Penney, Richard K. (PDO) | Moosa, Riyadh (PDO) | Shahin, Gordon Thomas (Shell) | Hadrami, F. (PDO) | Kok, A. (PDO) | Engen, G. (PDO) | van Ravesteijn, O. (PDO) | Rawnsley, K. (Shell) | Kharusi, B. (Shell)
Significant volumes of heavy oil remain in fractured carbonate reservoirsworldwide. Some of these reservoirs are good candidates for the application ofthermally assisted gas-oil-gravity-drainage (TA-GOGD), a novel EOR technique.Unlike a normal steam flood, the steam is used as a heating agent only toenhance the existing drive mechanisms. The elegance of TA-GOGD is that thefracture network is both used for the distribution of steam (heat) and therecovery of the oil. The number of wells can therefore be kept to a minimumcompared to conventional steam floods. Following encouraging pilot results in afield in Oman, a steam injection project is heading for implementation, a firstof its kind on this scale. Studies to date indicate that recovery factors of25-50% with Oil-Steam-Ratios of 0.2 -0.4 m3/ton of steam are feasible. Thesuccess of the project is critically dependent on the field-wide presence ofconductive fractures and the ability to characterize them. Both stochastic anddeterministic studies were tried, but the latter method is now favoured as itallows the use of geological and dynamic understanding as input to themodelling and honours existing faults, deformation mechanism and the conceptualmodel. Fracture characterisation is to some extent still an art and outputs are'only static scenarios'. Therefore results should be validated with dynamicdata as much as possible. The dynamic models are thermal and dual permeability,with compositional dependencies: a complexity that is rarely encountered.Explicit fracture block models are used to verify that the heating rate andGOGD are captured properly, in particular for irregularly shaped fracturepatterns. A new fully integrated workflow of fracture characterisation withstatic and dynamic modelling has enabled uncertainties and risks to be managedin a scenario based approach.