Rapid Model Update - Enabling Fast, Structured Dynamic Model Updates Leveraging Automation

Abstract

Abstract The quality of a reservoir simulation model is essential for reliable field development planning. To reduce uncertainties on production forecast, the data integrated into the model should incorporate all wells drilled along with their completion data and historical production and injection data and reflect accurately the latest understanding of the subsurface based on all the latest measurements performed in the field. Additionally, the model should be able to reproduce the historical production and injection to a good accuracy. To improve the reservoir engineers’ ability to keep the reservoir simulation models up-to-date, a fully automated framework was developed to facilitate the process of gathering data and reviewing the quality of the dynamic reservoir model by connecting directly to production databases from the geomodelling platform, integrating all required data for dynamic reservoir modelling, and providing automated data quality check tools to assess the quality of the input data to the simulation model, as well as the quality of the simulation results against historical data. The framework enables a rapid update of the static modelling data, relative permeability and capillary pressure data, fluid model and initial conditions of the reservoir. Additionally, a direct channel to production databases (e.g., Oracle database, OFM database) is open to retrieve the latest well trajectories, completion data, and associated historical measurements to ensure that the latest available data is integrated with the model without introducing human errors through manual data manipulation. Additionally, provides quality control tools enable checking the consistency in multiple building blocks of the dynamic reservoir model, such as the compatibility of drainage and imbibition relative permeability curves and capillary pressure curves used in modelling hysteresis processes, potential issues in black oil fluid model tables. A streamlined process to evaluate the quality of the model calibration to historical data is also integrated and provides flexible metrics and visualization to assess the quality of the history match of standard properties such as production/injection rates, cumulative production/injection volumes, water cut, gas-oil ratio, static and flowing pressures, and enables the assessment of the model quality against additional advanced metrics such as water breakthrough time, and match to surveillance logs (e.g. saturation logs, pressure logs, production logs). The framework was applied on a giant onshore carbonate oilfield with a large well count. Quality control was performed to assess the consistency of the reservoir simulation model input data against production databases, followed by an update of the model and the assessment of the quality of the history match following the update. The process that the engineer could spend weeks to perform can now be performed in days, with a higher quality and accuracy through automation, with a fast understanding of the model quality to perform reliable field development plan optimization.