RESERVOIR CHARACTERIZATION AND MODELING OF AN UPPER MIOCENE DEEPWATER FAN RESERVOIR, GENDALO FIELD, KUTAI BASIN, OFFSHORE EAST KALIMANTAN

摘要

The Gendalo Field is the largest of several deep-water gas discoveries in offshore East Kalimantan. The primary reservoir, the Upper Miocene Gendalo 1020, contains thin-bedded turbidites [median sand bed thickness is less than 3 inches/7.5 cm; average sand bed thickness is approximately 15 cm] deposited at the base-of-slope as a largely unconfined fan. Three internal units were mapped based on 3D seismic data and four wells.An object-based facies modeling and experimental design approach were used to: a) build static geocellular models, b) generate gas-in-place distributions, c) understand the key uncertainties, and d) provide models for dynamic flow simulation. Static and dynamic models for this reservoir incorporate lobe and channel geobody sizes from an analog Pleistocene fan located 170 km to the northeast. Three depositional styles were modeled: (1) 90% lobes with 10% channels, (2) 70% lobes with 30% channels, and (3) 100% channels. The models were populated with petrophysical properties from well data including net-to-gross, porosity, permeability, water saturation, and gas formation volume factor. Net-to-gross and structure were found to be the most important parameters in determining the gas-in-place distribution.To capture the uncertainty of reservoir connectivity, three levels (or scales) of compartmentalization were used. Each level of compartmentalization reduces cell connectivity using an independent transmissibility multiplier that reduces fluid flow. The three levels of compartmentalization are: (1)cell to cell lamination effect, (2) geobodies (channel and splay elements), and (3) possible larger-scale compartmentalization based on seismic discontinuities. The calibrated full-field models were consistent with Gendalo well drill-stem test data.Transmissibility multipliers were found to be an important and effective modeling tool for development optimization and production forecasting. Experimental design work suggests that in addition to net-to-gross and structure, all three levels of compartmentalization are critical factors for reserves uncertainty, whereas the depositional style (channel or lobe) is not an important parameter to reserve uncertainty.