Application of stochastic simulation for hydrocarbon in place estimation

Abstract

Original hydrocarbon in place estimation is very important in oil field development. The volumetric method which requires the estimates of porosity and water saturation is one of the techniques to determine original hydrocarbon in place. This study focuses on the estimation of porosity and water saturation distributions based on available data. Using three different stochastic simulation techniques which are Monte Carlo Simulation (MCS), Sequential Gaussian Simulation (SGS), and Sequential Gaussian Cosimulation (SGCOSIM), porosity and water saturation probability distribution functions were generated and used to estimate original hydrocarbon in place and quantify its uncertainty. A set of field data obtain from locations along three directional wells drilled in a gas reservoir was selected for this study. Then, the three stochastic methods were used to determine original gas in place (OGIP) for the selected area of gas field. Results from the three methods were compared to determine the best methods. Based on the criterion that the best algorithm should provide the least variance of OGIP estimate, SGCOSIM was determined to be the best method. To quantify the influence of spatial structure variables on hydrocarbon in place estimation, variogram parameters such as nugget value, search distance, and correlation coefficient were varied. Considering the mean and variance of OGIP, the most sensitive spatial structure variables is the nugget value followed by correlation coefficient and search distance.