We present a dynamic network model for modeling two-phase flow. We account for wetting layer flow, meniscus oscillation, and the dynamics of snapoff. Interfaces are tracked through pore elements using a modified Poiseuille equation for the equivalent hydraulic resistance of the fluids between the pore element centers. The model is used to investigate the effects of capillary number and viscosity ratio on displacement patterns and fractional flow in primary drainage. We show that the amount of snapoff increases with increasing capillary number and decreasing wetting phase viscosity. For capillary numbers lower than approximately 10(-5), the pore-scale fluid distribution and fractional flow are similar to those obtained using a quasistatic model that ignores viscous forces. The contribution of oil transport from ganglia, formed by snapoff, is negligible except for very large capillary numbers, greater than around 0.1.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractDuring a large water flood study on a cluster of fields in South Oman it became apparent that relative permeability constituted one of the major uncertainties impacting reserves in the cluster. At the onset of the study, only two experimental measurements were available that had been acquired with the currently recommended approach of wettability restoration and a combination of steady state and centrifuge experiments. Therefore, the team proposed to core five wells and embarked on a large scale special core analysis (SCAL) program, covering all predominant rock types, in order to get a better handle on the relative permeability characteristics.This paper presents a case study of using a properly measured set of relative permeability data to replace the previously used analogue database and hence reduce uncertainties of waterflood recovery predictions. The experimental programme followed a recommended procedure of wettability restoration and a combination of steady-state and centrifuge experiments. When the experimental data became available, they were reviewed and numerically interpreted using the state-of-the art simulation techniques.This has led to several insights that were missed in earlier field studies which used a set of simplified correlation functions/parameters for the cluster of fields without adequate special core analysis data calibration. Based on the new results the field's relative permeability characteristics are divided into two categories linked to rock types thus significantly reducing the uncertainty range. In this paper we will also highlight the procedure that was used to generate the new SCAL experimental dataset and the analysis that has been done to arrive at this conclusion.The simulation effort and the subsequent analysis have reduced the uncertainty in relative permeability by a factor of three resulting in a significant improvement in the robustness of the development plans.

No abstract

Steam Assisted Gas-Oil Gravity Drainage (SAGOGD) trial is planned for a limited area of a giant producing light oil field in Oman. Oil production from this oil-wet fractured carbonate reservoir commenced in 1967, and recovery factor currently stands at approximately 20%. The SAGOGD process in a light oil fractured reservoir is complex and is comprised of numerous recovery mechanisms, with a number of these being uncertain and poorly understood. Very little world analogue data is available [1], and that, combined with large recovery process uncertainties make this ‘large pilot scale’ Phase 1 essential to mitigate the downside risk in a full-field development. During Phase-1 it is planned to inject 2000t/d of steam by means of 4 vertical steam injectors. Oil, gas and condensed steam will be produced by 7 horizontal producers and 5 vertical back-up producers. The magnitude of the SAGOGD production response is highly uncertain. Having the capability to accurately measure the incremental oil production response over this wide uncertainty range was considered to be a key success factor for the Phase 1 project. To accurately measure the incremental response required that a ‘no steam’ production response could be confidently projected into the future for a minimum of two, and up to five years. This task was made considerably more complex by the fact that historical GOGD well production profiles were often relatively unstable. This paper describes the work carried out within PDO to ensure that one of the key Phase 1 success criteria – that being to measure the incremental oil due to SAGOGD – can be achieved over a primary evaluation period of two to five years. The discussion will include a description of efforts linked to optimization of cold GOGD performance (optimum oil rim management), well production stabilization (via installation of new production control hardware) and accurate measurement of total and individual well production levels (dedicated bulk and well-test facilities), and how this all came together to yield a stable cold production baseline which could be confidently projected into the future.

An integrated study on a cluster of 23 fields in South Oman was performed in the Petroleum Development Oman (PDO) Study Center, in order to derive a coherent view on the cluster development as well as to write (waterflood) FPDs for the individual fields. The study combined conventional and state-of-the-art workflows and was conducted by a large integrated team over a period of three years. The study was conducted in four phases: screening, full field modelling, infra-structure development and FDPs for the 6 large fields and the writing of appraisal / development plans for the remaining 17 fields. This paper describes the workflow and learnings of the study. Introduction The cluster of fields in South Oman holds a significant near-term growth potential through active waterflood implementation and later, through polymer flood recovery (Fig.1). The cluster development strategy addresses 23 clastic fields with a significant STOIIP base. A large integrated team of PE staff have spent 3 years to address and progress the major fields therein to FDP. The Cluster consists of green and brown fields characterised by high gross liquid production, rapid watercut development (>80%), variable well performance and marked pressure decline. The fields are medium to small salt withdrawal-related structures often with stacked reservoirs, varying oil quality (7–400cP) and low to medium fault density. To date, only a small percentage of the STOIIP has been recovered. This marks the cluster with high development potential for large-scale waterflood. The study was undertaken to support further development decisions, to identify the value of remaining reserves including growth and SFR opportunities and to create life cycle development plans. Key challenges to unlock the value of the cluster was to provide a clear scope maturation plan for each field by understanding surface and subsurface controls leading up to a development concept selection and culminating later with the roll out of 23 field development plans. The study started in Oct 2003 and by end 2005 delivered 6 value-assured FDPs for the large fields in Phase 1 covering 2/3rd of STOIIP. The remaining satellite fields were pursued in Phase 2, and delivered 17 FDP's by the end of 2006 covering the remaining 1/3rd of STOIIP. For evaluation of capital intensive waterflood developments the operating asset team required that the subsurface models be robust with relatively long shelf life. Workflows and synergies were built across the asset-study team interface, to not only deliver quality FDPs but also through embedment, ensuring successful execution. Fig.2 shows the cluster study work flow. The study clearly demonstrates that a blend of conventional methodology and state-of-the-art stochastic techniques can efficiently model waterflood behaviour. The Screening Phase In the early stage of the study it was essential first to get a high level overview of the fields and their underlying uncertainties in order to focus more in-depth work at the subsequent stages of the project. Some of the fields had been studied in the years before, but these studies were performed by different teams, at a different maturity level of the fields and with different objectives. The current study provided the first cluster wide attempt for development planning. A multi-disciplinary effort went into the screening phase to help identify data gaps and formulate clear appraisal strategies. This was built upon the following steps: Seismology: Existing seismic interpretations were inventorised and QC'd. Obvious problems and inconsistencies were fixed and a first pass estimate of the top structure map uncertainty was provided for all fields.