Integrated reservoir simulation model (IRSM) is a common working tool in multidisciplinary teams nowadays. Usually such models constructed and used before field production startup, on the pre-project stage. But authors are commonly missing part about how good is the prediction ability of the model after first production and pressure data acquired? History matching, calibration of all parts of integrated model (reservoir, well, surface pipelines) requires constant efforts of multidisciplinary team or even building new model from scratch. In the scope of the paper the preparation of dynamic integrated reservoir simulation model and calibration based on 2 years of Phase 1 field production is covered.
Commercially available reservoir simulator with fully coupled network infrastructure was used for modeling purposes. Multi-reservoir static model grid was linked with networks (oil production, gas and water injection). Real time data from well downhole gauges (including smart wells) and MPFMs were used for reservoir models HM. Additional data from surface gauges and export meter were used to calibrate pressure drop in wellbore and surface pipelines.
Initially created IRSM was not able without HM stage and calibration to accurately predict the reservoir performance and pressure in the surface production system. Exceeding complexity of the model was decreasing the accuracy of the forecast even more. Only after history matching step for all wells using daily pressure/rate data and implementation of real pipeline layouts, the model was used for different purposes such as well placement and business planning. Highly efficient prediction tool was constructed in the scope of the study.