Lithium Batteries (for the rest of us)

[Lagoon4us]

Originally Posted by roetter

Zanshin may mean the 'intelligence' that a battery monitor has. I.e. 'knowing' when the batteries are full (low amps going in and over a certain voltage), accumulating the amps going out over time am showing them as Ah out, calculating percentage of charge and estimate time remaining, utilizing Peukert and so on.
That is just a 'simple' battery monitor that everybody, lithium or not must have.

Absolutely agree a monitor is necessary hence we use a BEP unit as well as a solar monitor, however i don't agree an automatic management system is needed for four cells and i say that from practical experience.
Cheers.

[OceanSeaSpray]

Going to lithium batteries and charging "gently" without the option of safely equalising would mean that cells will become imbalanced over time without a lot of options to rectify the issue and performance would drop severely. All the benefits from high charge rates would be gone too.

On the other hand any kind of fast charge requires active charge termination. In the case of sealed AGMs (most common solution), overcharging of individual cells results in some gassing and recombination and no further issues.
Lithium chemistry elements are different, they must not be overcharged and don't tolerate it well. Some of the failure modes can be very unsafe indeed. For this reason, charging and monitoring is performed on a per-cell basis. It would require quite a complex charging circuit with several isolated outputs to feed different currents and implement different cut-offs for each of the cells. This means individual power converters etc.
As long as the circuit doesn't perform sharp disconnects and the nominal voltage of the bank is close enough, then a standard alternator should be able to feed it. One issue is that the voltage on board could rise very significantly during a fast charge, potentially more than some of the equipment can typically handle. High charge rates require high cell voltages as charging gets near its end.

So it is technically possible, but is there really a gain to be found besides light weight?
AGMs can also take very high rates of charge as long as they are indeed charging: technically you can fully or near-fully charge an AGM in an hour or so as long as it is kept from gassing, without the issues and complexity arising with lithium. It is just that manufacturers don't recommend charge rates that can exceed the ability of the battery to recombine gases due to risk of shortening their life etc, i.e. about 0.3C, but algorithms have been developed to charge AGMs at 1C and over.
The issue with doing it properly is with active charge termination and monitoring again. If you are blindly pouring a steady 140A into a 100Ah AGM and something is preventing the battery from absorbing the charge, the power has to go somewhere else and the outcome could be unhealthy.
Still, AGM is a much easier and more tolerant technology to manage because it suits charging elements in series, some overcharging doesn't cause damage due to gas recombination and it is easy to equalise as a result.

At the moment, I think lithium has its place in applications where weight and energy density are problematic and then a great deal of power electronics need to be added to manage the cells and system voltage during charging. This is not affordable yet for large banks.
On board for cruising, I don't really see the value. I would rather put the effort into implementing a safe fast charge system for AGMs. This is a lot easier and cheaper and it would deliver just about the same in the end, just heavier.
Fast charging large banks is all very well on the paper, but one also needs to find the current and voltage to do it. Besides I don't know of very many battery devices that fast-charge in less that about 1 hour, regardless of battery chemistry and sophistication of the charging circuit... that's a charge rate of 1C more or less. Faster doesn't seem to be sustainable.

[familycruisers]

Think Dewalt charges their batteries in less than an hour, and have great cycle life. Oh wait thats lifepo4. Half the weight, half the volume, thirty percent more usable capacity . Hobby industry has the charging chemistry including cell monitoring, charger termination, and balance figured out already. Adapting would not be difficult.

[CharlieJ]

OceanSeaSpray-
I have to disagree with your entire post.

Going to lithium batteries and charging "gently" without the option of safely equalising would mean that cells will become imbalanced over time without a lot of options to rectify the issue and performance would drop severely. All the benefits from high charge rates would be gone too.

Since house banks we charge and discharge at fractional C rates, by actual, real world experience, the cells do not drift and stay remarkably balanced. A top balance every year or so may be warranted.

Lithium chemistry elements are different, they must not be overcharged and don't tolerate it well.

This is true of AGMs also.

For this reason, charging and monitoring is performed on a per-cell basis. It would require quite a complex charging circuit with several isolated outputs to feed different currents and implement different cut-offs for each of the cells. This means individual power converters etc.

This is just plain misleading. A simple BMS will terminate charging prior to hitting the high voltage knee and it monitors the cells that are paralleled as a single cell.

As long as the circuit doesn't perform sharp disconnects ...

The prudent designer does not disconnect the output from the alternator to terminate its charging. A $4 relay in the B+ to the alternator regulator does just fine.

...then a standard alternator should be able to feed it.

