The error looks like it's normally caused by the use of fills where a surface appears at more than one level, say defining a fill and also in the cell description of the fill universe.
The VisEd errors will match the output file, but the output file should have more of an explanation; which offending cells that surface is in and where they are in the chain. The offending surface 2 is composite which is why we have decimals. Fills are normally done into a bigger volume and the filled cell acts as a cookie cutter, filling with the same size would produce this because edges would be coincident. It's normally a warning. I have no idea why this is a fatal error.
Just as an aside, VisEd X_ 22S corresponds to X 2.6 - circa 2008.
If we get more info, maybe we can help solve this. If the OP has solved this, I would hope we're told.