TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractNew methods have been developed to initialise a compositional model for a giant Middle East reservoir where the initial H 2 S mol% varies laterally from close to 0 to around 20 and vertically from close to 0 to around 20, being highest at the lower parts of the reservoir.Conventional initialisation methods assume equilibrium, assigning the same composition to any grid block at a given depth. Two different new method, triangulation and surface-fit, are used to capture observed lateral nonequilibrium variations. Vertical variations in fluid composition are simulated using a gravity segregation model as in the conventional method. The nonequilibrium methods use data on the variation in fluid composition and pressure derived from 18 representative fluid samples collected from wells early in the field history.

Reservoir management requires tools that can (a) provide fast track and accurate assessment of a large variety of operations, while (b) are capable of quantifying uncertainties associated with management decisions. Reservoir managers must be able to compare and contrast a large number of development scenarios, while taking into account the uncertainties and risks involved with each scenario, in a relatively short period of time. To achieving this important task with traditional technologies one must either sacrifice the accuracy or the speed.

As reservoirs mature, subsurface flow complexity and surface production operation challenges increase. This brings the necessity of making capital-intensive decisions to sustain or increase reservoir potential in an optimum way. However, subsurface uncertainties affect decision success. Reservoir surveillance, a process that involves data acquisition, validation, analysis, integration opportunity generation and execution, can mitigate the outcome of such decisions in the presence of uncertainties. Although Value of Information (VOI) is a well-known process for justifying data acquisition, engineers struggle to extract the relevant information from historical data to apply Bayesian approach. The objective of this paper is to illustrate a methodology for identifying the value of information in reservoir management, in particular for deriving the conditional probabilities of success when new and imperfect data are acquired. A methodology to assess the value of reservoir surveillance is supported by two cases. In the first case, the incremental value of Real-Time Reservoir Characterization (RTRC) in underbalanced drilling (UBD) was nearly 100 times the cost of the services; in the second case, the incremental value of permananet downhole gauges (PDHG) was near 230 times the cost of installation and services. Reliability of facquired data, among other uncertainties, resulted to be a key success factor for both cases; however, in worst-case conditions, the incremental value was always positive.

Fast track development projects, with timely data acquisition plans for development optimization, are very challenging for tight and heterogeneous carbonate reservoirs. This paper presents the challenges and key learning from initial stages of reservoir development with limited available data. Focus of this study is several stacked carbonate reservoirs in a giant field located in onshore Abu Dhabi. These undeveloped lower cretaceous reservoirs consist of porous sediments inter-bedded with dense layers deposited in a near shore lagoonal environment. The average permeability of these reservoirs is in the range of 0.5-5 md. Mapping the static properties of these reservoirs is difficult since they are not resolved on seismic due to the low acoustic impedance contrast with adjacent dense layers. Petrophysical evaluation of thin porous bodies inter-bedded with dense layers in highly deviated wells pose significant challenges. Laterolog type LWD resistivity measurements which are less affected by environmental effects, offer more accurate formation resistivity compared to propagation type measurements. With limited suite of logs, some of the zones with complex lithology had to be evaluated innovatively as detailed in the paper. Integrated studies are initiated to improve reservoir description by carrying out accurate permeability mapping, SCAL, geomechanical and diagenesis & rock typing studies. Significant challenges exist regarding the development of thin, tight and highly heterogeneous reservoirs, in terms of recovery mechanism, well architecture, well count, drilling, well completion and economics. Static and dynamic models were used extensively to evaluate different development scenarios and conduct sensitivity studies to bracket uncertainties. Various geo-steering options were discussed and the paper also details maximizing the reservoir productivity using long reach MRC (Maximum Reservoir Contact) wells. Tight and heterogeneous reservoirs call for extensive and real time reservoir surveillance activities to assess well performance and reservoir connectivity. This paper highlights how these challenges are overcome through upfront surveillance planning and proactive well completion strategy.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractNew methods have been developed to initialise a compositional model for a giant Middle East reservoir where the initial H 2 S mol% varies laterally from close to 0 to around 20 and vertically from close to 0 to around 20, being highest at the lower parts of the reservoir.Conventional initialisation methods assume equilibrium, assigning the same composition to any grid block at a given depth. Two different new method, triangulation and surface-fit, are used to capture observed lateral nonequilibrium variations. Vertical variations in fluid composition are simulated using a gravity segregation model as in the conventional method. The nonequilibrium methods use data on the variation in fluid composition and pressure derived from 18 representative fluid samples collected from wells early in the field history.