Abstract

Abstract

Behavior of the Coal Observed with Injection Liquid /Supercritical CO2 and N2 on Stress Constraint Conditions
Journal of MMIJ, 126, 148-155, 2010
Tamotsu Kiyama, Soshi Nishimoto, Nago Masao, Masashi Fujioka, Ziqiiu Xue, Yoji Ishijima

Coal bed CO2 sequestration, which separates and collects the CO2 generated by a large-scale point source and then injects it into a deep underground coal bed, can be regarded as a viable form of carbon capture and storage (CCS). However, many uncertainties are still involved, due to the fact that coal reacts with CO2 in a variety of ways. In verification tests at Yubari, CO2 was injected from an injection well into the coal bed at depth of 900 m, and methane was produced from an production well. Since the injection rate of CO2 was one tenth of magnitude that estimated by the preliminary analysis, N2 was injected in an attempt to improve its performance. While the injection rate of CO2 increased temporarily, it later decreased again in a short time. To try to clarify the phenomena observed in the Yubari trials, we conducted two types of laboratory tests under stress constraint conditions. In the test I, we injected liquid CO2 into a water-saturated core specimen and heated it to make supercritical CO2. This is based on the assumption that during the initial CO2 injection at Yubari pores of the coal bed were saturated with water. In the test II, we injected supercritical CO2 into the specimen saturated with N2, and then repeatedly injected N2 and CO2. This test corresponded to the N2 injection and CO2 re-injection at Yubari. During the test I, we observed a swelling strain of 0.25 to 0.5% after injecting CO2; during test II the swelling strain was 0.5 to 0.8% after injection of supercritical CO2. Following the further injection of N2 in the test II, slow shrinkage was observed. At effective conf ining pressure of 2 MPa, permeability of the water-saturated specimen was 2 ×10.6 darcy. In contrast, the permeability of the N2-saturated test specimen was originally ranged during 5 × 10-4 to 9 × 10-4 darcy, and after injection of supercritical CO2 it decreased to 2 × 10.4 darcy. Further injections of N2 and supercritical CO2 caused little subsequent change in permeability. It seems that when liquid CO2 is injected into the water-saturated specimen, it does not completely replace water in the coal matrix, based on residual amounts of CO2. To explain this behavior, we developed a model in which the CO2 permeating the coal is distributed in clefts shaped like flat cracks and in fine pores in the matrix interior.