LiFePO4 Batteries - Okay Tear Me Apart ;-)

[hellosailor]

"I got my AGM batteries (160AH) at $140 total "
gecko, that's dirt cheap, even for three or four years ago. At today's lead prices, I'd expect to pay double that.

[seandepagnier]

Well I think, for some people, lead acid is still overall much cheaper because they are more common and I can get better deals because of that. But lithium iron still seem very interesting.

[s/v Jedi]

My Trojan L16H's do 500 cycles at 80% depth of discharge but double that to 1,000 cycles at 50% depth of discharge. If the LFP is happy with the 80% for at least 2,000 cycles, you only need half the Ah battery capacity and at twice the cycles, the LFP can be 4 times the price and still beat LA.

cheers,
Nick.

[electric1]

Guys, as I am trying to better understand the use of marine batteries to design simple and affordable LFP/BMS drop-in replacement, I have some questions. I might be generalizing some stuff and making it too simple, but bear with me please.

When you run the alternator from gas/diesel engine you have 2 scenarios:
1 - engine runs for propulsion and alternator output is charging the battery bank. What happens when battery is full and not taking any more charge? Do you somehow disconnect the alternator or just let it run since its not putting any load on the system since voltage between alternator and the battery is equalized? I suppose since the engine is running for other reasons, alternator is feeding the loads and spares the battery work, right? For this reason I assume it not wise to ever disconnect alternator from the circuit?
2 - engine is started for the sole purpose of running the alternator to charge the battery bank. How do you know when to stop the engine? Assuming at some point charging current will reduce to a point of diminishing return once the bulk charging is done and you are mostly wasting fuel on finishing stage. How do you determine that point when engine should be shut off? Is there any automation or is it manual process? I suppose that load level is part of that decision since alternator offloads some work from the battery?

To summarize my questions - are there times when alternator is physically disconnected from the bank, either by field switch or main switch or both?

Biggest concern people had with my diagram is the fact that alternator diodes can be damaged if contactor is open while alternator is running ( although logically this is not possible under normal circumstances since BMS should not trip when battery is charging with 14.5V sources ). So, my thought process is to keep alternator permanently connected to the battery and have contactor positioned such that it only breaks the load circuit, not the charging circuit.

I also have an idea to introduce 2-3 minutes time delay between LVC buzzer alert and opening contactor. This would give you time to go below the deck and start the engine/alternator without losing power. Since your loads are usually less than 1C and LVC is set to somewhat conservative levels, there is time to act on LVC alert without damaging the battery.

Your thoughts and feedback are appreciated.

[s/v Thea]

Electric1,

I think most boats have the alternator connected to a bank all the time.

Depending on the setup may be more difficult to break only the load, I think interrupting the alternator field with a a parallel relay is a better way of going.

Personally I don't think the LVC is a big worry, at 25C your manual says it interrupts at 2.6V or 10.4V on bank, no one should let the batteries get that low, if they do they don't have any capacity left in them anyway.

When running the engine just to charge, I normally figure charging to 85% or so, run them down to 50%, so using 35% of batt. cap, if the LPF cells can be charged quickly to 90% and run down to 20% that's double the capacity of the LA.

My biggest concerns are:
1) How long will the batteries really last?
2) Can they really be charged from 20% to 90% at constant rate of about 0.3C with LA chargers/regulators?
3) Will they be damaged with keeping a long term float voltage of about 3.3V/cell on them?

Doug

[s/v Jedi]

Hi Electric1,

1. charging while motoring: 99% of the boats that would consider LFP bank have an external smart regulator that will switch to float charge when the batteries are full. The connection between alternator output and batteries is never disconnected.

2. for this scenario of running propulsion engine just for charging batteries, we're talking about a smaller group and I would estimate it at under 50% of potential LFP buyers. The rest will use other forms of charging like wind generator, solar panels, genset or a combination of those. Charging is often stopped at 80% SOC as indicated by the battery monitor or if they don't have that by the charge current (that gets low).

