Analytical equations for an ideal reservoir have been used to calculate the pressure profile for modified isochronal tests. The errors made by applying the usual deliverability plots are shown to be quite high for low permeability reservoirs. Reservoir complexities are expected to aggravate the problem. It is proposed that gas well deliverability can be estimated by numerical simulation reservoir parameters obtained from single point test.
It is a requirement that deliverability tests be performed on gas wells to assess their flow potential against local pipeline pressures. One of the earlier test designs was known as the back pressure test, and the testing procedure required that wells be flowed at a number of constant rates until the bottom hole pressure became constant. Deliverability plots were then made from the stabilized rate and pressure data. Since the time required to reach stabilization is dependent on formation permeability, this test can be length for low permeability reservoirs.
In the interest or saving time and reducing the amount of flared gas, the testing procedure was changed to the isochronal test and later to the modified isochronal test. Both of these tests require the well to be flowed at a number of constant rates over equal time intervals. Stabilization of the bottom hole pressure is not required. In the isochronal rest the well is shut-in between flow periods until the pressure returns to equilibrium; whereas the modified isochronal rest requires equal time interval shut-in periods. It has been recognized that the unstabilized rate and pressure leads to an overestimation of the well's deliverability potential; therefore, a final extended rate is required to make a correction to the unstabilized deliverability plot. This correction is made by drawing a line through the extended rate parallel to the unstabilized deliverability line.
Aziz1 evaluated the dependence of the slope of the deliverability plot on the length of the isochronal intervals. He found that for laminar flow the error In the slope calculation is acceptable. Matear and Lin2 considered the same problem over the turbulent flow regime, and arrived at a similar conclusion. In both of these studies the duration of the extended rate was not addressed. Furthermore there are no guidelines reported in the literatures regarding criteria an operator can use to determine when flow rates have stabilized. A common practice in Alberta is to use 4 h for the isochronal rates end 72 h for the extended rate. It is one of the objectives of this study to show, using simulation examples, that this testing procedure can lead to serious errors. In the lowest permeability reservoir considered, the over estimation of the absolute open hole flow potential is more t.han 100%. It is expected that the error resulting from the analysis of field data will be greater than it is for the ideal case.
Examples of incorporating the well bore storage effect are used to demonstrate this point.
The other objective of this paper is to recommend the use of the well established single point drawdown and buildup teat to estimate reservoir parameters.