VRLA battery undervoltage and sulphation - safety hazard?

[dkjfnvfd]

Hi chaps,

Just wondering if I could get some independent people to confirm my thinking about a problem?

I have a "temporary" installation that's been deployed for the last 1.5 years since the proper charger and battery failed. It's always been slightly undervoltage at about 116V for 9 x 12v batteries (so should be on float of around 9 x 6 x 2.25 ~= 121.5V), the supplier of the temporary installation pointed out the low voltage would lead to premature sulphation and have mentioned it in all of their 3M maintenance reports. The sulphation makes sense to me because even a small undervoltage will accumulate a lot of debris on the weakest cell over a long period of time. I gather that debris will only accumulate on the undervoltages parts of a battery cell, so the maximum possibly reduction in capacity is a reflection of the reduction in float voltage, but again, without substantiation I would be inclined to mark this system down as "not good enough".

In the last 6 months the impedance measurements of the blocs has gone totally off the charts, from something in the 1 - 3 mOhm range to literally off the charts (>10 Ohm).

I've flagged this with the "responsible" engineer and he seems to be putting up a fight. He says the undervoltage is not a problem and it cannot be sulphation, and he autonomy tests the batteries himself to prove they're safe. Now, he won't let me see the batteries, he won't let me take my own measurements, and he says the problem is the maintenance contractor must not be measuring the impedances properly themselves.

My gut feeling is that I should be quite worried about somebody trying to autonomy test batteries that have internal impedances that have already increased by more than 10,000% and I have asked him if he knows the load current going through the blocs, the response has been "I didn't notice any heat during autonomy". My thought is that a successfull autonomy (NOT discharge test) test is not that useful at telling you if something will work the next day, and if we don't get a handle on this system now when it's already clearly in a bad state then next time it's really needed it could either open circuit or blow.

 

[itsmoked]

You appear correct in your assessment.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[FacEngrPE]

This is the first time I have seen "Autonomy" used in this context, so O looked it up.

https://www.constantpowerservices.com/questions/17-what-is-ups-battery-autonomy-and-how-is-it-calculated[/URL]]What Is UPS Battery Autonomy And How Is It Calculated?

Battery autonomy is the period of time (in minutes or hours) a UPS battery will last for at a specified load level in the event of a power outage. In essence, it is how long the inverter will run off battery power.

Autonomy can also be referred to as backup, discharge time, or runtime. It is a function of battery charge state, load size, and capacity.

I am not a battery expert, a parameter change as large as you cite above is an obvious indicator of trouble.

Ask your battery maintenance contractor if the electrolyte specific gravity indicates an actual charge level. That might give some evidence of your UPS not being able to perform, prior to attempting a Autonomy test. Otherwise you will find out the next time you have a power hickup.

 

[dkjfnvfd]

Thanks for the replies - I really appreciate it because I did wonder if I was missing something.

The responsible engineer finally agreed to investigate after I started cc'ing a lot of important people in my emails to him. It looks as if there is something (I don't know what, this guy won't let me see them) about these batteries that means it's very hard to take impedance readings, so the contractor has just been sending us readings of a non-conductive surface(although they appear to have got it right the first time they ever did it, when the impedances were quite low).

Still don't really know what the impedance is, but at least we're getting somewhere, even if progress is glacial.

An autonomy test is just when you turn the mains off and allow something to run off battery power for a period of time. In this case it's emergency lighting so he would be turning the power off for 3 hours. I often have to explain (over and over again) to enginers at work that an autonomy test tells you nothing about if the batteries will continue to work.

As far as I'm concerned, running batteries for 3 hours when you can't explain high impedance measurements is pretty irresponsible, but in this case nobody cares because nothing actually blew up.

In the words of Kurt Vonnegut - so it goes.

 

[MacGyverS2000]

it's very hard to take impedance readings, so the contractor has just been sending us readings of a non-conductive surface

Falsifying data... to what, save face? Save time (in the short run)? I'd fire someone for such a thing... at the very least, dump them from the project.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[dkjfnvfd]

Well, what I described was based on what the recalcitrant engineer told me. I've phoned up the contractor and they said it's conceivable so they are investigating & will get back to me soon.

I doubt they will get into trouble for it, that doesn't seem to be the way things work where I am.

