Interesting info. Must've read something wrong. How do you provide an equalizing charge every few weeks exactly?
That is the nature now to make things that don't last and create a lot of waste. Many of these UPS are bought and used by business owners and they normally switch these out every 3 years due to the suggestions of their IT department.
I don't think the recommendation to replace batteries in typical UPS systems every 3 years is bad advice at all. For maximum system reliability, this seems like prudent practice. Especially given that most UPSs do not go to any great lengths to maximize their battery life.
The life of most of these batteries depends on a lot of factors. Temperature, how "properly" they're charged, whether or not they're actually discharged much or not, etc.
Higher temperatures reduce life. More discharge cycles reduces life. Deeper discharge reduces life.
I remember reading that you should only discharge a SLA as deeply as the application demands, and only as often as necessary. Most rechargeable batteries have a finite "charge-discharge" life. So purposely discharging any rechargeable battery is usually a bad idea.
The idea that purposely discharging cells is "good" came from an old wive's tale about Nickel Cadmium cells. But even with NiCds, this was a myth for almost all practical uses.
This legend began when NiCds were used in geosynchronous satellites. The satellites recharged the NiCds when their solar panels were exposed to sunlight, and then discharged the cells in use when the satellite was in darkness. Naturally, this happened like clockwork every day. I read an article written by one of the satellite engineers many years ago in a Ham Radio magazine where he described the phenomenon that they observed.
What he said was that the incredibly regular and perfectly-repeated charge and discharge cycle that their NiCds were exposed to created an effect known as "cell voltage depression". This was NOT, and never was "battery memory". He said such a phenomenon never existed, and is a myth. Instead, what they observed (again, with an incredibly regular and perfectly repeated charge and discharge regime) was a situation where the cell voltage of the NiCds in question would decrease over time. To combat this reduced cell voltage, they would take one battery off line about once per year, and discharge it deeply, but carefully, being sure to not allow it to discharge TOO deeply such that the stronger cells in the battery could reverse charge the weaker ones.
This once-per-year "sort of deep" cycling of one battery at a time recovered the cells in that battery from the state of reduced cell voltage that had accumulated over the past year from the daily, and extremely exact charge-discharge that the battery had experienced.
He made the point that in no typical consumer use would a battery ever be exposed to such a perfectly repeated charge-discharge pattern every day for a year! And purposely discharging any battery was a waste of cell "cycle life" and in many cases would damage the cells.
This is especially true for any lead-acid battery. Lead acid batteries, as anyone who has accidentally left the lights on in their car knows, are easily damaged by being allowed to discharge too deeply. Many modern lead-acids are greatly improved in their tolerance for deep discharge. But even with these improvements, there is nothing to be gained, and considerable danger, in deep discharge of any lead acid. Often, protection systems are built into systems that use Lead Acids to prevent them from being over-discharged. If nothing else, purposely discharging a battery of any type, will "waste" some of its charge-discharge life.
Now it is true that when a battery is new, it will often gain capacity with the first few charge-discharge cycles. But it is usually best to just use the device normally and allow this to happen, rather than to purposely cycle the battery in some "non use" way. You might as well be getting some good out of even those first charge-discharge cycles.
This guy's recommendation, for NiCds, was to simply use the device (tool, radio, flashlight, etc.) as you normally would, and pay attention to its operation. When the device begins to show reduced performance, that's the time to recharge its battery. Don't run it after the performance is obviously diminishing. To do so, you risk discharging the pack too deeply, and allowing the "stronger cells" to reverse-charge the weakest one(s). This is why it's often good practice to have more than one battery pack for any radio, tool, flashlight, etc. You can swap packs when the drill (for example) begins to slow down, and set the dying pack aside for charging at your earliest convenience.
Anyhow, the myth of "battery memory" was born from the satellite "voltage depression" story as well as another interesting problem often seen with early NiCd packs, particularly in hand calculators.
A problem with NiCds is that if you trickle charge them continuously, "dendrites" of metal can grow, and these dendrites can puncture the cell plate separators, creating high-resistance short circuits. So things like calculators that one might leave on trickle charge for far too long, or even continuously, will likely eventually develop these partially-shorted cells. The symptom of this is that the battery pack, once removed from the charger, will self-discharge whether or not you're using the calculator or other device. So it appears that the battery pack has developed "memory", and won't give full capacity. But in reality, you've ruined the pack by leaving it on a trickle charger constantly, or for too long for each cycle.
These trickle chargers were very cheap to manufacture, often using a transformer with a "too small" core to limit the current, and a single diode to rectify the output. There was no voltage regulation, and no timer or voltage sense system to shut off the charging once the NiCd pack was fully charged. As a result, people wrecked millions of these NiCd packs by constantly charging them.
Deep cycling them was often claimed to remove this "memory", but it never did! It was all well-distributed hokum.
