Long hydraulic fractures are usually required to optimize recovery from low-permeability gas reservoirs. Since these fractures can be quite expensive to create and since there is still a great deal of "art" associated with fracture design and creation, it is frequently helpful to use pressure-transient tests to determine created fracture properties. In this way, optimal fracture treatments can be developed for a given area. Fast et al. 1 and Kozik and Holditch presented examples showing the potential benefits of such an approach. In addition, Veatch demonstrated the effectiveness of combining the efforts of operations and research personnel to characterize and then to improve stimulation treatments in specific areas.
The specific task of the engineer is to estimate propped fracture length and effective fracture conductivity for treatments in a given geological formation, If the reasons for success or failure of previous fracture treatments can be determined, the engineer can then do a better job in the future when designing fracture treatments for the same area.
There are currently four basic techniques used to analyze postfracture pressure-transient tests:
semilog (pseudoradial flow) analysis,
square-root-of-time (linear flow) analysis,
type-curve analysis. and
reservoir simulator history matching.
The strengths and weaknesses of each technique were discussed by Lee and Holditch. No existing technique is without problems or possible ambiguity in some applications; thus, there is a need for still other techniques that may succeed in some situations for which existing techniques are inadequate.
The purpose of this paper is to introduce such a new technique for analyzing postfracture pressure-buildup tests. This technique can be particularly helpful when analyzing data from wells in which a finite-conductivity fracture has been created.
The pressure-buildup test analysis technique proposed in this paper requires that the analyst prepare a semilogarithmic graph and a square-root-of-time graph of the test data. The technique also requires use of two correction curves developed from the analytical solutions for finite-conductivity fractures presented by Cinco-Ley et al. The analyst chooses a single straight line on each of the graphs prepared for the test data and solves for fracture length, formation permeability, and fracture conductivity with an iterative procedure.
As a first step in understanding the basis for the proposed method, consider Fig. 1.