 

[dkjfnvfd]

This situation has dragged on a bit so I'm wondering if anybody can check my reasoning here.

It turns out the measurements were wrong - very wrong - and it nows appears to be "impossible" to take a correct impedance measurement as some kind of buildup or corrosion has happened on the terminals. It's not clear why the contractor didn't flag this up, but based on conversations with them I think that they did think it was at least possible that the blocs were in that state due to the undervoltage.

The responsible engineer now says he has been taking extra measurements to mitigate the bad impedances all along but as usual he refuses to tell me what they are. I'd say he needs to take frequent voltage measurements and observe all of the blocs and links using a thermal imaging camera during autonomy testing, but despite what he's saying I doubt he has - because if he was then he'd have told me already.

Now, bad impedance or not, extra measurements or not, I think the undervoltage is still an issue that needs adressed. I get that this charger and blocs are a temporary solution but "temporary" is never ever just that (it's already been in a year longer than planned) so I think a plan needs to be in place to deal with a failed bloc when it happens to make sure the emergency lights can continue to work. The RE says all sulphation does is reduce capacity, but I am saying that even such a small undervoltage *coud* cause a bloc to fail.

The logic I have for that assertion is this:

[ol 1]
[li]A number of blocs have an average per cell voltage of <= 2.14 Vpc[/li]
[li]When a bloc is undervoltage, it is concentrated on the weakest cell[/li]
[li]When a cell discharges, sulphation accumulates on the plates in proportion to the % discharged, however if the bloc is recharged then the sulphation reverses[/li]
[li]If a cell is left in a discharged state for some time (let's say starting at 3M without a topup then the sulphation begins to coalesce and crystalise, at which point it is very difficult to reverse[/li]
[li]Hard sulphation can spread, and can do this very rapidly, either open circuiting a cell (bloc) or causing a dangerously high impedance[/li]
[/ol]

So it would be OK to partially discharge the blocs to the rate they're at and then leave them for a short period of time (<M3), but to be in a position where it could be for years is unsafe?

I have asked a couple of people at work to peer check but I honestly think this situation is at a point where the technical depth is beyond what any of us really know, although my understanding based on my own research aligns with what the contractor told me.

Another issue (not really technical I guess) is I'm told we have these quality standards at work, one of which is "no explosions, no fires, no unplanned outages" - so we are meant to be 100% watertight on issues, now we have an SOP for what a good or bad battery looks like and this system (and another couple that belong to this RE) are way out of spec. When I go to standards and safety people I get told this stuff is a problem and needs sorted, but I'm beginning to feel like when I talk to the operations and engineering team I'm being treated as if I'm a pain in the @$$, and I can't just ignore it either because for some reason I'm held at least partially accountable for any battery out of spec.

 

[FacEngrPE]

dkjfnvfd You probably also have a standard that states "No Lost Time Accidents" but while that is a really desirable goal accidents are often a lagging indicator of problems being created elsewhere.

Could a periodic manual equalization charge help?

 

[dkjfnvfd]

I think a periodic refresh could help - although if there's been any permanent sulphation then it won't come off.

Before that can happen, the RE has to agree that there's a problem.

 

[MacGyverS2000]

Before that can happen, the RE has to agree that there's a problem.

Agreed... otherwise you're just pissing into the wind.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[FacEngrPE]

Equalization charges should be part of a large battery preventive maintenance program. A justification is that the wrong charger is currently installed. The correct charger would take care of the periodic equalization charge.

 

[sunnysky]

Respectful Opinions from Experience.

False: it nows appears to be "impossible" to take a correct impedance measurement
- a pulsed load test easily measures ESR at to levels for the double-electric charge layer model for the weakest link (cell).

(Is a ) VRLA battery undervoltage and sulphation (a) safety hazard?
Negative: But it's a battery failure mode from high ESR and low C.

Large series strings need to be balanced for ESR and C [kF or Ah] and a cell monitor balancer achieves better overall reliability.

A smart periodic built-in ESR battery tester, detects aging of battery.
A fast rise time 5W 25kHz pulse injector will extend battery life greatly for stand-by batteries by reducing and reversing sulphation.

In my mind the only failure was lack of a good customer requirement list for reliability and maintainability as these problems are systemic for VRLA in long term standby mode.
Peukert's law is a better way to describe "Autonomy".