Abstract
The Horseshoe Canyon (HSC) CBM play of the Western Canadian Sedimentary
Basin is unique to low-rank coal reservoirs because of lack of water
production; the production characteristics are qualitatively similar to
conventional low-pressure dry gas reservoirs. However, the complex geological
history of the coals and non-coal interbeds has imparted strong vertical and
lateral heterogeneities that make the play difficult to characterize using
conventional methods.
Recently, advances in production data analysis (PDA) methodologies have been
made for CBM wells; techniques developed for conventional oil and gas
reservoirs have been adapted by incorporating some CBM reservoir properties.
For example, the popular flowing material balance (FMB) technique, as well as
production type-curve and pressure transient analysis (PTA) have been modified
to include relatively simple CBM reservoir behaviour (ex. equilibrium
desorption). These methods, however, are primarily restricted to the analysis
of single-layer reservoirs; significant errors in estimates of
original-gas-in-place (OGIP) and other reservoir properties may occur if strong
contrasts exist from layer-to-layer. In this work, multi-layer analysis tools
are discussed, including analytical simulators that are used to history-match
layer-allocated rates and pressures, and layer-specific FMB, which is used as a
PDA method for individual layers.
The applicability of advanced PDA methods to the quantitative assessment of
HSC reserves was investigated. Single-layer and multi-layer analysis tools were
first tested against simulated data. Next, single-layer-equivalent analysis was
performed on >40 real wells using type-curve, FMB, and analytical
simulation. Finally, a more rigorous multi-layer analysis was performed on a
subset of wells where spinner surveys and individual-seam pressure buildup data
were available. Analysis of these wells included PTA of the individual seams,
individual seam material balance, and multi-layer analytical model
history-matching of total commingled flow rates, individual coal zone rates
estimated from spinner surveys and shut-in pressures. The single-layer-
equivalent analysis appears to yield conservative estimates of OGIP compared to
the more rigorous multi-layer analysis in the cases analyzed.
Future work will include continued comparisons of multi-layer vs.
single-layer PDA, investigation of additional constraints on input reservoir
properties used in multi-layer history-matching process, and time-lapse PTA
work to quantify changes in layer permeability and skin during depletion.
© 2009. Society of Petroleum Engineers
View full textPDF
(
3,515 KB
)
History
- Original manuscript received:
29 March 2008
- Meeting paper published:
18 June 2008
- Revised manuscript received:
24 January 2009
- Manuscript approved:
29 August 2009
View Cart
