Carbon monoxide going off - possible causes?

[PaulRainbow]

Had similar problems this week. Marina called to say boat was sounding CO alarm. Put new batteries in but yesterday that device and the main CO/LPG alarm I'd wired in was beeping.

The only thing powered up on shore power is the battery charger. Switched that off and will do some battery testing.

Not wishing to give any egg sucking lessons, but remember, if a battery is gassing enough to set the Co alarm off, the slightest spark will cause an explosion.

Check the batteries for temperature, if one is hot it's the culprit. If they have been sitting with no charging, carefully separate them and check the voltages. If one is around 10v, that's the culprit.

 

[PaulRainbow]

In the past, on these forums, i have been told that i'm basically making things up when i say that gassing batteries can explode.

I present exhibit A :

 

[Bouba]

In my motor home, the dogs used to fart the alarms on every night

 

[wonkywinch]

To be honest, I'm surprised half the posts on this forum don't set my gas alarms off!

 

[ylop]

In the past, on these forums, i have been told that i'm basically making things up when i say that gassing batteries can explode.

I present exhibit A :


View attachment 180897

Certainly possible to get explosive concentrations of hydrogen from charging a dud battery but this assertion

remember, if a battery is gassing enough to set the Co alarm off, the slightest spark will cause an explosion.

Is considerably more certain than reality.

Most domestic CO alarms trigger for H2 concentrations far below the LEL - roughly about 1/20th of the concentration. Of course you probably don’t have the CO alarm mounted in the exact space where the batteries are, and so if you have enough to trigger a CO alarm in the galley your battery space might be explosive and if you can’t find a CO source on the boat and mute the alarm you might well get to explosive levels in the whole saloon. If you have a gas alarm too you’d expect that to trigger too.

H2 LEL 4%
Sensitivity of common CO meters to H2 about 1:5-1:10 of the CO response.
Alarm threshold for domestic CO alarms varies, but most will alarm if exposed to 30ppm CO for a reasonable period (or higher signals for less time); therefore they will be triggered by 150-300 ppm H2 - ie orders of magnitude below the explosion limit.

You definitely shouldn’t ignore it, but I’d challenge the claim it WILL explode with a spark.

 

[PaulRainbow]

Certainly possible to get explosive concentrations of hydrogen from charging a dud battery but this assertion

Is considerably more certain than reality.

Most domestic CO alarms trigger for H2 concentrations far below the LEL - roughly about 1/20th of the concentration. Of course you probably don’t have the CO alarm mounted in the exact space where the batteries are, and so if you have enough to trigger a CO alarm in the galley your battery space might be explosive and if you can’t find a CO source on the boat and mute the alarm you might well get to explosive levels in the whole saloon. If you have a gas alarm too you’d expect that to trigger too.

H2 LEL 4%
Sensitivity of common CO meters to H2 about 1:5-1:10 of the CO response.
Alarm threshold for domestic CO alarms varies, but most will alarm if exposed to 30ppm CO for a reasonable period (or higher signals for less time); therefore they will be triggered by 150-300 ppm H2 - ie orders of magnitude below the explosion limit.

You definitely shouldn’t ignore it, but I’d challenge the claim it WILL explode with a spark.

Co levels are totally irrelevant, it's not Co that's triggering the alarm, its hydrogen.

That's usually caused by a short circuited cell in a battery, which means you've basically got a 10v battery being charged at 14+V.

Considering that charging a battery with an old style charger, then removing the leads with the charger still turned on often causes an explosion, messing around in a battery compartment on a boat, with a battery that's being so badly overcharged isn't something any same or sensible person should be doing.

 

[PaulRainbow]

No matter what you post on these forums, someone comes along with a smart reply. I despair.

 

[penberth3]

......Most domestic CO alarms trigger for H2 concentrations far below the LEL - roughly about 1/20th of the concentration. Of course you probably don’t have the CO alarm mounted in the exact space where the batteries are, and so if you have enough to trigger a CO alarm in the galley your battery space might be explosive and if you can’t find a CO source on the boat and mute the alarm you might well get to explosive levels in the whole saloon. If you have a gas alarm too you’d expect that to trigger too.

H2 LEL 4%
Sensitivity of common CO meters to H2 about 1:5-1:10 of the CO response.
Alarm threshold for domestic CO alarms varies, but most will alarm if exposed to 30ppm CO for a reasonable period (or higher signals for less time); therefore they will be triggered by 150-300 ppm H2 - ie orders of magnitude below the explosion limit.

You definitely shouldn’t ignore it, but I’d challenge the claim it WILL explode with a spark.

Basically correct, until you get into semantics and contradict yourself.

Alarms sound at less than the LEL, yes. There WILL be higher concentrations of gas nearer the source, yes. So you can't assume there isn't an explosive atmosphere somewhere, or that there is no source of ignition.

