Efficient reservoir modeling-statistical approach for evaluation and development of future waterdrive undersaturated oil reservoir performance

For achieving optimal reservoir management, an insight view towards good understanding of the reservoir behavior should be attained. Then, an efficient design of integrated reservoir modeling is a crucial requirement for planning of an optimal future development especially in mature oil fields. The field under study is South Rumaila oil field/main pay. It is a mature oil field located in South of Iraq with around 59 years of production history. It has 40 producing wells and 20 injection wells located only at the east flank to maintain the reservoir pressure since there is an infinite acting edge water aquifer and its strength from the west flank is much effective more than the east flank. A reservoir simulation study has been done to determine the optimal future reservoir performance in regards to the effectiveness of water injection to improve oil recovery. In this paper, multiplicity scenarios have been done to investigate the efficiency of water injection and consider the streamline-based simulation results for optimal water injectors optimization. This study has considered the cumulative oil production (NP) and net present value (NPV) as an objective functions to find out the optimal decision. In order to find the accurate attainment, the first future prediction scenario has been performed without water injection by running the black-oil reservoir model for future period production of 12 years. Then, full factorial design (FFD) considering water injection rates and number of water injectors as factors has been adopted. Twenty-five experiments have been carried out for the two factors with five levels. The full factorial design (FFD) results showed that the oil recovery is getting higher at the lowest levels of water injection rate and number of injectors; however, the injection rate has the highest effect. The optimal scenario has shown higher cumulative oil production when activate nine wells with 4000 STB/DAY. Never the less, the FFD has shown that the number of activated injectors has the highest influence when the NPV is the objective function and NPV has the highest value at the lowest water injectors. Consequently, the optimal scenario is either considering one injector with 2000 STB/DAY injection rate or to continue with base case of non-injection as there is very small difference in NPV. This optimal injector has connections with 15 producers and should support the reservoir pressure. The validity of the two linear models has been done by t-test, p-value, and F-Statistics & Analysis of Variance fANOVAV. Copyright © 2014, International Petroleum Technology Conference.