Abstract

This paper summarizes the results of a joint EAGE/AAPG research conference that was convened in Almeria, Spain in October 1998. The theme of the conference was how to better produce deep-water reservoirs based on lessons learned from the past 25 years. A repeated message at the conference was that there is more complexity than anticipated in turbidite reservoirs, contrary to the expectations of many geoscientists. Such complexity may go unnoticed during initial depletion, and only be observed during secondary injection of fluids. Early recognition of shale occurrences and geometries, bed continuity, and stratigraphic variations in net-to-gross ratios appear to be the main issues related to maximizing well performance.

Introduction

With the increased emphasis on exploration for, and production in, deep-water reservoirs during the past 25 years, there is now a sizable amount of production information from fields. Increased drilling in the twenty-first century makes these an important target for exploration through enhanced recovery. In many of the newly discovered fields, the current trend is for companies to fast-track field development, with minimum drilling, no workovers and intervention, increased use of subsea tiebacks, optimal individual well rates and ultimate recoveries, and expanded perforated intervals. To accomplish this requires a good understanding of the architecture of reservoirs and prediction of their production history based on the use of multiple integrated datasets, often with a minimum number of data points from wells. The most commonly used datasets are 3D seismic, wireline logs, borehole images, cores (possibly) and quantitative (where possible) reservoir models from both subsurface and analogue outcrop data.

To address these many topics in the development of deep-water reservoirs, we convened a joint EAGE/AAPG research conference in Almeria, Spain on 3–9 October, 1998, titled ‘Developing and managing turbidite reservoirs: case histories and experiences’. Some 65 participants from 40 companies, universities and government agencies attended. Twelve countries were represented. The symposium provided a forum to discuss and present experiences and technology applications related to the development and management of turbidite reservoirs worldwide.

A total of 25 oral presentations and 13 posters were given. Programme elements included: new fields; fast-track developments; key geoscience and reservoir engineering technologies and their integration; plateau and mature field management; geological modelling and reservoir performance prediction; reservoir description and outcrop analogues; and the integration and application of static and dynamic data. The main fields described in the conference are shown in Fig. 1 and summarized in Table 1. Also included in Table 1 are the presentations that described other aspects of turbidite reservoirs. Kick Kleverlaan and Peter Haughton led a two-day field trip to the Tabernas Basin.

This paper reviews the important results of this conference for the AAPG/EAGE membership because of the increased emphasis on deep-water reservoirs in the twenty-first century and their impact on future industry development. We specifically emphasize the salient points stemming from discussion groups during the last morning of the conference. Four specific topics are addressed: (1) lessons learned from maturing and mature fields; (2) data collection from exploration to late stage field development; (3) developing a common language between geoscientists and reservoir engineers to improve efficiency of reservoir teams; and (4) outcrops as reservoir analogues and for reservoir modelling. There is some overlap between the topics because of the nature of the material.

Lessons Learned From Maturing and Mature Fields

Maturing fields

A key message repeated often at the meeting was that in maturing fields, the production history turned out to be considerably different than that originally predicted. For many fields, surprises in later production occurred that did not appear during initial production. Several fields from different basins were identified as having surprising, and often negative, results during their development. For example, water breakthroughs occurred in unexpected places owing to early, simplistic reservoir models. Often, initial production, if examined carefully with other datasets, did indicate potential problems with reservoir development (e.g. Magnus Field).

In thin-bedded, channel-levee-overbank reservoirs, such as at Tahoe and Popeye fields, an early key issue was thin-bed continuity. As new wells were drilled, pressure drawdowns indicated uniform, widespread depletion and long distance continuity. However, differential vertical depletion in some instances indicated the presence of baffles or significant permeability variations. In other cases, differences in elevation of fluid contacts on opposing sides of channels implied that flanking levees are either partially or completely separated by mudstone channel-fill.

Other lessons learned from maturing fields such as Bullwinkle are that high rate, high ultimate production wells are essential, cycle time from discovery to start-up must be dramatically reduced, development costs must be dramatically lowered, and companies must be able to operate economically in more than 5000 feet of water. Small fields are now being tied back to larger fields, for example in the Gulf of Mexico, Angus to Bullwinkle, Europa and King to Mars, Tahoe to Bud and Macaroni to Auger (Shell 1999).

Auger Field provided an example of conservative planning for production. Excellent lateral continuity of thick, permeable sands in the S Reservoir resulted in higher-than-anticipated flow rates of up to 12 000 BOPD with cumulative production from April 1994 (start-up) through August 1998 of 89 × 106 barrels of oil equivalent (BOE).

In Garden Banks 236 Field, the key to effective reservoir management was recognition and characterization of reservoir quality, strong water drive, and shale barriers of variable length which resulted in reservoir compartments and perched oil/water contacts.

