Summary Effective permeability to gas with various degrees of brine saturation has been measured in the laboratory for several very tight sandstones from the Spirit River formation of Alberta, Canada. Gas permeability as low as 20 × 10 -9 darcy was measured successfully with a pulse-decay permeameter with nitrogen as the mobile fluid. Results show that gas permeability depends very strongly on the degree of saturation, with 40% saturation causing permeability to decrease an order of magnitude relative to the dry rock. Therefore, accurate knowledge of in-situ saturations is crucial before natural-gas production rates can be estimated in these formations. The experiments also show that confining pressure causes significant permeability reduction in these sandstones. A Spirit River sample recovered from 2133.6 m and subjected to in-situ levels of pore pressure and confining pressure shows a seven fold reduction of gas permeability. It also was found that as brine saturation increased, the sensitivity of the rock's permeability to small changes in confining pressure increased. Introduction In recent years the importance of natural gas contained in low-permeability rock has increased tremendously. Reservoirs that once were dismissed as too tight now are being reevaluated in the light of new technology, such as massive hydraulic fracturing. In some cases, very tight gas-saturated rock is found directly above or below higher-permeability gas reservoirs, which creates the possibility that gas from the tight sand can migrate vertically into the producing formation. In any case, the permeability of the tight rock is of utmost importance in estimating the future recoverable reserves and the recovery rate. The most common method of determining permeability in cores recovered from depth has been to measure air permeability in systems that apply only enough pressure to the core to prevent gas flow around the jacket. The results of this study indicate that gas permeability of tight sandstones determined by this method can be an order of magnitude higher than when in-situ levels of confining pressure are applied. Using steady-state flow methods to measure permeability of these samples with confining pressure becomes quite difficult because long times are required to establish equilibrium conditions. The addition of partial water or brine saturation further reduces gas permeability, compounding the problem. For these reasons, we have constructed an apparatus that uses a fast and accurate pulse-decay technique and allows independent control of confining pressure (pc), pore pressure (pp), and fluid saturation (Sw). Permeability as low as 20 × 10 -9 darcy can be measured quite easily with this equipment. The samples studied here are from the Spirit River member of the Fahler formation in western Alberta, Canada. This formation is believed to contain large volumes of natural gas in low-porosity, tight sandstones. Spirit River samples are denoted by SR followed by a number that corresponds to the depth (in feet) from which they were recovered. Detailed mineralogy data from thin-sectioned, X-ray diffraction, and scanning electron microscope (SEM) methods are given in Table 1. Experimental Technique A pulse-decay technique was used to make the large number of permeability measurements required for this experiment in a reasonable period of time. The concept of this method has been described by a number of experimenters, including Brace et al., Sanyal et al., and Zoback and Byerlee. JPT P. 930^