Mixed AGM / LIFEPO Charging

I have two 8D batteries for starting and six 8Ds for the house. I'm happy to say that they're all in need of replacement..... ugh!

It's now cheaper for me to replace the house 8Ds with LIFEPOs than AGMs. 12V 280Ah Eco Worthy units are now just $500 each as opposed to the best price I could find for 250 Ah AGMs at $568. Plus I get all that more usable Ah. Using 50% state of discharge for the AGMs I would have 750 usable Ah from all 6, and more than 950 Ah from just the four LiFePo. I will replace the two 8D start batteries with 4D AGMs because the cost of LiFePo start capable batteries is still too high. The 4D units should spin the Detroit 6-71s easily, and I'm not interested in humping any more 160lb batteries aboard at my age....

I have twin Xantrex SW3012 inverter/chargers that feed the batteries through a total of 3 Blue Seas ACRs (auto-charging relays). They connect port / stbd / generator start batteries to the charging output, and then of course parallel the charging connection to the house bank when the starting battery is full.
With these 'mixed' battery types I plan to use a modified AGM charging profile so that: 1) I protect the AGM batteries, and 2) I get a reasonable state of charge without damaging the LiFePo units.

The Xantrex 3-stage charge cycle for AGMs is Bulk = constant current, 14.3V; Absorb = constant voltage 14.1V; and Float = 13.4V. I can create a custom charge profile to set the voltage and duration of the Absorb phase, among other things. The Eco Worthy specs call for a charging voltage of 14.6v. The max charging current of the Xantrex is 120A which is well within the 160A max battery specification.

My main questions are:
- Will the lower charging voltage of 14.3V cause any harm to the LiFePo batteries? I assume not, and that I will just not ever reach 100% SOC.
- If I shorten the Absorb phase timer will that cause an issue for the AGMs? If I leave it alone will that affect the LiFePos?
- Does the 13.4V float charge harm the LiFePos at all? I think that they are not supposed to be kept at 100% SOC anyway?

Thanks in advance for any replies.

markpj23 said:

I have two 8D batteries for starting and six 8Ds for the house. I'm happy to say that they're all in need of replacement..... ugh!

It's now cheaper for me to replace the house 8Ds with LIFEPOs than AGMs. 12V 280Ah Eco Worthy units are now just $500 each as opposed to the best price I could find for 250 Ah AGMs at $568. Plus I get all that more usable Ah. Using 50% state of discharge for the AGMs I would have 750 usable Ah from all 6, and more than 950 Ah from just the four LiFePo. I will replace the two 8D start batteries with 4D AGMs because the cost of LiFePo start capable batteries is still too high. The 4D units should spin the Detroit 6-71s easily, and I'm not interested in humping any more 160lb batteries aboard at my age....

I have twin Xantrex SW3012 inverter/chargers that feed the batteries through a total of 3 Blue Seas ACRs (auto-charging relays). They connect port / stbd / generator start batteries to the charging output, and then of course parallel the charging connection to the house bank when the starting battery is full.
With these 'mixed' battery types I plan to use a modified AGM charging profile so that: 1) I protect the AGM batteries, and 2) I get a reasonable state of charge without damaging the LiFePo units.

The Xantrex 3-stage charge cycle for AGMs is Bulk = constant current, 14.3V; Absorb = constant voltage 14.1V; and Float = 13.4V. I can create a custom charge profile to set the voltage and duration of the Absorb phase, among other things. The Eco Worthy specs call for a charging voltage of 14.6v. The max charging current of the Xantrex is 120A which is well within the 160A max battery specification.

Click to expand...

markpj23 said:

My main questions are:
- Will the lower charging voltage of 14.3V cause any harm to the LiFePo batteries?

Click to expand...

No.

markpj23 said:

I assume not, and that I will just not ever reach 100% SOC.

Click to expand...

No. It just takes longer, but you can get to 100% SoC in the 13.6-13.8V range with longer absorption time. 14.3V will only require a slightly longer absorption time - maybe 30-60 minutes to be sure.

markpj23 said:

- If I shorten the Absorb phase timer will that cause an issue for the AGMs?

Click to expand...

Yes. They need long absorption times.

markpj23 said:

If I leave it alone will that affect the LiFePos?

Click to expand...

Again, charging to lower peak voltage means you need longer absorption times, but not much.

The longer absorption times will not be the best thing for your LFP batteries, but they shouldn't be too harmful.

markpj23 said:

- Does the 13.4V float charge harm the LiFePos at all?

Click to expand...

No. 13.5V is a little better, but 13.4V will just allow a little discharge.

markpj23 said:

I think that they are not supposed to be kept at 100% SOC anyway?

Click to expand...

That's the official story, but LFP is far less sensitive to being stored at 100% than other Lithium chemistries. The thing you want to avoid is an excessively high float voltage that actually over-charges the LFP. 13.4-13.5V is fine.

sunshine_eggo said:

No.



No. It just takes longer, but you can get to 100% SoC in the 13.6-13.8V range with longer absorption time. 14.3V will only require a slightly longer absorption time - maybe 30-60 minutes to be sure.



Yes. They need long absorption times.



Again, charging to lower peak voltage means you need longer absorption times, but not much.

The longer absorption times will not be the best thing for your LFP batteries, but they shouldn't be too harmful.



