There are many coalbed methane reservoir simulators in the market. The two most important characteristics of these models that, for all practical purposes, put them out of the reach of most of the independent producers are their price and complexity of use. In order to effectively use these commercial simulators, the average CBM producer must have at least one dedicated engineer, preferably with a graduate degree, on the staff that is willing and capable of performing effective modeling and simulation. It seems that the commercial simulation packages have been designed for majors, large independents and consultants. This leaves the majority of independent producers that are in the business of CBM production in a class by themselves. The model that is presented in this paper targets this class of CBM producers.

A new CBM reservoir simulator has been developed at West Virginia University in order to address this specific need of our industry, specifically in the northeast of the United States where new interests in coalbed methane have been growing steadily. The WVU CBM model uses a previously developed legacy code by King and Ertekin at PSU as its point of departure. The nearly twenty year old King & Ertekin's formulation which has been verified and proven to be robust, forms the foundation of WVU CBM model. The WVU CBM model is PC-Windows based and has a user friendly and graphical interface. The PSU model used a commercial solver that if was used in this model it would have made it prohibitively expensive for the independent producers. A new solver was developed and is used in this model. The new WVU CBM model is a single-well radial model with a graphical input interface that includes default values for the Pittsburgh coal for all parameters in case some of the values are unavailable. It generates graphical outputs in form of gas and water production rate vs. time, cumulative production vs. time, pressure, saturation and gas desorption distribution throughout the reservoir. It allows for production scheduling for detail history matching process. It lets the user to plan hundreds of runs in the form of a batch file that can be run over night.

The model has a unique feature that allows user to provide input parameters in the form of probability distribution functions with minimum, maximum, and, if they choose, most likely values instead of crisp numbers. The output in such cases will be probability distribution functions rather than crisp numbers. The WVU CBM model will be available to the industry through our web sites.

You can access this article if you purchase or spend a download.