UPS - practical maintenance advices to help end-user keeping batteries (classic lead acid models) in good shape

7

4

I have read that most (maybe all?) UPS devices for personal computing have lead-acid batteries inside.

As long as its chemistry, compared to lithium-ion, is very different, I would like to know what general end-user maintenance cares should I remember in order to make my UPS keep working as its best for the longer possible time.

Please include the possibility of changing batteries. I am and end-user, but capable of using a screwdriver and changing things inside my computer, so I can do it too inside my UPS. Oh, yes, maybe it is not worth the effort for some models (if a new UPS is not expensive), but I am not sure what kind of batteries should I buy and all that stuff.

Sopalajo de Arrierez

Posted 2014-04-21T01:42:18.570

Reputation: 5 328

Answers

10

I am not a battery engineer, but I have been working for more than 15 years in computing assistance with end-users, and I have some practical knowledge that I would like to share.

First of all, it is good to remember that battery engineering is an art (a complex one). Conventions seem not to be completely scientific. It is a curious thing to find how often different engineers have arrived to different conclusions about it, sometimes even from inside the same company.

I have only worked with off-line/standby (the more widespread and less expensive in end-user market) models. I hope anyone knowing about the line interactive or online/double conversion can confirm if these advices are appropriate for them too:

  • When replacing batteries:
    • Buy a "deep cycle" model for your UPS. Never buy a model intended for another proposal (i.e: car/bike usage). Their internal design is different enough, and your battery will probably last only a few months.
    • Buy the fresher battery you can. They all should have a manufacturing date. Remember that batteries are a time-perishable product.
    • Buy a model with same voltage and amperage. Voltage is a need or you could break the device. Amperage is advised, because the rest of the circuitry is designed counting on it (I have tested batteries of 9Amp replacing 7Amp models, and results were only good in the first months) (*).
  • Avoid small discharges. Batteries don't like 3% discharges. Drain at least a minimum of 20% (that is, reduce the 100% charge to 80%). Smaller discharges generate dioxides and dendrites that reduce the lifetime of the batteries. Suggestion: when there is a power outage of only a few seconds, that will be a good time to start a (longer, in fact) battery draining, so it will not become a small discharge.
  • Avoid small charges. One of the reasons that short trips are not good for your car/bike is the incomplete reload (while motor is running and recharging it) of the (big amount: many Amperes) battery you wasted for the high-power consumption task of starting the engine. So, reload your UPS completely, as you cannot (you shouldn't be capable of) overcharge it.
  • Temperature: keep refrigerated at about room temperature. 17ºC (62ºF) approx. seems to be a good one for batteries. The graphic below (extracted from Interstate Batteries) shows lead-acid lifetime battery as a function of environment temperature (well, it is for car batteries, but the technology is very similar):

Battery Lifetime as a function of environmental temperature

  • Make it work. They like some "jogging". Keeping the lead battery unused reduce its lifetime. Once a month is a good idea. Two or three months is acceptable too. Make it coincide with a drainage if you want. Many UPS in the world stay unused during longer periods; their batteries can easily last no more than 1-2 years.
  • Draining the UPS:
    • This could sound like a dumb idea, but a switch for only one power plug is a neat trick to convince end-users (the hardest part, I swear) to drain batteries, because they only need to push a button (the switch). This example is a Schuko European model.

Schuko power switch

  • Drain to 50% approximately if possible. Will not be harmed if it goes down to 20% remaining (the level at where the UPS circuitry uses to turn off the full device). Better go down to 20% than making no drainages at all (like it uses to happen to most users: they forget about it).

  • Estimated life should be hundreds (more than 300-600) of cycles whichever the discharge depth will be. So, assuming a drainage/month, you will (theoretically) have a working battery for about 10 years (speaking only of life cycles). For those interested, see the example table below (extracted from AINO Micro Manufacturer's Data Sheet for its lead-acid deep cycle batteries), showing the estimated life (how many charge-discharge cycles the battery will be capable to give, and its corresponding charge retain capacity) is better for not too big discharges than for bigger ones (30% depth discharges will give you several hundreds of lifecycles more than 50% discharges): Number of cycles expectancy life as a function of discharge depth

  • Annoying beeping during drainages: choose a UPS model capable of disabling this beeping. In fact, there are often two beepings: the slow one (battery still has enough charge) and the fast one (careful: battery is running down). The idea behind is being capable of disabling the slow one so you can keep using your computer (and stay beside it) without having a headache, and the fast one will remind you to give back the electrical power to the UPS.

