foreland, consistent with the low pressures recorded in the Kanau-1 well.

Strongly Folded Belt Where the gently NE-dipping limb of the giant Darai Plateau meets the frontal fold belt structures, such as Zongwe (Figs 1 & 3), strong linear reflectors were observed on seismic at c. 3 s (see Fig. 9 in Hill et al. 2008). These are interpreted to be Magobu Coal Measures overlying basement. Using the depth conversion methods outlined above, these reflectors record a consistent dip of c. 68 for over 12 km to the NE beneath the Strongly Folded Belt suggesting a planar, relatively undeformed Jurassic sequence above basement such that the overlying Zongwe, Ai-io and SE Hedinia structures are detached within the sedimentary section (Fig. 3). The Zongwe anticline is interpreted to be the leading edge of the thin-skinned thrusting, in that there is a Darai Limestone repeat at surface along a fault detached within the Ieru Formation. This is thought to be underlain by a reactivated basement thrust creating a large gentle fold in the Toro reservoir. The evidence for the basement thrust from seismic data is equivocal as the data quality in that area is poor. However, on the synthetic aperture radar image (Fig. 1), the Zongwe and Ai-io anticlines together can be interpreted as a 3–5 km wide asymmetric structure that resembles a mini Darai Plateau. The SE Hedinia anticline has been drilled by three wells and the detachment is inferred to be near the top of the Koi-Iange section. The SE Hedinia wells show that the anticline is tight, probably with an overturned forelimb, or with a forelimb sheared out by thrust faulting. Both are typical

40

K. C. HILL ET AL.

structural styles along strike. Due to the steep dips, the core of the structure is not effectively imaged on seismic data. Between SE Hedinia and Kutubu East (Figs 1 & 3), there is a consistent SWdipping panel from basement through to the basal Darai Limestone indicating uplift of basement to the NE on a significant basement fault. Based on seismic interpretation and section balancing, the basement faulting is interpreted to deform the top Koi-Iange Formation detachment so that at least some of the basement faulting occurred after the overlying thin-skinned deformation. The Kutubu East anticline was drilled by two wells, which encountered minor gas and very high pressures, as opposed to normal to low pressures in the Strongly Folded Belt to the SW. The conundrum of having a breached structure that preserves very high pressures is currently being investigated, but it clearly shows a sealing fault underlying the Kutubu East anticline. Regionally, this fault separates the high pressure belt to the NE from the adjacent SE Hedinia Gasfield and Kutubu Oilfield to the SW.

Imbricate Belt Over a distance of 10 km to the NE of the Kutubu East anticline, interpretation of the seismic data (Hill et al. 2008) indicates a planar, gently NE-dipping panel of strata from basement to top Darai Limestone. Geological maps show that this panel is overlain by thrust repeats of Darai Limestone and thin upper Ieru Formation, the start of a major Darai duplex that crops out over a band that is 24 km wide from Lake Kutubu to the Wage anticline. Beneath the mapped Mubi anticline the step-up in basement, inferred from seismic data, appears to fold the overlying Darai Limestone duplex, so the basement thrusting occurred after the thin-skinned deformation. The Darai Limestone duplex exposed at surface represents considerable shortening in the Darai Limestone and upper Ieru Formation. Below, or to the NE, this must be balanced by equivalent shortening in the lower Ieru Formation to Koi-Iange Formation. On the cross section, this has been represented as a simple duplex forming the Mount Castle and Wage anticlines. However, numerous other interpretations are possible, including basement involvement. A recently acquired regional seismic line across these structures may resolve the subsurface geometry.

Centrifuge analogue modelling Background In order to create a more accurate representation of the subsurface structure of the Papuan Fold

Belt, and to improve exploration success, scaled physical analogue models were employed. These centrifuge models simulate mechanical stratigraphy and can help determine which factors control the deformation and thus predict the structural style, the deformation sequence and the geometry of hydrocarbon traps. Dixon (1996) showed that when a pre-existing dip-slip fault was inserted into the model to simulate an old basement or extensional fault, it was reactivated early in the deformational history, prior to thin-skinned deformation. Recent centrifuge modelling of facies changes and reefs within fold and thrust belts, such as the Canadian Rocky Mountains, clearly demonstrated that mechanical stratigraphy not only affected the structural style, but also the sequence of structural deformation (Dixon 2004). This modelling also demonstrated how the strength of the basal dećollement surface influences the style of deformation. A dećollement of moderate strength produces forelandverging folds in weak, basinal facies and thrusts in competent platform facies. A weaker basal detachment promotes upright folding in the basin facies and forethrusts and backthrusts associated with upright buckling on the platform (Dixon 2004). Dixon (1996) applied generic analogue models to the Papuan Fold Belt, and showed detached or loosely linked structures in competent beds, depending upon the relative strength of the intervening weak layer. He illustrated the potential for complete detachment between the reservoir and near-surface structures. He also showed that pre-existing faults in a ‘basement’ analogue are reactivated early in the deformation and remain as important features. However, the amount of reactivation varied inversely with the dip of the pre-existing fault such that reactivation was barely noticeable if the initial preexisting fault was steeper than about 458. Dixon’s (1996) pre-existing faults extended through the analogue Mesozoic section, with growth in the competent reservoir analogue, but not in the underlying weak Jurassic syn-rift section. Dixon (1996) also discussed the limitations of the modelling. He pointed out that it cannot simulate variables such as geothermal gradient, pore-fluid pressure and syntectonic erosion. Furthermore, the modelled stratigraphic sequence rests on a rigid baseplate that is not involved in the deformation.

PNG model parameters The centrifuge experiments were performed in the Experimental Tectonics Laboratory at Queen’s University, Canada. The centrifuge modelling technique used is discussed in detail by Dixon & Summers (1985), Dixon & Tirrul (1991), Liu & Dixon (1991) and Dixon & Liu (1992). The centrifuge employed in these experiments (Fig. 5) can


Fig. 5. Centrifuge modelling. (a) An oblique view of the centrifuge used for analogue modelling, showing the location of the chamber within the centrifuge rotor. (b) Close-up of the centrifuge rotor, indicating how the model fits into the chamber. (c) Schematic diagram of the initial model configuration. The rigid base plate represents the basement of the model, and consists of aluminium plates. Shortening of the foreland stratigraphic succession is caused by the gravitational collapse of the hinterland wedge.

subject the model to a centripetal acceleration up to 20 000 g, such that it simulates the Earth’s gravity. All experiments were subjected to an acceleration of 4000 g (where 1 g ¼ 9.8 m/s2, normal Earth

41

gravity). Each model underwent two to four deformation stages, each stage lasting for five minutes at the maximum acceleration of 4000 g, with an additional seven minutes for the acceleration and deceleration of the centrifuge. The models were photographed in plan view and cross section after each stage. Table 1 outlines the model scaling ratios used, after Liu & Dixon (1991). The models were constructed of plasticine modelling clay and silicone putty, with internally layered units of differing mechanical strengths. These materials exhibit a contrast in their competencies and are suitable analogue materials for the different rock types in the Papuan Fold Belt (Dixon & Summers 1985). In the models, the plasticine represents competent units such as limestone and sandstone, and a combination of silicone putty and plasticine represents incompetent units such as shale. The strength of a mechanical unit can be altered by changing the ratio of plasticine to silicone putty, with an increase in plasticine corresponding to an increase in strength. Building on Dixon’s (1996) work, the models presented here were designed to specifically simulate the known stratigraphy (Fig. 2). From a mechanical point of view, this comprises the strong Miocene Darai Limestone; the weak Cretaceous Ieru Formation, intermediate strength Upper Jurassic Iagifu, Hedinia, Digimu and lowermost Cretaceous Toro sands (here collectively termed Toro) and the weak Jurassic clastic sequence, mainly Imburu Formation (Fig. 2b). The thicknesses and competencies of the layers were varied in each model to demonstrate how different mechanical stratigraphies, and different combinations of preexisting faults, control the overall structural style of the fold belt (Table 2). Importantly, the thickness and strength of the Jurassic syn-rift section was varied to see if it resulted in different structural styles. If a characteristic style could be attributed to a specific syn-rift thickness, it may be possible to predict the location of old normal faults in the Papuan Fold Belt by analysing structural style. The initial Papuan Fold Belt modelling focused on changes within the overall mechanical stratigraphy, the strength and thickness of the lower detachment horizon and the competency of the Ieru Formation. Further centrifuge experiments incorporated previous modelling work by Dixon (1995) and Dixon et al. (1996), that show how primary extensional faults can disrupt the typical deformation sequence of fold and thrust belts. The models created for these experiments contained a pre-cut fault in the lower unit. The fault was cut at an angle of 188 dipping to the hinterland. This very low angle was used as steeper dipping faults were found to exhibit less reactivation and then lock up (Dixon 1995, 1996; Dixon et al. 1996).

42

K. C. HILL ET AL.

Table 1. Model scaling ratios used for this study, after Liu & Dixon (1991) Ratio (model: prototype)

Equivalence (model ¼ prototype)

lr ¼ 1.0 1026 rr ¼ 0.6 tr ¼ 1.0 10210 ar ¼ 4.0 103 sr ¼ rr lr ar ¼ 2.4 1023

1 mm ¼ 1 km 1.60 ¼ 2.67 (bulk value of stratigraphic column) 1023 s21 ¼ 10213 s21 (for example) 4000 g ¼ 1 g Calculated from other ratios

Quantity Length Specific Gravity (mass) Time (strain rate) Acceleration Stress

Table 2. Construction details for plane-layered models KL10, 12, 17 and 19 and for pre-cut fault models 20 and 22 Model

Prototype unit

Construction materials

Number of internal laminae

Total unit thickness (mm)

P:SP thickness ratio

Total model thickness

KL10

Darai Ieru Toro Imburu Darai Ieru Toro Imburu Darai Ieru Toro Imburu Darai Ieru Toro Imburu Darai Ieru Toro Imburu

P P/SP P P/SP P P/SP P P/SP P P/SP P P/SP P P/SP P P/SP P P/SP P P/SP

8 16 4 16 8 16 4 32 8 16 4 32 8 16 4 32 8 16 4 32

1.5 1.0 0.5 1.0 1.5 1.0 0.5 2.0 1.5 1.0 0.5 2.0 1.5 1.0 0.5 2.0 1.5 1.0 0.5 2.0

1:0 1:1 1:0 1:0 1:0 1:1 1:0 1:0 1:0 2:1 1:0 1:0 1:0 2:1 1:0 2:1 1:0 2:1 1:0 1:0

4 mm

KL12

KL17

KL19

KL20 KL22

5 mm

5 mm

5 mm

5 mm

Abbreviations: P, plasticine, SP, silicone putty.

Vaseline petroleum jelly was applied to the fault surface to prevent the hanging wall and footwall from re-adhering after being cut. The initial model configurations for these experiments are shown in Figure 2 and Table 2. Regrettably, the seismic data in PNG were not of adequate quality in the basement, typically at 3 –6 s two-way time (twt), to determine the dip of faults to compare with the model (see discussion).

Plane layer, mechanical stratigraphy modelling The initial model had a relatively thin basal unit, simulating c. 1 km of Jurassic section as observed beneath parts of the Fly Platform (Fig. 3). Forelandverging imbricate duplex structures developed as thrusting initiated early during deformation and

only low-amplitude folds were able to form prior to the development of break-thrusts (Fig. 6a, KL10). Two dećollement surfaces exist within the Imburu and Ieru analogues, creating differential shortening and thrust spacing between the competent units. The second model shows an increase in the thickness of the basal unit from 1.0 mm to 2.0 mm, simulating c. 2 km of Jurassic section above rigid basement, as observed beneath the Darai Plateau (Fig. 3). This modification decreased the amount of foreland vergence and created more upright structures (Fig. 6b, KL12). There was an increase in folding versus thrusting, with increased fold amplitude in both the Toro and the Darai analogue units. The dećollement horizon within the Ieru analogue was not active consistently throughout deformation, allowing regions of both harmonic and disharmonic deformation to develop.


43

Fig. 6. Centrifuge analogue modelling of intact, plane-layered strata with initial conditions as shown in Figures 2 and 5 and Table 2. Note the variation in structural style as the thickness and relative mechanical stratigraphy are varied. Observed faults have been highlighted with thin black lines. Model (a) has a 1-mm thick Imburu analogue as opposed to 2 mm for models (b) to (e). In model (c) the mechanical strength of the Ieru is increased and in model (d) the mechanical strength of the Imburu is increased. Model (e) has a pre-cut fault dipping at 188 from the base Imburu to the base Darai analogue. See text for discussion.

The third model incorporates an increase in the mechanical strength of the Ieru analogue, creating a mechanical linkage between the overlying Darai and underlying Toro competent units (Fig. 6c, KL17). By forcing the linkage of the competent

units, the upper three units act as a single beam deforming above a weak, relatively thick, dećollement horizon. The upright, fold-dominated structures are harmonic between the competent units, with surface structures situated directly above the

44

K. C. HILL ET AL.

structures at depth. However, in detail it can be seen that the Toro analogue is dominated by tight folding, the Ieru analogue manifests substantial changes in thickness and the Darai analogue records both open folding and thrust faulting. The Ieru analogue acts as a local or diffuse detachment zone connecting the folds and minor faults in the Toro analogue to the faults in the Darai analogue. The structural style is very similar to that recorded in the Moran and Usano structures (Franklin & Livingston 1996; Davis et al. 2000; Lingrey 2000) and is discussed further below. The fourth model includes an increase in the mechanical strength of the Jurassic Imburu analogue (Fig. 6d, KL19). This strong lower dećollement horizon produced a thrust-dominated structural style, containing duplexes with foreland vergence. Upright structures did not develop due to the enhanced foreland vergence. Early developed folds had very low amplitude, and thrusts developed early in the deformation. A significant second dećollement surface developed within the Ieru analogue, allowing the structures between the Darai and Toro analogues to become laterally displaced from each other. Varying the mechanical stratigraphy used in the fold and thrust modelling of plane-layered strata resulted in four different structural styles

(Fig. 6a–d). The structural style that most resembles that recorded in the Papuan Fold Belt is shown in Figure 6c, perhaps with elements from Figure 6d. Thus the modelling suggests the presence of a relatively thick and weak Jurassic section and a slightly more competent Ieru section. The models are discussed further below when compared to cross sections of the Papuan Fold Belt.

Modelling of pre-existing faults Recently acquired and/or reprocessed seismic data across the Kutubu Oilfield suggest a low-angle thrust fault with 4–8 km of displacement in that area of the Papuan Fold Belt (Bradey et al. 2008), as illustrated in Figure 7. The Kutubu seismic line (Fig. 7) also shows an unusual ‘double-hump’ structure in the hanging wall with the Toro Sandstone and Darai Limestone both gently folded, except in the forelimb where overturned Darai was encountered at the base of one well. The structure has been confirmed by 43 wells drilled over the field (Bradey et al. 2008). The double-hump structure in part results from thrust displacement over a series of ramps and flats, but it requires a relatively large thrust displacement. No such large-offset thrust or double-humped anticline was recorded in the plane-layered models presented above. It has

Fig. 7. Seismic line PN88-IAG-1 across the Iagifu and Hedinia anticlines that comprise the Kutubu Oilfield. See Figure 1 for location. Note the inferred relatively large offset of the Toro and two-humped nature of the Toro anticline, similar to that in Figure 6e. Note also the required internal deformation within the Ieru in the anticline cores, as shown by tighter folds in the Darai than in the Toro.


long been suspected that Jurassic extensional faults were a controlling factor in the structural evolution in the Papuan Fold Belt (e.g. Buchanan & Warburton 1996) and that the Kutubu Oilfield is underlain by a Jurassic normal fault that may have been reactivated. These concepts were tested by models that included one or two faults cut into the pre-deformation model, each of which was lubricated by Vaseline petroleum jelly smeared along the fault surface. In order to obtain the large thrust offset, the faults were pre-cut at 188 to bedding and lubricated as steeper pre-cut faults tend to lock up during deformation (Dixon 1996). A model was run with a single pre-cut fault from the base of the Imburu analogue to the base of the Darai analogue (Fig. 6e, KL20). The stratigraphy used was the same as that in Figure 6c (Table 2). The pre-cut fault was reactivated early in the deformation and accommodates much of the shortening. The dip of the pre-cut fault evolved to become steep as it passed through the break in the Toro analogue, shallow as it continued through the incompetent Ieru analogue, and steeper again where it propagated through the Darai analogue to the surface of the model. Much of the shortening was channelled to the incompetent Ieru unit, creating laterally displaced structures between the Darai and Toro analogues. A second large fault with a shallower dip developed in the Darai analogue on the foreland side of the reactivated fault. The slip on the pre-cut fault at Toro level is roughly equal to the sum of the slip on these two Darai faults indicating that the pre-cut fault splays upwards, linking to both faults. Furthermore, some of the displacement was channelled towards the foreland along the Ieru dećollement, creating greater shortening and a different structural style within the Darai analogue than in the Toro analogue. The differential shortening laterally displaced the Darai analogue further towards the foreland with respect to the structures that developed in the Toro analogue. This model bears a strong resemblance to the structure of parts of the Kutubu Oilfield (Fig. 7) and to that shown in the SE Hedinia area on the regional cross section (Fig. 3). In particular, the model shows a two-humped Toro anticline in the hanging wall of the pre-cut fault, as seen in the Kutubu anticline and large displacement along the fault. Further, the model shows a detachment in the Ieru, with splays cutting through the Darai towards the foreland of the Toro structure, as recorded in the Zongwe and Ai-io structures on the regional section (Fig. 3). To test the possibility of multiple normal faults beneath the Papuan Fold Belt, two faults were pre-cut in the model, each penetrating from the base Imburu to half way through the Ieru

45

analogue unit. Displacement along the pre-cut faults was simultaneous within the resolution of the experiment, each with equivalent shortening that was channelled to both competent units (Fig. 8). The region between the two pre-cut faults acquired the least amount of deformation, creating a large zone of uplift where the panel between the faults was transported up the frontal fault towards the foreland. Initial thrusting along the pre-cut fault can be seen after only 5% total shortening (Fig. 8a). At 16% total shortening, the faults remain as the dominant feature of the model (Fig. 8b), but as deformation progresses to 44% shortening, younger fold– thrust structures develop independent of the reactivated faults (Fig. 8c). Importantly in these models, the pre-cut faults were reactivated first creating anticlines akin to inversion structures and the fold and thrust structures formed subsequently. If true in the Papuan Fold Belt, it would have important implications for migration and charge of structures (see discussion). Comparing the model shown in Figure 8c with the regional cross section (Fig. 3) it is apparent that there are some broad similarities. Perhaps the structure above the leading pre-cut fault is equivalent to the broad SE Hedinia –Kutubu East culmination, in other words the Strongly Folded Belt. The long hinterland-dipping limb above the pre-cut fault in the model is very similar to that interpreted on seismic data (Hill et al. 2008) used to construct the regional cross section. Similarly, the structure above the more hinterland pre-cut fault in the model could equate to the broad culmination defined by the Mount Castle and Wage structures on the regional section. If this comparison is valid, then it suggests that the two culminations may be associated with pre-existing faults, perhaps Jurassic extensional faults.

