Battery USB backups incase of power outs

[Streetwise]

We use heavy APC units for every network closet at all of our sites. We need to keep internet and VoIP phones online.

My primary home UPS is for my cable modem, router, AP.

For aquarium technology, I have learned that you want the best sine wave power output.

I also have an Apex that can detect when power is out even while connected to a UPS, so it can optimize what remains on while under battery backup.

[CT_]

On 3/23/2022 at 9:21 AM, Greg Stewart said:

The reason the primary unit needs to supply all the power to all the batteries and all devices is that the secondary/tertiary/etc units won't know that mains power is out and won't disconnect draw from "mains" (the primary unit) to avoid cascading that power draw. Mains disconnection comes from the mains supply, and by serially connecting units you provide "mains supply" to the next unit from the battery on the unit before it in series.

If they all start fully charged you may not loose that much efficiency.  But I agree over all you shouldn't connect them in series, you should just size one appropriately.  My biggest fear would be engineering realities causing unknown problems (eg one UPS might not like the quality of power its getting from another UPS.)

 

On 3/23/2022 at 12:49 AM, StevesFishTanks said:

You just have to make sure they output a pure sinewave. Its crucial to the linear pumps design.

Is it?  genuine question.  my understanding of how they work (100% derived from this picture below), leads me to believe a "simulated sine wave" from the cheaper ones would be fine.  As for synchronous AC motors (like the kind in HOBs) i'm 80% sure they'd be fine, other AC motors are all black magic to me so idk.  But i'm fairly out of practice as an electrical engineer these days, and I never really worked with electromechanical devices.

 

 

On 3/23/2022 at 2:09 PM, Streetwise said:

For aquarium technology, I have learned that you want the best sine wave power output.

I'm curious, what has happened with the cheaper "simulated sine" ones?

[Widgets]

Here is a FAQ on a pure sine wave vs modified sine wave.

https://theinverterstore.com/faq-items/what-is-the-difference-between-modified-pure-sine-wave-inverters/

[Greg Stewart]

On 3/23/2022 at 5:02 PM, Odd Duck said:

I will only have the air pump on back up and nothing else.  So 14 watts should give me decent run time even if not for as long as the Tetra-100.  Tetra says 5 watts on its underside, is that what that stands for, the Rated Input Power?  I know an older 1500 lasted about 8.5-9 hours when we had a power out a month or so ago when they were replacing a bad transformer nearby.  Maybe I would get somewhere between the 5 and 9 hours, I’m guessing?  When I finally found a chart that showed my model, it said 2 hrs 14 min for 50 watts.  That seems very short compared to what you’re saying.  Doesn’t seem to match my irl experience, either, but I know I can’t directly compare 14 W with 5 W.

Any recommendation for a better brand or different model?  Go big or go home?

Which chart did you find that said only 2 hours? Which model do you actually have?

If your experience is 8 to 9 hours with a 5W draw, 7 or so hours on a 14W draw doesn't seem off. But, battery health will impact that, and get "worse" over time. Your operational runtime will depend entirely on how good your batteries are at the time the power goes out.

Deciding which UPS to get, if you replace the ones you have, will depend on what, exactly, you need to get out of them. UPS units are not really designed to be long term power banks. They are really designed to allow you time to address a power outage before equipment goes down hard.

If you really need long term uptime for power on your fish room, if might be a better expense considering a backup generator with an automatic transfer switch. The larger the UPS you get, the larger the expense you're looking at. And, if you're prone to multi-day outages, you won't get good coverage from a UPS.

[Greg Stewart]

On 3/23/2022 at 5:27 PM, CT_ said:

If they all start fully charged you may not loose that much efficiency.  But I agree over all you shouldn't connect them in series, you should just size one appropriately.  My biggest fear would be engineering realities causing unknown problems (eg one UPS might not like the quality of power its getting from another UPS.)

