Linking Well-Test Interpretations to Full Field Simulations

Abstract

The objective of this study was to investigate a workflow where well test data could be used more effectively in history matching of full-field reservoir simulation models and also in situations where existing simulation models could be used in well test interpretation. The need to effectively use information available from well test analysis in full-field simulation has long been recognized. However, only limited benefit could be obtained by reconciliation of the analytical well test model with the numerical full-field model. We present a more complete approach where a more integrated approach using a common model is advocated.. The benefits of such a workflow can be summarized as follows: Existing commercial well test and near wellbore modeling software packages, were used to carry out this task. These packages provide engineering interfaces to the simulator that make their respective workflows easy. However, there were a number of situations where manual intervention and workarounds were necessary. We would like to propose an easy to use software that can provide a complete workflow for regular use. Introduction The link between reservoir simulation and well testing has been growing steadily for several years. Raghavan, et al1. have reviewed its development in their work. They found that there are three broad areas where there are growing industry interests: The first has been the subject of interest for some time. Most well testing software with numerical option focus on this aspect. There are a number of papers that has been published in the last few years 2,3,4 Currently, we are interested in a consistent and integrated workflow for the latter two areas. Corbett et. al.5. used a numerical model of braided fluvial reservoirs to calculate well test responses. A "geo-skin concept" was developed from these data to be used in a full-field model. He and Chambers6 presented a methodology to verify and update geostatistically-based full-field reservoir simulation models using numerical well testing. The authors pointed out the difficulty in history matching during a simulation study due to the large uncertainties associated with reservoir properties in flow simulation models particularly for permeability. This is because of limited core sampling from the reservoir and difference in scale between grid block and core permeability. Well test data is ideal for bridging the gap between core and grid block permeabilities because it samples the reservoir on the scale of the grid block size. The authors recognized that well tests would not fully cover the full field model. Therefore, they found it useful to calculate property multipliers around each well and then interpolate multipliers to untested areas in the reservoir.

Reservoir Characterization and Flow Simulation of a Low-Permeability Gas Reservoir: An Integrated Approach for Modelling the Tommy Lakes Gas Field

“…The deviation between the core/logderived permeability and the in-situ permeability can be two orders of magnitude for a tight gas reservoir. Previous literature has discussed numerical well testing and the linking to the full-field reservoir model (Settari 1990;Puchyr 1991;He and Chambers 1999;Mlacnik and Heinemann 2003;Kamal et al 2005;Al-Thawad et al 2007). An integrated workflow developed in this paper that explains the field-based corner-point grid model is based on matching the well-test buildup data:…”