 

[ylop]

Co levels are totally irrelevant, it's not Co that's triggering the alarm, its hydrogen.

That's usually caused by a short circuited cell in a battery, which means you've basically got a 10v battery being charged at 14+V.

Considering that charging a battery with an old style charger, then removing the leads with the charger still turned on often causes an explosion, messing around in a battery compartment on a boat, with a battery that's being so badly overcharged isn't something any same or sensible person should be doing.

CO is totally relevant. Off the shelf CO sensors are actually reactive to a wide variety of gasses. They are set to trigger at a particular threshold - that value is usually something like the current the sensor produced for 30ppm of CO in contact with the sensor got 60 minutes (I may be wrong on the exact numbers as I haven’t looked up the data sheet). Now another gas, H2, CH4, C3H8 etc can trigger the same surface reaction in the sensor and thus the alarm incorrectly “think” it is detecting CO. The sensitivity level to different gasses is different - depending on the particular sensor component about 5-10x less sensitive for H2 than CO. It will therefore trigger the CO alarm at roughly a couple of hundred ppm H2 - well below the explosion limit for hydrogen.

No matter what you post on these forums, someone comes along with a smart reply. I despair.

No matter how adamant you are that you are right - sometimes you are wrong. If your batteries trigger your CO alarm you do not necessarily have an explosive gas mixture (certainly not in the area where the CO sensor is installed). You said you absolutely did - that’s wrong; and I’m not saying that from a purely theoretical basis either - I had a battery fail last year whilst i was cooking with gas, it set off the CO detector but didn’t cause an explosion. We obviously stopped cooking and ventilated the space but when it kept happening and then started triggering the hydrocarbon sensor even though all gas and similar sources were eliminated that we worked out the cause.

If you meant to say “if you have triggered a CO alarm through a battery charging problem, you should make sure you ventilate the area particularly the area around the battery before disconnecting the terminals as you could have explosive gas mixtures” that would have been helpful, relevant and accurate. What you wrote was misleading.

Basically correct, until you get into semantics and contradict yourself.

Alarms sound at less than the LEL, yes. There WILL be higher concentrations of gas nearer the source, yes. So you can't assume there isn't an explosive atmosphere somewhere, or that there is no source of ignition.

But that’s precisely the point I was making - rainbowpaul is wrong to say there is always an explosive mix if your alarm goes off. Certainly if you have a badly ventilated battery space there could be.

 

[PaulRainbow]

CO is totally relevant. Off the shelf CO sensors are actually reactive to a wide variety of gasses. They are set to trigger at a particular threshold - that value is usually something like the current the sensor produced for 30ppm of CO in contact with the sensor got 60 minutes (I may be wrong on the exact numbers as I haven’t looked up the data sheet). Now another gas, H2, CH4, C3H8 etc can trigger the same surface reaction in the sensor and thus the alarm incorrectly “think” it is detecting CO. The sensitivity level to different gasses is different - depending on the particular sensor component about 5-10x less sensitive for H2 than CO. It will therefore trigger the CO alarm at roughly a couple of hundred ppm H2 - well below the explosion limit for hydrogen.

No matter how adamant you are that you are right - sometimes you are wrong. If your batteries trigger your CO alarm you do not necessarily have an explosive gas mixture (certainly not in the area where the CO sensor is installed). You said you absolutely did - that’s wrong; and I’m not saying that from a purely theoretical basis either - I had a battery fail last year whilst i was cooking with gas, it set off the CO detector but didn’t cause an explosion. We obviously stopped cooking and ventilated the space but when it kept happening and then started triggering the hydrocarbon sensor even though all gas and similar sources were eliminated that we worked out the cause.

If you meant to say “if you have triggered a CO alarm through a battery charging problem, you should make sure you ventilate the area particularly the area around the battery before disconnecting the terminals as you could have explosive gas mixtures” that would have been helpful, relevant and accurate.

Not going to waste my time responding to most of the nonsense above. But, if there is enough hydrogen to set off the CO alarm, there will be a greater concentration at the batteries. That should be pretty obvious.

Saying "I had a battery fail last year whilst i was cooking with gas, it set off the CO detector but didn’t cause an explosion" is ridiculous. A gassing battery isn't going to produce enough hydrogen to fill the whole boat to a level where it's likely to explode.

What you wrote was misleading.

What i wrote might save someone from being injured by an exploding battery.

But that’s precisely the point I was making - rainbowpaul is wrong to say there is always an explosive mix if your alarm goes off. Certainly if you have a badly ventilated battery space there could be.

There will almost certainly be an explosive mixture at the battery, no matter how well ventilated the battery space is.

