When implementing any Enhanced Oil Recovery (EOR) or tertiary recovery project it is critical to understand the key variables that influence the success of the scheme -reservoir parameters and fluid properties -and the behaviour of the existing development scheme. The additional components of miscible fluid fronts in heterogeneous carbonates, sources of miscible injectant and the systems needed to manage a miscible EOR development will generate additional uncertainty to that already present during primary and secondary recovery.This paper describes the reservoir on which evaluation of CO 2 miscible EOR has been carried out, the reasons it is a candidate for CO2 EOR and identifies the key uncertainties that will have a major impact on the behaviour of a CO2 development scheme; namely fluid distribution, reservoir structure, compositional variation and reservoir property distribution. A key element of the paper is describing how the uncertainties are recognised and managed by the modelling philosophy, given the relatively immature status of the reservoir development and the limited historical data available.The paper describes the interpretation of incremental benefits of CO 2 EOR over existing development schemes. A small component of the paper discusses options of appraising CO 2 EOR further, the approach and need to pilot injection patterns and concepts of phasing development of CO 2 into an existing pattern development. Uncertainty range will reduce over time and the paper describes how work emphasised the importance to recognise when this range is acceptably small to make decisions that involve larger capital investment, such as new development options involving additional wells and facilities.Although the majority of work done in this study was on subsurface elements, this paper briefly touches on infrastructure needed for CO 2 EOR and implications of the scheme on existing facilities. Also considered are the external influences that impact EOR value, such as the source of CO 2 , price of fluid streams entering and leaving the system, and potential for the scheme to be a CO2 storage option once EOR benefit is exhausted.At present in Abu Dhabi, CO 2 or Miscible Injectant (MI) EOR has been recognised as the EOR development of choice by a number of different studies. The fluid types and pressure/temperature regimes in Abu Dhabi reservoirs mean that miscible flooding is likely to be the most appropriate mechanism to further increase recovery factors. Using CO 2 also has the benefit of demonstrating CO 2 capture and likely storage at a time when the UAE is seeking to curb carbon emissions. Improving the understanding of how CO2 can play a role in Abu Dhabi reservoirs is therefore highly topical.

Simulating a high-resolution multimillion cell model brings many benefits, by enabling reservoir engineers to use the best grid size for accurate representation of water and gas movement in the reservoir, essential for advanced field management, Enhanced Oil Recovery or complex well design studies. To improve the characterization of a giant heterogeneous carbonate reservoir and enhance the quality of field development plans, new high-resolution static and dynamic models have been used to study one of the largest oil fields in Abu Dhabi. A detailed static model of over 50 million grid cells was constructed, using a unique water saturation modeling approach, without upscaling to a dynamic simulation, using hysteresis for both relative permeability and capillary pressure. The reservoir has over 50 years of history, with hundreds of vertical and horizontal wells. Large volumes of data from well logs, cores and other measurements were used to populate the static model, define dynamic rock types and match well log water saturation and water capillary pressure profiles. The concept of wettability change with depth was introduced, with an oil-wet system at the crest, graduating to a water-wet system near the thin transition zone. A geological resolution grid was used for reservoir simulation studies, after testing input data consistency and stable behavior. A stability test was performed by running the simulation with no wells for 50 years after equilibration and showed no movable fluids. This verified the consistency of the reservoir static properties, rock types, water saturation, relative permeability and fluid model. A history matched case was developed with over 850 wells using the same fine grid, to meet the objective of completing each simulation run within one day. After history matching, a compositional simulation model was built, to investigate the impact of grid resolution on future production forecasts. This is the largest dynamic model built by the company and demonstrates the benefits of rigorous attention to the quality of the static data, while using modern simulation workflows to avoid compromising the detailed model by upscaling. The methodologies presented in this paper will be adopted as best practices for future similar projects.