Application of multivariate statistical methods for characterizing heterolithic reservoirs based on wireline logs - example from the Carpathian Foredeep Basin (Middle Miocene, SE Poland)

摘要

Principal Components Analysis (PCA) and Cluster Analysis (CA) were applied for well log data derived from heterolithic intervals drilled in two boreholes (Mrowla-1 and Cierpisz-2) in the Miocene fill of the Carpathian Foredeep. Both boreholes penetrated similar basement elevations conductive for structural trapping of hydrocarbons in an overlying thin-bedded heterolithic reservoir, which produces gas in commercial quantities in one borehole. The PCA was used to reduce data space preserving sufficient amounts of parameters for a differentiation between thin layers of sandstones and mudstones and between gas- and water-saturated horizons. In both boreholes, the number of logs was reduced to four significant principal components (PCs). Differences between gas-saturated and water-saturated layers were found. CA was used for the classification and grouping of data according to natural petrophysical features of the analysed rocks. The group corresponding to gas-saturated zones was found in the Cierpisz-2 borehole. It is concluded that PCA and CA can provide useful information for a more reliable identification of gas-saturated horizons.