So, the alternator is never disconnected from the batteries. Your BMS should drive a signal into the smart alternator regulator to cut alternator output, or, if that function isn't available (I think every regulator can do it but I didn't play with every one), drive a relay that interrupts field current. The Ample Power and Balmar regulators have the on/off input and I would think that is the majority. There's a group that uses a Xantrex regulator but I don't know that one.

You should not break the load circuit at HVC, only at LVC. A buzzer warning ahead of time is nice for those who do not have a battery monitor which I think is not going to happen because everyone willing to put $$$ in LFP already has a good monitor that can sound a low voltage alarm.

No, the thing that worries me is the balancing/shunting in combination with not fully charging the bank. I understand that this balancing mechanism only works when fully charging.

cheers,
Nick.

[hellosailor]

Electric, I would say that your alternator is both always and never connected to the batteries and ship's mains.

Always, physically. Never, in the sense that the connection to the battery/mains does't really matter because that connection is always UNDER REEGULATOR CONTROL and if the regulator says "No output" then the output spools down to zero--even though the connection is always there and never physically broken.

A proper battery management system (BMS) is a microprocessor controlled system which is constantly looking at amperage (in and out) and voltage, at charge state, at alternator output, at load, at every piece of information it can get. In fact if you look at laptop computers a number of them can actually tell you the number of charge cycles that have been inflicted on a battery, as well as the percent of original power the battery is still capable of, among the other information they have gathered and are using.

You won't be building a "battery management system" as such, unless you're an electronics whiz designing one. A charge controller, maybe. More likely, a smart regulator is what you need, coupled with and sized to match a proper alternator.

[Intentional Drifter]

I think Nick is absolutely correct. The market segment is liveaboards and long distance cruisers, not week-end warriors and folks that tie up in a marina every night. That segment will already have a reasonably sophisticated battery monitor system. They are also very, very likely to have charging sources other than their propulsion engines, and perhaps more than one. Gensets, solar panels (with MPPT) and wind generators will be common, if not typical.

The other market segment will be those people doing refits in order to do liveaboard and long-range cruising. They will be designing new systems and will want to maximize capacity, minimize weight and complexity and integrate their system.

Charging from the propulsion engine will mainly be happening when motor sailing or just motoring. They will try and avoid running the main engine just for charging, but would do so in the event of a failure in some other part of their charging system.

The "traditional" model when cruising is to live between 50% and 80%, so we have big banks. Being able to reduce the size of those banks would be a big plus for lots of boats, especially multihulls where weight is always the enemy.

The idea of being able to double my daily usable amp/hrs, while simultaneously halving my charging time, and reducing the weight to one-third of conventional, is very attractive and would likely get a fair number of customers.

As a vendor (and, please, go ahead and change your status to reflect that), one of the things that you could do would be to offer plug & play battery and management systems that would play nicely with the other charging sources. I'd even suggest that you put packages together, pre-configured for the most common scenarios.

Being in Tampa, you're geographically convenient and close to where lots of business would be possible.

ID

[hellosailor]

Short article in today's newspaper, some company believes they've found a way to increase the # of cycles in a conventional LiOn battery by some huge amount by a simple process involving tin coating or treatment to the elements. I don't recall the specifics, I read it pre-coffee, but if they're right we'll see conventional LiOn batteries with more than 10x the current charge cycles at nearly the same cost--real soon.

Of course for the past 100 years everyone has been saying "real soon" and "huge increase" for batteries while not much has really changed.

"The market segment is liveaboards and long distance cruisers," That's still a tiny tiny market, sort of like "Who owns a navy blue 1973 Dodge Duster with a slant six that still runs?"

Which means hand bespoke systems at custom built prices.

[Intentional Drifter]

Yeah, I wouldn't be planning for 4-figure production runs quite yet.

