diy solar

diy solar

Use a LiFePO4 battery pack inside UPS??

P

Painas

New Member
Joined
Sep 21, 2020
Messages
3
Hi there,

I have a couple of UPS that are completly funcional but they dont have batteries.

Instead of buying Lead acid batteries I was thinking of making a LiFePO4 battery pack and use it instead of the lead acid batteries.

I have done some research but I do not have enough information yet so:

- I know that the UPS runs at 12V;
- From what I have read, it will cutoff as safety at 12.9V (2.15V per cell) and will consider a max charge at 14.1V (2.35V per cell);
- Is it safe to make a battery pack, connect it to a LiFePO4 BMS and put it inside? Will this simple conversion cause any problems in the future (to the battery)?
- Could the charging of the UPS hurt the batteries (from what I've read, both are charged via CC-CV ).

Although I know I wont take advantage of the full potential of the pack there is no problem. From what little I would lose it is still better that the original lead acid battery.

Thanks in advance :)
 
UPS batteries are typically AGM. AGM like to be floated for standby (UPS) applications at about 13.8V. 13.8/4 = 3.45 - above the recommended float voltage for LFP.

If you can verify the UPS floats at or below 3.4V/cell, then I see no issue.

Other considerations:
  • the extremely slow charge rate of most UPS. They may take several hours to replace what they used in 15 minutes. A big battery might take days to charge.
  • The duty cycle of consumer UPS is generally pretty low, i.e., the inverter is not designed to run much longer than the original battery permits. Substantially higher capacity may cause the inverter to fail due to prolonged operation.
 
Hi,

From what I've understood, float is the voltage that is constantly applied to the battery in order to maintain it correct?

Are LiFePO4 compatible with that or is it like "Yeah they can do that but dont expect them to last long"?

Considering that it would cause no harm to the batteries, if the UPS outputs to the battery a maximum of 13.8V, there should be no problem, correct?
 
Painas said:
Hi,

(1) From what I've understood, float is the voltage that is constantly applied to the battery in order to maintain it correct?

(2) Are LiFePO4 compatible with that or is it like "Yeah they can do that but dont expect them to last long"?

(3) Considering that it would cause no harm to the batteries, if the UPS outputs to the battery a maximum of 13.8V, there should be no problem, correct?
Click to expand...

(1) Yes.

(2) If the float isn't too high, yes. If the float is above 3.4V/cell - bad for LFP.

(3) The UPS may charge to higher than 13.8V, which is fine (14.7V max). If it floats over 13.6V, it's not good for LFP.

In short:

If a UPS charges to less than 14.7V and it floats at 13.6V or less, LFP should be fine.
 
snoobler said:
(1) Yes.

(2) If the float isn't too high, yes. If the float is above 3.4V/cell - bad for LFP.

(3) The UPS may charge to higher than 13.8V, which is fine (14.7V max). If it floats over 13.6V, it's not good for LFP.

In short:

If a UPS charges to less than 14.7V and it floats at 13.6V or less, LFP should be fine.
Click to expand...

You sir, deserve a medal. I cant thank you enough for your time to help me understand this :D
 
snoobler said:
(1) Yes.

(2) If the float isn't too high, yes. If the float is above 3.4V/cell - bad for LFP.

(3) The UPS may charge to higher than 13.8V, which is fine (14.7V max). If it floats over 13.6V, it's not good for LFP.

In short:

If a UPS charges to less than 14.7V and it floats at 13.6V or less, LFP should be fine.
Click to expand...
Also planning to do battery conversion on a ups.

But how to check the charging voltage?
And how to check the floating voltage?
 
If it isn't a smart ups of some type, see other threads about APC upses on the forum, you will have to check the battery voltage with a multimeter. Some meters have a peak hold feature. If yours does, hook the probes to the battery terminals and start the peak hold function. It'll record the maximum voltage the battery reaches. Some meters have a min/max feature too and that would work. Same process, connect meter leads to battery terminals, start min/max and observe voltages at a later time.

You can then determine if it does an absorption stage charge and what the float voltage is once that is completed.
 
This is the info / thread I was looking for!

I was REALLY hoping I could drop-in / plug-and-play a LiFePO4 12v battery replacement in exchange from my 12v SLA / lead acid battery in my basic UPS, but I guess I need to measure the "trickle charge" and cross my fingers it isn't above 14.7 V!
 
My UPS is CyberPower 1000PFCLCD. I built an 18Ah LiFePO4 battery. The battery is made of 26650 cells in 4s5p configuration. It currently extends UPS runtime. My UPS float voltage is 13.45V. Have another UPS 850PFCLCD with float voltage of 13.5V.

As UPS use SLA, you really don't need to worry about equalization as that is typically only for FLA. The adsorption voltage on UPS models is unlikely to exceed 14.4V.
 

