Fuzzy Optimization Model Of Oilfield Development Program And Its Application

The production optimization is the main content of oilfield development planning, which aims at obtaining the maximization of productivity and benefit or the minimization of oilfield development cost under a limited investment. Production optimization can be further divided into production composition, production distribution and stimulation production composition. Generally, mathematic programming method is applied to solve the restrictiveness problem between goal and constraint during optimization process. However, most of the methods were established under definite condition and can not solve the actual production problems with fuzzy characters in parameter, variable, constraint and goals. In fact, during production optimization of oilfield development, fuzzy characters are widely existed in resource evaluation, oil price predicting, market demanding and so on. The definite models which ignore fuzzy factors during production optimization can not reflect the actual process. So, it’s important and meaningful to study oilfield production optimization based on fuzzy programming theory.Firstly, this paper made a comprehensive analysis on the uncertainly during oilfield development process, and defined the fuzzy factors and their manifestation in production optimization process. The fuzzy factors mainly include the flexibility of constraints, the fuzzy objective and the fuzzy number of parameters coefficient or variable in the model.Secondly, according to fuzzy programming theory, we established fuzzy objective resource models and fuzzy coefficient models based on the different fuzzy manifestation in the model..Each model was further classified according to planning period (annual or multiple years) and optimization target (single or multiple objectives). When the oilfield was considered as a whole, we established fuzzy objective resource and fuzzy coefficient production optimization models with various planning period and optimization target. In addition, two-level fuzzy objective resource models and fuzzy coefficient parameters models with different optimization targets were established when the oilfield was considered as a two layer system.Thirdly, membership function and triangular fuzzy number were applied to convert fuzzy models to deterministic models based on fuzzy theory. Then, the established fuzzy models were solved by the solution of deterministic optimization process and various programming plans were obtained. Compared to deterministic production optimization method, fuzzy production optimization were more suitable to the production practice by introducing membership and fuzzy numbers which reflected the changeable interval and the degree of probability.As the programming plans varied with optimization target and method, it is necessary to establish an effect evaluation method to obtain the best plan. In this paper, a non-linear fuzzy comprehensive evaluation method with prominent impact factors was set up and this method combined the characteristics of prominent impact factor and overcome the limitation of tradition fuzzy comprehensive evaluation method. In addition, both the non-linear fuzzy comprehensive evaluation and linear fuzzy comprehensive evaluation method were applied to the evaluation of fuzzy coefficient production composition models. The evaluation results were consistent partly, but the result of non-linear fuzzy evaluation reflected the effect of prominent impact factors and was more suitable for the oilfield production optimization, which possesses prominent impact factors in evaluation indicator system.