This article, written by Technology Editor Dennis Denney, contains highlights of paper SPE 102699, "Boundary Effects and Depth of Investigation From Long Buildups Following Short Flow Tests," by L. Larsen, SPE, Statoil ASA, prepared for the 2006 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 24-27 September.
Slug tests and other tests for which the buildup period greatly exceeds the flow period have been addressed in the literature by several authors over the last 30 years. This paper examines such tests to investigate the possibility of drawing conclusions about boundary effects.
Although some of the standard techniques used to analyze slug tests can, in principle, detect boundaries at distances far beyond the distance affected by the flow, special approaches must be used to avoid the need for a highly accurate estimate of the initial formation pressure. Often, the pressure change relative to the initial formation pressure is used in analyses. For cases with small pressure changes at the end of the buildup, the characteristics of the pressure response are highly sensitive to the choice of formation pressure, making it difficult to distinguish boundary effects from effects of erroneous input.
Straightforward derivative methods can be used to analyze buildup data following short flow tests without the initial formation pressure as a key input parameter. The full-length paper details the use of two approaches to analyze the data and reduce the uncertainty in the results. The methods use different data sets, but both require high-quality data to reduce uncertainty and maximize the depth of investigation. Results from well-known derivative methods deviate to some extent from expected results. The derivative approach does not work for all types of models in a theoretical sense, but it does work for all cases of practical interest. The results apply to any data set with a shut-in period that is much longer than the flow period—for instance, with days of flow and months of shut-in data. Gauge resolution is particularly important in analyses to identify effects of distant boundaries. Noise in the pressure data and disturbances in the data from wellbore effects also are critical.
For successful analysis, the method is highly important when late-buildup data that approach the initial pressure are used. Pressure differences relative to an estimated initial pressure normally are used in analyses, and such analyses can be highly sensitive to the accuracy of this parameter. Because the initial pressure is not known exactly, alternative methods that use only measured data are advocated here.
When drawing conclusions about observed boundary effects and the radius of investigation in general with or without boundary effects, the pressure derivitives become critical. A high estimate of the initial pressure in a standard slug-test analysis can cause late data to suggest the presence of a no-flow boundary. A low estimate of the initial pressure in such an analysis can cause late data to suggest the presence of a constant-pressure boundary. The deviating trend also can be taken as an indication of an erroneous value for the initial pressure, but only if it is known that no boundary effects exist.