Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Cooling Jacket Two-Phase Fire Case Relief

Status
Not open for further replies.

TiCl4

Chemical
May 1, 2019
631
I have a pressure vessel that has 3 stamped dimpled jacket segments on it. The jackets are all rated to 150 psig. I'd like to walk through my initial sizing methodology and assumptions and get feedback on the assumptions and things I might have missed.

Dimpled Jacket, required relief rate
MAWP: 150 psig @ 400 F
Location: Vessel bottom head.
Diameter: 92". Free passage space for flow is 0.3285"
Contents: Water when cooling, normally 60 psig.
Governing case: Fire, jacket full of water. Interior vessel empty - no credit for vessel contents providing heat sink to jacket. Assuming two-phase flashing saturated liquid flow.
Area, A, exposed to fire: 46.16 sq.ft.
Heat Input: 923,279 BTU/hr (NFPA 30 Eqn B2a, Q = 20,000*A)
Allowable overpressure: 21%
Relieving Pressure: 196.2 psia
Relief to atmosphere through short tailpipe, flow is critical.
Enthalpy of vaporization at relief conditions: 845 BTU/lb
Required relief rate: 1093 lb/hr vapor

Valve Sizing
I chose the two-phase methodology, API 520 C.2.3, due to the initial required flow (after thermal expansion), whereby steam evolving from the heat transfer surfaces will push the hot, saturated liquid to the inlet of the PRV, where it will flash to two-phase flow through the relief valve. Thus the 1,093 lb/hr vapor relief will push out 42.6 gpm of water at saturation conditions. Liquid will be pushed to the PRV initially due to essentially no disengagement space in the jacket. In addition to the nozzle on the picture below, there is an identical nozzle at the bottom of the jacket that I could use to essentially force the relief to be liquid-only. The spreadsheet I used to calculate this is attached.

The result of the calculation is a 0.53 sq.in. orifice

Discussion/Questions

[ul]
[li]This 0.53 sq.in. would call for an API "H" orifice size. However, the ASME "G" orifice has an 0.589 sq.in. orifice size. Would using a "G" orifice be possible if I re-ran the calcs using specific discharge coefficient from a manufacturer?[/li]

[li]For two-phase flow, should a relief valve be a dual-certified relief valve?[/li]

[li]The inlet losses do not appear to exceed 3%. The relief path (see picture below)connection to the jacket is 8" long and 3" wide, for a total area of 7.2 sq.in. The connecting pipe is 2" NPS, and a normal "H" valve that I've found has a 1.5" inlet. At this flow rate, pressure drop is something like 0.15 psi/10 equivalent feet. Comments?[/li]

[li]Lastly, is this exercise even useful? The jacket has an approximate volume of 10 gallons (92" diameter, void space is 0.3285" high), and will likely empty in 5-7 minutes, tops. After that, the vessel will contain vapor at 182 psig, but temperature will continue to climb until equilibrium is reached. I don't see the jacket surviving the combination of fire temps and 182 psig. What's the point of providing overpressure protection if the vessel will fail anyways due to high temperature?[/li]

[li]If the before point/question is valid, should I just document this case in the relief design record and state that the fire case will always result in jacket failure, and just put a much smaller thermal relief valve on the jacket?[/li]
[/ul]

This was a long question. Thank you for your time.


image_axdmui.png



 
Replies continue below

Recommended for you

I'd install small PSV for liquid thermal expansion, and for code compliance, and be done with it. The hours spent in engineering design, and the resulting PSV sized for 2-phase relief, are a waste of time and money which adds no value from a safety or a code-compliance perspective.

During a fire, the jacket will quickly be emptied of liquid and fail. That will occur regardless of whether you install a small thermal PSV or a big 2-phase PSV. Not to mention the fact that a 2-phase design will probably require a modification to the jacket (it's very unlikely you have a nozzle that's equal to or larger than the inlet to the 2-phase PSV).
 
That is the direction I was leaning too. However, in this case the nozzle is adequately sized for the two-phase PRV - we have a 2” and only 1.5” is required. The 3% loss rule is met as well.

Not installing the two-phase relief would solely be on the basis of imminent jacket failure upon vessel emptying.
 
