After more than 20 years of production, the 800 million STB Bokor field, offshore Sarawak, Malaysia, is set to undergo a revitalization to increase production rates and recovery factors. A joint PETRONAS Carigali (PCSB)-Schlumberger team was formed to review the field and develop the first full field simulation model, which will be utilized as a ‘live’ model for identifying further potential and for future reviews. This paper outlines the modeling workflow and processes developed to allow the study to be completed within a shorter duration than with a conventional approach, as well as the key study results.
Made up of some 130 vertically stacked reservoirs over a 6,000-ft interval, evaluation of the field is hampered by the lack of any useful seismic images over the hydrocarbon zone, owing to shallow gas and other anomalies. Understanding the reservoir behaviour has always been a challenge, with the degree of sand consolidation in the reservoirs varying from totally unconsolidated at 1,500 ft subsea, filled with 10 cP oil, to consolidated deeper sands, with 0.1 cP oil at 7,500 ft subsea. In addition, the limited pressure depletion owing to the presence of a very strong aquifer, and the fact that many fluid contacts have not been penetrated, means that significant uncertainty still exists over the likely stock-tank oil initially in place (STOIIP).
By fully utilising all available data and an iterative team-based approach to history-matching the geostatistical models with production data, an understanding of the key parameters was gained. This revealed that in many reservoirs, what was previously considered poor-quality rock was in fact a major contributor to reservoir flow, being neither as poor in permeability nor as high in water saturation as previous interpretations suggested.
Through the increased understanding gained during the review, robust predictions have allowed new facilities to be appropriately sized, allowing the field to begin its new lease of life.
Many of the challenges to redeveloping the Bokor field are related to continuing uncertainty about the initial oil in place (OIP). As detailed elsewhere,1 the estimation of the STOIIP has more than tripled since the original decision to develop was taken. Even with the extensive well penetrations and production period, significant uncertainty still remains because of the existence of a shallow gas cloud (Fig. 1) that masks much of the faulted crestal area, especially in the deeper intervals in which the seismic imaging is particularly poor (Fig. 2). In addition to the uncertainty in bulk volume, the extensive vertical interval (more than 130 reservoirs between 1,500 and 7,500 ft subsea) and limited flank wells means that for many sands no initial fluid contact was established. These factors result in a wide band of uncertainty on remaining OIP, with substantial upside potential, especially in the deeper sands. Add to these structural issues a 12-month project time frame, severe hole washout in many sands, internal gravel packs making individual sand offtake determination difficult, a low gas/oil ratio (GOR), high-viscosity oil, and a very strong aquifer, and the complexity of the task becomes clear.
Whilst a few individual reservoir models have been built over the field life to review sweep and infill locations, no full field model had been constructed previously. Production forecasts were the result of decline curve analyses and were therefore not readily updatable with any change in operating conditions such as increased lift gas injection. Therefore, a key deliverable was a full field reservoir model within which various redevelopment scenarios could be considered before any major capital expenditures are finalised.