On Fri, 13 Jun 2014 20:45:29 +0000 (UTC),
(Andrew Gabriel) wrote:

In article ,
Johny B Good writes:
On Fri, 13 Jun 2014 16:40:22 +0100, tony sayer
wrote:



The UPS manufacturers could have so easily addressed this issue but
it simply wasn't in their interest to do so (UPS supplied SLA
batteries are their equivilent to the inkjet printer manufacturer's
cash cow of inkjet refill cartridges).

Their argument for high charge rate is that the UPS needs to be ready
for use again within a short time.

That isn't a valid excuse to carry on cooking the batteries once they
are charged, which is what used to happen in one we had at work.

We used to get 3 years viable life from an APC SmartUPS UPS, or 4
years life to completely dead.

OTOH, in another datacentre, we had a central 80kW UPS (Chloride
Gaedor?) with separate batteries which were 10+ years old and still
as good as new.

I've noticed on all the APC ones they seem to cook their SLA batteries
after a while ....

Well, after I posted that, I took a look at the charging circuit for
the 2000 and it seems incredibly complex for its function. The whole
circuit diagram for the UPS is spread over 6 sheets (the charger is
sheet6 BTW). In this case, the power comes from transformer terminals
1 and 2 on sheet 3 and following the trail takes you to the inverter
powerfet stack circuit (a full bridge cct attached to the very same
terminals that are used to feed the 5A rectifier diodes on sheet 6 -
the inverter transformer does double duty as a charging transformer).

I gave up pondering the problem any further and decided to google to
the wiki on Lead Acid battery technology with regard to best charging
regimes where it all became ever so complex over the issue of choosing
a charging voltage that's high enough to avoid sulphation yet low
enough to minimise corrosion.

Interestingly there are three different float charge volts per cell
figures for Gel, AGM and flooded cell types (2.23, 2.25 and 2.32 to
within a figure of +/- 0.05v[1] respectively) for 20 deg C temperature
with a temperature compensation figure of -.0235v per deg C rise per 6
cell battery).

Reading the article suggests it might not entirely be the UPS
designer's fault but more the terrible limitations of Lead Acid
battery technology (but that doesn't explain the 6 month life under
the benign management scheme of a UPS versus the 5 to 10 years life in
the harsher conditions of starter battery use).

Car batteries are not normally run down at all. Starting the engine
requires only a tiny proportion of their capacity, and they're
normally recharged from this within a minute or two even on tickover.
They are excellent at providing high current, but running them
flat kills them very quickly. Their capacity drops fairly linearly
over their life, but as you only need a tiny fraction of their capacity
to start the car, you normally won't notice until they are will under
10% left.

I agree with that assessment but I think there was more to it than
just that. Despite the trickle charge tailing off to miliamps (when
new), this did eventually lead to high self discharge rates and cell
voltage imbalance which, in turn, led to higher trickle charging
current, accelerating the process of deterioration to the point where
the battery became a liability after a mere 6 months or so.

Considering the much higher float charge voltage used by alternators
(typically 14.0 to 14.2 volts) and the 150 odd amps discharge when
cranking the starter (and even higher current draw in wintertime
conditions), it's a rather surprising finding.

I've learnt the hard way (or, if you prefer, the easy way) that car
batteries aren't suited for such usage. Testing the batteries when new
to determine the endurance time to exhaustion 2 or 3 times in the
first week didn't suggest the relatively deep discharge was causing
any noticable loss of capacity.

The UPS would start charging the batteries within minutes of
completing each test which should have avoided the sulphation issue.
Admittedly, it needed a good 16 hours or more to complete the charging
process (normally a matter of 8 hours for the originally specified
17AH SLAs) but I don't think this was a material factor.

The normal advice over totally discharged Lead Acid batteries is to
not leave them longer than 24 hours in this state before putting them
back on charge. Perhaps I've misinterpreted this advice and it was
meant to say they should be fully recharged within 24 hours. Even on
the basis of this interpretation, the batteries were being fully
charged within this 24 hour period.

The problem seems to be exclusively down to the float charging
conditions since they'd deteriorate without being called upon to power
the inverter during a blackout.

One thing's for certain, I won't be entertaining the use of car
batteries as a substitute for SLA batteries. It's either SLA or Deep
discharge depending on how cheaply I can get my hands on a set of
four.
--
J B Good