Long term transient pressure from a permanent down-hole gauge (PDG) is a combined response of the reservoir and many other sources. These include interference from other wells nearby, multi-phase flow and skin changing along with the production process, which generated a non-linear pressure signal, making well test analysis a challenging task for engineers.

If the transient pressure is non-linear, the traditional well testing principles, based on linear reservoir system, such as superposition and deconvolution are no longer valid for the analysis. In this case, engineer has three options in solving such well test problem: (1) Linearising the nonlinear problem, for example, using so called pseudo-pressure approach (2) Choosing a specially designed approach for this problem such as numerical well testing (3) Identifying or diagnosis for the cause given the nonlinearity of the reservoir response, then decompose the response to separate the linear and nonlinear components. Thereafter each of these can be analysed using the corresponding methods respectively.

This paper presents primary study results on the last option (3) mentioned above. Two cases causing the non-linear reservoir response were investigated. One is the changing skin problem. Another is the PDG pressure influenced by interference.

The study has demonstrated that before proceeding for any analysis, the diagnosis of the nonlinearity is essential to ensure the correct selection of the analysis method, particularly for handling long term transient data from a PDG. Examples include the changing skin case, if the resulting data are analysed using deconvolution algorithm with predefined constraints; the nonlinear response of the reservoir system can be smoothed out completely, leading to the wrong interpretation. While transient pressure combined with interference, deconvolution algorithm can't be applied at all.

This nonlinearity diagnostic key step should be included in well testing best practice of the industry.

You can access this article if you purchase or spend a download.