7 replies to this topic

[mm111]

Hi friends,
During exterior fire, all liquids inside the liquid fill vessel is vaporized and cause over pressure and results in relief. For single component, the due as well as bubble point would be same. So the relief temperature would be the bubble point at relief pressure (21% over set pressure), but for multi component the due temperature and bubble point temperature would be different. What temperature would be at relief pressure?
Thanks for your view.

[gvdlans]

For multicomponent mixture, the relief temperature is not constant but will increase over time, as light components will vaporize first resulting in a heavier liquid phase. Note that because of this effect also the molecular weight and latent heat of vaporization will change over time.

Relief temperature will be the bubble point temperature at relief pressure and the liquid composition. Since liquid composition will change over time, you have to find liquid composition by flashing out 2%, 4%, 6% etc of light-components vapor. Note that the maximum required orifice area for the relief valve will not be at start of relief, but will go through a maximum.

If fire would continue, in the end (most of) the liquid will be vaporized, and relief temperature will reach the flame temperature. Of course, the vessel will have probably failed before that, which explains why having a relief valve only can not be considered sufficient protection against vessel failure in case of an external fire. Additional measures (e.g. cooling vessel surface with water, depressurizing systems, fireproofing) will be required as described in API 521, section 3.15.4.

My recommendation is to get a copy of article in Chemicle Engineering by Wing Y. Wong, "Fires, Vessels and the Pressure Relief Valve", May 2000, pages 84-92.

[Art Montemayor]

mm111:

I've read your post with concern and was relieved to see gvdlans address the important issues right away. His advice should be heeded and followed.

I've witnessed the results of a pool fire within a processing area and what it has caused and the reason for my concern is that many engineers state what you've stated regarding an exterior fire depleting all the liquid within a vessel, and not blink an eye. Let me state this:

1) Once you have removed any heat sink (the liquid in the vessel) and offer no barrier (insulation) or sacrificial sink (water sprinklers) from a pool fire, your equipment is DOOMED. Forget about the relief scenario, the relief valve, or any orifice sizing. The vessel will fail (and fail catastrophically) should the fire continue. This is a situation that you, as an engineer, should not allow to develop. If it does, you've not done your job and you & all others are no longer needed; the site should be evacuated and nature should be allowed to resolve the issue. This has to be the saddest part of any equipment fire, because it signals the complete defeat of any engineering attempt.

2) For those (& who hasn't?) who have seen the live video of the twin towers demolishing themselves on 09/11/01, the action is typical of what fire does to structural steel when it is subjected to open flames without any heat sink or insulation to help out. That same effect will occur to your vessel !!

So, what I would comment to you is that contemplating the temperature of the relief valve AFTER the heat sink is removed is a moot and academic point that has no engineering value.

As gvdlans says, a depressurization system is more in order to debate and discuss. However, the depressurization system does not prevent the 09/11 effect from happening. The vessel will fail, except it will not explode, but perhaps keel over in a fused state. Of what value is the relief temperature under such circumstances? Perhaps you have additional information or there is another part to this story that we haven't been told.
Please let us know, you may have a valid and practical reason for trying to find out the vapor temperature.

Art Montemayor

[mm111]

Thanks gvdlans and art.
Since a year I am doing the relief valve analysis and calculations. I came across to many scenarios. But for fire case I always got question in my mind. I would like to be clearer and to go more in depth. I am more concern about the relief temperature is because the DIERS calculations is required the flash temperature for evaluation of relief device. For exterior fire case, for multi-component contents the first bubble will generate at bubble point temperature. But whole liquid would not vaporize at this temperature. By fire, the temperature keeps increasing and reaches to dew point temperature, when all the content would be at vapor phase. And as per API, exterior fire can cause relief when all the liquid vaporize and cause vapor or 2 phase relief. So I think the relief temperature would be the due point temperature. And at the due point temperature we will get the maximum required orifice area as well as maximum flow instead of bubble point temperature. As I totally agree with Art, that fire is a case when any thing can happen. So it doesn’t mean anything to think about relief valve when tank is on verge of burst. But my point of view is towards the calculations part rather than over all practical view. Here in my question I assumed that fire will not burst the tank but just cause the relief valve open.

Again gvdlans, I will get the copy of Wing Wong’s article about the fire case in a day or two. Thanks for good suggestion. Thanks Art about the practical point of view of Fire scenario. I do appreciate your any comments.

[gvdlans]

mm111,

I just want to add the advice that you shouldn't try to re-invent the wheel... Many people have already looked at this issue, see for example the article of Wing Y. Wong, but there is more (e.g. many discussions on eng-tips forum, research reports by UK HSE executive (www.hse.gov.uk), report by Norwegian company Scanpower (www.scandpower.org), articles in Chemical Engineering Progress).

Other thing is that you cannot simply say that maximum flow and maximum required orifice area is at liquid dew point. As I already wrote, during relief not only relief temperature, but also Molecular Weight and Latent Heat of Vaporization change. You will have to look at what happens by doing stepwise calculations/simulations as I described before.

[avsp]

mm111,

Please get hold of this excellent article in the Chemical Engineering Progress by Per Salater et al, Norsk Hydro, "Size depressurisation and relief devices for pressurised segments exposed to fire", September 2002, pages 38-45. This article has been extracted from Norsk Hydro's Best Practice on Depressurisation and Fire Relief Design.

Also, this article forms a basis for the international guideline "Guidelines for the Design and protection of Pressurised system to withstand severe fires", Institute of Petroleum, UK as well as the corresponding Norwegian Guideline for protection against fire.

Cheers,

[gvdlans]

The article in CEP as mentioned by avsp can be downloaded via following link:

http://www.cepmagazi.../pdf/090238.pdf

This article can be seen as a summary of the report from Scandpower, that is basis of (future) Norwegian standards on the subject.

[mm111]

Thanks gvdlans
I am sorry i couldn't reply you on time because i was ill since 2 weeks. Thanks for quick reply.
I got the articals you refered and i m looking on it.
Thanks to all for review and reply.

MM111