Ricardo,

The option #1 would be the best approach in my opinion - it is a matter of including the nozzle flexibilities per WRC107/297 to have more accurate understanding and assurance of the stress conditions in the vessels and the piping. Along with the two nozzle flexibilitites, there would be added flexibility from flange joints with PTFE flange faces and any gaskets. A single PTFE envelope gasket with its compressible core would be able to absorb the 1/16" total thermal growth. Two gaskets could be better. If you want to include a very small gap for piping fit-up, then the cold spring effect could offset the theorectical forces of millions of pounds. Since the short PTFE lined spool does not accomodate piping fit-up tolerances between the equipment nozzles, you might want to include a solid PTFE spacer or two for adjusting by thickness or taper.

The probable two phase flow out of the reboiler would 'hammer' an expansion joint, so it is understandable that the client does not allow an expansion joint. I would avoid an expansion joint for almost all corrosive services.

The designers did the right thing to support the two equipment items at a common elevation. There was another separation column and reboiler of zirconium that had the reboiler supported 30 ft up in structure. The column was skirt supported at grade. Vertical differential growth between the reboiler and column would have required an flexible expansion joint - there would not have been sufficient space for large piping loop. It was early enough in design that the column was revised to be lug supported and structural steel was revised to carry the weight of the column and reboiler at the same elevation near the top of the reboiler. The reboiler had a vertical outlet with elbow over to the column for a small amount of piping flexibility.