THE ROLE OF RESERVOIR SIMULATION IN THE DEVELOPMENT OF THE CHALLIS AND CASSINI FIELDS

Abstract

The Challis oil field development was approved in 1987 with marginal reserves (for an isolated offshore project) of 22 MMbbl. The initial development envisaged three subsea production wells connected via a riser to a floating production facility with one water injector also being required to maximise recovery. However, due to additional potential in the vicinity of the field, the production system was designed to accommodate up to 10 production/injection wells.

Further appraisal in 1988/1989 doubled the reserves to 43 MMbbl and increased the number of initial production wells to seven from five reservoirs. The appraisal results also confirmed earlier concerns as to the structural complexity of the field. Analytical interpretations of the production tests performed on the wells could not be fully reconciled with the available well log, core and seismic data. Furthermore, the analytical models developed from these interpretations could not fully match the test results.

Reservoir simulation was used to resolve, where possible, the discrepancies. Individual reservoir models were calibrated with the production test results and used to quantify the major uncertainties and their potential impact on production performance. The simulation results indicated that water injection may not be required. However, the degree of internal reservoir communication and the extent of the expected aquifer support were identified as the two principal unknowns.

Production policy and monitoring procedures were structured to resolve these uncertainties as quickly as possible during the production start-up phase. Comparative forecasts of expected performance were developed for each reservoir with various levels of aquifer support. A surface controlled interference test was designed to investigate the extent of internal reservoir communication in the main reservoir.

The success of the interference test and the results of the early well performance have confirmed the simulation predictions. Simulation modelling was successful in quantifying the range of expected pressure response (to production) for each reservoir and was able to quickly confirm the degree of pressure support present in each reservoir.