u tube heat exchanger.

[Merzi]

During insertion of a U tube bundle, it was observed that because of the ovality in the shell it was difficult to push the bundle inside. In this case what is the best solution.
a) to grind the baffles ( which i think is not the best thing to do) or
b) grind the areas in the shell

Merzi

 

[OwaisGillani]

What makes you think that grinding the baffles is not the best thing?

Do not play with pressure retaining parts.

 

[metengr]

Most unfortunate. You can do both, provided you evaluate the remaining thickness of the shell (pressure boundary) in the areas where you have removed metal locally, and do not fall below minimum wall thickness. I would suggest verifying the remaining shell thickness using ultrasonic thickness testing (UT).

Re-rounding of the shell might be possible with local heat, but could be tricky to do in the field.

 

[codeeng]

A lot of end users increase the TEMA clearance of the baffle/shell for this reason. In your case I wouldn't touch the shell base metal, just grind the baffles and/or any excess weld build-up.

 

[marsich]

increase clearance of the baffle/shell is ok only where clearance has no significant effect on shell side heat tranfer. See temaRCB 4-3.

 

[deanc]

If you grind the baffle you should make sure the tubes will not contact the inside of the shell during operation. Sure path to "boating" and leaks.

I would suggest a "porta-power" and jack it back into place from the inside or the same manner from the outside. You will need to yield the shell twice as much as you need to keep it.

It would be possiable to follow metengr's suggestion. Contact a "R" stamp firm.

There may be other problems with this item such as excesive loads that caused it to go out.

 

[SnTMan]

Doesn't help much now, but good practice is to "plug" the shell with a couple of discs the same OD as your baffles bolted a foot or so apart, before saying the shell is done.

That way you know if you're going to have a problem.

Good luck

Mike

 

[panduru]

@Merzi:

You have not indicated to what extent the ovality is - in other words, the amount of linear mismatch in the radii.

If the difference between the minimum internal radius of the shell and the radius of the baffle plates is not much, then you may attempt to fit the tube bundle under differential temperature (expand shell, insert bundle, cool), though this requires extreme care and discretion. I do not know, but this may even coerce the shell into a near-circular section.

Grinding the plates may effect the shell side heat transfer as marsich has pointed out.