Abstract
Knowledge of reservoir fluid properties is very important in reservoir engineering calculations such as reserves estimates, material balance equation, and numerical reservoir stimulation etc. Laboratory measurements are the most dependable methods for pressure-volume-temperature (PVT) analysis of reservoir fluid properties. During reservoir performance evaluation, the measurements of PVT data are not readily available, hence engineers resort to the use of existing empirical correlations to obtain PVT data. Often time, the data obtained from correlations do not agree with the experimental data, thereby yielding a considerable amount of error. To correct that error in this study, Kartoatmodjo and Schmidt correlation for estimating the oil formation volume factor was optimized using the particle swarm optimization (PSO) algorithm to minimize the error in estimating the oil formation volume factor from Kartoatmodjo and Schmidt correlation at various depletion pressure. The optimized correlation is a function of bubble point pressure, API gravity, gas gravity and reservoir temperature. Therefore, to validate the optimized correlation, PVT data from differential liberation test in the laboratory was used as case a study and the result obtained with the optimized correlation matches closely with the experimentally values at various depletion
pressure. Also, the Kartoatmodjo and Schmidt optimized correlation was validated with Standing, Kartoatmodjo and Schmidt, Glaso and Petrosky-Fashad correlations. The optimized Kartoatmodjo and Schmidt gave the least average relative error of ±0.00065 and the highest correlation coefficient of 0.99996 after 154th successive iterations with the PSO algorithm.