Mature fields

Mature fields are recognized to be especially important as analogues for geological models, but most importantly in terms of learning about optimization of development strategies. This was a repeated theme at the conference. Development plans for many fields should be based upon past experiences in more mature fields; unfortunately such experiences are not always utilized.

A real need was also expressed to develop appropriate criteria for choosing the correct analogue, be they mature fields or outcrops. Experiences were related about development plans for the Bullwinkle and Auger fields in the northern Gulf of Mexico (Holman & Robertson 1994; McGee et al. 1994). For both fields, some of the early development plans, such as waterflood, were never used. At Bullwinkle, reservoir models from fluvial-deltaic fields were used in setting the platform. In this case, the unanticipated high production rates reflected the continuous nature of sheet sand reservoir rocks in a confined intra-slope basin.

Lessons learned from three North Sea turbidite fields were also presented. For these fields, the goal was to minimize appraisal by using proper datasets, then to move quickly into subsea development. The business decision was made to build flexible reservoir models instead of large models. The flexibility enables changes and modifications in response to production history in the field. Second-order heterogeneities in the reservoir had their greatest negative impact when the fields were coming off plateau production. Some first-order models proved incorrect. A lesson learned from these fields was that data collected for one purpose early in the life of the field might be used later to solve an entirely different problem. Also, wells that are producers today may not be producers in the future, and entire new suites of wells will be drilled before field life ends. In some older fields, all of the production today is from new penetrations, and many early development wells have been shut-in or sidetracked to exploit new targets. Initial development wells cannot necessarily be expected to maintain high production rates throughout the life of the field.

The first example, the Paleocene Forties Field was discovered in 1970. By 1974, it was producing 0.5 × 106 BOPD, which was considered peak production. The lower portion of the reservoir comprises a high net-to-gross sandstone, whereas the upper part of the reservoir comprises lower net-to-gross strata. The original development plan incorporated the field as a ‘tank model’, so that flank wells could be used for water injection. Following a period of stable production until 1984, there was gentle decline until 1987, then rapid decline from 1987 to 1992. In response to the need for an infill programme, a 3D seismic survey helped identify the main lithofacies, fluid distribution and zones of unswept oil, and new wells were drilled to produce from the lower net-to-gross sands. The lessons learned were that close reservoir surveillance is required, there will be unexpected reservoir performance characteristics through time, and a good infill programme should be planned and updated as more data are obtained.

The second example cited was the UKCS Magnus Field, discovered in 1983. It was the first major Jurassic submarine fan field discovered in the North Sea, with a 1000+ft. thick oil column. Although it is a high net-to-gross reservoir, there were always difficulties in correlating sands across the field. Initially this was thought to have a minor impact on production. In time, though, a number of shale units and faults were seen to act as major permeability barriers to fluid flow despite the high net-to-gross nature of the reservoir sandstone. The impact of such shales can only be recognized with careful and consistent reservoir modelling and surveillance. A field-wide debris flow acts as a pressure barrier between the upper and lower portions of the reservoir. A major cross-field fault also was found to act as a barrier to waterflood. Early wells have now been sidetracked and today’s production is largely from new or sidetracked wells.

The third example used was the Paleocene Andrew Field, which also has a high net-to-gross ratio, but with upward increasing shaliness and heterogeneity. Lessons learned from the older Forties field development were applied to this field. The field was marginally economic (130 × 106 BOE), and there was a real need to isolate the gas cap and underlying aquifer encroachment during production due to the presence of a thick gas cap, active aquifer and relatively thin oil column (180 feet). A series of ten horizontal wells were drilled, each with a reach of 6000 feet. Localized fine-scale biostratigraphic zonation, together with fine-scale geocellular models were combined with innovative completions and a comprehensive surveillance strategy to optimize well placement and manage conformance. Well performance is strongly influenced by the geometry of shales in the reservoir, which impact gas migration. The key was to understand their nature so as to manage production when it comes off its peak.

The Miocene Yolumne Field in the San Joaquin Basin of California was described as a mature field with remaining, bypassed oil being trapped in flanking levee thin-beds. Technical challenges were decreasing reservoir thickness and quality of levee thin-beds, as well as laterally continuous shales that vertically isolate reservoir sandstones. Exploitation of this oil was accomplished by a slant well drilled into the thin-bedded facies and artificial fracturing to break shale barriers.

