Solar chargers & BMS cutoffs

[krid2000]

There's a lot of discussion here around BMS cutoffs and alternators. One particular concern is that after a BMS cutoff that hasn't been announced, the alternator might cause a very short (milliseconds) high voltage spike that can not only destroy the alternator, but also the other electronic systems.

I have 610Wp solar installed and frequently see > 45A charge current in my 12V system, about par with the maximum I see from my alternator.

Why are we not concerned about BMS cutoffs and then solar chargers causing similar voltage spikes? Does it not happen?

Thanks,
Dirk

[markcouz]

If you set your MPPT to a reasonable voltage (I use 13.8v for both bulk and float) then you will never reach an overcharge or imbalance situation that would trigger the BMS.

[donradcliffe]

Because the MPPT controller does not spinning parts.

[tanglewood]

Because there isn't a large magnetic field built up with nowhere to dump the stored energy when it collapses.

[wholybee]

The coils in the alternator are large inductors. Inductors resist changes in current. That is, if you turn off the current, as the magnetic field collapses, the inductor will continue to provide current for a brief time.

An inductor is the opposite of a capacitor. A capacitor will allow AC to flow through it, but not DC. A capacitor allows constantly changing currents but blocks a continuous current. An inductor blocks AC, but allows DC. So, when the current stops, the inductor holds the DC current until the field has collapsed.

With nowhere for that current to go, it builds voltage in what we call the load dump. The higher the charging current at the time of the load dump, the higher the resulting voltage will be.

Other charge sources (MPPT, shore chargers, etc.) do not operate with a large inductor as the power source. So, load dumps are not possible.

[noelex 77]

This is a concern.

If the BMS activates, disconnecting the battery, the voltage will be poorly controlled. Without a battery in the circuit, high voltage spikes will occur if charge sources such as solar are active. If electronics are connected to this circuit there is a risk systems will be damaged.

This risk can be eliminated by arranging the wiring of sensitive electronics so that they disconnect when the BMS is activated, but this is not always practical.

[HeywoodJ]

This is a common design flaw of both battery systems and BMSs. The BMS should have separate actions for HV cutoff and LV cutoff. And those actions should do different things. An HV cutoff should only disconnect charge sources, and allow consumers to remain online and remove the over-voltage situation as quickly as possible. Lithium batteries don't like sitting at the elevated voltages that caused the disconnect.

LV cutoffs should still allow charge, while doing whatever they are programmed to do to ameliorate the low voltage situation.

As others have said, an HV cutoff of the solar controller poses lower, but not zero risk of damage to the charging components. There is a lot less "inertia" than there is in the collapsing magnetic field of an alternator, so it is easier to deal with in electrical design. Ideally, however, an HV cutoff would disconnect the solar input from the controller, rather than the controller output to the batteries. The risk appears to be low enough, and the complications of installing high enough, that this approach does not get implemented often, instead cutting off charge current at the battery.

[krid2000]

Thanks y'all for clarifying!
Dirk

[CaptainRivet]

@sailingharry: yes i do have real life experience.
I currently run 4p4s with 272AH Lishen a+ grade so 1088AH bank.
I have electric galley and sometimes i pull 400A+from the bank, my system is tested till 600A cont.
I have electrodacus BMS which has a 300mA passive balancer and already balances before 3,4V with a special logic when cell devation is too big.
If the bank gets out of balance by 80mV which happens due to sometimes pulling high loads then the BMS would need approx 5 weeks operational on a 1088AH bank to balance this due to creator of the BMS i can agree with that.
So an active 5A balancer is on the way to me that fixes that.
In the upper knee of 3,4V a 250-300AH cell has 5 AH capacity, 80-100mv devation are about 1AH for one cell with 4 its 4AH that a 300mA has to burn off. Simple math as the balancer does that sometimes on 3 cell packs in parallel the 300mA are divided, so maybe 80mA is balanced then per cell.

[CaptainRivet]

Originally Posted by HeywoodJ

This is a common design flaw of both battery systems and BMSs. The BMS should have separate actions for HV cutoff and LV cutoff. And those actions should do different things. An HV cutoff should only disconnect charge sources, and allow consumers to remain online and remove the over-voltage situation as quickly as possible. Lithium batteries don't like sitting at the elevated voltages that caused the disconnect.

