Carbon Dioxide (CO 2 ) Enhanced Oil Recovery (EOR) process is most likely to be the future hydrocarbon recovery process in Abu Dhabi [1] . In November 2009, Abu Dhabi Company for Onshore Oil Operations (ADCO) has commissioned the first CO 2 -EOR pilot in the Middle East. The pilot project was implemented successfully on a fast track basis with a broad objective to test the concept of CO 2 flooding in Abu Dhabi complex carbonate oil reservoirs.The project comprises three wells, a single CO 2 injection well, an observation well and a single oil producing well [1] . An intensive reservoir monitoring plan was implemented pre-and post-CO 2 injection to evaluate the uncertainties related to the reservoir fluid and rock properties, CO 2 injectivity, miscibility, and sweep efficiency.Available reservoir static and dynamic data was employed to construct a fully compositional simulation model to history match the pilot performance based on remarkable field observations. Effort was made to ensure that all reservoir data were quality checked, re-interpreted and validated before being utilized in the simulation model.Conventional history matching was conducted to match the model response to the pilot performance. This is complemented with Software Assisted History Matching to further quantify and assess the range of uncertainties for different reservoir parameters that would impact the quality of the history match. Sensitivity analysis was carried out to identify those parameters that have the most effect on the pilot performance. This paper outlines the main findings for the first miscible CO 2 -EOR pilot in Abu Dhabi based on well calibrated compositional model considering the most critical reservoir uncertainties affecting the process.

This paper describes the implementation of the solvent model in streamline simulation and its application to WAG injection optimization for a producing field. The solvent model was implemented as an extension to the 3-component black oil model, with solvent as an additional component, to model miscible displacement process. The relative permeabilities and fluid properties of the oil and gas phases are modified based on the fraction of the solvent component, the reservoir pressure and the Todd-Longstaff mixing-parameter (i.e. an empirical treatment of the effects of physical dispersion between miscible components). The solvent model was applied to a Middle Eastern oilfield, which is currently under development using a miscible WAG process. PVT analysis has been done to prepare the properties for the reservoir gas and the injection solvent. Streamline simulation models with and without solvent model were run to compare the results with a reference finite difference compositional reservoir simulator and the effect of miscibility has been validated. Results of streamline simulation models with and without solvent model were compared against a reference finite difference reservoir simulator. The comparison shows the streamline simulation model with solvent model has much better agreement with the reference finite difference compositional reservoir simulator which shows that the miscible displacement process is properly simulated in the streamline simulation model with solvent model. The solvent model presented in this paper advances streamline simulation technology, combining the intuitive and unique properties of streamlines and the capability of simulating miscible recovery mechanism. It allows simulating both immiscible and miscible displacement within the same simulation. The solvent model considers the effects of miscibility by considering relative permeabilities and fluid properties adjustments based on pressure and solvent concentration. The technology will help effective simulation of miscible recovery process, assist optimum solvent allocation and improve unified sweep.