Comparing Single-Point Pressure Buildup Data With Reservoir Simulator Results

Abstract

Matching simulator results with historical reservoir data is an important part of a numerical reservoir simulation study. A good match is generally interpreted as indicating the validity of reservoir data used in the simulation and as justification for predictions of reservoir performance. One common matching technique is to compare simulator-calculated pressures with observed pressure data. Ideally, pressure data from observation wells (never produced or injected into) would be used for such matching. Unfortunately, such data are rarely available. Generally, most reservoir pressure data come from relatively short-term pressure data come from relatively short-term pressure buildup tests. Occasionally, only one pressure is pressure buildup tests. Occasionally, only one pressure is reported, even though the well may have been shut in for days or weeks. This note presents a method for comparing this kind of pressure data with simulator calculations. To keep simulation costs reasonable, reservoir simulation studies usually use 30- to 90-day time steps. Since pressure buildup tests normally last only a few days, it generally is not practical to include them in simulation runs. Thus, simulator results do not directly provide pressures for comparison with buildup-test pressure. Van Poollen et al. present a method for comparing observed pressure buildup data with simulator-calculated node pressure (not simulator calculated well pressure) for the time step during which the buildup test was run. This node pressure should agree with observed buildup pressure at a buildup time of ...........(1) where = buildup time at which buildup pressureshould be compared with nodepressure, hours= porosity, fraction= total system compressibility, psi= grid size, ft= total system mobility near the well, md/cp. Total compressibility, ct, and total mobility (k/u)t, are defined by Matthews and Russell." Eq.1 applies to a well in the center of a square node and assumes that the semilog buildup plot (shut-in pressure, Pws, vs log of shut-in time, log t) is linear at t*. If this semilog plot is not linear at t*, the linear portion of the plot should be extrapolated to t*. The pressure att*, taken from the semilog straight line (extrapolated if necessary) should be compared with simulator-calculated node pressure. If the complete buildup curve is not available for instance, if only one pressure is observed - Eq. 1 cannot be used. In this situation, however, we can calculate a pressure to compare with simulator-calculated node pressure. .............(2) P. 711