煤矿瓦斯管理的模拟气体回收试验外文文献翻译

英文原文Simulation of an enhanced gas recovery field trial for coal mine gas managementRussell Packham a,*, Yildiray Cinar b, Roy Moreby aa School of Mining Engineering, University of New South Wales, Sydney, NSW 2052, Australiab School of Petroleum Engineering, University of New South Wales, Sydney, NSW 2052, AustraliaArticle infoArticle history:Received 30 September 2010Received in revised form 27 November 2010Accepted 27 November 2010Available online 13 December 2010Abstract: Coal mine gas management has evolved from being predominantly dependant on mine ventilation systems to utilising sophisticated surface based directional drilling for pre-drainage of coal seams. However the advent of enhanced gas recovery techniques in the coalbed methane industry has provided an opportunity to address gas management objectives hitherto impractical. Specifically: achieving very low residual gas contents to mitigate against frictional ignitions and fugitive emissions; the means to accelerate gas drainage to accommodate mine schedule changes; and to enable pre-drainage of coal reserves with very low permeability. This article examines a possible enhanced gas recovery field trial at an Australian mine site. Production data from four surface to inseam medium radius gas drainage boreholes was modelled and history matched. The resulting reservoir characteristics were then used to model the performance of the boreholeusing an enhanced recovery technique. One of the boreholes is modelled as an (nitrogen) injection well and two flanking wells are modelled as production wells.The model results suggest that accelerated gas flow rates as well as very low residual gas contents are achievable using typical coal mine gas drainage infrastructure and goaf inertisation systems.Keywords: Enhanced gas recovery , Coal mine gas management , Pre-drainage , Methane drainage , Nitrogen1.IntroductionAs underground coal production rates increase, and the seamsworked become deeper and gassier, the practise of gas drainage has been progressively adopted. Gas drainage involves capturing the seam gases, before they enter the mine ventilation system, and using a reticulation system to dispose of the gases safely. Gas drainage includes pre-drainage of gas prior to mining commencing and post drainage(also known as goaf or gob drainage) of gas from fracturedand de-stressed strata associated with coal extraction.Adoption of medium radius drilling for surface pre-drainage of coal seams has allowed drainage lead times in excess of 3 years. Typically this degree of pre-drainage achieves residual gas levels that do not impede mining operations. There are, however, some scenarios that still provide difficulties from a coal mine gas management perspective:Localised regions of low permeability requiring extended drainage time not compatible with mine development and extraction schedules.Changes in mine development schedules requiring accelerated drainage.Lowering residual gas levels to very low levels as a possible mitigation of frictional ignitions and fugitive emissions.A promising technique referred to as “enhanced gas recovery” was described by Puri and Yee (1990). The technique was proposed for use in the coalbed methane industry to both increase cumulative production of methane from coal seams and to improve the rate of recovery. Recognition of the potential value of this technique with respect to coal mining has largely been overlooked by coal mining industry although two references to the possibilities have been found(Brunner and Schwoebel, 2007; Thakur, 2006).The technique involves injecting a gas, which is different to the seam gas, into the coal seam to stimulate methane (or other seam gas) production. The injectant is introduced into the coal seam via an injector borehole and the seam gas is collected at separate production borehole(s).Four field trials involving using nitrogen as an injectant gas have been identified. The Tiffany unit trial (Reeves and Oudinot, 2004) which ran for 3 years demonstrated a 5-fold increase in gas flow rate arising during the trial. The Alberta Fenn Big Valley micro pilot trial (Mavor et al., 2004) demonstrated increases of absolute permeability from initial conditions of 1.2 mD to 13.8 mD resulting from nitrogen injection. During the Yubari CO2 sequestration trial (Shi et al., 2008),in an attempt to improve the CO2 injection rate, N2 was introduced to the reservoir. Modelling of the results indicated that an improvement in well block permeability of 0.1 mD to 40 mD accrued.One trial involving nitrogen injection has been conducted at an underground coal mine in China (Yang et al., 2010). The trial involved injection of nitrogen at 500 kPa into closely spaced, 1520 m long gas drainage boreholes at the face of a development heading. The gas flow rate from the production boreholes demonstrated a 2 fold increase after 16 h of injection.The stimulation of the gas production and seam permeability is brought about by several mechanisms which are described below. The objective of this paper is to determine whether enhanced gas recovery can be applied to coal mine gas drainage systems, theoretically and practically. The specific research questions are: Can enhanced gas recovery be used to lower residual seam gas content to negligible levels? Can enhanced gas recovery be used to significantly increase gas drainage recovery rate?2. Background fluid flow in coal2.1. Coal structureGas in coal exists predominantly as molecules adsorbed on the internal surface of the coal matrix (Zuber, 1996). The adsorbed gas exists within the coal matrix at near liquid density (Yee et al., 1993). Unlike natural gas reservoirs in which gas exists in the pore space of the reservoir, coal seams