7 replies to this topic

[Tadic zmaj]

Hello,

I am a mechanical engineer from Serbia. I have a few question about cooling adjacent tanks.

Our standard (SRBS) says that every oil tank larger then 300 m³ must have a fixed foam system for fire protection and EVERY oil tank must have fixed water spray system for cooling. It also says that, when there is a fire on one tank, you must cool the shell and roof of adjacent tanks. Based on that standard, I am to calculate the amount of water that I need, water speed, size of pipes, valves etc. Then, the inspectors look at my project and, if it is ok, give me permission to build this system.

But now I have a problem. We have 15 tanks in one dike (size of dike 90 m x 70 m x 2m) which is partitioned by lower walls. They do not know which tank is adjacent tank so they told me to cool every tank that is in range of 2*D (D is diameter of the tank on fire). So, in a worst case, I need 450 m³/h of water (for extinguishing fire and cooling). The part about a 2*D range is not in our standards, so they made it up. They said that if I can find anything about this in API, NFPA, or DIN, I can put that in project and they will not demand a 2*D ruling.

So I read API and NFPA (I do not know German, so I can not read DIN) and it says nothing about this. API 2021 even says:
"Cooling water should be used with discretion. Fire fighters are trained to use cooling water on exposures; confronted with a tank fire, they may respond by cooling all adjoining tanks. This is usually not needed and may adversely affect attempts to extinguish the fire. Depending on the size of the fire, tank spacing and wind, cooling of adjacent tanks is typically unnecessary unless there is direct flame contact or sufficient radiant heat to scorch the paint. The “hose stream heat test” can be used to determine the need for cooling, as discussed in 8.5.3. When this test shows that tanks do require cooling, water streams should be fanned on the sides and roofs for best results."

Do you have any ideas about this? Thanks in advance for your time.

Edited by Art Montemayor, 23 August 2012 - 01:43 PM.

[breizh]

http://www.epa.gov/o...vkinpresent.pdf

Check whether this link can support your query.
Hope this helps

Breizh

[Tadic zmaj]

Dear Mr. Breizh,
Thank you for quick answer. I read NFPA 30 2000 Edition (which is very similar to our standard) earlier. We did everything that the standard says about distances from other objects, distances of adjacent tanks etc. However, I did not find anything in NFPA 30 that tells me which tank to cool with water if adjacent tank is on fire (some zone in which we need to cool the tank like 2*D or 1/3 of the sum of those 2 tanks or amount of heat per m2 over which it need to be cooled etc.). Any ideas?

[kkala]

Sorry for late response, I have just found (by chance) a point related to 2*D distance in relevant Greek legislation, Decree 34458-1990, Gov gazette 846-B-31/12/90, Technical specifications on layout, design, construction, safe operation and fire fighting for refineries and rest petroleum industries (official title in English may differ somehow). A rough summary of this point is attached in the "2D.doc".
Local refinery legislation for firefighting is influenced by NFPA, yet I am not specifically accustomed to the latter. Strange that NFPA does not report something similar. I could suppose influence from EU regulations but there is no such hint in the Decree.


Note: For tanks within a single dike, look at http://www.cheresour...diked-enclosure '> http://www.cheresour...diked-enclosure . 15 tanks in a single dike seem excessive.

Attached Files

•  2D.doc   23KB   65 downloads

Edited by kkala, 29 December 2012 - 03:34 PM.

[S.AHMAD]

My current company's practice is to consider downwind quadrant that is within 2D radius of the burning tank where D is the burning tank diameter. If any part of an adjacent tank is located within this quadrant, it is considered to require cooling.

I am not aware of any standards w.r.t. this topic. My suggestion is for you to propose to management based on others local practice. Otherwise you may stuck forever.

Edited by S.AHMAD, 01 January 2013 - 08:52 PM.

[kkala]

Practice mentioned by S.AHMAD complies with local legislation, considering distance 2D from center of tank in fire (post No 4). But the latter (ie legislation) does not clarify status of dike area .
Local legislation is understood to concern fire on one tank, not having undergone loss of containment (*). Whenever the latter occurs, there will be pool fire over all dike area, so all tanks in the dike need water cooling.
Probably tanks of adjacent dike, downwind of mentioned pool fire, need also cooling (case judged rear, tanks in other dike can be far enough).
Any recommended practice concerning pool fire extended to dike area would be appreciated.

(*) or probably tanks of different dikes

Edited by kkala, 02 January 2013 - 06:17 AM.

[S.AHMAD]

1. Good points by Kkala.
2. I went through local Design Practices, I could not find anything related to cooling requirement of tanks for pool fires. However, after giving a deep thoughts on the pool fire scenario, I believe the tanks engulfed in a pool fires are not to be considered as "adjacent tanks", they should be treated as the "burning tanks".
3. Further more, water shall not be used for fighting pool fires since the water may leads to "boil-over" and exaggerates the pool fires. Only foam shall be used.

Edited by S.AHMAD, 02 January 2013 - 07:48 PM.

[kkala]

Thanks S.AHMAD for the care and useful response; no info on pool fire has been detected in local legislation, either (post No 4). Following from an operating manual (elaborated by us, concerns a fire water pump station) is noted.
"Fire water is the most important extinguishing agent in a refinery. It feeds hydrants, monitors, sprays and any other firefighting equipment , with an intent to cool metal surfaces, control fire intensity, prevent ignition or formation of flammable vapor clouds, participate in foam making. Nevertheless water shall not be used for materials reacting with it, electrical fires, LPG fires (it promotes vaporization). Besides it should be used with caution on liquid hydrocarbon fires (foam is generally preferable), since difference in density promotes fire expansion".
Indeed, foam shall be spread over pool fires developed in dikes. However foam is assumed inefficient for cooling of the tank walls. Eventual collapse of another tank would throw more "oil into fire".
In such a case water cooling of all tanks in the dike should be realized despite the risks (if possible by using installed sprays, to limit water flow), simultaneously with foam application on dike area. This is advice from a colleague dealing with fire fighting design (*). Fire is ulikely to expand out of the dike. Water will fall down in contact with metal wall of the tanks, generally being far from dike wall.
I also think damage from any boil-over will be less in a shallow hydrocarbon pool (as in dikes).

(*) Colleague has read it somewhere, source not recalled

Edited by kkala, 03 January 2013 - 05:10 AM.