On brand new batteries, fully charged with no memory, across the board on all units, yeah. There won't be too much loos of efficiency. But, we don't live in a perfect world. And, all batteries deteriorate over time. They are constantly being topped off, and it's the moment you have an outage that's the most critical. Batteries are usually running, draining, getting topped, for months, and by that time, you no longer have your brand new freshly charged batteries.

Line conditioning is also only part of it. You add load with every UPS and attached device you add serially, and that load is taken in its entirety by the primary unit in series. Unnecessarily overloading the head unit unnecessarily reduces its service life. I just don't see a win, there, in any case.

[CT_]

On 3/23/2022 at 4:45 PM, Greg Stewart said:

On brand new batteries, fully charged with no memory, across the board on all units, yeah. There won't be too much loos of efficiency. But, we don't live in a perfect world. And, all batteries deteriorate over time. They are constantly being topped off, and it's the moment you have an outage that's the most critical. Batteries are usually running, draining, getting topped, for months, and by that time, you no longer have your brand new freshly charged batteries.

I've only ever worked on lithium battery circuits.  Are you saying that lead acid needs higher trickle charge current as their health deteriorates?  random tangent: i hate lead acid the more I learn about them the more undesirable they seem.  i wish you could find more lifepo4 battery powered UPSs 😞 

[Greg Stewart]

On 3/23/2022 at 5:09 PM, Streetwise said:

I also have an Apex that can detect when power is out even while connected to a UPS, so it can optimize what remains on while under battery backup.

Ain't you special 😉

I've actually thought about the Apex, but my wife doesn't yet know how much I've spent on all the rest of my gear, and I'm saving up for my dream of an auto dosing system that i don't need to calibrate every three months, like a Neptune DOS. Their products are just so darned 'spensive.

[Streetwise]

@Greg Stewart, I had grand plans for CO2, lights, etc. The monitoring, alarms, and leak detection kits are what I really appreciate today.

[Greg Stewart]

On 3/23/2022 at 7:52 PM, CT_ said:

I've only ever worked on lithium battery circuits.  Are you saying that lead acid needs higher trickle charge current as their health deteriorates?  random tangent: i hate lead acid the more I learn about them the more undesirable they seem.  i wish you could find more lifepo4 battery powered UPSs 😞 

Most consumer grade/price point UPS systems use lead-acid. There are some Li-ion units on the market, but they're more limited. Lead acid does retain memory, but it's also non-flammable. Li-ion batteries can suffer from thermal runaway, and with the size of UPS batteries, the potential for ignition makes a possible fire a far bigger deal. Li-ion fires are difficult to put out.

[CT_]

On 3/23/2022 at 5:15 PM, Greg Stewart said:

Most consumer grade/price point UPS systems use lead-acid. There are some Li-ion units on the market, but they're more limited. Lead acid does retain memory, but it's also non-flammable. Li-ion batteries can suffer from thermal runaway, and with the size of UPS batteries, the potential for ignition makes a possible fire a far bigger deal. Li-ion fires are difficult to put out.

Sorry.  I sort of jumbled things in my last post.  I meant, do worn out batteries need more current to float charge?

 

Also lifepo4 batteries are pretty safe especially compared to the more typical lithium ion chemistries which is why I wish they were more available. 

[Widgets]

This is the first I have heard of lead acid batteries having a memory effect; but considering the misunderstanding that the public has on the memory effect, it should not be that surprising. 

From my understanding memory effect is limited to NiCad batteries. If they are repeatedly discharged down to the same level and "properly" recharged, they will develop a very small knee drop-off in the voltage discharge profile. You need sophisticated gear to even see the memory effect.

What most people know as the memory effect is really electrolyte crystallization due to overcharging. The consumer level battery chargers were driven by a simple charger with no circuitry to stop the charge when the batteries were full. If the charger was designed to full charge a fully depleted battery in 8 hours, the battery would get 8 hours of charge regardless of the starting charge level. The extra charging after the battery was full would crystallize the electrolyte and reduce the available battery capacity. Folks would see the reduced capacity and call it the memory effect.

