Between 2003 and 2008, there was no testing. There was a lot of calculating. This is a fixation of the regulator, that you can calculate everything. Unfortunately, there becomes a limit beyond which you cannot calculate.
I could give technical details of why the calculations are very difficult, but I'm afraid that might entail too much technical detail.
We used BNL to be independent reviewers of what we had done, and then we asked BNL to do it independently. Again, this was to try to satisfy the regulator's request that it be demonstrated by calculation. In the end, we needed to do the tests, because you couldn't calculate some of these effects. You already have to know the answer to get the code to tell you the answer. That's the way it works with thermomechanical codes; that's the way it works with thermohydraulics codes. Neutronics is probably the closest to getting an independent answer, but with the thermohydraulics and the thermomechanical, you have to tell them the answer before the code will give you the right answer. So you already have to know what the answer is.
This fixation on a calculated solution--