Devising an Optimum depletion plan for the fourth largest gas field in the United States required a detailed volumetric study for each of the twelve Katy reservoirs. All of the reservoirs have different water drive strengths, so the erect of producing rate on ultimate recovery is different for each reservoir. This is a key consideration in total-field optimization. Introduction The Katy gas field is presently under a 20-year gas sales contract, begun in 1967, and is in the last stages of gas cycling, begun in 1943. The field is located 35 miles west of downtown Houston, covers an areal extent of 50 sq miles, and comprises 12 reservoirs in the Cockfield formation, located at depths ranging from 6,000 to 7,400 ft subsea. Each reservoir has a different water-drive strength, depending on rock permeability, sand thickness, ratio of aquifer radius permeability, sand thickness, ratio of aquifer radius to reservoir radius, and reservoir geometry in general. Thus, the effect of the producing rate on ultimate recovery is different for each reservoir, and is a key consideration for total field optimization. Reservoirs with weak water drives exhibit high recoveries as a result of the combined action of pressure depletion and water influx over a long production life. Reservoirs with strong water drives are more sensitive to the producing rate, and should be depleted very quickly in order to maximize ultimate recovery. To determine the optimum production and investment schedules for the entire Katy gas field it was first necessary to develop mathematical models for each reservoir from pressure-production performance over the past 27 years. Anomalies caused by interreservoir communication and pressure interference from other fields producing from the Cockfield formation were considered in developing these models so that reasonably accurate predictions could be made regarding future reservoir pressure and well deliverability. Water influx data for reservoirs that have already undergone a significant amount of blowdown was determined by volumetric balance calculations. For the reservoirs that have experienced little or no depletion, however, these factors had to be estimated. After reasonable reservoir models were determined, it was necessary to impose certain constraints on the long-range blowdown scheduling. Two of the reservoirs have an associated thin rim-type oil column that must be essentially depleted before full-scale gas blowdown can occur. In addition, certain reservoirs are separated by very thin shale barriers and therefore must be depleted simultaneously to prevent large pressure differentials across these thin barriers. This precaution minimizes potential interreservoir communication and thus conserves potential interreservoir communication and thus conserves reservoir energy. Since most of the existing wellbores penetrate the productive areas of all the Katy reservoirs, it was productive areas of all the Katy reservoirs, it was necessary to develop models for estimating, the time at which wells water out and are available for recompletion to other zones. Wells now used for gas cycling in four of the upper reservoirs are also made available for recompletion as they drop out of the cycling program. Other factors to be considered for total field optimization are liquid recovery factors for each reservoir and development of sufficient deliverability to meet contract sales rates and peaking requirements. peaking requirements. JPT P. 145