Enhancing gas storage wells deliverability using intelligent systems

Gas storage fields have numerous wells that are used for both injection during low demand periods and withdrawal during high demand periods. As these wells age, their deliverability declines due to several factors. Stimulation treatments (hydraulic fracturing of the formation) are routinely used in gas industry to improve gas well productivity. This study was conducted on a large natural gas storage field located in Northeastern Ohio. The formation is tight gas sandstone and is called the Clinton Sand. All of the storage wells were initially stimulated by hydraulic fracturing. Restimulation is considered a last resort method of deliverability enhancement in this storage field. However, some wells are selected to be restimulated each year based on maintenance history, past fracture response, years since previous stimulation and overall deliverability potential. Since 1970, an average of twenty-five wells have been refractured (restimulated) each year for a total of around 600 refracturing treatments. Since most wells in the field have been refractured (restimulated), some up to three times, the need for post stimulation well performance estimates and optimal fracture design is very important to maximize deliverability gains. The experience with the Clinton Sandstone indicates that hydraulic fractures grow vertically out of the zone, regardless of rate and fluid viscosity. Therefore, it appears critical to use high proppant concentrations in a viscous fluid to create a conductive fracture in the pay interval. Treatment designs for the storage field currently include a 25 to 30 pound linear gel with maximum sand concentrations from 3 to 4 pounds per gallon (ppg) (McVay et al., 1994). Several well testing methods are available for predicting hydraulically fractured well performance including type curve matching and computer simulation (Millheim and Cichowicz, 1968; Gringarten et al., 1975; Cinco-Ley et al., 1978; Agarwal et al., 1979; Hopkins and Gatens, 1991). In addition, twoand three-dimensional computer simulators are frequently used for fracture design. Use of these tools, however, requires access to several types of reservoir data. Reservoir data necessary for hydraulic fracture simulation include porosity, permeability, thickness and stress profiles of the formation. Experience has shown that given the aforementioned data and assuming availability of a good geologic and structural definition of the reservoir, hydraulic fracturing simulators can predict the outcome of the hydraulic fracturing process with reasonable accuracy.