--> Construction of stochastic model for Ordovician karsted carbonates reservoirs of Tahe oil field, China: dimensional and geometric implications and structural considerations

AAPG Asia Pacific Region, The 4th AAPG/EAGE/MGS Myanmar Oil and Gas Conference:
Myanmar: A Global Oil and Gas Hotspot: Unleashing the Petroleum Systems Potential

Datapages, Inc.Print this page

Construction of stochastic model for Ordovician karsted carbonates reservoirs of Tahe oil field, China: dimensional and geometric implications and structural considerations


The paleokarst reservoirs from Ordovician Yingshan Formation represent relevant storages for oil and gas production in Tahe oil field, western China. Since the decade of 2000s, the oil exploitation of these very deep reservoirs located at north of prolific Tarim Basin, have been very significant. The complexity for karstic patterns modeling, essentially of caves or sinkholes, motive the needing of apply different techniques to honor the geological model. Hence, this research aims the construction of a 3D model using the acoustic impedance cube and traces seismic as inputs in stratigraphic grid, to propagate the karst facies, besides of physical properties. Highlighting the tectonic features occurred during the Caledonian and Hercynian orogeny that produced the exhumation of platform and configured its current position, and oriented the karstification process. According to Fey et al. (2016), those karst fillings are compositionally heterogeneous, siliciclastic and carbonates mixed. For this case, 61 faults (between thrust, reverse and sinistral strike-slip faults) and 3 stratigraphic horizons at depth around from an area of 245 Km² were interpreted and validated. The displacement and erosion of thrust (main structure) generated by compressive episode during the Caledonian orogeny it honored in grid, thus as, the strike–slip and reverse faults formed at Hercynian orogeny, even the displacements of those are less. The physical properties values and discrete facies data of paleokarst and host rock obtained from petrophysical evaluation of 163 wells, as well as from previous sedimentological analysis. The effective porosity and permeability average values calculated are around of 8% and 3 mD, respectively. Moreover, the 3D seismic data (32 bit) of frequency content from 0 to 80 Hz a predominant frequency of 30Hz at 6,700 m. From the seismic data applied the relative acoustic impedance seismic attribute using the values under -40.500 km/s*g/cm³ as 3D proportion data in model, to discretize patterns with similar characteristic to paleokarst, with tubular and roughly circular forms randomly distributed into carbonate host rock associated probably to run-off or vadose zone. To better facies upscaling and low connectivity between patterns, it created very thin a grid geological to cover karstic patterns size <2 m. The upscaling of karst facies in model is around 90% effectiveness. Additional, it generated a seismic traces map classified in 8 seismic facies between unconformities horizons of 474 - 476 ma (interval karsted), associating the facies seismic 6 and 7 to paleokarst partners. The seismic attributes and stratigraphic well sections allow identifying paleocave chambers with similar dimension and geometries of surface rivers and streams channels, subsequently this data is transferred and modeled in grid. The paleocave chambers size in model fluctuate between 40 - 50 m of wide and 20-30 m of high, being thinner in branches and sinkholes, 4x2 m. The sinkhole geometry is randomly distributed and sometimes isolated. Using the facies seismic model can recognize two main karstic systems related to run-off zone orientated at northwestern and southeastern, being it also honored in model. Appling the kriging with external drift statistical process in porosity and permeability values improve effectively the propagation of these properties in karst zone. The geological integration demonstrates the paleocaves hold on the same trends of main faults. Finally, the stochastic model shows partially very good continuity of paleocave chambers according to depth, besides support the geological data about the paleokarst distribution and dimension - geometry descripted in outcrops by Wang (2017) in Yingshan Formation. In addition, the model exhibits paleocaves not drilled yet which could be considered as reexploratory prospects for oil and gas industry.