  • Desulfators: I have several 12Volts and 24Volts desulfators (example: PPL24 for 2 serial 12Volt batteries). Technically, they are fantastic to make the lead-acid batteries last for more years, but my results (maybe due to wrong installation) have given contradictory results; furthermore, the technology is very recent (commercial models are in the market since 2006, I think), so I don't know much about them. Sorry, but I don't know if I must advise you a desulfator. Anyway, they require manual installation, they are a bit awkward to attach to an already-fitted UPS case (the photo below show the ring connectors that must be adapted to the battery plug), and you could even need to do some cable solderings: this question is for "end-users", but for some others it could be worth to mention.

24Volts Desulfator (2x12Volt serial batteries)

As said, these are the advices based on my experience. As long as chemical engineering is a slow-evolving technology, I think they keep being valid. Anyway, if somebody more informed can make amendments, I will correct whatever was wrong.

(*) I have no more info than my personal tests about the Amperage detail. If someone can confirm, please report. I have found nothing on the web. References welcome.

  • Further readings (worthy! I advise them, at least the first one) for those interested:

  • History: Changes in this document (read only if this is not the first time you visit it):

    • 06-05-2014: Added info and image about desulfators. Yet unsolved the doubt.
    • 03-05-2014: Added Life Expectancy as a Function of Discharge Depth chart.
    • 02-05-2014: Small charges considered as harmful since now on. Removed request for confirmations.
    • 01-05-2014: Added "room temperature" (refrigeration too much is not good) as more appropriate description, according to my last researches.
    • 29-04-2014: Changed "Drain to **50% if possible**. Will not be very harmed if it goes to 20%" to "Drain to **50% if possible**. Will not be harmed if it goes to 20%" due to last researches.
    • 25-04-2014: Added "Avoid small charges" advice.

Sopalajo de Arrierez

Posted 2014-04-21T01:42:18.570

Reputation: 5 328

Much of what you've written about "Avoid small discharges", "avoid small charges", discharging to 50%, etc., is lore concerning NiCd batteries. It has very little to do with lead-acid types in UPS service. – Jamie Hanrahan – 2016-05-19T15:26:15.093

@JamieHanrahan, I don't think small discharges are harmless. I remember I read about it on the papers at the Further readings section. I don't know exactly where, but I remember having read some thing like "Car batteries are not so needed of deep discharges, as long as the great power needed to start a vehicle motor is already a deep discharge". Maybe someone over here could confirm with more exact links. – Sopalajo de Arrierez – 2016-05-19T21:33:33.717

The energy needed to start a car is not at all a deep discharge for a car battery in good condition. A typical starter motor for a small car might pull 100 A for 3 s.: less than a tenth of an A-h. The same car's battery would likely have a capacity of around 45 A-h. So the start takes about 1/500 of the capacity, hardly "deep". btw: some UPSs run a periodic quick self-test, running the load on the battery for about 10 seconds every week. (You can configure it but this is the default.) If this was damaging to the battery, don't you think they'd do something else? – Jamie Hanrahan – 2016-05-19T22:07:31.977

Well, @JamieHanrahan , the usual warranty period for a UPS battery is just 6 months. If the small discharge of the self-test shorten its lifetime, I think it is a great business for sellers. In fact, I don't use to find UPS devices with batteries at good health state after 2 years, or even less time. – Sopalajo de Arrierez – 2016-05-20T00:53:09.027

@JamieHanrahan, concerning the 1/500 capacity draining to start a motor, I can not rebate your numbers, but I dont think my bike could try to start (i.e: when the motor is to cold) 500 times, or even 50. – Sopalajo de Arrierez – 2016-05-20T00:54:55.290

The batteries in my UPSs, despite all those little self-tests, are usually good for three years at least (known because they do a full discharge self-test every month or every other month, depending on model). Don't see how a battery-damaging self-test protocol would help the UPS makers as most people don't buy their replacement batteries from the UPS maker; aftermarket ones are far cheaper. – Jamie Hanrahan – 2016-05-20T01:36:41.743

