Omitting PSV on pressure vessel 6

[dennisr]

It is well known that a PSV does not provide any protection against failure due to fire on an unwetted vessel. Many recognised companies practice not to install PSVs on unwetted vessel if fire is the only contingency.

Does anyone know of any officel reference, ASME clause or other codes that support the above practice?

 

[don1980]

CC 2211 (now UG-140(a))can indeed be used in a fire case. In fact it was a fire case for a vapor filed vessel that originated CC 2211.

This provision is applicable when there are no credible scenarios that can cause the vessel pressure to exceed it's MAWP. For vapor filled vessel, exposed to fire, the pressure will rise as the gas heats and expands, but that pressure increase will usually not reach the MAWP before the vessel fails. Even if the pressure does reach the MAWP, or set pressure of the PSV, the vessel will not be adequately protected because the heat is not being carried away by a boiling liquid.

 

[don1980]

dennisr.....my information comes from the company that made the original inquiry that initiated CC 2211.

Regards

 

[JoeWong88]

don1980,
Thanks for your information.

I am still wondering how a HIPS system can help in a vessel expose to fire...

JoeWong
Chemical & Process Technology

 

[don1980]

Joewong....We're discussing UG-140(a) (used to be CC 2211) rather than HIPS. Your right in saying that HIPS can't effectively defend against a fire scenario, but that's another subject. Vapor filled vessels, exposed to fire risks, are candidates for UG-140(a).

From your question, I'm guessing that you might be referring to UG-140(b) which allows the use of instrumentation to build a case for "overpressure by system design." That instrumentation can be HIPS but it doesn't have to be. I've heard others that interpreted it that way too, but that's an overly restrictive interpretation. It can be HIPS or it can be non-HIPS.

Personally, I think the wording could have been a little clearer. I think a lot of people are going to misinterpret UG-140(b). I'm also very surprised to find practically zero discussion on UG-140 when I search the internet. This was published by ASME last July, but I don't think a lot of folks have noticed it yet.

Regards.

 

[dennisr]

I have just received the new revision of API-521 (2007). In the new revision it is clearly written that, if accepted by local regulation, relief valves on unwetted vessels may be replaced by other means of protection such as water spray or depressurization.

Also the new revision distinct between pool fire and gas jet fire. In case of gas jet fire only (gas plants) the relief valve does not provided any protection and should be replaced by other means of prtection.

Also ASME VIII UG-140 allows relief valves to be omitted if allowable overpressure cannot be exceeded. This clause also refer to API-521.

Conclusion: In our synthesis gas plants we will consider to delete relief valves on unwetted vessels (reactors) and install depressuring valves instead.

 

[na5rc]

When there is no basis for sizing, it becomes difficult to justify selection. Some vendors don't want to supply a valve without a basis and without a calculation. I have seen process engineers make up an imaginary liquid for fire sizing when no other case exists and still use the properties for the gas! This is not a valid calculation.

Or another basis that does not work is "thermal expansion" for gas. In this case, Carbon Steel will often melt before thermal expansion of the gas can reach relief temperature. Thermal expansion of gas is almost never a viable case. Using it when it is not viable raises questions.

. . .and so the question still arises; "what is the basis?". The basis is "Nominal valve selected without sizing because no viable relief case exists. Valve supplied because ASME Section VIII requires it." Maybe this will work.

I know that made up cases and calculations don't work when they are scrutinized by approval agencies, et al.