Naslov
Improvements in reservoir characterization applying geostatistical modelling (estimation & stochastic simulations vs. standard interpolation methods), Case study from Croatia

Autori
Smoljanović, Saša ; Malvić, Tomislav

Izvornik
Nafta : exploration, production, processing, petrochemistry (0027-755X) 56 (2005), 2; 57-63

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
reservoir characterization; geostatistical modelling; Kriging; stochastic simulation

Sažetak
There are several larger gas-condensate fields in Croatia, important for present and future hydrocarbon production. These fields are characterised with complex lithology and hardly predicted petrophysical properties of reservoir. Analyzed field, Stari Gradac-Barcs-Nyugat, is characterized with four reservoir lithofacies and unique gas/water contact. Input set included 16 mean porosity values (well locations) values, particularly for each lithofacies. Improvements in reservoir characterization were performed using geostatistical approach (Kriging and conditional stochastic simulations) and given results were compared with "Inverse Distance Weighting" (IDW) maps. Kriging was selected as interpolation method with the least estimation risk, also depending variogram model (here spherical theoretical model was used). Away from data, and regarding variogram model, Kriging does not take any risk, i.e. estimation is replaced with simple mean value. Quality of Kriging porosity maps was checked in each lithofacies also calculating MSE value too. Conditional stochastic simulations were used as tool for better estimation of cell values away from control points (wells) and obtaining 100 "equally-valuable" porosity maps. Conditional simulation maps were slightly noisy, and Kriging map was fairly smoothed and elongated in direction of structure. New porosity maps led to new OGIP (Original Gas In Place) values. Old value calculated from IDW maps was 19, 918, 304 m3. OGIP based on Kriging maps was decreased on 17, 648, 378 m3. Using of conditional stochastic simulations yielded to lower OGIP in range min=16, 329, 468 m3, median=16, 538, 900 m3 and max=17, 113, 130 m3. More precise porosity maps (based on geostatistical technique) led to more precise OGIP calculation.

Izvorni jezik
Engleski

Znanstvena područja
Geologija

POVEZANOST RADA