SPE logo, 37K

SPE Comparative Solution Project
questions and updates

Q In the model 1 (the cross-sectional problem) how many pore volumes of injection do you want reported and on what frequency?
A We'd like the problem run to 7560 days, which is almost 3pv, and a reporting time of every 180 days.
Q Is the depth of 12,000 ft in model 2 the top or the bottom of the formation or somewhere in between?
A 12,000 ft is the depth to the top of model 2.
Q The second model does not have a block centred grid. This will be a problem with some reservoir simulators. Having the producers on the corners is not a problem, but having the injector directly in the centre of the model is with the grid you propose.

A This depends on whether you upscale to an even or an odd numbered coarse grid. There's just no way round this with upscaling on regular grids. Options to consider are: tensor product grids, where you have variable dz, dy on the coarse grid, and local grid refinement.
Q What is the filling convention for the datasets in model 2?
A The filling convention is top to bottom.
Q What well radius should I use in model 2?
A The well radius is 5 inches.
Q What is the rock compressibility for model 1?
A We used 6.0e-5/bar, or 4.138e-6/psi.
Update There is an error in the fine grid porosity and permeability file for model 2. The y, z permeabilities for k = 35, 70 duplicate those for k = 1, 35.

Please download the corrected file.