I am unsure of your intent with that D1=250mm. Your approach might report the proper position of the this node but I'm not confident that the loads and displacements elsewhere are correct. If you model results (throughout) match the field observations, you are OK in terms of structural response. I would still question any code-defined stress evluations for this system.

Here's how I would approach it... The pipe and restraint change their relative positions. You are moving the pipe, I would move the restaint instead. Place a CNode on that +Z restraint and set a +Z displacement (D1) at that new node equal to the distance that the pipe had to move to fall off the column. Include D1 in all analyses.

The problem I have with all this is the pipe apparently "walked" off the support through several thermal cycles. If so, the "ratcheting" has changed the structural state of this system - something somewhere has yielded. We work with an elastic model and cannot accommodate this plastic response. Of course, you can model plastic hinges in CAESAR II (with bilinear rotational restraints) but it would be difficult (for me) to have high confidence in the results. You may gain this confidence by manipulating your CAESAR II model so that it matches the observed response in the field (loads & displacements).