Well,
I would say that the problem is, in reality, that a pair of relieve valves would never open at the same time. You could model as it will happen but chances, (again, in my opinion), are very slim. I think that in this kind of arrangements, one valve is more sensitive than the other, and whoever is more sensitive is going to pop up before and release the overpressure on the system. If the pressure is not relieved by one valve (which I certainly doubt) then, the second would open, but again the time-force curve will be different.
Since you never know who is the more sensitive one, I would suggest analyzing one at a time. This would be the worst case since the load is totally unbalanced.
Regards,
Pedro Massabié
_________________________
P Massabie