In order to demonstrate the importance of using observed pressure data for history matching an early-stage waterflooding process, a novel workflow for experimental design, assisted history matching and probabilistic analysis, was applied to analyze laboratory displacement test measurements. In order to achieve matches between experimental and model data, an evolutionary global optimization technique was used and three different strategies were investigated: matching pressure data only, matching production data only and matching both pressure and production data. For the purpose of generating model displacement data, a one dimensional, two-phase simulator was run.

Tornado plots on the matched data reflect that calculated model pressure values appear to be more sensitive to changes in the dynamic calibration parameters (uncertainty parameters) than calculated cumulative fluid values. This implies that use of the pressure drop information enhances the ease with which acceptable history matches can be obtained. Additionally, the results obtained while history matching pressure only were better than those from matching volumetric data only. The results were validated by crosschecking the probabilistic forecasting with real postmortem observed values and by comparing calibrated uncertainty parameters with core petrophysical properties (laboratory analysis).

You can access this article if you purchase or spend a download.