Quality Control of Gas Cylinders

[ QUOTE ]
I suspect that there may be a contaminant gas in the bottles, probably nitrogen, which will preferentially fill the gas volume in the bottle. In some cases bulk supplied LNG/LPG will have other gases in it, which will "gas" out of the bulk liquid. If you get a supply of this into your bottle, the nitrogen in the top of a new bottle will mess up the stoichiometry of your flame, essentailly reducing the flame speed and producing the behaviour you describe. Venting the gas of the top of the bottle will draw off most of the nitrogen build up and restore the composition to normal.


[/ QUOTE ]

That sounds very feasible although I'm not sure what you mean by flame speed? What happens is that it lights but the flame is blown too far away from the burner so that the flame doesn't touch the falme failure sensor and thus it goes out once the overide is released.

Is the shape of the Gaz cylinder a factor perhaps? It is a much flatter top than a Calor one so less room for a gas space?
 
[ QUOTE ]
Robin,
On a charter boat we probably get through more cylinders per year than most and never experienced your problem.

Can I suggest that when it happens next you take the cylinder to anothr boat and see if that cooker lights.

If it does then process of elimination between your regulator or cooker.

You are a sensitive soul and I think your cooker/regulator is reflecting this!!


[/ QUOTE ]

Hi Roger

The problem is now resolved for us by our changing to Calor cylinders and a 2 bottle changeover setup (which included a new regulator) but I raised it for information to others. Two Calor cylinders will just about last us for our usual summer 5 week cruise if we start out with both full. We will still carry 2 Gaz full spares which we can connect in if needs be via the Gaz/Calor adapters but this will be rarely needed. Interestingly the cost of the switch to the Gaslow changeover are recouped in just one season of using the gas in it's cheaper Calor format.

I guess that your charter boat cylinders are Solent sourced rather than like ours through PYC so this may well be a local problem or just more prevalent here? Also, if you are using Gaz on the charter boat, think how much you could save by using Calor at less than half the price!

I know I can give some very frosty stares (I'm advised by she who notices) but enough to blow out the gas? Watch out Mr Geller I could be bending spoons next!

Robin
 
Flame speed is the speed at which a flame will propagate into a gas stream. So, for example, if your gas flows out of the cooker at 30 cm/s, and the flame speed is 40 cm/s, the flame will burn back into the gas stream and "attach" itself to the burner. If the flame speed drops, by being too lean, or having too much nitrogen in it, to say 20 cm/s the gas stream will blow the flame off the cooker and it will never attach.

The complete flame dynamics are more complex due to the velocity profile of the jet and the slowing of the flow once some standoff distance has been achieved. It may be possible get a flame to stabilise at a small distnace off the burner if the flame speed is close to but just below the gas exit velocity. The actual flame surface area (cone like) balances the fuel consumption with the fuel flow rate, so bulk jet velocities at the burner exit in reality are likely to be higher than the flame speed. If this was not the case, there would be some risk of the flame flashing back into the fuel supply line. Not desirable, it tends to lead to loud pops and bangs and the fuel pipes melting. /forums/images/graemlins/smile.gif

For some absolute numbers, the flame speed of methane is about 40 cm/s. I could dig out propane etc if you really need to know!
 
[ QUOTE ]
Flame speed is the speed at which a flame will propagate into a gas stream. So, for example, if your gas flows out of the cooker at 30 cm/s, and the flame speed is 40 cm/s, the flame will burn back into the gas stream and "attach" itself to the burner. If the flame speed drops, by being too lean, or having too much nitrogen in it, to say 20 cm/s the gas stream will blow the flame off the cooker and it will never attach.


[/ QUOTE ]
Thanks for that information! That is exactly what it seems to do.
 
Sorry .... did I read ... shaking the cylinders ?

I would advise against such action - as that would splash liquid gas onto valves etc. and possibly inhibit correct operation oif regulator especialy if any liquid was carried by gas when asked to supply to appliance ...

