MDMT - Calculation Affected Due to SRV Throttling 1

[RPRad]

Strange question....one of my clients is being questioned by the AI with regard to whether the vessel MDMT reflects the low temperature which may result locally in the shell and nozzle in the event of a SRV opening, somebody (not I) has done calculations showing that theoretically the SRV throttling temperature could get as low -49F (vessel normally operates at 425 Psig / 300F). Hard for me to believe that this would be anything but a very localized condition, but admittedly I have never stuck around to see if frost is forming when a large SRV goes off!

Firstly, is this even a Code consideration when calculating vessel MDMT, being a rather abnormal intermittent pretty localized operating condition in itself...although admittedly Autorefrigeration is mentioned in UG-20. Secondly I have never heard of a failure resulting from such a thing (there have certainly been failures due to fatigue and / or high vent piping stresses and I guess perhaps low material toughness could have played a part in accelerating a failure if the local material temperature due to throttling has occurred, but I dont believe I have ever heard of it directly being the cause of a failure). The shell material in question of course is a 1970's vintage SA515 Gr70 which can have legitimate toughness issues at low temperatures...but this is not a Cold Start up situation we are talking about and the vessel in question is not located outside

Has ASME ever put out an interpretation on such a thing? Seems a bit beyond the scope of what the purpose of the MDMT calculation is all about

R

 

[SJones]

By the time that the material gets to -49F, what is the pressure? The AI can then satisfy himself against the definition of MDMT in UG-20 (b). Seems a classic case of confusing MDMT with MAT (API 579-1).

Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/in/drstevejones

All answers are personal opinions only and are in no way connected with any employer.

 

[DriveMeNuts]

It is a valid question relating to the SRV. It is for this reason that SRV's tend to be made from Stainless Steel.
However if the concern is for the remainder of the vessel, it is odd.
MDMT stands for Minimum Design Metal Temperature.
Perhaps your AI is getting confused with Minimum Fluid Temperature.
The fluid may get down to -49F due to depressurisation, however I doubt the mass of fluid has enough energy potential in it to cool the vessels entire metal shell by more than a few degrees. And it would take time for the fluid to metal heat transfer to occur, by which time there is no pressure in the vessel.
Perhaps your AI is trying to be too clever.

 

[hacksaw]

Since your dealing with a safety relief valve, it will reseat when the pressure drops if not it stays open until the tank contents boil away eventially at atmosperic pressure.

 

[r6155]

Anything can happen. Consider that brittle fracture may occur during the hydraulic pressure test, see UG-99.
Many PV broke during test.
What is the MDMT indicated on the nameplate?

Regards

 

[weldstan]

Your AI may have seen many PVs whose MDMT has been based on Joules Thompson cooling when SRV goes off; however, that is not always done because the likelihood of the metal temp being reduced to the calculated fluid temp is often nil (based on mass heat transfer calcs performed) and the pressure is negligible at the calculated fluid temp.

 

[RPRad]

Thank you all for response, yes I believe this seems to be a bit over the top in terms of issues to worry about, as indicated by your response above most of the effects would be localized more to the SRV. The posted MDMT on the vessel is -20F, but thats not really correct when you do the calculation its more in the range of 30F...-20F probably came from a UG-20 exemption clause, I havnt tracked down a copy of the 1970 Code to see what it said actually with regard to MDMT at the time.

The AI of course claims that there have been (albeit a small number) cases of cracking at the shell / nozzle weld interface and that the cause was at least partially due to low temperatures generated by the SRV throttling....no he was not forthcoming with actual cases. My guess is that the material toughness in these cases was likely poor to start with (and of course a weld zone could exacerbate the problem) and throttling gradients were not really the primary cause of the failure or caused a sudden change in material toughness. I just cant see the a throttling temperature gradient extending down a flanged nozzle to create a severe MDMT issue at the shell/nozzle interface nor can I see this happening when the vessel is chugging along at full operating pressure and temperature...I would put higher odds on it occurring during a cold start up when the metal temperature has not yet reached operating temperatures. I am also guessing that if this has occurred its probably in older vessels utilizing materials with a more coarse grained structure and I dont know how you would even be able to remedy it after the fact (I guess you could move the SRV further away from the vessel to appease the AI..doable but not ideal)

 

[DriveMeNuts]

SRV's are designed so that the low temp outlet material where cooling occurs is a long distance from any carbon steel. The warm process fluid flows up through the SRV, warming the base of the SRV, preventing the cold from conducting down towards the vessel. It is part of the SRV design, otherwise the SRV could ice over and seize up.
Your other option is to find another AI.

 

[RPRad]

That would be my assessment on both points....unfortunately the latter is not within the current realm of possibilities!

 

[georgeverghese]

Leaking SRVs' would ice up down to the inlet flange and even further upstream. So check the piping materials for the inlet flange and piping upstream. Presumably this SRV is not direct mounted on the exit nozzle of the vessel?

 

[r6155]

Be careful with residual stresses.

"99 Diseases of Pressure Equipment: Brittle Fracture"
See

https://inspectioneering.com/journal/2007-01-01/3683/99-diseases-of-pressure-equipm

Regards