Rule number one. Do not automatically place an anchor at the place where you figure the pipe will stop moving axially.
The idea is to model enough straight pipe so that the analysis will be able to accumulate enough friction to stop itself. The VAL is an estimate of the length of pipe that can do this. With this straight length in the line, the upstream and downstream portions are structurally detached for analyses at or below the temperature/pressure for which the VAL was calculated.
Now, the VAL calculation in CAESAR II assumes a stick/slip response to axial friction - rigid axial restraint or sliding with a resisting load. But your CAESAR II soil model probably has a some sort of deflection associated with the estimated ultimate axial load - that will give a nonrigid stiffness to our "stick" in the stick/slip model. Therefore, some movement must be absorbed to build up the friction load that is accumulated down the line. For this reason - I suggest you run at least 2 VALs down the line before truncating the system.
I don't really care if it's 2 or 3 VALs, if I have that straight length in my model, it's just one pipe that you build and CAESAR II will take it from there. The important thing here, if you have less than 2 VAL between your last modeled thrust point (e.g., a change in direction or branch) and the next thrust point that you are trying not to model, I would code through that next thrust point.
I think 2 VAL should be enough but I have not looked at enough buried models to practically reinforce this concept.
Back to my inert section of pipe mentioned in the previous reply... I am not looking at a single point on the line that does not move axially; you will find a single such point in almost any run of hot pipe - buried or unburied. I am looking for a series of nodes that do no move axially. Now, in your buried pipe, our "Zone 3" (friction controlled) automatic spacing between added restraints is 100ODs. They cover a lot of pipe distance. And zero displacement, in my opinion, need not be 0.0000, If you get my implication.