LFP system design with interconnected LFP windlass bank

[pretzel]

This is for an Island Packet 380 sailboat. It’s currently outfitted with 3x AGM’s for the house, and 2x AGM’s for the windlass/thruster.
Everything is 12v.

Here’s the system I’ve been sketching out:
House battery: 4P4S EVE LF304Ah LFP cells
Windlass/thruster battery: 1P4S EVE LF304Ah LFP cells

House and windlass batteries to be joined by an interconnect conductor 6m (20’) (one way) of 8 or 6AWG
With appropriate breakers at batteries at each end - 60-75A for 8awg or 100A for 6awg (correct?)
- Though I might up the size to support 150A for the windlass at least, to have a backup for that.

The rated output of the EVE LF304Ah cells is 2C for ~30sec
1P4S should be ok for thruster’s 680A max load especially if supplemented with house bank via interconnect
(I'll measure actual Amps with a clamp meter when I have an opportunity)
Windlass is a Simpson Lawrence (Amperage TBD, but probably roughly <150A)

Advantages of this dual-bank configuration:

• Conforms to ABYC requirement for redundant backup bank
• House bank can be taken offline for servicing during which time the windlass battery will provide operating power at only marginally reduced capacity
• Supplemental thruster load sharing from the house bank
• Selection via OFF-1-2-BOTH battery switch to the house load
  • Adding a simple LED digital mili-voltmeter between the bank positives to determine that it’s ok to parallel
  • And a length of resistance wire to manually equalize banks if necessary

Charging:
Sources:

• Upgraded alternator on the Yanmar diesel
• Shore power / genset 110v charging
• Solar (eventually)

In a simplified charging scenario, charging is mainly to the house bank, and due to the inherent resistance of the bank interconnection, the house battery will take the lead, but both batteries will achieve full and equalized charge together.

Bonus if one or more charging sources can be branched separately to both banks to equalize independently if necessary.
There exists potential for conflict if paralleled charging circuits bypass interconnect, thus, branched and isolated bank charging is preferred.
(Is there an alternator that can charge dual banks?)

Anything in this seem off?
Sounds like a great idea?
This is a work in progress so I really appreciate feedback on this design idea.
Thanks!

 

[pretzel]

[updates added here as appropriate]

- Interconnect fused at both ends near batteries.

 

[codfish]

This is for a 40’ sailboat. It’s currently outfitted with 3x AGM’s for the house, and 2x AGM’s for the windlass/thruster.
Everything is 12v.

Here’s the system I’ve been sketching out:
House battery: 4P4S EVE LF304Ah LFP cells
Windlass/thruster battery: 1P4S EVE LF304Ah LFP cells

House and windlass batteries to be joined by an interconnect conductor 20’ (one way) of 8 or 6AWG with an appropriate breaker - 60-75A for 8awg or 100A for 6awg (correct?)

The rated output of the EVE LF304Ah cells is 2C for ~30sec
1P4S should be ok for thruster’s 680A max load especially if supplemented with house bank via interconnect
(I'll measure actual Amps with a clamp meter when I have an opportunity)
Windlass is a Simpson Lawrence (Amperage TBD, but probably roughly <150A)

Advantages of this dual-bank configuration:

• Conforms to ABYC requirement for redundant backup bank
• House bank can be taken offline for servicing during which time the windlass battery will provide operating power at only marginally reduced capacity
• Supplemental thruster load sharing from the house bank
• Selection via OFF-1-2-BOTH battery switch to the house load
  • Adding a simple LED digital mili-voltmeter between the bank positives to determine that it’s ok to parallel
  • And a length of resistance wire to manually equalize banks if necessary

Charging:

Battery charging by:

• Upgraded alternator on the Yanmar diesel
• Shore power / genset 110v charging
• Solar (eventually)

In a simplified charging scenario, charging is mainly to the house bank, and due to the inherent resistance of the bank interconnection, the house battery will take the lead, but both batteries will achieve full and equalized charge together.

