8 replies to this topic

Dear Sir,
I am sizing PSV for fire case of compressor KO drum. The KO drum lenght is 10 m (Dia 4 m) and its TL elevation from ground is 3 M. As there is no liquid in the drum, I am sizing the PSV for fire case considering unwetted area. But I am not sure how much exposed surface area i should considered. Should i consider
1. Total KO drum surface area
2. Suface area upto 25 ft (7.6 m) as per API (which we considerd during fire case of wetted area).

If share your experience and idea for the same.

Regards,

Bharat Deshmukh

[joerd]

Most companies use your option 2, to consider only the area up to 25 ft.

[Art Montemayor]

Bharat Deshmukh:

This is a relatively LARGE compressor suction drum (4 meters diam. x 10 meters T-T) that is elevated at a height of 3 meters. And if there is a pool-type fire generated below it, it will subject the gas-filled vessel to surpassing the maximum allowable temperatures and reaching such low allowable material stress levels that the vessel will fail - by rupture/fusion - if subjected to a direct flame long enough.

This scenario is stressed in API 521, a document you should have studied before attempting to resolve this type of engineering safety relief problem. If you haven't been given a copy of this document, then request it. If you are refused a copy, then get one from a technical library, buy one, or borrow one. It is THAT important and serious to know what you are dealing with. This type of scenario does not involve high science or mathematics. It merely involves some practical horse-sense applied to the situation. I have discussed this application many times before on these forums and it is a simple engineering recognition that a gas phase has probably the WORST thermal conductivity properties in the world. That's why gases serve as INSULATORS in industrial practice. This, in the practical world, means that you really should not rely on the gas inside a pressure vessel to transfer heat effectively and quickly raise its own pressure due to elevated outside temperatures. You could easily, if given enough time, reach the fusion (or rupture) point of the vessel's shell material - and under pressure! - without reaching the set relief pressure of your PSV. This is a very dangerous hazard that must be evaluated and circumvented.

There are practical, and logical ways to design a protection system against this type of worse scenario. You can:

• Install vessel insulation; (in practice, not very effective)
• Install water sprinklers that bathe the vessel with a deluge of water in case of a fire;
• Install a blow-down system on the entire compressor system. You have to be able to shutdown the compressor immediately in case of a fire and you should have a blowdown of the trapped vapor in order to shutdown the compressor without any load. Study the shutdown requirements for the compressor and you will find that you should be shutting down the compressor without any load on it - which means that you should have blown down any trapped gas. This then, means that there is no pressure in the vessel(s) during the fire and no overpressure scenario. This doesn't mean you don't need a PSV; you still require one on the vessel because the ASME (or similar) code requires it for a pressure vessel.

The 3rd option above is, in my opinion, the most effective in the case of a gas-filled pressure vessel.

I urge you to obtain a copy of API 521 (& both parts of API 520) and study them carefully and with interest.
I hope this helps you out.

[fallah]

Dear Sir,
I am sizing PSV for fire case of compressor KO drum. The KO drum lenght is 10 m (Dia 4 m) and its TL elevation from ground is 3 M. As there is no liquid in the drum, I am sizing the PSV for fire case considering unwetted area. But I am not sure how much exposed surface area i should considered. Should i consider
1. Total KO drum surface area
2. Suface area upto 25 ft (7.6 m) as per API (which we considerd during fire case of wetted area).

If share your experience and idea for the same.

Regards,

Bharat Deshmukh

In my opinion the point you raised above is one of vague points in API 521 that should be clarified by whom are responsible for that.
In equation (8) of API 521 (latest ed.) you will see A' is "the exposed surface area of the vessel",but contrary to partially liquid filled vessel there is no criteria for taking an estimation for exposed surface area based on elevation from fire base level.

Anyway,in the cases i have came across total surface area (your first option) have been selected and also i personally follow this procedure.

[ashetty]

Dear Sir,
I am sizing PSV for fire case of compressor KO drum. The KO drum lenght is 10 m (Dia 4 m) and its TL elevation from ground is 3 M. As there is no liquid in the drum, I am sizing the PSV for fire case considering unwetted area. But I am not sure how much exposed surface area i should considered. Should i consider
1. Total KO drum surface area
2. Suface area upto 25 ft (7.6 m) as per API (which we considerd during fire case of wetted area).

If share your experience and idea for the same.

Regards,

Bharat Deshmukh

I agree with Mr.Montnemayor,

Failure of the shell from overheating would happen even if a PSV was provided.You might want to use devices such as rupture disc or fusible plug, to depressure the vessel upon actuation and reduce the risk of vessel failure due to overheating.You might also try isulation etc.

