RELIABILITY OF HORIZONTAL WELL PERFORMANCE ON A FIELD-SCALE THROUGH AUTOMATIC HISTORY MATCHING

An efficient procedure for the determination of the error bounds (95% confidence intervals) of forecasted horizontal well production rates i s presented. In particular, we advocate the use of automatic history m atching (based on non-linear regression) to arrive at the best reservo ir parameter values (typically porosities and permeabilities) for a gi ven reservoir structure (i.e., for a given grid block representation). If th Gauss-Newton method is used for the parameter search, the sensi tivity coefficients required for the estimation of the standard error of the parameters and the well production rates are computed as part o f the parameter search algorithm. Using typical techniques from statis tical analysis (and the statistical properties of multivariable non-li near regression estimates), one can readily compute the confidence int ervals of the predicted production rates. These confidence intervals a round the expected reservoir behaviour incorporate the uncertainty in all reservoir parameters. As a result, a realistic measure of the unce rtainty in the model predictions of future reservoir performance is ob tained. The proposed procedure is illustrated with a case study.