A Machine-Learning Approach to Predict Gas-Oil Ratio Based on Advanced Mud Gas Data

Mud gas data from drilling operations provide the first indication of hydrocarbons in the reservoir. It has been a dream for decades in the oil industry to predict reservoir gas and oil properties from mud gas data because it would provide knowledge of the reservoir fluid properties in an early stage, continuously for all reservoir zones, and at low costs. Previous efforts reported in the literature did not lead to a reliable method for quantitative prediction of the reservoir fluid properties from mud gas data. In this paper, we propose a novel approach based on machine learning that enables us to predict gas-oil ratio (GOR) from advanced mud gas (AMG) data. The current work is based on a previous successful pilot in unconventional (shale) reservoirs to predict GOR from the C1 to C5 compositions present in AMG data. We aim to extend the results of the pilot study to conventional reservoirs. In general, the prediction of reservoir fluid properties is more challenging for conventional reservoirs than for unconventional reservoirs due to the complexity of petroleum systems in conventional reservoirs. Instead of building a model directly from AMG data, we trained a machine-learning model using a well-established database of reservoir fluid with more than ,000 pressure-volume-temperature (PVT) samples. After a thorough investigation of the compositional similarity of C1 to C5 between PVT samples and AMG data, we applied the model developed from PVT samples to AMG data. The predicted GORs from AMG data were compared with GOR measurements from corresponding PVT samples to assess the accuracy of the GOR predictions. The results from 22 wells with both AMG data and corresponding PVT samples show a large agreement between prediction and measurement. The accuracy of the predictive model (MAPE 35%) has a significant improvement compared to the results from existing empirical correlations reported in the literature (MAPE 60%). In addition, a quality check (QC) metric was developed to efficiently flag low-quality AMG data. The QC metric is vital to give a confidence level for GOR prediction based on AMG data when PVT samples are not available. This study confirms that AMG data can be used as a new data source to quantitatively predict continuous reservoir fluid properties in the drilling phase for conventional reservoirs. The new method can provide early access to reservoir fluid distribution for optimization of wireline operations for exploration wells. GOR prediction based on the new method can be used for real-time decisions on well placement, perforation, completion, and potential sidetracks for production wells. After high-quality PVT data become available from the wireline logging or production phase, the continuous GOR prediction can be further improved and used for reservoir management and production optimization.