Review/Feedback: LiFePO4 upgrade

[aybabtme]

Hey folks,

I'm looking for a 2nd set of eyes on a diagram I've been working on (PDF here, picture below) (using KiCAD and made up symbols). I'm not a pro at KiCAD and electrical diagrams, so pardon my abuse and ignorance of conventions. I hope the diagram is understandable!

I'm about to start an overhaul of the main power system in our boat. When I started this process, we had the following criteria in mind:

1. High power: our boat comes with CNG for the galley, and it needs to go. I don't want to replace it with propane, I'll go for an electric galley instead. I'm starting with 3kW inverter but want to be able to go bigger.
2. Removable: I want to be able to remove the equipment fairly easily and bring it on another boat later, or on some other off-grid system.
3. Prototype/test bed: we see our current boat as a test bed for the future boat we'd like to have. We're not afraid to test ideas with it. We're happy to be adventurous (and learn from our mistakes).

Note that from the design below, we already own the LiFePO4 cells (24x CALB CA100Fi), the Orion Jr BMS, Victron 100-50 MPPT and Victron Multi 24/3000KVA. They're sitting under a settee for now while I finalize the wiring diagram and figure out the safety components (fuse, breakers, contactors). We don't yet have the Victron Orion-Tr battery-to-battery (B2B) chargers, and I'm not sure about that section which interacts with the 12V circuit and the engine alternator and battery.

For context, I'm building the LiFePO4 bank as a separate circuit from what's already there, using 24V instead of 12V because of the size of the PV array (charge controllers able to receive >900W at 12V are very expensive) and to be able to power additional heavy consumers in the future. I'd have gone with 48V but it's too much of a hot subject/troll magnet, and there's not a lot of marine equipment that supports 48V.

The existing 12V system is mostly unchanged, aside for:

1. Replacing the AGM cells with a single small start battery.
2. Disconnecting the engine alternator from the 12V, connecting it to the 24V bank through a B2B charger.
3. Using the 24V bank to regulate charge the 12V circuit.

By keeping all the 12V in place mostly unchanged, I can always remove the 24V system and the LiFePO4 later. I'd only need to reconnect the alternator directly to the 12V circuit.

In the diagram below, I attempted to incorporate all the safety features that the individual devices demand (from their manual and specifications). I'm also going for a separate port (charge, discharge) setup like described in Nordkyn Design's Lithium battery systems series. Despite this, I'm sure I messed something up.

FYI: the 24V part is sure to raise a lot of eyebrows. This area is pretty much settled, so don't waste too much time trying to change my mind.

[john61ct]

Originally Posted by aybabtme

High power:
… electric galley instead. I'm starting with 3kW inverter but want to be able to go bigger.

Inverter is the least of it, the real key is energy inputs, lots and lots of Ah coming in every day.

IMO keep the working CNG galley until the electric appliances and Ah-per-day requirements have been well proven for months of cruising.

How many days at a time are you away from shore power?

Good thread on the LFP storage side: http://www.cruisersforum.com/forums/...ey-201795.html


> Removable: I want to be able to remove the equipment fairly easily

If you mean the LFP bank, that is do-able, but not optimal with a big one, will take extra care in design and a fair bit added expense and complexity.

Another approach is your main "mothership" bank mostly stays on the boat, but sub-pack "modules" get taken out and plugged back in, e.g. for powering an electric dinghy motor.



> Orion Jr

I would be **very** interested to learn more about your experience with their BMS, and especially (whether it is possible / howTo) integrate with the above "mothership / dinghy module" concept.

One thing I really like is their ability to accept J1772 input and control Elcon chargers via CANbus signal in a failsafe manner.

Such a discussion likely worth its own thread.


> Victron Orion-Tr battery-to-battery (B2B) chargers

Have you got compelling reasons to choose those over the Sterlings?



> I'd have gone with 48V

Also worth its own thread, personally I'm very interested in that, and as electric propulsion becomes more mainstream we're going to need to work out the howTo details anyway.

48V is great for storing high-power (many kW) mains/ICE AC charge inputs, due to all the world-class gear from ex-telecom / server room infrastructure available cheap as chips.