Structural style from well and seismic data The Moran anticline The Moran anticline was first drilled in September 1996 and was found to contain an 800 m oil column in Digimu and Toro sandstone reservoirs. Davis et al. (2000) stated that ‘the Moran structure is a narrow, elongate SW-vergent fault-bounded anticline with a moderate to steeply dipping (308 –508) backlimb and an overturned near vertical forelimb’. Davis et al. (2000) and Lingrey (2000) presented cross sections through the Moran 1X, 2X, 1XST and 2XST wells, two of which drilled through the hanging wall across the thrust and tagged the footwall. The sections indicated that the structure was folded first then a break-thrust broke

46

K. C. HILL ET AL.

Fig. 8. Centrifuge analogue modelling of plane-layered strata with two faults pre-cut from the base Imburu to middle Ieru as shown in Fig. 2b. Initial conditions are recorded in Table 2 and Fig. 2. Note that the pre-cut faults are reactivated first making large anticlines (a and b) and the thin-skinned deformation encroaches at higher degrees of shortening (c). See text for discussion.

through the stretched, faulted and boudinaged forelimb. The structure is strongly compartmentalised (Hill et al. 2008), with a fault across the centre separating an 800 m oil column to the west from an equivalent higher-pressure water column at the

same elevation to the east (shown by the abrupt eastern edge of the field boundary on Fig. 1). Here, we present two seismic lines across the Moran anticline and adjacent structures and a cross section based on well data, surface dips and


seismic data and utilizing the centrifuge analogue modelling presented above. Figure 9 shows uninterpreted and interpreted versions of recently reprocessed, migrated seismic line PN96–302 across the Moran anticline through the Moran 1X, 2X, 1XST and 2XST wells. The c. 5 km seismic line illustrates the difficulty in acquiring good quality data across this steep, mountainous, jungle-covered terrain with karstified limestone at surface. Data acquisition is exacerbated by crooked line paths, air-filled caverns, deep fissures, an irregular low velocity weathered zone and a velocity inversion from the Darai Limestone to the underlying Ieru Shale (Lingrey 2000). Seismic acquisition is also limited by the current cost of US$ 100 000/km, in part due to the necessity of helicopter-supported operations. Although the seismic data quality is only poor to moderate, it is still useful in helping to interpret the structure when used in conjunction with surface geology. Indeed, now that the obvious surface anticlines in the Papuan Fold Belt have been drilled, such seismic data are vital in future exploration. On the uninterpreted section of Figure 9, the broad form of the strongly reflecting Darai Limestone can be seen, revealing the position of the Moran Thrust. The inclination of the backlimb can also be determined, both at surface and at depth, away from the core of the structure. It is also possible to infer a potential sub-thrust structure at Toro level, as shown, although this is not proven and relies in part on structural and modelling analogues. Unfortunately, below the Darai Limestone in the core of the structure, almost all of the reflectors are spurious, as shown from the interpretation post-drilling. The steep dips and probable small-scale internal faulting of the beds make imaging of the core of the structure almost impossible. In areas without well control, interpretation of the core relies on structural and analogue models. Figure 10a shows a structural interpretation of the Moran and Paua anticlines as well as the NE limb of the large Mananda anticline that is imaged on the seismic line (Fig. 9). The Moran section is similar to those presented by Lingrey (2000) and Davis et al. (2000), except that the borehole dips have all been reprocessed, there are additional surface dips and the seismic data have been reprocessed. The Mananda part of the section is constrained by additional seismic lines and eight wells on the Mananda anticline along strike and the Paua structure is constrained by two wells projected onto the section. It should be noted that the number of dips shown is only a small sample of those used to construct the section and that the Mesozoic stratigraphic zonation is very fine, with many more horizons correlated than can be

47

shown. Thus the hanging wall anticlines are well constrained. The Middle Jurassic to basement portion of the section is largely unconstrained and schematic. A key part of the interpretation is the ‘regional’ level of the relatively undeformed Middle and Lower Jurassic section. Projecting those beds downdip from the foreland suggests that the top of the Middle Jurassic should be at c. 5 km subsea, consistent with the regional section (Fig. 3). However, interpretation of the seismic data suggests that the top of the Middle Jurassic is at 3–4 km subsea beneath Moran and modelling of earthquake seismic data in the area (Hill et al. 2008) indicates high velocity basement is at 6 km subsea. Therefore, reverse faults in basement have been inferred on the interpretation, as shown. These may be reactivated extensional faults, but this is unproven. As can be seen from comparison of Figure 10a, b, there are strong similarities between the section interpretation and the centrifuge model shown in Figure 6c. Both sections show open folds in the Darai but tight to overturned folds in the Toro– Iagifu (reservoir) section, with large changes in thickness of the intervening Ieru Formation. Both show brittle faulting in the Darai, but more folding in the reservoir section. Furthermore, a detachment within the Ieru is manifested in both sections, connecting the faulting in the Darai above Moran to the thrust fault underlying the Paua anticline. Such structures may well have formed out-of-sequence as the Moran anticline developed in front of and below the Paua anticline. A significant difference between the interpreted and analogue model sections (Fig. 10a, b) is the level of detachment beneath the reservoir. The Moran wells and a few other wells drilled to the fault in the Papuan Fold Belt, combined with the seismic data, strongly indicate a detachment in the Imburu Formation just above the Koi-Iange Formation, c. 800 m below the top of the Toro Sandstone. In contrast, in the analogue model the fundamental detachment is scaled to be 2.5 km below the top of the Toro. However, it should be noted that the level of detachment varies in the Papuan Fold Belt and that there is good evidence from seismic data (better quality than in Fig. 9) that the detachment is deeper, c. 1.5–2.0 km below the top Toro, in the core of the Mananda anticline, the core of the Kutubu Oilfield (Fig. 7) and the northeastern part of the regional section (Fig. 3). Another difference is the degree of thickness changes in the Ieru, which is considerably greater in the analogue model. It may be that the interpreted cross section underestimates Ieru thickness changes in the synclines; for instance the Darai keel between the Moran and Mananda

48

K. C. HILL ET AL.

Fig. 9. Seismic line PN96-303 through the Moran-1 and Moran-2 wells, blank and interpreted. The well ticks are horizon tops, not dips. The data illustrate the limitations of seismic acquisition and interpretation in PNG. However, it is possible to infer the shape of the Darai Limestone, the location of the Moran Thrust and potential sub-thrust structures. See text for discussion.


Fig. 10. (a) Structural cross section of the Moran and Paua anticlines based on detailed well dips and stratigraphy, surface mapping and seismic data. The well ticks are representative dips. This section has not been balanced. (b) Expanded view of the middle of Figure 6c showing a strong similarity between the interpreted section and the analogue model.

49

50

K. C. HILL ET AL.

anticlines could be shallower, as it is only constrained by poor quality seismic data. Detailed analysis of the Moran cross section (Fig. 10a) reveals several features that suggest a structural evolution (discussed later). In the core of the structure a low-angle, folded thrust fault was encountered at three locations in the boreholes, with c. 50 –100 m offset. The wells that drilled the forelimbs of both the Moran and Paua structures found a relatively complete stratigraphic sequence, but 30 –50% of the normal thickness and cut by faults. The Moran Fault was interpreted to comprise a zone of faults through these stretched beds and is itself offset by a backthrust encountered in three wells. Within the Ieru Formation, the Bawia Mudstone is a weak and mobile bed in which the thickness is highly variable, probably due to tectonism. This horizon is interpreted to be the main detachment horizon within the Ieru Formation as dips are highly variable above it with common thrust splays but dips below are relatively consistent. One further feature is the interpretation of ‘out-of-the-syncline’ thrust faults (Dahlstrom 1970), for which the main evidence is offsets and increased thickness of the Darai Limestone. Combining those features suggests the following structural evolution: (1) Minor low-angle thrust faults developed through parts of the structure. (2) A fault-propagation fold developed with a stretched overturned forelimb, including folding of some of the existing low-angle thrusts.

(3)

With continued shortening, the fold evolved into a break-thrust. (4) The next structure towards the foreland, the Mananda anticline, was thrusted and folded, jacking up the existing structures and generating out-of-the-syncline thrusts and backthrusts due to the resultant space problems. In terms of hydrocarbons, it seems likely that the 800 m oil column is preserved, in part, due to tight folding of the reservoir so that it is encased in mudstone of the Ieru Formation, the regional seal. The forethrusts and backthrusts have effectively made the crest of the anticline a pop-up structure that may have helped to isolate and preserve the hydrocarbon column. It is notable that the adjacent Paua structure encountered a residual oil column, minor gas and high pressures and that the same was true for the eastern half of the Moran anticline across an important tear fault (Hill et al. 2008). The Mananda structure to the southeastern preserved a small oil column at low pressure in an isolated crest at its SE end, the SE Mananda field (Fig. 1).

Moran – Agogo structure Figure 11 shows c. 9 km of an interpreted version of a recently reprocessed, migrated seismic line PN07–505 across the Agogo Oilfield and the southeastern part of the Moran anticline. This SE part of the Moran structure tested water at high pressure instead of oil. Although imaging of the core of the Moran structure is still poor, probably due to

Fig. 11. Structural interpretation of seismic line PN07-505 across the Agogo oilfield and SE Moran anticline. See Figure 1 for location. The Agogo anticline, the adjacent syncline and even the underlying basement are relatively well imaged.


steep dips and faulting, the seismic and nearby Moran-3X well data (Fig. 1) suggest that the structure is a thrust ramp, with a less well developed forelimb than the oil bearing part of the structure shown on Figure 10. The SE Moran structure is perhaps more like the thrust Toro structure near the centre of Figure 6e than the folds with stretched forelimbs near the centre of Figure 6c. The Agogo anticline is along strike, but slightly en echelon to the SE Mananda anticline, hence is further away from the Moran Thrust. Thus the syncline between the Moran and Agogo structures is preserved rather than faulted out as it was on Figures 9 and 10. In consequence, the Darai Limestone to Jurassic section is much better imaged, particularly within the Agogo structure. It is also possible to infer some underlying basement structure as shown on the section. It is notable that the structural style on this seismic line is not exactly like any of the centrifuge analogue models presented here. In particular, the consistent large thrust offset of Toro Sandstone and Darai Limestone differs from the models, but resembles the Toro offset near the centre of Figure 6e and towards the right hand side of Figure 8c.

Discussion Centrifuge analogue modelling The influence of mechanical stratigraphy upon structural style was demonstrated by the experiments presented here, as shown by the different structural styles in Figure 6. The thickness of the competent beds remained the same in these experiments, so the variations in structural style were due to subtle changes in the relative strength and thickness of the weaker layers. Thin alternating strong–weak –strong– weak layers with large competence contrasts resulted in imbricate thrusts in the strong layers with thrust spacing proportional to the bed thickness (Fig. 6a). The intervening weak layers acted as ductile detachments that accommodated the variable strain, so that thrusting in the strong layers was disharmonic, as previously recorded by Dixon (1996). Doubling the thickness of the lower weak layer above ‘basement’ profoundly changed the structural style with more buckle folds developing that were variably harmonic and disharmonic in the overlying competent units (Fig. 6b). The basal weak layer above basement in PNG is the syn-rift sequence, so modelling thickness changes as in Figure 6a, b suggests how structural styles might change across a syn-rift growth fault. Unfortunately it was not possible to run a model with a step in basement. However, Dixon (2004) modelled lateral changes in mechanical stratigraphy

51

to represent reef versus off-reef facies, showing abrupt changes in structural style at the boundary. The modelling presented here indicates that in PNG and elsewhere, the structural style would similarly change dramatically across an old extensional fault with thick, weak syn-rift strata on one side (Fig. 6b) and thin equivalent strata on the adjacent high (Fig. 6a). The large folds preferentially formed over the syn-rift strata (Fig. 6b) would be more prospective as hydrocarbon traps. Slightly strengthening the upper weak layer in the models produced dominantly buckle folds in the middle competent layer, the Toro reservoir analogue, and thrusting in the upper, thick competent layer, the Darai analogue (Fig. 6c). These models were most like the known fold belt structures, such as in the Moran anticline. Significantly, the models were able to produce tight, overturned folds in the Toro reservoir section with break-thrusts through the attenuated forelimb as recorded in the Moran, Paua and SE Hedinia anticlines (this paper) and the Usano and Hedinia anticlines (e.g. Lamerson 1990; Franklin & Livingston 1996). Furthermore, the models recorded open folds in the more competent Darai and common faults cutting the Darai, many of which were splays from a Ieru detachment that rooted back to the previous Toro anticline, towards the interior of the fold belt. The models also recorded dramatic thickness changes within the Ieru as proven in many wells. The importance of the models is not in the area of known structures, but as an aid to interpretation in areas under exploration with little subsurface data and poor to moderate quality seismic data. When a lubricated pre-cut fault dipping at 188 through the Mesozoic analogue was introduced to the model (Figs 6e & 8) it was reactivated early in the deformation and recorded a large displacement. In the hanging wall an overturned fold was generated at the leading edge (Fig. 8). This large offset, particularly in the centre of Figure 8c, is similar to the structure interpreted within the Kutubu Oilfield, the largest oilfield in PNG (Fig. 7). The modelling did not otherwise produce faults with large offset. A comparison of Figure 6c and e shows the impact of the pre-cut faults in sections that otherwise had identical stratigraphy. With the exception of the pre-cut fault the structural style is the same. This suggests that a pre-existing fault or weakness was necessary to produce the Iagifu –Hedinia anticlines that contain the Kutubu Oilfield (Fig. 7). It is debatable whether or not a pre-cut fault dipping at 188 can represent the effect of an old extensional fault, which would be expected to have a dip of 45 –608. Dixon (1996) modelled the reactivation of pre-cut faults, but found that with steeper pre-cut faults the amount of reactivation

52

K. C. HILL ET AL.

decreased, such that there was minimal reactivation of faults dipping steeper than c. 458. An important consideration is that Dixon’s pre-cut faults did not have thicker, weak (syn-rift) beds in the hanging wall so did not simulate a Jurassic growth fault offsetting basement. However, Dixon’s later experiments with an abrupt facies change from weak to strong beds across a vertical contact (Dixon 2004) did generate a fault with large offset at the contact, albeit at relatively high levels of shortening. Importantly these structures had an overturned fold in the hanging wall, as recorded by drilling at the leading edge of the Hedinia anticline in the Kutubu Oilfeld (Bradey et al. 2008). Combining the results from the models shown in Figure 6a, b with the pre-cut fault models and Dixon’s (2004) facies change models, it seems very likely that an old extensional fault with a thick, weak syn-rift section would be reactivated early in the deformation as a thrust fault. Furthermore, this fault would propagate to have a large displacement compared to other faults in the area. The Hedinia Thrust beneath the Kutubu and SE Hedinia oil and gas fields (Figs 3 & 7) is probably such a fault.

PNG structural style and hydrocarbon prospectivity In the introduction, the Timor Sea area was proposed as an analogue for the pre-deformation architecture of the Papuan Fold Belt (Fig. 4; Cooper et al. 1996). In order to assess hydrocarbon prospectivity, this concept is reviewed further here, incorporating the results from the structural sections and centrifuge modelling experiments. The Timor Sea lies between the Bonaparte Basin to the NE with gas reserves of 28 TCF and the Browse basin to the SW with gas reserves of 30 TCF (Australian Government 2007). Within the Timor Sea, the Vulcan Sub-basin (Fig. 4) contains a handful of medium-sized oilfields with total reserves of 357 million barrels (Longley et al. 2002), similar in size to those of the Papuan Fold Belt. Rifting to form the Vulcan Sub-basin occurred in Late Jurassic to Early Cretaceous times (Pattillo & Nicholls 1990) following Late Triassic to Mid Jurassic rifting in PNG (Home et al. 1990), as part of the same break-up of the north Australian margin (Veevers 2000). Kennard et al. (1999) concluded that oil generation and expulsion were restricted to Oxfordian– Kimmeridgian syn-rift source rocks principally within the Swan and Paqualin grabens and the deepest (SW) portion of the Cartier Trough (Fig. 4). Thus the oil source rock was focussed in local deep graben within an otherwise gas-prone area.

The regional cross section presented here (Fig. 3) illustrated a number of different structural styles across the Papuan Fold Belt, which can be related to features within the Vulcan Sub-basin. The section showed that the Darai Fault (Fig. 3) is an important basin-bounding fault across which there is significant growth in several stratigraphic units demonstrating long-lived extensional activity. This is akin to the major basin-bounding fault that abuts the Londonderry High such that the sediments beneath the Darai Plateau may be comparable to those in the Skua syncline (Fig. 4b). In PNG, the basin-bounding extensional fault is interpreted from seismic data to continue along much of the front of the Papuan Fold Belt, although occasionally offset across tear faults such as the Bosavi Lineament (Fig. 1). The Paqualin Transfer across the Vulcan Sub-basin (Fig. 4a; Woods 1992) is considered to be a basement-controlled feature that probably resembled the Bosavi Lineament prior to compressional deformation. The Darai Fault was probably inverted in Early Tertiary times, was definitely inverted in Pliocene times and remains active, as indicated by compressional earthquakes. Importantly, the current basement inversion occurred prior to any thin-skinned deformation as the nearest thin-skinned structure is 40 km to the north (Fig. 3). This is consistent with the results of the centrifuge analogue modelling with a pre-cut fault (Fig. 8 and Dixon, 1996) in which the existing faults were reactivated in compression at low amounts of shortening prior to the onset of thin-skinned deformation. If inverted early in the deformation sequence, it would be reasonable to expect structures such as the Darai Plateau to trap any subsequent hydrocarbon charge, yet all wells in that area are dry with few oil shows. As the structures to the north are charged, there must be a barrier to migration between the Strongly Folded and Gently Folded Belts (Figs 1 & 3). The Strongly Folded Belt contains all of the commercial oil reserves found in PNG, with large gas discoveries to the NW at Hides and to the SE in the Gulf of Papua. The belt is underlain by a large-offset thrust fault and has accompanying large hanging wall anticlines (Figs 3, 7 & 11) that appear to be less common elsewhere in the fold belt. The centrifuge analogue modelling suggests that a pre-existing weakness is required to generate such large-offset faults. It seems likely that, prior to compressional deformation, the Kutubu, Agogo and SE Hedinia oil and gas fields (Figs 1, 3, 7 & 11) were underlain by a significant normal fault, perhaps the southern bounding fault of a deep graben similar to the Swan Graben on Figure 4 (see hypothetical graben on Fig. 3). The graben could have been the source kitchen for all the oil and been responsible for the development of the


large-offset fault that created the Strongly Folded Belt, hence supplying both trap and charge. The significant normal fault may be the barrier to migration into the Gently Folded Belt structures to the SW.

