The method presented here can be used to analyse a build-up, a drawdown or a test encompassing one or more combinations of flow and shut in periods. The method is based on the general flow equation for radial flow. To account for variable flow, the principle of superposition was applied to the equation. A matching procedure is used to approximate performance. The data required to interpret a tandard test is used. In particular, this paper describes ways in which the method appears successful in interpreting and predicting future performance of wells in a low permeability reservoir. It is difficult or may be impossible to employ the standard build-up and AOF tests on such reservoirs. The method is illustrated through several field examples, some utilizing test data on wells completed in the Milk River and Medicine Hat zones of Southeastern Alberta.
Over the last few years, the development of shallow gas zones in Southeast Alberta has created new challenges in engineering. These zones are of low permeability, often less than a millidarcy, but may still be economical to develop. The problem is not to find the gas, as general maps have been prepared showing its distribution. Rather, the main challenge is to determine whether the gas will flow at rates that yield satisfactory return on investment if put on production.
In Alberta, the modified Isochronal back pressure test is generally well accepted both by industry and the Energy Resources Conservation Board. It normally provides information which can be used successfully to forecast rates from a gas well. However, the method has a serious drawback when applied to a well completed in a zone of low permeability. One of the requirements of the test is that the extended flow period should be long enough for the well to reach stabilization. Such time in these cases may be in the order of a year or even more. Obviously, this would be both wasteful and impractical. To deem an extended flow test stabilized before it actually is, would only result in rate predictions that would be too optimistic. Figure 1 presents data from a Medicine Hat well that was flow tested for five days. The plots clearly indicate that the flow is in a transient state. Both pressure and flow rate are declining at a significant rate. A method which can interpret the test during this period and also be used to forecast future production will be outlined.
The well should be cleaned by blowing it to atmosphere. This is often necessary to get rid offrac fluids remaining in the formation after a stimulation. The time required should be recorded and used as part of the interpretation.
An extended flow period of 3 to 7 days may be necessary to get sufficient data. The reason for this is that frac jobs often are very successful in drastically increasing initial flow rates.