Abstract

It wasn't always this complicated! In 1990, Seidle and Arri1 demonstrated how to easily adapt a conventional black oil simulation model for use in coalbed methane simulation. These days, when setting up a coalbed methane simulation model, one needs to worry about a single or multi-component gas description; depletion or enhanced coal bed methane recovery; single, dual or triple porosity; vertical, horizontal, or multi-lateral wells; coals only or a mix of coals and sands; and coalbed methane or coal mine methane? Correctly determining what to model is almost as daunting a task as the simulation work itself. All of these variations add complexity to the task and, in some cases, require specialized simulation models to adequately characterize the problem.

This paper discusses how the exploitation and development of coalbed resources throughout the world is changing, and with it how our approach to reservoir simulation of the process is changing as well. The paper will provide a valuable resource to engineers and geoscientists faced with developing predictive tools to assist them in evaluating the optimum strategy to exploit these valuable resources.

Introduction

There is probably not a more appropriate place to start a discussion on the evolution of coalbed methane reservoir simulation than to point out that one really doesn't need a coalbed methane reservoir simulator to perform coalbed methane reservoir simulation. Any conventional black oil simulator will do the job.

The idea of modifying a conventional black oil model to simulate the performance of coalbed methane wells was first presented by Amoco¹. The technique is quite simple. One first initializes the model with a small immobile oil saturation. The magnitude of the oil saturation is not important, however, it is important that the oil be immobile and the flow of the other fluids not be impacted. To accomplish this, the porosity and fluid saturations (gas, water) are adjusted accordingly based on the oil saturation:

• Equations

In the above notation, ‘eff’ refers to the modified values, while ‘act’ refers to the original values. The impact of the above modifications is to adjust the pore volume to maintain the proper initial fluid volumes, and to shift the fluid saturations to maintain the original relative permeability relationships.

After the above modifications, the only remaining change required is to supply an effective dissolved gas relationship with pressure which will mimic the gas content isotherm one would normally use to describe a coalbed methane system. In essence, the dissolved gas in the oil replaces the Langmuir isotherm function. This is done via:

• Equation

In the above equation, ‘V’ is the original coal gas content at any given pressure expressed in Scf/ft³ and the oil formation volume factor is normally set equal to 1.0.

As was pointed out in the original reference, the modified black oil representation works well if the release of gas from the matrix to the cleats is fast compared to the flow of gas and water in the cleats. This is because the modified black oil technique implicitly assumes that the sorption time is instantaneous. Potential problems can occur if the actual sorption time in the coals is unusually long, or if the permeability of the cleat system is extremely high.