engleski

Indicator vs. Gaussian geostatistical methods in sandstone reservoirs – Case study from the Sava Depression, Croatia

Geostatistical methods are very often used in reservoir modeling, but the hardest question is which method should be used to obtain the most realistic solution. Many geologists prefer determinism despite of the fact that just one solution to the problem can be obtained. Simulations in contrary provide as many realizations/solutions as one would want, so with 5-10 realizations a quick inside in the uncertainties of mapped area can be obtained. Lower Pontian sandstone reservoirs of the Kloštar Field (the Sava Depression) are mapped by deterministic and stochastic methods. Input variables were porosity, thickness and depth. Applied deterministic methods were Ordinary Kriging (OK) and Indicator Kriging (IK) while applied stochastic methods were Sequential Gaussian Simulations (SGS) and Sequential indicator simulations (SIS). Presented results showed that in some cases stochastic methods gave better solutions, while in others deterministic ones were more appropriate. Results also showed that the most important question is not which approach to use, rather what we want to get with mapped variables? If we want to acquire some knowledge about depositional environments and lithofacies changes by porosity mapping, best results were obtained by IK and SIS. On the other hand SGS showed better solutions structural mapping, but also in mapping the original petrophysical properties inside the reservoir.

simulations, determinism, sandstones, Kloštar Field, Sava Depression

Publikacija sadrži odabrane prezentacije sa skupa The 15th Hungarian and 3rd Hungarian Croatian Geomathematical Convent, održanog svibnju 2011., Mórahalom, Mađarska.