LFP disconnect/connect with chargers/loads

[rgleason]

I would like to have good sound reasoning for these questions.

Configuration

1. Alternator: 160a LFP charge parameters, de-rated to 100a max
2. Solar PV: Small MPPT chargers
3. Charger/Inverter: Small 120v, Portable connected to charge bus when needed.
4. AGM/FLA: 70a min, provides high voltage disconnect Alternator protection via BMS/LFP operation of external relay.
5. LFP: 200a
6. BMS:
   1. Charge Disconnect: Control of an external relay which disconnects Alt/AGM/FLA from LFP on a pre-alarm condition.
   2. Charge Disconnect: Control of MPPT Solar chargers to disconnect or switch off on an overcharge pre-alarm condition.
   3. Discharge Disconnect: Control of an external relay which disconnects all Loads from LFP on a pre-alarm condtion.

Operation

1. During Alternator charging no issue with discharge of LFP to AGM/FLA. (Requires current limiting LFP regulator with alt temp sensor, or BMS must have current limiting to prevent Alt burnout.)
2. During Solar PV charging no issue with discharge of LFP to AGM/FLA.
3. Pre-alarm or Charge Disconnect LFP disconnected from Alt/Solar/AGM.
4. House loads draw LFP down when Alt/Solar/AGM/FLA are disconnected.
5. Alternator/Solar can continue to charge AGM/FLA
6. When the LFP cell voltage drops to user set (3.1-3.7v cell) and the BMS closes charge relay.
7. The LFP (between 3.2-3.7v) and AGM/LFP (12.8v at rest, or lower) are then connected and it is possible for the LFP with higher potential, to charge the AGM/FLA.
8. EG: Victron Values
   1. Cell undervoltage Pre-alarm (3.10v) 12.4v (Disconnect all Loads)
   2. Cell undervoltage Disconnect (2.8v) 11.2v EOD (Disconnected)
   3. Alternator Charge Absorption (3.5-3.6v) 14-14.4v
   4. Alternator Float (3.375v) 13.5v

Questions:

How much power is lost per hour due to Item #7 with the LFP charging the FLA?
13.2v-12.8=.4v V=IR
.4v=I (400mohm)=.4/400,000,000ohms=0.000000001 amps

Is this about right?
If so, it isn't enough to worry about, is it?

When a load is suddenly disconnected from the LFP, I understand it can damage any FETs in the BMS. How can this be prevented/protected from?

[T1 Terry]

Does the solar and alternator charge both the LFP and AGM/FLA battery at the same time and the BMS just isolates the LFP when a high cell voltage is sensed?

If this is the case, no damage to the mosfets will occur because there is still a current path, no voltage spike.
As far as item 7, if the AGM/FLA batteries are as new or highly sulphated, the 12.8v will be maintained with very little current flow required. Once they age or the sulphation is cleared, a constant current flow will be required to maintain the 12.8v. The voltage differential will increase because the rested AGM/FLA voltage will gradually drop.

In our experience with motorhomes where a permanent link between the lithium house battery and the AGM/FLA start battery, the drain gradually increases over time until it drains a 200Ah lithium house battery over night. Clearly there would have been other loads from the house system, but in general the observed effect was a drop from 60% SOC in the morning to 30% in the morning (over time) to the system sounding the 20% SOC alarm in the early hrs and a LVD event if the motorhome was unattended.
Removal of the fixed link saw the LFP house battery recover and hold better than the 60% SOC previously recorded first thing in the morning when the system seemed to be working well.
As long as the start battery was charging during the day, the motor restarted the next morning for a few mths, then starting getting harder and ....... In the end, a switchable link to start the motor with assistance from the house battery saw them through until they reached some where they could cost effectively swap out the start batteries for new ones. It can be 3 times the price if not more for batteries in the less populated areas than the main cities.

T1 Terry

[Adelie]

Originally Posted by rgleason

I would like to have good sound reasoning for these questions.
Only you can explain why you have questions. You just need to look inside yourself.

Configuration

1. Alternator: 160a LFP charge parameters, de-rated to 100a max
2. Solar PV: Small MPPT chargers
3. Charger/Inverter: Small 120v, Portable connected to charge bus when needed.
4. AGM/FLA: 70a min, provides high voltage disconnect Alternator protection via BMS/LFP operation of external relay.
5. LFP: 200a
6. BMS:
   1. Charge Disconnect: Control of an external relay which disconnects Alt/AGM/FLA from LFP on a pre-alarm condition.
   2. Charge Disconnect: Control of MPPT Solar chargers to disconnect or switch off on an overcharge pre-alarm condition.
   3. Discharge Disconnect: Control of an external relay which disconnects all Loads from LFP on a pre-alarm condtion.