Bonus if one or more charging sources can be branched separately to both banks to equalize independently if necessary.
There exists potential for conflict if paralleled charging circuits bypass interconnect, thus, branched bank charging is preferred.
(Is there an alternator that can charge dual banks?)

This is a work in progress so I really appreciate

Fused ( or breakers) both ends sized for the AWG rating. You have two high current sources. And ideally, in addition something like a Class T or Mega fuse right on each battery.

 

[codfish]

I also have a 40’ sailboat, and that is a Lot of ah. I converted to a single 280ah Eve a couple years ago ( and 100 watts of solar). the battery seems to be enough. Solar is too small, but tops / floats the battery when away. The Big difference is charge acceptance. Our usage pattern seems to have enough incidental engine time, and the battery takes All those amps, so that it works out. If you are a live aboard, and run everything electric then you would need more.

 

[Horsefly]

I'm helping someone convert the house battery on their 40' Island Packet to LFP. We're doing 4S2P, with two 150A BMS's. After looking at the amperage draw on everything, we decided to keep the bow thruster on a separate AGM bank. The LFP could handle the ~500A required for the bow thruster, but the BMS couldn't. The bow thruster bank will get a 15A DC-to-DC charge from the house bank. I wanted to put the windlass on the bow thruster bank, but I lost that argument, so it will be on the house bank. Should be OK though.

I don't think I understand the concept of simply connecting the house bank to the bow thruster bank. What's the point of having the two separate banks if they are just connected to each other?

If you are doing 4P4S 304Ah cells for the house bank, what kind of BMS are you using? You are going to have a BMS, right? Same thing on the bow thruster bank.

 

[pretzel]

I'm helping someone convert the house battery on their 40' Island Packet to LFP.

Awesome - this is for an IP 380

We're doing 4S2P, with two 150A BMS's. After looking at the amperage draw on everything, we decided to keep the bow thruster on a separate AGM bank. The LFP could handle the ~500A required for the bow thruster, but the BMS couldn't.

Have you looked into an external contactor for the BMS?
And yea, the current draw over that distance makes for a very large conductor - that's another reason I have for two joined banks.

The bow thruster bank will get a 15A DC-to-DC charge from the house bank. I wanted to put the windlass on the bow thruster bank, but I lost that argument, so it will be on the house bank. Should be OK though.

That's going to be a large conductor for the windlass, right?

I don't think I understand the concept of simply connecting the house bank to the bow thruster bank. What's the point of having the two separate banks if they are just connected to each other?

The banks are connected through an OFF-1-2-BOTH switch for flexibility.

The advantages of interconnection include:

• Conforms to ABYC requirement for redundant backup bank
• House bank (either bank for servicing,) can be taken offline for servicing during which time the windlass battery will provide operating power at only marginally reduced capacity
• Supplemental thruster load sharing from the house bank
• ALSO: identical battery chemistry for both banks means they can simply be paralleled, and the charging profile(s) are identical, too.

If you are doing 4P4S 304Ah cells for the house bank, what kind of BMS are you using? You are going to have a BMS, right? Same thing on the bow thruster bank.

Yes, BMSs will likely be REC or similar, and both will use an external contactor, so system amperage won't be an issue.

Thanks for the feedback! I'd like to compare more notes as we go. -cheers

 

[Horsefly]

That's going to be a large conductor for the windlass, right?

I won't be there when he does the install, but I suggested 2/0 AWG, which should be plenty. Still lots smaller than the bow thruster.

 

[pretzel]

More thoughts on the bow battery:

The advantages I’m seeing for LFP for the fore bank include:

• Getting 2x AGM heavy monster batteries off the bow.
• The thruster in its current configuration is barely supported by the AGM’s. We are supposed to only run the thruster for 4 seconds at a time to avoid overheating things, apparently.
• Interconnection with the house bank for unified charging and backup bank, etc., will effectively increases the house bank capacity by another 25%.

I think the BMS issue will be covered with an external contactor fed from the BMS. Amperage then shouldn’t be an issue.

