Low-Permeability Laboratory Measurements by Nonsteady-State and Conventional Methods
Abstract Evaluation of tight gas reservoirs requires an accurate but rapid and practical method to determine permeability. Such a method is presented for determining both specific and effective gas permeability in the 0.0001- to 0.35-md range for plug-size core samples. Equipment is described that meets the requirements for calculation of nonsteady-state flow and incorporates the capability of simulation, high net overburden pressures by either hydrostatic or triaxial confining pressures with ease of operation. The time required to collect data and calculate Klinkenberg permeability is typically less than 6 minutes per sample. Values normally differ by less than + 5 % from those obtained by steady-state methods. This method is well suited for routine laboratory determinations of permeability on samples from reservoirs with tight or very low gas permeability. Effective gas permeabilities on samples containing nearly irreducible water saturations and the water permeabilities presented are closer to the Klinkenberg permeability values in low-permeability samples than most previously reported. Introduction Substantial price incentives exist in the U.S. to make it attractive for producers to recover gas from tight formations that are less than 15,000 ft 14572 mi deep and have no more than 0.1 md in-situ permeability. This incentive, plus the need for a rapid method to obtain accurate laboratory data on low-permeability samples for well completion and gas reservoir engineering, made it desirable to develop the subject equipment and test method. Various methods used to determine limiting permeability were investigated. The conventional method of determining three specific gas permeabilities and using the Klinkenberg relation to determine a limiting permeability is laborious. Methods involving numerical solutions of one-dimensional (ID) gas-flow equations such as those proposed by Aronofsky and Jenkins and Bruce et al. involve solutions by finite differences. This approach required long calculation times, which made it too cumbersome. Methods such as those proposed by Brace et al. and Walls et al. require pore pressures of the sample to be brought to equilibrium at values close to the reservoir pressure before analysis of the sample, and thus excessive time is required in approaching equilibrium. Jones suggested accounting for the non steady mass flow through the sample during an upstream pressure drawdown test. Such an approach may be used with relatively low mean pore pressures ( - 100 psig ( 690 kPa). The number of calculations was not large, while the reported accuracy was good. The method described in this paper accounts for the nonsteady mass flow through a sample during a downstream pressure-buildup test. The downstream approach allows the smallest possible downstream volumes to be used and ensures flow through the sample. These small downstream volumes allow the detection of very small flow rates in a relatively short time. SPEJ P. 928^