Incorrect summation in contour plot

[EdmundStephens]

See attached model TEST_GRID including an .rsp project file. The model has FOIP = 9.527 MMstb oil in place. When a contour plot is made showing Cell Result = SFIPOIL and Map Projection / Result Aggregation = Sum, then the Sum given for "Current Time Step and Visible Cells" does not equal the FOIP. Testing in version 2018.11.2, the Sum depends on the Sample Spacing factor: gives 3.1288e+07 with spacing 0.75 and 1.9053e+07 with spacing 1.0. If a range filter is made to isolate a single cell, e.g. cell 3,2,1 has FIPOIL = 376811 bbl, the contour plot with the same range filter gives Min/Mean/Max = 376811, but Sum = 753622 bbl.

It seems logical as well as having practical application for the sums (and averages) to reproduce the same result as could be derived from the grid array. I was trying to use the contour plot in this way to look at the STOIIP density by region, for individual layers and in a high perm facies (via property filter on OPERNUM). I noticed that the summation did not match the STOIIP. So this looks to be a bug, but if not it is not clear what is the meaning the present summation?

TEST_GRID.zip](https://github.com/OPM/ResInsight/files/2872431/TEST_GRID.zip)

[lindkvis]

I think we need to remove the Sum from the info box for the contour map, because as it is, it is just misleading and is mainly used as a control by us developers to check on things we know should match up.

This is the reason why you don't get a matching sum in this particular case is not a bug, but that the sum isn't showing the same as the sum in the 3d Model:

• The sum for the contour map is a sum of the values of the square 2D cells that make up the contour map.
• This is not weighted in any way, and depending on the sample factor, there may be multiple 2D elements per cell.
• In the particular 1-cell example set up in the provided example, you get two 2D elements each with the same value and thus the sum of them is twice the value of the 3d cell!

I will soon check in a change that removes the "sum" from the Contour Map info box. The min, mean and max still seem useful as they should closely match the 3d result.

[EdmundStephens]

Hi, Thanks for looking into this, removing this will make the interface more consistent and not introduce a confusing result in the display.

However, I would suggest something like this is a useful enhancement. It is very useful to be able to extract sum / averages / volumetrics from a map, and to work from all the interface filters available. For example to look at average porosity and total STOIIP in specific regions broken down by combinations of facies and unit. This is already in the 3D grid selecting "Current Time Step" and "Visible Cells", could you not use that method to display the same sum on the linked contour plot (but with the filters set on the contour plot) ?

So 3 possible improvements:

1. Add Sum to the contour plot statistics using the same method as for the 3D grid
2. Add P90 and P10 to the contour plot Info Box display similar to the 3D grid
3. Suggest "Current Time Step" and "Visible Cells" should be the default setting for the 3D view info box, "All Steps" and "All Cells" are interesting but I struggle to see why these should be default.