LV cutoffs should still allow charge, while doing whatever they are programmed to do to ameliorate the low voltage situation.

As others have said, an HV cutoff of the solar controller poses lower, but not zero risk of damage to the charging components. There is a lot less "inertia" than there is in the collapsing magnetic field of an alternator, so it is easier to deal with in electrical design. Ideally, however, an HV cutoff would disconnect the solar input from the controller, rather than the controller output to the batteries. The risk appears to be low enough, and the complications of installing high enough, that this approach does not get implemented often, instead cutting off charge current at the battery.

What you describe is the classic load and charge bus design. Besides doing exactly what ypu describe it also cleans up your install. The only problem here are charger/inverter combos like victron multi, best solution adress them to a 3rd combo bus, they are most likely the biggest current load or charge source in your system and very close to the bank.

How the cut off happens is dependent on BMS. Some switch the source off on the low current side with factory remote (i prefer them) or CAN bus or you have relay of all sorts cutting off the high current side. Low current side has much lower operational risk (also to get dumb loads) and less to no costs involved as also the remote function is tested by factory and present.
Best example is victron MPPT, you can cut input, output side or remote, remote is present and low risk as tested to work 20000times. Well a good blue system 300A relay has 300 cycles due to manual, saw and had some failing much earlier when contacts melted close.

[CaptainRivet]

Originally Posted by HeywoodJ

This is a common design flaw of both battery systems and BMSs. The BMS should have separate actions for HV cutoff and LV cutoff. And those actions should do different things. An HV cutoff should only disconnect charge sources, and allow consumers to remain online and remove the over-voltage situation as quickly as possible. Lithium batteries don't like sitting at the elevated voltages that caused the disconnect.

LV cutoffs should still allow charge, while doing whatever they are programmed to do to ameliorate the low voltage situation.

As others have said, an HV cutoff of the solar controller poses lower, but not zero risk of damage to the charging components. There is a lot less "inertia" than there is in the collapsing magnetic field of an alternator, so it is easier to deal with in electrical design. Ideally, however, an HV cutoff would disconnect the solar input from the controller, rather than the controller output to the batteries. The risk appears to be low enough, and the complications of installing high enough, that this approach does not get implemented often, instead cutting off charge current at the battery.

What you describe is the classic load and charge bus design. Besides doing exactly what ypu describe it also cleans up your install. The only problem here are charger/inverter combos like victron multi, best solution adress them to a 3rd combo bus, they are most likely the biggest current load or charge source in your system and very close to the bank.

How the cut off happens is dependent on BMS. Some switch the source of on the low current side with factory remote (i prefer them) or CAN bus or you have relay of all sorts cutting off the high current side. Low current side has much lower operational risk (also to get dumb loads) and costs involved as also the remote function is tested by factory and present.

[sailingharry]

Originally Posted by CaptainRivet

What you describe is the classic load and charge bus design. Besides doing exactly what ypu describe it also cleans up your install. The only problem here are charger/inverter combos like victron multi, best solution adress them to a 3rd combo bus, they are most likely the biggest current load or charge source in your system and very close to the bank.

There is another problem with the split bus. It's a manageable problem but still a problem.
On my boat (still AGM), I have a 400 amp main fuse. This fuse serves as a supply when my 3kw inverter is online (theoretically 300 amps of discharge current). It also serves as a charge when my 270A of alternator is online. That will soon be supplemented by 70A of solar, allowing upwards of 330A of charge. I intend in my lithium conversion to have a single contactor between the battery and "the bus." There will still be a single fuse and a single wire to the battery. All of the charge sources will be disabled via control wires, rather than a disconnect. As a result, I will only need a single contactor, a single fuse, and a single large battery cable.

This has a small downside. Should I have a low voltage disconnect, that will also disconnect charge sources (but if I have a low voltage disconnect, I probably do not have any functioning charge sources!). I will have an alarm at 15% SOC, so I do not believe that a low voltage disconnect will occur in the life of the battery.