An important message from mature fields was the importance of continuing to collect and analyse data throughout their life. Breaks in data collection will lead to later, unexpected problems, as exemplified by the Magnus Field. An example was provided of the Long Beach Unit of Wilmington Field in the Los Angles Basin. To date, this field has produced 2.5 × 109 BOE from about 5000 wells. Even with so many wells, there was considerable bypassed oil, so an aggressive infill drilling programme was initiated. Old well files fortunately had been retained, and were used after considerable manipulation, to successfully implement a horizontal well and alkaline water/steam flood project.

In summary, a good rule of thumb should be that complexity is always greater than anticipated in turbidite reservoirs. This is contrary to most people’s expectations. Turbidite reservoirs are usually considered less complex than other types of reservoirs. Clearly all reservoir types reveal their true complexity late in field life. Such complexity may go unnoticed during initial depletion, and only be observed during secondary injection of fluids. Early recognition of shale occurrences and geometries, and stratigraphic variations in net-to-gross ratios appear to be the main issues related to maximizing well performance. They should be the basis for focused data collection and surveillance strategies.

Data Collection from Exploration to Late Stage Field Development

A major direction that companies have been moving towards is the fast-tracking of fields. The move toward rapid subsea completions does not allow for production intervention and workovers. The need to minimize costs also is of paramount importance. This direction, however, is potentially in conflict with important development decisions that must be made throughout the life of a field. There is a real need for a frank discussion of the kinds of data needed to be collected to maximize long-term production, but minimize time between discovery and start-up. Presentations at the conference emphasized that much of the data that are collected will be field- and basin-dependent. More mature basins, where generic issues concerning the development of different play types are well known, require different datasets and analytical approaches, compared with new frontiers or lightly explored plays; e.g. the approach to a new development of the Paleocene North Sea reservoirs will be different from the new fan plays in deepwater West Africa, northern Gulf of Mexico or Brazil.

Two types of data – static and dynamic – were recognized as important to maximize long-term production. Both are needed from the early stage of a discovery’s appraisal and should be fully integrated to get maximum value from their information. Two general considerations for appropriate data collection and management in a field are:

  1. good practice data are data collected at various stages in field life that produce good investment returns; and

  2. problem-solving data can provide appropriate information for contingencies when something goes wrong in the production history of a well

Static data

Static data are a key data source early in the development of a field, especially when dynamic data are often sparse and equivocal indicators of major reservoir heterogeneities. Rock data include cores, sidewall cores, and full suite conventional wireline logs. The importance of specialized logs such as borehole images and nuclear magnetic resonance logs is just now being realized. Initial potentials (IPs) and fluid distribution and properties, pressure data, as well as productivities and chemistry are also important data types to capture.

3D seismic data, which are recognized as a must for development, are not absolutely required at the exploration stage. There is a cost trade-off in many basins between 2D and 3D seismic data at the wildcat stage. Many of the fields that were discussed at the conference were discovered with 2D seismic data because they are within well-defined structures. However, there was agreement that 3D seismic data is essential to field development, especially for fast-track developments. The Zafiro Field in Equatorial Guinea, discovered in March 1995 by 2D seismic data, had first oil production 18 months later. 3D seismic data were collected and processed on the seismic ship within 7 months of discovery and used throughout the planning for production. In this case, 3D seismic data helped to identify future well locations. Initial production is from a Floating Production Storage Offloading (FPSO) facility; by the year 2003, a platform will be set. Two other techniques were used in the development of this field: interference testing between wells to determine the lateral continuity of reservoir strata; neural network techniques which distinguished oil from gas and thick- from thin-bedded sandstones.

Dynamic data

The need to collect dynamic data during the life of a field was considered of paramount importance, particularly to maximize production as the field matures. In the future, the need for pressure monitoring through downhole gauges will be essential to subsea tiebacks and similar production scenarios that cannot have intervention. The use of time lapse 3D seismic imaging for monitoring fluid movement is increasing, and has been used successfully in areas only where favourable reservoir properties exist. The monitoring can be accomplished in two primary ways: by placing permanent geophones on the sea floor and acquiring data at different times (e.g. Foinaven Field, West of Shetland, UK) or by conducting repeated 3D seismic surveys using similar acquisition parameters (e.g. Forties Field, UK North Sea).

A major hurdle in collecting dynamic data is cost and associated assessment of the commercial value of the data. The relative merit and cost of having data, versus not having a particular piece of data, must be presented in any budget review. For example, the platform for the Shell Bullwinkle Field in the northern Gulf of Mexico was designed for more slots than were needed owing to the better-than-anticipated performance. This over-design of facilities may have been preventable if a strategy of dynamic data collection and analysis had been pursued.

Repeat Formation Tests (RFTs), which are often the lower-cost choice for dynamic data gathering, may not solve all reservoir problems. The relative merits of short vs. extended well tests also need to be evaluated during appraisal. An unspoken, but extremely important aspect of not collecting sufficient dynamic data is associated with making a wrong decision, which often cannot be changed without costly intervention or facility upgrades. Can the mistake of not spending money early in dynamic data collection be justified in the context of reduced reserves, poor well rates or ultimate recoveries from a well or field?

