Simulation of THAI (Toe-To-Heel Air Injection) Process

TianXiang Xia (cestxx@bath.ac.uk) has been developing advanced EOR/IOR processes for heavy oil recovery and in situ upgrading, as part of the Bath IOR Group's activities. Together with Group head Malcolm Greaves (M.Greaves@bath.ac.uk), he presents a simulation view of THAI, from the laboratory and field-scale perspectives.

THAI (Toe-to-Heel Air Injection) is a new IOR technique which has the potential to revolutionise how we extract heavy oil and bitumen. The Bath IOR Group has conducted extensive laboratory studies of the THAI process since 1990, using 3-dimensional physical models. THAI creates a "short-distance displacement" pathway, achieving stable propagation of a high temperature combustion front through the oil-bearing matrix. High oil recovery (up to 85% OOIP) and substantial in situ upgrading (6-8^o API) are achieved (http://ior.rml.co.uk/issue2/rd/universities/bath.htm).

The formation of a coke zone inside the horizontal producer is one of the main mechanisms for the stability of THAI. Experiments have shown that the coke inside the HW acts as a gas seal, forcing the injected air to flow directly to the advancing combustion front. Oxygen is therefore consumed by high temperature oxidation reactions when it contacts the fuel ahead of the combustion front. The produced gas contains virtually no oxygen, or only a very low, safe level. Potential benefits of THAI applied to heavy oil reservoirs are:

• Stable, robust process, due to controlled gas overriding and gas seal in the horizontal producer well
• Achieves high sweep efficiency and very high oil recovery, up to 80-85% OOIP
• Produces significantly lighter oil (100 mPas at 20^o C) from heavy crude (48,000 mPas at 20^o C), achieving in situ upgrading of 6 to 8^o API
• Reduces sensitivity to reservoir heterogeneity, because of more limited reservoir effect, compared to a conventional displacement processes
• High air injectivity, maintaining the process in HTO mode, avoiding low temperature oxidation (LTO)
• Significant environmental benefits due to in situ removal of sulphur and heavy metals
• Easy front tracking via its "toe-to-heel" propagation

Numerical Simulation

A numerical model of THAI needs to integrate all of the chemical reaction kinetics (thermal cracking and oxidation of coke and heavy residual) together with multi-phase fluid flow and heat transfer in the reservoir. The well geometry used for THAI involves either VIHP (vertical injector-horizontal producer) or 2VIHP (two vertical injectors and one horizontal producer) per section. VIHP is direct line drive and 2VIHP is staggered line drive. In order to capture the essential features of THAI, a 3-dimensional model is required, as in the 3D cell experiments

The Bath IOR Group was the first to perform a numerical simulation of a THAI 3D combustion cell experiment, using the advanced thermal reservoir simulator STARS (Computer Modelling Group, Canada ) in 2000 [1]. It was successful in achieving stable propagation of the combustion front, without oxygen breakthrough, but the numerical model employed a "sleeve-back" numerical technique to artificially control the flow into the HW. Further progress was made by including coke (fuel) as a pseudo-component in the oil, in order to overcome a limitation of the thermal cracking model, dispensing with the "sleeve-back" control.

A big simulation advance was achieved by using a discretised wellbore to represent the horizontal producer in STARS , instead of a "source-sink" [2]. The numerical model utilised 8000 grid blocks (40 ´ 20 ´ 10) to represent half of the 3D combustion cell, measuring 0.60 ´ 0.40 ´ 0.10 m (Figure 1a). The simulation results are for a heavy oil of 10^o API, with the same initial conditions used in the 3D THAI experiment.

Results of Numerical Simulation Model

The numerical model predicted that the THAI process operates in a stable manner and no oxygen is produced at the production well (Figure 1b). Using an improved thermal cracking kinetics model, the in situ upgrading effect observed in 3D combustion cell experiments is accurately predicted by the STARS simulation (Figure 1c), also confirming that the thermal cracking model is correct. The 6.7^o API upgrading is very significant. The predicted oil recovery was 80.5% OOIP (Figure 1c), almost the same as in the experiment.

