Devonian Shale Gas Production: Mechanisms and Simple Models

Abstract

Summary We show that, even without consideration of their special storage and flowproperties, Devonian shales are special cases of dual porosity. We also showthat, while neglecting these properties in the short term is appropriate, suchneglect in the long term will result in an under-estimation of production. Introduction Tremendous accumulations of gas occur in the Devonian shale formations ofthe eastern U.S. Because the formations are so tight, gas can be produced onlywhen extensive networks of natural fractures exist. In areas where fracturingis sufficient to promote production, the gas basically will flow theproduction, the gas basically will flow the short distance from storage in thetight rock matrix to the fracture network and then through the fracture networkto the producing well. Only minute quantities of gas flow directly from therock to the well. These production characteristics make shale gas productioncharacteristics make shale gas reservoirs examples of so-called dual-porositybehavior. The problem with the term "dual-porosity" in shales, however, is that shales have very little open porosity. Unlike conventional gasreservoirs, where gas is stored in the open pore space of the rock, shalesstore an enormous amount of gas in a sorbed (adsorbed or absorbed) state. Thebehavior of fluid flow in the shale matrix also deviates from that of aconventional gas-reservoir/rock matrix. The permeability of conventionalreservoir rock is permeability of conventional reservoir rock is really amacroscopic average of viscous open-channel flow occurring on a pore level. Inshales, the conduits are generally so small that only a few molecules can slipthrough the openings at any given time. Open-channel flow cannot exist for themost part; thus, flow through the shale matrix results predominantly frommolecular diffusion. predominantly from molecular diffusion. The objectives ofthis paper are to clarify the roles that these unusual storage and flowproperties have in the general dual-porosity properties have in the generaldual-porosity system; to estimate their effects on the production behavior ofcommon Devonian shale production behavior of common Devonian shale wellsheuristically; and to show that even without the unusual storage and flowbehaviors, Devonian shale gas reservoirs are special cases of dual porositybecause of the tight nature of the rock matrix. This paper does not present newpredictive models for the behavior of shale reservoirs, nor does it usesophisticated numerical models normally required for predictions of theperformances of these nonlinear systems. The objectives of this paper areachieved through the application and intuitive comparison of the most simplemodels available for shale reservoirs. Theory Dual-Porosity Equation Formulation. Devonian shale gas reservoirs areexamples of the classic dual-porosity, fractured-reservoir model. In thisdual-porosity model, the reservoir is composed of matrix elements andfractures. Fig. 1 shows two examples of very idealized fractured reservoirs. The matrix, represented by individual blocks in Fig. 1, is a portion of thereservoir that can store large quantities of gas, but it does not have theconductivity to transport gas for long distances. The fractures, whichpartition the matrix elements, can transport the partition the matrix elements, can transport the gas but have limited storage capability Production fromdual-porosity reservoirs differs from the production of conventionalsingle-porosity systems, where gas flows directly from the formation to thewell. In fractured reservoirs, gas flows through the fracture network to thewell. The fracture network, in turn, is being constantly recharged by flow fromthe matrix elements. Given a sample control volume in a fractured reservoir(Fig. 1a or 1b), a control volume mass balance can be performed. − = 0, Ac t ..................(1) where v f = average flow velocity through the fractures with respect to anentire cross section, A c = external surface area of the control volume, qt =total mass flow rate from the matrix to the fractures within the controlvolume, and delta x delta y delta z = magnitude of the control volume. The plussign in Eq. 1 results from the assumption that flow from the matrix to thefracture is a positive quantity. For conceptual simplification, it is assumedthat density and pressure are approximately constant in the control volume. Because the total matrix recharge includes flow from all the individual matrixelements, the total matrix flow rate can be rewritten as qt = N (q ma), where qma is the average volumetric flow rate per element, and N is the total numberof elements in the control volume.