Proposed structural evolution Combining the structural observations from the cross sections presented and the analogue modelling the following structural evolution is suggested. † The pre-compression architecture of the margin consisted of a stable platform in the south bounded by large Late Triassic to Mid Jurassic extensional faults with thick syn-rift sediments beneath the future Darai Plateau. Further north, beneath the future Strongly Folded Belt, a second major normal fault is inferred, bounding the area to the north that contained the main oil source rocks. This may have been a deep graben (Fig. 3). † During early compressional deformation, in the Late Miocene, there was probably minor reactivation of basement faults as shown on Figure 8a, including the fault beneath the Strongly Folded Belt. † Thin-skinned thrusting occurred along an Imburu detachment that was in some areas between the Iagifu and Koi-Iange sandstones (Fig. 10a) and in other areas below the Koi-Iange (Figs 3, 7 & 6c). The transition from one detachment level to another probably occurs across transfer zones that may be old basement faults similar to the Paqualin Transfer on Figure 4. † In the Strongly Folded Belt, the inverted normal fault was reactivated as the Hedinia Thrust and accommodated 4– 8 km of shortening, building the Iagifu and Hedinia anticlines that comprise the Kutubu Oilfield (Figs 1 & 7). † As the thin-skinned structures propagated over the source kitchen beneath the Strongly Folded Belt, oil was generated and expelled. † In areas such as Moran, a fault-propagation fold developed with a stretched overturned forelimb, including folding of some of the existing thin-skinned thrusts. This evolved into a break-thrust. † Thrusting and folding of the next structure towards the foreland occurred, jacking up the existing structures and generating out-of-thesyncline thrusts and backthrusts due to the resultant space problems. † As orogenesis propagated further towards the SW, inversion occurred creating the Darai Plateau and was probably accompanied by renewed basement reverse faulting beneath the fold belt, perhaps reactivating old extensional faults.

53

Future work To determine the hydrocarbon prospectivity of the Papuan Fold Belt, understanding the structural style is vital and this paper attempts to address that issue. However, equally as important is understanding which faults seal and which faults leak and/or breach the structures. Furthermore, the timing of fault sealing with respect to hydrocarbon charge is important as fault seal parameters will change with the changing stress regime (e.g. Castillo et al. 2000). Crowhurst et al. (1997) argued that the compression direction was NE –SW in the Late Miocene, but changed to more E-W in the mid Pliocene. Such a stress rotation would have a significant effect on the sealing capabilities of faults, particularly cross-cutting faults. The importance of sealing or breaching faults is demonstrated by the Moran and Paua structures (Figs 1 & 10). The Moran oilfield resides in the NW half of the anticline where there is an 800 m oil column, separated by a sealing cross fault from the SE portion of the structure that tested water at elevated pressures. To the NE the Paua anticline encountered very high pressures, a residual oil column and minor gas and is thought to have been breached and then resealed. Clearly the dip-slip faults between Moran and Paua must seal, yet may previously have been open to allow hydrocarbon charge. Resolution of this issue is beyond the scope of this paper, but is the aim of ongoing studies.

Conclusions (1)

(2)

(3)

(4)

(5)

Centrifuge analogue modelling of intact, plane-layered strata determined that the mechanical stratigraphy and the thickness of weak strata above the lower dećollement horizon exert the greatest control on the changing structural styles of the Papuan Fold Belt. The models most like known structures had a thick, incompetent Jurassic shale sequence, a competent Toro reservoir sequence, an intermediate competence Ieru sequence and a competent Darai sequence. These models produced tight overturned folds in the Toro, thickness changes and detachments in the Ieru and open folds and thrust faults in the Darai, all recorded in the Moran anticline. Centrifuge analogue modelling with pre-cut faults in the Mesozoic section produced early inversion anticlines as recorded in the Papuan Fold Belt by inversion of the Darai Fault. With continued shortening the pre-cut faults accommodated much of the shortening as large-offset thrust faults that did not otherwise occur in the models.

54

(6)

(7)

(8)

(9)

K. C. HILL ET AL.

The large-offset faults produced a ‘doublehump’ hanging wall similar to that in the Kutubu Oilfield, PNG’s biggest oilfield with .350 million barrels original oil in place. The undeformed PNG margin probably comprised a small number of large basin-bounding or graben-bounding faults across which there was substantial growth in the stratigraphic section. Away from those faults the Mesozoic section probably thickened gradually towards the NE. The seismic data, structural sections and analogue modelling defined the important elements of the PNG structural style. These are: (a) Inverted basement faults. (b) Thin-skinned faults detached in the Imburu at various depths of 1–2 km below the top Toro. (c) Tight and overturned anticlines in the Toro with long, continuous, steep backlimbs and stretched and boudinaged forelimbs cut through by a zone of break-thrusts. (d) Highly variable thickness within the Ieru, with the major detachment within the incompetent Bawia Member. This detachment is often reactivated out-ofsequence as the next structure towards the foreland forms. (e) Open folds in the competent Darai Limestone, cut by numerous thrust faults linking back into the Ieru detachments. (f) Cross-cutting or tear faults in places linked to old basement faults. These faults often seal and compartmentalise the reservoirs. The transition from one detachment level to another probably occurs across these faults. Understanding the structural style is only the first step in determining the hydrocarbon prospectivity. The next, more difficult, step is to determine the hydrocarbon charge, fluid pressures and sealing and/or breaching nature of faults.

Centrifuge analogue modelling was made possible by the support and guidance of Dr John M. Dixon and the research facilities at Queen’s University. The modelling research was supported by the Ontario Graduate Scholarship programme, the Natural Sciences and Engineering Research Council of Canada (NSERC) and the industry sponsors of the Fold– Fault Research Project (FRP). Peter Hamilton and Simon Skirrow of Oil Search Ltd are thanked for excellent drafting. Oil Search Ltd and Joint Venture partners gave permission for publication of this paper and use of any proprietary seismic data, sections and images within it. This study builds on the previous work of many geologists and geophysicists, particularly at

Chevron, BP, Esso, Mobil and the PNG Geological Survey. The paper greatly benefited from the helpful comments of Chris Elder and an anonymous reviewer. The interpretations presented here are strictly those of the authors and do not necessarily reflect the views of any of the companies involved.

References APC 1961. Geological results of petroleum exploration in western Papua. Journal of the Geological Society of Australia, 8, 133. AUSTRALIAN GOVERNMENT 2007. Australian Liquified Natural Gas (LNG) – clean energy for a secure future. Australian Government Publication, 20p. Bally, A. W., Gordey, P. L. & Stewart, G. A. 1966. Structure, seismic data and orogenic evolution of the southern Canadian Rocky Mountains. Bulletin of Canadian Petroleum Geology, 14, 337 –381. Bradey, K., Hill, K., Lund, D., Williams, N., Kivior, T. & Wilson, N. 2008. Kutubu oil field, Papua New Guinea – a 350 mmbl fold belt classic. In: Blevin, J. E., Bradshaw, B. E. & Uruski, C. (eds) Eastern Australasian Basins Symposium III. Petroleum Exploration Society of Australia, Special Publication, 239–246. Buchanan, P. G. & Warburton, J. 1996. The influence of pre-existing basin architecture in the development of the Papuan Fold and Thrust Belt: implications for petroleum prospectivity. In: Buchanan, P. G. (ed.) Petroleum Exploration, Development and Production in Papua New Guinea. Proceedings of the Third PNG Petroleum Convention, Port Moresby September 1996, 89–109. Castillo, D. A., Moos, D., Barton, C. A., Zoback, M. D. & Peska, P. 2000. Reservoir geometrics applied to drilling and completion programmes in hostile and complex environments: North West Shelf, Timor Sea, North Sea and Columbia. In: Buchanan, P. G., Grainge, A. M. & Thornton, R. C. N. (eds) Papua New Guinea’s Petroleum Industry in the 21st Century. Proceedings of the Fourth PNG Petroleum Convention, Port Moresby, 487–508. Chambers, J. L. C., Carter, I., Cloke, I. R., Craig, J., Moss, S. J. & Paterson, D. W. 2004. Thin-skinned and thick-skinned inversion-related thrusting – a structural model for the Kutai Basin, Kalimantan, Indonesia. In: McClay, K. R. (ed.) Thrust Tectonics and Hydrocarbon Systems. AAPG Memoir 82, Chapter 31, 614– 634. Cole, J. P., Parish, M. & Schmidt, D. 2000. Sub-thrust plays in the Papuan Fold Belt: the next generation of exploration targets. In: Buchanan, P. G., Grainge, A. M. & Thornton, R. C. N. (eds) Papua New Guinea’s Petroleum Industry in the 21st Century. Proceedings of the Fourth PNG Petroleum Convention, Port Moresby, 87–99. Cooper, G. T., Hill, K. C. & Baxter, K. 1996. Rifting in the Timor Sea and New Guinea: a template for compressional forward modelling. In: Buchanan, P. G. (ed.) Petroleum Exploration, Development and Production in Papua New Guinea. Proceedings of the

STRUCTURAL STYLES IN THE PAPUAN FOLD BELT Third PNG Petroleum Convention, Port Moresby September 1996, 133–146. Crowhurst, P. V., Hill, K. C. & Foster, D. A. 1997. The structural and tectonic development of the Frieda River mineral district, NW Papua New Guinea. In: Hancock, G. E. (ed.) Proceedings of the PNG Geology, Exploration and Mining Conference. 1997 Madang, Australian Institute of Mining and Metallurgy, Melbourne, 51– 60. Crowhurst, P. V., Maas, R., Hill, K. C., Foster, D. A. & Fanning, C. M. 2004. Isotopic constraints on crustal architecture and Permo-Triassic tectonics in New Guinea: possible links with eastern Australia. Australian Journal of Earth Science, 51, 107–122. Dahlstrom, C. D. A. 1970. Structural geology in the eastern margin of the Canadian Rocky Mountains. Bulletin of Canadian Petroleum Geology, 18, 332–406. Davis, K., Pederson, K., Todd, B. & Wali, K. 2000. Integrated geological and engineering evaluation of central Moran field, Papua New Guinea. In: Buchanan, P. G., Grainge, A. M. & Thornton, R. C. N. (eds) Papua New Guinea’s Petroleum Industry in the 21st Century. Proceedings of the Fourth PNG Petroleum Convention, Port Moresby, 397 –425. Dixon, J. M. 1995. Centrifuge modelling in the Foothills Research Project, 1995: thrust reactivation of normal faults and facies boundaries. Foothills Research Project, Research Report, University of Calgary, 1, 5-1– 5-22. Dixon, J. M. 1996. Physical model investigation of the influence of early extensional (growth) faults on fold–thrust structures, with application to the Papuan fold and thrust belt. In: Buchanan, P. G. (ed.) Petroleum Exploration, Development and Production in Papua New Guinea. Proceedings of the Third PNG Petroleum Convention, Port Moresby September 1996, 147–160. Dixon, J. M. 2004. Physical (centrifuge) modelling of fold-thrust shortening across carbonate bank margins – timing, vergence, and style of deformation. In: McClay, K. R. (ed.) Thrust Tectonics and Hydrocarbon Systems. AAPG Memoir 82, Chapter 12, 223–238. Dixon, J. M. & Liu, S. 1992. Centrifuge modelling of the propagation of thrust faults. In: McClay, K. (ed.) Thrust Tectonics. Chapman & Hall, London, 53–70. Dixon, J. M. & Summers, J. M. 1985. Recent developments in centrifuge modelling of tectonic processes: equipment, model constructions techniques and rheology of materials. Journal of Structural Geology, 7, 83– 102. Dixon, J. M. & Tirrul, R. 1990. Centrifuge modelling of fold–thrust structures in a tripartite stratigraphic succession. Journal of Structural Geology, 13, 3 –20. Dixon, J. M., Blackburn, J. & Mastromatteo, D. 1996. Physical modelling of fold–thrust processes: reactivation of extensional (growth) faults, linkage of shortening between structural levels, and formation of pop-ups and triangle zones. Foothills Research Project, Research Report, University of Calgary, 2, 8-1– 8-47. Eisenberg, L. I. 1993. Hydrodynamic character of the Toro Sandstone, Iagifu/Hedinia area, Southern

55

Highlands Province, Papua New Guinea. In: Carman, G. J. & Carman, Z. (eds) Proceedings of the Second PNG Petroleum Convention, 1993. Port Moresby, June 1993, 447–458. Findlay, A. L. 1974. The structure of the foothills south of the Kubor Range, Papua New Guinea. Australian Petroleum Production and Exploration Association Journal, 14, 14–20. Franklin, S. P. & Livingston, J. E. 1996. Development of an infill well program to maximise economic return from the Iagifu– Hedinia Field: Part I. Integrated structural, stratigraphic and reservoir attribute modelling as input to reservoir simulation and well targeting. In: Buchanan, P. G. (ed.) Petroleum Exploration, Development and Production in Papua New Guinea. Proceedings of the Third PNG Petroleum Convention, Port Moresby September 1996, 573– 590. Hill, K. C. 1991. Structure of the Papuan Fold Belt, Papua New Guinea. Bulletin American Association of Petroleum Geologists, 75, 857– 872. Hill, K. C. & Gleadow, A. J. W. 1989. Uplift and thermal history of the Papuan Fold Belt, Papua New Guinea: Apatite Fission tack analysis. Australian Journal of Earth Science, 36, 515–539. Hill, K. C. & Gleadow, A. J. W. 1990. Apatite Fission Track Analysis of the Papuan Basin. In: Carman, G. J. & Carman, Z. (eds) Petroleum Exploration in Papua New Guinea. Proceedings of the First PNG Petroleum Convention, 1990, Port Moresby, February 1990, 119– 136. Hill, K. C. & Hall, R. 2003. Mesozoic–Tertiary Evolution of Australia’s New Guinea Margin in a West Pacific Context. In: Hillis, R. R. & Muller, R. D. (eds) Evolution and Dynamics of the Australian Plate. Geological Society of Australia Special Publication 22 and Geological Society of America Special Paper 372, 265– 290. Hill, K. C. & Raza, A. 1999. Arc– continent collision in Papua Guinea: constraints from fission track thermochronology. Tectonics, 18, 950– 966. Hill, K. C., Norvick, M. S., Keetley, J. T. & Adams, A. 2000. Structural and Stratigraphic Shelf-Edge Hydrocarbon Plays in the Papuan Fold Belt. In: Buchanan, P. G., Grainge, A. M. & Thornton, R. C. N. (eds) Papua New Guinea’s Petroleum Industry in the 21st Century. Proceedings of the Fourth PNG Petroleum Convention, Port Moresby, 67– 85. Hill, K. C., Bradey, K., Iwanec, J., Wilson, N. & Lucas, K. 2008. Structural exploration in the Papua New Guinea Fold Belt. In: Blevin, J. E., Bradshaw, B. E. & Uruski, C. (eds) Eastern Australasian Basins Symposium III. Petroleum Exploration Society of Australia, Special Publication, 225–238. Hobson, D. M. 1986. A thin skinned model for the Papuan thrust belt and some implications for hydrocarbon exploration. The Australian Petroleum Production and Exploration Association (APPEA) Journal, 26, 214– 224. Home, P. C., Dalton, D. G. & Brannan, J. 1990. Geological Evolution of the Western Papuan Basin. In: Carman, G. J. & Carman, Z. (eds) Petroleum Exploration in Papua New Guinea. Proceedings of the First PNG Petroleum Convention, Port Moresby, 107– 117.

56

K. C. HILL ET AL.

Hornafius, J. S. & Denison, R. E. 1993. Structural interpretations based on strontium isotope dating of the Darai limestone, Papuan Fold Belt, Papua New Guinea. In: Carman, G. J. & Carman, Z. (eds) Petroleum Exploration in Papua New Guinea. Proceedings of the First PNG Petroleum Convention, Port Moresby, 313–324. Jenkins, D. A. L. 1974. Detachment tectonics in western Papua New Guinea. Bulletin of the Geological Society of America, 85, 533–548. Johnstone, D. C. & Emmett, J. K. 2000. Petroleum geology of the Hides gas field, Southern Highlands, Papua New Guinea. In: Buchanan, P. G., Grainge, A. M. & Thornton, R. C. N. (eds) Papua New Guinea’s Petroleum Industry in the 21st Century. Proceedings of the Fourth PNG Petroleum Convention, Port Moresby, 319–335. Kennard, J. M., Deighton, L., Edwards, D. S., Colwell, J. B., O’brien, G. W. & Boreham, C. J. 1999. Thermal history modelling and transient heat pulse: New insights into hydrocarbon expulsion and ‘hot flushes’ in the Vulcan sub-basin, Timor Sea. The Australian Petroleum Production and Exploration Association (APPEA) Journal, 39, 177– 207. Lamerson, P. R. 1990. Evolution and structural interpretations in Iagifu/Hedina field, Papua New Guinea. In: Carman, G. J. & Carman, Z. (eds) Petroleum Exploration in Papua New Guinea. Proceedings of the First PNG Petroleum Convention, Port Moresby, 283– 300. Lingrey, S. 2000. Structural interpretation and modelling of seismic data from the Moran and Paua area, PNG Foldbelt. In: Buchanan, P. G., Grainge, A. M. & Thornton, R. C. N. (eds) Papua New Guinea’s Petroleum Industry in the 21st Century. Proceedings of the Fourth PNG Petroleum Convention, Port Moresby, 385–396.