I also am concerned that two different gases may be involved here ... my experience before was that Calor based on high % butane ... and GAZ based on Butane / Propane mix ... the two are not same and used to require different regulators .... I assume that this is not the case now and that you are using correct regulators for the gases and that GAZ may be same as Calor now then ?

In all the years I caravanned ... boated in UK with Calor .... I have holiday home on south coast with propane ... never ever had trouble with using full cylinders small and large.

??
 
[ QUOTE ]
I suspect that there may be a contaminant gas in the bottles, probably nitrogen

[/ QUOTE ] I have a sneaking suspicion that you are on the right track. I wonder if any purging or pressure testing of bottles is carried out as part of the refilling process. Any nitrogen left in the bottle would be entirely in the gaseous phase as it cannot be liquefied above its critical temperature and it would have to be vented from the bottle before a stream of butane gas that could be ignited was produced.

It would be interesting to follow the refilling process through a depot.

Robin have you observed whether you get a stream of gas that won't light or whether you get no flow of gas at all when this problem strikes. If the former then it supports a theory that an inflammable gas is present,. If the later then it supports the theory that liquid is somehow causing the regulator to interrupt the flow .
 
[ QUOTE ]
[ QUOTE ]
I suspect that there may be a contaminant gas in the bottles, probably nitrogen

[/ QUOTE ] I have a sneaking suspicion that you are on the right track. I wonder if any purging or pressure testing of bottles is carried out as part of the refilling process. Any nitrogen left in the bottle would be entirely in the gaseous phase as it cannot be liquefied above its critical temperature and it would have to be vented from the bottle before a stream of butane gas that could be ignited was produced.

It would be interesting to follow the refilling process through a depot.

Robin have you observed whether you get a stream of gas that won't light or whether you get no flow of gas at all when this problem strikes. If the former then it supports a theory that an inflammable gas is present,. If the later then it supports the theory that liquid is somehow causing the regulator to interrupt the flow .

[/ QUOTE ]

I cannot comment as to whether UK refiller use inert gas to purge cylinders before refilling ... but out here they do not ... they fill direct from main storage without purging first.

Second ... even if they did purge with inert gas first - on refilling initial volume into cylinder would be vapour at pressure that would vent out the volume already there - ie the inert gas ... as the filling line and cylinder cool and liquid volume takes over - the cylinder would fill with flammable volume .... bring any residual inert gas to near zero content .... IMHO.

The LPG content on passing to burner would far exceed any inert gas anyway and mixed with air as it should be - the LPG / air mix should burn ....

It is possible but I think extremely unlikely.
 
I wasn't considering that the bottles may be purged, but rather that the nitrogen that exists naturally in LPG was coming out of the liquid LPG preferentially and putting an inert layer in the top of the bottle. The behaviour of the flames, and the solution of venting the gas off the top of the bottle both support this. It would only take 15% by volume in the gas in the top of the bottle, to significantly alter the flame speed. In a full bottle, this would be a very small quanity of nitrogen. As long as the nitrogen comes into the bottle in solution with the LPG, there is little reason why this should not occur. Nitorgen levels of tenths of percent in the LPG would be more than enough.
 
I won't argue ... but my lab disagrees !!

Will check a few of our analysis - as we are the only LPG testers in FSU Baltic .. and we look after domestic and import / export quality of Butane / Butane + Propane / Propane only etc. gases .....
Nitrogen doesn't ring any bells with me on this ....
 
I was just having a hunt through some of our reports and a search on line, but turned up nothing. However, I was sure that I had seen something both on gas certs in Oz and in data here that suggested trace levels of N2. I'll dig some more tomorrow if I am not too busy... /forums/images/graemlins/smile.gif
 
Nigel you obviously have some knowledge of the lpg industry. How are the cylinders pressure tested? They must surely be tested at intervals during their lives.
I can't believe it is with the butane itself, water seems unlikely too and I would not have thought that air would have been used as that would lead to some air being left in the bottle unless it was removed by vacuum.
 