If you do require a PRV you can get away with a smaller size.

I took the 1093 lb/hr as steam flow representing the worst case. If I run an api-520 steam calculation using 150 PSIG with 21% overpressure, 400 DEG C relief temp, I calculate that an 'E' Orifice will be suitable.

Depending on the operating pressure, you can use a 1" x 1.5" PRV with screwed connections (flanges optional, you'll also need a lifting lever to meet ASME VIII) - this selection is not API-526 but follows ASME VIII. You just need to watch that your operating pressure is typically no more than 85 % (of set pressure) as these types of PRV have a longer blowdown.

I took the angle that the liquid portion would be relieved first - the 'E' orifice from one vendor is rated for approx. 70 USGPM water at a set pressure of 150 PSIG.

You'll find that the established major PRV vendors offering this size valve have their PRV Trim designs certified for both liquid and vapour.



Per ISO-4126, only the term Safety Valve is used regardless of application or design.
 
Obturator - thanks, but I believe that the water will flash during the fire case because it is already at the bubble point when the PSV opens, causing two-phase flow through the PSV - this the higher required area.

There is no ASME requirement to provide over-temperature protection, is there?
 
On the feed and exit lines for this jacket, can you fit in an RV for this 2phase flow case at either of these locations?
 
Georgeverghese,

Yes. It can be mounted on either side. Mounting on the inlet will force majority liquid flow. However, in the fire case the jacket will still fail due to over temperature. So why protect it against fire overpressure when fire overtemp will occur and ruin the jacket anyways?
 
In design codes, the relief temp for firecase, is not relevant to the justification argument for relief provisions. Relief temp is only relevant for operating (non fire) relief cases.
If you can mount it on the inlet or exit line piping, then that would be better for this 2phase relief, and you can walk in on a safety audit, if and when it happens, without worries. May be better to have this RV on the elevated exit so dirt doesnt bung up this RV when it triggers, plus the RV would have excess capacity when it flows all vapor.
 
Georgeverghese, I guess my point is that the jacket will still exceed its MAWP because the temperature would increase above the rated coincident MAWP temperature - see my OP for rational.

Thus, the two-phase relief valve doesn’t actually protect against overpressure because the jacket will suffer overpressure due to pressure being at MAWP AND higher than design temperature.
 
There are no clauses in the ASME 8 or API that reflect the same observation you have here. In many all gas pressure vessels where only firecase gas expansion is applicable, the relief temp is well in excess of design T.
Unless you can find a clause in the ASME or relevant APIs' that reflect your opinion which argues the case for not providing a firecase RV, an RV here is unavoidable.
 
Georgeverghese,

I looked back through ASME Section VIII and you are right - UG-125 (c)(1)(2) requires relief during fire case to no more than 21% overpressure, with no exceptions.

However, what do you suggest for the other two jackets on the vessel? In their case, I calculate a required "J" orifice due to relieve the flashing liquid flow. The inlet of such a valve is 2.5", and we only have 2" ports available. In this case, all requirements of code cannot be met.

I think it best to take other protective measures. The building has firewater supply to it, so a sprinkler system can be installed to cool each jacket. I also am going to explore the water supply/return valves. We can probably relocate them in such a way that any pool fire would melt the supply tubing to the actuators, causing them to fail open. That would certainly eliminate the fire case from even causing overpressure.

 
Had a think about this yesterday; there may be a way to limit the relief to just single phase steam vapor:
If you install the RV on the jacket exit piping and elevate it some 1.5-2m higher than the top of the jacket, with an oversize feed line to the RV, it can be argued that only vapor will get to the RV. The 1-1.5m high oversize feedline will act as a disengagement drum to prevent liquid from getting to the RV.
Same approach for other cooling jackets. Run some calcs to estimate what size pipe you'll need for this oversize feedline.
An RV is mandatory for all coded pressure containment equipment, regardless of what instrumented failsafe devices you install to minimise overpressure during a fire - you may be arguing all day long with the govt inspector for pressure vessels if he/she turns up at your place.
 
Georgeverghese,

Overpressure protection via rupture disc or relief valve is mandatory, but I believe I can make the fire case non-credible. Thus, only a smaller thermal relief valve would be required.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor