History Matching and Predicting GOR Behavior Predicting GOR Behavior With Analytical Functions
A study initiated in 1982 examined gas production mechanisms in Prudhoe Bay. The results we report represent only a part of this effort Prudhoe Bay. The results we report represent only a part of this effort and deal mainly with developing analytical techniques for GOR prediction. Traditionally, a well's producing GOR is history matched and prediction. Traditionally, a well's producing GOR is history matched and subsequently predicted by numerical reservoir simulation. We demonstrate how GOR history in gas underrunning and coning wells may be described by rate- time-, and cumulative-offtake-dependent analytical functions. The coefficients in the equations are determined by a nonlinear parametric least-squares regression approach. For each well, after an empirical GOR match is achieved, the analytical functions are used for short-term GOR forecasting as an integral part of a comprehensive production availability forecast (PAF) computer model. We conclude that the parametric studies have produced viable analytical models for the matching parametric studies have produced viable analytical models for the matching and predicting of GOR in wells that have a history of high GOR.
In oil reservoirs with an initial gas cap and complex stratification such as Prudhoe Bay 1 (Appendix A), the ability to identify the various gas influx mechanisms and the capability of predicting the dynamic response in GOR owing to a selected offtake strategy are important concerns of reservoir management. Gas production at Prudhoe Bay has increased steadily since field startup in 1977. With time, the distribution of wells producing gas assumed a discernible pattern and appeared concentrated in certain areas. Different gas production mechanisms appeared to dominate in specific areas production mechanisms appeared to dominate in specific areas of the field. The predominant gas influx mechanisms were identified by qualitative interpretations of the production history characteristics, openhole and repeat compensated neutron logs (CNL), shale correlations, gas breakthrough times, GOR response to rate changes, repeat formation tests (RFT), pressure buildup results, and the current understanding of regional pressure variations. The principal technique to detect gas saturation changes at principal technique to detect gas saturation changes at Prudhoe Bay, however, has been the use of baseline (initial) Prudhoe Bay, however, has been the use of baseline (initial) and repeat CNL measurements. Fig. 1 shows a typical well in which gas/oil contact (GOC) movement has occurred between baseline and repeat measurement, Gas underrunning shale complexes, evolution of solution gas, and coning were the main reasons wells went to high producing GOR'S. These three gassing mechanisms may occur independently or in combinations simultaneously throughout the field (Figs. 1- and 3). The first objective of this study was, therefore, to determine the predominant mechanisms that control excess gas predominant mechanisms that control excess gas production on a well-by-well basis. These classifications were production on a well-by-well basis. These classifications were used to group wells of similar type for the parametric studies to be described. As a final stage, a proper set of analytical equations describing the wells' GOR behavior is incorporated into the PAF model with a variety of other features. The PAF may then be run to investigate different oil offtake scenarios to arrive at an optimum short-term reservoir management strategy.
About 50% of the active oil producers exhibited excess gas production as of Jan.1983. The split of the various mechanisms is shown in Fig. 4 (Table 1 explains the notation). As shown in Fig. 2, 59% of the high-GOR wells are caused by gas underrunning shale complexes or tonguing (Romeo shales not shown), 25% are caused by the evolution of solution gas, and only 9% are classified as coning wells. The remaining high-GOR wells produced excess gas because of other mechanisms (Table 1). Note that these percentages do not reflect actual production volumes attained from each mechanism. Table 2 shows the petrophysical and geological factors influencing the various gassing mechanisms.