During last decades the major companies increasingly exploit such hydrocarbon resources as heavy oil, bitumen or oil shale. The production of oil from such reservoirs is frequently done through reservoir heating, most often via steam injection. However it becomes more and more evident that for some cases there exist better suited technologies and one of them is the electromagnetic (EM) heating assisted oil recovery. According to the EM field frequency different physical mechanisms may underlie the heat release. The radio-frequency (RFH) or dielectric heat generation results from rotation with friction of polar molecules in the EM field.

The advantages of the RFH-based technology are related to the in-situ heating and steam generation from the connate water which eliminates the problems associated with low well injectivity, well injection pressure, the water supply and hence, the water treatment etc. They include also the possibility to apply the heating in shallow or thin reservoirs and also in case of initially mobile heavy oil to optimize the production efficiency using both principal production mechanisms: the gravity drainage and the steam drive.

The large-scale RFH models have been recently developed for numerical simulation of realistic EM power distributions. These models provide a means to critically analyze the processes of oil recovery. In the current study we take advantage of two dedicated simulators and the methodology of loose coupling between them to deliver the calculated EM heating power field to dedicated reservoir simulator.

The combination and competition between two principal production mechanisms for different well patterns, spacing and orientation are considered; their influence together with other operational conditions impact on the efficiency and application limits of the oil recovery method are analyzed.

The analysis of different mechanisms of RFH assisted heavy oil production is an important feature in process design considerations.

You can access this article if you purchase or spend a download.