Summary

Well test analysis has been used for many years to assess well condition and estimate reservoir parameters. The conventional interpretation methods (straight lines, log-log pressure) plots were limited to the estimation of well performance by estimating permeability, skin and drainage radio. With the introduction of pressure-derivative analysis in 1980’s well test become a powerful tool for reservoir characterization with the development of complex interpretation models that are able to account for detailed geological features such as heterogeneities, different flow geometries and boundary effects. In the early 2000’s the deconvolution was introduced in the technical literature with the aim of maximizing the information provided by well test analysis trough the extension of the investigated reservoir volume.

The data collected at the well bottom hole may contain information about the reservoir at larger radius of investigation than that accessible to conventional derivative analysis, which is limited to the interpretation of single flow periods at constant rate. In other hand, deconvolution methods do not suffer from this constraint because they are designed to perform well test analysis at variable flow rate. In this work different case study synthetic model were generated by means of Eclipse where reservoir and fluid (dead oil) properties were provided. Several well test scenarios were simulated “considering short shut-in periods” with the aim of comparing the result provided by conventional Pressure Transient Analysis and the potential additional information obtained by deconvolution technique. Moreover, initial reservoir pressure was varied and random error was added to the simulated data to evaluate the behavior of deconvolved response and test the performance of the deconvolution algorithm respectively.

From the results, well test interpretation using deconvolution technique reveals characteristic feature which are not evident in derivative analysis. Furthermore, deconvolution shows to be a good tool to deal with error. However, it is very sensible to changes of well behavior and initial reservoir pressure that can lead to the erroneous results.

Keywords: Reservoir parameters, Pressure Transient Analysis, Deconvolution, Reservoir Characterization, Reservoir Boundary.