Hydrogen dispersion 2

[imatasb]

Hi,

I am working in the area classification of a plant which involves some hydrogen pipes.

I am using the sources of release qualification method.

I have some formulas to calculate the horizontal extent of the zone. For example:

X = (920*Q/LEL)^0.55 (1)
Release from high level point source where X horizontal distance in meters, Q gas flowrate in m3/s and LEL lower explosive limit in vol %.

For a release from ground level there is a similar formula

X = (1849*Q/LEL)^0.55 (2)

Alternatively I am also using which calcultes the extent based on the degree of ventilation avalaible (ventilatin efficiency and number of air changes.

In the case of the a vent from a pipe placed outside both approaches reach a similar number.

My problem is that the the formulas 1&2 are not adequate for a emission which happens indoors and if I use the ventilation approach it results in number like a persistence time of 19 hours for a hypothetical leak of 0.16 g/h (leak through a 0.05 mm hole in a 1050 mbara line).

Sincerely I don't think that the hydrogen is going to take 19 hours for the hydrogen to dipserse.

It looks to me that these formulas are not applicable for the hydrogen case and its huge differencein density with the air.

I have been looking in some books I have from Dr. Frank Lees where he talks about dispersion models and I have found models for passive dispersion and for dense gas but nothing about light gases.

Does anyone know a formula applicable for calculation of distances for area classification in the hydrogen case indoors.

Thanks

 

[owg]

I will make a couple of qualitative comments, and look forward to comments from others. Hydrogen released from an elevated outside source like a vent pipe will either ignite due to friction and burn fairly quietly or disperse without ignition. It will not accumulate and explode unless released in massive quantities, like a rocket motor failing to burn properly. Hydrogen released indoors will also tend to ignite due to friction. If it doesn't it could accumulate and explode on finding a source of ignition, although I cannot quote any case histories of this happening. I don't recall any recommended air changes per hour for hydrogen compressor houses but there may be some guidance on this somewhere.

HAZOP at

http://www.curryhydrocarbons.ca

 

[25362]

The following link may be applicable:

http://conference.ing.unipi.it/ichs/Papers/320121.pdf

 

[mbeychok]

Imatasb:

Please tell us how you calculated or determined your leak rate of 0.16 g/h. If that was not done correctly, then all subsequent dispersion calculations would be incorrect.

Milton Beychok
(Visit me at www.air-dispersion.com)
.

 

[imatasb]

Hi mbeychok,

For calculating the release I used the next formula:

W = 0.8 * A *((2*densitiy* (p1-100,000))^0.5

where w mass flowrate in kg/s
A cross sectional area of the orifice in m2
P1 pressure upstream in N/m2 (abs)
density of the gas in kg/m3

Cheers

 

[mbeychok]

Imatasb:

The equation you used to calculate the leak flow is completely incorrect. It is the equation for the flow of a liquid through a leak and you are dealing with a gas leak. Incidently, you also used 100,000 N/m² as the atmospheric pressure in that equation ... and that is also incorrect. Atmospheric pressure (at sea level) is 101,325 N/m². Don't make the mistake of thinking that atmospheric pressure is 1 bar.

You gave the hydrogen gas pressure in the pipe as 1050 mbar. I assume that means millibar and therefore the hydrogen pressure in the pipe is 1.05 bar. Since atmospheric pressure is 1.01325 bar, the hydrogen pressure in the line is only slight above the atmospheric pressure. Therefore, the gas flow will be non-choked. You can find the correct equation to use for non-choked flow at:

http://www.air-dispersion/msource.html#Gas

Discharge




Milton Beychok
(Visit me at www.air-dispersion.com)
.

 

[ScottyUK]

owg,

The following might be of interest as an example of an indoor hydrogen explosion. Definitely interesting to us electrical types who have large battery systems. Link is 171k for those on dialup.

http://www.detrick.army.mil/bulletin/safety/work/batteryexplosion.pdf

----------------------------------

I don't suffer from insanity. I enjoy it...

 

[imatasb]

Hi,

sorry for the delay but I have been out a couple of weeks.

Dear Milton,

Thank you very much for your information.

I have been looking in your webpage I have seen all that information you are showing in it.

I know that the formula I was using is the same one used for a liquid release but in the engineering standars of the my company, it is also the recommended formulae if the gas is released at a pressufre below 2 bar.

Anyway I feel more comfortable in my calculatios with the right formula and not the one I was using which looks like a simplified version.

In any case there is not a big difference between both values:

Original release: 4.5 e-8 kg/s
New calculation: 4.0 e-8 kg/s

Note: I completely agree that 100,000 is not the the atmospheric pressure but I just wrote the formula as shown in the standard.

My question now that the release rate has been calculated, how can I calculate the distance from the origin at which the concetration falls below the lower explosive limit.

This is really the number I would like to calculate for zoning purposes. As I said on my first writing I haven't been able to find a way to calculate it for light gases as hydrogen.

Thanks

 

[paalr]

Where did you obtain equasions for the two different types of releases? I have a simular problem from the hydrogen generation from batteries.

 

[mbeychok]

imatasb:

I know it is quite long time since you asked your question about hydrogen dispersion. If it is not too late, you might try reaching the Gard Analytics company at

http://www.gard.com/index.htm

Milton Beychok
(Visit me at www.air-dispersion.com)
.