Prediction of reservoir properties using inverse rock physics modelling in the Kanywataba Exploration Area, Albertine Graben

Abstract

In this study we used the concept of inverse rock physics modelling to analyse reservoir properties of the Kanywataba Exploration Area, with a focus on their lateral distribution away from the Kanywataba well. The procedure employed rock physics models calibrated for the basin constrained by seismic inversion data, where non-uniqueness and data error propagation issues were also taken into account. Both seismic and well log datasets were used in the data calibration. The procedures enabled us to obtain the most likely estimate mean, weighted mean and posterior mean of the reservoir properties. We obtained a good match between measured and modelled porosity values. Any misfit between the observed and predicted lithology was mainly attributed to uncertainties in defining the correct mineral properties. The integrated approach revealed that high porosities correlate with low clay volumes and, furthermore, indicated two distinct reservoir units in the basin, which were interpreted as the Oluka and Kakara Formations. Fluid saturation data were less successfully predicted but this was most probably due to a result of lack of real saturation logs for use in the calibration of the rock physics model; instead, predicted saturation logs based on Archie's law were used in the calibration process. This analysis is first of its kind in this basin and therefore exhibits a high level of novelty in the determination of reservoir properties in this area.