Now, we are quick to pull out memory effect as the cause for capacity degradation due to overcharging. Modern chargers should have circuitry to prevent overcharging, but cheaper chargers will sometimes use the old designs to cut costs.

All battery technologies have pros and cons. The trick is to match the battery technology to the use. Lithium ion batteries can suffer capacity degradation when kept long term at a high or low state of charge, and are best stored at 50% charge. For our UPS use, we all want to have fully charged batteries at the ready. We don't have advance warning to charge the battery before an outage. Because of this Lithium ion batteries are a poor choice for UPS, but consumer demand is pushing them onto use and making more money for the manufacturers.

[Greg Stewart]

@Widgets You're right. I was mis-using the term "memory" for lead-acid batteries. lead-acid suffers sulphation crystalisation due to inconsistent charging, and over time this results in a reduction of capacity. And, "home use" is generally inconsistent (industrial use is frequently not that great, either) so, for reliability I automatically factor in that inconsistency.

[Greg Stewart]

On 3/23/2022 at 11:39 PM, CT_ said:

Sorry.  I sort of jumbled things in my last post.  I meant, do worn out batteries need more current to float charge?

I believe the charging capacity/retention is unrelated to the supplied current. But, I could be wrong on this. Higher current would charge a battery faster, but not necessarily "better". Older batteries typically charge faster due to capacity loss.

[StevesFishTanks]

I have a 55ah lifepo4 battery in my truck that was used for my solar powered fridge/freezer.  In my opinion lithium iron phosphate is the best battery when you factor weight, and depth of discharge to its rated capacity.

I will likely tear down that project and integrate into my linear piston automatic battery backup. At the least replace the lead acid battery with the lifepo4

I run on wall power and charge/maintain the battery with a charger but it would be fun to have a totally off grid central air pump using the MPPT some panels and the lifepo4. There may be an update to that post in the future!

Edited March 24, 2022 by StevesFishTanks

[CT_]

On 3/24/2022 at 6:19 AM, Greg Stewart said:

I believe the charging capacity/retention is unrelated to the supplied current. But, I could be wrong on this. Higher current would charge a battery faster, but not necessarily "better". Older batteries typically charge faster due to capacity loss.

so then battery health shouldn't be a factor in UPS efficiency when passing through current right?

[Odd Duck]

On 3/23/2022 at 6:37 PM, Greg Stewart said:

Which chart did you find that said only 2 hours? Which model do you actually have?

If your experience is 8 to 9 hours with a 5W draw, 7 or so hours on a 14W draw doesn't seem off. But, battery health will impact that, and get "worse" over time. Your operational runtime will depend entirely on how good your batteries are at the time the power goes out.

Deciding which UPS to get, if you replace the ones you have, will depend on what, exactly, you need to get out of them. UPS units are not really designed to be long term power banks. They are really designed to allow you time to address a power outage before equipment goes down hard.

If you really need long term uptime for power on your fish room, if might be a better expense considering a backup generator with an automatic transfer switch. The larger the UPS you get, the larger the expense you're looking at. And, if you're prone to multi-day outages, you won't get good coverage from a UPS.

I have BX1500M which doesn’t pop up when I search that specific model,on the APC website, but when I search bx1500m, and pic the first option, then go to the run times table, I eventually found my way to this.  Mine is listed right at the bottom, but the 1500’s in the chart are all similar.

https://www.apcguard.com/BR1500MS.asp?utm_source=vgsearch&utm_term=Bx1500m

We have a couple back up generators that will be used for more long term support, but when I work, my shortest shifts are typically 10 hours.  I might sometimes be scheduled for 8, but reality turns out to be 10.  I just got done with a 14 hour overnight that turned into 15.5 hrs and just had to leave without all my charts finished or I would have been too tired to drive home.  Same for hubs.  So I need something that will keep at least some kind of filter running on the tanks for at least 8 hours or so, until we can get home and get the generator running.  We don’t have a good way or place to set up a generator with auto-start.  The gennies are stored either in the garage or in our storage unit, depending.  We do want to get an auto-start, whole house gen, but that will likely wait until next year.  In the mean time, we get by with pawn shop, rebuilt specials that hubs buys cheap and does carb rebuilds on them (usually all they need).

