Compositional Simulation for Effective Reservoir Management: The Brady South Weber Pressure-Maintenance Project

Summary

This paper discusses the problems encountered, the logic used to address their solutions, and the compositional models developed to history match a partial pressure-maintenance project in a rich retrograde gas reservoir. Residue gas from the tailgate of a lean-oil refrigerated plant is being reinjected to optimize condensate recovery.

Since the discovery of the South Weber reservoir in Nov. 1972, several compositional modeling studies were completed to determine the proper plan of development and reservoir management. However, these studies were carried out without the benefit of any producing history. In late 1975 the reservoir was placed on production. A study was begun in March 1978 to develop mathematical models that were consistent in matching observed performance (at this time about 21 % of the original gas in place had been produced). This paper concentrates on that study. To utilize all available data in the history match, compositional models were developed to simulate reservoir performance, tubing flow behavior, and performance of the primary and test separators, the stabilizer, and the plant. These models then were used (1) to update rate-time forecasts under current operations. (2) to investigate the feasibility of infill drilling or changing the flood pattern, and (3) to determine the benefits of changing to nitrogen injection for partial pressure maintenance.

The results of this study demonstrate utility of the compositional simulator to supplement and reinforce measured production data and to test various hypotheses efficiently that might explain observed performance. Furthermore. the study demonstrates that the simulator can be used to predict the sensitivity of the project performance to different reservoir characteristics or various operating scenarios.

Introduction

The Brady South Weber reservoir is in southern Wyoming, 30 miles southeast of Rock Springs. It was discovered in Nov. 1972 with a successful test of the Weber in the Brady Deep Unit Well 1 W. Reservoir fluid analyses of recombined surface samples indicated that the fluid was a very rich retrograde gas. The gas is sour, having about 1.2 % H2S and 34 % CO2. Because of the complex fluid, the formation depth, and the well cost, mathematical modeling has been used throughout the project as a key to effective reservoir management. Table 1 summarizes objectives and major conclusions of previous simulation studies. All these studies were undertaken without the benefit of any producing history in an attempt to identify an optimal exploitation plan. With completion of a lean-oil refrigerated plant, production began in Nov. 1975. The field-wide rate from seven producing wells averaged 34 MMcf/D of primary separator gas and 6.600 B/D of stabilized condensate. Residue gas from the South Weber, the South Nugget, and more recently the North Weber is being reinjected into the South Weber reservoir. An average of 37 MMcf/D of residue gas has been reinjected into six peripheral wells. Cumulative production to Dec. 1979 mas 10 MMbbl, or about 21 % of the original condensate in place. Gas production have been less than predicted in earlier studies. However, six of the seven producing wells have not been fracture stimulated. The initial GOR was lower than reservoir fluid analyses indicated. Increasing GOR's (presumably the result of residue gas breakthrough) were premature at some wells and late at others. On a field-wide basis of GOR vs. cumulative production, the reservoir performed better than predicted. Water production at Wells 1OW and 16W, in the southwest portion of the reservoir. was higher than expected.

JPT

P. 635^