An investigation on carbon dioxide sequestration in Indian coal seams

The term ‘geological sequestration’ stands for the capture of CO2 directly from anthropogenic sources and disposing of it deep into the ground for geologically significant periods of time (Bachu, 2002). Coal seam sequestration as described by White et al. (2005) is “the storage of CO2 from anthropogenic sources into deep, unminable coal seams for geologically significant limits with or without concomitant recovery of natural gas”. Methane is native to coal and has its origin in coalification process. Coal is the source as well as reservoir to significant quantities of methane, a potential energy resource. Coalbed methane (CBM) a bonus non-conventional energy source is generated mainly due to geochemical transformation of the organic matter by catagenesis. CO2 injection reduces the partial pressure of methane and enhances desorption of methane from the matrix. CO2 has an additional effect compared to other gases that it is preferentially adsorbed onto coal surfaces, displacing methane from adsorption sites. CO2-ECBM sequestration is a value addition project in management of increasing atmospheric concentration of greenhouse gases (GHG) as it recovers the cost of capture, processing, transportation and storage of CO2 by production of methane. The primary aim of this research was to develop an understanding on the coal-fluid interaction pertinent to carbon storage in coal seams with special reference to Indian coal basins. Damodar valley coalfield where production of coalbed methane is being carried out was chosen for this study. The objectives of this study included detailed knowledge of geotechnical characterization of coal measure rocks, strength characteristics of coal under saturation in different media, swelling and stress induced changes in flow behaviour of coal in multiple phases of carbon dioxide and estimation of gas storage potential of Indian coal basins. After the geological field work, petrographic and geomechanical characterization of coal measure rocks was carried out. Sandstone and shale along with varying degrees of their intercalations were identified in thin section studies under the microscope. The pertinent geomechanical characteristics of the rocks and coal were determined using laboratory tests according to the International Society of Rock Mechanics (ISRM) standards. Some data generated during these tests were used as inputs for prediction of Uniaxial Compressive Strength (UCS) and P-wave velocity using soft computing. An important aspect of coal seam sequestration is the influence of CO2 saturation on strength parameters and failure characteristics of the host. Fluid saturation experiments were carried out to explore the effects of CO2 adsorption on natural as well as moisture-equilibrated coal. Water saturation of coal led to high reduction in its strength implying that the seam pressure in water saturated sinks must be managed at lower levels during sequestration as compared to the dry reservoirs. Injection of CO2 leads to sorptive weakening of coal. The scenario was worst when CO2 was injected in wet coal. CO2 saturation induces swelling in coal and creates or enhances the fracture lines along coal as also evident by the early crack initiation in the acoustic emission results. Coupled stress strain data during uniaxial compression were recorded and analyzed. Introduction of water and CO2 caused early failure of coal samples at a lower value of load alongside inducing larger amount of strain at same stress values. The brittle nature of coal became less pronounced upon water saturation leading to larger time for crack initiation. To overcome the structural heterogeneity in coal, reconstituted coal specimens of uniform grain size were prepared using moulds and their strength was calculated under dry and saturated conditions. The reconstituted coal specimens were developed at different stresses and the properties of coal developed at 22 MPa and 24 MPa were found to be closest to that of natural coal. The magnitude of strength reduction in these samples was less as compared to the natural specimens. This could be due to high compaction and elimination of flow paths for fluids. Sample interiors remained largely intact and least affected by the fluids and hence, the strength remained closer to unsaturated reconstituted coal specimens. Permeability in coal is one of the most vital issues that determine the production or sequestration performance of coal. The permeability of CO2 in different phases in coal samples under various possible scenarios was estimated using a newly developed, high precision P-T controlled triaxial set up. The permeability of coal decreased exponentially with increasing effective stresses for Indian coal and new empirical equations accounting for stress behaviour of coal permeability were proposed. The rate of coal matrix swelling reduced in 30–40 hours of gas injection after which coal permeability gradually