Avoiding damage cause by fumes from batteries

[Dave]

The sealed lead acid batteries are not totally sealed and do vent some
mist of sulphuric acid.

The manufacturers say not to charge in a confined space, but the biggest
problem as I can see is that this is quite corrosive to electronic
components, with several reports of it eating pcb tracks, attacking a
crystal in a very high quality oscillator so much it could not be put
onto frequency etc.

Has anyone found an answer to this problem? Sometimes you want battery
backup in a case with electronic components, but it seems they don't
exist too well with each other.

I've been thinking whether there is any checmical solution, such that
you place a sacrificial chemical that will be more reachive with the
gases, leaving less to attack your components. Of course, any byproducts
would have to be non-corrostive.

I've also thought about sealing the batteries in a plastic bag. The bag
and expand/contract as and when the batteies vent cases, but they are
kept inside the bag. I'm not sure if the batteries would like this
though. Any you still have the problem of sealing around the leads.

 

[Mebart]

The problem is not just for lead acid type batteries. ALL batteries
emit corrosives. In battery powered appliances that run at low power,
the effect can be observed on the metal plating of the clips that
maintain electrical continuity between the battery and the
electronics.

We see this on our battery powered wall clocks and on our electronic
thermostat that runs our heating system. Every 1-2 years, these fail
to operate even though the battery shows full voltage under load.
Clean up the contacts, reinstall the battery and they operate agian
for 1-2 years before failing again.

I just noticed it happening to a laser poiter I have, same pattern
although I can't examine the internal battery contacts as well as I'd
like to. So, can't be absolutely sure it's the same problem.

M

 

[Mac]

The sealed lead acid batteries are not totally sealed and do vent some
mist of sulphuric acid.

Which kind of batteries are you talking about? There is a type of lead-
acid battery sometimes called a "gel cell" which uses a gel electrolyte
and which does not leak any kind of gas at all under normal circumstances.
At least that is my understanding. It is possible for the gel cell to vent
hydrogen gas if it is overcharged, but the vent does not normally open
unless there is pressure inside the battery container, and such
overcharging wrecks the battery anyway, so has to be avoided.

There is another lead-acid battery technology called absorbed glass mat
(AGM) or something like that. These are similar to gel cells, but the
electrolyte is absorbed in a glass mat, as I understand it.

Both of these battery types can be mounted in almost any orientation, and
don't leak AT ALL.

The manufacturers say not to charge in a confined space, but the biggest
problem as I can see is that this is quite corrosive to electronic
components, with several reports of it eating pcb tracks, attacking a
crystal in a very high quality oscillator so much it could not be put
onto frequency etc.

I don't think this spewing of sulphuric acid mist applies to gel cells and
AGM batteries. If you think it does, please provide more specific
references.

Has anyone found an answer to this problem? Sometimes you want battery
backup in a case with electronic components, but it seems they don't
exist too well with each other.

You could use some other battery technology. Li ion, or NiCad or Nickle
metal hydride.

I've been thinking whether there is any checmical solution, such that
you place a sacrificial chemical that will be more reachive with the
gases, leaving less to attack your components. Of course, any byproducts
would have to be non-corrostive.

I'm sure there is some type of conformal coating which will render your
board immune to corrosive atmospheres.

I've also thought about sealing the batteries in a plastic bag. The bag
and expand/contract as and when the batteies vent cases, but they are
kept inside the bag. I'm not sure if the batteries would like this
though. Any you still have the problem of sealing around the leads.

This is silly. I'm sure you don't need to do this.

--Mac

 

[Ross Herbert]

The sealed lead acid batteries are not totally sealed and do vent some
mist of sulphuric acid.

The manufacturers say not to charge in a confined space, but the biggest
problem as I can see is that this is quite corrosive to electronic
components, with several reports of it eating pcb tracks, attacking a
crystal in a very high quality oscillator so much it could not be put
onto frequency etc.

