Numerical simulation of heavy oil and bitumen recovery and upgrading techniques

As a result of the increasing energy demand but a heavy dependence on easy-to-produce conventional oil, vast reserves of recoverable heavy oil have been left untapped. According to the International Energy Agency, IEA, fossil fuels – oil, coal, natural gas – will still predominate, despite a decline in their overall share, towards meeting the increasing world energy demand. While heavy oil has been predicted to account for an increasing share, contributions from conventional light oil have been predicted to drop from 80% to 53% in the next two decades (IEA, 2013b). Therefore, the large reserves of the under-utilised heavy oil, if extracted cost-effectively and in an environmentally friendly manner, will facilitate the meeting of both the short and long term energy demands. In this work, different thermal heavy oil recovery processes were reviewed with particular attention given to the air injection processes. In-situ combustion, ISC, has been identified as the most efficient and environmentally friendly technique used to recover heavy oil. Until the last decade, there was only a small interest in the conventional ISC. This is due to the complex nature of the processes taking place during ISC and the lack of success recorded over the years. The successful pilot scale testing of the Toe-to-Heel Air Injection, THAI, by Petrobank has revived interest both industrially and in the academic environment. Experimentally, THAI has been consistently proven to exhibit robust and stable combustion front propagation. Among the advantages of THAI is the ability to incorporate the in-situ catalytic upgrading process, THAI-CAPRI, such that further catalytic upgrading is achieved inside the reservoir. To realise the theoretical promise offered by THAI-CAPRI, there is a need to develop a reliable numerical simulation model that can be used to scale laboratory experiments to full field scale. Even for 3D combustion cell experiments, only one such model exists and it is incapable of predicting the most critical parameters affecting the THAI process. Therefore, the subject of this work was the development and identification of an accurate and reliable laboratory scale model that can then be used to develop field scale studies and investigate the effect of reservoir geology on the THAI process. However, because of the significant uncertainty introduced by the kind of kinetics scheme used and the fact that the main mechanism through which fuel deposition takes place is still a contentious issue, three different kinetics schemes, based on Athabasca bitumen, have been tested for the model of the 3D combustion cell experiment. All the models offered an insight into the mechanism through which oxygen production begins. They revealed that oxygen production was as a result of the combustion front propagating along the horizontal producer (HP). They also showed that the presence of coke inside the horizontal producer is an essential requirement for stable combustion front propagation. It was also observed that LTO is not the main mechanism through which fuel is deposited as oxygen does not bypass the combustion front. The models also showed that the temperature around the mobile oil zone (MOZ), where catalytic reaction in the CAPRITM is envisaged to be located, will not be sufficient to make the hydro-treating catalysts effective. Therefore, it is concluded that some form of external heating must be used in order to raise the temperature of the catalyst bed. Two out of the three different Arrhenius kinetics schemes that were successfully used to history-match the 3D combustion cell experiment were adjusted and implemented in field scale simulations. This is because the kinetics parameters obtained from the laboratory scale model cannot be used directly for the field scale simulation as they led to excessive coke deposition. A comparative study, between the two kinetics schemes, showed that the adjusted direct conversion kinetics predicts higher oil rate, and higher air rate can be injected right from the initiation of the combustion compared to in the case of the split conversion kinetics. The direct conversion kinetics was then used to study the field performance because it provided a more realistic representation of the physicochemical processes than the split conversion kinetics. The study revealed that even if the combustion front swept the whole reservoir length, it has to propagate along the horizontal producer for oxygen production to take place. It was observed that the combustion zone does not only have to cover the whole reservoir length but also has to expand laterally in order to produce the whole reservoir. For heterogeneous reservoirs, the THAI process was found to have larger air-oil ratio (AOR) in reservoir containing a discontinuous distribution of shale lenses compared to the homogeneous model. However, overall, the THAI process is only marginally affected in terms of cumulative oil recovery. The combustion front was found to propagate in a stable manner just like in the homogeneous model. However, further study is needed to investigate the effect of different permeability distributions would have on the THAI process. This should allow the optimum location of the wells to be determined. Studies of the effect of bottom water (BW) on the THAI process have shown that the oil recovery is heavily affected depending on the thickness of BW zone. It was found that the location of the HP well relative to the oil-water interface significantly affects the oil production rate and hence the cumulative oil produced. More oil is recovered when the HP well is located inside the BW zone. It was found that a ‘basal gas layer’, just below the oil-water interface, is formed when the HP well is located in the BW zone. The study has shown that there is a limit to BW thickness above which the THAI process cannot be applied to a BW reservoir. However, future work is needed to determine this BW thickness. The reservoir cap rock, depending on it is permeability and porosity, only marginally affects the oil recovery in the THAI process. It was found that the cap rock aids in heat distribution to the extent that most of the upper oil layer is mobilised. However, the effect is observed to be less pronounced with increased permeability and porosity. Future work should look into whether longer operation period has an adverse effect on the stability of the combustion front, and thus on the overall performance of the THAI process