Part of the interconnect design that I like is that it serves as a backup bank. That functionality is not baked in when configured for a DC-DC charger, though it can be wired around.

But another option is to keep the AGM windlass batteries, add a dc-dc charger, as well as running an interconnect to allow the fore battery to serve as backup for house loads through a simple 1-2 switch.

This still leaves us with a huge amount of dead lead at the bow, which isn't the best.
And I do like the simplicity of letting the two banks with identical charging profiles supplement loads for each other via running parallel.

All that said, certainly the complexity and cost of a second LFP bank with contactor is a factor, too.

I suppose I could split the project into phases and wait on the LFP bow battery conversion for now.
But given high inflation, maybe it's better to spend now!

Food for thought!

 

[Horsefly]

But another option is to keep the AGM windlass batteries, add a dc-dc charger, as well as running an interconnect to allow the fore battery to serve as backup for house loads through a simple 1-2 switch.

I think I have more problem with this than your original thoughts of an interconnect between two LFP banks. Others here seem to disagree, but I don't like the idea of an AGM bank connected on a normal basis with an LFP bank.

Your bow thrusters are a bit more than the ones on my target boat. His are only max 480A. That sounds like a ton, but it is quite a bit less than your 680A.

Help me understand the concept of operations of two battery banks interconnected: Will the two banks be interconnected under normal circumstances? You said it was 1/2/both, but I can't fathom what that means for two battery banks for two mostly separate loads. What does the switch connect or disconnect? When you are going into a marina and need the thrusters, is than when you connect the two banks? What happens if they are at dramatically different state of charge at that point? When do you disconnect them?

It seems like REC may be a good idea on a boat, but... My experience was only with FET-based BMS's, and I don't think I knew enough to recommend something as expensive as the REC. So I installed two JBD 150A FET-based BMS's, one each on the two 4S x 280Ah batteries. I will be sure to point him to what you are planning.

 

[pretzel]

I think I have more problem with this than your original thoughts of an interconnect between two LFP banks. Others here seem to disagree, but I don't like the idea of an AGM bank connected on a normal basis with an LFP bank.

Your bow thrusters are a bit more than the ones on my target boat. His are only max 480A. That sounds like a ton, but it is quite a bit less than your 680A.

Help me understand the concept of operations of two battery banks interconnected: Will the two banks be interconnected under normal circumstances? You said it was 1/2/both, but I can't fathom what that means for two battery banks for two mostly separate loads. What does the switch connect or disconnect?

1-2-BOTH selects source for the house load.
Will be in BOTH unless a bank is down for some reason.
So both banks usually share loads in parallel.

When you are going into a marina and need the thrusters, is than when you connect the two banks?

Bow battery is always connected directly to the windlass/thruster via bms/contactor regardless of switch setting.

What happens if they are at dramatically different state of charge at that point? When do you disconnect them?

From what I'm seeing, it's not a big deal with LFP's since the voltages are very similar even with greatly varying SoC.
Another post was showing on the order of a dozen amps between banks. But, I don't know for sure! That's just my hunch for now.

Incidentally, because interconnect offers some resistance, house bank leads for house loads and bow leads for windlass/thruster.
(I may end up sizing the interconnect larger if the thruster trips the interconnect breaker, but we'll see and I'll have a better idea after measuring actuals.)
But in the end they equalize with each other since they are in parallel.

It seems like REC may be a good idea on a boat, but... My experience was only with FET-based BMS's, and I don't think I knew enough to recommend something as expensive as the REC. So I installed two JBD 150A FET-based BMS's, one each on the two 4S x 280Ah batteries. I will be sure to point him to what you are planning.

I like it, thanks for the mention.
Cheaper FET BMS feeding a contactor is probably what I will go with for the bow bank. Something fancier for the house bank.

