Assessment of shrinkage–swelling influences in coal seams using rank-dependent physical coal properties

Abstract

Characterization of coal reservoirs and determination of in-situ physical coal properties related to transport mechanism are complicated due to having lack of standard procedures in the literature. By considering these difficulties, a new approach has been developed proposing the usage of relationships between coal rank and physical coal properties. In this study, effects of shrinkage and swelling (SS) on total methane recovery at CO₂ breakthrough (TMRB), which includes ten-year primary methane recovery and succeeding enhanced coalbed methane (ECBM) recovery up to CO₂ breakthrough, and CO₂ sequestration have been investigated by using rank-dependent coal properties. In addition to coal rank, different coal reservoir types, molar compositions of injected fluid, and parameters within the extended Palmer & Mansoori (P&M) permeability model were considered. As a result of this study, shrinkage and swelling lead to an increase in TMRB. Moreover, swelling increased CO₂ breakthrough time and decreased displacement ratio and CO₂ storage for all ranks of coal. Low-rank coals are affected more negatively than high-rank coals by swelling. Furthermore, it was realized that dry coal reservoirs are more influenced by swelling than others and saturated wet coals are more suitable for eliminating the negative effects of CO₂ injection. In addition, it was understood that it is possible to reduce swelling effect of CO₂ on cleat permeability by mixing it with N₂ before injection. However, an economical optimization is required for the selection of proper gas mixture. Finally, it is concluded from sensitivity analysis that elastic modulus is the most important parameter, except the initial cleat porosity, controlling SS in the extended P&M model by highly affecting TMRB.

Introduction

ECBM recovery technique increases methane production significantly and leads to CO₂ sequestration. As being the most important parameter affecting both methane recovery and CO₂ sequestration, absolute fracture (cleat) permeability depends on the balance between effective stress in the cleats and shrinkage & swelling (SS) of coal matrix. Around production wells, depletion of methane begins within cleats. Subsequently, adsorbed methane inside the coal matrix desorbs and diffuses to feed the cleats due to concentration difference. During this process, the effective stress in the cleats increases and coal matrix shrinks as a result of desorption. Former decreases permeability by getting closer the cleat surfaces and latter does the opposite. As for injection well, CO₂ results in swelling owing to relatively higher affinity of coal to CO₂ than CH₄, which decreases the absolute cleat permeability. On the other hand, depending on its magnitude, injection pressure can reduce swelling effect by expanding the fracture aperture mechanically (Mavor and Gunter, 2004).

Relative permeability has also an important role on the production of coalbed methane. As the reservoir pressure is reduced in the cleat system by production of water, methane desorbs into the cleat system. At this point, and for the remainder of the life of the producing wells, two-phase flow occurs in the cleat system. Under two-phase flow conditions, the relative permeability relationships between gas and water control the relative flow of gas and water in the reservoir. Thus, it is important to determine the relative permeability characteristics of the coal being analyzed. Furthermore, changes in fracture porosity due to shrinkage and swelling during application of ECBM recovery technique controls fluid saturations that in turn lead to changes in relative and effective gas permeability (GRI, 1996).

In the literature, there are parametric simulation studies (Smith et al., 2003, Davis et al., 2004, Maricic et al., 2005) investigating the effects of physical coal properties on primary and enhanced coalbed methane recoveries. The trend is generally to change a model parameter in its given range and to observe its impact on TMRB. In this study, however, a new approach was followed during preparation of input data for a commercial compositional simulator, CMG (Computer Modeling Group)/GEM module (CMG, 2007). Instead of using a real field or a hypothetical data set, rank-dependent coal properties in the literature were gathered to construct a database. This database enabled us to acquire more realistic outputs from the simulator. Impact of SS on TMRB and CO₂ sequestration were studied by using this database. In the following section, methodology followed up during preparation of rank-dependent coal properties is explained.