Determine MAWP of in Process Exchanger

[Headsup]

I am looking for a way to determine the MAWP of exchangers that have no design or nameplate information. The catch is that the exchanger will not be opened up for internal UT. These exchangers have been in process for a number of years. Can I rate this exchanger by performing calculation on external components (Shell and heads) that can be UTed and performing a pressure test UG-101M (stopping at 4 times operating pressure)?

If there is another way to determine MAWP in this case please let me know as I don't like the thought of testing at 4 times operating.

Thanks

 

[deanc]

Strongly suggest you contact your local Jurisdiction and Insurance carrier before you spend more time on this.

 

[PAN]

I understand that you can calculate MAWP. In this case, corrosion allowance should be zero.

 

[TBP]

I agree with deanc - without documentation & nameplate data, I doubt you have anything to work with, whether they get opened for an internal inspection or not. I suspect that the inspection people tell you that you have HXs that are suitable for only low pressure service, and that you require safety/relief valves/rupture discs to ensure that they are.

You won't test at 4 times operating pressure anyway. It'll likely be 1.5 times (max) of the safety valve setting. I would NOT do a pneumatic test on these things, just hydrostatic.

 

[robsalv]

Regulatory issues aside, get yourself a copy of API RP579, the fitness for service guide - failing that, ASME VIII [or local equivalent] and probably TEMA. By the way, you'll never know with certainty until you open the hX but you may be able to improve your confidence that the hX as installed, is at least OK for the pressures it's seeing.

Do you have drawings?? Get some! Or Make some. Get a competent drafty to site measure and detail up as much as possible about the equipment including flange ratings and carry out a detailed UT survey.

You are going to have to reverse engineer this baby as best you can - but best to work this issue from both sides.

Side one: If you don't know the DP & DT, then, tough luck, you have to set one. What DT & DP does the process need from this piece of equipment? You have to do this for both sides of the hX.

Side Two: Do you know the construction quality?? If not assume the worst joint efficiency.

Do you know the materials of construction? If not, then assume some poor quality version of what you think this is. (May be you could get a spark test done to determine broadly the material spec.)

The type of hX will have a large bearing on the confidence of the calcs. Is this a fixed tubesheet hX? Or a removable bundle floating head?? U tube?? Do you know the tube diameter? Tubesheet thickness?? Your calc will be softest in this area - you will have to tread warily since sight unseen, you are making assumptions about the condition of the tubesheet(s), tubes, floating head etc

Now, settle on a nominal thickness for all accessible components[RP579 can help here]. Be aware though, that a UT survey might not pick up pitting. Probably better to use "time of flight defraction", or large structure inspection type techniques for more confidence.

If there's corrosion, then set a corrosion allowance, because calcs based on the full thicknesses are only as good as the next bit of corrosion that occurs!

Anyway, having done all that, now open RP579 [especially if you have corrosion - determine the effective thickness] or use the pressure equations in your pressure vessel code, but reverse them to solve for pressure.

Determining a basic MAWP for a clear area of shell is relatively easy, but understand, that it might be the nozzle or some other component that limited the shell design. Only a more detailed reverse eng calc will show this up. However, as a preliminary calc, and given that most designers design for the shell, then beef up nozzles to suit... you can have some confidence in the number. Be conservative: Use low joint efficiency, use the allowable stress for the poorest steel you can have confidence for at the DT you determined, use a conservative wall thickness and take off the corrosion allowance.

For tube side, the channel/bonnet/head/<insert descriptor> can be calc'd a similar way for the tubeside DT, but care is needed for tubes and tube sheet calcs. Tubes are a guess unless you have a drawing or history about a past retube. Tubesheets can be estimated from flange thicknesses [fixed tubesheet] or exposed tubesheet between channel and shell flanges. [Allow corrosion allowance on tubesheet too] Floating heads are a stab in the dark.

Tally up all the determined pressures and then compare them to the DP you said you needed.

Hopefully your hX is better than the DP you need and might take the heat off shutting the plant down until a suitable window presents itself.

Hope this helps.

Rob