Planning, budgeting and gathering of data are an interdisciplinary affair. For both exploration and production, plenty of thought, discussion and planning are required. In particular, all disciplines must be represented at the planning stage. Also, data should be quality controlled while it is being gathered: e.g. the engineer on the rig is as essential to the successful collection and integration of data as is the geologist or planner. It is also important that all of the data are used after collection. It does no good to collect a borehole image log, then file it away. A few examples like this provide fodder for future budget denials!

In summary, both static and dynamic data are crucial to maximizing hydrocarbon recovery as a field matures. It is essential that teams understand and clearly demonstrate the value of and need for such data and understand their impact on the reservoir depletion plan. It is important that different disciplines participate in evaluating the cost-benefit of acquiring such data, and the potential loss of value by not acquiring an adequate static and dynamic dataset.

Developing a Common Language among the Disciplines to Improve Efficiency of Reservoir Teams

One of the major common ‘soft’ challenges in the development of turbidite, and other types of fields, is the need to develop a shared understanding of rock and fluid properties and reservoir geometries to explain and predict reservoir behaviour. In essence, when working in integrated teams, verbal and written communication of concepts, interpretations and facts is a real problem between geologists, geophysicists, petrophysicists, reservoir engineers and modellers. Team members commonly come from different backgrounds and perceptions of subsurface problems may be quite varied. Commonly, geologists find that a ‘team’ field trip to study outcrops, where clear stratigraphic relationships are exposed and concepts can be discussed while looking at rocks, is a good way to bridge the communications gap. Discussing the rocks at a ‘watering hole’ is also successful for team building.

In a similar vein, maintaining team continuity can pay real dividends. There commonly is a hiatus between the exploration and development stage of a field, and a long time interval between field start-up and later stage development. Team continuity and longevity are vitally important so that there is a historical record of decisions for the field. Too much critical data and levels of understanding are lost because of transfers or employment changes within a team.

At the conference, the breakout group that discussed the communication issue came from six different companies who worked development projects from many basins and fields. Their common experiences led to some simple communication rules for integrated teams: (1) be simple and clear; (2) be consistent with definitions of words and terminology (including geometry); (3) be disciplined in the use of terms and avoid jargon wherever possible; (4) be precise; (5) be clear on your own uncertainty; (6) be extremely clear of the ultimate goal of a project and its component parts; and (7) make geological work relevant to the necessary tasks and decisions at hand.

These relevant tasks and decisions generally include: (1) initial reserves estimation; (2) production profiles; (3) facilities design; (4) development well locations, costs and design; (5) depletion strategy; (6) infill and enhanced recovery strategies; (7) managing decline; and (8) the decision to sell or abandon the field near the end of its life. It was also deemed critical that subsurface personnel have appropriate equipment and support personnel to meet their needs.

Outcrops As Reservoir Analogues and For Reservoir Modelling

The use of outcrops as analogues to reservoir models, specifically for helping to build reservoir models, has become an increasingly common practice during the past several years. There are probably 12–15 world-class turbidite outcrops that are of significant lateral and vertical size to be described and documented in a semi-quantitative fashion for analogue reservoir modelling of fields. Figure 2 shows the most common outcrops used by different companies for these purposes. Table 2 lists the key outcrops and selected references. Many internal and industrial-sponsored research projects are occurring on critical outcrops globally. Conference participants were in unanimity as to the importance of outcrop characterization, although the relative importance of different outcrops in reservoir models was debated. Everyone agreed that a better synthesis of all of the different imaging systems for turbidites (outcrop and subsurface) must be achieved.

Exploration-scale applications

When using an outcrop as an analogue for a reservoir, the larger basinal, stratigraphic, and depositional context must be kept in mind. A major shortcoming of many outcrops is that they rarely provide insights into the entire depositional system: we are rarely able to view the nature of the sediment delivery system into the basin, which plays an important part in the reservoir architecture. The classification system of Reading & Richards (1994), and follow-up work by Richards & Bowman (1998) emphasized that the nature of the turbidite sediment delivery system exerts a fundamental control on the geometry, architecture and internal character of the deposit. Therefore, there is a real need to understand both updip and downdip terminations of fields, and how these effect reservoir performance. Also, the nature of coarse-grained heterolithic canyon fill interpreted to be coeval to considerably higher net-gross downslope channel elements raises questions about the nature of some feeder systems.