Attachments

  • 20210205_103436.jpg
    20210205_103436.jpg
    67.6 KB · Views: 48
Positronium said:
My UPS is CyberPower 1000PFCLCD. I built an 18Ah LiFePO4 battery. The battery is made of 26650 cells in 4s5p configuration. It currently extends UPS runtime. My UPS float voltage is 13.45V. Have another UPS 850PFCLCD with float voltage of 13.5V.

As UPS use SLA, you really don't need to worry about equalization as that is typically only for FLA. The adsorption voltage on UPS models is unlikely to exceed 14.4V.
Click to expand...
Is it possible to grab an off the shelf 50 to 100ah lifepo for a 1000va UPS and assume there is enough absorption voltage and the float is in that range you specified? I have a newer APC that just went out of "warranty" range and we're having more and more power fluctuations so I want to be ready by converting over to lifepo. I have enough knowledge to be dangerous and i'm most familiar with my overlanding rig with dc to dc chargers and inverter.
 
snoobler said:
(1) Yes.

(2) If the float isn't too high, yes. If the float is above 3.4V/cell - bad for LFP.

(3) The UPS may charge to higher than 13.8V, which is fine (14.7V max). If it floats over 13.6V, it's not good for LFP.

In short:

If a UPS charges to less than 14.7V and it floats at 13.6V or less, LFP should be fine.
Click to expand...
Most lifepo4 battery banks will have a BMS module that is supposed to prevent over charging, over discharging, over and under temp shutoff, cell balancing and so on. Shouldn't this allow us to use a inverter/charger with the lifepo4 batteries? I have an older tripplite aps1250 that has a 13.5 float voltage, not sure what the bulk charge voltage is.
 
ToddBailey said:
I have an older tripplite aps1250 that has a 13.5 float voltage, not sure what the bulk charge voltage is.
Click to expand...
It is Float that can shorten the life of Lithium batteries if it can not be turned off. A voltage 3.375 volts per cell is high enough above resting voltage of 3.3 unless there is a constant load on the batteries. That is something most BMSs are not designed to prevent. Some UPSs power the loads so that there is no switchover delay and that load may be enough to prevent the cells from getting stressed.
 
Hi @Ampster

This weekend I replaced my AirPump UPS for Koi with LiFePO4, my concern about the use of LiFePO4 as you mentioned is the float voltage, at 14.5volts (rest voltage/nominal at 13.8) it will damage the cells? The BMS will not stop it or the UPS circuit trickle charge the LiFePO4 (not applicable to LiFePO4).

I just ordered a SmartBMS and set the max voltage to 3.55 to prevent such issue. I am also trying to look for some over charge disconnect module, still no luck so the SmartBMS is my last choice, your thoughts on this?



Regards,
 
We have 6 (six) 350W ish UPSs of various flavours (but all el-cheapo) running PCs, TVs and the network kit.

All have LiFePO4 packs made from 4S x 32650 cells with the cheapest 40A BMSs I could find just dropped in, no modifications.

All work just fine and the oldest has been doing its thing for 18 months / 2 years or so.

In reality they just have to bridge over the gap whilst the genset starts but they do that unerringly. The cells are cheap so if they die so be it, they are still cheaper than the equivalent gel packs.
 
Last edited:
Update: Since the UPS is made for SLA, it maintains the battery to 14.4V (BMS Cut off). Since I don't want that kind of setup what I did is I added a Relay that would close if the power if off (AC), but this is not suitable for computers. Mine is forgiving because it is the built-in UPS of the Halea CP60 Airpump which only consume 12-18Watts (variable).
I rested the LiFePO4 to 13.5Volts, currently uses JBD BMS.
I am not sure if keeping the LiFePO4 in charging voltage (14.4-6) is healthy for long term use.
 
So I built a LiFePO4 12v pack for my APC UPS and all I can say is that it only seems to charge up to 41% (based on the BMS readings), doesn't charge past 13.5v... so the benefits of the LiFePO4 don't get realized. I'll likely just put another SLA battery in and use the LiFePO4 elsewhere.
 
3.375v per cell should be near full, and safe to float there indefinitely from what I've researched. I'm converting a 24v APC Smart UPS to 8s LiFePO4 and it floats at 27.2v or 3.4v/cell which could be borderline depending on who you listen to.

Are you sure your BMS is correct? At 41% if cells are at 3.35v, it sounds miscalibrated to me.
 
You must log in or register to reply here.

Similar threads

P
Replacing a 120v 3000VA UPS
Replies
12
Views
425
robbob2112
robbob2112
Azurisz
UPS - Replacing lead-acid with LifePo4
Replies
4
Views
605
bolhuijo
bolhuijo
R
APC Smart UPS 3000RM NET can it use LiFePo4 batteries
Replies
3
Views
413
patrickcasavant
P
S
Anyone ever use cells in parallel in a battery pack?
Replies
12
Views
494
pollenface
pollenface
A
UPS alternatives?
2 3
Replies
49
Views
2K
archjeb
A
Back
Top