Liu, S. & Dixon, J. M. 1991. Centrifuge modelling of thrust faulting: structural variation along strike in fold–thrust belts. Tectonophysics, 188, 39– 62. Longley, I. M., Buessenschuett, C. et al. 2002. The North West Shelf of Australia – a Woodside perspective. In: Keep, M. & Moss, S. J. (eds) The Sedimentary Basins of Western Australia 3. Proceedings of the Petroleum Exploration Society of Australia Symposium, Perth, WA, 27–88. Pattillo, J. & Nicholls, P. J. 1990. A tectonostratigraphic framework for the Vulcan Graben, Timor Sea region. Australian Petroleum Exploration Association Journal, 30, 27– 51. Smith, J. G. 1965. Orogenesis in western Papua and New Guinea. Tectonophysics, 2, 1 –27. Thornton, R. C. N., Emmett, J. K., Lyslo, J. A. & Gottschalk, R. R. 1996. Integrated structural and stratigraphic analysis in PPL 175, Papuan Fold Belt, Papua New Guinea. In: Buchanan, P. G. (ed.) Petroleum Exploration, Development and Production in Papua New Guinea. Proceedings of the Third PNG Petroleum Convention, Port Moresby September 1996, 195–216. Van Wyck, N. & Williams, I. S. 2002. The age and provenance of basement metasediments from the Kubor and Bena Bena blocks, central Highlands, Papua New Guinea – constraints on the tectonic evolution of the northern Australian cratonic margin. Australian Journal of Earth Sciences, 49, 565–577. Veevers, J. J. 2000. Billion-year Earth History of Australia and Neighbours in Gondwanaland. GEMOC Press, Sydney. Woods, E. P. 1992. Vulcan Sub-basin fault styles – implications for hydrocarbon migration and entrapment. Australian Petroleum Exploration Association Journal, 32, 138–158.

Ductile duplexes as potential natural gas plays: an example from the Appalachian thrust belt in Georgia, USA BRIAN S. COOK1,2* & WILLIAM A. THOMAS1 1

University of Kentucky, Department of Earth and Environmental Sciences, 101 Slone Building, Lexington, Kentucky 40506, USA 2

Southwestern Energy Company, 2350 N. Sam Houston Pkwy E. Ste. 125, Houston, Texas 77032, USA *Corresponding author (e-mail: [email protected]) Abstract: In a well-defined small-scale recess in the Appalachian thrust belt in northwestern Georgia (USA), two distinct regional strike directions intersect at c. 508. Fault intersections and interference folds enable tracing of both structural strikes. Around the recess, tectonically thickened weak stratigraphic layers – shales of the Cambrian Conasauga Formation – accommodated ductile deformation associated with the folding and faulting of the overlying Cambrian– Ordovician regional competent layer. The structures in the competent layer are analogous to those over ductile duplexes documented along strike to the SW in Alabama, where gas production has been established from the deformed shale. The analogy with structures in Alabama suggests a ductile duplex and natural gas potential within the recess in Georgia. The tectonic thickening of the weak-layer shales is evident in palinspastically restored cross sections, which demonstrate a nearly 100% increase in volume over the restored state cross sections. The dominant cause of the surplus shale volume is likely pre-thrusting deposition of thick shale in a basement graben that was later inverted. The volume balance of the ductile duplex is critical for palinspastic reconstruction of the recess, and for the kinematic history and mechanics of the ductile duplex.

In the southern part of the Appalachian thrust belt in eastern North America, recently established gas production from ductile duplexes in Cambrian shales in the state of Alabama has generated interest in further exploration. In the southern Appalachians, the regional dećollement is in Cambrian shales near the base of a Cambrian –Pennsylvanian stratigraphic succession above Precambrian crystalline basement rocks. In Alabama, the depositional thickness of the Cambrian shale was controlled by basement grabens, the boundary faults of which later localized frontal thrust ramps (Thomas 2001). Beneath some frontal thrust ramps in Alabama, the Cambrian shale is tectonically thickened in ductile duplexes (mushwads), above which the overlying regional competent layer (Cambrian –Ordovician massive carbonate rocks) is non-systematically deformed. A ductile duplex is formed by a thick, weak stratigraphic layer that is deformed in incompetent horses between floor and roof thrusts (Thomas 2001). Exploration for gas in the Gadsden mushwad in Alabama (Fig. 1) began in 1985, and production was established in 2005. Fourteen wells have been drilled into the Cambrian shales (Conasauga Formation); initial production test rates ranged from 26 to 233 thousand cubic feet per day (Mcfd) (Williams 2007). In August 2007, eight wells produced a total of 6.8 million

cubic feet (MMcf), and the most productive of these wells produced 2.684 MMcf (Alabama State Oil & Gas Board 2007). The purpose of this paper is to consider possibly analogous structures along strike to the NE in the state of Georgia as potential targets for natural gas exploration.

Regional setting of southern Appalachian structures in Georgia Bends in the gross-scale structural trend of the Appalachian thrust belt have been recognized for well over 100 years (Willis 1893). Regionally, the Appalachian thrust belt includes the gradually curved Tennessee salient, convex toward the craton in the direction of thrust translation, and the more angular bend of the Alabama recess, concave toward the craton (Thomas 1977). At a small-scale recess in northwestern Georgia, north-northeastward-striking thrust faults and related folds in the southern arm of the Tennessee salient intersect east-northeastwardstriking thrust faults and related folds that diverge from the predominant strike of the eastern arm of the Alabama recess (Fig. 1). The Sequatchie anticline (Fig. 1), along the northwestern structural front of the southern Appalachian thrust belt (Thomas & Bayona 2005), has


58

B. S. COOK & W. A. THOMAS

Fig. 1. Structural outline map of the Appalachian thrust belt in Alabama and Georgia, modified from Thomas (2007). The black rectangle shows the location of the more detailed map in Figure 2. Names of faults are in all capital letters. The Floyd synclinorium is labelled as Fs, Gadsden mushwad as Gm. The label ‘Birmingham’ shows the location of both the surface thin-skinned Birmingham anticlinorium and the subsurface Birmingham basement graben.

a remarkably straight axial trace trending c. 040 and extending from the front of the Alabama recess on the SW to a tangent near the apex of Tennessee salient on the NE. The straight trace crosses the foreland with no deflection in strike at the smallscale recess in Georgia. Parallel to and SE of the Sequatchie anticline, the frontal Appalachian structures are characterized by narrow anticlines and broad flat-bottomed synclines, the southeasternmost of which is the Lookout Mountain syncline (Fig. 1). In contrast, along the trailing edge of the Appalachian sedimentary thrust belt, the Cartersville and Great Smoky faults mark the leading edges of metamorphic thrust sheets and intersect at c. 708 (Fig. 1). In Alabama, the Cartersville and related Talladega faults generally parallel the regional 040 trend of the thrust belt; but in Georgia, the Cartersville fault bends to 070 and intersects the Great Smoky fault, which trends c. 000 (Fig. 1). The intersection of the Cartersville and Great Smoky faults is the most pronounced surface expression of the two regional structural trends in the small-scale recess in Georgia (Thomas & Bayona 2005).

In the trailing part of the Appalachian sedimentary thrust belt (in the immediate footwall of the bend in the Cartersville/Great Smoky fault system), the trend of the Eastern Coosa fault bends abruptly from 020 on the north to 070 on the SW, framing the small-scale recess (Fig. 1). Where the fault bends abruptly in strike, several trailing splays extend southward, continuing along the direction of strike of the north-northeast-striking leading fault (Fig. 2). The intersection between the Eastern Coosa fault and the trailing splays in the hanging wall defines a clear interference pattern between the two dominant strike directions of the leading fault. Further southwestward in Alabama, the 070-striking segment of the Coosa fault merges into the predominant 040-trending Appalachian structures (Fig. 1). In intermediate structures between the sharply bent Eastern Coosa fault and the nearly straight frontal structures (e.g. Lookout Mountain syncline), the bend in strike is absorbed by various intersecting and interfering folds and thrust faults in the Kingston– Chattooga composite thrust sheet (Figs 1 & 2). The distinct structural intersection in

DUCTILE DUPLEXES AS POTENTIAL NATURAL GAS PLAYS

59

Fig. 2. Geological map of the small-scale recess in Georgia, compiled from field data of authors, as well as Butts & Gildersleeve (1948), Cressler (1963, 1964a, b, 1970, 1974), Georgia Geological Survey (1976), Thomas & Cramer (1979), Coleman (1988), Osborne et al. (1988) and Thomas & Bayona (2005). Plunge directions of fold hinges are denoted by closed arrows. The Kingston– Chattooga anticlinorium is the structurally high outcrop area dominantly of Units 1 and 2 between Lookout Mountain syncline and Taylor Ridge monocline. The Floyd synclinorium (including Little Sand Mountain, Rock Mountain and Judy Mountain synclines, as well as other unnamed folds) encompasses the entire outcrop area of Unit 4 SE of the Kingston and Chattooga faults.

these intermediate structures, at an angle of c. 508 between two distinct elements of regional strike, characterizes the small-scale recess in Georgia. Structures striking 020 in the southern arm of the Tennessee salient and striking 070 in the eastern arm of the Alabama recess plunge from opposite directions into the depression of the Floyd synclinorium in the trailing part of the Kingston – Chattooga composite thrust sheet (Figs 1–3). The Rome thrust sheet, consisting of Cambrian shale-dominated facies of the Conasauga Formation, bounds the southern and eastern sides of the small-scale recess (Figs 1 & 2). Trailing the eastern side of the small-scale recess, the Rome fault has a highly sinuous trace, indicating a folded, subhorizontal fault surface (Fig. 2). Along the southern side, the Rome fault trace has an average trend of c. 090 but is highly sinuous in detail (Cressler

1970), indicating a subhorizontal envelope of folds of the fault surface that cuts obliquely across several thrust ramps and folds in the footwall. In addition to the irregular map trace, the shallow dip of the Rome thrust sheet is evident from the lack of seismic imaging of the near-surface fault (Thomas & Bayona 2005). The Rome fault truncates footwall folds that are coaxial with the folds of the fault surface; however, the fault-truncated footwall beds are folded more tightly than is the fault surface. The map relationships show that older footwall folds were truncated by an out-of-sequence Rome fault, and that the footwall folds were subsequently tightened, folding the Rome thrust sheet along with the footwall beds (Thomas & Bayona 2005). Further to the west in Alabama, the trace of the Rome fault curves to parallel the large-scale Appalachian structures (Fig. 1).

60 B. S. COOK & W. A. THOMAS Fig. 3. Geological cross sections illustrating the structural geometry of the small-scale recess in Georgia. The locations of the cross sections are shown on the inset map. Blue lines above present topographic surface show inferred pre-erosion extent of contacts between lithotectonic units.

61

Fig. 3. (Continued).


62


Structure and stratigraphy of the smallscale recess in northwestern Georgia An interference pattern in the structural intersection in the small-scale recess in Georgia, between east-northeastward and north-northeastward-striking folds and faults enables the tracing of both strike directions through parts of the intersection. Fold trains of both structural trends plunge into the depression of the Floyd synclinorium in the trailing part of the Kingston –Chattooga composite thrust sheet (Figs 1– 3). Unique structural expressions distinguish three structural domains: (1) the Kingston– Chattooga anticlinorium, which includes the frontal structures of the thrust sheet; (2) the Little Sand Mountain– Horn Mountain fold train, which trends c. 020 in the northern part of the Floyd synclinorium; and (3) the Simms Mountain –Horseleg Mountain fold train, which trends c. 070 in the southern part of the Floyd synclinorium (Figs 2 & 3). The Palaeozoic strata are divided into four lithotectonic units (Thomas 2001, 2007; Thomas & Bayona 2005) on the basis of general stratigraphic characteristics and mechanical behaviour during deformation (Fig. 2): Unit 1, the regional dominant weak layer, containing the regional dećollement, encompasses Lower to lower Upper Cambrian finegrained clastic rocks and minor thin-bedded limestones (Rome and Conasauga Formations); Unit 2, the regional dominant competent layer, which controls ramp geometry, is an Upper Cambrian – Lower Ordovician massive carbonate unit (Knox Group); Unit 3, a relatively thin, laterally variable, heterogeneous Middle Ordovician to Lower Mississippian succession of limestone, shale, sandstone, and chert; and Unit 4, an Upper Mississippian– Pennsylvanian synorogenic clastic wedge dominated by shale in the lower part and generally coarsening upward into sandstone and shale. The detachment of the Kingston–Chattooga composite thrust sheet is persistently in shale-dominated facies of the Middle to lower Upper Cambrian Conasauga Formation (Unit 1). In northwestern Georgia, Units 3 and 4 primarily are deformed passively over the underlying regional competent layer (Unit 2). Topography in northwestern Georgia is largely controlled by stratigraphy: most ridges are on Unit 3, and topographic flats are predominantly on shale-dominated strata in Units 1 and 4. Interestingly, the idea of dividing the regional stratigraphy into layers on the basis of relative rigidity and the inferences of how they affect structures in the southern Appalachians were first discussed by Hayes in 1891.

the Lookout Mountain syncline (Figs 2 & 3). Only a very gentle concave-cratonward curvature of the Kingston fault corresponds roughly to the more angular recess between the fold trains within the Floyd synclinorium further to the SE. The Chattooga fault and a leading imbricate parallel the trailing limb of the Kingston thrust sheet and end northeastward along strike, indicating that the Chattooga fault is a splay in a composite thrust sheet from the detachment of the Kingston fault (Fig. 2). The leading part of the Kingston– Chattooga composite thrust sheet forms the structurally high Kingston –Chattooga anticlinorium exposed in Units 1–3 (Figs 2 & 3). The anticlinorium is deformed by internal folds and faults. The trailing limb of the anticlinorium (the Taylor Ridge monocline) dips southeastward beneath the relatively deep Floyd synclinorium, which plunges into a regional depression within the recess between the oppositely plunging fold trains (Figs 2 & 3). The Taylor Ridge monocline is expressed at the surface primarily in Unit 3, striking c. 025.

Little Sand Mountain – Horn Mountain fold train The flat-bottomed Little Sand Mountain syncline, which is expressed at the surface in a sandstone in Unit 4, parallels the southeastern (downdip) side of the Taylor Ridge monocline, trending c. 020 (Fig. 2). The northwestward-verging Clinchport fault ramps through the trailing (SE) limb of the Little Sand Mountain syncline, and northeastward along strike, obliquely truncates the 000-trending Dick Ridge anticline (Fig. 2). Johns Mountain anticline is a cylindrical ramp anticline in the hanging wall of the Clinchport fault exposed in Units 2 and 3, trending c. 020. The Johns Mountain anticline ends in a southwestward-plunging, apparently conical fold associated with the southwestern end of the Clinchport fault (Fig. 2). Northeastward along strike, the Johns Mountain anticline merges with the up-plunge part of the 000-trending Horn Mountain anticline, which splits southward into a pair of southward-plunging anticlines (Fig. 2). Turkey Mountain anticline in the hinterland of the southwestern end of Johns Mountain anticline is a doubly plunging anticline exposed in Unit 3, trending 015. All of the anticlines rise steeply above the flat-bottomed synclines and have amplitudes of c. 650–1500 m. Spacing between the anticlines is c. 4– 7 km.

Kingston – Chattooga anticlinorium

Simms Mountain – Horseleg Mountain fold train

The northwestward-verging Kingston fault and a leading imbricate bound the southeastern limb of

On the southern side of the small-scale recess, eastnortheastward-plunging flat-bottomed synclines and


narrow, steep-sided anticlines diverge from the north-northeastward-striking Chattooga fault and Taylor Ridge monocline, and plunge northeastward into the depression of the Floyd synclinorium (Figs 2 & 3). Simms Mountain anticline plunges 075 into the deepest part of the Floyd synclinorium and, southwestward up-plunge, shows distinct fold interference with the Taylor Ridge monocline (Fig. 2). The flat-bottomed Rock Mountain syncline is expressed at the surface in sandstones of Unit 4 and trends 067. Lavender Mountain anticline is a cylindrical fold, forming a ridge of Unit 3, trending 064, and ending in a northeastward-plunging conical fold (Fig. 2). The southwestern up-plunge end of the Lavender Mountain anticline shows fold interference with the Turnip Mountain anticline, which is a ramp anticline exposed in Units 2 and 3, trending c. 020, roughly parallel with the Taylor Ridge monocline in the footwall (Fig. 2). Judy Mountain syncline, which is expressed in a sandstone within Unit 4, trends 067. Horseleg Mountain anticline is exposed in Unit 3 and trends 059 (Fig. 2). The anticlines rise steeply above the flat-bottomed synclines and have amplitudes of c. 650 –1000 m; spacing between the anticlines is c. 4 –7 km. Strawberry Mountain anticline, which is c. 17 km NW of Simms Mountain anticline and on the opposite end of the Little Sand Mountain syncline, trends c. 059, parallel with other anticlines in the Simms Mountain –Horseleg Mountain fold train. Strawberry Mountain anticline ends in both directions along strike by interference with the Taylor Ridge monocline to the SW and with Dick Ridge anticline and the Clinchport fault (part of the Simms Mountain –Horseleg Mountain fold train) to the NE (Fig. 2). Although the Strawberry Mountain anticline has the orientation of the Simms Mountain– Horseleg Mountain fold train, it is isolated within the Little Sand Mountain –Horn Mountain fold train, clearly showing interference between the two fold sets.

