Multiple vessels in fire relief case ? 1

[joerd]

If two or more vessels are protected by a single relief valve, should we:
1. add the wetted area of all vessels and then apply the API RP521 equation Q=21000*F*(A^0.82), or
2. apply the equation to each vessel and then add the total heat input to calculate the relief load?

The second option gives a higher total heat input (and therefore relief load), because A1^0.82 + A2^0.82 is larger than (A1+A2)^0.82.
However, one can argue that the first method is right, since multiple vessels exposed to fire are "less likely to be completely exposed to the flame of an open fire" (quote from API RP521 re. the origin of the 0.82 exponent).

What do you think?

Cheers,
Joerd

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Guidoo]

The API formulas are empirical formulas, obtained by curve fitting data from a small number of experiments. Most data were from 1944 Rubber Reserve Company Tests on cylindrical vessels 5 ft diameter x 33 ft length, 85% full with water, atmospheric pressure, open 8 inch vent. The vessel was exposed to an engulfing type gasoline pool fire.
The exponent of 0.82 was obtained by fitting a straight line through data points plotted on double logaritmic paper. When looking at the data, even on double log paper they are clearly not on a straight line.
Since the API formula was derived from tests on single vessels, my choice would be to calculate the heat load for each vessel, and then add the results to get the total heat load. This is your second option. It may be conservative, which is good for relief valve sizing...
It would be good when next API revision would incorporate more sophisticated tools, that can be used for other types of fires (e.g. jet fires, confined pool fires).

 

[CHD01]

Apply the calculation for each vessel; then add the loads.

I would do this because the vessels may be different heights, and if greater than 25-30ft would not be included in relief rate calculations. Also, if there are different liquid levels and volume hold-up, this will affect time required before a release occurs. You could even have one tank relieve and the other remain secured.



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

 

[EGT01]

Joerd,

I have to agree with the others, my preference is to determine the heat load for each vessel separately.

What I find in API RP-521, 3.15.2.1 Heat Absorption Equations, is that "large vessels are less likely than small ones to be completely exposed to the flame of an open fire." The way I interpret this is to say small vessels are more likely to be completely exposed to the flames. Just the past couple of days, I saw a news report of a fire at the Nova Chemical Plant located in Pasadena, Texas. It looked as if multiple vessels could have easily been completely engulfed in flames.

Now in some cases where you may want to account for piping associated with a vessel, I think it would be reasonable to add the piping area to the vessel area. See API RP-521, 3.15.1.1 Effect of Fire on the Wetted Surface of a Vessel, where they suggest "It may be appropriate to add a percentage of the vessel area to account for vapor generation in piping associated with the vessel under consideration."

 

[JoeNava]

Joerd,

I would also calcuate each veesel, probably different liquids are vaporizing, and different latent heats shoudl be used.

 

[joerd]

Thanks all for sharing your thoughts, there seems to be good consensus on this matter. EGT01, where did you read the article on Nova Chemicals?

Cheers,
Joerd

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[EGT01]

The Nova Chemical report I referred to was actually a video clip that I caught on the evening news. I did happen to find a link to the Houston Chronicle web site that has the report and a video. Unfortunately, our network administration and configuration won't allow me to view the video so I don't know if it presents the same video view I saw early. Give it a try but try soon because I don't know how long that link will be good.

http://www.chron.com/cs/CDA/story.hts/metropolitan/1950420

 

[EGT01]

The above Houston Chronicle link referred to KHOU television in Houston. They also have a slideshow on the fire.

http://www.khou.com/slideshows_archive/#