This paper discusses the application of a numerical model to investigate the effect of water coning on the sweep efficiency and ultimate recovery from gas reservoirs with active aquifers similar to those found in the Middle Devonian of north east British Columbia. Reservoir properties typical of this type of pool were selected to cover the range of properties normally found. The properties used in the numeric simulator were systematically varied to produce a complete range of reservoir performance that might be expected for a standard gas production rate. The parameters that were varied were those indicated by literature and previous studies to be factors that contribute to the coning of water into a gas well. The resulting data were analysed to yield an equation that evaluates sweep efficiency and consequently ultimate gas recovery from given values of the reservoir parameters.
The study shows that the ultimate recovery of gas from a water driven Middle Devonian gas pool in northeast British Columbia can vary from 36% to 70% of the original gas in place.
The effect of water influx into a gas reservoir has been poorly documented. Many papers deal with the problem piece meal. The Petroleum Resources Branch of the British Columbia Ministry of Energy, Hines and Petroleum Resources decided to conduct studies to gain a betterunderstanding of the gas recovery from the Middle Devonian gas fields. To this end a study was made on the Clarke Lake gas field.1 The results of that study indicated that gaswas trapped by the encroaching water and that reservoir volume cut off by coning (poor sweep efficiency) was not a significant factor in the ultimate recovery of gas.
A second study was made on the Cabin Slave Point ‘C’ gas pool.4 The purpose here was to evaluate the effect of a poorer rock type in a thinner reservoir with a stronger water drive. In this study coning of water was a significant factor in the recovery of gas and because of the strong water drive very little depletion of the reservoir took place by pressure reduction.
The results of the two studies seemed to conflict. In the case of the Clarke Lake study it appeared that more gas could be recovered by drawing down the reservoir as quickly as possible to maximize the depletion effect. On the other hand in Cabin a high rate of drawdown adversely affected ultimate recovery. It was decided that further studies should be conducted.
The prior Branch studies and a literature search indicated that overall gas recovery is a function of formation thickness, rock properties and the strength of the encroaching aquifer. It should be noted here that the manner in which the strength of the aquifer and sweep efficiency affects the ultimate gas recovery is shown in Appendix A which is derived from Reference NO. 1. Obviously a method to predict sweep efficiency is required in order that an estimation of recoverable reserves in a variety of water driven gas fields can be made.
