Uncertainty Assessment in Geologic Modeling and Sensitivity Analysis of Static and Dynamic Models in Umiat: A Frozen Shallow Oil Accumulation in National Petroleum Reserve of Alaska

摘要

Current high oil price and availability of new technologies allow re-evaluation of oil resources previously considered uneconomic. Umiat oil field is one such resource: a unique, shallow (275?1055 feet), low-pressure (200?400 psia) reservoir within the permafrost zone with no initial gas cap located north of the Arctic Circle, 80 miles west of Trans Alaska Pipeline System (TAPS) with an estimated 1.5 billion barrels of oil-in-place. A static model was built based on reinterpretation of original log and core data and seismic information. A permeability anisotropy ratio of 0.45 was incorporated into the geologic model. A Monte Carlo simulation was conducted to estimate the different degrees of uncertainty in the original oil in place (OOIP) estimates. To cover the wide permeability range (0-500 md), three sand groups (rock types) were defined and assigned appropriate capillary pressure and relative permeability curves. These were included in the dynamic model along with measured PVT data and gas-oil relative permeabilities in the presence of ice to evaluate the performance of immiscible gas injection using a multilateral wagon wheel well pattern with horizontal well length of 1500 ft. The simulation results show that with 50 years of gas injection, recovery factors for the base case (400 psia injection pressure) and a case with 600 psia injection pressure are 12% and 15%, respectively, keeping other parameters constant. Those recovery numbers reduce by 18% and 6% when producing GOR is restricted to 5,000 scf/STB and 10,000 scf/STB, respectively. The potential effect of natural fractures was modelled by considering the effect of permeability anisotropy (K_v/K_h).. Simulation results indicate that lower anisotropy ratios will reduce oil recovery, probably due to inhibition of downward gas movement. The results obtained in this study contribute to the understanding of uncertainties in resource estimation and evaluating ranges of oil recovery in reservoir modeling. Despite limited data and lack of production history to tune the model, the results demonstrate that the proposed development plan bears a high degree of uncertainty and risk. These findings strongly encourage the operator to include in development plan strategies to reduce the risk by enhancing the quantity and quality of simulation input data.