"Distributed Power Systems" and "Point of Load" regulation, all key googling terms. Basically, you have your Storage / Backbone Buss voltage up high, then multiple lower-voltages outputs are delivered close to the PoL, using modular DC-DC converters as needed.

Often implemented with 270Vdc (range, up to ~310V) from high output rectifiers these days, which is also great for efficient hybrid propulsion off a genset.

In which case 48Vdc becomes just a bank charging output, and another output for loads that require it, or intermediate to the 12/24 ones.

> Disconnecting the engine alternator from the 12V, connecting it to the 24V bank through a B2B charger

Consider replacing that with a 24V alternator with a good external VR, direct to House. Eliminate the costly inefficiency and current bottleneck of DCDC at that key high-amp input source.

Starter batt can stay 12V, takes very little input to keep topped up. Or go to a 24V starter one day. . .

[aybabtme]

Originally Posted by john61ct

> Victron Orion-Tr battery-to-battery (B2B) chargers

Have you got compelling reasons to choose those over the Sterlings?

Price and I like Victron, and don't own anything from Sterlings. Their 12 to 24V charger could be interesting since its higher amperage, but I don't think my current alternator can handle the load. I think my alternator is 80A but I'm not sure. The Orion-Tr with 15A at 12V should be gentle enough for it. My idea has been to just have a way to recharge from the alternator if needs be, not to optimize this source. If later I need more from the alternator, I'd put a 24V one with an external, programmable regulator (if such thing exists at 24V).

[aybabtme]

correction: 40A -> 80A

Originally Posted by john61ct

> Disconnecting the engine alternator from the 12V, connecting it to the 24V bank through a B2B charger

Consider replacing that with a 24V alternator with a good external VR, direct to House. Eliminate the costly inefficiency and current bottleneck of DCDC at that key high-amp input source.

Starter batt can stay 12V, takes very little input to keep topped up. Or go to a 24V starter one day.

Another thing I thought is to use a regular 24V to 12V step down converter, and use an ultracapacitor bank for the starter battery. This way I'd get a regulated 12V, and not have to worry about charging algorithms. Plus the ultracapacitors require no maintenance, recharge pretty fast and can crank a bunch of amps. It might even help smoothen loads from the 12V regulator. But this is unknown territory for me and I haven't researched this.

[john61ct]

Originally Posted by aybabtme

this is unknown territory for me and I haven't researched this.

I really can't emphasize this enough.

You need solid reliability.

To the extent you depend pioneering hitech stuff, risk of failure at the most critical time is increased by orders of magnitude.

Make safety your #1 priority.

At least have traditional well-understood and supported alternatives as spares / backup / reserve.

KISS is even better, do the R&D on shore.

[noelex 77]

Antoine, I think your overall approach is excellent, but be careful with your power in/power out sums. 915w of solar on small boat (where there are more shadow issues) combined with a small engine alternator, seem to be your only sources of power when away from shore power. This may be just enough to run electric cooking, but you will need to cruise areas of good solar insolation and be careful about your other electrical loads if you want to avoid being dependent on shore power. Make sure these restrictions match your requirements.

A lot will depend on the details, particularly how much cooking you do. Do you like plenty of hot drinks? etc etc. I would do a careful, realistic energy audit before ditching the non electric cooking option with this sized solar array.

On the solar array, three strings need better fusing on the solar panel side. You may wish to give some thought to installing three separate controllers, one for each of your panels. This will slightly improve the output and it will provide some redundancy. This is especially important, as you will not be able to heat any food if your solar regulator fails. Three small regulators are often not much more expensive than a single larger unit.

[john61ct]

Originally Posted by aybabtme

Price and I like Victron

Is Price a name, or you mean money?

> don't own anything from Sterlings

not a good reason, buying best of breed for each component is better than blind brand loyalty.

Yes they make quality gear but not all their lines are market leading designs.

For voltage converters sure Victon's been around for yonks, the development of the DCDC charging aspect is very recent not yet out in the community getting widespread feedback, nor thorough resting by specific trusted members.

