Self Ignition of Hydrogen 6

[petepsingpy]

I see in the literature for handling hydrogen gas that there's often a warning that hydrogen can "self ignite" when allowed to spray through a nozzle or leaking flange.

I'm trying to understand the mechanism of this self ignition. I did see some previous discussion on this topic awhile back. Someone pointed out that hydrogen does heat up as it expands through a nozzle, but wouldn't it need to get hot enough to reach its auto ignition temp of 500C? It seems unlikely that that could happen.

Hydrogen apparently has a very low minimum ignition energy, maybe < 0.05 mJ. So is it a static spark phenomenon?
Thanks,
Pete

 

[kbander]

H2 does heat up, but I believe you need a pretty extreme pressure drop to go from say 0 C to over 500 C (approx 9500 atmospheres using Peng Robinson).

As you stated, the minimum ignition energy for hydrogen is very low (less than 10% of natural gas), so it wouldn't take much of a spark to ignite it.


Regards,

Bob

 

[25362]

One problem is that hydrogen leaks are invisible. Sources that can provide the small minimum ignition energy are many and varied: electric, such as wiring of motors, friction of bearings, hot surfaces (boilers, lamps), burner flames, cutting and welding, etc. In a large plant, potential ignition sources may run in the hundreds.

 

[kenvlach]

Hydrogen self-ignition has nothing to do with expansion --gases cool when expanding into a lower pressure.

The danger is that a sharp or jagged metal edge will catalyze the ignition of the hydrogen.

The first application of catalysis was nearly 200 years ago as a lighter. Hydrogen was stored in a capped, teapot-like container having a bit of platinum (gauze or powder) at the spout. When a gentleman wanted to light his cigarette, he squeezed a lever-type handle which let the hydrogen escape over the Pt where it ignited.

I believe I read about this in a Johnson-Matthey Platinum Metals Review some 25 years ago. Some results of an on-line search on the ddiscovery of catalysis:
&quot;1817 H. Davy studies the oxidation of methane on platinum wires
E. Davy studies the oxidation of methane on platinum dust

1823 J. W. Döbereiner discovers that metals glow in contact with air and combustible gas.

1824 W. Henry studies oxidations catalysed by platinum adsorbed on clay pellets.

1825-1833 Michael Faraday studies the ignition of hydrogen in air at platinum surfaces.&quot;

from

http://www.nottingham.ac.uk/~eczehl/catalysis/history.htm

 

[kbander]

kenvlach,

Unfortunately it is not true when you make the statement that gases cool with expansion. At high pressure hydrogen is on the other side of the Amagat line (negative JT coefficient)...and it heats up when it expands.

Since hydrogen is sometimes kept at extreme pressures, this is sometimes brought up as a concern.

By the way...As you increase pressure at a constant temperature, most real gases will reach a point where the JT coefficient will go from +ve to 0 then to -ve. It's just that we rarely see other gases at such high pressures ...so the phenomenon isn't noticed.



Regards,

Bob

 

[kenvlach]

kbander,
Thanks much for the info, I learned something interesting (or perhaps forgotten since phsical chemistry & thermodynamics courses many years ago):
&quot;The temperature below which an expansion produces cooling is called the inversion temperature; it is high for most gases, but for hydrogen, it is about 193 K. Hydrogen, therefore, must be cooled below this temperature before it is expanded.&quot;

http://www.hq.nasa.gov/office/pao/History/SP-4404/app-a1.htm

Your earlier post indicates that compressed hydrogen at 2500 psi & 25[sup]o[/sup]C leaking into atmospheric air does not create heat enough from this effect to self-ignite. The auto-ignition temp. is 565.5°C (1050°F), from MSDS. However, cylinders come with a warning to not ‘crack’ the valve open as self-ignition may result. So, I still believe the danger, absent any external ignition source, is still catalytic ignition. The light-off is a result of the heat release due to catalytic hydrogen oxidation on the metal surface (e.g., while exiting cylinder valve), especially if the metal edge is fine enough that heating can occur. Certainly, the Joule-Thomson heating of the hydrogen would contribute.
I have seen similar warnings re self-ignition of acetylene at sharp metal orifices.

A description and photo of the early, room temperature catalytic lighter I referred to earlier can be found at

http://barbarina.ingentaselect.com/...ocpdf/rpsv/cw/matthey/00321400/v43n3/s13/p122

Ken

 

[25362]

I should have added that hydrogen flames are practically invisible because of the lack of soot and short, because of hydrogen's large buoyancy and diffusibility, having a length of less than about 500 leaking hole diameters. However, because of their high temperature, sometimes a &quot;mirage&quot; effect can be sensed because of air stratification. A small leak from a gasket in a pipeline may ignite by static electricity and be near invisible to workers, who can get burned by the &quot;torch fire&quot;.

 

[quark]

Is it anything to do with presence of oxygen? In an atmosphere of 7.5% oxygen, hydrogen can autoignite (or I heard so) and we used to purge Nitrogen in Hydrogenation reaction.

 

[kenvlach]

Yes, this thread concerns hydrogen leaking into air, and whether high pressure can cause self-ignition.

kbander calculated an extremely high initial hydrogen pressure would be necessary to reach the autoignition temperature (for homogeneous combustion), 25362 cautioned against sparks, heat and other ignition sources, and I mentioned that catalysis on a metal surface can ignite hydrogen at low temperatures (ambient temperature with Pt catalysis).