It is true that you won't get 500 starts-without-recharge out of a car or bike battery, but remember that a battery's A-h capacity is normally quoted at the 10-hour discharge rate, not the 30-minute rate! You should, however, get at least 10, which I think pretty well refutes the "one start is a deep discharge" claim. – Jamie Hanrahan – 2016-05-20T01:38:34.650

I would say, @JamieHanrahan, that maybe the batteries in your UPSs are usually good for three years at least precisely because they do a full discharge self-test every month or every other month. This is supposed to clean dioxides and dendrithes. Anyway, full discharges are not supposed to be good for lead-acid batteries (see the graph from AINO Micro Manufacturer above). – Sopalajo de Arrierez – 2016-05-20T20:49:28.180

Another possibility for motor batteries: 1.- You start your car (-10% battery). 2.- You drive to the nearby (it uses to happen on cities) so the fuel motor recharges your battery +3% and stop the motor. 3.- You start again the motor (-10% again, so battery goes to 83%). 4.- You go to some near place again... etc. In this case, you could left your battery at some 60% remaining charge, that is more close to deep discharge. – Sopalajo de Arrierez – 2016-05-20T20:52:42.853

It would be fine to have some measurement device integrated on lead-acid batteries, but the last time I asked for (about year 2009) the price of this tool was about 8.000$ (LOL :-O ) . So, most of our debates must be just especulations. – Sopalajo de Arrierez – 2016-05-20T20:54:16.413

Back when I had cars with actual ammeters (added by me) I noticed that charging after a normal start was usually complete within a few minutes of normal city driving. And I've run local errands "all day long" with no trouble starting the car at the end. So I don't think the "gradual approach to deep discharge" scenario is plausible, not if your charging system is working right. – Jamie Hanrahan – 2016-05-20T22:43:35.117

My goal with UPSs is not to coddle the batteries for maximum calendar time before replacement, but rather to USE them to supply power during outages. Confidence that they'll do the job does require testing the "run to failure" time periodically. If that means I have to replace the batts a little more often, well, so be it. Better that than wait until I need them to learn that they've gone bad. – Jamie Hanrahan – 2016-05-20T22:55:24.150

Aino's chart doesn't show much impairment (below 80%) until after 200 deep cycles. At one test per month, which is far more than the number I go through from actual power failures, that's 16+ years; I think I can stand it. But whether or not the deep-cycle test helps or hurts the battery, batteries do sometimes fail prematurely, so periodic tests are required if the batteries are to be relied upon. – Jamie Hanrahan – 2016-05-20T23:07:21.793

What about calibrating a newly replaced battery? Should I be concerned about that? – RogUE – 2016-09-30T08:56:09.257

@RogUE , I think calibration, at leats for lead-acid batteries, is a method to make the circuitry of the UPS know about the battery it has inside (via full charge and discharge), so it should be UPS device specific. – Sopalajo de Arrierez – 2016-10-01T10:03:05.690

It would probably be a bad idea to use a desulfator on a gell type cell in the common consumer UPS, work ok for big wet batteries, the art of desulfating can (does not have to) create gas, gas in the gell never leaves, just ruins them. A (slightly) higher amperage battery should not present a problem, an overrated one, could just be as bad as its fake rating. These gell cells are not good for long times or numbers of cycles at all, you wont get 100 50% out of them and not 25 deep. They should be replaced A) anytime it doesnt work well enough B) after about 4-5years. – Psycogeek – 2014-04-21T06:29:53.697

1I wouldn't bother to drain a gell cell, in a wet cell things can "move" a bit, again in these gelled electrolyte cells you cant "slam stuff back onto the plates with power" (my words). They will last as long on the proper maintainance charge as they will by cycling them any ammount. I have used about 70 of the little gems and the ones stuck in the UPS on its maintance charge lasted the longest in time, but indeed are becomming weak and a 50% discharge is not going to hurt, but I wouldnt be doing it as a regular thing. – Psycogeek – 2014-04-21T06:41:39.073