Cross sections and subsurface structure For this study, field measurements of structural orientation data and stratigraphic thickness have been obtained and compiled with structural data from other studies in the region (Butts & Gildersleeve 1948; Cressler 1963, 1964a, b, 1970, 1974; Georgia Geological Survey 1976; Thomas & Cramer 1979; Coleman 1988; Osborne et al. 1988; Thomas & Bayona 2005) to constrain construction of palinspastically restorable cross sections (Fig. 3). The subsurface geology is interpreted from seismic reflection profiles and projection of surface data. The depths to basement and thickness of a basal weak layer are measured from seismic reflection

63

profiles, and structures of the overlying units are constructed by extending surface measurements (i.e. stratigraphic thickness and strike/dip, etc.) into the subsurface. The seismic profiles show two distinct packages of clear layered reflectors in most places. The lower package of layered reflectors corresponds to Unit 1, and the base of the package is near the base of the sedimentary cover above Precambrian crystalline basement (Fig. 4). The top of the lower package of layered reflectors marks the top of Unit 1. The upper package of layered reflectors evidently corresponds to Unit 3, and also defines the top of Unit 2. The elevation of Unit 2 (regional competent layer) illustrates structural highs and lows across the Kingston –Chattooga composite thrust sheet. On the NW, the Kingston–Chattooga anticlinorium is a broad structural high bounded on the SE by the Taylor Ridge monocline, which dips into the deeper Floyd synclinorium. Unit 2 is structurally lower within the Floyd synclinorium, which is partitioned on the SW by east-northeastward-plunging anticlines of the Simms Mountain–Horseleg Mountain fold train and on the NE by south-southwestwardplunging anticlines of the Little Sand Mountain– Horn Mountain fold train, including the Clinchport thrust ramp (Johns Mountain anticline). Seismic reflection profiles show that the top of Precambrian crystalline basement dips very gently southeastward and is broken by small steep normal faults (e.g. Thomas & Bayona 2005). In cross section, the difference in elevation between the base of Unit 2 in the Kingston –Chattooga thrust sheet and the top of basement constitutes a large area to be filled (Figs 3 & 4), requiring an interpretation of subsurface structure. Previous interpretations have consistently included imbricate thrust sheets of Units 1 and 2 in the core of the Kingston–Chattooga anticlinorium, as well as blind thrusts in the cores of the anticlines of the Little Sand Mountain–Horn Mountain and Simms Mountain–Horseleg Mountain fold trains (e.g. Thomas & Bayona 2005). The dual fault traces of the Kingston fault and leading imbricate and the Chattooga fault and leading imbricate have been interpreted to be the surface expression of long imbricate thrust sheets in the core of the anticlinorium. Although this structural configuration satisfies the geometric form of the structures, other observations suggest that this interpretation may not be appropriate. The Chattooga fault and leading imbricate both end along strike, suggesting a relatively small magnitude of displacement. Furthermore, the Kingston fault and leading imbricate apparently terminate southwestward along strike and extend into an unfaulted detachment anticline (Thomas & Bayona 2005). Seismic reflection profiles clearly image the southeastern limb of the

64


Fig. 4. Seismic reflection profile of the ductile duplex, interpretation shown in lower panel. Location of the profile is near the northwestern end of cross section A –A0 in Figure 3, and the profile is oriented approximately parallel to the section.

Lookout Mountain syncline in the footwall of the Kingston fault; however, the profiles are ambiguous SE of the Kingston fault, where no coherent reflectors are shown above the basal package of layered reflectors above the basement (Fig. 4). The seismic reflection profiles lack resolution of any possible imbricate thrust sheets of Unit 2 beneath the surface-exposed thrust sheet (Fig. 4). Information applicable to the resolution of structural style in Georgia may be obtained by analogy from structures along strike to the SW in the Appalachian thrust belt in Alabama. In Alabama, deep drilling in the Gadsden mushwad (Fig. 1) has documented a minimum thickness of 2835 m of intensely deformed and tectonically thickened dark-coloured shale and thin-bedded limestone of the Middle to lower Upper Cambrian Conasauga Formation (Unit 1) (Thomas 2001). Seismic reflection profiles image dipping reflectors of the regional competent layer (Unit 2) both NW and SE of the Gadsden mushwad; however, the profiles show a distinct lack of coherent reflectors within the mushwad (Thomas 2001, fig. 7). The internal structure of the mushwad is inferred to include thrust faults that partition the ductilely deformed mass into internally deformed horses. Observations of outcrops and shallow core holes document disharmonic, tight, smallscale folds (amplitudes and wavelengths on the scale of a few metres) broken by faults of uncertain displacements. The mushwad structure is interpreted to be a ductile duplex beneath a roof thrust sheet of the regional competent layer (Unit 2) and a structurally attached uppermost part of Unit 1. The roof of the Gadsden mushwad has been eroded leaving the core of the duplex exposed; however, the structure of the roof can be inferred from bounding structures across strike (Thomas 2001). Further to the SW in Alabama, the crest of the Birmingham anticlinorium (Fig. 1) includes multiple thrust faults and folds, as well as backthrusts, exposed in

Unit 2 (Thomas 2001; Thomas & Bayona 2005). These structures form the roof of a separate subsurface mushwad, which is also shown in seismic profiles as a zone lacking coherent reflectors. Interestingly, prior to drilling of the first well into the Gadsden mushwad in 1985, the common interpretation was that the structurally high rocks at the top of the exposed Unit 1 reflect a subsurface stack of imbricate thrust sheets of Unit 2 and younger rocks (Thomas 1985, 2001, fig. 9). By analogy with ductile duplexes that have been documented along strike in the Appalachians in Alabama (Thomas 2001), the subsurface structure beneath the Kingston– Chattooga anticlinorium, as well as beneath the trailing part of the composite thrust sheet, is interpreted here as a ductile duplex. In this new interpretation, the mapped Kingston fault and leading imbricate, as well as the Chattooga fault and leading imbricate, are interpreted to be relatively low-magnitude thrust faults limited to the roof of the ductile duplex (Fig. 3). An interval of layered reflectors beneath Unit 2 shows that some strata in the uppermost part of Unit 1 are attached to the competent layer in the thrust sheet, and that the detachment of the Kingston –Chattooga composite thrust sheet is within Unit 1. The roof thrust of the ductile duplex places Unit 1 strata in the Kingston– Chattooga composite thrust sheet over ductilely deformed, tectonically thickened Unit 1 in the ductile duplex. The ductile duplex fills the space between the base of the thrust sheet and the top of the autochthonous lower part of Unit 1 overlying Precambrian basement beneath the dećollement (Figs 3 & 4). Seismic reflection profiles show that the leading edge of the ductile duplex forms a tectonic wedge under Unit 2 in the NW-dipping limb of the Kingston –Chattooga anticlinorium (common limb with the Lookout Mountain syncline) (Figs 3 & 4); a similar wedge is documented for the leading edge


of the Gadsden mushwad in Alabama (Thomas 2001, figs 5 & 7). The folds of the Little Sand Mountain –Horn Mountain and Simms Mountain– Horseleg Mountain fold trains are interpreted to be exaggerated detachment folds in the roof of the duplex, with the exception of the Clinchport faultrelated fold (Johns Mountain anticline) and Horseleg Mountain anticline, which are interpreted to be translated detachment folds in the roof of the duplex (Fig. 3).

Volume balance in the ductile duplex In the cross sections (Fig. 3), a large volume of ductilely deformed Unit 1 (Cambrian Conasauga shale) is shown to fill the space beneath the roof thrust at the base of the Kingston–Chattooga thrust sheet. A simplistic iteration of a palinspastically restored cross section, which employs only a linelength balancing of the competent layer (Unit 2), outlines the implications for an area-balanced reconstruction of the weak layer (Unit 1) (Panel 1 of Fig. 5). Such a reconstruction is applicable to the evolution of a detachment fold, in which the regional weak layer is tectonically thickened to fill the cores of detachment anticlines as the overlying competent layer is translated. In this palinspastic reconstruction (Panel 1 of Fig. 5), however, the restored area of Unit 1 is only about 50% of the deformed area of Unit 1 in the ductile duplex (Panel 2 of Fig. 5), clearly requiring a different explanation for the large excess in the area of Unit 1. Two end-member solutions may be suggested for the excess volume of Unit 1 in the deformedstate cross sections. First, deformation/flow of the weak-layer shales from out of the cross section planes could supply local excess volume. Secondly, a complex history of basement-fault movement may have resulted in the sedimentary accumulation of locally thick weak-layer rocks as a source for a ductile duplex. Tectonic thickening of Unit 1 as a result of outof-plane flow requires convergence of material into the tectonically thickened ductile duplex. The intersection of the two structural trends (defined by the Little Sand Mountain –Horn Mountain and Simms Mountain –Horseleg Mountain fold trains) suggests possible convergence from the depression of the Floyd synclinorium into the Kingston–Chattooga anticlinorium. The tectonic thickening of c. 100% in the ductile duplex would require withdrawal of ductile rocks from an area as much as twice the size of the mapped area of the ductile duplex. Such a withdrawal would likely generate structural depressions (e.g. structures in the competent layer plunging away from the centre of the recess), which are not recognized in the present outcrop

65

geology. The documented plunge into the depression of the Floyd synclinorium, in contrast, suggests divergent (rather than convergent) flow. Given these observations, out-of-plane flow does not seem likely to account for more than a small fraction of the excess volume of Unit 1 in the ductile duplex.

Analogy with structures in Alabama A complex history of basement-fault movement has been demonstrated to be integral to the formation of the ductile duplexes in Alabama, where the boundary faults of the Birmingham basement graben are clearly imaged in seismic reflection profiles (Thomas 2007). Large-scale frontal ramps rise northwestward over down-to-southeast basement faults, and thick disharmonic ductile duplexes (mushwads) underlie anticlinoria in which the competent-layer roof rocks are non-systematically faulted (Thomas 2001). Palinspastic restorations of thrust belt structures provide a framework to interpret stratigraphic variations in the context of episodic reactivation and inversion of the basement faults. In palinspastic location, the Middle to lower Upper Cambrian Conasauga Formation includes a shale-dominated facies greater than 2000 m thick in the basement graben, and a much thinner carbonate facies that is less than 800 m thick outside the graben (Thomas 2007). The differences in facies and thickness indicate synsedimentary fault movement, and the sedimentary variations document the time and magnitude of fault movement. Upper Cambrian massive carbonate deposits (Unit 2) overstep the graben boundary faults, indicating cessation of fault movement during deposition of Unit 2 carbonate rocks (Thomas 2007). The upper part of the Cambrian–Ordovician Knox Group (Unit 2), however, is unconformably absent over the palinspastically restored Birmingham graben. The unconformity is marked by a karstic palaeotopography with tens of metres relief, as well as sporadically distributed chert-clast conglomerate at the base of the Middle Ordovician cover stratigraphy (Thomas 2007). Middle Ordovician limestone units onlap the erosionally truncated Unit 2 and thin over the graben. These relationships indicate tectonic inversion of the Birmingham graben in the Middle Ordovician during Taconic tectonic loading (Bayona & Thomas 2003; Thomas & Bayona 2005). The amount of truncation of upper Unit 2 strata, palaeotopography and thinning by onlap combine to indicate as much as 700 m of reverse slip on the basement faults during inversion of the graben (Thomas 2007). Stratigraphic and sedimentological data indicate some minor episodic movement of the Birmingham graben faults during Silurian– Mississippian time, followed by .900 m of normal slip during

66 B. S. COOK & W. A. THOMAS Fig. 5. Simplified palinspastic restoration (Panel 1) based on line-length balance of the competent layer (Unit 2). Note that restored area of Unit 1 in the ductile duplex is c. 50% of the area of the ductile duplex in the deformed-state cross section (Panel 2), showing that this interpretation of palinspastic restoration cannot be area balanced.

DUCTILE DUPLEXES AS POTENTIAL NATURAL GAS PLAYS 67

Fig. 6. Sequential cross sections illustrating a basement graben that is interpreted to be the source of the surplus volume of Unit 1 shales in the small-scale recess in Georgia. Panel 1 illustrates the thick Unit 1 succession in a synsedimentary graben, overlain by a uniform thickness of Unit 2 shallow-marine carbonates. Panel 2 illustrates graben inversion, leading to elevation of thickened Unit 1 and erosional truncation of the top of Unit 2 over the former graben. Panel 3 illustrates the deformed-state cross section (cross section A– A0 from Fig. 3) in which the thickened Unit 1 is at the centre of a ductile duplex. The dashed red lines in panels 1 and 2 represent future trajectories of thrust faults, including the floor and roof thrusts bounding the ductile duplex. The area of the ductile duplex is equal in all three diagrams.

68


deposition of Upper Mississippian– Lower Pennsylvanian clastic strata. The ultimate composite vertical separation on the basement fault is c. 2255 m (Thomas 2007). Propagation of Palaeozoic, thin-skinned Appalachian thrust faults at a regional dećollement in Unit 1 encountered the thick, mud-dominated facies (Conasauga shale) in the basement graben, as well as a basement-fault buttress at the northwestern boundary of the graben. Ductile deformation generated thick mushwads beneath large-scale frontal thrust ramps of the regional competent layer (Unit 2) (Thomas 2001). The maximum structural relief on the roof of the mushwads is as much as 4500 m, indicating c. 3:1 tectonic thickening of the depositionally thickened Conasauga Formation in the mushwad (Thomas 2007).

Interpretation for structures in Georgia No large-magnitude basement faults are seismically imaged in the region of the Kingston –Chattooga composite thrust sheet in Georgia; however, minor disruptions in the basal reflector package show the locations of faults that presently have small displacement of the top of the basement. By analogy with the history of mushwads (ductile duplexes) in Alabama, the present fault offset of the top of basement may reflect a composite of successive displacements, some of which are inverted. Assuming a history similar to that of the Alabama mushwads, an area balance of the ductile duplex beneath the Kingston –Chattooga composite thrust sheet (Kingston–Chattooga anticlinorium and fold trains) requires an original depositional thickness of the Conasauga Formation c. 500 m greater than that in the foreland to the NW (Fig. 6). Accommodation of the greater thickness indicates a basement graben c. 500 m deeper than present basement elevation. Later inversion of the graben would have reversed part of the original slip. By analogy with stratigraphy in Alabama, inversion during Taconic (Middle Ordovician) loading may be recorded in erosion of the upper part of the Knox Group, Unit 2 (Fig. 6) (Bayona & Thomas 2003). Previously unexplained observations in Georgia include a local lack of the upper components that regionally comprise the Knox Group; specifically the Lower Ordovician Chepultepec Dolomite is unconformably absent in northwestern Georgia (Coleman 1988). The same stratigraphic unit (Chepultepec Dolomite) is unconformably absent in the area of the Gadsden mushwad and along the Birmingham anticlinorium in Alabama, where the top of the Knox Group is marked by chert-clast conglomerates. Similar chert conglomerates are found sporadically at the top of Unit 2 in northwestern Georgia. These observations suggest that inversion occurred along basement

faults in Georgia. Although the Birmingham graben shows subsequent reactivation in Alabama during late Palaeozoic (Mississippian–Pennsylvanian) thrusting and tectonic loading (Thomas 2007), this later episode of basement-fault reactivation is not documented by stratigraphy in Georgia. The maximum structural relief on the roof of the mushwad in Georgia is c. 2500 m, indicating c. 2:1 tectonic thickening of the depositionally thickened Conasauga Formation. Sequential diagrams (Fig. 6) illustrate the interpreted origins of stratigraphic variations necessary to area balance the ductile duplex beneath the Kingston– Chattooga thrust sheet. The deformed-state cross section (Panel 3 of Fig. 6) shows the present location and geometry of the interpreted ductile duplex. The cross section in Panel 1 of Figure 6 illustrates the depositional framework of a thick Unit 1 succession in a synsedimentary graben; after the end of fault movement, Unit 2 was deposited across the graben with uniform thickness. The top of Unit 2 is drawn nearly horizontal to reflect the interpreted shallow-marine shelf deposition of the carbonate rocks. Area-balance restoration of the deformed state of the ductile duplex requires Unit 1 to be c. 1700 m thick in the graben, in contrast to a regional average of 1200 m. The depositional thickening requires c. 500 m of vertical separation along the normal fault boundary of the graben (Panel 1 of Fig. 6). Panel 2 of Figure 6 shows inversion of the graben to elevate the thick graben fill (Unit 1) and cover (Unit 2), leading to erosion of the upper part of Unit 2. In this interpretation, the thickness of Unit 2 in the deformed-state cross section (c. 600 m) constrains the thickness of the eroded upper part of Unit 2, which is of the order of 300 m. The amount of truncation, plus palaeotopography and onlap, indicate c. 500 m of reverse slip during inversion, and that magnitude of inversion places the top of basement at the present structural level (Fig. 6). As a final note, the volume balance of the ductile duplex is critical for palinspastic reconstruction of the recess, and the understanding of the kinematic and mechanical history of the local structures. The intersection and fold interference exemplify a long-standing problem in volume balancing of palinspastic reconstructions of sinuous thrust belts. Cross sections generally are constructed perpendicular to structural strike, parallel to the assumed slip direction. An array of cross sections around a structural bend may be restored and balanced individually; however, restorations perpendicular to strike across intersecting thrust faults yield an imbalance in the along-strike lengths of frontal ramps. Similarly, the restoration leads to an imbalance in the surface area of a stratigraphic horizon. The inverted basement


graben provides a solution to the volume balance problems encountered in palinspastic restoration of the cross sections around the small-scale recess in Georgia.

Conclusions Around the small-scale recess in northwestern Georgia, tectonically thickened weak stratigraphic layers of the Cambrian Conasauga Formation accommodated ductile deformation associated with the folding and brittle faulting of the overlying Cambrian –Ordovician regional competent layer. Ductile deformation of the underlying structurally thickened weak layer allows the shales to fill the cores of anticlines in the competent layer. The ductile duplex in the core of the Kingston – Chattooga anticlinorium represents an excess volume of Unit 1 shales that elevates the structural level of the Kingston–Chattooga composite thrust sheet. The trailing limb of the anticlinorium is marked by the Taylor Ridge monocline, which dips into the structurally lower Floyd synclinorium. In the Floyd synclinorium, two fold trains of broad synclines and narrow anticlines plunge into the depression of the synclinorium with two distinct structural trends. Low-amplitude folds, which are the plunging ends of the fold trains, characterize the centre of the abrupt bend in Appalachian structural trends in the recess in Georgia. The area of the mushwad in deformed-state cross sections is approximately twice the area of the corresponding Unit 1 in the restored cross sections, and cannot be explained solely by tectonic thickening parallel to the direction of apparent shortening of a conventional palinspastically restored cross section. This imbalance may result from some combination of two mechanisms: transport of Unit 1 shales into the plane of cross section, and activation/inversion of a basement graben. The out-of-plane transport of material implies an as yet unrecognized deficit in Unit 1 thickness elsewhere in the thrust belt to balance the surplus in the ductile duplex. A new interpretation proposes that a basement graben accommodated deposition of a locally thicker Unit 1 succession (c. 1700 m, in contrast to c. 1200 m to the NW of the graben) prior to thrust deformation, analogous to the Birmingham graben along strike to the SW in the Appalachian thrust belt in Alabama. Subsequent Middle Ordovician reactivation/inversion of the graben, related to Taconic loading, resulted in uplift and the erosion of the upper part of the overlying Unit 2 (Bayona & Thomas 2003). Finally, thrusting and accretion of the weak layer into the ductile duplex occurred during tectonic shortening in late Palaeozoic

69

times. The exposed competent-layer structures in Georgia are analogous to those over shaledominated ductile duplexes (mushwads; Thomas 2001), which are being developed for natural gas in the Appalachian thrust belt in Alabama; however, the total thickness of the Unit 1 shale-dominated ductile duplex in Georgia is somewhat less than in those in Alabama. Finally, the interpretation of a basement graben yields a solution to volume balance encountered during palinspastic restoration of the array of cross sections around the small-scale recess in Georgia. Field mapping by Cook (2007– 2009) was funded by support from the EDMAP component of the National Cooperative Geologic Mapping Program of the US Geological Survey, the American Association of Petroleum Geologists, the Geological Society of America and the Ferm Fund from the University of Kentucky Department of Earth and Environmental Sciences. Acknowledgement is made to the donors of the Petroleum Research Fund, administered by the American Chemical Society, for partial support of this research. The authors thank Tim Needham, Terry Engelder, and special editor Graham Goffey for their constructive reviews of the manuscript.

