Automated Reservoir Model Calibration for Field Development Plan Evaluation Under Subsurface Uncertainty Applied to a Complex Multi-Zones Heavy Oil Field

Abstract

Abstract Rubiales is a major heavy oil field in Colombia with an OOIP larger than 5000 MSTB (Stanko, and others, 2015). The field produces from six zones mainly with horizontal wells. Production is driven by a strong aquifer which causes tilted oil-water-contact and early water breakthrough. Fully integrated reservoir modelling for field development optimization under subsurface uncertainty has been a major challenge so far. This paper presents an automated calibration process, probabilistic infill well ranking and location optimization. An automated reservoir characterization workflow was developed to generate multiple history matched models on field and well level. Static reservoir characteristics and contacts where parameterized for sensitivity assessments and calibration update steps. Variations of dynamic reservoir characteristics with an impact on model forecasting behavior were applied to alternative history matching solutions to create an ensemble of reservoir models for uncertainty assessment. Economic success criteria and a simulation opportunity index were defined for a probabilistic well ranking and optimized well location assessment. The workflow was applied to a sector of the full field including approximately 300 producer wells. Multiple history match solutions were created with 80% of the producer wells matching on well level. Quality assurance measures were applied to verify geological consistency of implemented model updates. The ensemble of forecasting models was used to deliver a probabilistic well ranking based on a well Net Present Value model. Infill well candidates with a robust performance delivery across the ensemble were identified. Results showed that a well placement scenario with half of more than 100 well candidates delivered above the economic threshold criterion and a similar recovery compared to reference field development plan. Probabilistic sweet spot maps based on a simulation opportunity index were used to efficiently identify well locations for more than 30 alternatives well candidates. The method produced robust results above the economic success criterion. Methodology and workflow design developed in this work successfully delivered a field development evaluation under subsurface uncertainty for a large heavy oil field with complex geological characteristics, long production history and large number of wells. The workflow design is applicable for other fields with similar characteristics and delivery objectives. The developing of this advanced workflow combined the application of a last-generation High-Resolution Reservoir Simulator (HRRS) and an Innovative Collaboration Environment (ICE) (Schlumberger 2020) which combines domain expertise and advanced digital technologies (ADT) enhanced quality and time results for history matching (HM) scenarios and bring the opportunity to execute several uncertainty cases for forecasting analysis allowing us to consider a wide range of results for final FDP proposed