Only if you have a means of reducing the voltage to somewhere in the neighborhood of 13.8VDC (3.45VPC). And, since the internal resistance of LFP's is so low, the load looks almost like a short to the alternator so it will put out max current for the whole cycle. Most stock, automotive type, alternators cannot operate continuously at max output.

One issue is that the voltage on board could rise very significantly during a fast charge, potentially more than some of the equipment can typically handle.

A simple BMS will open a charging contactor if the native control features of the charging device (solar controller, inverter/charger, charger, alternator, etc.) fail to control the device.

High charge rates require high cell voltages as charging gets near its end.

Not so.

If you are blindly pouring a steady 140A into a 100Ah AGM and something is preventing the battery from absorbing the charge, the power has to go somewhere else and the outcome could be unhealthy.

And how does one accomplish this? Ohm's Law remains has not been revoked! A LA battery will accept only as much charging current as the charging voltage and internal resistance of the battery dictate.

...and it is easy to equalise as a result.

Most AGM manufacturers specifically advise against equalizing!

I have designed three LFP bank projects; two have been deployed and are working exceptionally well. The third project is a 2,000Ahr bank comprised of 100Ahr LFP cells in a 10P8S configuration. We have been testing them in the shop for the last couple of weeks and are more and more impressed by this technology. With a Peukert exponent of about 1.01, reduction in size, reduction in weight and absolutely no maintenance, LFPs are to LA batteries what fiberglass was to wood in boat construction.

I suggest you read the other LFP threads on CF and note the real world experience of T1 Terry, DeckOfficer, MainSail, ebaugh, Lagoon4Us in designing, testing and using LFP house banks.

[Maine Sail]

Originally Posted by OceanSeaSpray

Going to lithium batteries and charging "gently" without the option of safely equalising would mean that cells will become imbalanced over time without a lot of options to rectify the issue and performance would drop severely. All the benefits from high charge rates would be gone too.

On the other hand any kind of fast charge requires active charge termination. In the case of sealed AGMs (most common solution), overcharging of individual cells results in some gassing and recombination and no further issues.
Lithium chemistry elements are different, they must not be overcharged and don't tolerate it well. Some of the failure modes can be very unsafe indeed. For this reason, charging and monitoring is performed on a per-cell basis. It would require quite a complex charging circuit with several isolated outputs to feed different currents and implement different cut-offs for each of the cells. This means individual power converters etc.
As long as the circuit doesn't perform sharp disconnects and the nominal voltage of the bank is close enough, then a standard alternator should be able to feed it. One issue is that the voltage on board could rise very significantly during a fast charge, potentially more than some of the equipment can typically handle. High charge rates require high cell voltages as charging gets near its end.

So it is technically possible, but is there really a gain to be found besides light weight?
AGMs can also take very high rates of charge as long as they are indeed charging: technically you can fully or near-fully charge an AGM in an hour or so as long as it is kept from gassing, without the issues and complexity arising with lithium. It is just that manufacturers don't recommend charge rates that can exceed the ability of the battery to recombine gases due to risk of shortening their life etc, i.e. about 0.3C, but algorithms have been developed to charge AGMs at 1C and over.
The issue with doing it properly is with active charge termination and monitoring again. If you are blindly pouring a steady 140A into a 100Ah AGM and something is preventing the battery from absorbing the charge, the power has to go somewhere else and the outcome could be unhealthy.
Still, AGM is a much easier and more tolerant technology to manage because it suits charging elements in series, some overcharging doesn't cause damage due to gas recombination and it is easy to equalise as a result.

At the moment, I think lithium has its place in applications where weight and energy density are problematic and then a great deal of power electronics need to be added to manage the cells and system voltage during charging. This is not affordable yet for large banks.
On board for cruising, I don't really see the value. I would rather put the effort into implementing a safe fast charge system for AGMs. This is a lot easier and cheaper and it would deliver just about the same in the end, just heavier.
Fast charging large banks is all very well on the paper, but one also needs to find the current and voltage to do it. Besides I don't know of very many battery devices that fast-charge in less that about 1 hour, regardless of battery chemistry and sophistication of the charging circuit... that's a charge rate of 1C more or less. Faster doesn't seem to be sustainable.

Your post is so off kilter I can't even begin to address each and every point....

FYI many of us, including myself, are using LiFePO4 in the real world, on real boats and do not see the issues you allude to at all.

My bank cost about the same to build as buying the same Ah capacity in Odyssey TPPL AGM's but it charges to FULL significantly faster, holds voltage significantly better, and our equipment LOVES it. At 80% DOD they are also rated at over 2000 cycles and the Odyssey AGM's at just 400....

