Methane resaturation in Barnett Formation core plugs and new approach for determination of post-coring gas loss

Abstract

Understanding the physiochemical mechanisms that control the loss of gas during coring processes is critical to accurately determining gas-in-place (GIP) resource assessments of unconventional shale-gas plays. Our study uses an experimental approach, utilizing methane (CH4) adsorption isotherms and degassing curves of methane-resaturated Barnett Formation core plugs, to determine the amount of lost-gas based on mass-balance analysis at different CH4 re-saturation pressures and varied exposure times. Several readily available empirical methods for estimating lost-gas were evaluated, quantified, and compared with the mass balance–derived lost-gas values in our experiments.A CH4 isotherm measurement on 3/8-inch Barnett Formation core plugs was performed at 35.4 °C; the amount of gas adsorbed in excess was then quantified and fitted to the modified Langmuir equation to determine the Langmuir maximum, Langmuir constant, and adsorbed gas-phase density. Two sets of CH4 gas-resaturation and degassing measurements, one varying saturation pressures and the other varying exposure times, were performed on 3/8-inch Barnett Formation core-plugs at an isothermal temperature of 35.4 °C. Degassing curves, the plot of the released gas yield versus the square root of degassing time, display three stages that correspond to different gas-releasing mechanisms. The rapid increase of released gas yield at the beginning of degassing represents that nonlinear gas expansion is dominant and that degassing evolves into a linear desorption-dominated phase over time. Experimentally derived values for lost gas were determined by subtracting the sum of the desorbed and retained gas at the peak of the degassing curve from the amount of gas initially charged into the samples at equilibrated resaturation pressure. Lost gas varies linearly with increasing gas-resaturation pressure and nonlinearly by a greater magnitude with increasing exposure time, indicating that lost gas is more sensitive to exposure time.The uncertainty evaluation of lost gas determined by three empirical methods was conducted by direct comparison with mass-balance-derived lost-gas values from our experiment. Nonlinear least-squares extrapolation overestimates, and both linear extrapolation and polynomial equation fitting underestimate, mass-balance lost-gas control points. Among the three empirical methods, the polynomial-fitted lost-gas values most closely agree with mass-balance lost gas, revealing that polynomial fitting to degassing curves is a viable way to accurately estimate lost gas and, more importantly, to estimate GIP values with up to 85% accuracy. As free gas is the dominant storage mechanism in the Barnett Formation, the trends of experimentally measured cumulative desorbed gas content versus the square root of elapsed time match neither the calculated curves derived from USBM Direct Method nor Amoco Method.