Automatic battery pack cycling has even been built into special chargers sold to RC car and plane enthusiasts, and there IS some validity to cycling NiCd packs to gain uniformity and higher cell voltage to some degree. But in general, purposely discharging any battery in an uncontrolled way is just throwing away cycle life from your battery.
And with more modern batteries, there are often quite sophisitcated charge controllers built right into the battery packs themselves. Especially for Lithium Ion rechargeables which can be quite dangerous if treated "badly". So combination protection and charge-control ICs are usually included in these LiIon battery packs. And that's a good thing!
But back to Lead Acids:
No battery type has suffered more from the myth of battery "memory" than Lead Acids. Lead Acids HATE to be discharged. Yeah, that's their purpose, but still, you should only discharge them as deeply as you must, and only as often as you must. People wreck lead acids all of the time by running them down too far. And they reduce their life by running them down for no reason.
But another detail that is often overlooked in the care of Lead Acids is the importance of equalizing the cells within a battery.
Years ago, I worked on old systems that used banks of lead-acid or lead-calcium cells to create 48 Volt systems. These were used for telephone systems and even computers of the day. These huge batteries consisted of banks of huge glass containers, each of which held a giant lead-acid "wet cell". The cells were connected in series (24 two-volt cells) to form 48 Volt batteries.
The chargers for them kept them at the proper float voltage. They had temperature probes that were located between the cells, and they adjusted the float voltage to compensate for the cell voltage at different temperatures. They also had a mechanical timer that you set by hand to apply an "equalizing charge" every so often. You just turned the timer dial, and gave the battery bank an equalizing charge every so often.
What that equalizing charge is, is simply charging the battery to a substantially higher voltage than the nominal "float voltage" for that cell at that temperature.
By doing this, you make sure that all of the cells in the series-connected string get fully charged every so often, so you don't develop cells that are under-charging.
With, for example, a 12 Volt Sealed Lead Acid battery, you have six cells connected in series. Because you cannot access the terminals of each cell separately, and you cannot therefore apply the exact proper float voltage to each cell independently of each other, it is helpful to "equalize" the cells periodically.
Since you cannot replace the water in the cells of a "sealed" lead acid battery, you must be very careful to not "boil off" the water in your cells. Water can be lost from the cells if you overcharge the cells at too great of a current such that the hydrogen and oxygen that is dissociated by electrolysis cannot be recombined within the cell, and it "gasses off". That water is lost forever, and damages the battery.
So the technique I got many years ago, and incorporated into many lead-acid charger designs is this:
You raise the voltage applied to (for example) a 12 Volt SLA battery to approximately 14.7 Volts (2.45 Volts per cell). The current is limited to 1/10th the amp-hour rating of the battery. But you also monitor the current that the battery is accepting, and when it drops to 1/100th the Amp Hour rating of the battery, you terminate this equalizing charge, and drop back to the proper "float voltage" for the battery (usually, around 13.65 Volts (2.275 Volts per cell) for a 12 Volt SLA at room temperature).
Then you keep the battery at that proper float voltage until the next equalizing charge.
In one of the chargers I designed for an automatic hearing tester, the small SLA got an equalizing charge any time it needed to be charged at greater than a certain current. That meant that the cells get equalized any time the unit is used long enough to discharge the battery deeply enough to trigger this. That made for an automatic way to guarantee that the battery received an EQ every so often. The instructions we gave the users was to plug the units in to charge ANY TIME the units were not being used. Period. Always have them plugged in unless they were being used in the field.
In the years that we had these devices in the field, not ONE battery needed to be replaced, ever, with the exception of one that a guy left discharged in a storage unit for several years, not charging.
Anyhow, equalizing the cells is a good idea. But it's best done by an automatic charging system that performs this task for the user, safely and correctly.
We had telemetry systems that we'd go out in the field and find that they had SLA batteries that were 20 years old and doing fine. On the other hand, we had people who would switch off the breaker or otherwise shut off the charging to the telemetry's battery, and let it run down. In every case, the solution to that was to replace the battery, which was basically wrecked. One customer, in particular, did this without fail every year like clockwork. When they called to complain that the telemetry had died, we'd just take a replacement battery with us. They shut off this breaker because there was something else on that circuit that they wanted to disable every year. We offered to rewire their panel to put the telemetry on its own separate circuit, but they always said they'd remember not to shut it off. Then we sold them a new battery that next spring, again, without fail, for ten years or more! The poor things discharged, then froze and swelled up nice and fat. These telemetry units used entirely CMOS logic, and would run for months and months on the charge in these 20AH SLAs. So nobody knew anything was wrong until the damage was done. We should have put a "power fail" alarm in that system as one of the telemetry alarms!
Here is PowerSonic's writeup on the care and maintenance of their SLAs. This document has likely been improved since I read it back in the late 1970s, but most of the information is the same. This is where I learned a lot of what I needed to know when designing our chargers for SLAs all those years ago.
https://www.power-sonic.com/wp-content/uploads/2018/12/Technical-Manual.pdf
