Using the Computer Modeling Group's ISCON simulator, three Athabasca Oil Sands combustion tube tests were matched: two wet combustion tests with water-air ratios of 61.1 and 2.25 kg water per m3 (ST) air and one dry test. Presented in the paper are laboratory measurement of oil viscosity, density and thermal cracking kinetic parameters as well as a description of the crude oil characterization technique employed.

The simulated results, using the above data and characterization, compared very well with the experimental runs. The only problem area pertained to matching the superheated steam bank in the second wet combustion run. A quantitative comparison between empirical and simulated temperature profiles, production histories and fuel lay-down is included.

With the parameters developed in the combustion tube matches a one-dimensional simulation, using reservoir parameters and large grid blocks was conducted. The effects of grid block size on the overall combustion mechanism and the parameters affected are discussed. Suggestions for using numerical simulators to predict field performance are also presented.


The Computer Modelling Group's "Iscon" simulator was used to simulate three combustion tube tests of which the experimental results were presented in three reports dated between December 1978 and August 1979. Based on tile parameters developed to match these combustion tube runs a one- dimensional field scale simulation was undertaken to ascertain the effects of large grid block size on the overall modelling procedures.

This paper summarizes the input data, the difficulties experienced, and compares the simulated results with the experimental data.


A computer program, owned and maintained by the Computer Modelling Group (CMG), was used to simulate the three Athabasca combustion tube experiments.


Of the numerical simulators that have been built, 1–6, there are many different features that have been incorporated by some and omitted by others. These numerical models can be divided into two classes based on how the oxygen consumption is calculated. One type assumes complete oxygen utilization and the others calculate oxygen consumption from kinetic parameters in the input date. In addition, all the models except Hwang et al 4 have the combustion front remaining within a block until all the fuel is consumed, (except in wet combustion) then the combustion process proceeds to the next block; by contrast the model of Hwang et al has a continuously moving front.

The Computer Modelling Group's ISCON simulator is a fully implicit model and is as generally applicable and rigorous as any of the simulators cited. The results obtained, using the program, are therefore expected to be representative.

Grid Configuration

The experimental combustion-tube tests were run in a combustion tube with a diameter of 18cm (3.92 in) which minimized any radial concentrations, saturation and pressure gradients. As described in the experimental guard heaters which produce

This content is only available via PDF.
You can access this article if you purchase or spend a download.