@Psycogeek, your argument about gas generation due to desulfation makes sense for me. I have the feeling that desulfation isn't doing much good to the UPS batteries in all my tests until now. But just a precision: resellers report that UPS batteries model use to be chrystalized gel, instead of just gel. I don't know how much does this affect. – Sopalajo de Arrierez – 2014-04-21T10:04:22.573

Yea the different types gets a little complicated. There is AGM (Absorbed Glass Mat) and "starved electrolyte" also of that same smaller size of lead acid and probably others. The AGM being completly different , but not likely to be used in the cheaper consumer stuff. All the ones in the 4 UPSes here are the same kind of gell cell you get at the hobby shop, and electronics store. They are Always way easier to replace than it might look from the outside. luckey it isnt an array of lower voltage cells. – Psycogeek – 2014-04-21T10:32:45.233

There is also compatably made li-fe-po batteries, I wish I could afford. the li-fe would love to be deep discharged, would last a full 10 years , uses very robust li-ion types of cells, and would not Dip in voltage so much under these loads, which could improve runtime by lots in high power need. Draining these gell cells in 15-20 minutes is harsh on them, the li-fe-po cells they are making psudo 12V batteries with would handle that better too. – Psycogeek – 2014-04-21T10:36:36.717

Respecting to discharges (to 50% suggested, 20% maximum), I am not proposing it as a "regular thing", but rather once each a few months, in order to clean dendrites, @Psycogeek. – Sopalajo de Arrierez – 2014-04-21T14:00:07.000

2

I'll offer some insight from someone who hasn't worked a lot with a UPS, but with way too much time spent with lead acid battery from a work-perspective. This is not UPS specific, but applies to Led/Acid accumulators in general:

  • Keep them at a decent temperature. Room temperature should be fine. Lower temp causes decreased performance
  • Contrary to popular belief, a Led/Acid battery that discharges and recharges is a healthy battery. Most beliefs that this is harmful stems from the fact that a discharged battery that stays like that will lose it's capacity to hold charge. Think of a car battery - It gets discharged and recharged fairly often. What eventually kills the battery are short trips where it gets partially discharged during start and not fully recharged. Running it down and charging it back up will prevent ion crystalization on the poles.
  • If the UPS will be inactive/switched off for a while, you should see what the manufacturer recommends as a storing charge. Most battery types have an optimal store-charge at which it will degrade slower. I believe that Lead/Acid is close to max, but i do not remember exactly (Li/Po is something like 40%, if my memory serves me right).

Jarmund

Posted 2014-04-21T01:42:18.570

Reputation: 5 155

@Jarmund "Most beliefs that this is harmful stems from the fact that a discharged battery that stays like that will lose its capacity to hold charge." You're thinking of "memory effect", which was associated with NiCd cells. It is flatly not an issue with lead-acid types. (It was never actually an issue with consumer NiCds either. Dumb trickle chargers did cause "voltage depression", which had a similar result of reduced runtime. But true memory effect was only ever seen in very large (aerospace) NiCds that were discharged a great many times to within a percent or two of the same point.) – Jamie Hanrahan – 2016-05-19T23:57:26.670

Indeed, @Jarmund, car/bikes batteries seem to last much more than UPS batteries, in my experience. It could be due to frequent charge/discharge cycles. And I think your report about the problem with partial (incomplete) charge reloads is an interesting one. When having more info about it, I will probably add it to my resume of advices (my experience seems to show this problem too, but I prefer to wait and see what does the forum think about). – Sopalajo de Arrierez – 2014-04-21T20:58:41.730

@SopalajodeArrierez I cannot remember the exact science behind cyclinc being healthy it, but it has something to do with crystalization in the acid. If the battery has been discharged and seemingly dead, a last ditch effort that often works is to overcharge it for a while on slightly higher than recommended voltage (PS: Dangerous, emergency use only. May cause explosion, fire, acid burn, plague, rodents, nazis, and other calamities) – Jarmund – 2014-04-21T21:55:56.990

Thanks, @Jarmund. I have added info about the small discharge matter (still as "suspected" until we found some more confirmations of it; remember: battery engineering is a complex science). – Sopalajo de Arrierez – 2014-04-21T22:22:14.710

Asked: 2014-04-21T01:42:18.570

Viewed: 2 334 times

Active: 2018-05-02T03:41:03.343