Petrophysical heterogeneity of the early Cretaceous Alamein dolomite reservoir from North Razzak oil field, Egypt integrating well logs, core measurements, and machine learning approach

• Published in: Fuel (Guildford) Vol. 306; p. 121698
• Main Authors: Sen, Souvik, Abioui, Mohamed, Ganguli, Shib Sankar, Elsheikh, Ahmed, Debnath, Akash, Benssaou, Mohammed, Abdelhady, Ahmed Awad
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 15.12.2021; Elsevier BV
• Subjects: Alamein dolomite; Algorithms; Carbonate rocks; Core analysis; Cretaceous; Dolomite; Drilling; Heterogeneity; Intervals; Learning algorithms; Machine learning; Macroporosity; Microporosity; Oil; Oil and gas fields; Oil fields; Performance measurement; Permeability; Petrophysical characterization; Pores; Porosity; Prediction models; Pressure drop; Quality assessment; Reservoir quality; Reservoirs; Storage capacity; Water storage; Well logs; Egypt; Heterogeneity; Alamein dolomite; Reservoir quality; Petrophysical characterization; Machine learning
• ISSN: 0016-2361; 1873-7153
• DOI: 10.1016/j.fuel.2021.121698

•Petrophysical attributes of the Alamein dolomite reservoir are characterized.•Machine learning approach by Random Forest Regression delivered better permeability prediction than conventional methods.•A new petrofacies zonation scheme is proposed in the heterogenous reservoir.•Inferences are drawn on oil production and optimum drilling-completion strategy. Capturing the petrophysical heterogeneities within a reservoir has a critical influence on reservoir deliverability as well as field development programs. In this study, we report a comprehensive petrophysical evaluation of the oil-producing Aptian Alamein dolomite reservoir from the North Razzak field, Western Desert of Egypt. Integration of wireline logs and routine core analysis indicates that the Alamein reservoir has an extremely wide range of porosity (1–23%) and permeability (0.01–7000 mD), contributed by the early diagenetic dolomitization history and complex distribution of vugs. Petrophysical assessment by reservoir quality index (RQI) and flow zone indicator (FZI) infers that the megaporous rock types offer very good to excellent reservoir qualities and macroporosity dominated intervals are of fair to good quality. Further, we developed a permeability prediction model in this challenging carbonate rock based on Random Forest (RF) regression, and tested its efficacy and generalizability by well-defined performance metrics. The RF-based algorithm provided a more confident permeability prediction (R2 = 0.937) compared to conventional methods. Based on the petrophysical attributes; six distinct petrofacies (PF) associations are identified. PF-1, PF-3, and PF-5 provide excellent reservoir qualities with superlative storage capacity and hydraulic flow potential contributed by connected vugs, while the microporosity-dominated impervious PF-2 and PF-4 intervals act as intra-reservoir permeability barriers. We suggest that the higher initial oil production rate was mainly contributed by the larger connected pores and vuggy spaces. As reservoir pressure drops, hydrocarbon flows restrict to the smaller pores causing accelerated production weakening. Based on this comprehensive analysis, a suitable drilling and completion strategy is recommended for the future reservoir development program.