Has anyone found an answer to this problem? Sometimes you want battery
backup in a case with electronic components, but it seems they don't
exist too well with each other.

I've been thinking whether there is any checmical solution, such that
you place a sacrificial chemical that will be more reachive with the
gases, leaving less to attack your components. Of course, any byproducts
would have to be non-corrostive.

I've also thought about sealing the batteries in a plastic bag. The bag
and expand/contract as and when the batteies vent cases, but they are
kept inside the bag. I'm not sure if the batteries would like this
though. Any you still have the problem of sealing around the leads.

You don't mention the application or other requirements of the battery
supply so it is difficult to be specific.

If the application has low power requirement you can use NiMH or NiCd
cells.

If the power requirement is high such that it demands a lead acid type
battery for heavy duty then you shoulds look at one which is capable
of external venting. Such a battery is OPTIMA D31A
http://www.1st-optima-batteries.com/yellow_top_battery.asp

If you don't want to or can't vent externally then you have a problem.

If your application is not one requiring deep cycle, or deep
charge/discharge operation, then you shouldn't be charging to the
extent where gassing occurs. Gassing generally only takes place when a
heavy charge current is applied such as when boost charging. If
normal float operation is all that is required then you should select
a battery type suited to this type of operation and gassing shouldn't
occur if you are using the correct charger.

Protecting pcb's and mounted components for use in volatile
atmospheres is generally accomplished by proper housing of the
electronics components in conjunction with the application of
conformal coatings.

 

[Ban]

The sealed lead acid batteries are not totally sealed and do vent some
mist of sulphuric acid.

The manufacturers say not to charge in a confined space, but the
biggest problem as I can see is that this is quite corrosive to
electronic components, with several reports of it eating pcb tracks,
attacking a crystal in a very high quality oscillator so much it
could not be put onto frequency etc.

Has anyone found an answer to this problem? Sometimes you want battery
backup in a case with electronic components, but it seems they don't
exist too well with each other.

I've been thinking whether there is any checmical solution, such that
you place a sacrificial chemical that will be more reachive with the
gases, leaving less to attack your components. Of course, any
byproducts would have to be non-corrostive.

I've also thought about sealing the batteries in a plastic bag. The
bag and expand/contract as and when the batteies vent cases, but they
are kept inside the bag. I'm not sure if the batteries would like this
though. Any you still have the problem of sealing around the leads.

You are mostly wrong with your assumptions. Sulphuric acid mist is only a
byproduct when gassing occurs. What gets produced is hydrogen and oxygen.
Together in a bag or in a small room this is a potentially lethal condition.
Get Gel-cells if you charge in a confined space, like the hull of a boat or
a small room.

 

[Dave]

You don't mention the application or other requirements of the battery
supply so it is difficult to be specific.

OK,
to be more specific, is to generate an accurate frequency standard, for
a home project, using a combination of

1) A 10MHz oven controlled crystal oscillator - which has an oven using
more current during startup, but less when running. Needing 24V

2) A 10MHz rubidium atomic source - again with an oven, taking more
during startup, and less during use. Needing 24V.

3) A GPS recieiver to lock both of them to 10MHz, which will accept 24V.

I want to keep this on 24 hours per day, as that way the frequency stays
more accurate.

To keep these within the current of my power supply (2.3A), I will
sequence the starting up of these.

Once running, it total current consumption will be about *900mA*. I have
no need to start this on batteries, so the batteries will only supply a
little under 1A @ 24V during mains power failure.

If the application has low power requirement you can use NiMH or NiCd
cells.

Not reallly economic (for a home project anyway) at 24W of power
consumption, with a design time of 7 hours.

I have bought two "sealed lead acid" but not the gel type 12V 7Ahr
batteries. Now I am begginging to wonder if that was the right decision
or not - I think the answer is not.

If the power requirement is high such that it demands a lead acid type
battery for heavy duty then you shoulds look at one which is capable
of external venting. Such a battery is OPTIMA D31A
http://www.1st-optima-batteries.com/yellow_top_battery.asp

I'm in a no mans land here. The power is too much for economic use of
NiCd/NMiH, but too small for using the large lead acids that allow
external venting.