[sailingharry]

Originally Posted by CaptainRivet

@sailingharry: yes i do have real life experience.
I currently run 4p4s with 272AH Lishen a+ grade so 1088AH bank.
I have electric galley and sometimes i pull 400A+from the bank, my system is tested till 600A cont.
I have electrodacus BMS which has a 300mA passive balancer and already balances before 3,4V with a special logic when cell devation is too big.
If the bank gets out of balance by 80mV which happens due to sometimes pulling high loads then the BMS would need approx 5 weeks operational on a 1088AH bank to balance this due to creator of the BMS i can agree with that.
So an active 5A balancer is on the way to me that fixes that.

My install will not be far from yours. I am looking at closer to 800aAh of LFP at 12V. I will not be electric cooking, so my discharge loads will rarely break 100A (the coffee pot). I am currently leaning strongly towards the electrodacus.

I'm not sure if you are saying that you have actually had out of balance cells that you cannot bring into balance, or if you are concerned that that might happen. In other words, is your active balancer a preemptive fix or solving a problem?

[CaptainRivet]

Originally Posted by sailingharry

There is another problem with the split bus. It's a manageable problem but still a problem.
On my boat (still AGM), I have a 400 amp main fuse. This fuse serves as a supply when my 3kw inverter is online (theoretically 300 amps of discharge current). It also serves as a charge when my 270A of alternator is online. That will soon be supplemented by 70A of solar, allowing upwards of 330A of charge. I intend in my lithium conversion to have a single contactor between the battery and "the bus." There will still be a single fuse and a single wire to the battery. All of the charge sources will be disabled via control wires, rather than a disconnect. As a result, I will only need a single contactor, a single fuse, and a single large battery cable.

This has a small downside. Should I have a low voltage disconnect, that will also disconnect charge sources (but if I have a low voltage disconnect, I probably do not have any functioning charge sources!). I will have an alarm at 15% SOC, so I do not believe that a low voltage disconnect will occur in the life of the battery.

read my post again. important is a clean load and charge bus seperation. the combined charger/inverter like mutli go to a 3rd combo bus i call it. Each bus goes directly to bank with a fuse, that splits the current and makes install cleaner and you can faster trace issues.
3 reasons for combo bus:

1) i hardly plug into shore so 99,5% that is my heavy load bus with 8x70sqm (4xplus/4minus to 2 inverters), every cable is fused with 200A NH1. the combo and load bus (with a BP220 as disconnect relay) are connected to a 630A NH3 fuse as main load fuse thats connected to main shunt of electrodacus. That 630A NH fuse can be easily pulled under full load as manual load emergency disconnect

2) combo bus on/off via internal remote of inverters (very reliable), no last resort relay has these 400A high current relay cause more problems then being relaible and being ridicoulsly expensive
3) that keeps the load and charge bus managable both running at max 150A, i use Victron BP220 which is a 220A SSD as last resort and LCV/load and HVC/charge relay. the charge bus having solar and DC2DC is connected to a NH2 250A fuse that goes on PV shunt of electrodacus as this are my 99.5% charge sources.

[CaptainRivet]

Originally Posted by sailingharry

My install will not be far from yours. I am looking at closer to 800aAh of LFP at 12V. I will not be electric cooking, so my discharge loads will rarely break 100A (the coffee pot). I am currently leaning strongly towards the electrodacus.

I'm not sure if you are saying that you have actually had out of balance cells that you cannot bring into balance, or if you are concerned that that might happen. In other words, is your active balancer a preemptive fix or solving a problem?

welll active balancer solving a problem and being lacy.
problem i pull 500A and not often in balancing range, so its normal to get devations.
lacy: an active balancer can spare you to tear the bank appart and top balance it. to top balance, I just disconnect all loads (load bus off) when at 80% and let solar charge till full and have active balancer running. when at balancing range above 3,4V i reduce chrage current so the bank stays long in the balancing range and the 5A active balancer doing the top balancing job.

a 5A heltec is 40Euro, so cheap. I would just get it and have it, and if its just for occasional top balance.



what cells are you using? 100A on a 800AH bank should not cause deviations if good quality cells, if they are not so well match B what most really get you will need an active one. i source mine via an NPO that gets all 2 weeks 2 container cells for EV and power walls.