Permeability enhancement in coal seam gas reservoirs
Coal seam gas (CSG) is a type of unconventional gas resource and it is mainly stored in the coal matrix in the form of adsorption. Gas/water flow in underground coal seam is controlled by a set of orthogonal fractures known as cleats. Permeability of an underground coal seam is dynamic and can change significantly during the production life of the well. The change in permeability is attributed to two competing factors: shrinkage of the coal matrix and rock compaction. Rock compaction tends to decrease fracture width while matrix shrinkage enlarging the fracture aperture. The cumulative impact of matrix shrinkage and rock compaction determines the degree and the functional form of permeability change during the life of a CSG well. Figure.1 shows a set of cleats and the two contributing factors (matrix shrinkage and compaction) on permeability change.
In under-saturated CSG reservoirs (cleats are 100% saturated with water) water extraction, known as de-watering operation, can result in coal compaction and reduction of permeability. Matrix shrinkage depends on the amount of gas desorbed from coal matrix. Since very little gas is produced from coal seam during the de-watering stage, matrix shrinkage has negligible impact on permeability change at this stage. Hence, rock compaction becomes the dominant factor reducing coal permeability. As more gas is extracted from coal, matrix shrinkage effect starts to play a more significant role. Coal permeability may rebound when matrix shrinkage effect cancels out the compaction effect and the pressure at which permeability starts to increase is known as the rebound pressure. Figure.2 shows the theoretical permeability change versus average reservoir pressure for an under-saturated CSG reservoir.
Low porosity coals are more prone to permeability enhancement than higher porosity coals. I will make a simple example to describe this. For a coal sample with a cleat porosity of only 0.1%, matrix part of the sample is 99.9%. If matrix shrinks only by a small amount due to gas desorption, (let us say from 99.9% to 99.8%), cleat porosity doubles. Based on the Carmen-Kozeny equation, permeability of the fracture network should increase by a factor of 8 (0.2/0.1)^3. Now let us assume that the cleat porosity of the coal sample was 1% (99% matrix). To achieve the same permeability increase (8 times), matrix should shrink from 99% to 98%, which is one order of magnitude higher than the previous case (0.1%). Higher shrinkage is needed in high porosity coals to achieve permeability enhancement. One particular field example is the Fairview Field, a prolific CSG reservoir in eastern Australia, where multiple coal seams are intersected in the Bandanna Formation. Some coal seams have high porosity (2-3%) while others show lower porosity. Rate transient analysis (RTA) and time-lapse pressure transient analysis (PTA) were conducted to investigate permeability enhancement in two wells. The well that intersected high porosity coal seams exhibits permeability enhancement by a factor of ~3, despite huge cumulative gas production, while the one intersected lower porosity coals shows permeability enhancement by more than one order of magnitude.
Time-lapse PTA, production history matching and rate transient analysis have been performed to investigate magnitude and functional form of permeability enhancement for producing wells in different Basins (See Table.1). All of these techniques require a CSG well to produce gas for a sufficient time to identify permeability change. For newly drilled well, mathematical models, incorporating shrinkage and geo-mechanical data can be utilized to predict permeability enhancement as a function of reservoir pressure similar to Figure 2 (the permeability data of this figure were calculated using Palmer-Mansoori model).
The magnitude of permeability change reported for CSG wells in the San Juan and Bowen Basins ranges from very small number up to 2 orders of magnitude. The functional form of permeability change for 28 CSG wells, located in the Fairway of San Juan Basin, is exponential. The functional form of permeability can be obtained by performing time-lapse PTA (at least 3 points during the life of the well) or production history matching techniques. The former is a more robust technique while the latter is subjected to uncertainty in flowing bottom-hole pressure data and other reservoir parameters.