Operation

1. During Alternator charging no issue with discharge of LFP to AGM/FLA. (Requires current limiting LFP regulator with alt temp sensor, or BMS must have current limiting to prevent Alt burnout.)
2. During Solar PV charging no issue with discharge of LFP to AGM/FLA.
3. Pre-alarm or Charge Disconnect LFP disconnected from Alt/Solar/AGM.
4. House loads draw LFP down when Alt/Solar/AGM/FLA are disconnected.
5. Alternator/Solar can continue to charge AGM/FLA
6. When the LFP cell voltage drops to user set (3.1-3.7v cell) and the BMS closes charge relay.
7. The LFP (between 3.2-3.7v) and AGM/LFP (12.8v at rest, or lower) are then connected and it is possible for the LFP with higher potential, to charge the AGM/FLA.
8. EG: Victron Values
   1. Cell undervoltage Pre-alarm (3.10v) 12.4v (Disconnect all Loads)
   2. Cell undervoltage Disconnect (2.8v) 11.2v EOD (Disconnected)
   3. Alternator Charge Absorption (3.5-3.6v) 14-14.4v
   4. Alternator Float (3.375v) 13.5v

Questions:

A. How much power is lost per hour due to Item #7 with the LFP charging the FLA?
13.2v-12.8=.4v V=IR
.4v=I (400mohm)=.4/400,000,000ohms=0.000000001 amps

B. Is this about right?
C. If so, it isn't enough to worry about, is it?

D. When a load is suddenly disconnected from the LFP, I understand it can damage any FETs in the BMS. How can this be prevented/protected from?

A. You can't calculate that unless you put a shunt between the 2 and measure the current. Current and voltage will be a changing values as the LA batteries charges so the power lost will also be a variable. But I don't think that's really what you want to know, I think you want to know what this costs you in Round Trip Efficiency (RTE), or how much do you get out compared to how much you put in? I just found a really good source for battery life and efficiency. The RTE for LFPs was 87-90%(88% ave.) cycling 10-100% for most batteries tested. The one FLA tested was 87% cycling 50-100%. Let's say 85% for the FLA cuz N=1 isn't very good statistically. 88% x 85% = 75%. Since not all of the charge discharge cycle involves one battery pushing the other, it's probably somewhat better than that, call it 80%. So for every 120Ahr put in you will probably get 100Ahr back out if the FL is drawing down to 50%. If it's not cycling down that far, it's RTE is lower. Hard to say with the lower DoD.

B. In terms of the amps lost, I doubt it, 400M-Ohms seems rather high. If the battery was that resistive you couldn't charge it.
C. If that is really your amps lost which implies a very small loss of power then you're fine. But I don't think so.

D. I don't know about loads being disconnected from LFPs being bad for the FETs but I do know that large inverters (4kW & up) will have a huge current draw (1000A give or take) when you connect them to any battery, EVEN IF THE INVERTEER IS OFF. This is because there are big capacitors and they suck power like the dickins when connected. The draw is only for milliseconds but it cast a habit of tripping the overcurrent protection in the BMS.
BattleBorn has a surge limiter just for this case. https://battlebornbatteries.com/prod...surge-limiter/
On the demand side they have an soft starter too. I think this is more to protect motors but maybe it protects the FETs too. https://battlebornbatteries.com/prod...-soft-starter/

[rgleason]

Two great answers/observations. Thank you both.

@TI Terry - The answer to your first question is "Yes". Very practical explanation. Perhaps when we leave the boat, we should make sure the LFP is not fully charged and in a good "storage" state and simply manually disconnect from the AGM/FLA leaving the AGM/FLA battery to be more fully charged by the solar PV and isolating the LFP. Then when we get back on the boat reconnect the LFP.

@Adelie - Thinking about this problem as RTE Round Trip Efficiency is interesting. The RTE declines as the AGM/FLA becomes sulfated (see TI Terry's description) because more LFP power is used. With a good AGM/FLA battery, it appears the RTE is good and the power used to charge the AGM/FLA from SOC 80-100% is not very high and RTE is good. With a sulfated battery this situation gets worse over time.

Thanks very much..

Rick wrote:

How much power is lost per hour due to Item #7 with the LFP charging the FLA?
13.2v-12.8=.4v V=IR
.4v=I (400mohm)=.4/400,000,000ohms=0.000000001 amps

Is this about right?
If so, it isn't enough to worry about, is it?

Adelie wrote:

B. In terms of the amps lost, I doubt it, 400M-Ohms seems rather high. If the battery was that resistive you couldn't charge it.
C. If that is really your amps lost which implies a very small loss of power then you're fine. But I don't think so.