If you don't want to or can't vent externally then you have a problem.

I don't mind venting externally, and hence have been thinking about
possibly making some sort of gas tight seal around the batteries.
However, its only a 3U high box, so some of that will probably get
sucked back in via a fan. It's not going into a rack (will sit on its
own), so there should not be too much

If your application is not one requiring deep cycle, or deep
charge/discharge operation, then you shouldn't be charging to the
extent where gassing occurs. Gassing generally only takes place when a
heavy charge current is applied such as when boost charging.

There is no reason to fast charge, but under load, I will be taking
about 0.9A from a 7AHr battery, so a little over 0.1C.

Will gassing occur during discharge?

If
normal float operation is all that is required then you should select
a battery type suited to this type of operation and gassing shouldn't
occur if you are using the correct charger.

I have selected one for normal float operation and can design the
charger for it myself. It only needs a constant voltage, which must be
temperature compensated.

I will be limited to supplying about 2.2A of current, due to the
capacity of my power supply. (Yes, I know that is less than the 3A
startup current of this, but I will ensure the heaters in the two
devices that have heaters are not both starting together)

Protecting pcb's and mounted components for use in volatile
atmospheres is generally accomplished by proper housing of the
electronics components in conjunction with the application of
conformal coatings.

These are OEM units - a Stnaford PRS10 rubidium source, an old HP ovened
oscillator, but none are hermetically sealed.

A company made a similar unit, and it is well known for basically
destroying itself. I don't know if it was bad design of the charger, or
just that you can't avoid this. Apparently the batteries and electronics
were sepparated by a metal

 

[Guy Macon]

to be more specific, is to generate an accurate frequency standard, for
a home project, using a combination of

1) A 10MHz oven controlled crystal oscillator - which has an oven using
more current during startup, but less when running. Needing 24V

Far less accurate and stable than your rubidium oscillator.
Any reason why you need several sources?

2) A 10MHz rubidium atomic source

That rubidium oscillator shuoldn't vary more than 10^-10 from
-20 C to +65 C. Are you sure you need an oven?

3) A GPS recieiver to lock both of them to 10MHz

Like this? [ http://denk-stein.com/en/meinberg-mt_transport.htm ].

I have bought two "sealed lead acid" but not the gel type 12V 7Ahr
batteries. Now I am begginging to wonder if that was the right decision
or not - I think the answer is not. ....
I'm in a no mans land here. The power is too much for economic use of
NiCd/NMiH, but too small for using the large lead acids that allow
external venting.

The kind of batteries used in commercial alarm boxes put out the
kind of power you have been talking about, and they do a quick
discharge when the bells start ringing and the lights start
flashing. I have never seen any commercial alarm box with
external venting.

 

[martin griffith]

You are mostly wrong with your assumptions. Sulphuric acid mist is only a
byproduct when gassing occurs. What gets produced is hydrogen and oxygen.
Together in a bag or in a small room this is a potentially lethal condition.
Get Gel-cells if you charge in a confined space, like the hull of a boat or
a small room.

ISTR that you can get the battery caps with a piece of platinum wire
fixed in them. This was said to force the H and 0 to recombine to
water. The main idea was to prevent water loss, in remotely located
systems


martin

 

[Roger Lascelles]

The sealed lead acid batteries are not totally sealed and do vent some
mist of sulphuric acid.

The manufacturers say not to charge in a confined space, but the biggest
problem as I can see is that this is quite corrosive to electronic
components, with several reports of it eating pcb tracks, attacking a
crystal in a very high quality oscillator so much it could not be put
onto frequency etc.

Has anyone found an answer to this problem? Sometimes you want battery
backup in a case with electronic components, but it seems they don't
exist too well with each other.

