In reservoir engineering, a variety of data is needed to accurately estimate reserves and forecast production. Fluid characterization consist of reservoir rock analysis and fluid analysis. The determination of gas condensate dew point pressure, gas specific gravity and producing yield is essential for fluid characterization, gas reservoir performance calculations and the design of production systems. The importance of gas condensate reservoirs has grown continuously and the condensate fluid has gained increased importance.
Traditionally, the dew-point pressure of a gas condensate is experimentally determined in a laboratory in a process called constant mass expansion (CME) test using a visual window-type PVT cell and the constant volume depletion test (CVD). However laboratory measurements, though most reliable, have been found to be laborious, costly and still subject to errors. Hence, the need for other simple and accurate methods of predicting the dew-point pressure (DPP) for a gas condensate. Thus using different equations of states (EOS), several correllations have been developed to predict the DPP of a gas condensate fluid.
This paper therefore seeks to predict the dew point pressure of a gas condensate using genetic algorithm. This is an improved, faster, simple and most accurate method of predicting the dew point pressure.
Genetic algorithm is an adaptive heuristic searching methodology, introduced on the evolutionary ideas of natural selection and genetics, which follows the Charles Darwin's principle of "Survival of the Fittest". Genetic Algorithm represents an intelligent development of a random search within a defined search space (also called as population) to obtain the optimum solution of the problems. Thus, GA would be very effective in carrying out this multi variable optimization.