Thermal Relief Valves

[npf]

All:

I am sizing thermal relief valves for liquid LNG. Typically we do not size those and use default, but the client wants to see the calcs. These valves are located on 26" diameter long and short run of piping.
The pipe length is approx 1000ft and is insulated.
Do we have to consider external fire case as a possible scenario for relief valves? If that is the case are we considering double jeapordy along w/ the blocked in the condition.
Also during fire all the liquid flow is stopped to the fire risk area and it would essentially be blocked in.

If I have to take a quick shot at the fire case, what pipe length should I assume to be exposed to fire?

Your comments would be appreciated.

Regards,
npf

 

[molecularman]

I think you are right that this is a double jeopardy case to be blocked in and have a fire but if you feel you need to calculate a fire case I would think API's fire circle area of 2500 - 5000 ft2 would make sense (I'm sure 2500 is mentioned and think the range goes up to 5000 ft2).

Therefore calculate the diameter of a 2500 ft2 - 5000 ft2 circle and use that length of pipe for the fire case. Just a thought.

Note that at our refinery we don't consider piping fire case scenarios and I don't think many others would do so either.

Also, if you needed to calculate the relief rate for the thermal expansion case, you'd need to calculate the volume of the pipe line in question and assume some sort of heat absorption to the pipe from some hot ambient condition and then apply the coefficient of thermal expansion to see what the rate of volume increase would be.

 

[SeanB]

I would second that we don't normally consider fire case for piping. You also have to evaluate whether or not you have a case for thermal expansion. You say your pipe is insulated - is it heat traced? What is your normal operating temperature? What you need to decide is if your pipe is blocked in full of liquid is it possible for the fluid to increase in temperaure? If so, then you have a thermal relief case.

 

[dcasto]

I normally include a 5% to 10% added capacity of a vessels fire case to include lines.

As for solar thermal, you can use the solar thermal calculation on the line and then calculate boil off rate just like any liquid line. The thermal relief valve will be a 1/2" to 3/4" valve OR you can provide a check valve around one of the block valves that would relieve the boiled up liquids back to a tank or vessel that has a thermal relief valve on it.

 

[ggordil]

I do not agree that assuming that a blocked outlet during fire is necessarily double jeopardy. It really depends on the system. For example, if there is a check valve upstream of you length of pipe and there is a control valve that is expected to fail close during fire downstream, then the liquid is blocked in during a fire. This isn’t so important for pipes but it is when there are vessels in the system.

You need to evaluate your system and make a determination.

 

[Bill3752]

Molecularman, am interested in where the 2500 - 5000 sq ft number is cited in the API. I just posted a question on that issue before reading this thread.

 

[djack77494]

I concur with ggordil. If you have a fire in an area containing LNG, I would think the reaction might be a lot of people running around shutting valves. These are not water or even oil lines where we might have little concerns for the prescribed situation. Instead we are talking of a facility that has probably the highest concentration and largest quantity of stored AND easily released energy in an industrial facility. Liquified gases really are a special situation, and methane is arguably the most dangerous owing to its high energy content and its very low boiling point. I would very carefully evaluate whether or not a nearby fire might lead the operators to block-in what would be normally open piping. It would take quite a lot of good reasoning for me to dismiss this as a case of double jeopardy. (Some companies and some locations insist that a safe condition be maintained EVEN in the event of a double jeopardy.) Hmmm.

 

[JoeWong88]

My 2-cents of opinions...

Double jeopardy
Don't think it is double jeopardy. As fire is occurred, all shutdown system is activated (as planned) and result inventory trapping in the system. All are in sequence and expected.

Piping to be included in fire load calculation
Piping within plant would normally be included as it is concentrated, piping inventory and associated risk could be big Discussed in Requirement of overpressure protection devices on system design to PIPING code and relief load can be managed in many occasion. Nevertheless, cross country pipeline which involved plant BL, it is always ended up a long discussion as some may argue that piping/pipeline code do not demand for pressure relief, BL definition, ...

Area to be considered
If fire scenario is credible, idea brought up by molecularman (Chemical) is one quick, normal & code "compliant" way to "limit" the area. If turn out to be excessive, qualitative risk assessment (QRA), fire modeling could be employed.

 

[npf]

All:

Thnaks for the comments. I am planning to take a quick shot at the fire case and look at the relief load.
SeanB: The pipe is insulated and the operating temp is around -260F and this relief valve is placed between 2 fail closed control valves.

Regards,
npf

 

[molecularman]

Sorry, I thought I had read about the 2500 ft2 fire circle in API but it turns out it is an internal standard.

 

[JoeWong88]

molecularman (Chemical),

Please refer to API Std 521, Jan 2007, Fifth edition, section 5.15.1.1 & 5.20.2.2. Read more...