Anyhow if you want the equations for relief load of "dry vessels" please refer API 521 (Jan 2007) Sec 5.15.2.2.2

Regards.

[djack77494]

I have two comments regarding this thread.

Failure of the shell from overheating would happen even if a PSV was provided.You might want to use devices such as rupture disc or fusible plug, to depressure the vessel upon actuation and reduce the risk of vessel failure due to overheating.

Though you (ashetty) make some valid points in the post, you got this point wrong. The concept that was intended to be conveyed was that a pressure relief device would not offer effective protection against vessel rupture in a fire situation. It has nothing to do with the type of pressure relief device employed. So, a PSV would not be effective, but neither would a rupture disc, rupture pin, etc. The fusible plug could offer additional protection, but these devices are not that common and the codes do not address their use for fire mitigation purposes.

My second point is that standards like API 520 and 521 provide guidance on how to set up scenarios and outline the mechanics of doing relief calculations. Still much is left to the discretion of the design engineer. In the case of a K.O. Drum, I would urge you to imagine scenarios wherein the vessel was at least partially filled with liquid. Do not assume it won't happen just because a simulation says there is no liquid. If there is ANY possibility of there being liquid in the vessel, then the liquid-containing scenario should be evaluated. There's a pretty good chance that it will govern the PSV sizing.

[ashetty]

I have two comments regarding this thread.

Failure of the shell from overheating would happen even if a PSV was provided.You might want to use devices such as rupture disc or fusible plug, to depressure the vessel upon actuation and reduce the risk of vessel failure due to overheating.

Though you (ashetty) make some valid points in the post, you got this point wrong. The concept that was intended to be conveyed was that a pressure relief device would not offer effective protection against vessel rupture in a fire situation. It has nothing to do with the type of pressure relief device employed. So, a PSV would not be effective, but neither would a rupture disc, rupture pin, etc. The fusible plug could offer additional protection, but these devices are not that common and the codes do not address their use for fire mitigation purposes.

Agreed. Altough i have to say,i have seen Fusible plugs been used on more than one occasion. Even if we could not prevent failure of the shell, we would like to depressurize and prevent an exposion etc.

My second point is that standards like API 520 and 521 provide guidance on how to set up scenarios and outline the mechanics of doing relief calculations. Still much is left to the discretion of the design engineer. In the case of a K.O. Drum, I would urge you to imagine scenarios wherein the vessel was at least partially filled with liquid. Do not assume it won't happen just because a simulation says there is no liquid. If there is ANY possibility of there being liquid in the vessel, then the liquid-containing scenario should be evaluated. There's a pretty good chance that it will govern the PSV sizing.

Agreed.

Dear Sir,
I am sizing PSV for fire case of compressor KO drum. The KO drum lenght is 10 m (Dia 4 m) and its TL elevation from ground is 3 M. As there is no liquid in the drum, I am sizing the PSV for fire case considering unwetted area. But I am not sure how much exposed surface area i should considered. Should i consider
1. Total KO drum surface area
2. Suface area upto 25 ft (7.6 m) as per API (which we considerd during fire case of wetted area).

If share your experience and idea for the same.

Regards,

Bharat Deshmukh

In my opinion the point you raised above is one of vague points in API 521 that should be clarified by whom are responsible for that.
In equation (8) of API 521 (latest ed.) you will see A' is "the exposed surface area of the vessel",but contrary to partially liquid filled vessel there is no criteria for taking an estimation for exposed surface area based on elevation from fire base level.

Anyway,in the cases i have came across total surface area (your first option) have been selected and also i personally follow this procedure.

Dear Sir,
I am sizing PSV for fire case of compressor KO drum. The KO drum lenght is 10 m (Dia 4 m) and its TL elevation from ground is 3 M. As there is no liquid in the drum, I am sizing the PSV for fire case considering unwetted area. But I am not sure how much exposed surface area i should considered. Should i consider
1. Total KO drum surface area
2. Suface area upto 25 ft (7.6 m) as per API (which we considerd during fire case of wetted area).

If share your experience and idea for the same.

Regards,

Bharat Deshmukh

In my opinion the point you raised above is one of vague points in API 521 that should be clarified by whom are responsible for that.
In equation (8) of API 521 (latest ed.) you will see A' is "the exposed surface area of the vessel",but contrary to partially liquid filled vessel there is no criteria for taking an estimation for exposed surface area based on elevation from fire base level.

Anyway,in the cases i have came across total surface area (your first option) have been selected and also i personally follow this procedure.