High Current lithium charging with drop-in

[80MB]

Hello all. I am seeking some advice on a high-current alternator charge solution for lithium setup. I have read threads like this and more, but still had questions at the end.

I have upgraded much of the current electrical system in my 42' sailboat, and have been educating myself on lithium options for some time now. I'm ready to choose cells, but have a couple unanswered concerns. Here are the basics of what I have so far:

all 12v system on the DC side:
4x190w HQST Panels
2xVictron 12v MPPT
Multiplus 12/3000
Cerbo GX
SmartShunt 500a

The house battery bank, Multiplus, Alternator, and load panel all hang off of a 4 post bus bar, aka the "charge bus"

The goal is to have 6-800ah of available lithium, so likely 3-4 cells connected in series. Since the continuous charge rate of this parallel pack will likely be >200a, I would like to charge it as fast as possible with a large alternator, to conserve fuel and potentially eliminate the need for a generator.

Ready to install: Balmar XT 12v250a and MC-618 regulator, which supports a LiFePo4 charge profile. I intend to limit the alternator output to 80% max field [~200a], and to cut off charging well before 100% SOC (~90 or so, let the solar finish the last bit). I understand that using voltage is a poor way to gauge lithium SoC, any advice on a better option is welcome.

However, I am concerned about handling a load dump scenario due to pack disconnect. The alternator was too expensive to be replacing, even once. The "best" way to do this is with a BMS that can shut down the alternator. The MC618 should power-down the alternator safely if it loses 12v, according to Balmar, but we still need a way to trigger that action. Drop-in replacement batteries provide no access to the BMS, which may or may not have any way to interface with it reliably to disconnect the regulator. So this doesn't seem like much of an option.

It seems there are a few paths:

In the do-it-right category, we have:
The pricey route:
Victron Lithium
Victron Lynx SmartBMS
Wakespeed WS500

... at 2x the cost!

The DIY route:
Source raw cells
Rec BMS
Wakespeed WS500

... 2x the complexity.

To continue with "drop-in" cells (specifically, SOK 12v206p), we have options like:
Use a battery isolator (ArgoFET, etc.) to provide an "alternate path" to your lead-acid starter bank /for the alternator-frying voltage spike which occurs during pack disconnect. Problems here are that you have no separate charge profile for the start bank. I've also heard that it "might not always work", but mostly from retailers. If you know specific reasons why this solution may fail, please share!

Use an "alternator protection device/module" -- limited to 60v spike, seems questionably undersized, would probably become a (~$70/ea!) consumable. Could these be wired to achieve higher voltage protection?

Use [multiple parallel] DC-DC chargers - The primary feature with this option seems to be providing a LiFePo4 charge profile, which is covered by the MC-618 regulator. Even with the largest known commercially "available" options (i.e. Sterling 120, CTEK Smartpass 120s), 2 would be required of them to reach the 200a goal. There may be a sterling 180a, which might be acceptable. However, I can't even find these for sale, so pricing information is unknown. And there is shipping to US from the UK to consider. These seems like a poor band-aid.

It occurred to me though, that if using a DC-DC charger on the lithium helps shield the alternator from a pack disconnect, perhaps it would also work to have an isolated DC-DC charger hanging off the charge bus and controlling the charge profile for the FLA starter bank. This would provide an alternate path to absorb the load spike, while also isolating the 2 banks and keeping them charged with an ideal profile for each. With this, the APD/APM might actually give the system a chance to absorb a voltage spike

Having said all that, there seem to be a handful of conditions that would trigger a pack disconnect, and most of them are accounted for:

1. Under temp -- Handled by MC-618
2. Over temp -- Handled by MC-618
3. Under-voltage -- Handled by MC-618. Also probably not a concern for the alternator, if we are low enough to cut off the pack, the alternator seems not likely to be running/charging
4. Over-voltage -- Seems this could be addressed by programming the charge profile to stop charging well before the voltage cutoff point. Perhaps this could be due to an individual cell-imbalance, though. Concerns like this make the DIY/Rec option more attractive.

So, my questions are basically:
Is there a more intelligent way to shut down the alternator regulator with drop-ins/no bms access?
With a externally regulated alternator, would DC/DC charging the starter bank provide a viable alternate path for spikes during a load dump scenario?
Is there even much of a concern for a pack disconnect/load dump if temps are watched by the regulator, and the alternator is only charging to say 90% SOC?
Are there other options I have not considered?
Any other thoughts?

I have been working on a system diagram I would be happy to share as well, if that is any help.

Thanks for taking the time to read this, hopefully it all made sense.

 

[Goboatingnow]

The 618 does not have a remote disconnect , but does have a “ force to float “ input. A bus could trigger this before abrupt disconnect

 

[80MB]

The 618 does not have a remote disconnect , but does have a “ force to float “ input. A bus could trigger this before abrupt disconnect

From the MC-618 manual:

The preferred method of turning off the regulator is disconnecting the regulator’s ignition (brown) wire, but if used as an EMERGENCY ONLY shutdown, disconnecting the regulator’s power input (red) wire in addition to the ignition wire has a very low chance of damaging the regulator.