References ALABAMA STATE OIL AND GAS BOARD. 2007. An overview of the Conasauga Formation shale gas play in Alabama, November 2007. Bayona, G. & Thomas, W. A. 2003. Distinguishing fault reactivation from flexural deformation in the distal stratigraphy of the Peripheral Blountian Foreland Basin, southern Appalachians, USA. Basin Research, 15, 503 –526. Butts, C. & Gildersleeve, B. 1948. Geology and mineral resources of the Paleozoic area of NW Georgia. Georgia Geological Survey Bulletin, 54. Coleman, J. L. Jr 1988. Geology of the rising Fawn CSD. Alabama Geological Society Guidebook, 25th Annual Field Trip, 12– 40. Cressler, C. W. 1963. Geology and ground-water resources of Catoosa County, Georgia. Georgia Geological Survey Information Circular, 28. Cressler, C. W. 1964a. Geology and groundwater resources of the Paleozoic rock area, Chattooga County, Georgia. Georgia Geological Survey Information Circular, 27. Cressler, C. W. 1964b. Geology and ground-water resources of Walker County, Georgia. Georgia Geological Survey Information Circular, 29. Cressler, C. W. 1970. Geology and groundwater resources of Floyd and Polk Counties, Georgia. Georgia Geological Survey Information Circular, 47. Cressler, C. W. 1974. Geology and groundwater resources of Gordon, Whitfield, and Murray Counties, Georgia. Georgia Geological Survey Information Circular, 39. GEORGIA GEOLOGICAL SURVEY. 1976. Geologic map of Georgia. Georgia Geological Survey. Scale 1:500 000.

70


Hayes, C. W. 1891. The overthrust faults of the southern Appalachians. Geological Society of America Bulletin, 2, 141– 154. Osborne, W. E., Szabo, M. W., Neathery, T. L. & Copeland, C. W. Jr (compilers) 1988. Geologic map of Alabama, NE sheet. Geological Survey of Alabama Special Map 220, scale 1:250 000. Thomas, W. A. 1977. Evolution of Appalachian – Ouachita salients and recesses from reentrants and promontories in the continental margin. American Journal of Science, 277, 1233– 1278. Thomas, W. A. 1985. Northern Alabama sections. In: Woodward, N. B. (ed.) Valley and Ridge Thrust Belt: Balanced Structural Sections, Pennsylvania to Alabama (Appalachian Basin Industrial Associates). University of Tennessee Department of Geological Sciences Studies in Geology, 12, 54–61. Thomas, W. A. 2001. Mushwad: Ductile duplex in the Appalachian thrust belt in Alabama. American

Association of Petroleum Geologists Bulletin, 85, 1847– 1869. Thomas, W. A. 2007. Role of the Birmingham basement fault in thin-skinned thrusting of the Birmingham anticlinorium, Appalachian thrust belt in Alabama. American Journal of Science, 307, 46–62. Thomas, W. A. & Bayona, G. 2005. The Appalachian thrust belt in Alabama and Georgia: Thrust-belt structure, basement structure, and palinspastic reconstruction. Geological Survey of Alabama Monograph 16. Thomas, W. A. & Cramer, H. R. 1979. The Mississippian and Pennsylvanian (Carboniferous) Systems in the United States – Georgia. United States Geological Survey Professional Paper 1110-H, H1– H37. Williams, P. 2007. Conasauga saga. Oil and Gas Investor. September 2007, 77–80. Willis, B. 1893. The mechanics of Appalachian structure. United States Geological Survey Annual Report, 13, Part 2, 211 –281.

Controls on lateral structural variability along the Keping Shan Thrust Belt, SW Tien Shan Foreland, China SEBASTIAN A. TURNER1,2*, JOHN W. COSGROVE1 & JIAN G. LIU1 1

Department of Earth Science & Engineering, Imperial College London, South Kensington Campus, London SW7 2AZ, UK 2

Present address: BP Exploration & Production, Chertsey Road, Sunbury-on-Thames, Middlesex TW16 7BP, UK *Corresponding author (e-mail: [email protected]) Abstract: Lateral structural variability and partitioning of fold –thrust belts often reflects lateral variations in the stratigraphy of the deforming foreland and interaction with inherited structures. The Keping Shan Thrust Belt, NW China, was initiated during the late Cenozoic and is a spectacular example of contractional deformation in a foreland setting. The belt is characterized by a series of imbricate thrusts which form a broadly arcuate salient and deform the thick (3– 6 km) Phanerozoic sedimentary succession of the NW Tarim Basin (SW Tien Shan foreland). Abrupt lateral changes in the thickness of the sedimentary succession are associated with a series of major preexisting basement faults which cross-cut the belt and which were probably initiated during early Permian times. These lateral variations in the basin template have impacted strongly on the structural architecture of the superimposed thrust belt. Variations in the thickness of the sediment pile affect the spatial distribution of thrusts, which increase in abundance where the sediment is thinnest. The inherited cross-cutting basement faults and the associated abrupt changes in sediment thickness combine to generate partitioning of the thrust belt.

Lateral structural variability and partitioning within foreland fold–thrust belts is commonly associated with lateral variations in the deforming basin. An increasing number of studies into the evolution of foreland fold–thrust belts have examined the interplay between pre-existing structures, variations in the thickness and rheology of the sediment pile, and lateral differences in the type and thickness of the detachment horizon (e.g. Liu et al. 1992; Marshak et al. 1992; Lawton et al. 1994; Macedo & Marshak 1999; Sepehr & Cosgrove 2004, 2007; Butler et al. 2006). Such variations have important consequences on the structural architecture of the fold–thrust belt, causing lateral variations in horizontal shortening, deformation style and the spatial organization of structures. Furthermore, such lateral variations are often accommodated by the formation of lateral ramps and strike-slip (transfer) faults which are oblique or perpendicular to the general structural trend of the fold–thrust belt. The interplay between the pre-existing basin template and a later fold–thrust belt therefore has important implications for hydrocarbon exploration in compressional belts, causing lateral compartmentalization of reservoirs and structural complexity. The Keping Shan Thrust Belt is one of several fold–thrust belt salients which have evolved in the foreland of the SW Tien Shan, NW China, during

late Cenozoic times. The Keping Shan is characterized by a spectacular series of imbricate thrusts, which deform a predominantly Palaeozoic sedimentary pile (Fig. 1). The thrusts have trends varying from east –west to NE– SW across the belt and broadly form an arcuate salient (Figs 1 & 2). The internal structural architecture of the Keping Shan is complex. A series of major strike-slip faults, which are oblique or perpendicular to the general trend of the thrusts, partition the belt into a series of structural domains, characterized by lateral variations in horizontal shortening and the spatial organization of structures. The aim of this paper is to examine the structural architecture of the Keping Shan Thrust Belt and to identify the underlying causes of lateral (alongstrike) structural variability. By examining the preexisting structure and stratigraphy within the NW Tarim Basin, relationships between the architecture of the late Cenozoic thrust belt and the inherited basin template can be examined and assessed in order to provide a new model for the long-term structural evolution of the region. The data used in this study were obtained through satellite image interpretation and field-based mapping. A regionalscale geological interpretation map (Fig. 1) was developed using Landsat ETM and high-resolution (5 5 m) SPOT satellite images. These maps were enhanced with structural measurements and


72

S. TURNER ET AL.

observations made at outcrop scale during two field seasons to the Keping Shan. These methods combined provide the basis for analysing the large and small-scale structure and investigating the stratigraphic framework of the region. The latter yields information concerning inherited structures and lateral variations in the rheology and thickness of the sediment pile, which could prove crucial in unravelling the causes of structural variability in the Keping Shan.

Geological setting The Keping Shan Thrust Belt actively deforms the 3–6 km thick Neoproterozoic –Recent sedimentary succession of the NW Tarim Basin, an area that equates to the foreland of the SW Tien Shan (Fig. 1). The sedimentary succession records the complex and protracted history of the Tarim Basin, which began in the Neoproterozoic. At this time, the underlying Tarim Craton rifted from Australia during a widespread rifting event (Li & Powell 2001; Chen et al. 2004). Subsequently, the Tarim Craton accumulated a thick passive margin succession prior to collision with the developing Eurasian margin during Late Devonian to early Carboniferous times (Carroll et al. 1995, 2001). The collision resulted in the formation of the Tien Shan orogenic belt (Burrett 1974; Burtman 1975; Coleman 1989; Jun et al. 1998) and sedimentation within the NW Tarim Basin occurred in a foreland setting (Carroll et al. 1995). A short but important phase of extension occurred in the early Permian (c. 275 Ma, Zhang et al. 2008), which was associated with magmatic activity and the generation of major normal faults, which impacted on the thickness of the sedimentary succession across the region. As a result, the NW Tarim Basin remained an intrabasinal high throughout the Mesozoic and did not receive any sediment during this period (Li et al. 1996). Thick Mesozoic successions are recorded in two isolated depocentres in the western and eastern Tarim Basin, recording a series of smaller collisions at the southern margins of the growing Eurasian continent during the formation of the Tibetan collage (Watson et al. 1987; Hendrix et al. 1992; Sobel 1999). The collision of India and Eurasia in the early Cenozoic marked the onset of renewed contraction across Central Asia. Ancestral mountain belts including the Tien Shan were reactivated and rejuvenated, shedding large quantities of sediment into the Tarim Basin. Flexure at the margins of the basin resulted in accumulations of Palaeogene– Neogene sediments which locally exceed 10 000 m (Bally et al. 1986; Yang & Liu 2002). Reactivation of the Tien Shan did not begin until c. 20 Ma, and has accelerated since c. 10 Ma (Abdrakhmatov et al. 1996; Sun et al. 2004). During this time the

belt has accommodated around c. 200 (+ 50) km of crustal shortening (Avouac et al. 1993; Abdrakhmatov et al. 1996). The initiation of folding and thrusting within the SW Tien Shan foreland (NW Tarim Basin) probably began during or shortly after this time (?10– 5 Ma). Seismicity along the boundary zone between the Tarim Basin and the Tien Shan indicates that folding and thrusting is presently active (USGS 2009), while geodetic (GPS) measurements suggest shortening rates of 8 (+3) mm a21 across the Keping Shan and adjacent Kashgar Fold Belt. This corresponds to c. 40% of the total shortening rate across the whole Tien Shan belt (Reigber et al. 2001) and serves to demonstrate the importance of foreland fold–thrust belts in accommodating crustal shortening across orogenic belts.

Structure of the Keping Shan Thrust Belt The morphology of the Keping Shan is characterized by major fault zones which were generated (or reactivated) during contraction in the late Cenozoic (Yin et al. 1998; Allen et al. 1999). Following the format of Sepehr & Cosgrove (2007), these fault zones are categorized for the purpose of this study according to structural trends. Major fault zones within the Keping Shan can be broadly divided into two categories: (1) belt-parallel fault zones; (2) belt-oblique fault zones (Fig. 2).

Belt-parallel fault zones Belt-parallel fault zones comprise NE –SW to east – west trending faults which are parallel to the trend of the Tien Shan orogenic belt (Figs 1 –3). Without exception, all the fault zones in this category are thrusts, which predominantly verge to the south towards the interior of the Tarim Basin. Allen et al. (1999) proposed that the thrusts detach onto a thin upper Cambrian salt horizon. Palaeozoic and Cenozoic strata of the basin are exhumed in the hanging walls of the thrusts, forming topographically prominent ridges, which rise up to 1200 m high relative to the piggyback basins that have developed between thrusts. These basins are narrow (6–15 km in width), internally draining, and filled with Quaternary sediments. The general strike of the thrusts varies across the Keping Shan, giving the belt an overall arcuate salient geometry (Figs 1 & 2). Thrust trends vary from NE–SW in the east to east –west in the west. The eastern end of the Keping Shan tapers into a 50 km-wide recess which we have termed the Aksu Re-entrant. To the west, there is a gradual transition from the Keping Shan Thrust Belt into the Kashgar Fold Belt (Fig. 1). In contrast to the Keping Shan, the Kashgar Fold Belt is dominated

Fig. 1. Geological map of the Keping Shan Thrust Belt showing the master structural elements described in this paper. Belt-parallel faults are characterized by east–west to NE–SW trending thrust faults that define a broad, arcuate thrust belt. The belt is partitioned by a series of belt-oblique (strike-slip and oblique-slip) faults that predominantly trend NW–SE. The northern margin of the Keping Shan Thrust Belt is defined by the South Tien Shan Fault, which separates the metamorphic rocks of the Tien Shan from the sedimentary rocks of the Tarim Basin.

KEPING SHAN THRUST BELT, NW CHINA

73

Fig. 2. 3D perspective view of the Keping Shan Thrust Belt, generated by draping a Landsat ETMþ (bands 321, 30 m resolution) satellite image over a digital elevation model. Major thrusts are identified by the surface expression of their hanging walls, which form long, arcuate ridges that exhume a predominantly Palaeozoic stratigraphic succession. Laterally, thrusts interact with major belt-oblique fault zones (marked) that partition the thrust belt into a series of structural domains.

by simple detachment folds which only affect Cenozoic strata and there is little surface expression of thrusts (Scharer et al. 2004). South Tien Shan Fault. This fault acts as the major bounding fault which separates metamorphic rocks of the Tien Shan from the sedimentary cover succession of the Tarim Basin to the south (Fig. 1). Cenozoic activity on the fault and related exhumation of the Tien Shan mountains began around 24 Ma, at the Oligocene –Miocene boundary (Sobel & Dumitru 1997). The fault is a steep (40– 508) north-dipping thrust that trends ENE –WSW. Within the metamorphic mica-schists immediately north of the fault, there is an abundance of shear structures and minor folds, which have east –west trending fold axes and axial planes that dip 50– 608 to the north (Fig. 4a). These features are attributed to intense ductile deformation associated with the fault zone. The fault juxtaposes the metamorphic rocks against the Phanerozoic sedimentary cover succession of the Tarim Basin. It is therefore assumed that the South Tien Shan fault zone continues to substantial depth (c. 20– 30 km), acting to allow the Tarim Block to be underthrust beneath the Tien Shan.

Keping (Frontal) Fault. This fault separates the present alluvial Tarim Basin from the Keping Shan. It is the most southerly of the belt-parallel thrusts and has the most topographically prominent hanging wall (Figs 1 & 2). In addition, it is the most seismically active of all the belt-parallel faults, with focal mechanism solutions indicating relatively pure thrust displacement (USGS 2009). These seismogenic and geomorphic attributes suggest that the Keping Fault is the youngest beltparallel fault and that each respective thrust to the north is progressively older, and that the thrust belt has largely evolved as a simple forelanddirected (piggyback) series (Dahlstrom 1970; Butler 1982). Along much of its length, cliff-forming Cambrian– Ordovician limestones which dip gently to the north define the base of the thrust hanging wall. Only in a few localities are the remnants of fault-related folding preserved, but it is postulated that folds of similar form were once continuous along the mountain front (Fig. 3). The southern limbs of these folds are steeply dipping and often overturned, while the northern limbs dip between 20 and 408. Where present, the cores of fault-related folds are characterized by internal deformation, and minor thrusts and folds are abundant (Fig. 4c, e).

74 S. TURNER ET AL. Fig. 3. Balanced cross section across the Keping Shan (A0 –A00 , see Figure 1 for line of section). Major thrusts detach onto a middle Cambrian evaporite surface (sensu Allen et al. 1999), predominantly dipping to the north and verging to the south. Once balanced, the horizontal shortening across the section is 33%. This section assumes that deformation is completely thin-skinned but in the absence of subsurface data we cannot determine whether basement faults are involved.


75

Fig. 4. Field photographs illustrating aspects of structural deformation in the Keping Shan: (a) Minor folding of interbedded quartzites and schists within the shear zone of the South Tien Shan Fault; (b) South-verging thrust duplex in Palaeogene conglomerate immediately south of the South Tien Shan Fault; (c, d) Deformation within Ordovician limestones in the Mystery Canyon, in the hanging wall of the Keping Fault, showing a south-verging thrust and intense deformation above a minor thrust respectively; (e) North-verging backthrust in the southern limb of a thrust-related anticline above the Keping Fault; (f) Middle Devonian red sandstones juxtaposed against middle Ordovician limestones along the northern segment of the Piqiang Fault, near Piqiang.

Minor folds have axes which are subparallel to the local trend of the Keping Fault.

Belt-oblique (cross) fault zones Major fault zones which are oblique (by .458) or perpendicular to the general trend of the Keping Shan are termed ‘belt-oblique’ and comprise

oblique-slip and strike-slip (transfer) faults. Several of these faults have a prominent surface expression, while others are more subtle. In either case, belt-oblique faults have an important role in partitioning the thrust belt into a series of structural domains which are characterized by variations in the spatial distribution of belt-parallel thrusts. The initial formation of these faults pre-dates the late

76 S. TURNER ET AL. Fig. 5. Stratigraphic correlation panel across a series of stratigraphic sections taken from the hanging wall of the Keping Fault and restored to the base Cenozoic unconformity. The correlation panel was constructed by measuring the thickness of the sediment pile at nine sections along the hanging wall of the Keping Fault. Major faults across which the sediment thickness changes abruptly were interpreted from the surface expression of structural lineaments (cf. Fig. 1). There is a dramatic thinning of the Palaeozoic sediment pile across the central part of the Keping Shan, an area referred to as the Bachu Uplift. The depth to the middle Cambrian detachment layer varies from more than 6 km to just over 2 km between the Piqiang and Sanchakou Faults. Much of the thickness change occurs in the lower Permian succession, but further work is required to determine whether this is a syntectonic feature. An additional major fault is proposed for the west, between the Sulphur Canyon and Yingan sections, where sediment thickness increases substantially despite a lack of belt-oblique structures with surface expression in this area.