I've been thinking whether there is any checmical solution, such that
you place a sacrificial chemical that will be more reachive with the
gases, leaving less to attack your components. Of course, any byproducts
would have to be non-corrostive.

I've also thought about sealing the batteries in a plastic bag. The bag
and expand/contract as and when the batteies vent cases, but they are
kept inside the bag. I'm not sure if the batteries would like this
though. Any you still have the problem of sealing around the leads.

Dave, from bitter experience, I can agree that AGM "sealed" batteries do
leak acid vapour. We have a product with AGM battery and temp compensated
charger set for a fairly low charge voltage and we get leakage.

If there is very slight leakage and we find it in time, then we can wash the
board and put it back in service. Once the acid gets through the board
coating, the the board must be replaced due to electrical leakage.

I heard on this newsgroup that batteries sitting with valves at the top are
less troublesome than batteries at other orientations, especially inverted
batteries, because liquid is less likely to issue forth. Unfortunately, our
product has a battery on its side.

In future, we will place batteries - of whatever type - in a separate
compartment with an external door, and seal the wires or terminals off from
the innards with silicone rubber. NiCads and other types can leak too.

Leakage depends on brand, age and charge-discharge history. We also suspect
the practice of charging multiple new batteries in parallel at our factory -
we think a more discharged battery gets a lot of current from the better
charged batteries, exceeding the battery's max charging rate.

Gell batteries may be better. Others may know if and why.

If you prise off the top cover of an AGM battery, you find puny little
rubber valves, which look as though they vent at low pressure.

There are other lead acid batteries - eg the Cyclon type, which are much
better sealed. There are alternative chemistries also, which we now prefer.

Roger Lascelles

 

[Roger Hamlett]

Dave, from bitter experience, I can agree that AGM "sealed" batteries do
leak acid vapour. We have a product with AGM battery and temp
compensated
charger set for a fairly low charge voltage and we get leakage.

If there is very slight leakage and we find it in time, then we can wash
the
board and put it back in service. Once the acid gets through the board
coating, the the board must be replaced due to electrical leakage.

I heard on this newsgroup that batteries sitting with valves at the top
are
less troublesome than batteries at other orientations, especially
inverted
batteries, because liquid is less likely to issue forth. Unfortunately,
our
product has a battery on its side.

In future, we will place batteries - of whatever type - in a separate
compartment with an external door, and seal the wires or terminals off
from
the innards with silicone rubber. NiCads and other types can leak too.

Leakage depends on brand, age and charge-discharge history. We also
suspect
the practice of charging multiple new batteries in parallel at our
factory -
we think a more discharged battery gets a lot of current from the better
charged batteries, exceeding the battery's max charging rate.

Gell batteries may be better. Others may know if and why.

If you prise off the top cover of an AGM battery, you find puny little
rubber valves, which look as though they vent at low pressure.

There are other lead acid batteries - eg the Cyclon type, which are much
better sealed. There are alternative chemistries also, which we now
prefer.

Roger Lascelles

One other alternative, is to look at 'manifold vented' batteries. These
have a chamber over the cells, with a seperate vent pipe. You then connect
this to a rubber hose, and have this fed outside the casing. You can even
feed the vent pipe to a seperate chamber with alkali filter elements.

Best Wishes

 

[Don Baker]

Gas Recombinant type batteries with AGM construction are used in critical
medical backup systems and vent 0 gas unless overcharged. I have used these
in several applications, including a rubidium backup and GPS system based on
BALL/Efratom oem OSCILLATORS (Rb) for long term remotely located
applications.. The critical step is a well designed power supply to strictly
control charge rates. Since these cells are a lead acid technology your
power density sounds realistically obtainable.

Don

 

[Dave]

Dave wrote:




Far less accurate and stable than your rubidium oscillator.
Any reason why you need several sources?

It's not quite as simple as that. The rubidium standard itself has a
crystal:

http://www.thinksrs.com/products/PRS10.htm

as you see it says "10 MHz rubidium-disciplined crystal oscillator"

There seems no way if having getting around having a decent crystal if
you want a decent frequency standard, as the short term stability of a
crystal is better than anything else. So that rubidium has an oven - I
am not putting it in any oven.