A good battery, being charged on a bench in a workshop, with an old style charger can explode just by disconnecting the charger. A battery with a shorted cell will be boiling vigorously and getting very hot, the one in the picture above was in a battery box beneath a berth and no-one was anywhere near it.

There is always a presence of an explosive gas, or the Co alarm would not go off, but, of course, not every battery, in every circumstance will definitely explode every time. I suspect anyone reading this thread understands this, but not everyone felt the need to nit-pick.

 

[ylop]

Saying "I had a battery fail last year whilst i was cooking with gas, it set off the CO detector but didn’t cause an explosion" is ridiculous. A gassing battery isn't going to produce enough hydrogen to fill the whole boat to a level where it's likely to explode.

Any yet as usual you are too argumentative and pig headed to go back and read your original post and recognise that this is exactly what you were implying. If the reader is assumed to have your “expertise” there is no need for the post - so presumably it was intended for the wider audience - your post said “, if a battery is gassing enough to set the Co alarm off, the slightest spark will cause an explosion.” Your statement was misleading - you don’t need to behave the way you do whenever anyone challenges something you write.

There is always a presence of an explosive gas, or the Co alarm would not go off,

Its only an explosive gas if it is present it ~4-75% mixture with air. The point is - a CO alarm responds to levels at <0.1% of h2. Knowing that a CO alarm responds to other gasses is useful - believing it means your boat is now a bomb is not.

I suspect anyone reading this thread understands this,

I suspect they don’t.

I also suspect most people getting a sudden and unexplained CO alarm going off from a battery charging are not using a portable charger with crocodile clips. Your safety advice may have been valid but you didn’t actually tell anyone how to avoid a tiny spark if they were using a more typical installation. Your advice would imply switching off the mains to the charger might be a risk (and it might depending on location of switch relative to battery).

 

[PaulRainbow]

Any yet as usual you are too argumentative and pig headed to go back and read your original post and recognise that this is exactly what you were implying. If the reader is assumed to have your “expertise” there is no need for the post - so presumably it was intended for the wider audience - your post said “, if a battery is gassing enough to set the Co alarm off, the slightest spark will cause an explosion.” Your statement was misleading - you don’t need to behave the way you do whenever anyone challenges something you write.

I did not imply that a gassing battery would cause an explosion anywhere on the boat, it happens at the battery.

Its only an explosive gas if it is present it ~4-75% mixture with air. The point is - a CO alarm responds to levels at <0.1% of h2. Knowing that a CO alarm responds to other gasses is useful - believing it means your boat is now a bomb is not.

Once again, i never said the boat was a bomb, or that it would explode if there was a spark. It's the battery that explodes if there is a spark near it.

I suspect they don’t.


I also suspect most people getting a sudden and unexplained CO alarm going off from a battery charging are not using a portable charger with crocodile clips.

They don't need to be, if they remove a battery clamp whilst the charger is on it can cause a spark.

I don't think anyone reading this is stupid enough to think a gassing battery will cause their boat to explode, rather than just the battery.

Your safety advice may have been valid but you didn’t actually tell anyone how to avoid a tiny spark if they were using a more typical installation. Your advice would imply switching off the mains to the charger might be a risk (and it might depending on location of switch relative to battery).

I didn't give any advice.

 

[wonkywinch]

Had a chance to investigate yesterday. The boat has been isolated from shore power, the Cristec charger turned off and left for several days with only a tiny load connected (gas alarm, auto bilge pump).

I disconnected all the batteries and checked them individually. All showed >12.5v after their few days "rest" and that voltage stayed good even when loaded with a 21 watt automotive lightbulb for a few minutes. Physically all were in good condition and acid was visible in the "magic eye" if I rocked the battery.

Engine battery - 1x Exide ER350 flooded cell 80 A/hr
Domestic batteries - 3x Exide ER350 flooded cell 80 A/hr

The engine and domestic batteries are mounted together but wired separately. When I bought the boat last year, the dealer said "the batteries have recently been replaced". The engine/domestic batteries certainly look new.

Bow thruster batteries - 2x Exide EP450 dual AGM 50 A/hr

These are under the forward bunk in separate battery boxes.

I then connected the charger (which showed a charge rate of 38 amps, slowly decreasing, on the NASA BM). I placed a CO alarm on top of the batteries whilst they were charging and checked them carefully for any sound/heat. Nothing abnormal.

I didn't want to leave the charger on so depowered everything and came here to scratch my head about what to try next.

 

[Snowgoose-1]

Exide ER350 flooded cell 80 A/hr
Domestic batteries - 3x Exide ER350 flooded cell 80 A/hr

The engine and domestic batteries are mounted together but wired separately. When I

Bow thruster batteries - 2x Exide EP450 dual AGM 50 A/hr

These are under the forward bunk in separate battery boxes.