Cenozoic evolution of the Keping Shan Thrust Belt. A 280 km stratigraphic correlation panel was constructed by measuring the thickness of the sedimentary pile at nine sections along the hanging wall of the Keping Fault (Fig. 5). When hung from the base Cenozoic unconformity, the correlation panel demonstrates the impact of the inherited faults on the thickness of the sediment pile during an earlier phase of tectonism in the Tarim Basin. Although the event responsible for the inherited faults remains the subject of debate, it is likely that they formed during a brief phase of extension that affected the NW Tarim Basin during the early Permian. This period was characterized by substantial basaltic magmatism, resulting in the emplacement of NW– SE trending dykes and extrusive basalt flows within the lower Permian stratigraphic succession. The basalts yield ages of 274 + 2 Ma (Zhang et al. 2008). Analysis of borehole data from the interior of the Tarim Basin has revealed that the basalts cover a total area of around 250 000 km2 and are thought to have been caused by a short-lived mantle plume in the early Permian (Jia et al. 2004; Jiang et al. 2004; Zhang et al. 2008). Incidentally, the area in which the sediment pile is thinnest in the Keping Shan (Fig. 5), across a structure known as the Bachu Uplift, correlates to an area proposed to represent the source region of the early Permian basalts which flooded much of the Tarim Basin (Zhang et al. 2008). Furthermore, Chen et al. (2006) identify this region as the central part of a much larger area affected by substantial crustal doming from the late Cisuralian to the Guadalupian (c. 270 –260 Ma). An alternative possibility is that the inherited faults relate to a later transtensional event which affected the NW Tarim Basin during the Jurassic (Sobel 1999; pers. comm., J. Suppe). Tectonism during this period was characterized by NW–SE trending strike-slip faults which formed deep and narrow transtensional basins in the western Tarim Basin (Sobel 1999). A complete absence of Mesozoic sediments within the Keping Shan means that it remains speculative as to which event was responsible for the formation of the inherited faults, although we maintain that early Permian extension was the most likely cause. In either case, given that the Keping Shan remained an intrabasinal high throughout the Mesozoic (Li et al. 1996) it is likely that a substantial amount of erosion occurred across individual fault blocks and the Bachu Uplift prior to the early Cenozoic, accounting for the substantially reduced thickness of Palaeozoic strata over the central parts of the Keping Shan. Piqiang Fault. This fault is expressed for more than 70 km as a prominent structural lineament that has a dramatic effect on the structural architecture of the Keping Shan Thrust Belt (Figs 1, 2 & 6). The

77

fault trends approximately NNW –SSE (340 –3508) and is defined by a series of segments which either offset or completely decouple the east –west trending belt-parallel thrusts to either side, thereby acting to partition the thrust belt. The trend of the fault is subparallel to the SSE-oriented thrust transport direction. Examination of the individual segments of the Piqiang Fault reveals that the faulting mechanism changes along strike at c. 20 –25 km intervals, locally acting as either a strike-slip fault or as a lateral ramp. In plan view, the Piqiang Fault has an impact on the spatial organization of the belt-parallel thrusts to either side of it, expressed as a change in the number and spacing of thrusts from west to east (Fig. 6). To the west, there are three major thrusts (with surface expression), one of which terminates against the Piqiang Fault as a lateral ramp. To the east, there are five major thrusts which are more closely spaced, creating narrower piggyback basins. The greater abundance of thrusts to the east implies that the total horizontal shortening to the east of the Piqiang Fault is marginally greater than to the west. Examination of the stratigraphy across the fault zone indicates that there is a net loss from west to east (Fig. 7). To both the east and west of the fault, the Palaeozoic megasequences are separated from the Cenozoic megasequence by a major unconformity which spans the Mesozoic. The anomaly lies in the age of the youngest Palaeozoic strata on the eastern and western sides of the Piqiang Fault. To the east, the unconformity separates the Middle Devonian from the Palaeogene, while to the west, it separates the lower Permian from the Palaeogene (Fig. 7). The upper Carboniferous and lower Permian sediments that are absent from the eastern side of the fault are shallow-marine to fluvial carbonates and sandstones which were deposited in the late Carboniferous foreland basin that developed adjacent to the ancestral Tien Shan (Carroll et al. 1995). On both sides of the fault, Palaeogene sediments are interbedded fluvial sandstones and mudstones which maintain the same thickness across the fault, suggesting they were deposited onto a flat, peneplain surface in the early Cenozoic. In total, we estimate that c. 800 m of sediment is absent from the eastern side of the Piqiang Fault. Saergan Fault. Similar to the Piqiang Fault, the Saergan Fault is a prominent structure which crosscuts belt-parallel thrusts (Figs 1 & 2). It can be traced for c. 40 km across the belt, and acts as a major right-lateral strike-slip fault. Unlike the other faults described in this section, however, there appears to be little or no stratigraphic discontinuity across it. In addition, the orientation of the fault is notably different from other belt-oblique fault zones, which generally follow trends of NW–SE to north –south. This suggests that unlike many other

78

S. TURNER ET AL.

Fig. 6. Structure of the Piqiang Fault: (a) Geological map derived from the interpretation of (b), Landsat ETMþ (bands 321, 30 m resolution) satellite image. Along strike, the faulting mechanism appears to change, acting either as a strike-slip fault (southern and northern segments) or as a lateral ramp (central segment).

belt-oblique faults, the Saergan Fault formed during the late Cenozoic evolution of the Keping Shan and was not part of the early Permian fault population. The fault may act to accommodate lateral changes

in horizontal shortening which are not suitably attained through other reactivated belt-oblique structures, but this will require a further, detailed study of the structure.


79

Fig. 7. Stratigraphic correlation across the Piqiang Fault, showing the net loss of nearly 800 m of Middle Devonian, upper Carboniferous and lower Permian strata from west to east, and the unaffected Palaeogene –Neogene strata above the base Cenozoic unconformity.

Sanchakou Fault. This fault zone has little surface expression but correlates to an important change in sediment thickness and causes localized disruption to the Keping Fault. The surface expression of the fault crops out in the village of Sanchakou, to the south of the Keping Fault. Tracing the fault to the north, the fault interacts with the Keping Fault and causes it to branch into two faults to the west (Fig. 8). Based on the stratigraphic correlation across the fault (Fig. 5), the net loss of stratigraphy is 500 –550 m from east to west. Most of the stratigraphic loss occurs within the Middle Devonian, and only a few kilometres to the west a progressive thickening of upper Carboniferous strata is observed in what would have been the original downthrown block. The interaction of the Sanchakou and Keping Faults is characterized by a zone of structural complexity in which upper Neoproterozoic to lower Cambrian sediments are exposed (Fig. 8). These are the oldest sediments within the central part of the Keping Shan, and they have been exhumed from beneath the middle Cambrian detachment layer.

Yijianfang Fault. The surface expression of the Yijianfang Fault is prominent within the Tarim Basin to the south of the Keping Shan, representing one of the few structural features that penetrate the present-day basin surface. The interaction with the Keping Shan is less obvious, but tracing the fault zone to the north it is apparent that it interacts with the Keping Fault and causes a kink in the NE–SW trend of it (Fig. 1). In addition, the structural dip of beds within the hanging walls of belt-parallel thrusts are substantially reduced for several kilometres eastward, from the average 358 dip values recorded across much of the Keping Shan to c. 258. Across the Yijianfang Fault, the change in stratigraphic thickness is c. 500 m and is most prominently demonstrated by the substantial increase in the thickness of lower Permian sediments to the east (Fig. 5).

Discussion Lateral variations in the structural architecture and partitioning of the Keping Shan Thrust Belt

80

S. TURNER ET AL.

Fig. 8. Structure of the Sanchakou Fault: (a) SPOT-5 (bands 431, 5 m resolution) false colour satellite image; (b) Geological map based on the interpretation of (a). The Middle Devonian thickens abruptly across the Sanchakou Fault. On interacting with the Sanchakou Fault, the Keping Fault branches across a zone of structural complexity in which late Neoproterozoic sediments are exhumed from beneath the middle Cambrian detachment layer.


correlates with major lineaments and fault zones which are oblique or perpendicular to the general structural trend of the thrust belt. These fault zones pre-date the late Cenozoic thrusting and previously acted as basin-bounding faults which controlled the thickness of the sediment pile. Partitioning of the belt correlates directly with changes in the total thickness of the sediment pile above the middle Cambrian basalt detachment surface. These changes occur across major beltoblique faults, which, as noted above, suggests they were active prior to thrusting. Plausibly, these faults were generated during an extensional phase associated with a mantle plume in the early Permian (Zhang et al. 2008) or during a transtensional phase in the Jurassic (Sobel 1999). Studies from other fold–thrust belts and analogue experiments have demonstrated the impact of sediment thickness on the structural architecture of superimposed compressional structures. Liu et al. (1992) show through analogue sandbox experiments that thicker sediment piles produce thrust systems

81

in which major thrusts are widely spaced, creating wider piggyback basins. Conversely, thinner sediment piles deform as a series of more closely spaced thrusts separated by narrower piggyback basins (Fig. 9a). This arises because when a sediment pile is thick, fewer thrusts are required to attain the topography that satisfies the critical angle within the deforming wedge, than when the sediment pile is thin. When applied to the Keping Shan, this theory certainly seems to be applicable. The impact is best explored across the Piqiang Fault, the most prominent belt-oblique structure which causes substantial lateral discontinuity between the thrusts to either side and acts to partition two parts of the thrust belt. The sediment pile to the west of the Piqiang Fault is c. 4 km thick, while to the east it is c. 2 km thick. This is expressed in the deforming sediment pile as fewer thrusts to the west, and a greater number to the east (Fig. 9a). Furthermore, given that the Piqiang Fault was a pre-existing structure over which this stratigraphic discontinuity occurs, it provides an ideal plane of weakness that

Fig. 9. Impact of sediment thickness on thrusting: (a) Abrupt and substantial lateral change across a major pre-existing fault zone, which results in the reactivation of the fault and lateral partitioning of the thrust belt; (b) Abrupt but small change across a pre-existing fault zone, causing a kink or branching of the superimposed thrusts but without the need to reactivate the fault. These models are based on the direct observations from the Keping Shan and supported by theoretical models proposed by Liu et al. (1992).

82 S. TURNER ET AL.

Fig. 10. Schematic model of the Keping Shan illustrating the impact of lateral variations in the thickness of the sediment pile, major pre-existing fault zones, and the structural architecture of the superimposed (late Cenozoic) thrust belt.


can be reactivated as a strike-slip (transfer) fault which effectively decouples the thrust belt to either side of it, accommodating the abrupt lateral change in horizontal shortening and spatial organization of structures. Where the change in sediment thickness appears to be less substantial, it follows that the lateral variation in the thrust belt will also be less substantial. Changes in sediment thickness in the order of several hundred metres, such as that observed across the Sanchakou Fault, still cause disruption to the belt-parallel thrusts. However, rather than creating major strike-slip fault systems that completely partition the thrust belt, the belt-parallel structures accommodate these changes by branching and splitting into two thrusts (Fig. 9b). Plausibly, this pattern continues to apply where the net loss of sediment thickness is smaller or the change is less abrupt, such as across the Yijianfang Fault. Such belt-oblique structures have only a minor impact on the superimposed thrust system, such as a minor kink in the trace of belt-parallel thrusts. There is no requirement to reactivate the fault to accommodate this lateral change and were it not for the surface expression of the structure to the south (Fig. 1), within the Tarim Basin interior, such structures would probably go unnoticed within the thrust belt. Taking these observations into account, and applying the theoretical models (Fig. 9), we present a schematic model for the whole of the Keping Shan (Fig. 10). The model shows that lateral partitioning and structural variability is strongly affected by lateral variations in sediment thickness above the detachment horizon, which in turn are associated with the presence of major pre-existing structures that were potentially most active during the early Permian. Comparison with other studies demonstrates that the primary causes of lateral structural variability in the Keping Shan are well documented in other settings. Within the Cordilleran Fold– thrust Belt, western USA, Lawton et al. (1994) show that abrupt discontinuities that form between segments (compartments) of the thrust belt arise because of rapid lateral changes in the thickness of the sedimentary section. Where the change in sedimentary thickness is less substantial, diffuse transition zones form without the requirement to generate major strike-slip fault zones. A similar case from the Apennines, Italy, is presented by Butler et al. (2006), where major belt-oblique lineaments which coincide with substantial changes in the thickness of Mesozoic to lower Cenozoic strata have impacted on the plan-view architecture of the more recently imposed fold– thrust belt. In the Zagros Fold– thrust Belt, Iran, major pre-existing structures such as the Kazerun Fault have long-lived histories

83

and acted to control not just the sedimentary thickness but also the sedimentary facies. Early activity during the Cambrian controlled the distribution of the Hormuz salt, which has created a major lateral change in the detachment within the belt. Subsequent activity of these earlier structures during the Cretaceous impacted substantially on the thickness of the sediment pile, which has further enhanced lateral variability in the Cenozoic fold– thrust belt (Sepehr & Cosgrove 2004, 2007; Sepehr et al. 2006). These studies in other fold– thrust belts compare closely to the observations and interpretations we have drawn from the Keping Shan.

Implications for hydrocarbon exploration Lateral structural variability within foreland fold– thrust belts has important implications for hydrocarbon exploration in similar settings, which, as discussed throughout this volume, are likely to become of increasing importance in the future. To date, there have been no discoveries of hydrocarbons in the Keping Shan, which reflects the lack of a suitable source rock within the stratigraphic succession. However, the Keping Shan provides an ideal analogue for similar settings which may be rich in hydrocarbons. Abrupt discontinuities within fold–thrust belts, such as those described in this paper and in analogous settings such as the Cordilleran fold– thrust belt, the Apennines and the Zagros (Butler et al. 2006; Lawton et al. 1994; Sepehr et al. 2006), act to partition the fold–thrust belt into a series of structural domains and may thereby compartmentalize structural reservoirs of hydrocarbons. In addition, lateral variations in the spatial organization of structures between different compartments can enhance the reservoir potential of certain parts of the fold–thrust belt over others, by enhancing the concentration of structural traps. An important consideration, and one which has formed the basis of much of this paper, is the preexisting basin template and the presence of major lateral variations in the pre-existing stratigraphic framework. This may have additional implications for hydrocarbon distribution, locally removing important reservoir or seal formations from certain parts of the fold–thrust belt. Thicker areas of sediment may also be subjected to higher temperatures at depth, thus increasing the thermal maturity of source rocks and enhancing hydrocarbon generation in certain parts of the belt relative to others. In a fold-thrust belt such as the Keping Shan, where sediment thickness varies on the order of several kilometres along strike, the presence and state of the key play elements (source, maturation, reservoir, seal and trap) could vary significantly and abruptly across belt-oblique fault zones.

84

S. TURNER ET AL.

Conclusions Partitioning of belt-parallel thrusts within the east – west trending Keping Shan Thrust Belt occurs across major north– south to NW –SE belt-oblique fault zones and lineaments. Although these fault zones have developed coeval to thrusting during late Cenozoic contraction, there is strong evidence of earlier activity linked to either a period of extension during the early Permian or regional transtension in the Jurassic. This extension generated a series of major north –south to NW– SE trending faults across which there are substantial variations in the thickness of the sedimentary succession. late Cenozoic contractional deformation, associated with foreland deformation adjacent to the Tien Shan orogenic belt, generated the Keping Shan Thrust Belt. The lateral structural variability and partitioning of the belt is strongly controlled by the pre-existing faults and associated variations in sedimentary thickness. In several examples, the inherited structures have been reactivated during thrusting either as oblique-slip or strike-slip (transfer) faults which cross the thrust belt and help to accommodate abrupt lateral changes in the structure. Understanding how inherited structures and their impact on the sediment pile combine to influence the partitioning of the fold– thrust belt has important implications for hydrocarbon exploration in foreland settings. The Keping Shan serves not only to provide an insight into earlier phases of tectonism within the hydrocarbon-rich Tarim Basin, but can be applied as an analogue to fold–thrust belts worldwide. This study was funded by NERC and undertaken as part of the PhD research by S. Turner. SPOT datasets were provided by CNES through the OASIS programme. We are especially grateful to Nick Brook, Gareth Morgan and the Xinjiang Seismology Bureau for logistical and scientific support in the field during the summers of 2007 and 2008. We also thank Richard Phillips and an anonymous reviewer for constructive and helpful reviews.

References Abdrakhmatov, K. Y., Aldazhanov, S. A. et al. 1996. Relatively recent construction of the Tien Shan inferred from GPS measurements of present-day crustal deformation rates. Nature, 384, 450–453. Allen, M. B., Vincent, S. J. & Wheeler, P. J. 1999. Late Cenozoic tectonics of the Kepingtage thrust zone: interactions of the Tien Shan and Tarim Basin, northwest China. Tectonics, 18, 639– 654. Avouac, J. P., Tapponnier, P., Bai, M., You, H. & Wang, G. 1993. Active thrusting and folding along the northern Tien Shan and Late Cenozoic rotation of the Tarim relative to Dzungaria and Kazakhstan. Journal of Geophysical Research, 98, 6755–6804.