Also, the phase noise performance of the HP10811A oscillator I will be
using has a spec that is 100x better than the phase noise on that
rubidium disciplined crystal.

I will get someone with a cesium source or hydrogen maser to compare
them for me.

That rubidium oscillator shuoldn't vary more than 10^-10 from
-20 C to +65 C. Are you sure you need an oven?

The rubidium has its own internal oven. I'm not putting any external one.

3) A GPS recieiver to lock both of them to 10MHz

Like this? [ http://denk-stein.com/en/meinberg-mt_transport.htm ].

Yes, pretty much like that. Just building it myself, although with the
Stanford PRS10 locking to GPS is a doodle, as the PLL is built into the
box. For the crystal, I will build a PLL, based on this design.

http://www.rt66.com/~shera/index_fs.htm

The kind of batteries used in commercial alarm boxes put out the
kind of power you have been talking about, and they do a quick
discharge when the bells start ringing and the lights start
flashing. I have never seen any commercial alarm box with
external venting.

I guess it depends on the design life. The alarm is probably not
designed to last 20 years, but that rubidium standard has a design life
of 20 years. Hence I suspect I need to be a bit more careful not to
destroy the thing.

 

[Dave]

Gas Recombinant type batteries with AGM construction are used in critical
medical backup systems and vent 0 gas unless overcharged. I have used these
in several applications, including a rubidium backup and GPS system based on
BALL/Efratom oem OSCILLATORS (Rb) for long term remotely located
applications.. The critical step is a well designed power supply to strictly
control charge rates. Since these cells are a lead acid technology your
power density sounds realistically obtainable.

Don

Any particular manufacturers you are aware of? I can't find any that say
zero leakage, only that it is small.

 

[Ross Herbert]

OK,
to be more specific, is to generate an accurate frequency standard, for
a home project, using a combination of

1) A 10MHz oven controlled crystal oscillator - which has an oven using
more current during startup, but less when running. Needing 24V

2) A 10MHz rubidium atomic source - again with an oven, taking more
during startup, and less during use. Needing 24V.

3) A GPS recieiver to lock both of them to 10MHz, which will accept 24V.

I want to keep this on 24 hours per day, as that way the frequency stays
more accurate.

To keep these within the current of my power supply (2.3A), I will
sequence the starting up of these.

Once running, it total current consumption will be about *900mA*. I have
no need to start this on batteries, so the batteries will only supply a
little under 1A @ 24V during mains power failure.


Not reallly economic (for a home project anyway) at 24W of power
consumption, with a design time of 7 hours.

I have bought two "sealed lead acid" but not the gel type 12V 7Ahr
batteries. Now I am begginging to wonder if that was the right decision
or not - I think the answer is not.


I'm in a no mans land here. The power is too much for economic use of
NiCd/NMiH, but too small for using the large lead acids that allow
external venting.


I don't mind venting externally, and hence have been thinking about
possibly making some sort of gas tight seal around the batteries.
However, its only a 3U high box, so some of that will probably get
sucked back in via a fan. It's not going into a rack (will sit on its
own), so there should not be too much


There is no reason to fast charge, but under load, I will be taking
about 0.9A from a 7AHr battery, so a little over 0.1C.

Will gassing occur during discharge?


I have selected one for normal float operation and can design the
charger for it myself. It only needs a constant voltage, which must be
temperature compensated.

I will be limited to supplying about 2.2A of current, due to the
capacity of my power supply. (Yes, I know that is less than the 3A
startup current of this, but I will ensure the heaters in the two
devices that have heaters are not both starting together)


These are OEM units - a Stnaford PRS10 rubidium source, an old HP ovened
oscillator, but none are hermetically sealed.

