Wetted vs Unwetted Fire Relief

[tenac]

We have a couple of gas/liquid separators. We calculated relief requirements due to fire considering the wetted surface heat absorption rate (API 521 eq. 3).

Were wondering about the possibility that the liquid could completely evaporate. After that, expanding vapour would have to be relieved (e.g.: API 521 eq. 8).

In some cases, this can lead to larger orifice sizes...

Has anyone heard about this design practice??? All companies I know determine wheter there is wetted or unwetted surface and 'stick' to that selection.

Of course, if there are possibilities of having vessels emptied (some scrubbers) we would consider the largest of both relief requirements, but this would not occur during normal operation for this vessels.

Thanks a lot.

 

[Montemayor]

tenac:

A Pool Fire scenario is a serious and very hazardous condition imposed on a pressure vessel. That’s why this PSV Forum serves such an important function for all engineers.

Your query is not complete and detailed in the important items involved when considering a PSV for the Pool Fire case. For example:

1) You don’t’ state the conditions under which the gas/liquid separators will be under when the scenario occurs. Will the separators be functioning in a steady state or will they be blocked off? I presume they will be blocked off by ESD valves.
2) You don’t state the relative sizes of the equipment with regards to the amount of liquid in the vessels. I suspect the liquid inventory (when the fire scenario occurs) will be approximately 10-15% of the total cylindrical height of the vessels. In other words, the majority of the vessel will be vapor/gas filled.
3) You don’t state the actual expected heights of the liquid level or the separators above the expected pool of fire. I suspect that these separators are relatively “small” and no more than 6 to 8 feet in total height. If this is the case, then it means that the vessels would be totally enveloped in a resulting pool fire. The entire vessel would be surrounded by the fire and its damaging impact and radiation. The amount of liquid inventory serving as a heat sink would be of relatively minimal amount and help.
4) You don’t state if your vessels are insulated or have a water spray available.
5) You don’t identify the fluids in the separators and their flammability/toxicity/ etc.
6) You don’t mention if the vessels are subjected to a de-pressuring action in response to the fire.

I believe you should be reviewing the essential, basic criteria in RP API 521. Specifically, you should be focusing on:

a) 3.15 External Fire; and,
b) 3.19 Vapor Depressuring

I also recommend you refer to thread124-143554. Unfortunately, there are a lot of informative and important PSV queries that are inserted in other Forums instead of being put in this PSV Forum. This is a very important subject and one that should be thoroughly understood as to hazards, dangers, and repercussions by the engineer(s) designing the proper and safe relief devices. It is good that you are employing the API 521 Recommended Practices. It is also important that you should understand the intent and the spirit under which the API writes these RPs. If you have a pressurized vessel that has a relatively large gas volume exposed to a direct pool fire, it will undergo the danger of mechanical failure or rupture in spite of having proper PSV protection. The gas within the vessel is incapable of acting as a “heat sink” fast enough to keep the vessel walls cool – much like a boiling liquid would do. That is why the API has had to introduce the technique of depressuring such vessels. In fact, I believe the API has been working overtime on this subject and has revised this section various times, seeking to improve the methodology and the vessel protection.

If you depressure the separators (& perhaps employ insulation and water spraying) you might save them during the fire. If you maintain pressure within the separators and the fire intensity and duration is enough to exhaust (“boil”) the liquid inventory, you would undergo a subsequent major vessel failure or rupture while it was under pressure (at the PSV set pressure). You don’t want to go there. This is the reason for the API 521 sections dealing with gas vessels during the fire case. As gas/liquid separators your vessels might fit under the category of gas vessels and this is the reason I would be concerned.

I hope this helps you out.

 

[monaco8774]

Tenac

Consider that if you have an oil gas seperator at a given pressure (the relief pressure) the vapour temperature of your remaining liquid gets increasingly high (in other words the fire will melt the vessel before it boils the liquid off) This is also why overating design pressures on vessels and using an elevated set point on the relief valve is not smart.

If your set pressure is a normal margin above operating and you have heavy end liquids (normal crude field) in the vessel then it is appropriate to consider liquid wetted surfaces, since your relief valve will do nothing to protect you against the high temperatures that would be required to boil off the heavy ends.

If you have a gas condensate field with water and light ends only I would simply treat as unwetted.

In either case you should have a system blowdown the vessel on fire detection to remove hydrocarbons.

 

[tenac]

Montemayor/monaco8774,

Thanks for your replies. Thanks for the recomendation about thread124-143554 - I had read part of it, but not the last and more interesting posts.

I only wanted to check others opinions and some general criteria, instead of boring you with numbers, levels, operating conditions, fluid properties, etc. Maybe I didn't express this clearly.

Anyway,
It's an horizontal vessel, normal liquid level =45% diameter, diameter about 6 ft, length 24 ft. The vessel will be blocked off by SDVs. Of course we have included a depressuring valve. Operating pressures is about 1000 psig and PSV set pressure is 10% higher (a very usual margin at this levels of pressure). It handles condensate and light ends.

The vessel will be insulated, but insulation specifications don't guarantee that it will resist fire (according to API 521 criteria). There will be a water spray, but API 521 doesn't allows taking credit for this. So, I am not reducing PSV size for any of this factors.

Time to boil the liquid, considering the initial wetted surface, is about 40 minutes. Of course, liquid level and wetted surface would decrease, but this are just heat absorption models. I'm ignoring water spray and insulation effects).

But we need to have a properly sized PSV as a last layer of protection. And my doubts were about the correct sizing of this PSV.

I have NOT seen additional verifications of this valves for the expansion of the remaining vapors in other companies... but I was in doubt about the need to also check that case. In some separators, dry fire calcs can lead to larger orifices than wetted fire calculations.

I do not agree with using the addition of the wetted and unwetted fire relief flowrates the as long as there is a considerable volume of liquid.

What's your opinion about the sizing case of the PSV?