Reliefs on pressels - Need clarification 1

[UtilityLouie]

I know that all vessels have to be protected from over pressurization, but what about these scenarios:

Air receiver rated at 150 psig, but the maximum discharge pressure of the air compressor is 125 psig. Am I correct in thinking because the system cannot ever exceed the design pressure of the tank, there should be no relief required.

Steam to glycol heat exchanger rated at 150 psig. I know there is a relief required on the water side, but the steam side (also rated at 150 psig) has 60 psig steam supplied to it. At the steam generator, the 60 psig system has reliefs installed with a maximum relief pressure at 80 psig. I would guess again that no relief is needed on the steam side.

Am I correct in these assumptions?

Thanks for the response.

 

[TD2K]

If the 125 psig is the compressor's maximum discharge pressure at no flow, then blocked discharge is not a relieving case.

However, air receivers can contain water and you would typically have a relief case if the vessel was blocked in and you had a fire vaporizing the water. I've seen some companies use operating procedures in this case and take credit for the normal outlet flow as a source of relief(essentially, if the receiver is ever to be blocked in, the vessel would be drained and depressured, thus if there was a fire, there would be no overpressure).

For your glycol exchanger, I would agree with your rationale on the steam side. Again, you could say there would be a relief case if the exchanger was blocked in and you had a fire around the exchanger since the steam side would have condensate that could be vaporized by the fire. However, that is frequently handled by operating procedures (isolate and drain any condensate if the exchanger is taken out of service).

 

[spector]

They are both pressure vessels designed to a MAWP. They must be protected accordingly. The relief device is required to prevent failure of the vessel, not to protect the system or generating components. Keep it simple. Don't talk yourself into a dangerous place.

 

[TD2K]

But you need to have a relieving case.

I'm currently working on a project where the client is not installing thermal relief PSVs on the diesel filters and going with operating procedures to have the outlet left open on the standby filter. I've seen other clients not install a PSV on the cold side of a S&T heat exchanger and depend on operating procedures and this is a major oil company.

Now, I agree with a pressure vessel like this, not having a PSV would be unusual. But what sizing case would you go for? Blocked discharge isn't a problem in this case IF the compressor truly can't put out more than 125 psig and the vessel is designed for 150 psig.

I'm not sure I would feel a lot safer with a small PSV on it sized for a fire case as that is the only case left that I see as a likely possible case with the information we have been given. If you had a fire around that vessel, a PSV is not going to prevent the vessel from failing in any event when the steel gets hot enough from the fire that it doesn't have sufficient strength.

The same argument as to the value of a PSV on the shell side of the steam exchanger applies.

 

[TD2K]

Couple of other comments here. ASME section 8, Div 1 requires all pressure vessels to be equipped with a pressure relieving device, no exceptions. CHD01 has commented that there is a code case where the code committee has agreed that if there is NO relieving case, you don't need to provide a PSV. The kicker can be if your state mandates ASME Section 8, Div 1 as law but does not recognize code cases as being legally binding (I think CA is one state, you can do a search on this site). If you are in such a state, you'll need to provide a PSV in any event or some other recognized pressure protection sized for 'something'.

Personally, I think adding a PSV or something similar to a piece of equipment with no real basis provides no protection to the equipment or personnel but you may not have a choice. On other hand, you need to be careful that you don't accept the risk of overpressuring equipment by depending excessively on operating procedures or saying 'such and such won't happen'. As a rule, I won't depend on procedures unless the consequences are not serious. For example, if you have a liquid filled diesel filter and it's blocked in and warms up, you will very likely leak at a gasket versus failing the filter housing (and being liquid filled, there isn't a lot of stored potential energy in the filter compared to a gas filled vessel).

Only, enough rambling for a Saturday morning.

 

[jte]

TD2K -

An interesting thread - which has been touched on before, but is worth rehashing. You are right as far as various states using / mandating various codes and standards. California mandates the latest version of VIII and accepts Code Cases on a case-by-case (no pun intended!) basis. No one has set the precedent (as far as I know) for the code case which allows overpressure protection through system design in California, so it is not legal here. One of the wonders of having a bunch of lawyers making laws is that another code which the State mandates is NBIC 1992. They managed to specify the edition instead of "latest edition" and we are stuck with it. Any use of later editions of NBIC '92 is at the discretion of the California Pressure Vessel Unit, again on a case-by-case basis.

jt

 

[EngineerBSME]

KISS
Keep it simple stupid! THINK!
The safety relief valve is always required.
No exceptions in all states where the law of physics applies.

Suppose, for example, a filled vessel is closed. The sunlight shines through the window, a nearby fire occurs, an electrical or chemical hazard develops. The vessel temperature rises. The pressure increases and exceeds the design pressure of the vessel. A catastrophic event occurs.

The original ASME B&PV code was developed because household coal fired hot water heaters ran out of water and got cherry red. Occupants panicked. Attempt to cool the hot water heater by adding water blew the entire house to smithereens. The inlet water pressure was 60 psig. The tank design pressure was rated at 2000 psig. No problem yet. The available pressure was upwards of 40,000 psig when the first drop of water contacted that cherry red vessel.