The scale of most outcrops precludes understanding turbidite systems at the larger exploration (seismic) scale. The Tabernas Basin field trip provided an example of difficulties in relating outcrops to subsurface reservoirs owing to scale variations. There were numerous, excellent small-scale architectural features which might affect reservoir performance, but comparison with larger-scale reservoirs, such as Miller Field in the North Sea, was problematic. Aerial photography can sometimes be of use in documenting larger-scale features of outcrops.

Thin-bedded turbidite reservoirs are particularly problematic, and it was stated at the conference that our understanding is still evolving. Typical well spacings in the mature fields in the northern Gulf of Mexico are 500 m. Many recent discoveries would be sub-commercial at such well densities, emphasizing the need to exploit technologies to optimize well production and placement. Outcrops of thin-bedded turbidites are quite valuable for understanding thin-bed architecture, but most outcrops typically are the size of one grid cell in a reservoir model. Lateral bed measurements are useful, but small-scale faults and erosional cutout can limit the utility of continuity data for reservoir modelling.

Sediment bypass in a basin, which is commonly recognized on sub-regional 3D seismic data, when integrated with lithostratigraphy and biostratigraphy, is considered a major process which controls how turbidite systems stack regionally. This process was illustrated at the conference in several presentations from the North Sea, Gulf of Mexico, West Africa and Brazil, but is difficult to recognize at the scale of most outcrops.

It may be that due to these issues and the greater reservoir and fluid complexity in many new discoveries, recovery factors will be lower than historical field data suggest.

A goal for outcrop workers is to develop facies criteria for predicting sedimentary bypass in outcrop. This has been done in the Carboniferous of West Ireland (Fig. 2, Table 2) with surprising results: thin-bedded facies outside of channels, interpreted in the subsurface of the Gulf of Mexico as overbank facies, can be unequivocally demonstrated to be the expression of bypass zones of downdip sand depocentres.

The role of syndepositional (salt, fault and shale deformation) tectonics on basin topography can be imaged at a meso-scale to the point where bathymetric barriers can be seen to affect stratal patterns on 3D seismic horizon slices. When applied to outcrop studies, one is often limited regarding the identification of potential barriers and bathymetric control; nevertheless, emphasis should be given to how this might be an influence on reservoir architecture and sand body geometry. The nature and aspect of basin topography needs to be more fully explored in outcrop studies.

One aspect that has plagued turbidite workers is the difficulty in comparing characteristics of ancient turbidite systems with modern submarine fans, as determined from side-scan sonar, shallow penetration seismic data and cores. The spectacular seafloor images that have been collected in the past 15 years of different modern turbidite systems provide the need to reexamine outcrops in the light of processes revealed from these modern studies. 3D seismic horizon images along the Gulf of Mexico and West Africa margins make us pause to reflect on how we examine outcrops. There now exists extensive 3D seismic databases from many margins of the world. There is a real need to study the upper one second of these data, coupled with seafloor images, to learn more of sedimentary processes, geometries, etc. In addition, high-resolution, shallow-penetration seismic data from shallow hazards studies in intra-slope areas offer an excellent new database for improved, detailed seismic characterization.

Reservoir scale

The emphasis in much current outcrop work is in turbidite elements, their architecture and geometry. There was a repeated emphasis at the conference on using a hierarchical approach to identifying and understanding these elements and their bounding surfaces in outcrops, and how these are used for building reservoir models. This is particularly important for object-based modelling.

Little emphasis currently is given to shale beds and how they control reservoir performance. Reservoir development and well performance often are a direct function of the frequency of occurrence and geometry of shales. Hierarchical schemes for sandy turbidite elements are being developed. There is a real need to develop a parallel hierarchical scheme for shales. Recognition criteria need to be developed to differentiate shales that are barriers from those that are merely baffles, from those that begin as barriers and break down as production matures. Sub-regional condensed sections, autocyclic shales and debris flow mudstones are examples of important, but physically different shale types. Some work of this type is being done now in the Tabernas Basin.

A lack of adequate understanding of turbidite and related sedimentary processes was repeated at the conference. Simple bed forms can be most important in predicting overall architecture of a reservoir as well as determining flow behaviour. Yet, there clearly remains considerable debate as to the nature of certain deep-water processes.

Another value of outcrop studies is in the identification of key flow units in a reservoir analogue and how these will impact a reservoir simulation. Quantification of lateral continuity and vertical connectivity are essential, as is the role of small- to large-scale erosional scour in erasing depositional continuity trends. Two-dimensional and, preferably, 3D permeability trends and predictive functions need to be developed from outcrops for input into reservoir simulators. Hard data on bed lengths and thicknesses are critical for both grid-based and object-based modelling, and need to be collected to build large, statistically reliable databases.

A major issue throughout the conference was upscaling fine-scale architecture and pore-scale reservoir quality to field scale. Though often difficult, upscaling is essential to the success of building reliable reservoir models. Using outcrop petrophysical information for upscaling is particularly difficult since outcrops have experienced burial diagenesis unlike that to which Cenozoic, and especially Neogene, reservoirs have been subjected. The main value of outcrops is in revealing architectural and geometrical details, as well as relations between relative petrophysical and facies properties. On the field trip, examples were shown of permeability trends in Tabernas Basin turbidites, and facies control on permeability. The effects of sorting on permeability were particularly well displayed.

The value of behind-outcrop coring and logging (particularly borehole image logging) of turbidite strata for comparison with field attributes was demonstrated by the use of a New Zealand outcrop in the improved understanding of the Ram/Powell Field (Gulf of Mexico). Goals of behind-outcrop coring and logging were to characterize details observed on borehole images for prediction of lateral bed attributes, as well as to place outcrop information into a format more familiar to subsurface geoscientists and engineers. Similar studies are now being conducted in the Ainsa channel of northern Spain, the Jackfork turbidites of Arkansas and Lewis Shale turbidites of Wyoming.

In summary, outcrops appear to be a key element in promoting successful communication and interaction between geoscientists and modellers, as well as engineers. There is still much to be learned about what parameters need to be measured for upscaling and fluid flow simulation. In addition, outcrop work trains people to think in terms of rocks.

Final Comments

The conference was a tremendous success for several reasons.

  1. A group of working specialists was assembled in one place, all of whom openly shared their experiences from different fields and outcrops globally. There were abundant opportunities to compare and contrast deep-water reservoirs. The chemistry between the group was fantastic and everyone learned from the experience.

  2. A suggestion was made to develop an internet site or electronic bulletin board where information and experiences could be shared between companies.

  3. The conference was valuable in providing contacts and leads to others with similar issues and interests, and for information exchange. This was viewed to be a major help in future cooperation and work between companies. Everyone agreed that each benefited from the public presentations and information exchange.

  4. A co-operative industry project, in the general vein of the successful Deep Star project in the northern Gulf of Mexico (Burton 1997), is clearly needed to share technology problems associated with developing reservoirs in deep-water.

  5. There was consensus for the real need to compile data from the key outcrops around the world into some large atlas, similar to the atlas of Pickering et al. (1995). The purpose is to have a clear understanding of the reservoir geometries and issues regarding upscaling. Also, sidescan sonar data need to be updated and made more readily available through publication to help us understand depositional processes and the geometries of turbidite elements. Such a publication would be enhanced if it was in electronic format.

It is worth noting that future key turbidite conferences, or those with significant turbidite components, include: the 1999 AAPG/PESGB Birmingham Convention, the 2000 AAPG New Orleans Convention, the 2000 GCSSEPM Houston Research Conference, and the 2000 AAPG/IPA Bali Conference. The recent 1999 AAPG National Convention in San Antonio also displayed a significant amount of current turbidite studies, particularly in poster format. It appears that within the next two years, significant advancements in our understanding of this important type of oil and gas reservoir will emerge through sharing of technology and scientific advances.

We thank Eveline Schut and Anja Kroon of the EAGE for their logistical support throughout the planning of this conference. We thank the following participants of the EAGE/AAPG Research Conference for their discussion and presentations that form the basis for this review paper: Vitor Abreau (Unocal), Nadia Al-Abry (Edinburgh University), Yousef Al-Aufi (Edinburgh University), Philippe Bourges (Elf), Wayne Camp (Anadarko), Mark Chapin (Royal Dutch-Shell), Mike Clark (Arco), Julian Clark (Heriott-Watt), Don Clarke (City of Long Beach), Dominique Claude (Elf), George Clemenceau (BP-Amoco), James Coleman (BP-Amoco), Steve Cossey (consultant), Bryan Cronin (Aberdeen University), Bob Davis (Schlumberger), Peter Diebold (Royal Dutch-Brunei), Shirley Dutton (BEG-Austin), Trevor Elliott (Liverpool), Franco Fonnesso (Agip), Mike Gardner (Colorado School of Mines), Tim Garfield (Exxon), Tim Good (Mobil), Frank Goulding (Exxon), Adrian Hartley (Aberdeen University), Peter Haughton (Dublin), Neil Humphreys (Mobil), Liz Jolley (BP-Amoco), Phillipe Joseph (Elf), John Kantorowicz (Amoco), John Kendrick (Shell), Kick Kleverlaan (consultant), V. Kolla (Elf), Markus Leishmann (Amerada Hess), Walter Maciel (Petrobras), Wenceslao Martinez (Repsol), Ole Martinsen (NorskHydro), Xaveir Mathieu (Total), Mike Mayall (BP-Amoco), William McCaffrey (Leeds University), Trey Meckel (Royal-Dutch/Shell), Chris Perry (Conoco), Ronal Pattinson (International New Ventures), Henry Pettingill (Repsol), Kevin Pickering (University College), Gonzalo Ruiz (Schlumberger), Kevin Schofield (Conoco), Will Schweller (Chevron), Randy Smith (Texaco), Chuck Stelting (Chevron), Dorrick Stow (Manchester), Susan Strommen (Statoil), Jose Torres (Repsol), Susana Torrescusa (Repsol), Richard Vaughan (Repsol), Jose Carlos Vincente (Repsol), Jacques Vittori (Elf), Roger Walker (Norsk Hydro), John Warrender (Conoco), Markus Weissenbaeck (RAG-Vienna), Jonathan Wonham (Elf). We thank Neil Hurley, AAPG editor, for his review of the manuscript.

