Estimation of formation hydraulic properties accounting for pre-test injection or production operations

Abstract

We propose to use regular monitoring data from a production or injection well for estimating the formation properties around the wellbore without interrupting the operations. Thus, instead of shutting-in the well for a substantial time period, we propose to select a portion of the pumping data over a certain time interval and then derive our conclusions from these data. A distinctive feature of our analysis is that we introduce an auxiliary parameter to account for the possible after-effects of pumping that preceded the test interval and, consequently, the non-uniform initial pressure conditions. We demonstrate that those effects influence not only the analysis of regular operations data, but also the analysis of a traditional pressure drawdown or pressure buildup test with a prior shut-in period. We show that phenomena usually attributed to wellbore storage or skin effects can be at least partially interpreted through the parameter we introduce. Unlike some traditional methods, our analysis utilizes almost the entire test time period for curve-fitting. It turns out that it produces good data matching even if the test period is short and the frequency of measurements is low. Another distinctive feature of the present approach is that the parameter estimation problem is reduced to a combination of quadratic criterion minimization and a search for the minimum of a one-variable function. Because we can obtain the solution to the quadratic problem analytically, we significantly simplify the problem and dramatically reduce the amount of computations required.