In a few years we may well hail them as the best, but for now Sterling is the tried & true, Victron's relatively unproven - only talking B2B niche here!.

Just IMO. . .

> Their 12 to 24V charger could be interesting since its higher amperage, but I don't think my current alternator can handle the load.

> The Orion-Tr with 15A at 12V should be gentle enough for it.

A DCDC charger only passes through the demand from the actual load, could be **capable** of handling conversion of 500A, but if the actual load device is pulling only 40A then that's what it will pull from the alt.

All under your control, like being rich or thin, no such thing as "too high" ampacity handling in infrastructure like DCDC chargers, same as fat wire gauges, or VSR/ACRs, or actual charge sources for that matter.



> I think my alternator is 80A but I'm not sure.

Find out, not just the rating but what current it will support continuous output in hot conditions.

> My idea has been to just have a way to recharge from the alternator if needs be, not to optimize this source. If later I need more from the alternator, I'd put a 24V one with an external, programmable regulator (if such thing exists at 24V).

If you want an electric galley, plan on that now. It will likely be the majority of your Ah per day input.

Or stick to gas-powered cooking, maybe convert to LPG.

Or consider a genset of some sort.

LFP lets you really get the most out of ICE sources, to the point that "on demand" energy sources will let you reduce bank size, not so dependent on massive solar panelage for your high power consumption needs.

[Dockhead]

Originally Posted by noelex 77

Antoine, I think your overall approach is excellent, but be careful with your power in/power out sums. 915w of solar on small boat (where there are more shadow issues) combined with a small engine alternator, seem to be your only sources of power when away from shore power. This may be just enough to run electric cooking, but you will need to cruise areas of good solar insolation and be careful about your other electrical loads if you want to avoid being dependent on shore power. Make sure these restrictions match your requirements.

A lot will depend on the details, particularly how much cooking you do. Do you like plenty of hot drinks? etc etc. I would do a careful, realistic energy audit before ditching the non electric cooking option with this sized solar array.

On the solar array, three strings need better fusing on the solar panel side. You may wish to give some thought to installing three separate controllers, one for each of your panels. This will slightly improve the output and it will provide some redundancy. This is especially important, as you will not be able to heat any food if your solar regulator fails. Three small regulators are often not much more expensive than a single larger unit.

This is really good advice.


To my eyes the weak spot here is the alternator. 80 amps @ 12v nominal is only a little more than 1kW at rated output, and car-type alternators never give their rated output for any length of time. You will not get much power out of that. I would add a second large frame school bus alternator (best) or upgrade that alternator to some truly bulk power rated 24v one (like Mark Grasser).


The solar array configured as Noelex suggests will be good, but the sun doesn't shine every day, and for some reason doesn't shine at all at night.

[aybabtme]

We're weekenders for the foreseeable future, so I'll heed the advice about the electric galley. We can leave the CNG in place and cook using only electric appliances and see how it goes. If it works, I can then take another look at removing the CNG. If it doesn't I can just fire up the CNG. We can always go back to the marina if anything seriously doesn't work out.

If it works as is with the 915W of solar, good. If I realize we're always short ~1000W/h, I can replace the current 305W panels with more modern 370W ones of the same size, and gain another ~200W of solar. If I realize I really need more power, I'll look at replacing the alternator with a beefy Balmar, 24V, externally controlled one, or something similar.

I'll add fuses on the solar array size of things. For now I'll stick to the single MPPT controller I already have, but will keep in mind the alternative idea of having 3 separate ones.

On the battery to battery charger from the alternator to the 24V bank, I think I'll punt on the Victron/Sterlings question. I have no idea yet if I need a little (small Victron), a decent amount (big Sterlings), or a ton (new Balmar alternator) of support from the alternator. I'll keep my money in my pocket until I can quantify this decision better. I want to prioritize the renewables first and see how far I can get with it.

[aybabtme]

Originally Posted by john61ct

Is Price a name, or you mean money?

The money part. The victron Orion-Tr are half the price of the Sterlings. They're more than half the power as well, but if I can get away with just a bit of alternator support, all the better.

