Explosion Pressure from a gas-air mixture in a closed vessel 1

[GasSensor]

I am designing a hazardous gas sensor housing that must be explosion-proof and follow CSA standard C22.2 No. 30. One test that is used in testing of the standard is a gas-air mixture is flowed into the housing through the stainless steel flame arrester. Once inside, the gas is ignited. I would like to know what pressure that would produce? I have found a value of 21 bar (305 psi). The gas mixture is 17% acetylene, 22% oxygen, 61% air. Once I have a pressure value, I am modeling the housing as a thick-walled cylinder with closed-ends.

Thanks

 

[cthame]

There are a number of factors that will affect the pressure seen inside of the enclosure. The fuel air mixture, the volume of the gas sensor housing, the ambient temperature for the test, the types of joints the enclosure uses, location of the ignition source, (18 in away or inside of the enclosure), and its geometry. Without knowing these other variables its hard to fix any pressure value, but this can be said: the enriched oxygen mixture will definitely give greater pressures than if you were going straight air and fuel. Acetylene is a very volitile test gas, and you most likely will be getting your greatest pressures with this gas than with hydrogen, ethelyne, or propane. An explosion is a non quasi equilibrium process, and its very hard to find simple mathmatical empirical equations to spit out an expected pressure value.

 

[owg]

Based on experimental data from Senecal and Beaulieu (1997) and from relations given by Zabetakis, the maximum pressure rise for most deflagrations is typically: P2/P1=8 for hydrocarbon - air mixtures and 16 for hydrocarbon - oxygen mixtures. Some risk analysts use 10 and 20 to be conservative. This puts your 305 psi at the high end of the range of possibilities. 20 fold happens to coiincide with the value from Lees (1986) for the pressure increase for a detonation, initially at atmospheric pressure.

HAZOP at

http://www.curryhydrocarbons.ca

 

[GasSensor]

Thanks cthame and owg.

I have gotten CANMET in Ottawa Canada to do the necessary tests. For the gas mixture is 17% acetylene, 22% oxygen, 61% air, the maximum experimental pressure was 130 psi. I used 305 psi is the design. I also design the sensor housing to UL 1203 and that standard said the min wall thickness is 1/8" which produces a very large safety factor with the type of aluminum alloy I used. The sensor housing has now pass all the necessary tests.

Thanks again for your guys help,
GasSensor

 

[owg]

Thanks for posting the results of the Canmet tests. Was that a paid service? How quickly were the results available after the request was made?

HAZOP at

http://www.curryhydrocarbons.ca

 

[crysta1c1ear]

Surely the heat released is completely independent of the rate of combustion and depends just on the mixture: the gas mixture is 17% acetylene, 22% oxygen, 61% air. From that you could calculate the heat released, and use gas laws to determine the pressure.

Just like an automotive engineer saying petrol is 43MJ/kg and knowing the heat released by multiplying mass of fuel by 43MJ, you could find the heat released by burning the acetylene and work out the pressure when the heat is applied to the gas in its container.

Correct?

 

[j2bprometheus]

Hi-
I did a quick constant volume, constant internal energy calculation using EQUIL (part of Chemkin) and it predicted P_final = 13.2 atm

Here are the initial conditions that I used:

P = 1 atm
T = 300 K
17 % C2H2
34.8 % O2
rest is N2



j2bprometheus

 

[owg]

crysta1clear - I just noticed your point about the heat released being independent of the rate of combustion. I believe that you are correct. However I think that the faster burning rate during detonation causes localized short term pressure spikes while not breaking any of the laws of thermodynamics. I had not thought about it until now and would welcome further discussion.

HAZOP at

http://www.curryhydrocarbons.ca

 

[25362]

I may be wrong, but it appears to me there isn't sufficient oxygen for complete combustion, when considering that the oxygen volume should be 2.5 times that of acetylene from a stoichiometric viewpoint.

I agree with owg in that in a matter of a fraction of a millisecond pressures may peak and then stabilize at lower values. Although the resulting pressures may not be excessively high, the rate of pressure rise in a confined detonation may be some hundreds of bar/s, and the maximum rate may not necessarily appear at the maximum pressure. The rate also appears to be a function of the vessel volume.

 

[1952mike]

This thread is almost like the one I posted the other day on "backfire" issue. I was using wrong term for it. It should have been "Exhaust Explosion".

"Exhaust Explosion" is not a controlled event like this one.
The question I have is what parameters are involved?
In Q=m*c*(Delta T), Q is heat of explosion of m lbs of un burnt methane & air mixture that didn't ignite in ignition chamber (mis fire?) and c at almost constant volume is approx. that of air & CH4, and then, one may find T2 and therefore P2. Am I right? Or am I making a lot of wrong assumptions[/color red] considering thermodynamics & fluid dynamics?

Thanks

 

[25362]

The following link may be of help:

http://www.hse.gov.uk/foi/internalops/hid/din/539.pdf

 

[sailoday28]

Assumimg a homogenous mixture the
Maximum and minimum pressures from the instantaneous ignition of the mixture is obtainable from solution of
"Chapman-Jouguet detonation and deflagration waves"
Regards

 

[GasSensor]

owg-Yes it was a paid service at CANMET. They are cheaper than CSA but the last time we used them, they did not do all the necessary tests for standards CSA C22.2 No.30 and UL 1203. So, we are paying CSA to do the rest of the testing. Note: CSA should honour the testing done at CANMET but talk to CSA first; we had a problem.

GasSensor