Post-well stimulation allocation of commingled production using geochemical fingerprinting techniques in unconventional reservoirs: A review of methods, and a case study of the Montney Formation, Western Canadian Sedimentary Basin

Abstract

The methods used in allocating commingled production in conventional reservoirs are similar to those that are effective in performing the same task on unconventional reservoirs. However, the protocols to follow can vary a great deal. The presence of distinct endmembers in the former allows the use of the method of production allocation using peak height ratios and mixing curves and using linear regression of peak heights. Due to the possible contribution from multiple intervals in the same formation or even from different formations as in the case of fracture stimulated unconventional reservoirs, the combined application of methods that compare the quantities and carbon stable isotopes of selected compounds (such as saturate and aromatic hydrocarbons) and other parameters (such as API gravity) was employed. This was done based on a series of samples presumed to represent the endmembers via their HRGC and GCMS oil fingerprint, followed by the determination of the contribution from each sample by using an algebraic solution of simultaneous linear equations. A review of the two methods is provided.The aforementioned method for unconventional resources is demonstrated in a case study of production allocation that was performed on three produced oils sampled at different times from three separate wells, “A” “C”, and “D”, located in the Western Canadian Sedimentary Basin. A total of 25 core extract samples representing two producing zones (end members) of the Montney Formation (i.e., the Middle and the Lower Montney) from well “A” and “B” were used. Results of GC and GCMS analyses of the samples were evaluated; rigorous filters, cluster analysis (dendrograms), and Principal Component Analysis (PCA) were applied to identify any clustering or variation between the samples representing possible contributor layers and the commingled oil. Then, using proprietary software and statistical techniques, the fingerprint of selected compounds was qualitatively compared and their quantity in each of the rock extracts and the produced oils was determined in order to allocate the contribution from the two end members that each extract belongs to.Results from the method for unconventional resources were compared to other data (such as GC trace patterns) for consistency. The case study demonstrates that a combined approach that accounts for the entire fingerprint (i.e., GC and molecular markers (including biomarker and non-biomarker parameters)), produces the best results and minimizes uncertainty.