Originally Posted by john61ct

not a good reason, buying best of breed for each component is better than blind brand loyalty.

Yes they make quality gear but not all their lines are market leading designs.

For voltage converters sure Victon's been around for yonks, the development of the DCDC charging aspect is very recent not yet out in the community getting widespread feedback, nor thorough resting by specific trusted members.

In a few years we may well hail them as the best, but for now Sterling is the tried & true, Victron's relatively unproven - only talking B2B niche here!.


Just IMO. . .

All the Victron stuff works well together and has good software. The Orion-Tr don't have VE.Direct, so that point is moot, but the Bluetooth connectivity is more interesting to me than what Sterlings seems to offer. Also, Sterlings AFAICT doesn't charge LFP batteries properly in the final CV phase. The spec of lithium cells (LFP or not in general) requires that charging stop when the current at CV drops to a certain level (e.g. stop when charging current goes down to 500mA at 3.2V). As far as I recall, Sterlings simply charges at CV for a specified duration, irrespective of the current. This means it could potentially overcharge cells by floating them at CV for longer than they should, which as I understand, is bad.


But I haven't interacted with the Sterlings equipment, so I might be mistaken. I also haven't interacted with Victron's Orion-Tr and they might do the same thing, FWIW.

Originally Posted by john61ct

A DCDC charger only passes through the demand from the actual load, could be **capable** of handling conversion of 500A, but if the actual load device is pulling only 40A then that's what it will pull from the alt.

All under your control, like being rich or thin, no such thing as "too high" ampacity handling in infrastructure like DCDC chargers, same as fat wire gauges, or VSR/ACRs, or actual charge sources for that matter.

The LFP bank will receive 1000s of amps (300Ah cells with max charge rate of 10C), way more than the Sterlings B2B can push, and way more than alternator can produce. So in that scenario, if my alternator can't sustain being continuously loaded with 70A, but the Sterlings is seeing a bottomless load on the output side, I gather that it'll pull 70A from the alternator for a long while. That's where I think having a smaller B2B charger might be better.


* * *

But all that said, I think I'll punt on the alternator->B2B->LFP part of the circuit for now. Now that I'm writing it, I think the Victron Orion-Tr's lack of VE.Direct connectivity is a blunter. One great thing with the Orion Jr BMS and CCGX's CANBUS connectivity is that the BMS cann tell the CCGX how many amps its able to accept for charging. In turn, the CCGX will throttle the MultiPlus and MPPT if needed. If the Orion-Tr aren't part of that game, it's rather strange.

[john61ct]

No, LFP for maximum longevity should not be charged anywhere near the stressful maximum specs promoted by industry.

CC-only, no Absorb is perfectly fine, can just use a dumb PSU, converter whatever, and a good HVC, and get **lots** longer lifespan than bothering with holding until an endAmps spec.

That's just "lead thinking".

In normal usage cycling that is, benchmarking capacity does require a precise termination point, e.g. "hold 3.45V until amps drop to 0.05C". But few owners do that more often than once a year, if ever.

_____
Any DCDC charger is current limited, by definition, and usually delivers well below its max rating, especially with overtemp conditions.

It is indeed foolish to go over 0.4C, especially when temps are getting cool, below 15°C my limit reco is maybe 0.2C

But no point in delivering **too low** a rate, unless your ICE source is running anyway and you're often terminating charge early on.

______
I would be **very** wary of letting the BMS control charging, unless completely separate failsafe protection devices are added.

Better IMO to select charge sources with solid controls to do what I want, and let the BMS do its main job, which is protecting against when those primary controls fail.

Just my 2¢

[aybabtme]

I'm willing to consider what you're saying about charging profiles, but I've read otherwise from primary sources (manufacturer spec sheets). Would you mind sharing primary references you might have that further explain what you describe? I would love to know more.

In the case of the cells I'm working with (CALB CA100Fi) the spec sheet is rather thin and doesn't specify anything about the charging regime, aside from a vague "4h standard charge time" and a max charge current/voltage. Other chemistries I worked with had much better spec sheets from LG and Samsung...


