View article

Final goal of reservoir management is finding a high net present value during the forecast period of reservoir assessment. The first step in reservoir management is history matching. Even if we have a good history matched model, without a robust production optimization algorithm, high value of NPV cannot be found. Field-scale optimization problems consist of a highly complex reservoir model with many control variables as unknowns. So finding a high value for NPV in a reasonable time depends highly on the efficiency of optimization algorithm. There are many optimization algorithms and in this paper we study three efficient algorithms for doing optimization. These are steepest ascent (SA), sequential quadratic programming (SQP) and interior point (IP) methods. In petroleum, steepest ascent is a very popular method in maximizing NPV of reservoirs due to the small optimization cost. This method is an unconstrained optimization and due to the nature of this algorithm, it cannot find a high NPV. In contrast, sequential quadratic programming (SQP) is a robust constrained optimization algorithm and it can find a high NPV after about 20 iterations, but computation cost of each iteration is significant and it is not comparable with steepest ascent method. In this paper we introduce another optimization algorithm ie interior point method. Although final NPV in this method is not as high as SQP, but it is much more than steepest ascent method and computation cost of each iteration is similar to the steepest ascent method. For better realizing advantages and disadvantages of each algorithm, we consider application of these methods to the Brugge field which …