Abstract
Abstract
This paper presents evidence that the unipore model is inadequate for describing diffusional fluxes from coal over the entire timescale of desorption. Field desorption of methane from coal pieces shows pronounced curvature, which is attributable to the bidisperse pore structure generally found for coat. A model is presented that accounts for the bidisperse pore structure of coal and predicts diffusion rates for both field and laboratory predicts diffusion rates for both field and laboratory desorptions. Transient diffusivities are determined for Pittsburgh bituminous and Madrid (NM) anthracite coals Pittsburgh bituminous and Madrid (NM) anthracite coals by using a pulse tracer technique coupled with Fourier analysis of the elution curves. This technique allows the rapid determination of the diffusion parameters with minimal experimental effort. In addition, steady-state diffusivities are determined to identify mechanisms of diffusion. Diffusion was found to be a combination of bulk, Knudsen, and surface diffusions, depending on the coal pore structure and gas pressure.
Introduction
Over the last decade, increased consideration has been given to coalbed methane as a resource, a result of rising overhead gas paces and declining U.S. conventional natural gas reserves. Recent estimates for the quantity of recoverable coalbed methane in the continental U.S. range from 30 × 1012 to 200 × 1012 cu ft [0.85 to 5.7 × 1012 m3]. Although coalbed methane is produced with conventional drilling and completion produced with conventional drilling and completion technology, the actual gas flow mechanism is quite different from conventional formations. The transport of methane through coal generally is assumed to be a two-stage process:diffusion of methane through the pore structure of the coal to naturally occurring cracks followed bythe simultaneous flow of water and gas through the crack structure.
The relative magnitude of these two steps has been studied by several investigators. However, these investigators, including Kissell and Kuuskraa et al., ignored the coupling of the diffusional and laminar flow processes. Also, both investigators used a "unipore" diffusion model, which we found inadequate for describing methane diffusion in coal. The major conclusions of both works were that pore diffusion could be ignored and that the production of methane from coal is controlled solely by the laminar flow of water/gas through the coal crack structure. Ancell et al. have acknowledged correctly that diffusion effects may not be neglected a priori. Diffusion effects were included by using the unipore diffusion model coupled to Darcy law equations for gas and water flow through the cracks. Although the mechanism of pore diffusion is included, no mention is made concerning the effect of diffusion on the gas/water production rates. production rates. The diffusion of an adsorbing gas through a porous solid may occur as a result of any combination of three mechanisms. These mechanisms are bulk diffusion (molecule/molecule interactions dominate), Knudsen diffusion (molecule/pore wall interactions dominate), and surface diffusion (transport of the adsorbed liquid-like film). To extrapolate findings safely from laboratory diffusion experiments to the high pressures associated with coalbed methane, it is important to have an adequate understanding of these mechanisms. These mechanisms have been identified for several coals and are presented in a following section. In addition, a model alternative to the unipore model has been found that describes desorption from coal over the complete time frame of interest. A straightforward experimental procedure is presented for determining the necessary procedure is presented for determining the necessary diffusion parameters for this model.
Methane Diffusion in Coal
Past investigations of methane diffusion through coal Past investigations of methane diffusion through coal have used two basic types of experiments. One technique is to flow methane or helium through a solid coal disk. With the measured pressure drop and flow rate, a permeability constant (and hence diffusion coefficient) permeability constant (and hence diffusion coefficient) may be obtained. In the second, the rate of desorption from small coal particles after undergoing a step-change in external concentration is used in conjunction with an appropriate model to determine a diffusion coefficient. Thimons and Kissell used the flow disk method for the study of methane and helium diffusion in three different eastern U.S. coals. The study at ambient temperatures and pressures ranging from 0.7 to 2.7 atm [70 to 270 kPa] resulted in the following conclusions.1. Surface transport of adsorbed methane is negligible compared to gas-phase transpose.2. Steady-state diffusion coefficients are on the order of 10–4 to 10–5 cm2/s.3. Transient diffusion coefficients ranged from 0.5 to 10 times the steady-state values. Sevenster conducted a similar study for a variety of gases on a British coal. In contrast to findings of Thimons and Kissell, a large surface transport contribution for methane was observed.
SPEJ
P. 529
Publisher
Society of Petroleum Engineers (SPE)
Cited by
101 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献