A company made a similar unit, and it is well known for basically
destroying itself. I don't know if it was bad design of the charger, or
just that you can't avoid this. Apparently the batteries and electronics
were sepparated by a metal

Thanks for the compehensive description. I would definitely use gel
cells for your project. Given the quite modest power requirements you
won't have any problems with venting of any volatile gases since they
won't exist. I would just use polyurethane conformal coating on
boards.

 

[Roger Lascelles]

If your application is not one requiring deep cycle, or deep
charge/discharge operation, then you shouldn't be charging to the
extent where gassing occurs. Gassing generally only takes place when a
heavy charge current is applied such as when boost charging. If
normal float operation is all that is required then you should select
a battery type suited to this type of operation and gassing shouldn't
occur if you are using the correct charger.

Ross, this just doesn't fit our experience with thousands of AGM cells of
about 3 major brands. Our product has a good temperature compensated
voltage regulated charger system which sits voltage at the low extreme of
the manufacturer's float charge voltage vs temperature graphs. This is not
a mickey mouse piece of engineering and product quality is good. Some units
come back with acid mist damage and the charger in those units checks out
perfectly.

Recently, we carefully voltage charged new batteries from stock, removed the
vent cover plate and observed moisture around the vents on 10% of that
particular battery stock.

The school of reality has forced me to accept that AGM batteries *do* vent
sometimes. Often its a tiny amont of gas which you would never notice unless
you enclosed the battery.

I suspect the factors like:

manufacturer and batch
temperature and rate of temperature variation, especially during charge
depth of discharge
rate of discharge
max available charge current
manufacturing defects
battery ageing
period of shelf storage
cell differences
wearout characteristics
charger algorithm for programmed charger cycles


Roger

 

[Dave]

Ross, this just doesn't fit our experience with thousands of AGM cells of
about 3 major brands. Our product has a good temperature compensated
voltage regulated charger system which sits voltage at the low extreme of
the manufacturer's float charge voltage vs temperature graphs. This is not
a mickey mouse piece of engineering and product quality is good. Some units
come back with acid mist damage and the charger in those units checks out
perfectly.

I think my approach is going to be to
b) Remove the plastic cover on the top of the battery (it just unclips,
exposing the 6 vents.

2) Drill a hole in the cover and attach some form of pipe fitting, so I
can vent the fumes out.

3) Seal the plastic cover back, using an adhesive - quite what I do not
know, as clearly any mist attacking the adhesive could cause a problem.

As long as I don't cover the vents in adhesive, this only leaves one
place for the fumes to escape, and that is outside the unit.

Comments?

 

[Roger Lascelles]

Roger Lascelles wrote:

I think my approach is going to be to
b) Remove the plastic cover on the top of the battery (it just unclips,
exposing the 6 vents.

2) Drill a hole in the cover and attach some form of pipe fitting, so I
can vent the fumes out.

3) Seal the plastic cover back, using an adhesive - quite what I do not
know, as clearly any mist attacking the adhesive could cause a problem.

As long as I don't cover the vents in adhesive, this only leaves one
place for the fumes to escape, and that is outside the unit.

Comments?

I think your idea is excellent.

Vent construction varies - some vents are rubber plugs held in place by the
welded on cover, which release when you remove the cover. Once you prise
off the cover, these batteries aren't really good for serious work, even if
you glue or tape the cover back down.

Looks like you have a battery with a better type of vent where the cover is
not doing work.

Silicone rubber (window - bathroom stuff from hardware store) handles acid,
holds quite well over an area and gives assurance of a seal. Also
polystyrene cement gets a bite on that battery plastic. Epoxy ??

Roger

 

[Ross Herbert]

Ross, this just doesn't fit our experience with thousands of AGM cells of
about 3 major brands. Our product has a good temperature compensated
voltage regulated charger system which sits voltage at the low extreme of
the manufacturer's float charge voltage vs temperature graphs. This is not
a mickey mouse piece of engineering and product quality is good. Some units
come back with acid mist damage and the charger in those units checks out
perfectly.