For Hydrogen in air at 1 atm total pressure:
FLASH POINT: Not applicable.*
AUTOIGNITION TEMPERATURE: 571°C (1060°F)
FLAMMABLE LIMITS (in air by volume, %):
Lower (LEL): 4.0%
Upper (UEL): 75.0%
Source:

http://www.airgas.com/documents/pdf/1026.pdf

*Note: Hydrogen has a only a theoretical flash point of 20[sup]o[/sup]K since there are no oxidizing materials, i.e., oxygen, in the gas state at this temperature, which is necessary to support combustion.

 

[25362]

to quark, if you are speaking of reactors, the hydrogenation catalysts therein may be able to reduce the activation energy of hydrogen ignition with oxygen to lower levels than normally needed for self-ignition; kenvlach's comments rounded up items on this subject.

 

[petepsingpy]

I don't know anything about build-up and discharge of electrostatic charge, so this may be a dumb question; but can a flowing gas generate it's own static charge that might later dissipate providing an ignition source?

 

[saxon]

petep, After a quick search of the literature, please check this site, http:albert.swl.rwth-aachen.de/anlagen/TH1//th1-re~1.htm. &quot;Typical research results on self-ignition and deflagration to detonation transition&quot;.

From this study, it becomes apparent that when hydrogen flows across small orifices extremely high velocities can be achieved (in the Mach range)this sets up shock waves within the gas stream that provide the necessary energy levels at various cell points that can cause deflagrations, detonations and the phenomena known as &quot;self ignition&quot;.

Hence the warning. Hope this helps.

saxon

 

[kenvlach]

saxon,
the study at the posted link is very interesting, but may not be relevant:

In the 1st part of the study, they studied self-ignition of combustible/explosive mixtures: “in H2-air-steam/CO2 mixtures under the following conditions: initial temperature (900-1350 K), initial pressure (0.3-1.7 MPa) and steam/CO2 concentration (0-40 %).” These are metastable mixtures above the auto-ignition temperature, where only pressure ‘nudge’ initiated combustion.

In the 2nd part of the study, H[sub]2[/sub]/air mixtures at an initial temperature of 293 K were accelerated down a tube having a constricted diameter area. The behavior (ignition, detonation or both) depended upon the Mach number in the pressure wave of the gas in the tube past the restricting ring. It seems to me to be a type of SCRAM-jet engine; not the situation if a high pressure hydrogen leak was dispersing into open, ambient air.

petepsingpy,
Re static electricity: In the USA, OSHA requires H[sub]2[/sub] systems to be electrically grounded:

http://www.osha.gov/pls/oshaweb/owa...able=STANDARDS&p_id=9749&p_text_version=FALSE

[note: paste entire URL into browser]
I know that flowing Freon inside PVDF & PTFE creates static electrical damage to the plastic, so metal piping is used. Don’t know whether hydrogen could create static on metal, but possibly it creates friction at the orifice/leak which could add to the heat of expansion. Air (containing oxygen) would be drawn in by Venturi effect, causing catalyzed ignition at the edge of the orifice.

Please tell us more about your hydrogen system: P, T, purity and what kind of nozzle & what gas and vessel it normally discharges into.

 

[saxon]

kenvlach, Although we all may not be in agreement as to what are the exact reaction dynamics/kinetics that can result in deflagation/detonation or self ignition of Hydrogen in every circumstance, I do believe that we can all agree that due to the wide explolsivity/flammability range of Hydrogen that it's prudent to handle it with extreme caution in every situation. Remember, Class/Division areas were set up for reason.

saxon

 

[kenvlach]

Agreed, take every precaution and use common sense.

Many safety regulations are the result of accidents, and I'm sure there are more to come.

Thanks to all for informative, interesting posts.

petepsingpy, Good luck (take care)!
Ken

 

[I2P]

Dear All who have responded to this thread,
Having experience in facing hydrogen related fires I would like to add my inputs:
1.0 H2 as we know has a negative JT effect and heats up on expansion,but again the auto ignition temperature being high,this temperature is not easily attainable by the pressure let down in case of leaks from high pressure to low pressure.

2.0 But the ignition energy is very low of the order of .05m joules.What happens is that when high pressure H2 leaks,there is very high degree of turbulence and this can generate sufficient static charge to ignite the H2 gas.As we know the explosive range of H2 is quite high.(4 to 74%)

3.0 In one of our vacuum gas oil plants we had a hydrogen fire due to flange leak upstream of the reactor.the pressure was 75 bar.The investigatiion done was very comprehensive and it calculated the auto ignition temperatures at this pressure letdown,and was found that the H2 did definitely not ignite on it's own heat.Subsequent simulations were done using softwares and the static cgarge generated considering the volume escaping,temperature rise etc,and the conclusion was that the very low ignition energy required was made available during the gas release by virtue of static charge which resulted in the fire.The location of the flange was such that there was no hot flange,or motors or any equipment which could have imparted the heat source for ignition.

4.0 Also purer the hydrogen more invisible the flame and during daytime is never visible.at night it will be bluis in colour.People have been known to have walked in hydrogen fire cloud resulting in fatalities also.Potential fire equipments falnges are provided with snuffinf steam rings since there is no other way of combating hydrogen fires(unless source of leak is isolated)
Hope this information is useful to those in the hydrogen handling industry.

 

[petepsingpy]

It was my first post and I'm very happy I wrote it...I ended up with a lot more information and responses then I expected.

Thanks for your replies.

Pete