Integrating Geomechanical Modeling and Production Data for Decision Support in Deep Gas Reservoirs in Oman

Abstract

Geomechanical modeling of hydraulic fracturing in deep gas reservoirs in Oman is complicated by high uncertainties in key parameters. This study aims to adopt a physics-based data analytics technique to model geomechanical behavior of rocks. The paper also presents the methodology of linking geomechanics with well performance. Finally, the integration of the results into a decision support system is discussed. The majority of deep gas reservoirs in Oman are tight. The permeabilities are in the sub-millidarcy range. Hydraulic fracturing helps to unlock the reserves. Meanwhile, proper hydrofracture design is required to optimize the development of these complex reservoirs. Due to high stresses, unclear processes governing hydrofracture propagation, and complex depositional and diagenetic histories, the applicability of standard hydrofracture modeling techniques becomes questionable. Proper surveillance design and in-depth analysis of the monitoring data assist in testing the range of applicability of the modeling tools. The analysis also aids in characterizing the influence of the processes not captured within the models. Recently the development of deep gas reservoirs in Oman started to benefit from horizontal well technology. Similarly to other horizontal developments, the question of proper well architechture and stimulation design was raised. In this study, the data pertaining to historical vertical wells was collated to understand the processes governing hydrofracture placement. The data indicated the presence of strong fracture barriers and of highly stressed zones which affect the ability to create sufficient fracture conductivity. Further, geomechanical models were calibrated to allow for realistic estimate of the contact area between the fracture and reservoir. Analysis of the production data indicated that productivity was often limited by the factors not captured in the models (e.g., suboptimal cleanup). For proper planning, risk factors may be chosen to reflect the loss of productivity. In the next step, the learnings from the vertical wells served for characterizing hydrofracturing in horizontals. Analysis of the data indicated that due to near-wellbore complexity and choking effect the productivity of an individual fracture in a horizontal well was only a fraction of that in a vertical well. As a final step all the data along with their interpretation are being incorporated into the library of hydrofracture scenarios. Future development will rely on searching for the analogs and selecting a design fitting all applicable scenarios. The paper presents an overview of the surveillance data analysis. The results of the analysis allow for creating a library of the development scenarios, which serve as a basis for a decision support system aimed at streamlining hydrofracture and well planning design.