I then connected the charger (which showed a charge rate of 38 amps, slowly decreasing, on the NASA

So the last few days the carbon monoxide alarm has been going off on my boat at night (Sealine S34) - I haven't been at the boat, only noticed via the security camera, and it had stopped by morning but then started again that night.

I was there all day yesterday, with heating on etc. from about 9am to 5pm and not a peep out of it (or any effects from co poisoning in myself), but noticed there was a reed diffuser right next to it, so thought it might be that, or maybe the unit was faulty. I fitted a second detector in the bedroom, can't be too safe, locked it all up and left, but then it went off again about 9pm that night (just the old one), so assumed it must just be a faulty unit, but then checking at 8am this morning the new one is also going off and reporting carbon monoxide issues.

The engines haven't been run for 2.5 weeks, the gas heating was on yesterday for a few hours (checked exhaust was working fine) but was switched off several hours before leaving, and prior to that hadn't been switched on for 4-5 days before the alarm started sounding.

The gas system is switched off, and hasn't been used since 3rd march.

The only things "on" on the boat is the battery charger, the sockets to run a small oil filled radiator, and the two fridges because we left food there.

The one that went off first is in the galley on the port side, the one that went off second is in the front cabin on the sb side.

Does anyone know what might be causing it? I'd read that dying batteries can emit carbon monoxide, but I spent 3-4 hours in the engine room yesterday doing jobs and there was no sign of "fizzing" or excess heat (or any ill effects for me, but I obviously had the hatch open). I can't think how an oil filled radiator could cause it. Could the galley fridge somehow be emitting either CO or some kind of gas that's fooling the detectors? Although I'm not sure why that would only manifest in the evening.

I'm pretty stumped, but obviously can't risk staying at the boat until we get to the bottom of it, which we're meant to be doing next weekend, so any thoughts would be greatly appreciated.

The alarms are a wonderful invention that must have saved many lives

I have 4T outboard in a well . The alarm comes on sometimes even though you can't see any smoke. It is located down below in the saloon

 

[westernman]

Has the alarm (not the battery) time expired?
Most are good for 5 years some for 10 years.

 

[wonkywinch]

Has the alarm (not the battery) time expired?
Most are good for 5 years some for 10 years.

No, the alarms are < 18 months old with new batteries. First alarm that started going off was the battery wall mounted one, next time it went off together with the hard wired installed CO/LPG alarm (Nereus system). The CO led was illuminated so certain it's triggered by gas that is building up whilst boat unattended.

Switching the battery charger off stops the alarms but as my long post above says, batteries seem OK at first inspection.

 

[LiftyK]

No, the alarms are < 18 months old with new batteries. First alarm that started going off was the battery wall mounted one, next time it went off together with the hard wired installed CO/LPG alarm (Nereus system). The CO led was illuminated so certain it's triggered by gas that is building up whilst boat unattended.

Switching the battery charger off stops the alarms but as my long post above says, batteries seem OK at first inspection.

Tricky. May I ask what the charging current drops to after a while, say a day or so? In my case, the alarm went off and it was indeed a faulty battery. The evidence that pointed to that was a much higher charging current than it should have been after 24 hours. I think it was in the region of 5-8 amps when it should have been close to zero.

 

[wonkywinch]

Back on the boat now. Put the CO detector on top of the batteries (bottom of cabin wardrobe) and turned the charger on. It went into alarm within 5 mins of the battery charger being on. Nothing feels hot, open circuit volts across each of the 4 batteries seem normal.

I did notice the set of 3 house batteries seem to drain in a shorter time than previously (according to the NASA BM2 which might normally say 85% left next morning at anchor, now reports say 65%).

Still head scratching.

 

[PaulRainbow]

Back on the boat now. Put the CO detector on top of the batteries (bottom of cabin wardrobe) and turned the charger on. It went into alarm within 5 mins of the battery charger being on. Nothing feels hot, open circuit volts across each of the 4 batteries seem normal.

I did notice the set of 3 house batteries seem to drain in a shorter time than previously (according to the NASA BM2 which might normally say 85% left next morning at anchor, now reports say 65%).

Still head scratching.

What's the charging voltages set to ?

Post #57 might be relevant, regarding current going into the batteries after a while.

Might also be worth testing them one at a time.

 

[B27]

Maybe connect a charger to one battery at a time and see whether the CO alarm goes off?

I would expect much better than 12.5 Volts a couple of days after turning off the shore power.

A 21W load for a few minutes is not a serious test for a 50Ah or whatever battery.
Having to rock the battery to see anything on the magic eye suggests electrolyte level is low.

38A charge current suggests one or more batteries is doing a lot of self discharge in a few days.

The real question might be whether the batteries have simply died of old age, or whether the charger or solar or whatever is overcharging them?