Appendix

The following is a list of the authors, their affiliations and their papers that were presented at the EAGE/AAPG research conference on ‘Developing and managing turbidite reservoirs: case histories and experiences’.

Camp, W. K. (Anadarko), Early drilling results at subsalt Mahogany Field: exploitation strategies for thin-bedded turbidite reservoirs, Gulf of Mexico, USA.

Chapin, M. (Shell UK E&P), Using turbidite outcrop data to influence deepwater development decisions.

Clark, M. S., Prather, R. K. & Melvin, J. D. (Arco), Reservoir characterization of a fan-shaped turbidite complex in an active-margin basin, Miocene Stevens Sandstone, Yowlumne Field, San Joaquin Basin, California.

Clark, D. D. & Phillips, C. C. (City of Long Beach, California), Subsidence and old data present unique challenges in aging turbidite oil fields: examples of successful technological solutions from the Wilmington Oil Field, California, USA.

Clemenceau, G., Colbert, J., Lockett, F. & Musso, B. (Amoco), Development drilling results from selected deepwater Gulf of Mexico Fields: implications on turbidite reservoir characterization and production performance prediction.

Coleman, J. L., Jr. (Amoco), Geology of the Miocene lower Rudeis deepwater sandstone reservoir, Gulf of Suez.

Cossey, S. (Consultant), Producibility of turbidite reservoirs: a comparison of the North Sea and Gulf of Mexico.

Cronin, B. T. & Hurst, A. (University of Aberdeen), Generic components of deep-water clastics: building blocks for the sub-surface from modern and ancient analogue data.

Dutton, S., Barton, M., Malik, M. (Bureau of Economic Geology), Asquith, G. B. (Texas Tech University), Cole, A. G., Pittaway, K. R. (Conoco) & Gogas, J. (Digital Prospectors), Characterization and development of turbidite reservoirs in a deep-water channel-levee and lobe system, Ford Geraldine Unit, Permian Bell Canyon Formation, Delaware Basin, USA.

Elliott, T. (University of Liverpool), A renaissance in the analysis of turbidite systems: implications for reservoir development and management.

Gardner, M. H., Johnson, K., Batzle, M., Sonnefeld, M. & Sinex, B.(Colorado School of Mines), Geologic building blocks for reservoir characterization: lessons learned from the Permian Brushy Canyon Formation, West Texas.

Garfield, T. R., Sullivan, M. L., Foreman, L., Liesch, A. & Jennette, D. (Exxon), Application of deepwater outcrop analogues to 3-D reservoir modeling: an example from the Diana Field, western Gulf of Mexico.

Gordon, A. F. (Elf Exploration – Aberdeen), Kolla, V. (Elf Aquitaine – Angola), Cooper, R. (Elf Exploration – Aberdeen), Claymore Field, North Sea: integrated seismic stratigraphy, depositional models and reservoir mapping.

Hartley, A. J. (University of Aberdeen), Hutchinson, A. (AP –Aberdeen), Hole, M. J. (University of Aberdeen) & Boyd, D. (BP – Aberdeen), A novel geochemical correlation technique: application to a turbidite reservoir.

Haughton, P. (University of Dublin), Subsurface lessons from Tabernas turbidites.

Imbert, P., Mouezy, A. (Total) & Parize, O. (Ecole des Mines), Post-burial deformation of clastic series, mechanisms and impact on reservoir geometry.

Joseph, P. (IFP), Gomes de Souza, O. Jr, (Petrobras), Eschard, R., Granjeon, D., Lerat, O. & Ravenne, C. (IFP), 3D architecture of turbidite reservoirs from outcrop analogues in the French Alps and application to Brazilian fields.

Kendrick, J. W. (Shell Deepwater), Turbidite reservoir architecture in the Gulf of Mexico – insights from field development.