I am confused. You think 0.000000001 amps is too much? OR do you think my use of 400mohm in the formula is too much?
Assuming a new AGM/FLA battery, how would you calculate this OR why is it not possible (Use a shunt)? -I just took the worst case potential difference divided by 400mohms (which is also worst case I think).
Curious: Is the current surge limiter another capacitor with some resistance?

[Adelie]

I wasn’t even thinking of sulfation.
In LA batts the RTE cycling 50-75% is really good, 95% ish.
75-100% is pretty poor, it gets worse the closer to 100%SoC you get. At 100% your RTE is 0%, the energy you are putting is heating the battery and electrolyzing water, energy you can never recover.
The net RTE 50-100% SoC is 87%.

The power being low as you near 100% doesn’t change that the efficiency gets really bad the closer you get.

I expect that sulfation will decrease overall efficiency. So you should equalize monthly to stave that off.

Regarding the amps and ohms question. The internal resistance of the battery is a moving target.
At 50% it is near zero.
At 75% it’s probably around 2ohms.
As you approach 100% it gets really high.
The following link has a chart on pg.13 but no numbers, just qualitative. https://idm-instrumentos.es/wp-conte...-batteries.pdf

You can’t figure it out analytically. You can measure it by putting a resister across the terminals and measuring voltage and amperage. You know the resistance of the resister. Knowing the V & I you can calculate actual resistance. The difference between actual resistance across the terminals and resistance of the resistor is internal resistance.

You could also measure the heat rise for a given amount of current and voltage in a limited time. Then divide that energy from the total energy input to figure how much is going into heat. A really indirect way to do it but it would work.

My understanding is that internal resistance is on the order of milliOhms near 50% and rising into the Ohms near 100%.

I just realized you are comparing the resting voltages of the two batteries. I don’t see you getting any info about dynamic performance from that. As soon as you connect the batteries they will be at the same voltage current will flow from one to the other and resting voltage on each will change as SoC changes.

You are trying to reason out what the batteries are doing from first principles. It can’t be done. The system is too complex, even understanding one battery is very complex, 2 operating together, can’t be done. You can test it though.

If you want to know the amps between the batteries at various states of charge you have to measure it with a shunt. Current will vary depending on SoC of both batteries and their states of health.
If you want to know internal resistance, you have to stop discharge and measure voltage and amperage across a known resistor across a single battery’s terminals.

If you want to know how much efficient the system is it will depend on state of charge and state of health of both batteries.

The closer you are to full with an FLA the less efficient. The more sulfation the less efficient.



I don’t know what the surge limiter is, a big inductor maybe. I’ll look into it.

[rgleason]

Adelie, Thanks that is very helpful. You are right, when the LFP and AGM/FLA banks are connected (without being charged by the Alternator), then the LFP are basically continuing to charge the AGM/FLA, albeit at a slow rate, because presumably the start bank stays close to 100% SOC.

Assuming these are new AGM/FLA, so there is no sulfation, you've explained and page 13 of your link shows that the resistance goes quite high as you get to 100% SOC.

Since the batteries are joined in this state, the voltages are quite close so the voltage differential will be small. This means that V-small = I x R-high so I=V-small/R-high. Therefore current will be very small and the amps used to charge the AGM/LFP from the LFP will be quite low.

As power is used the resistance of the AGM/FLA will be reduced and more current will flow from the LFP. I understand that it is a dynamic system.

It would be helpful to have some real world examples, measured by someone.

[Adelie]

Are you really interested in the amps or are you interested in the efficiency?

[rgleason]

I think I understand amps so I headed for that. I think it is very helpful to think about Efficiency (it how much heat is created) during the moving around of electrons from one battery to another. The efficiency will be lower when the AGM/FLA is closer to 100%SOC while connected to LFP, but the current is very low, so the cost will not be high.


Surge Suppressors use MOV
SURGE PROTECTION CIRCUIT PRINCIPLE AND DESIGN | ElectronicsBeliever

Transient suppression devices can take on many forms from arc contacts, to filters, to solid state semiconductor devices. Discrete semiconductor transient suppression devices such as the Metal-oxide Varistor, or MOV, are by far the most common as they are available in a variety of energy absorbing and voltage ratings making it possible to exercise tight control over unwanted and potentially destructive transients or over voltage spikes.

[Adelie]

Yeah, but even though the amps get small, efficiency is so bad the loss is still significant.

[hzcruiser]

Hello Rick,

I must've missed this thread last year...

Have you sorted out the LFP disconnect issue? Your math about losses seems a bit off there, maybe you're mixing up milli Ohm and mega Ohm ? Where are those 400 mohm coming from? The internal batt resistance maybe? That's typically in the 10 mOhm range, as in milli Ohm. Mega Ohm would be MOhm.

