Novel Method of Production Back-Allocation Using Geochemical Fingerprinting

Abstract

AbstractGeochemical fingerprinting using gas chromatography techniques is a proven alternative or additional tool to traditional approaches for the production back-allocation such as metering or production logging tools. It can be applied in various scenarios, from commingled reservoirs in a single well to allocation of multiple wells or entire fields produced via the same evacuation system. The approach is fast, cost-effective and does not require interruption of production, thus enabling frequent monitoring of production. The method is based on detailed comparison of fluid compositions obtained from gas chromatography of representative samples acquired from the point of interest (single reservoir, well, etc.), called further the ‘end-member’ and the ‘commingled fluid’ to be allocated. Production allocation using a geochemical fingerprinting approach has been successfully used across the globe with specific traction in North America, the North Sea region and the Middle East.Our method is based on analysis of ratios of heights of neighboring chromatographic peaks (compounds) rather than the single peak heights or areas that all the chromatograms have in common. Such approach reduces inconsistencies between light and heavy hydrocarbons due to some problems of reproducibility during the sampling or during the analysis. It also allows us to tackle issues related to the changes in compositions of end-members during production. In addition, the resolution manages the non-linearity of the equations derived from the physics of the mixtures. The non-Gaussian distribution of the errors is taken into account to comply with the maximum likelihood. Thus, a solid theoretical framework is established to avoid current issues encountered when peak ratios are utilized. Benefits of this method include firstly, a complete management of the uncertainties on the proportions of end-members and on each individual peak ratio employed. In addition to minimization of ‘calibration’ lab mixtures, elimination of manual peak selection (sometimes subjective). Finally, with this methodology employed heir in there is theoretically, no limitation on the number of end-members.In this paper we demonstrate our approach applied successfully on a series of case studies including biodegraded oils and ‘annoyingly’ similar fluids. We demonstrate that our approach can be successfully and cost-effectively applied to allow for more reliable reservoir/field management.