Estimation of matrix permeability is critical for evaluating the long-term economic viability of unconventional reservoirs. Currently, a variety of techniques (e.g. pulse-decay, crushed-rock) are used to measure gas permeability of tight rocks in the laboratory. However, these laboratory-based methods are not fully representative of different gas flow regimes that are encountered during field production. In this work, a new experimental setup, and previously-developed rate-transient analysis (RTA) techniques, are combined to mimic field-scale well-test/production data in the laboratory for the determination of representative matrix permeability and pore volume in tight rocks.
A new experimental set-up comprised of a vacuum pump, a series of (back-pressure) valves, a high-precision pressure transducer and flowmeter was developed to simulate gas well-test/production scenarios in the laboratory. The experimental procedure involves injection of gas (CH4) into one end of a core plug (monitoring the pressure until equilibrium), followed by constant flowing pressure gas production from the same end that gas was injected. For analysis of the data, a log-log plot of flow rate versus time is used to first identify the flow regimes during the production phase. Using previously-derived RTA algorithms, permeability is then estimated using the slope of the square-root of time plot (if transient linear flow is observed) and distance of investigation calculation (if the end of linear flow is observed).
In order to test the new apparatus, experimental procedures and RTA algorithms, two experiments under similar experimental conditions were conducted on a core plug from the Montney Formation, which was previously analyzed using more conventional methods. The flow-regimes identified during the production cycle were linear flow followed by boundary-dominated flow for the two experiments. The square-root of time plot yielded permeability estimates of 0.00067 and 0.00072 md from tests A and B, respectively - the distance of investigation (DOI) approach yielded a permeability estimate of 0.00069 md for both tests. The results of the two experiments are in reasonable agreement (maximum discrepancy < ± 20%) with permeability measured using the more conventional pulse-decay technique with methane as the analysis gas (0.00084 md).
Standard laboratory techniques for the determination of permeability in tight rocks are not fully representative of fluid flow mechanisms that occur during field-scale production. Routine laboratory-based methods either use samples that are not representative of the "in-situ " reservoir rock (e.g. crushed-rock samples) or represent only unidirectional fluid flow through the core plug samples (e.g. pulse-decay and steady-state techniques) that does not account for heterogeneity observed in the field. Integrating the previously-developed rate-transient analysis techniques with a new experimental set-up, the core testing procedure proposed herein represents a novel approach to the evaluation of tight rock permeability, better simulating field-scale production than previous approaches.