ID

[electric1]

Originally Posted by hellosailor

A proper battery management system (BMS) is a microprocessor controlled system which is constantly looking at amperage (in and out) and voltage, at charge state, at alternator output, at load, at every piece of information it can get. In fact if you look at laptop computers a number of them can actually tell you the number of charge cycles that have been inflicted on a battery, as well as the percent of original power the battery is still capable of, among the other information they have gathered and are using.

You won't be building a "battery management system" as such, unless you're an electronics whiz designing one. A charge controller, maybe. More likely, a smart regulator is what you need, coupled with and sized to match a proper alternator.

You are absolutely correct, if I was to start from scratch. However, that is not my intention. There are plenty of big companies with well established products designed for Lead Acid batteries. There is no sense in competing with them at this time. My idea ( well, not my original idea clearly since companies like Genasun already exist ) is to create simple, reliable and affordable drop-in kit to swap existing LA banks without touching anything else and with minimal learning curve for the cruiser.

Right now is the time when LFP is well established and proven to have significant benefits over LA, yet big companies are still playing wait and see games since battery market is so dynamic. Few years from now big companies which sell LA today will switch over to LFP or some similar technology and will flood the market. Yet, clearly based on how active this thread is , there are people today willing to spend a little more on LFP to get the benefits today. Like I said before, I am trying to create a Honda solution, not a Cadillac.

It just so happens ( pure dumb luck AFAIK ) that 4 LFP cells fit perfectly into 12V Lead Acid operating range from all angles, while providing huge benefits at reasonable cost. So, why would anyone want to reinvent the wheel and retrofit all those existing systems when you can just drop-in LFP kit with simple integrated BMS and enjoy immediate benefits?

BTW, I sent a request to moderators to change my status to a vendor.

I am also ordering sample 200AH LFP cells to build a prototype of a drop-in kit and hopefully will be testing it with assistance of a very experienced cruiser here in St.Pete area.

I will post my progress as I go along. Thanks everyone for excellent feedback!

[electric1]

Originally Posted by s/v Jedi

You should not break the load circuit at HVC, only at LVC. A buzzer warning ahead of time is nice for those who do not have a battery monitor which I think is not going to happen because everyone willing to put $$$ in LFP already has a good monitor that can sound a low voltage alarm.

No, the thing that worries me is the balancing/shunting in combination with not fully charging the bank. I understand that this balancing mechanism only works when fully charging.

cheers,
Nick.

HVC subject keeps coming up, but I just can't understand why. Its is not possible to overcharge LFP battery with equipment built for LA battery. There will not be HVC event, ever.

I understand your point of most people used to treating their LA banks well and they should never encounter hard LVC event either. In fact the BMS is mostly an insurance policy to protect the battery from neglect and abuse and under normal conditions BMS will never do anything at all, but it must be there to cover the seller's warranty claims. That is why I want BMS to be as simple and affordable as possible. I have seen plenty of fancy and expensive BMSs in EV world that are absolute waste of money.

Your don't need to worry about balancing much because balancing only matters when you use the battery close to steep edges. What is balancing anyway? Its a method of establishing a base line of SOC. Since LFP curve is flat, you can't accurately judge SOC between 30%-70%, so we use balancing to ensure that all cells are on the same SOC level. You can only do it at the top charge level or bottom charge level. Doing it at the top is easier and less risky, but its mostly done once when you get new cells and later its done by very small increments at the top of the charge. If you never use the battery outside of 20%-90%, then you don't need to worry about balancing. If for some crazy reason your 4 cells drift apart too much ( there is zero evidence of that in real world ), then BMS will start tripping LVC prematurely ( since one cell will be lower than others ) and that will bring your attention so you can service the bank, either yourself or paying electrician to do it for you. This is done easily by charging individual cell with 0-4V bench power supply. I have done it when marrying new cells into older packs, its a piece of cake.