[Greg Stewart]

@Odd Duck I get it. That's a tough situation to manage. The only thing I can really suggest is to spec out new units with the battery capacity to handle the duration, rather than spec'ing them according to the max load. And, then distribute your gear over the units to minimise load as much as possible.

Are there alternative pumps and other kit to handle your tanks that you can get that have lower power ratings (and, work just as well)? You might have to tackle this from both angles. Minimise power demand at the same time as maxmising battery runtime.

You could run tests on the units you have to gauge the actual runtime for your gear. But, keep in mind that every time you run the batteries down to null they'll get "older". And, the longer the span between charges, the more they'll likely lose capacity and be less reliable over time.

It seems to be a no-win situation without having the money or space to throw at it, unfortunately.

[Widgets]

This is anecdotal, but several years ago when I spec'd an APC UPS for my work computer I diligently calculated the load. Using the APC guides, I decided to get a unit that could handle twice the load to allow for growth/aging. When I installed the UPS it showed a load of 45% and a runtime of 35 minutes. After the batteries fully charged, I did the acid test of unplugging the UPS to simulate a power loss. The computer shutdown and the UPS alarmed that it was overloaded. After restarting the UPS still showed a 45% load. I then ran the APC self test which also indicated an overload condition. I contacted APC support to ask what the problem was. Their response was that I needed to buy a bigger UPS!?! I explained that I had already doubled the size that they indicated, and asked to return this for credit towards a larger unit. They offered a 50% trade in on my now used unit.

Since then, I have just ordered the biggest unit that I could budget, and test it to see what it would could handle. Sometimes I was OK, other times I had to ask for more budget to get a second UPS. I have found that the larger enterprise targeted UPS are generally closer to "as advertised" than the consumer targeted UPS.

Note that some of the enterprise UPS allow a secondary battery pack to extend the runtime.

Edited March 25, 2022 by Widgets
Stupid typos

[Odd Duck]

On 3/24/2022 at 9:02 PM, Greg Stewart said:

@Odd Duck I get it. That's a tough situation to manage. The only thing I can really suggest is to spec out new units with the battery capacity to handle the duration, rather than spec'ing them according to the max load. And, then distribute your gear over the units to minimise load as much as possible.

Are there alternative pumps and other kit to handle your tanks that you can get that have lower power ratings (and, work just as well)? You might have to tackle this from both angles. Minimise power demand at the same time as maxmising battery runtime.

You could run tests on the units you have to gauge the actual runtime for your gear. But, keep in mind that every time you run the batteries down to null they'll get "older". And, the longer the span between charges, the more they'll likely lose capacity and be less reliable over time.

It seems to be a no-win situation without having the money or space to throw at it, unfortunately.

I’ll probably get a larger unit just to be as safe as possible.  Even if we can get close to 8 hours, that’s enough to keep the tanks going and reasonably healthy.  The likelihood of the power going out practically the minute we walk out the door is low.  I tend to over-filter, plant heavy, and understock.  Once I finally have the fish room done, I will run a test on a day off to get an idea of how long the UPS will run the linear.

I appreciate all the info and your expertise.  For sure, eventually, we will get the whole house auto-start gen.  We live close enough to afire station we aren’t very likely to be out of power for very long.  We didn’t have anything more than a momentary  flicker in the power during the Texas Snowpocalypse last February when about 4.5 million homes were out of power in Texas.  Some people were out of power for up to 3 weeks even though it was only a few days for most.  After talking with others, we figured out they must have been routing what little power they had to medical and rescue services.  Being just a block or so from a fire station was our saving grace, I’m sure.