Optimizing a Complex, Uneven and Low-Resolution Reservoir While Using Dynamic Dataset for Structure Definition

Abstract

It is well established that the oil & gas industry has long surpassed its plateau for large discoveries. Thus, many companies have shifted their focus to cost-constrained policies for hydrocarbon discovery, making it difficult to have a strong sub-surface definition. Add the structural uncertainties and complexities of new discoveries in this scenario, the optimized field development becomes a real challenge. This study presents the development strategy established for a reservoir, with low seismic resolution, using the integrated dynamic data to define the reservoir structure and optimize its recovery. This paper focuses on a relatively new discovered formation in one of the oldest gas giants in Pakistan. The productive sandstone units are in beds with thicknesses ranging from 10-50m, separated by mudstone intervals. The low seismic resolution has posed a major challenge in finding the sweet spots for hydrocarbons above Gas-Water-Contact (GWC)-resulting in 60% well failures. Therefore, a workflow was developed to analyze various dynamic datasets in conjunction with the re-interpretation of seismic to delineate the reservoir structure. This included re-interpretation of MDT data and formation gradients, core re-evaluation, critical analysis of reservoir pressure variation, well failure analysis, thickness maps and PVT properties. The first four wells drilled in this reservoir had two successes, both in the Western compartment. However, sudden water production loaded up these wells only after 123 BCF production, which was a lower recovery as compared to GIIP estimates. After further geological evaluation and 3D seismic re-acquisition, more wells were drilled – revealing another deposition with a different GWC, but only 1 well was successfully completed as a producer, adding around 38% more reserve, while the others were again unsuccessful due to high structural uncertainty. This led to the development of a detailed algorithm for integrating dynamic data set with seismic re-interpretation and thickness mapping with the help of which two more wells were drilled and added around 22 % more reserves to the current mix. The dynamic data of these wells have now been further evaluated which revealed that the two compartments are in fact in communication with each other despite having a 60m difference in their GWCs. Finally, two more wells are now planned which will add around 10-20 % more recoverable volumes, giving an overall EUR of ~80% from these compartments. The main achievement of this workflow is a robust algorithm to integrate the dynamic data with geological interpretation to delineate a low-resolution reservoir since the seismic interpretation cannot be solely relied upon for developing such reservoirs. The paper also illustrates the robust engineering models and data analyses in a more systematic manner to ensure optimum locations for future wells to access, the otherwise, undrained locations.