I'll offer this much. For the purposes of discussion, I envision a line that is supported down low, and goes through a fairly long vertical rise before turning horizontal and crossing the support location where the +Y is. As a matter of course, the support at the bottom is going to force the thermal expansion to the top and give you your 3" liftoff.

From a technical standpoint, if you've run the case w/o the support and your sustained stresses are within allowables, then the 3" lift off is not an issue by itself.

But, I will offer a couple of other considerations.

1. Buckling - If you are supporting in this way, you need to make sure the vertical run is adequately guided, becuase supporting from down low makes your vertical run subject to column buckling and Caesar won't tell you anything about that. If you had the hard support at the top with either no support or a spring down low, assuming you have enough flexibiilty to absorb the 3" of growth, then the riser would not be subject to buckling. However, you would still want/need to guide it due to wind and such

2. Asthetics - One thing I've learned over the years is that if something looks "wrong" to some guy in the field, he's going to want to do something about it. I've gotten numerous calls from the field after startup from some guy telling me that a shoe isn't sitting. I look at the analysis to find that it was supposed to lift off and I assure the field that there is no problem. There are a lot of people in the industry who still don't believe that pipe grows when it gets hot. If a support is in a highly visible location, you may want to add a spring to provide peace of mind to the field guys. THis is not something I end up doing often, but it has happend.

3. Equipment - You should see it when you removed the support and ran the model. When that support lifts, the loads shifts to adjacent supports. If one of those is an equipment nozzle, especially rotating, your nozzle may be overloaded and this is often a limiting factor more so that stresses in the pipe itself. I've even had cases where the line does not actually "lift" but vertical thermal expansion is such that the support load is reduced effectively to zero and overloads a nearby pump. I've had to put springs on in such a case anyway.

So, the answer to your question is: Yes, it's OK, maybe :-)