We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Department of Mathematical Sciences


Publication details for Jonathan Cumming

Tung, Y, Virues, C, Cumming, J A & Gringarten, A C (2016), Multiwell Deconvolution for Shale Gas, SPE Europec featured at 78th EAGE Conference and Exhibition. Vienna, Austria, SPE.

Author(s) from Durham


In the last decade, single well deconvolution (von Schroeter et al., 2001) has become recognised as a powerful tool for reservoir characterization. Deconvolution transforms well test pressure data measured at varying rates into an equivalent unit rate single drawdown which has the same duration as the pressure history and can be analysed using conventional techniques to identify boundaries and assess connectivity between different compartments and layers. As its name suggests, single well convolution is only applicable when there is no interference from other wells. i.e. to exploration, appraisal and isolated production wells.

In 2013, a multiwell deconvolution algorithm was presented (Cumming et al., 2013) which enables deconvolution to be applied to groups of interfering wells. The algorithm yields a unit rate deconvolved derivative for every well, representing the well signature with interferences removed; and the interference derivatives between well pairs. An example of use with eight interfering North Sea oil wells was presented by Thornton et al. (2015).

This paper describes the application of the Cumming et al. multiwell deconvolution algorithm to eight wells in a multilayer shale gas reservoir. The main objective of the study was to verify the capability of the multiwell deconvolution algorithm to remove the effects of interference between horizontal wells and to assess the efficiency of the actual well spacing, which had been intended to avoid interferences. This study can have an impact on upcoming LNG projects in Canada and other countries which are evaluating LNG prospects.

An additional objective was to identify an optimal testing procedure to minimise uncertainties on the deconvolved derivatives. This was done by numerically simulating the behaviours of a number of interfering, multiply-fractured, horizontal wells (from 2 to 8) for various rate sequences. The impact of including non-interfering wells in the multiwell deconvolution was also investigated.

It was found that the accuracy of the multiwell deconvolution process was very much dependent on the rate sequence: producing and shutting in wells sequentially provides better results that having the same rate sequence on all the wells. In addition, to minimize errors on the derivatives, the start of interference effects should be evaluated correctly and non-interfering wells should be excluded from deconvolution. Both require a good understanding of the reservoir.