increased with increasing upstream pressure. Finally, at constant pore pressure, coal permeability reduced drastically with increasing confining stresses. A similar experiment was conducted using naturally fractured coal from the same seam while maintaining a low range of confining and injection pressures to ensure a complete gas phase CO2 flow. The permeability of coal reduced with increasing CO2 injection pressure in four out of five cases of confinement implying that complete closure of fractures due to coal swelling took longer time of CO2 saturation. However, permeability of coal reduced exponentially with increasing effective stresses. The next experiment investigated liquid and supercritical CO2 permeability of coal at varying confining pressures (15 – 24 MPa), corresponding to different depths of coal. The initial liquid CO2 permeability of coal at 10 MPa injection pressure reduced from 0.011 mD at 15 MPa confinement to 0.0004 mD at 24 MPa confinement. Further experiments revealed that the initial permeability of coal using supercritical CO2 (0.005 mD) was nearly half of that for liquid CO2 (0.011 mD). Nitrogen, used as a relatively inert medium, experienced a reduction in permeability in coal through which CO2 was passed. Maximum reduction in coal permeability was observed in supercritical CO2 flow, due to high sorption induced swelling of coal. After the role of confining stresses in different phase of CO2 was studied, the role of CO2 saturation period on coal permeability was studied at a fixed confining pressure of 18 MPa. Permeability of liquid as well as supercritical CO2 reduced after each period of 20 h saturation. Maximum permeability reduction took place by nearly 27% and 43% for liquid and supercritical CO2 saturation respectively, in the first out of three saturation periods. However, permeability of supercritical CO2 continued to reduce after each period of saturation while minimal reduction took place at the end of second and third period of swelling with liquid CO2. Similar trends were also observed in case of N2 since the cleats were effectively closed due to passage of different phases of CO2. This highlights that supercritical state of CO2 induces maximum adsorption related swelling and the subsequent reduction in coal permeability. Based on detailed experimental understanding of the coal samples and the field information on behaviour of coal seams, reservoir simulation studies were carried out using a commercial simulator called COMET 3. Model construction was followed by history matching of the CBM wells to establish and validate those before extending the cases for CO2 injection and enhanced recovery of the natural gas. The same established models were used to investigate parametric influence on production characteristics of coal. In a scenario of given block size with one injection well and two production wells, the spatial distribution and relative flow of the two gases were explored with change in time. The salient findings of these studies included: coals adsorbed CO2 quickly and surrounding matrix attained peak CO2 matrix concentration in one month due to high adsorption potential for CO2. The fracture gas saturation increased at first mainly near the production well due to sudden desorption of CH4 as a consequence of dewatering. Gradually, high gas saturation resulted in the block. The statistical data generated from this study revealed that over a period of 4000 days of observation, the coal block would take in approximately 7.7 bcf of CO2 and in lieu of which it would produce around 2.6 bcf of CH4 and a total of 12000 bbls of water. Similar exercise was carried out for two blocks in Jharia coalfields. The study on one block showed that over a period of 4000 days, the chosen coal block adsorbed approximately 7.75 bcf of CO2 in turn released 2.24 bcf of methane gas. The same variety of coal was simulated for a period of 20 years with an increased block size and it was found that for the chosen dimensions of Gondwana coal block in India, a total of 15.1 bcf of CO2 may be injected for permanent storage alongside recovery of not less than 5 bcf of methane. Numerical simulation for the established Jharia model was used to predict and understand the influence of sorption time on the production behaviour of coals. It revealed that the CO2 injection capacity into high sorption time coal was significantly higher than coal with low sorption time. Therefore, if not suitable for economical extraction of methane, these may alternatively be utilized as CO2 sinks. It was also shown that for coal belonging to Gondwana basin setting in India, sorption time less than 10 days followed equilibrium model of desorption. Although coals with higher sorption time exhibited non-equilibrium desorption and diffusion, they were finally observed to converge with the equilibrium model at later stages of gas injection/production. Some future works based on the new findings were suggested at the end of this research work. (...