The Impact of Porosity Uncertainty in Reservoir Modeling Using Multiple-Point Geostatistics

This study provides a systematic way to investigate the propagation of porosity uncertainty through the estimate of original and recoverable oil in place (OOIP and ROIP) of an original reservoir model. To perform the intended investigation, a multiple-point geostatistics (MPG) modeling methodology is adopted. Then, a synthetic fluvial reservoir model (original reservoir) is generated utilizing a given dataset. Next, different realizations of the original data are generated through Monte Carlo Simulations (MCS) based on quantified uncertainty values in the porosity. Following that, the original reservoir is remodeled using the generated realizations. Finally, the OOIP and ROIP of the new reservoir models and their percentage errors with respect to those of the original reservoir are computed and analyzed. The results show that the estimates of the OOIP and ROIP move from overestimation to underestimation as the uncertainty increases from 0.01 to 0.15. This clearly shows that the uncertainty of porosity has a great impact on the model outputs. Thus, it should be given great care to facilitate more accurate reservoir modeling applications.