We don't get involved in pressure testing cylinders - so cannot answer for definite ... but would suggest that water test is most likely.

Reason for water test is that water cannot be compressed and is a safe medium if the vessel bursts when pressure is exerted. It's way Fire Extinguishers etc. are done. The water is placed into vessel and small volume allowed for air pressure to be applied to raise pressure to required level.

I don't like to say this - but I have reason to assume that many cylinders may be on market based on visual inspection of condition ... but don't take me as gospel on this ... I would like to think that all vessels are tested.

But as I say - we are not involved in the cannister ... we have our own ASTM double ended sampling cannisters ... we are involved in the actual product - the LPG. Not only quality of product as supplied but also blending LPG's for various markets ....... basically do not assume that butane is 100% B ... nor Propane is 100% P ..... often they have %'s of each other dependent on market reqt's.
 
I to have suffered the same problems as you, bottles purchased at cobb,s obviously same supplier calor centre poole and did exactly the same as you vented the bottle on the pontoon would not like to do at sea.
 
Just been chatting to our Oil and Chem chap, who is ex tankers, and heavily involved with the LNG/LPG trade. Saudi spec for C4 gases is less than 10 ppm N2 content. He mentioned that N2 is rarely tested for unless there is a suspected problem. As it is often used to inert bulk tanks some adsorption into the LPG is expected. At the 10 ppm level, you could expect about half a litre (at STP or 0.03 grams) in a 4.5 kg bottle. This would be enough to muck up the stoichiometry of the flame if it has migrated to the small volume in the top of a full bottle. As the fuel air ratio for butane is about 1:30, a half litre of N2 contamination of the butane would lead to the first 15 litre through the stove being u/s assuming that all the N2 came out first. If it was only making up 20% of the fuel composition then it would be the first 70 odd litres.

Typically N2 content is less than 1 ppm, so wouldn't be an issue, but if you have a source which is on the limit, then maybe.
 
To be honest - forgot bto check with lab ... as we are into winter spec diesel blending at moment ... will have a word later ...

I know that we have never seen N2 as a product separating .. or noted it ... we do the full spectrum testing and it does not ring any bells ............. it is there in small volume - has to be - it does constitute one of the major gases around us !!
 
[ QUOTE ]
Will check a few of our analysis - as we are the only LPG testers in FSU Baltic .. and we look after domestic and import / export quality of Butane / Butane + Propane / Propane only etc. gases

[/ QUOTE ] You are presumably testing bulk supplies not small individual Calor or Gaz cylinders.

This seems to be an occasional problem and it appears to affect certain depots.

If the problem is caused by nitrogen, or any other non flammable gas for that matter, that is expelled by venting a little gas to waste then the only way any laboratory testing is going to find the answer is by testing the gas initially produced from a good number of individual cylinders. Not something you are going to be getting your lab to do I guess.

In the days when I was doing scientific investigation I think the first thing I would have done would have been to have visited a filling depot and become acquainted with all the operations carried out.
 
Reason I mentioned I would look at analysis is the claim of N2 being present in the blend ...

As to individual cylinders -you are right - we have no commercial interest in them or the filling. We look after bulk storage / import / export.

The interest on my part was to check the N2 content - and damn !! I was just talking to my lab about "bugs" and forgot to ask about LPG !!

FYI - a component test analysis for LPG could run to about 1000 Euros for one full spectrum test ......... so per cylinder ???? OOPs - don't think customer would accept that !!
 
You must log in or register to reply here.

Other threads that may be of interest

Chiara’s slave
Aesthetics. The eye of the beholder?
Replies
19
Views
523
Minerva
M
geem
Atlantic crossing 2025
2
Replies
39
Views
2K
geem
geem

Share this page

Top