Figure 1: 3D Simulation of THAI Experiment: (a) Numerical Model, (b) Produced Gas, (c) Oil Recovery and Upgrading (Click for larger view)

Stable high temperature (600-800^o C) combustion front propagation is predicted (Figure 2). In practise, the combustion front temperature can be controlled by water injection. The simulated oxygen profiles in the sandpack are shown in Figure 3. The injected oxygen stays in the top part of the oil layer throughout the simulation, as it does in the experiment. This "top-burning" effect, or controlled gas overriding, is an important factor for the stability of THAI. The predicted coke profiles in the oil layer (Figure 4) show that there is a stable process of coke burning upstream of the combustion zone, and steady laydown of coke ahead of it. The simulation predicts a coke deposit in the horizontal producer, which acts like a gas seal. At the end of the simulation run, the highest coke concentration was found at the "heel" of the horizontal producer. This is consistent with a post-mortem analysis of the sectioned HP in Run 2002-01 [2].

Figure 2: STARS Simulation of Temperature Profiles During 3D THAI experiment (Click for larger view)

Figure 3: STARS Simulation of Oxygen Profiles During 3D THAI experiment (Click for larger view)

Figure 4: STARS Simulation of Coke Profiles During 3D THAI Experiment (Click for larger view)

Field Scale Numerical Simulation

THAI is set for a pilot test in the Athabasca Tar Sands in Alberta , Canada , in the fourth quarter of 2004 ( http://ior.rml.co.uk/issue6/rnd/Universities/Bath/bath.htm ). Field scale simulation can provide the key parameters for project design. Scale-up of THAI, from laboratory to a field scale, is very complex. In fact, previously, there has not been any satisfactory 3D simulation of conventional in situ combustion. This was limited in part by computing power available at the time. We have recently performed a partial field-scale simulation of THAI using STARS. The model used 32,000 grid blocks to represent half of a section of an oil field (200 ´ 100 ´ 20 m) containing 10^o API oil. The well configuration was VIHP in direct-line drive. Compared to the 3D cell experiment, the volume of the field-scale model is about 33.3 million times bigger. Steam injection was employed to establish the initial communication between the vertical injector and the HW. After 180 days of steaming, auto-ignition was achieved as soon as air was injected. The simulation required about 1 month of computing time on a 3.09 GHz PC. The simulation was terminated after 370 days due to slow convergence.

As shown in Figure 5, the combustion front propagates from the air injection well to the "toe" position of the HW. The produced oil is in situ upgraded from 10^o API to about 16^o API, in good agreement with the laboratory numerical model, and experiment. The oil production rate would increase if air injection operation continues, as the combustion front expands laterally to cover the whole width of the section.

Figure 5: Field scale simulation of THAI: Temperature Profiles (a) 180 days, (b) 250 days and (c) 371 days and (d) Oil Production (Click for larger view)

Future Research Areas

Future research is targeted at modelling reaction kinetics for various heavy crudes, bitumen and also medium-heavy oils (especially in relation to the North Sea ). Other topics are reservoir heterogeneity, bottom water and gas cap reservoirs, and start up protocols. There are JIP proposals covering these topics and sponsors are being sought. Please contact Prof Malcolm Greaves ( M.Greaves@bath.ac.uk ) or Dr TianXiang Xia ( cestxx@bath.ac.uk ).

References

• Greaves, M., and Xia, T. X.: "Simulation Studies of THAI Process", Paper 2000-084, presented at the Petroleum Society's Canadian International Conference, Calgary, Canada, June 4-8, (2000)
• T. X. Xia, M. Greaves, and A. Turta.: " Main Mechanism for Stability of THAI - 'Toe-to-Heel Air Injection'", Paper CIPC 2003-30, presented at the Petroleum Society's Canadian International Petroleum Conference 2003, Calgary, Alberta, Canada, June 10-12 (2003) .

Disclaimer: The material available on this website is designed to provide general information only. Whilst every effort has been made to ensure that the information provided is accurate, it does not constitute legal or other professional advice.
Please note: The Department of Trade and Industry cannot be held responsible for the contents of any pages referenced by an external link.