Your on the right track. While, as stress engineers, we are concerned with the design of the piping system, we need to take into account the effect of attached equipment - especially thermal expansion. Towers, tanks, drums, exchangers all made of metal grow just like the pipe does.
One other bit I'll leave you with - you noted several times that the system is working in the field after one year of service. That's probably not long enough to know. Remember that the expansion stress limits are based on how many cycles the system can survive going from install to operating temperature and back again. Typical B31.3 basis is 7000 cycles (which is modified by changing the f value accordingly for different cycle life)
So, your piping system can be well over the allowable expansion stress values, and still survive a few cycles before failing. If the unit is one that is intended to come up and run at a fairly steady condition, you may just not have hit the (low) cycle limit that your "overstressed" system has.
In conclusion, when someone tells you it "works fine in the field" take it with some skepticism as that doesn't mean you have a code compliant design.
_________________________
Edward L. Klein
Pipe Stress Engineer
All the world is a Spring