can store comparatively large volumes of gas at low pressures. Coal seams also demonstrate an orthogonal fracture system known as cleat. The cleat system is comprised of face cleat and butt cleat, which typically exists at 90 to face cleat. Butt cleat is terminated by face cleat and both orientations tend to be normal to the bedding plane horizon.The cleat system provides a transport route for gas migrating through the seam. It should be acknowledged that two separate processes control the flow of gas from a coal seam reservoir to a gas drainage system: diffusion of gas from adsorbed state in the coal matrix to the cleat fracture system, and flow through the cleat fracture system as free gas. The cleat system is generally saturated with water before the gas drainage process begins.In order to explain the process of enhanced gas recovery it is desirable to understand the mechanisms affecting primary gas recovery. Two laws that control the rate of gas recovery from a coal seam are the Darcys law which governs gas and water flow through the cleat system and the Ficks Law which models the gas diffusion from the coal matrix into the cleat system.2.2. Diffusion Ficks law and adsorption isothermsThe mechanism by which gas migrates from the very small pores of the coal matrix to the cleat system is diffusion. Three types of diffusion can contribute to the overall diffusion rate namely, bulk diffusion, surface diffusion and Knudsen diffusion (Cui et al., 2004; Zuber, 1996). The diffusion coefficient, D, is a composite of these three type of diffusion. Diffusion of gas from the coal matrix into the cleat system is described by the modified Ficks law (Zuber, 1996): （1）where the gas production rate , is a function of matrix volume, , and the difference between the matrix gas concentration, , less the equilibrium concentration at the matrix cleat boundary . The diffusion coefficient, D, and fracture spacing, , are normally lumped together through the use of desorption time, , which may be derived from a gas content testing. （2） is defined as the time taken for 63.2% of the total gas to diffuse at constant pressure from a coal sample (King et al., 1986). can be derived from the gas content of a coal sample measured using a slow desorption method. Most Australian coal mines conduct pre-drainage for mitigation of potential outburst hazards. The outburst hazard is assessed during routine mine operation by gas content testing for which a fast desorption method is generally used. cannot be directly determined from a fast desorption gas content test. However a common feature of slow and fast desorption gas content testing is the measurement of initial desorption rate to determine lost gas (“Q1”).From the initial desorption rate of the core the volume of gas desorbed in the first 30min after the core was deemed to have started desorption is measured. This volume, referred to as “IRD30” in fast desorption testing, has been used to provide a correlation to (Williams and Yurakov, 2003, pp. 7172): （3）This correlation has been used to provide an indicative for the seam gas at the mine site subject to this investigation.The use of to characterise the desorption behaviour of the reservoir is simplistic. The simplification arises from the assumption that the fracture spacing of cleat and shape factor of the matrix in the coal sample are representative of the whole reservoir. The diffusion coefficient, D, is related to the pore size within the matrix and the mean free path of the gas molecule (Cui et al., 2004). Thus where an injectant gas is used to stimulate diffusion of a seam gas, a separate desorption time must be identified for each gas if the process is to be modelled.The concentration of gas in the coal matrix is defined by the coals characteristic adsorption isotherm and has a non-linear relationship with pressure. Yee et al. (1993) provide a good description of the various models available for defining isotherms. Pan and Connell (2008) have discussed the problems associated with various adsorption models in relation to potential errors and practicality of application. Of the models the Langmuir isotherm model is widely used for primary gas drainage modelling and can be modified to create binary or ternary gas mixture isotherms from pure gas isotherms as is necessary when examining the effects of multiple gas species adsorbing and desorbing in a coal matrix. The Langmuir model is described by (Zuber, 1996): The Langmuir model is described by (Zuber, 1996) （4）where V is the coal gas content at pressure p in equilibrium, is the Langmuir constant, and VL is the Langmuir volume defining the adsorption isotherm for a single gas in a specific coal seam.Where the Langmuir model is used in a mixed gas environment (such as enhanced gas recovery or where the seam gas composition is a variable mixture of methane and carbon dioxide) the Extended Langmuir Model (ELM) may be applied: （5）such that , VEi is the extended Langmuir storage capacity of component i, VLi is the Langmuir volume for pure component i; Yi, is the mole fraction of component i; nc is number of components, and i is the Langmuir constant for component i. There is an assumption in the use of ELM that there is no interaction between the different gas species arising from counter-diffusion as would occur in an enhanced recovery process.2.3. Darcys LawFor a 1-D, single phase flow through a porous medium (coal seam) in the following manner Darcys Law is described by: （6）Volumetric flux in the x direction, vx, is a function of seam permeability, kx, the fluid viscosity, , and the incremental pressure drop. The pressure drop relates to the difference in pressure at the gas drainage borehole and the seam gas pore pressure. In Australian