Kneller, B., McCaffrey, B. (University of Leeds), Knight, I. (Phillips Petroleum), Hailwood, E. (Core Magnetics Palaeomagnetic Services) & Bayes, R. (Phillips Petroleum), A multidisciplinary approach to sedimentological modeling and reservoir characterisation: Andrew A1 Sandstone, Central Graben, North Sea.

Kolla, V., Bourges, P. & Urruty, J. M. (Elf Exploration, Angola), Claude, D. & Morice, M. (Elf Exploration Production), Durrand, E. (Elf Petroleum – Nigeria) & Pirmez, C. (Lamont-Doherty Earth Observatory), Reservoir architecture in recent and subsurface, deep-water meandering-channel and related depositional forms.

Leishman, M. (Amerada Hess), Depositional elements and processes within a confined Palaeocene/Eocene deep-water depositional system – the Bittern Field, UK North Sea.

Leonard, A. & Bowman, M. (BP) Managing Turbidite Fields in the North Sea.

Maciel, W. B., Del Lucchese, C. Jr., Corá, C. A. G. & Pinto, A. C. C. (Petrobras), Development and management of turbidite reservoirs in Campos Basin, Brazil.

Martinson, O., Indrevaer, G., Dreyer, T., Mangerud, G., Ryseth, A. & Soyseth, L. (Norsk Hydro), Slumping, sliding and basin floor physiography: controls on turbidite deposition and fan geometries in the Paleocene Crane Field area, Block 25/11, Norwegian North Sea.

Pettinghill, H. S. (Repsol Exploracion), Turbidite giants: lessons from the world’s 40 largest turbidite discoveries.

Slatt, R. M. (Colorado School of Mines), Why outcrop characterization for reservoir studies?

Slatt, R. M. (Colorado School of Mines), Browne, G. H. (Institute Geological and Nuclear Sciences, New Zealand), Clemenceau, G. R. (Amoco), Davis, R. J. (Schlumberger – Jakarta), Young, R. A. & Anxionaz, H. (Services Techniques Schlumberger – Paris) & Spang, R. J. (Texaco), Behind-outcrop characterization of thin-bedded turbidites for improved understanding of analog reservoirs: New Zealand and Gulf of Mexico.

Slatt, R., Al-siyabi, A., Van Kirk, C. W. & Williams, R. W. (Colorado School of Mines), From geologic characterization to ‘reservoir simulation’ of a turbidite outcrop, Arkansas, USA.

Smith, C. R. (Texaco), Petronius project: reservoir characterization, modeling, and fluid-flow simulation.

Stelting, C. E., Schweller, W. J., Florstedt, J. E., Fugitt, D. S., Herricks, G. J. & Wise, M. R. (Chevron), Production characteristics of sheet and channelized turbidite reservoirs, Garden Banks 236 Field, Gulf of Mexico, USA.

Stephen, K. D., Clark, J. D. & Good, T. R. (Heriot Watt University), Modeling the effect of turbidite amalgamation on reservoir flow properties using outcrop data.

Stow, D. A. V. (University of Southampton), Deep-water sedimentary systems: new models for 21st century.

Strommen, S. K., Halvorsen, C., Langlais, V., Laursen, G. V., Nadeau, P. & Samuelsen, E. T. (Statoil), Sleipner Ost Field, a sand-rich Palaeocene (Ty Formation) gas-condensate reservoir offshore Norway; sedimentology, stratigraphy, heterogeneity and paleocontact influence on reservoir properties, flow and production.

van den Berg, A. (Shell Deepwater), Deepwater Gulf of Mexico development overview.

Weimer, P., Crews, J. R., Crow, R. & Varnai, P. (University of Colorado – Boulder), Atlas of the petroleum fields and discoveries in northern Green Canyon, Ewing Bank, and southern Ship Shoal and South Timbalier areas (offshore Louisiana), northern Gulf of Mexico.

Weissenback, M. & Kreczy, L. (RAG – Vienna), Examples of turbiditic reservoirs in the Oligocene (Upper Puchkirchen Formation) of the Upper Austrian Molasse Basin.

Williams, T. A., Humphreys, N. V., Monson, G. D. & Blundell, L. C. (Mobil), Technology application as an enabler for rapid development of the Zafiro Complex, Equatorial Guinea.

Wonham, J. P. (ELF UK), Jayr, S. (ELF UK, Imperial College), Chuilon, P. (Elf Exploration Prod) & Mougamba, R. (Universite des Sciences et Technologies de Lille), 3D Sedimentary evolution of a turbidite channel reservoir (early Miocene-age) of the Baudroie Marine and Baliste fields, offshore Gabon.