Thermal Relief

[foxymophandlpapa]

When calculating thermal relief PSV's do you typically include fire as a scenerio? I would think a fire would count as two failures since the first failure would be blocking in the liquid. But fire has such a higher heat input, than say solar, I think it would still be considered.

 

[ash9144]

Yes, external fire is normally a scenario to consider. The purpose is to eliminate bleve potential.

 

[dcasto]

Is this vessels or piping thermal relief? Since you mentioned blocked in flow, I'd assume piping. On piping, I never size for fire, thermal only and with thermal I never use a relief valve, well very very seldom.

 

[foxymophandlpapa]

This is for piping. I ask because say for example, in a batch operation valves are closed for non-usage. A fire could occur and heat the piping. I don't see how you cannot include fire.

What type of valve are you using then for relief?? We always, always, always use PSV.

 

[dcasto]

Inside a plant area we size the fire relief for the vessels and add 10 to 20% for the piping. If the pipe is away from a vessels we do no size for fire because a pooling fire will still cause the pipe to fail anyway. Its a risk management decision.

For pipe that can be blocked in and is subject to solar radiation or such, we also allow for thermal syphoning if part of the pipe in underground. I use a 1 to 1 ration. If the pipe is more than 50% underground, their will be no expansion.

If the pipe is above ground, I use 1/4" or 1/2" tubing check valves around the valve that is isolating the the line so that the fluid could expand and fail the line. If the isolated line gets a pressure higher than the line it is isolated from, then a 1 cc of fluid will go through the check valve and into the other line and the pressure will drop 100 psi. (just some example numbers not real). I've done this and cascaded several sections. I also use this on valve bodies (as a matter of fact most DBB valves use a check valve in liue of a relief like they use to use 20 years ago and I started doing this in the lated 1980's)

 

[foxymophandlpapa]

Ya I've heard of using the check valves, but our customer requires PSVs. All our piping is above ground thus subject to fire within 25ft. We have taken into consideration solar heat as well as ambient conditions and heat exchangers.

I wasn't sure if fire should be calculated where it is possible because that, in my mind, would be a double failure: 1) blocking in the liquid 2) fire. But during batch operations where its reasonable to shut off piping, for example a 1000 ft monomer line, fire would then be a 1st failure.

 

[dcasto]

A friend of mine told me about his client that "required" PSV's too. When he pointed out that for piping there is no code or standard for a fire case, blocked in or not, and that the 2 - 8" by 10" where going to cost $MM they accepted the recommendation.

The fire relief in your case will have to go into a flare header. And after some time period the line WILL fail. You mentioned that the liquid was "monomer", not noing which one, ther may be other issues with keeping it from escaping uncontrolled so the relief valve may be for that case more than for protecting the piping.

Where I worked on time, Exxon saw our check valve arrangement and did the same thing on their butadiene lines and meters because they hated all the little reliefs and vents and flares. They still ended up with 1- 3/4" relief, but it lowered their risk overall.

When you state 25', the code says vessels surface area more than 25' above ground need not be considered in the calculation for fire. If you have pipe you would have to consider the topology and the area where a pooling liquid fire would subject the pipe to. It could be more than 25 feet.

I've seen flowing lines fail when there was a fire near them and there was no pressure rise in the line that a relief valve would open on. You best defence for a pooling fire near pipes is to shut in the lines and let the pipe fail. If its stuff you do not want out of the line, then even with a relief valve you need a way to push the material, like a nitrogen system.