A BFA-CM OPTIMIZATION LOG INTERPRETATION METHOD

It is difficult to calculate reservoir parameters of tight sand reservoirs using conventional interpretation methods, due to their complex lithology and variable pore structure. An optimization log interpretation method is able to take full advantage of the log data and geological information. Therefore, it is an effective method to evaluate tight sand reservoirs. In this study, in order to calculate the reservoir parameters of tight sand reservoirs, an appropriate interpretation model needed to be first established according to the reservoirs’ characteristics. Then, the interpretation parameters were chosen, and the specific form of the objective function was determined. Next, an optimization algorithm was adopted to search for the optimal solution. A bacterial foraging algorithm (BFA) is a newly developed algorithm which has strong global search capabilities. It simulates the behavior of the colon bacillus which swims with flagella for food in the human gut. However, since it slowly converges in the later part of the optimization, it was combined in this study with a complex algorithm (CM) for constituting a BFA-CM hybrid algorithm, in order to improve the precision and efficiency of the search process. Also in this study, the unknown reservoir parameters of the optimization log interpretation method were determined using a genetic algorithm (GA), particle swarm optimization (PSO), BFA algorithm, and BFA-CM hybrid algorithm, respectively. The calculation results showed that, when compared with the GA and PSO, the errors of the porosity and the component content calculated by the BFA were minimal. However, the calculation result curves were found to be inconsistent. Therefore, by combining a BFA algorithm with a CM algorithm to constitute a BFA-CM hybrid algorithm for calculating reservoir parameters, the accuracy was improved, and the curves became more stable. The results of the BFA-CM optimization log interpretation method verified that the objective function value was F ≈ 0. Also, the sonic, neutron, and density log theoretical value curves (AC0, CNL0, DEN0) fell within the confidence interval, which indicated that a system deviation influence did not exist, and that the optimization results were reasonable and credible. When compared with the other algorithms, the BFA-CM hybrid algorithm displayed unique advantages during the process of calculating the unknown parameters with the optimization log interpretation method. Its calculation results were of high accuracy and stability, and the efficiency was also improved. The experimental results showed that the BFA-CM optimization logging interpretation method was able to accurately calculate the tight sandstone reservoir parameters, and could therefore be applied to actual production practices.