No. 13.5V is a little better, but 13.4V will just allow a little discharge.



That's the official story, but LFP is far less sensitive to being stored at 100% than other Lithium chemistries. The thing you want to avoid is an excessively high float voltage that actually over-charges the LFP. 13.4-13.5V is fine.

Click to expand...

Thanks! I can adjust the float voltage up to 13.5V. I assume the AGMs will be OK with that slight increase as well.

markpj23 said:

Thanks! I can adjust the float voltage up to 13.5V. I assume the AGMs will be OK with that slight increase as well.

Click to expand...

Most AGMS I've seen are 13.6-13.8V, so I wouldn't expect so.

Reading further into the Xantrex manual, it shows that the absorb phase last until either: a) the charging current drops to 2% of the battery capacity; or b) the absorb timer lapses (default is 3 hours). Since the LFP batteries will accept all the current that the charger can supply until they are full, I don't think the absorb phase will ever see current reduced to the 2% cutoff value. That means that the Absorb Timeout is the only controlling factor.
Is this a big deal? I expect the AGM batteries can only accept 'x' amount of current as they approach full charge even if the charging bus is still cranking out what the LFPs can take. Or will the BMS also reduce charging current as the LFPs near full charge?

I guess the question is will the AGMs be hurt by constant current during the absorb phase? Also, do I reduce the Absorb Timeout value? (I've seen that recommendation in a Lithionics charging guide, although the context was an all LFP scenario).

Don't want to 'bake' the AGMs while the charger satisfies the LFP bank.....

markpj23 said:

Since the LFP batteries will accept all the current that the charger can supply until they are full, I don't think the absorb phase will ever see current reduced to the 2% cutoff value.

Click to expand...

This reasoning is incorrect. When full, LFP completely stops taking current simply because there is no longer a voltage difference between the LFP and the source.

markpj23 said:

Or will the BMS also reduce charging current as the LFPs near full charge?

Click to expand...

In almost all cases, the BMS of a LFP battery is an on/off switch. It regulates nothing. If a limit is reached, it terminates the action, i.e., battery/cell voltage too high? no more charging. period. charge current too high? No charging. Voltage too low or discharge current too high, no more discharge.

The BMS is a safety feature. Its existence is only evident when something goes wrong by going out of limits. It is otherwise invisible.

markpj23 said:

Don't want to 'bake' the AGMs while the charger satisfies the LFP bank.....

Click to expand...

The opposite will happen. You will over-charge the LFP bank while undercharging the AGM, but this will only happen when recovering from deep discharges.

LFP takes on the vast majority of its charge below 13.6-13.8V and rapidly fills at higher voltages. AGM tend to take on another 20% once absorption is reached, and it is also very inefficient in that range. If you use 100Ah of AGM, you need to replace it with about 115Ah.

When you have those two banks operating together, the AGMs will do very little most of the time. AGM resting at 12.8-13.0V is fully/nearly fully charged. At that resting voltage level, your LFP will be at a low-ish state of charge - 20-30% (spitballing), so until you've mostly consumed your LFP bank, the AGM will just sit there as backup. Conceptually:



If you're lightly cycling your bank (not discharging the LFP below 30-40% of their capacity), the AGM stay nearly fully charged. Large loads that pull voltage down can draw from them, but for the most part, you will just be charging the LFP and will need a short absorption time at ~14.3V (30-60 minutes tops).

The real issue comes down to if you're "deeply" discharging the entire bank to where you are eating into the AGM capacity. At that time, you'll need to absorb longer to get the AGM fully/nearly-fully charged... likely the 3 hours.

sunshine_eggo said:

This reasoning is incorrect. When full, LFP completely stops taking current simply because there is no longer a voltage difference between the LFP and the source.




In almost all cases, the BMS of a LFP battery is an on/off switch. It regulates nothing. If a limit is reached, it terminates the action, i.e., battery/cell voltage too high? no more charging. period. charge current too high? No charging. Voltage too low or discharge current too high, no more discharge.

The BMS is a safety feature. Its existence is only evident when something goes wrong by going out of limits. It is otherwise invisible.



The opposite will happen. You will over-charge the LFP bank while undercharging the AGM, but this will only happen when recovering from deep discharges.

LFP takes on the vast majority of its charge below 13.6-13.8V and rapidly fills at higher voltages. AGM tend to take on another 20% once absorption is reached, and it is also very inefficient in that range. If you use 100Ah of AGM, you need to replace it with about 115Ah.

When you have those two banks operating together, the AGMs will do very little most of the time. AGM resting at 12.8-13.0V is fully/nearly fully charged. At that resting voltage level, your LFP will be at a low-ish state of charge - 20-30% (spitballing), so until you've mostly consumed your LFP bank, the AGM will just sit there as backup. Conceptually:

View attachment 206031

If you're lightly cycling your bank (not discharging the LFP below 30-40% of their capacity), the AGM stay nearly fully charged. Large loads that pull voltage down can draw from them, but for the most part, you will just be charging the LFP and will need a short absorption time at ~14.3V (30-60 minutes tops).

The real issue comes down to if you're "deeply" discharging the entire bank to where you are eating into the AGM capacity. At that time, you'll need to absorb longer to get the AGM fully/nearly-fully charged... likely the 3 hours.

Click to expand...

OK makes more sense now - thank you.