Using Borehole Image Logs to Refine Reservoir Architecture for Carbonate Reservoir Models, Korolev Field, Republic of Kazakhstan

Abstract The architecture of carbonate reservoirs can be complex due to depositional profiles with variable, locally steep dips, and to changes in stratal geometries through time resulting from the interplay of accommodation and carbonate sediment production. Capturing this architecture in reservoir models is important for constraining the geometry of flow units and baffles and for distributing reservoir quality. Where seismic interpretation is uncertain or core is limited, borehole image logs can provide valuable information to better constrain architecture. This study focuses on Korolev Oil Field, a late Paleozoic isolated carbonate platform in the Pricaspian Basin, Republic of Kazakhstan, and illustrates how borehole image logs can be used to refine the interpretation of horizon tops and facies boundaries for input into reservoir models. Bed dips interpreted from image logs provided useful quantitative information, and trends in dip datasets were revealed using plots of cumulative dip azimuth and magnitude. Integration of bed dip data with image log facies, seismic mapping, core description, biostratigraphy, and conventional logs helped identify the orientation of depositional slopes, the boundary between platform and slope facies, and the position of differential compaction related to antecedent platform margins. This updated reservoir architecture has changed the interpretation of facies (slope vs. platform) and horizons to which productive zones in some wells are attributed and has therefore impacted the distribution of reservoir quality in the model. This study underscores the value of commonly under-utilized bed dip interpretations for constraining carbonate reservoir architecture. Integration of bed dip trends from borehole image logs with other available datasets improved the selection of well tops, the interpretation of surfaces, and the position of facies boundaries, all of which are critical static reservoir model inputs.