Evaluation of PVT Comparisons and GOR Prediction Based on Advanced Mud Gas Data: A Case Study From Snorre Field

The use of real-time advanced mud gas (AMG) analysis in oil and gas exploration has been well known for at least 15 years. In recent years, its application in production fields has become more prevalent. The advances in mud gas extraction and analysis techniques combined with directional drilling tools provide more confident data, allowing the identification of formation vs. injection fluids, as well as movable contacts. This analysis during the drilling process results in more informed decisions during drilling, aimed at maximizing production value Over the last decade, the field of advanced gas analysis has developed a method to correct constraints upon the varying efficiency of extraction for specific species of interest associated with a constant volume, constant temperature mud gas extractor. This method has become recognized in the industry by the name extraction efficiency correction (EEC). EEC provides the quantitative composition of formation fluid in real time by analyzing methane through pentane components. These results are comparable to downhole pressure-volume-temperature (PVT) samples and have been used to optimize wireline tool runs and fluid sampling programs over the years, but this application is not limited to exploration wells. A technique presented by Yang et al. (2019a) has demonstrated the potential of machine-learning predictions, like gas-oil ratio (GOR) using EEC compositional data. More recently, results shared by Kopal et al. (2022) demonstrate the optimization of reservoir analysis when integrating petrophysical with the GOR prediction. This paper presents a case study based on the analysis of several zones in development wells of the Snorre Field in Norway. The consistent dynamic EEC data from AMG are demonstrated to successfully distinguish the types of fluids compared to PVT samples in the field. These data are presented as continuous logs, allowing for the evaluation of reservoir zone thickness. This information is available while drilling; with modern real-time data services, operators can access it from almost anywhere. The model of GOR prediction proposed by Yang et al. (2019a) is applied by the operator to this Snorre example. The results are demonstrated in the paper when compared to those of GOR production. Promising results from machine learning generate confidence in the applicability of quality AMG data to development wells for real-time petrophysical and operational decisions. The field case demonstrates a new and broad application area for advanced mud gas in production wells.