Figured the BMS could trigger a relay to open the brown wire connection, achieving a remote disconnect.

 

[svsagres]

Having been through the DIY battery route, it really wasn't all that complex. It's really nice to fully engineer it so you never get a disconnect. That said, there are other pre-fab batteries that the WS500 can work with.

 

[mikefitz]

Connect a lead acid battery in parallel with the lithium bank.
This will provide two useful features should the BMS disconnect, a load for the alternator and emergency DC power.
This is a fail safe system.

Mike

 

[Goboatingnow]

Connect a lead acid battery in parallel with the lithium bank.
This will provide two useful features should the BMS disconnect, a load for the alternator and emergency DC power.
This is a fail safe system.

Mike

Yes but it can damage the lead acid , and represents a safety risk

It’s also not ABYC or ISO compatible if it’s a boat for example. This has insurance implications

 

[mikefitz]

Why can it damage the lead acid?

 

[Goboatingnow]

Why can it damage the lead acid?

Any shorts in the LA and big currents can flow ,and the LA can discharge into the lithium resulting in PSOC issues.

 

[Vigo]

I intend to limit the alternator output to 80% max field [~200a], and to cut off charging well before 100% SOC (~90 or so, let the solar finish the last bit). I understand that using voltage is a poor way to gauge lithium SoC, any advice on a better option is welcome.

So, using voltage is a 'poor way to gauge' SoC, but it's not useless. It's perfectly good in the sense that 'if pack rises to ~13.3-13.4 im pretty dang sure 80-90% of it is in there'. So in that sense using a number like that to transition to a lower charge current or to cease charging is easy. The only thing you need to be careful of is depending on where you sense that voltage and what the wiring is like you may have to account for voltage drop across the wiring at a high charge current rate, throwing off that number by a couple of tenths. Pretty easy to navigate that as long as you realize it as a possibility to account for.

I would say if you are planning to have a large lithium bank and not planning to cycle it 100 to 0 to just stop charging at 13.3 or 13.4 and forget about the rest of it! Pack will last longer and it might simplify other things, see below:

Use a battery isolator (ArgoFET, etc.) to provide an "alternate path" to your lead-acid starter bank /for the alternator-frying voltage spike which occurs during pack disconnect. Problems here are that you have no separate charge profile for the start bank. I've also heard that it "might not always work", but mostly from retailers. If you know specific reasons why this solution may fail, please share!

So.. maybe there is a factor here that I'm not realizing but as far as i understand it you can directly parallel '12v nominal' lithium to 12v lead with very little current flow if you're not charging the lithium to 100%/14.6. It solves the 'spike' problem as far as im concerned, but it doesn't isolate the systems from each other for fault protection. One thing I've experimented with is simply having a diode between the lithium and lead. The diode will cause a ~0.7v drop which from 13.4, is 12.7 at which point basically nothing will flow into the lead, EXCEPT in the case of voltage spikes! 'Clamping diodes' are used in all sorts of common circuits for basically this purpose.

However, that single diode would only prevent the lithium from draining the lead, not the other way around. To get protection from a failed lead side draining/shorting the lithium, you could either A: just put a fuse in line with the diode on the thinking that whatever voltage/current spike occurs will not occur long enough to actually heat up and melt the fuse element (say, 30a+ rating might work for 250a at 10ms? Just spitballing..), and just accept that it wouldn't protect from a very slow drain which is under the fuse rating, or B: have the diode circuit be switched, perhaps by a relay from the load port of a charge controller or simple adjustable voltage switch (i have bought an adjustable voltage 'window switch' with a small relay built in for $4 on Amazon Prime, no kidding..), such that 'if lithium drops below X, lithium-to-lead connection goes open circuit'. Or do both! Likewise to the fuse idea, that whole circuit would not really need to be sized for 250a. It would only need to be sized to not melt INSTANTLY from 250a. Every 12v automotive relay experiences a ~60v spike and the 'clamping' circuit is NOWHERE NEAR the common 30a capacity of the actual switch in the relay. If it has a fuse in it i think you could make this from 14ga wire! I think if one were to experiment and find that point, it wouldn't need to be that large of components because the duration of the high current condition would be so short. Wouldn't hurt to go bigger, but i would guesstimate that an 8ga wire has 'no chance' of instantly smoking from that short 250a, and 10, 12, 14.. im not 100% sure because i don't know the duration of the high current, but if it's short enough you can use tiny wire!

As far as damaging the lead acid.. not likely? Any starter battery in any car is subjected to currents in that ~250a range for much longer than a 'spike', multiple times per day. Those are ~40lb batteries in the ~50-100ah range and they last years doing that. Walmart sells full size car batteries for $59, so even as a 'consumable' i imagine they'd last a lot more uses than the electronic gizmo you posted.. My experience with electronics is you go far enough past the limit ONE time and it's fried open circuit. A battery might just get worse and worse at being a battery, but im not sure how much worse it would actually get as a surge protector! Even a 'bad' 12v lead battery, in terms of internal resistance, is gonna dang well take a good amount of current at ~60v. I imagine a flooded battery is better for this than agm or gel because of something about making permanent gas pockets in the gelled electrolyte (or something like that) from the current spikes.