In your example, the cross current from LFP to LA with 0.4V difference between them and 400 milliOhm resistance of the batt would be: 0.4 V / 400 * 10^-3 Ohm = 1 A

And that's not "losses", that's charging the LA batts.

In the real world the current would be much lower as the wire and contact resistances also add up when you're dealing with such low resistances.

[rgleason]

Originally Posted by hzcruiser

Hello Rick,

I must've missed this thread last year...

Have you sorted out the LFP disconnect issue? Your math about losses seems a bit off there, maybe you're mixing up milli Ohm and mega Ohm ? Where are those 400 mohm coming from? The internal batt resistance maybe? That's typically in the 10 mOhm range, as in milli Ohm. Mega Ohm would be MOhm.

In your example, the cross current from LFP to LA with 0.4V difference between them and 400 milliOhm resistance of the batt would be: 0.4 V / 400 * 10^-3 Ohm = 1 A

And that's not "losses", that's charging the LA batts.

In the real world the current would be much lower as the wire and contact resistances also add up when you're dealing with such low resistances.

HZCruiser is this the post you are referring to? I believe so. https://www.cruisersforum.com/forums...ml#post3320076
I was trying to figure out the flow of current and how the two chemistries would work.

Thank you for clarifying the magnitude. I realize now that Adelie also pointed out this out sometime ago.

I put my LiFePo efforts on the back burner for the last couple of years. Still interested in doing LFP and I hope there is more knowledge and experience available than several years ago and perhaps better approaches.

[CarlF]

My cat has a house bank of 5 300AH Kilovolt batteries (wired parallel on equal length cables) connected through VSRs to two 80AH FLA start batteries. Because the Kilovaults stay above 13v the VSR is always closed even at night with the engines off. The start batteries and Kilovaults are always the same voltage. The alternators are wired to the start batteries but since they are always fully charged the current flows to the house through the VSR.

The Kilovolts each have bluetooth communication that shows voltage and current for the battery. These don't show any amps leaving the Kilovolts that is not accounted for by loads other than the start battery. Presumably, any amps used starting the engine is replaced within a few minutes and the start batteries are then essentially on a float charge from the Kilovolts drawing so few amps that it is not visible at the house bank.

The only downside is that all charge sources are set to the KiloVault spec (14.1 absorption, 13.4 float). This is less than ideal for the start batteries which is why I went with FLA instead of AGM as they aren't as picky about charge voltage. I'm going to be interested to see if the start batteries sulfate. I doubt it since they never even get to 12.8V. But if they do, they are cheap and easy to replace.

I also considered connecting a DC-DC charger to charge the start batteries from the house instead of a VSR. This is becoming very common on cats. The DC-DC chargers are always on and I haven't heard anyone complaining that the start battery charging is draining their house bank.

[hzcruiser]

Originally Posted by rgleason

HZCruiser is this the post you are referring to? [...]

I put my LiFePo efforts on the back burner for the last couple of years. Still interested in doing LFP and I hope there is more knowledge and experience available than several years ago and perhaps better approaches.

Yes, that's the one.

I'm a bit surprised but LFPs have come down in prices in the last year, at least looking at aliexpress prices for cells.

[hzcruiser]

Originally Posted by CarlF

My cat has a house bank of 5 300AH Kilovolt batteries (wired parallel on equal length cables) connected through VSRs to two 80AH FLA start batteries. Because the Kilovaults stay above 13v the VSR is always closed even at night with the engines off. The start batteries and Kilovaults are always the same voltage. The alternators are wired to the start batteries but since they are always fully charged the current flows to the house through the VSR.

So you only really need the VSRs to prevent accidental draining of the starter batts?

[...] I'm going to be interested to see if the start batteries sulfate. I doubt it since they never even get to 12.8V. But if they do, they are cheap and easy to replace.

I also considered connecting a DC-DC charger to charge the start batteries from the house instead of a VSR. This is becoming very common on cats. The DC-DC chargers are always on and I haven't heard anyone complaining that the start battery charging is draining their house bank.

And then move the alternator to the LFP bank?

[CarlF]

I now realize the VSR's are unnecessary. They were already installed when the house bank was FLA so it was easier to leave them. But the 1500AH lithium bank is never going to go dead - and has alarms if it ever falls below 20%. And if it somehow did die, the genset has its own isolated start battery so I could use that to get things going.

WIth the DC to DC chargers you connect the alternator directly to the LFP house bank. This is a better way to use DC to DC chargers than charging the house bank from the start battery since a typical DC-DC charger can only charge 30 amps. So most of a 120 amp alternator is wasted if the DC-DC charger is charging the LFP bank. But 30 amps is plenty to charge a start battery. The DC-DC charger is also able to absorb a spike from a BMS shutdown into the start battery protecting the alternator (although still best to have belt-and-suspenders with an alternator protection device too.