Arguably, unless I change the alternator for a massive one, nothing will be charging this pack above 0.4C (70A from MultiPlus, ~40A from MPPT). So all in all it's a moot concern.

* * *

The circuit includes manual switches for both the charge and discharge bus bars. I feel like that's enough. I can turn off individual charging devices in VenusOS. Could probably add a hardware switch on top of it if the desire arises.

[john61ct]

No as I said industry parrots those maximum stress specs as if that should be charge termination, almost across the board.

First off, recognize the cell manufacturers are huge billion-dollar companies, and all their focus is on their main customers, first the Chinese military, and secondly EV manufacturers. These have their own PhDs in battery science and DC power, and priorities are high-C-rate discharge for propulsion and range / per-use cycle capacity - expected lifespans are relatively short in those use cases, and optimizing for longevity is just not high on their list of priorities.

The purchase of LFP cells directly by consumers for use as "House bank" in a mobile context and for solar energy is at most .0001% of the cell manufacturers' sales, not even on their radar in economic terms, even as a niche sub-market, .

It is for these same reasons that industry and academic research is not targeted toward maximizing longevity, but oriented around those EV & other high C-rate use cases, so **very** different from our much gentler House bank cycling.

And thus no testing / different specs are developed by industry for the House bank use case; that has been left to the end users, and the more objective/technical vendors like MaineSail.

Consider also this: why would CALB want to teach its customers how to get 5, 8, even 10000 cycles out of their product?

Following those vendor charge specs is fine, **if** you only want the cycle lifetimes they advertise. You can however get **much** longer life by looking at the vendor's SoC vs Voltage chart, and "avoiding the shoulders" at both ends; stay well within the smooth, low sloped parts of the curve.

My usual stop-charge voltage setting for charging LFP is 3.45Vpc, which for 4S "12V" packs = 13.8V max. Note that is usually at an amps rate of .2 - .3C (.3C = 30A charge current per 100AH bank size). At higher rates, to shorten ICE run-times, it is safe to go to 3.5Vpc / 14.0V.

Note even at the "low" max charge voltage, letting the charge source continue to "push" even low currents long **past** the endAmps point is over-charging, in the sense of reducing potential lifecycles.

At low charge rates, as with many solar setups, under say 0.2C, I reduce termination voltage down to 3.40Vpc / 13.6V.

If your charge rate is **very** low, below say 0.05C, or very variable above and below that ballpark, then you are in real danger of overcharging - even at that seemingly low voltage of 3.40Vpc / 13.6V. In fact, monitoring voltage while charging will not even be meaningful wrt SoC.


For daily use cycling, best and simplest is to use a 0.2 - 0.3C rate, and "just stop" charging when your end-point voltage is reached. A long Absorb stage is holdover "lead thinking".

Summary:
Stop at 3.45Vpc / 13.8V for 4S, for amps rate of .2 - .3C.

At higher rates, to shorten ICE run-times, it is safe to go to 3.50Vpc / 14.0V.

At **very** low charge rates, as with many solar setups, back off to 3.40Vpc / 13.6V and do not hold Absorb / CV at all.

Originally Posted by aybabtme

The circuit includes manual switches for both the charge and discharge bus bars. I feel like that's enough. I can turn off individual charging devices in VenusOS. Could probably add a hardware switch on top of it if the desire arises.

Manual assumes you're always watching.

If the bank is expensive, more than one layer of automated protection is justified, especially against overcharging when unattended (busy sailing, cooking whatever)

[john61ct]

This thread is long but very informative
http://www.cruisersforum.com/forums/...nks-65069.html

make sure to give both Maine Sail and Ocean Planet your close attention.

Also MS' summary notes here
https://marinehowto.com/lifepo4-batteries-on-boats

**Everything** at that site is worth reading, very valuable, feel free to make a donation to help with those expenses https://marinehowto.com/support. He also has great articles in Practical Sailor.

[aybabtme]

If you look at the diagram, the manual switches are after two large normally open contactors to shut off charging or discharging, controlled by the BMS. So there's both manual and automatic disconnects, on top of a main fuse at 600A (2C).