External fire and failure of plant compressed air supply - is this a fair assumption? 2

[chemks2012]

Hello all,

I am writing a pressure relief case and making an assumption that plant compressed air supply will fail [stopped] in case of external fire.

Do you think this is the fair assumption? I believe, in case of fire, to stop the air supply would be one of the emergency actions. For example, failure of air supply will close all the raw material supply valves [if they are air fail close] and open cooling water supply valves [if they are air fail open valves]?

Please share your experience.
Thanks
KS

 

[zdas04]

Compressed air systems are sometimes shut down by the ESD. Other times they are left in service to allow the plant operators to attempt to restore control. It is just a decision in the analysis that shouldn't be pre-supposed. It is pretty easy to check to see if the ESD button will kill the air system or not. If not then your scenarios need to consider it as operable in any scenario except "loss of control air". If the ESD system kills power to the air compressors AND shuts a block valve, then you can you can assume the "fail on loss of control air" position of the control valves.

David Simpson, PE
MuleShoe Engineering

Law is the common force organized to act as an obstacle of injustice Frédéric Bastiat

 

[chemks2012]

zdas04,

Thanks,agree with you. All make sense.

 

[moltenmetal]

When there is a fire, you want the valves to go to their failure positions before their actuators get too hot to work- or melt off! That is why you DO NOT want to connect your air lines to your actuated valves with something like stainless steel tubing- you need something in the line which will melt and fail first during fire to allow the valve to travel to its failure position. We use durable polyurethane tubing for the drops, which not only makes the installation quicker and maintenance easier, it also provides this fire safety function.

It's also important to remember that a diaphragm-actuated control valve may lose its ability to remain in the fail position during prolonged fire. This is one of many reasons that control valves are often backed up with firesafe rotary on/off actuated valves in failure critical applications, i.e. isolating a source of fuel. The rotary valve's aluminum actuator may melt off eventually, but he valve will long ago have been taken to the failure position- and unlike a rising stem valve actuated with a diaphragm, the fluid provides no driving force to change the valve's position even once the actuator is removed.

 

[zdas04]

That is a very coherent and cogent description of one school of thought. A perfectly valid foundation for a design philosophy.

Another plant may take the position that it is not credible for the entire plant to ignite at once and retaining control gas gives them a better potential to minimize the extent of damage through control actions. This would also be a valid foundation for a design philosophy.

I guess my point is that long before you get to the point of specifying plant equipment you need to have thought about these things and made some foundational decisions. You are not going to get a one-size-fits-all answer from a free web site.

David Simpson, PE
MuleShoe Engineering

Law is the common force organized to act as an obstacle of injustice Frédéric Bastiat

 

[chemks2012]

@moltenmrtal and zdas04,

Thanks very much for your input, much appreciated.

 

[vladrath]

If you follow the API rules for sizing PSVs you take loss of air supply into account separate from the external fire case. If you were to account for "double trouble" in every case you would be mired in exceptions and possible what if scenarios.

So I would write it up stating that you take failure of instrument air in one scenario and the external fire case in a separate scenario.

 

[TD2K]

I just reviewed some PSV files where the group doing them basically used the approach where all the outlet valves went to the fail position (closed in this case) and the inlet stream(s) continued to flow. When I asked what the scenario was, they said they usually assumed that a small fire could take out the air lines supplying the air to the outlet valves in which case they would go closed.

I could see where they were coming from. Most of the time we don't design the instrument air lines other than the main supply header and maybe some sub-lines. The final lines, 1.5" or smaller are run by the field contractor and these valves were all in a similar area that a single line could be the common supply point. If you don't like the idea of a fire, someone closes a root valve or a forklift takes out the air line, the result is the same. In an existing facility knowing how the lines are actually run, this might not be the case you could decide this case was no longer credible.

Another factor is how large of a fire circle are you using? A 2500 ft2 circle is about 56' diameter. 5000 ft2 is 80 ft2 which just makes more things possible to be caught up in the fire.