Bally, A. W., Chou, I.-M. et al. 1986. Notes on sedimentary basins in China: Report of the American Sedimentary Basins Delegation to the People’s Republic of China. US Geological Survey, Open File Report 86–327, 108. Burrett, C. F. 1974. Plate tectonics and the fusion of Asia. Earth and Planetary Science Letters, 21, 181–189. Burtman, V. S. 1975. Structural geology of the Variscan Tien Shan, USSR. American Journal of Science, 275A, 157–186. Butler, R. W. H. 1982. The terminology of structures in thrust belts. Journal of Structural Geology, 4, 239– 245. Butler, R. W. H., Tavarnelli, E. & Grasso, M. 2006. Structural inheritance in mountain belts: an Alpine– Apennine perspective. Journal of Structural Geology, 28, 1893–1908. Carroll, A. R., Graham, S. A., Hendrix, M. S., Ying, D. & Zhou, D. 1995. Late Paleozoic tectonic amalgamation of northwestern China: Sedimentary record of the northern Tarim, northwestern Turpan, and southern Junggar Basins. Geological Society of America Bulletin, 107, 571 –594. Carroll, A. R., Graham, S. A., Chang, E. Z. & Mcknight, C. 2001. Sinian through Permian tectonostratigraphic evolution of the northwestern Tarim basin, China. In: Hendrix, M. S. & Davis, G. A. (eds) Paleozoic and Mesozoic Tectonic Evolution of Central Asia: From Continental Assembly to Intracontinental Deformation. Boulder, Colorado, Geological Society of America Memoir, 194, 47–69. Chen, Y., Xu, B., Zhan, S. & Li, Y. 2004. First midNeoproterozoic paleomagnetic results from the Tarim Basin (NW China) and their geodynamic implications. Precambrian Research, 133, 271– 281. Chen, H. L., Yang, S. F., Wang, Q. H., Luo, J. C., Jia, C. Z., Wei, G. Q. & Li, Z. L. 2006. Sedimentary response to the Early– mid-Permian basaltic magmatism in the Tarim plate. Geology of China, 33, 545–552 (in Chinese). Coleman, R. G. 1989. Continental growth of Northwest China. Tectonics, 8, 621–635. Dahlstrom, C. D. A. 1970. Structural geology in the eastern margin of the Canadian Rocky Mountains. Bulletin of Canadian Petroleum Geology, 18, 332– 406. Hendrix, M. S., Graham, S. A., Carroll, A. R., Sobel, E. R., Mcknight, C. L., Schulein, B. J. & Zuoxun, W. 1992. Sedimentary record and climatic implications of recurrent deformation in the Tian Shan: evidence from Mesozoic strata of the north Tarim, south Junggar, and Turpan basins, northwest China. Geological Society of America Bulletin, 104, 53–79. Jia, C. Z., Zhang, S. B. & Wu, S. Z. 2004. Stratigraphy of the Tarim Basin and Adjacent Areas. 2. Science Press, Beijing, 513. Jiang, C. Y., Zhang, P. B. et al. 2004. Petrology, geochemistry and petrogenesis of the Kalpin basalts and their Nd, Sr and Pb isotopic compositions. Geological Review, 50, 492–500 (in Chinese with English abstract). Jun, G., Maosong, L., Xuchuang, X., Yaoqing, T. & Guoqi, H. 1998. Paleozoic tectonic evolution of the Tianshan Orogen, northwestern China. Tectonophysics, 287, 213– 231.

KEPING SHAN THRUST BELT, NW CHINA Lawton, T. F., Boyer, S. E. & Schmitt, J. G. 1994. Influence of inherited taper on structural variability and conglomerate distribution, Cordilleran fold and thrust belt, western United States. Geology, 22, 339–342. Li, D., Liang, D., Jia, C., Wang, G., Wu, Q. & He, D. 1996. Hydrocarbon accumulations in the Tarim basin, China. American Association of Petroleum Geologists Bulletin, 80, 1587– 1603. Li, Z. X. & Powell, C. McA. 2001. An outline of the paleogeographic evolution of the Australasian region since the beginning of the Neoproterozoic. EarthScience Reviews, 53, 237– 277. Liu, H., McClay, K. R. & Powell, D. 1992. Physical models of thrust wedges. In: McClay, K. R. (ed.) Thrust Tectonics. Chapman & Hall, London, 71–81. Macedo, J. & Marshak, S. 1999. Controls on the geometry of fold –thrust belt salients. Geological Society of America Bulletin, 111, 1808–1822. Marshak, S., Wilkerson, M. S. & Hsui, A. T. 1992. Generation of curved fold– thrust belts: Insight from simple physical and analytical models. In: McClay, K. R. (ed.) Thrust Tectonics. Chapman & Hall, London, 83– 92. Reigber, Ch., Michel, G. W. et al. 2001. New space geodetic constraints on the distribution of deformation in Central Asia. Earth and Planetary Science Letters, 191, 157– 165. Scharer, K. M., Burbank, D. W., Chen, J., Weldon, R. J., Rubin, C., Zhao, R. & Shen, J. 2004. Detachment folding in the Southwestern Tian Shan– Tarim foreland, China: shortening estimates and rates. Journal of Structural Geology, 26, 2119–2137. Sepehr, M. & Cosgrove, J. W. 2004. Structural framework of the Zagros Fold– thrust Belt, Iran. Marine and Petroleum Geology, 21, 829–843. Sepehr, M. & Cosgrove, J. W. 2007. The role of major fault zones in controlling the geometry and spatial organization of structures in the Zagros Fold– thrust Belt. In: Ries, A. C., Butler, R. W. H. & Graham,

85

R. H. (eds) Deformation of the Continental Crust: The Legacy of Mike Coward. Geological Society, London, Special Publications, 272, 419– 436. Sepehr, M., Cosgrove, J. W. & Moieni, M. 2006. Role of the Kazerun Fault Zone in the formation and deformation of the Zagros Fold– thrust Belt, Iran. Tectonics, 24, TC5005. Sobel, E. R. 1999. Basin analysis of the Jurassic– Lower Cretaceous southwest Tarim basin, northwest China. Geological Society of America Bulletin, 111, 709–724. Sobel, E. R. & Dumitru, T. A. 1997. Thrusting and exhumation around the margins of the western Tarim basin during the India–Asia collision. Journal of Geophysical Research, 102, 5043–5063. Sun, J., Zhu, R. & Bowler, J. 2004. Timing of the Tianshan Mountains uplift constrained by magnetostratigraphic analysis of molasse deposits. Earth and Planetary Science Letters, 219, 239–253. USGS (UNITED STATES GEOLOGICAL SURVEY) EARTHQUAKE HAZARDS PROGRAM 2009. Available online, http://earthquake.usgs.gov. Watson, M. P., Hayward, A. B., Parkinson, D. N. & Zhang, Z. M. 1987. Plate tectonic history, basin development, and petroleum source rock deposition, onshore China. Marine and Petroleum Geology, 4, 205– 225. Yang, Y. & Liu, M. 2002. Cenozoic deformation of the Tarim plate and the implications for mountain building in the Tibetan plateau and the Tian Shan. Tectonics, 21, 1059. Yin, A., Nie, S., Craig, P., Harrison, T. M., Ryerson, F. J., Xianglin, Q. & Geng, Y. 1998. Late Cenozoic tectonic evolution of the southern Chinese Tian Shan. Tectonics, 17, 1– 27. Zhang, C. L., Li, X. H., Li, Z. X., Ye, H. M. & Li, C. N. 2008. A Permian layered intrusive complex in the western Tarim block, northwestern China: Product of a ca. 275-Ma mantle plume? The Journal of Geology, 116, 269–287.

The use of palaeo-thermo-barometers and coupled thermal, fluid flow and pore-fluid pressure modelling for hydrocarbon and reservoir prediction in fold and thrust belts F. ROURE1,2*, P. ANDRIESSEN2, J. P. CALLOT1, J. L. FAURE1, H. FERKET1,3,4,5, E. GONZALES1,6, N. GUILHAUMOU7, O. LACOMBE8, J. MALANDAIN1,8, W. SASSI1, F. SCHNEIDER1,9, R. SWENNEN4 & N. VILASI1,4 1

IFP Energies Nouvelles, 1 – 4 Ave. de Bois-Preáu, 92852 Rueil-Malmaison, Cedex, France 2

VU-Amsterdam, de Boelelaan 1085, 1081 HV Amsterdam, the Netherlands 3

IMP, Apartado Postal 14-805, 07730 Mexico DF, Mexico

4

KU-Leuven, Celestijnenlaan 2000, B-300 Leuven, Belgium 5

6

VITO, Boeretang 200, 2400 Mol, Belgium

Pemex, Av. Urano 420, Col. Ylang Ylang, Boca del Rio, CP 94298 Veracruz, Mexico 7

Museum nat. Histoire Naturelle, 61 rue Buffon, F-75005, Paris, France

8

University Pierre & Marie Curie, Paris VI, Laboratoire de Tectonique, 4 Place Jussieu, F-75252, Paris, France 9

Beicip-Franlab, 232 Ave. Napoleón Bonaparte, PO Box 213, 92502 Rueil-Malmaison, Cedex, France *Corresponding author (e-mail: [email protected]) Abstract: Basin modelling tools are now more efficient to reconstruct palinspastic structural cross sections and compute the history of temperature, pore-fluid pressure and fluid flow circulations in complex structural settings. In many cases and especially in areas where limited erosion occurred, the use of well logs, bottom hole temperatures (BHT) and palaeo-thermometers such as vitrinite reflectance (Ro) and Rock-Eval (Tmax) data is usually sufficient to calibrate the heat flow and geothermal gradients across a section. However, in the foothills domains erosion is a dominant process, challenging the reconstruction of reservoir rocks palaeo-burial and the corresponding calibration of their past thermal evolution. Often it is not possible to derive a single solution for palaeoburial and palaeo-thermal gradient estimates in the foothills, if based solely on maturity ranks of the organic matter. Alternative methods are then required to narrow down the error bars in palaeo-burial estimates, and to secure more realistic predictions of hydrocarbon generation. Apatite fission tracks (AFT) can provide access to time– temperature paths and absolute ages for the crossing of the 120 8C isotherm and timing of the unroofing. Hydrocarbon-bearing fluid inclusions, when developing contemporaneously with aqueous inclusions, can provide a direct access to the pore-fluid temperature and pressure of cemented fractures or reservoir at the time of cementation and hydrocarbon trapping, on line with the tectonic evolution. Further attempts are also currently made to use calcite twins for constraining reservoir burial and palaeo-stress conditions during the main deformational episodes. Ultimately, the use of magnetic properties and petrographical measurements can also document the impact of tectonic stresses during the evolution of the layer parallel shortening (LPS). The methodology integrating these complementary constraints will be illustrated using reference case studies from Albania, sub-Andean basins in Colombia and Venezuela, segments of the North American Cordillera in Mexico and in the Canadian Rockies, as well as from the Middle East.

Present geothermal gradients can usually be derived from BHT (bottom hole temperature) measurements. Seemingly, the overall distribution of conductivities in the overburden can be reasonably described by applying standard values for each

dominant lithology, provided the latter can be properly documented by means of well logs and extrapolated laterally by the use of seismic sequences and attributes. In nearshore segments of passive margins and in foreland basins, where lithosphere


88

F. ROURE ET AL.

and crustal thickness remained relatively constant and only limited erosion occurred, vitrinite reflectance (Ro) and Rock-Eval (Tmax) values measured along vertical profiles (i.e. geochemical logs in wells) are usually sufficient, when combined with 1D well modelling (burial v. time curves), to derive realistic values for the palaeo-thermicity of a given area. In contrast, calibration of petroleum modelling becomes more complex in areas where both crustal and lithosphere thickness have been strongly modified since the deposition of the source rock, either in the distal portion of continental margins near the continent–ocean transition or in the inner parts of the orogens, where slab detachment or asthenospheric rise could result in drastic changes in the heat flow. Large uncertainties are also recorded when addressing petroleum modelling in foothills domains, basically because of the lack of controls on the palaeo-burial estimates in areas which have been strongly affected by erosion, and where it becomes challenging to solve for each time interval and for each cell of the model two unknown parameters (i.e. both temperature and burial). This paper will first briefly describe the current state of the art and integrated workflow developed recently for addressing basin modelling in fold and thrust belts (FTB). It will then document, based on representative case studies, the use of various palaeo-thermo-barometers for reducing the error bars in petroleum modelling in such tectonically complex areas as FTB, where major erosional events prevent any direct access to the palaeo-burial.

Integrated workflow developed for hydrocarbon and pore-fluid pressure modelling in FTB Dewatering processes and coeval overpressures build up have been widely documented in modern accretionary wedges by means of seismic attributes and deep ODP–IODP (International Oceanic Drilling Program) wells. For instance, the increasing load of synflexural sediments deposited in foredeep basins results in a vertical escape of formation water and a progressive mechanical compaction of the sedimentary pile where pore-fluid pressures remain dominantly hydrostatic. However, this process ultimately induces a velocity increase of seismic waves from the surface down to a depth where the vertical permeability reaches a minimum, precluding any further escape of underlying fluids toward the seafloor. Undercompacted sediments occur beneath this compaction-induced regional seal, being characterized by slower seismic velocities and overpressures. Worth mentioning, this occurrence of an overpressured horizon in the foreland strongly

decreases the mechanical coupling and friction between deeper and shallower horizons, thus helping the localizing and propagating forelandward of the deformation front. Although FTB share many similarities with offshore accretionary wedges in terms of the modes of thrust emplacement and overall structural style, boundary conditions of these two geodynamic systems are rather different in terms of porosity/ permeability distributions and fluid flow regimes. This is due to (1) the age of the accreted series (usually restricted to the relatively young synflexural sequences in accretionary wedges, against dominantly pre-orogenic passive margin sequences in FTB), and (2) the origin of the fluids (mixing of sedimentary fluids with meteoric water in FTB, against entirely marine or basinal fluids in offshore accretionary wedges). Unlike in modern accretionary wedges, overpressures can usually not be detected by anomalies in the seismic attributes in FTB, making integrated basin modelling techniques an indispensable tool, as documented below, to predict the distribution of pore-fluid pressures and hydrocarbon (HC) potential before drilling.

Coupled kinematic, thermal and fluid flow modelling in the frontal part of Eastern Venezuelan FTB The El Furrial and Pirital thrusts developing at the front of the Eastern Venezuelan thrust belt have been the focus of a pilot modelling approach coupling various 1D (Genex) and 2D (Thrustpack and Ceres: Sassi and Rudkiewicz 2000; Schneider et al. 2002; Schneider 2003; Deville & Sassi 2006) basin modelling tools. A structural section was first compiled from the interpretation of seismic profiles, and integration of wells and outcrop data. This section was then balanced and restored to its pre-orogenic configuration, providing an accurate control on the initial spacing of future thrusts. Incremental 2D forward kinematic modelling coupling erosion/sedimentation and flexure was subsequently performed with Thrustpack by means of a trial and error process, until the result section of the model was consistent with (1) the present architecture of the El Furrial and Pirital thrusts, (2) the pattern of erosional surfaces and unconformities currently observed in the Morichito piggyback basin and adjacent Pirital High, as well as (3) the measured temperature proxies (Ro from wells and outcrops). Despite strong erosion on top of the Pirital allochthon, where late Miocene and Pliocene series of the Morichito thrust-top basin rest locally directly on top of Cretaceous series (Fig. 1), this thrust unit

+ 100

+

(b)

50

N

0 +

MATURITY RATES OF THE ORGANIC MATTER (Tmax and Ro values of potential source rocks)

50 0

(a)

ST3 (Pj.1,7km)

EL PILAR FAULT 0

S

LO

A SERRA

UR

IC

A

MARGARITA

FA

ULT

DEL

NIA

MO

HIT RIC SIN BA

INTER IOR

T FAUL

SAN

O

PIRI

TA

L TH

CARIBBEAN

RU

AN - JU SAN ULT FA

S BA JO S FA UL T

NTA

L THR

–150

EAS

TER

N

VEN

EZU

ELA

N

N

A1

Tmax 421 to 437C Ro 0,5 to 0,57

2

TOBAGO

0 2

Tmax 441 to 444C Ro 0,38 to 0,64

4

NORTH RANGE IN BAS –50

TRINIDAD

PARIA GULF

4

Ro 0,8

6

6

8

8

CARIPE BLOCK

–10 BERGANTIN BLOCK

Orinoco

10 km

10 km

–10 Maturin

SANTA ROSA BLOCK

River

ORINOCO0 10 DELTA

0

A'

10 0

0

1

0

10

Miocene Carapita shale (type II - III) Late Cretaceous Querecual F. (type II) Lower Cretaceous Barranquin F. (type III)

Immature Oil window Gas window

50 km

0

5 km

10

(c)

1

S

GUYANA

Stage 2 (

0 –4

–16

–16

–28

–28

–40 km

–40 km 0

25

50

75

S

2

100

125

Stage 3 (

150

175

200

225

15 M.a.)

–28

–40 km

–40 km

S

125

150

175

225

Stage 8 Present

0 –4

0 –4

–16

–16

–28

–28

–40 km

–40 km 0

200

25

50

75

100

TR

Hydrocarbons in Contractional Belts

Hydrocarbons in Contractional Belts (Geological Society Special Publication 348)

Fossil Hydrocarbons

Reactive Hydrocarbons in the Atmosphere

Insect hydrocarbons

Archean greenstone belts (Developments in Precambrian geology)

Intraplate strike-slip deformation belts

Densities of Polycyclic Hydrocarbons

Bioremediation of Petroleum Hydrocarbons in Cold Regions

Strained Hydrocarbons Beyond the

Toxicological profiles - Polycyclic Aromatic Hydrocarbons

Bioremediation of Petroleum Hydrocarbons in Cold Regions

Densities of Aromatic Hydrocarbons

Advances in the Theory of Benzenoid Hydrocarbons

Toxicological profiles - Total Petroleum Hydrocarbons

Metal-Catalysed Reactions of Hydrocarbons

Six Sigma: Advanced Tools for Black Belts and Master Black Belts

Handbook Timing Belts: Principles, Calculations, Applications

Aromaticity of Polycyclic Conjugated Hydrocarbons

Evaporites: Sediments, Resources and Hydrocarbons

Physical Properties of Hydrocarbons vol.2

Dark Nebulae, Dark Lanes, and Dust Belts

Characteristic Differentiation in the Spectra of Saturated Hydrocarbons

Insect Hydrocarbons: Biology, Biochemistry, and Chemical Ecology

Physical Properties of Hydrocarbons vol.1

Thermophysical Properties of Chemicals and Hydrocarbons

Hydrocarbons in Crystalline Rocks (Geological Society Special Publication No. 214)

Thermophysical Properties of Chemicals and Hydrocarbons

The Carcinogenic Effects of Polycyclic Aromatic Hydrocarbons

Cratons and Fold Belts of India (Lecture Notes in Earth Sciences)

Hydrocarbons in Contractional Belts