[s/v Jedi]

I just checked the Genasun marine product and found it is flawed. It will break the connection between AC charger and battery pack in case of HVC which will result in expensive damage to the charger (diodes will blow).
Instead, a HVC should turn the charger off, disconnect solar panels, switch wingen output to dump load and interrupt field current to alternators. Nothing more, nothing less.

I understand the problem in that every charger is different and some will not allow a remote shut-off signal. In that case one could insert a relay on the AC input of the charger and switch it off that way. If you disconnect charger output while charger is active, diodes blow just like with the alternator. I tested this (fuse blew, Freedom30 ruined).

HVC subject keeps coming up, but I just can't understand why. Its is not possible to overcharge LFP battery with equipment built for LA battery. There will not be HVC event, ever.

When all is okay, this won't happen, but we don't need the BMS for when all is okay; we need it when something is wrong. When we charge 4 LFP cells in series with 14.2V and one cell develops a short, the other cells WILL cause an HVC event while charger is at high output with a lot of damage as the result.

On balancing: Genasun claims on the fly balancing with 14.2V charge voltage for a 12V system. This translates to 3.55V per cell and I have the feeling that this is too low for balancing.

If for some crazy reason your 4 cells drift apart too much ( there is zero evidence of that in real world ), then BMS will start tripping LVC prematurely ( since one cell will be lower than others ) and that will bring your attention so you can service the bank, either yourself or paying electrician to do it for you. This is done easily by charging individual cell with 0-4V bench power supply. I have done it when marrying new cells into older packs, its a piece of cake.

See, you don't understand it yet. When there are just some Indians peddling their cayuko's around your boat, there is not going to be any electrician you can pay to fix it.
I agree that by following a well defined & written procedure, cruisers will be able to fix it themselves if they have that power supply aboard... which they don't have now. I also think a DC-DC converter would be much better than a bench PSU for a boat.

What I do like about the Genasun product is the "two batteries" setup. That might actually be their solution for interrupting charge circuit without blowing diodes because each battery has it's own set of relays on input/output.

I've been studying LFP & available solutions for the past days and I am convinced much better products for using it on a boat will come. I think I'll jump in and buy some cells to play with.

cheers,
Nick.

[Tinkerer007]

Lots of good information in this thread. Thank you.

I ordered 4 200 AH cells from EVComponents in February. They arrived last week, at which time I ordered the Mini BMS from Electric1, which arrived today. Now to install and start bench charging. I will have to figure out how to get to full charge. Maybe the equalization option on my Xantrex charger.

I decided to get the LiFePo4 batteries as I am racing to Hawaii in July, and the LA batteries I have are weak, and not sealed. I have an Olson 40 that was build in the early 80s. I am looking for a "drop in" solution to replace existing batteries.

It is interesting how you think you understand your electrical system, and then when you go to change it, all sorts of questions come up. This thread has been great at bringing up additional things that need to be looked at. I hadn't thought about the alternator being disconnected abruptly while charging. And just where exactly does that charging wire from the alternator hook into the system.

I think we should add racing boats to the list of users of this technology. They really care about weight.

It will be fun to get this installed and start watching charge and discharge usage. I have a link 2000 system, and a DME to watch these things.

I am in the SF Bay area if anyone wants to come kibitz during install.

[s/v Jedi]

Hi Tinkerer... your 1st post here, welcome!

You are right, racers are a group that will all switch away from LA.

Your link2000 allows you to change CEF and Peukert settings. Start by setting both to 1.00 to effectively disable these corrections to it's calculations. I think CEF will need sone other value but my current info varies between 85% and 100% efficiency which isn't accurate. You could test this easily by discharging until LVC and noting the link2000 reading for Ah at that point. Next you can calculate CEF so that the monitor will show empty when LVC occurs. The monitor will probably still show 20Ah available at that point in which case a CEF of 0.9 should do the trick.

Yes, I think the equalize charge feature is the way for a full charge. A lot of chargers/regulators/controllers allow you to set the voltage for equalize mode.

cheers,
Nick.