The vessel design pressure determines the setting of the relief valve. Pump size is not a factor.

 

[EGT01]

For those who are interested, I believe the Code Case TD2K and CHD01 refers to is Code Case 2211. I first came across this mentioned in the back of a Chlorine Institue pamphlet (I think is was on chlorine piping systems). I've since found it being referenced by those promoting Safety Instrumented Systems (SIS). Take a look at

http://www.clarkson.edu/~wilcox/Design/hipps.pdf

I've not officially obtained a copy of Code Case 2211 but as I understand it details under what conditions a pressure vessel may be provided with overpressure protection by system design in lieu of a pressure relief device as required by Sec VIII, Div 1, para UG-125 and Sec VIII, Div 2, para AR-100. Interestingly enough, one of the conditions of Code Case 2211 is "the vessel is not exclusively in air, water or steam service." Not sure what to make of that.

You may also want to check out what another forum has to say about it

http://www.cheresources.com/forums/forum5/messages/10050.shtml

(sorry, but the above link doesn't seem to post correctly, may have to copy and paste it into your browser address window)

 

[UtilityLouie]

I guess I never had any questions about putting the reliefs on the water side of the vessel. My big question was on the steam side.

If you do need a relief on the steam side ( and I guess now I will put one on just to do it, not because I believe that it is necessary for engineering reasons ) how would you size it?

The worst case for the steam side of the heat exchanger would be the steam valve being closed and the volume of the shell condensing to condensate, the condensate not draining, a vacuum breaker opening during the condensing process bringing the vessel back to atmospheric pressure and closing, and a fire that repressurizes and overpressurizes the shell. I could probably get away with a 1" relief valve, but if I sized it for total design steam flow through the unit, it would be an 8" valve. There are not always big cost implecations, but there could be.

 

[greg87]

Selection of a reasonable, yet conservative, basis for a relief system sizing criteria can be anything but simple. In addition to complying with what is “required” by the Code, the engineer should satisfy his/her team that the scenario analysis and the sizing criteria is thorough and reasonable and provides a safe system (and "safe" is anything but black and white). Knowledge, analysis, and documentation of your system is key to the process.

For “dry” vessels with steam or air service and having no credible relieving scenarios, selection of the sizing basis is a dilemma. For these vessels, actual failure in a fire will most probably not be caused by pressure buildup, but due to wall failure due to high temperature, which would fail before the relief pressure is reached.

For “dry” vessels with relatively low pressure ratings, small size, and low fire hazard (and no other relieving scenarios), I will provide a relief valve sized for the fire case, based on the assumption that there will be, or could be, some condensate in the vessel, using the entire surface of the vessel in the calculation. I feel that this complies with code (and an inspector's interpretation of code), provides a high degree of protection in the event of fire, and provides a measure of protection for an unanticipated event, without incurring high cost or complexity. (this seems to be the case with UtilityLouies' systems) For vessels with high pressure rating (and relief setting) and higher fire hazard areas, I may not be satisfied with this approach. In that case, the Case 2211 design approach and measures that are an option (or in addition to) relieving devices would be applicable.

 

[twjag]

For ASME VIII, 1 the requirements are laid out in UG-125 thru UG-137. UG-133 addresses capacity requirements including operating loads, fire protection, systems of interconnected vessels, internal failure for HX's, etc.

 

[Keef]

In answer to the question on how to select a safety valve on the steam side, it is usual to ensure that the safety valve is capable of relieving the full flowrate of steam that can reach it. This provides safety from the possibility of say an upstream pressure reducing valve failing open, allowing full boiler pressure onto your vessel. The maximum flowrate is usually restricted by the valve upstream of the heat exchanger and can be calculated by using the Kv of the valve and the upstream and downstream pressures. The set pressure of the valve should not be above the maximum pressure of the vessel but may be lower. Once you know the set pressure and maximum flowrate, you can use a safety valve manufacturers chart to select the appropriate valve. Most safety valve manufacturers will help you in calculating the maximum required discharge capacity if required, I know Spirax Sarco do.

Keef

 

[UtilityLouie]

Keef,

I would agree with that in a one boiler, one relief system. This system has a total of 12 relief valves on the header. I would say that some other criteria needs to be used in this situation. These valves require maitenance and if you get too big of valve just because it is easy, it ends of being an expense item that maybe you didn't need.

 

[Keef]

UtilityLouie,

It depends on the pressure rating of the vessel and your steam pressure. If you do reduce pressure before the vessel, it is usual to size the safety valve in this way. I should also add that inspectors usually like to see a safety valve mounted after a reducing valve, even if the vessel can take full boiler pressure.

I am wondering why you have 12 relief valves on your header. You have to have at least one per boiler (depending on what country's standards you have to comply with) but from there on you need them to protect equipment that cannot take full boiler pressure. What are these 12 valves for?

I fully agree that you do not want to install a safety valve larger than required (it can lead to problems in itself), but if your safety valve is too small, how much point is there in having it at all as the system pressure will build up above the level you have set as your limit.