Cold Start up of Pressure Vessels

[AndrewTravers]

Good Day,
I'm on the hunt for suggestions and hopefully some references on how to start up a pressure vessel from a state colder than design.
The vessel has an MDMT of -20 F, but the environmental conditions are such that parts of the vessel the vessel could be at an ambient temperature of -50 F when they go to start it. The vessel is insulated, with a heated building on one end. The areas actually exposed to ambient are large diameter nozzles, such as the manway and inlet connections.
The unit can't operate at this temperature for very long, as it's a gas/oil/water separator in a gas plant. Basically the customer is looking for assurances and recommendations to get the unit up to pressure long enough for the process to start moving, at which point the temperature of the process will keep the unit above MDMT.
Having double checked my design, all the components are good for -50 F or lower, due to UCS 66 allowances. The vessel has already been designed, registered, and built, so impact testing the welds isn't an easy option, and it appears that is the only area where we may need to be cautious.
My supervisor has already suggested restricting the design pressure as the temperature decreases, but we're hunting for references. I have been thinking about using values from Fig UCS 66.1 as a basic guideline.
If someone has a start up guideline for a non-fired pressure vessel in a similar situation, it would be a great help.

Thank you in advance for any help you can provide.

Regards,
Andrew Travers, C.E.T.

 

[owg]

Having commissioned a few vessels in the oil sands, where there was a good chance they had seen -40 deg F, I have one general suggestion. In addition to the precautions that you plan to take, I suggest that a hydrotest is important. In some cases the hydrotest can be substituted with a steam test but this should not be general practice. The benefit of a hydrotest is that if anything leaks or even breaks, only a bit of water comes out. With a steam or gas test, on failure, projectiles can be generated. Sorry I no longer have a complete procedure for you.

HAZOP at

http://www.curryhydrocarbons.ca

 

[dcasto]

keep the pressure low and do not pound on the vessel with a ballpean hammer until it is warmed up. Many vessels that were stamped -20 at 250 psig and now operate at -40 and 15 psia are in service. Look up in section VIII about chillers in the service. I believe as long as the vessel is under 2" thick, there is no problem.

The real problem will be in the piping as it warms up and grows longer.

 

[AndrewTravers]

Thanks to both of you.

OWG
I think the idea of hydrotesting the vessel just before start up would get discarded as being impractical, not least of which due to the temperatures involved (at least using water). However, if the liquid levels are then blown down by the process gas, the vessel wouldn't have the chance to drop much below freezing as the pressure increased.

dcasto,
We're already suggesting easing the pressure up, but we don't have a lot of resources as to what is considered a reasonable derating, other than UCS 66. In actuality, they could run it at the operating values and they wouldn't be more than 15% of the design pressure of the vessel, from the start. They would face a lot of issues though on the process side if things didn't heat up pretty quick.
I do concur that the real issue is the piping. I suspect this is why the end users are the ones who normally write these procedures, we don't have a lot of detail on that side of things from them.

 

[owg]

AndrewTravers, "hydrotesting" was done with a water/ethylene glycol mix, around 50/50 as I recall. I think that is good to about -45 deg F.

HAZOP at

http://www.curryhydrocarbons.ca

 

[AndrewTravers]

owg,
I'm not saying it's a bad idea, please understand. I'm just saying that it probably will be discarded as impractical, especially since they are fixated on -50 F and EG/Water mix isn't going to reach that far down. Just filling the vessel partway with EG/water may wind up being the answer (as it would facilitate heat transfer from the cold end to the hot end.

 

[SJones]

thread794-259202

Again, review the definition of MDMT. You need to analyse the stresses and temperatures achieved through a start up pressurisation and see whether you remain within the allowable envelope.

Steve Jones
Materials & Corrosion Engineer

http://www.oilandgaspeople.com/cv/11499664

http://www.linkedin.com/pub/8/83b/b04

 

[owg]

Good luck AndrewTravers. I see that 70% EG in water is good to -60 deg F, so you should be OK at -50. It turns to slush anyway when it freezes so it is not so bad.

HAZOP at

http://www.curryhydrocarbons.ca

 

[FMJalink]

AndrewTravers,
The problem lays in pressurizing a vessel leading to a specific stress level, when the wall temperature is below that of the acceptable ductility.

The problem occurs frequently and is normally caused by a misinterpretation of the design criteria (ambient temperature) and operation sequences (not continues service). Not only less or not insulated parts will get cold in sufficient time, but the entire vessel in the cold atmosphere.

If it is only for start-up, you can add heat tracing on the vessel, which will be difficult when the vessel is already insulated.

Another method is using a stream with a reduced pressure to heat-up the vessel. This should be safe guarded by sufficient measures as instrumentation etc. Heat tracing will normally be easier to imply.

Reducing the allowable working pressure as the temperature decreases can be quite difficult, pending on your position in the project. Normally a supplier will not be allowed to supply a vessel capable to less than ordered for.