I recently had a case where the cold shutdown condition caused the highest restraint loading. This is not correct because the line is empty of fluid in the cold case. I would typically set up my load conditions as:

1) W+T1+P1 (Hot operating)
2) W+T2 (Cold shutdown)
3) W+P1 (SUS)
4) W (pipe + fluid weight)
5) C1-C3 (Stress range from install to hot operating)
6) C 2-C4 (Stress range from install to cold)
7) C5-C6 (Stress range cold to hot)

I don't see anyway that I could change the load cases or fluid density to produce both hot and true cold restaint loading in a single run. Anyone have any ideas?