Summary
The unique properties and complex characteristics of coalbed methane (CBM)
reservoirs, and their consequent operating strategies, call for an integrated
approach to be used to explore for and develop coal plays and prospects
economically. An integrated approach involves the use of sophisticated
reservoir, wellbore, and facilities modeling combined with economics and
decision-making criteria.
A new CBM prospecting tool (CPT) was generated by combining single-well
(multilayered) reservoir simulators with a gridded reservoir model, Monte Carlo
(MC) simulation, and economic modules. The multilayered reservoir model is
divided into pods, representing relatively uniform reservoir properties, and a
“type well” is created for each pod. At every MC iteration, type-well forecasts
are generated for the pods and are coupled with economic modules. A set of
decision criteria contingent upon economic outcomes and reservoir
characteristics is used to advance prospect exploration from the initial
exploration well to the pilot and development stages. A novel approach has been
used to determine the optimal well spacing should prospect development be
contemplated. CPT model outcomes include a distribution of after-tax net
present value (ATNPV), mean ATNPV (expected value), chance of economic success
(Pe), distribution of type-well and pod gas and water production, reserves,
peak gas volume, and capital. An example application of CPT to a hypothetical
prospect is provided.
An integrated approach also has been used to assist with production
optimization of developed reservoirs. For example, an infill-well locating tool
(ILT) has been constructed to provide a quick-look evaluation of infill
locations in a developed reservoir. ILT, like CPT, is used for multiwell
applications, combining the single-well simulator with a multilayered reservoir
model and economics. An application of ILT to a CBM reservoir is provided, and
the results are compared with the predictions of an Eclipse reservoir
simulation.
Introduction
CBM reservoirs have a relatively short history of development compared to
conventional reservoirs; therefore, few analog fields may be relied upon for
extrapolation to new basins and new plays. Further, key reservoir properties
such as absolute permeability vary greatly within and between existing
developing basins, which complicates prediction of these parameters for new
plays. The production performance of CBM reservoirs in new plays or basins, in
which few reservoir data exist, is correspondingly difficult to predict.
Existing conventional reservoir fields cannot be relied upon as analogs for
CBM play analysis because coal-gas reservoirs differ from conventional
reservoirs in their fluid-storage and -transport mechanisms. Coals act as
source rocks and reservoirs to gas, and a significant amount of gas may be
stored in the adsorbed state relative to the free-gas state. CBM reservoirs are
often naturally fractured and may be modeled as dual-porosity, or even
triple-porosity, reservoirs. Gas-transport mechanisms vary depending on the
scale and location within the reservoir. For example, gas transport at the
scale of the matrix between natural fractures is caused by the mechanism of
diffusion, whereas Darcy flow occurs in the fracture system. Single- or
two-phase (gas and water) flow can occur, and, hence, relative permeability
characteristics are important.
Permeability and gas content are two critical parameters that dictate the
economic viability of CBM reservoirs. Unfortunately, there are many controls
upon these parameters. For example, gas content is a function of the amount of
organic matter within these rocks, the organic matter composition, organic
matter thermal maturity, in-situ PT conditions, gas composition, and matrix and
fracture gas-saturated porosity. Absolute permeability is dependent upon
natural-fracture existence, frequency, orientation (with respect to in-situ
stress), and degree of mineralization. Natural-fracture permeability is also
stress- and/or desorption-dependent. Although the range of expected parameter
values for a new unconventional play may be reduced by knowledge of basin
hydrodynamic characteristics, tectonic regime, local and regional stratigraphy
and sedimentology, local and regional structural geology, and existing
production within the basin, the uncertainty associated with key reservoir
variables is still likely to preclude a deterministic evaluation of reservoir
producibility and recoverable reserves.
Because of the variability in reservoir parameters that could be expected
when exploring for CBM in existing or new basins, it is natural to use a
statistically based (stochastic) approach in the prediction of gas in place,
recoverable reserves, well performance, and economic return. A comprehensive
study by Roadifer et al. demonstrated the use of MC simulation for screening
key parameters affecting CBM production.
Well performance is a key factor determining the economic viability of CBM
reservoirs. Accurate prediction of well performance is required for development
strategies such as optimized well spacing, completion gathering system, and
wellsite design.
The current work discusses how to integrate reservoir simulation and
economics for the purpose of optimizing CBM exploration and development
strategies. Central to the discussion is the use of single-well (multilayered)
simulators, which were constructed in Excel* and incorporate many attributes of
CBM reservoirs. These single-well (tank) models are discussed in the following
section and have some utility for exploration and development applications when
used on their own, but they are particularly powerful when integrated with
reservoir, surface, and wellbore models, MC simulation,7 and economics. Two new
tools (CPT and ILT) described in this work are examples of integrated tools for
application to exploration and development, respectively.
© 2005. Society of Petroleum Engineers
View full textPDF
(
1,302 KB
)
History
- Original manuscript received:
21 April 2004
- Revised manuscript received:
6 June 2005
- Manuscript approved:
11 August 2005
- Version of record:
15 December 2005
View Cart
