Abstract

In recent work, the authors (Clarkson et al 2008a; Clarkson et al 2007; Jordan et al 2006) demonstrated how modern production data analysis (PDA) methods, such as flowing material balance (FMB) and production type-curves, may be adapted to account for the unique reservoir characteristics of coalbed methane (CBM) reservoirs through the appropriate use of material balance and time transforms. Reservoir characteristics related to storage and fluid flow that were addressed included: adsorbed and free-gas storage; single-phase flow of water above desorption pressure (undersaturated coals); 2-phase flow of gas and water below desorption pressure (saturated coals); non-static absolute permeability during depletion; and multi-layer behavior. Example (field) applications of the new PDA methods were limited to vertical wells that were either openhole completed, or slightly stimulated with hydraulic fracturing methods.

In this work, new workflows and analytical approaches are provided for analyzing single- and multi-phase flow of CBM from vertical, hydraulically-fractured wells and horizontal wells. The workflow for single-phase CBM wells, which parallels that used in modern pressure transient analysis (well-testing), includes identification and analysis of flow regimes, followed by (model) type-curve matching and analytical (or numerical) modeling. Pseudotime is modified to account for instantaneous desorption. The analysis and methodology for 2-phase flow reservoirs is more complex, requiring additional modification of the pseudovariables to account for changes in effective permeability. A new workflow is introduced for 2-phase CBM wells that includes the transformation of the well production and flowing pressure data into dimensionless type-curve and flowing material balance coordinates using outputs from the simulator used to history-match the production data. Both simulated and field cases are analyzed to illustrate the new procedures and analytical techniques.

The primary contribution of the current work is the demonstration that modern production analysis methods, modified for CBM reservoir behavior and combined with analytical (or numerical) modeling, can be used to extract quantitative reservoir information from CBM reservoirs which exhibit a wide-range in production characteristics, and completed in a variety of styles. The methods are expected to be used as a pre-cursor to or in parallel with reservoir simulation, to assist with CBM development decisions.

Introduction

The most abundant data collected for coalbed methane (CBM) reservoirs is gas and/or water production data, so it is logical to maximize the amount of information obtained from those data. Significant advances in the analysis of production data from conventional reservoirs have allowed the engineer to extract key reservoir properties such as permeability-thickness product (kh) and hydrocarbon volumes-in-place, and stimulation information such as skin factor (s) or hydraulic fracture half-length (xf) and hydraulic fracture conductivity (kf wf) from fluid production and flowing pressure information. The analytical methods that have been developed for production data analysis (PDA) over the past several decades can be categorized primarily as:

  1. Type-curve methods (pressure-transient analysis analog): These methods involve the matching of production data (corrected in some cases for variable flowing pressure) to analytical/empirical solutions of flow equations, cast in dimensionless variable format. Much research has been performed in recent years on the use of pseudovariables and superposition time functions to allow real data to be transformed onto type-curves developed for constant-rate or constant-flowing bottomhole pressure production of slightly compressible fluids. Additionally, the type-curve solution set has been expanded to account for modern well geometries and completion techniques [Select references include: Fetkovich 1980; Fraim and Wattenbarger 1987; Palacio and Blasingame 1993; Agarwal et al 1999; Pratikno and Blasingame 2003; Ilk et al 2008a].

This content is only available via PDF.
You can access this article if you purchase or spend a download.