Recently, we carefully voltage charged new batteries from stock, removed the
vent cover plate and observed moisture around the vents on 10% of that
particular battery stock.

The school of reality has forced me to accept that AGM batteries *do* vent
sometimes. Often its a tiny amont of gas which you would never notice unless
you enclosed the battery.

I suspect the factors like:

manufacturer and batch
temperature and rate of temperature variation, especially during charge
depth of discharge
rate of discharge
max available charge current
manufacturing defects
battery ageing
period of shelf storage
cell differences
wearout characteristics
charger algorithm for programmed charger cycles


Roger

Roger,

I bow to your superior experience and I do agree that AGM batteries
can vent a small amount of gas during charging, particularly at higher
temperatures. I haven't any direct experience using AGM batteries in
enclosed spaces, my experience has been on larger 1000Ah and up,
flooded cell batteries operated in float conditions, and these
batteries don't issue gas except when being boost charged, which is a
rare occurrence.

I was considering the fact that the OP was using quite small capacity
batteries in a float situation and surmised that if he used a good
quality 3 stage charger then gassing shouldn't occur.

While it is beyond the scope of the project being developed by the OP,
the US defence forces specification requires AGM batteries which do
not gas even with heavy charge and discharge currents at elevated
temperatures. According to one particular manufacturer, Lifeline
advanced AGM batteries are the only ones which currently meet this
requirement, so hopefully their technology may eventually be applied
to smaller capacity batteries suitable for rack mounted applications.

http://www.lifelinebatteries.com/

 

[Dave]

I think your idea is excellent.
Cheers!

Vent construction varies - some vents are rubber plugs held in place by the
welded on cover, which release when you remove the cover. Once you prise
off the cover, these batteries aren't really good for serious work, even if
you glue or tape the cover back down.

I'll have to double-check the fact, but it did not look like my vents
had released, but I must admit I did not look too carefully. The cover
went back on without (at least as I noticed), it then being obstructed
by the vents, which looked to be in place.

But whether or not they had raised up 0.5mm or so I am unable to say.

I'll try to clarify this with the manufacturer. If nothing else, it
might make them consider the issue for small batteries, which never
seems to be done.

I am now in a position to know where I could drill without removing the
cover, so could in future do it without even removing the cover, if
removing the cover automatically releases the vents. But I don't think
it does in my case.

Looks like you have a battery with a better type of vent where the cover is
not doing work.

I hope so, but I will ask the manufacturer.

Silicone rubber (window - bathroom stuff from hardware store) handles acid,
holds quite well over an area and gives assurance of a seal. Also
polystyrene cement gets a bite on that battery plastic. Epoxy ??

I was going to look around the web to try to find what is most likely.
The had thought about silicon rubber and epoxy myself.

 

[Roger Lascelles]

Roger,

I bow to your superior experience and I do agree that AGM batteries
can vent a small amount of gas during charging, particularly at higher
temperatures. I haven't any direct experience using AGM batteries in
enclosed spaces, my experience has been on larger 1000Ah and up,
flooded cell batteries operated in float conditions, and these
batteries don't issue gas except when being boost charged, which is a
rare occurrence.

I was considering the fact that the OP was using quite small capacity
batteries in a float situation and surmised that if he used a good
quality 3 stage charger then gassing shouldn't occur.

While it is beyond the scope of the project being developed by the OP,
the US defence forces specification requires AGM batteries which do
not gas even with heavy charge and discharge currents at elevated
temperatures. According to one particular manufacturer, Lifeline
advanced AGM batteries are the only ones which currently meet this
requirement, so hopefully their technology may eventually be applied
to smaller capacity batteries suitable for rack mounted applications.

http://www.lifelinebatteries.com/

Reinforces the idea that some brands of small AGM will leak a less often
than others -even if none of them are up to the Lifeline standard. We have
just recently dumped a particular brand - lets hope the Century-Yuasa ones
we are trying will do better. A very useful insight for me. Thanks.

Roger