-cheers

 

[codfish]

I like the plan, and do something similar with house and starter battery ( both diy LFP). They often run connected but with a current limiting wire between them. They end up at the same voltage but each takes the majority of their local load.
I suspect you could get by with a much smaller thruster / windlass battery. I think I calculated retrieving 125ft of chain only consumes 3ah even assuming full windlass amps the entire retrieval. I went with Eve 100ah, but rated at 3C, so they can do the current. I also violated convention by eliminating the BMS, but added an active balancer. It has worked fine for 2 years. The 100ah Eve are much smaller and actually cost less than the lead battery replacement would have cost.

 

[Horsefly]

I like the plan, and do something similar with house and starter battery ( both diy LFP). They often run connected but with a current limiting wire between them. They end up at the same voltage but each takes the majority of their local load.
I suspect you could get by with a much smaller thruster / windlass battery. I think I calculated retrieving 125ft of chain only consumes 3ah even assuming full windlass amps the entire retrieval. I went with Eve 100ah, but rated at 3C, so they can do the current. I also violated convention by eliminating the BMS, but added an active balancer. It has worked fine for 2 years. The 100ah Eve are much smaller and actually cost less than the lead battery replacement would have cost.

Interesting. What exactly is a "current limiting wire"?

 

[codfish]

Interesting. What exactly is a "current limiting wire"?

A wire long enough that when you calculate the wires resistance and likely voltage difference limits the current to V / R. It is fused on both ends at less then the AWG rating.
Why? I can leave the boat on solar float and both batteries at the same voltage. And after a long motor the starter battery which would otherwise stay at needlessly high charge bleeds slowly down to the house battery float voltage.
I do have a dc-dc victron charger but these days is mostly turned off.

 

[Horsefly]

A wire long enough that when you calculate the wires resistance and likely voltage difference limits the current to V / R. It is fused on both ends at less then the AWG rating.
Why? I can leave the boat on solar float and both batteries at the same voltage. And after a long motor the starter battery which would otherwise stay at needlessly high charge bleeds slowly down to the house battery float voltage.
I do have a dc-dc victron charger but these days is mostly turned off.

Well, I'll have to pass all this on to my friend. I'm sure he will stick with the separate AGM banks for now, but he may want to reconsider later.

He's got a 100A Balmar alternator and a Wakespeed-500 that will be going directly to charge the house battery. The solar will also go to the house battery. Then he has two separate DC-to-DC chargers, one to the bow thruster AGM bank (200Ah) and one to the AGM engine starter battery (100Ah).

 

[codfish]

Sure if you have the stuff then use it. But I lean towards fewer things to break in safety critical spots. This works and only takes cheapo voltmeter to diagnose.

 

[pretzel]

I like the plan, and do something similar with house and starter battery ( both diy LFP). They often run connected but with a current limiting wire between them. They end up at the same voltage but each takes the majority of their local load.

Great to hear it's working for you!

I suspect you could get by with a much smaller thruster / windlass battery. I think I calculated retrieving 125ft of chain only consumes 3ah even assuming full windlass amps the entire retrieval. I went with Eve 100ah, but rated at 3C, so they can do the current.

The cost difference is minimal for 304ah cells which will be identical to the house bank, so better interchangeability.
Also, less risk of depleting the bank. Another bonus is that it supplements the house bank with an additional 25% capacity.

But the overriding consideration for sizing is the amperage limitation.
This type of LFP cells are rated, at 2C, so 608A - and that's only for 30 seconds.
So, hopefully the actuals show that the draw is within reason, and it will be supplemented a bit by the house bank, too.

I also violated convention by eliminating the BMS, but added an active balancer. It has worked fine for 2 years. The 100ah Eve are much smaller and actually cost less than the lead battery replacement would have cost.

Cool - that makes me think that a low-amp bluetooth connected bms might fit well for the bow bank, feeding a contactor for LVC protection.
I can monitor the cells balance for peace of mind, and/or add active balancing, too.
As long as the bow bank BMS has dual charge/load (non-common) connections, it will be protected for charging, too.

And since the bow pack is charged from the same source as the house bank, that house REC or whichever fancy BMS I end up with there will protect both banks for charging. <-- not sure how I'd need to wire it for this to work properly, so holding off on this statement for now