Rupture disk tail pipe routing?

[hollerg]

I have seen a tail pipe on a rupture disk directed within the same dike as the tank it protects. The tank contains class 1B flammables, it is the only tank in the dike and the disk will be set at 13 psig (15 psig tank). The only source filling the tank is from periodic transfers from a tank truck.

A fire that might result in the overpressure would be from a leak involving that tank or the pumping system associated with that tank. What makes this practice acceptable or not acceptable?

 

[EGT01]

API RP-521 has a detailed discussion about atmospheric venting, you may want to review that to see if it answers your question.

API RP-521 gives criteria to meet when venting flammable vapors such as ensuring dispersal so a flammable mixture isn't created at grade or in surrounding structures. Generally, it is not recommended to vent flammable liquids to atmosphere and there are other methods of disposal which are discussed for liquid relief events.

It is unclear from your description if the rupture disc tail pipe is directed upwards or downwards. I would have concerns if it were directed down even if there was not a possibility for liquid relief.

 

[CHD01]

hollerg: Boy, this is a good question. I'm going to answer you without checking API - so I might be full of hot air. But here goes, I think this is perfectly acceptable installation assuming that the dispersion of this liquid to atmosphere is not a concern and that splashing of the liquid is contained within the dike wall. The rupture disk is protecting against failure of the tank from overpressure (objective #1 complete), and the dike is containing the release (objective #2 complete). The dike is designed to contain the largest tank and since you have just one tank we meet that criteria. There is the issue of whether or not the fire case applies. In your case it may not since for a fire to occur it appears that the tank must first overpressure and then build up liquid in the dike and then have an ignition source to cause a fire. You will have to determine what the response time is to a liquid release from the rupture disk. This could be a double jeoprody (never could spell this damn word and I hate the TV show) case and fire would not apply; or you might as a company policy decide to protect against fire also even though it first requires a separate overpressure event. Just go through a good engineering analysis and document it with your other calculations.

Hope this helps.

The more you learn, the less you are certain of.

 

[CHD01]

hollerg: Boy, this is a good question. I'm going to answer you without checking API - so I might be full of hot air. But here goes, I think this is perfectly acceptable installation assuming that the dispersion of this liquid to atmosphere is not a concern and that splashing of the liquid is contained within the dike wall. The rupture disk is protecting against failure of the tank from overpressure, and the dike is containing the release. The dike is designed to contain the largest tank and since you have just one tank we meet that criteria. There is the issue of whether or not the fire case applies. In your case it may not since for a fire to occur it appears that the tank must first overpressure and then build up in the dike and then have an ignition source. You will have to determine what the response time is to a release from the rupture disk. This could be a double jeoprody (never could spell this damn word and I hate the TV show) case and fire would not apply; or you might as a company policy decide to protect against fire also. Just go through a good engineering analysis and document it with your other calculations.

Hope this helps.

The more you learn, the less you are certain of.

 

[hollerg]

Clarification -- The tail pipe is directed downwards.

Since the flammable could be from the pump seal, and fumes can find ignition sources, I presume a fire could heat the vessel (uninsulated) and burst the rupture disk.

Once popped, I would think the gas from the tail pipe would catch on fire and the fire might propogate into the vessel along with air, if the heat rate was not great enough to have a high velocity jet(or when a water hose begins to cool that tank. What happens next, deflagration?, explosion?

 

[EGT01]

Inadequate dispersal was the concern I had for venting flammable vapors to grade. But, Hollerg raises another good point about the rupture disk installation. Even if it were vented upwards to disperse vapor, once the rupture disk opens, you have an open path into the vessel. Would it be prudent to consider installation of a flame arrestor inline with the rupture disk?

 

[CHD01]

Valid points to consider; your point being that you can have a contained fire without first having an overpressrue event. Thus it may be better, in this case, to go to atmosphere with the vent, instead of down. In doing so you will have to make sure you are not discharging liquids via the vent that may need to be separated (maybe that is why they brought the vent down?). For the pump seal leak case you will also have the issue of how long it will take to heat the tank based on the pump seal leak rate that you select for analysis AND what is operations response time to correct the problem (considering interlocks and alarms available and remoteness of tank, etc).

On the flame arrestor, that depends on the flash point for the liquid, class of fluid, etc. It may be advisable.

The more you learn, the less you are certain of.