longwall mining environments the coal seams are typically almost level, as a consequence gravitational effects on gas flow are negligible. Seam permeability is a dominant parameter for gas production rates.2.4. Permeability variationsPermeability is known to be sensitive to effective (horizontal) stress brought about either by local tectonic or geological structures and also by the shrinkage or swelling of the coal matrix due to desorption or adsorption of gas.Gray (1987) described permeability of a coal seam in relation to the changes in effective stress in the coal seam, where, if water is removed from the cleat, the matrix blocks are less constrained and tend to compress the cleat. This process, referred to as cleat compression, leads to a reduction in permeability. As the fluid pressure in the cleat system falls, gas desorption occurs. The subsequent release of gas from the matrix into the cleat causes the matrix to shrink and a reduction in effective horizontal stress. In terms of permeability changes, the two processes, cleat compression and matrix shrinkage tend to cancel each other. Several analytical models have been developed to interpret the effect of cleat compression and matrix shrinkage on reservoir permeability. Palmer (2008) provides an exposition of the analytical models and discusses how the reservoir permeability may be determined from either change in volumetric strain/porosity or effective horizontal stress. Seidle et al. (1992) described how coal seam permeability changes in effective horizontal stress: （7）where ko is the initial permeability, cf is the cleat-volume compressibility and is the horizontal stress. Shi and Durucan (2004) developed a relationship to enable the calculation of change in horizontal effective stress resulting from changes in reservoir pressure and desorption of gas from the coal: （8）where l and P are the matrix shrinkage constants, and E are the Poissons ratio and Youngs Modulus of the coal, respectively. Initial or reference horizontal stress and pore pressure are 0 and p0, respectively. The two terms on the right hand side of the equation relate to cleat compression and matrix shrinkage respectively.Volumetric shrinkage strain, s, is considered in the Shi/Durucan formulation to be related to the Langmuir type relationship of matrix strain at the maximum adsorbed gas content and the gas content pressure at which half of the maximum strain occurs: （9）Shi and Durucan (2005) further developed this relationship to account for matrix swelling (as may occur where a gas adsorbs onto the coal matrix in enhanced gas drainage). Assuming the pressure of the free gas in the cleat is in equilibrium with the adsorbed gas, then: （10）where s is the volumetric shrinkage/swelling coefficient for a specific gas (i.e. a seam gas, methane or an injectant gas such as nitrogen), V corresponds to the gas content at the reservoir pressure, p; Vo is the gas content at the initial reservoir pressure po. V and Vo can be determined using the Langmuir isotherm (see Eq. (4).This allows the change in effective horizontal stress to be determined resulting from cleat compression and matrix shrinkage or swelling due to change in pore pressure: （11）In an enhanced gas drainage process using nitrogen as an injectant, the coal matrix desorbs one gas, generally methane or carbon dioxide, and adsorbs nitrogen. The net matrix shrinkage effect is thus determined by the volumetric shrinkage coefficients, s and the mixed gas adsorption isotherm characteristics for the desorbing and adsorbing gasses as may be determined from the extended Langmuir model, Eq. (5).2.5. Gas-water relative permeabilityGas drainage from coal seams includes typically gas and water. The water exists predominantly in the cleat space (pore volume) whereas the gas exists initially in an adsorbed state in the matrix. As desorption occurs gas becomes present as free gas in the cleat space thereby creating the second phase within the cleat system. Gas may also be dissolved in water. The solution of gas is negligible for methane in water but may be significant for carbon dioxide in water.At initial conditions the cleat space is typically saturated with Water and the effective permeability of the cleat to gas is zero. In order for gas to flow, the water saturation must be reduced. This is achieved by drawdown of the gas drainage well. Drawdown refers to pumping water out of a gas well and progressively reducing the well pressure to enable water to flow to the well and subsequently gas, see Fig. 1. Underground gas drainage holes achieve drawdown by drilling the hole from seam horizon at atmospheric pressure whereas surface gas drainage boreholes achieve drawdown by pumping water from the well.Relative permeability is important in relation to gas drainage from coal seams as the characteristic of the relative permeability curves determine how much water (%) must be removed from the seam before gas begins to flow.The gas flow, desorption/diffusion and permeability relationships referred to above may be modelled by means of finite difference numerical modelling (King et al., 1986). Several coal-bed-methane simulators are available to model these behaviours. Law et al (2002) compare 5 simulators as part of an investigation into improvements for enhanced coalbed methane modelling. One of the simulators, Simed2,has been used in this research (the windows version SimedWin).2.6. Primary productionPrimary production thus is brought about by pressure reduction at a production well causing fluid flow through the cleat system to the well. Fluid flow in turn causes a pressure reduction at the cleat/matrix interface resulting in desorption of gas within the matrix and diffusion of the gas to the cleat interface. The p