Optimization of Infill OilWell Locations Field-Scale Application

摘要

Determining the optimal locations of infill wells is a crucial decision to be made during a field development plan. It concerns implementing accurate reservoir modelling in order to precisely evaluate the reservoir behavior and predict its future performance. The reservoir modelling-optimization approaches were adopted on the main pay-Upper Sandstone formation of South Rumaila Oil Field, located in Iraq. First of all, comparing outcomes of different parameters with their measured values through history matching process attained the validity of the reservoir flow model. After that, two methods of optimization were adopted to find out the optimal number and locations of infill oil wells.The first method is manual optimization via spreadsheet and the second one is automatic optimization throughAdaptive Genetic Algorithm (GA). Both methods were done according to the aspects of net present value (economic evaluation) as objective function in the wells selection optimization procedures. GA depends on the principle of artificial intelligence concept of Darwin's theory of Natural Selection. The genetic program was coupled with the reservoir flow model to re-evaluate the chosen wells at each iteration until obtain the optimal choice. The genetic algorithm program gave results similar to the results that were obtained by manual method with much less computation time. Three different future predictions of oil production and NPV cases were studied to determine the optimal future scenario with respect to whether considering water injection or not in the available water injectors. The first one without water injection, the second and third with 7500 surface bbls/day and 15000 surface bbls/day water injection per well, respectively. According to the relationship between net present value and future production time, the abandonment time was estimated to be at the end of the 8th prediction year for all above cases. The optimal future scenario was with water injection of 15000 surface bbls/day; however, the current capabilities of surface injection facilities cannot handle this rate. Therefore, the optimal future prediction is to continue with water injection of 7500 surface bbls/day/well. The optimal number of infill wells for this case was three wells even though drilling more wells have led to increase the cumulative oil production. The incremental percent of NPV based on the optimized infill well location scenario are improved by 3.4% higher than the base case on no-infill wells.