Our cells have never once gone out of perfect balance, out to the thousandths of a volt using a Fluke DVM, and are still in perfect balance after approx 150 cycles... That 150 cycles is over 1/3 of the useful rated life of an Odyssey AGM.. I cycled them nearly 60 times before even installing them on the boat doing multiple experiments and deciding upon the charging parameters I wanted to use.. While some have been shallow cycles we have generally deeply cycled them, on purpose, because I want to see how they do....

As a marine electrician I have seen and been through the trials and tribulations of AGM's and they can not even come close. AGM's will only take the high current until they begin to sulfate, then they act more like flooded batteries.

I have a bank of four 6V Lifeline's on my bench right now that come up to 14.4V with just 50A of applied charge current at 50% SOC. It takes a very long time to charge them to 100% when they come up to 14.4V nearly immediately and with just 50A of charge current. This should not happen in a few years but very often does if you don't treat them with kindly..

They should take a heck of a lot more than that at 50% SOC before coming up to absorption voltage but they are now a few years old and have suffered from sulfation. The owner never knew about charging back to 100% and equalizing them to minimize the effects of sulfation. This is a very common theme I see and deal with on AGM's.. Lifeline AGM's can be condition charged, but most can not. The bank on my bench is going though that as I type.. AGM's also suffer from declining acceptance in the higher SOC range making getting back to 100% darn tough on a cruising boat and limiting your useful Ah capacity...

Our LiFePO4 bank charges easily back to 98-99% SOC and the last 2% takes about 15-20 minutes. We actually have tough time cycling this bank and very often have our alternator and solar system physically turned off because they charge so fast and I want to cycle them deeply...

[OceanSeaSpray]

This is not just in response to your particular comment, so please bear with me.

Being successful in some instances doesn't address the intrinsic issues around charge control. There is no doubt that one can achieve deeper cycling and fast charge rates with LiFePO4, but the question is how long is it going to last and what is going to happen? Where does charge control come from?
The fact that you haven't recorded any imbalance after that long is probably just an indication that you are "pushing" past the end of charge point. If you get away with it, it is great. It doesn't mean you will forever, or that it is the right thing to do.

Going back to AGMs, there is no such thing as "not overcoming Ohm's law". If you want to fast charge a battery that sees its internal resistance increase as charge builds up, you need to increase charging voltage to maintain the charge rate and - no - it won't cause the battery to gas as long as it is still absorbing the charge. But you need to drop that voltage again once charge is complete. That is the point of active termination. You may also want to have a read at
http://www.powerstream.com/SLA-fast-charge.htm, these guys build chargers.
You are achieving fast charge rates at "low" voltage with a LiFePO4 bank because it seems to have lower internal resistance throughout, that's all. It just creates a bigger challenge when it comes to determining when to stop. Yes you can try using an energy monitoring system and replace the missing capacity, but this works best in theory. The characteristics of the battery change as it ages, the calibration is never perfect and things drift.
A BMS is not a way of charging a battery in itself if it operates based on capacity. You need to go back to measurable properties, i.e. voltage, current and temperature. That's the way modern lithium batteries are charged, but cell-by-cell. Also a BMS designed to work with lead-acid technology would be completely unsuitable to terminate the charge of lithium elements. They don't respond the same.
In the toys and tools world the technology is everywhere indeed and it is not "difficult" to extrapolate, but it is not that easy on the kitchen table and it is very costly. Power electronics that handle a few amps are cheap, the ones handling hundreds of amps are not and additional complexity arises. If most cars ran on LiFePO4 banks, there would be cheap components and batteries available to use on boats, but it is not the case at the moment.
Very fast charge rates, below the hour, come at the expense of battery life, because the elements get hotter. What you do with toys is not automatically what you want to do with a large bank.
AGMs do suffer from sulfation when they are not recharged properly, it is a known problem, and in cycling applications they are seldom recharged past 80%, which is why they sulphate and people experience poor life. Now, if you bulk charged them properly as I propose, a lot of these issues would go away.
AGM manufacturers sometimes advise against equalising because they advise against anything that can cause the recombination rate to be exceeded - i.e. loss of water - but I know of at least one instance where Lifeline specifically advised doing exactly that to combat sulfation issues that were killing costly banks every 12-24 months only.
I have designed and built a number of charging systems, regulators and electronics over the years and I have seen many more, full of gadgets and BMS, that failed to perform.
The most common problem with marine battery charging is that the charge controller can't accurately track the current going in, built-in alternator regulators typically have no idea (most "smart" regulators have no external current shunt and no idea either), and the charge voltage is not allowed to rise high enough while the battery is still absorbing, because they don't know when to terminate the charge.
The best charging systems I know of are for large stationary solar installations. They use individual flooded 2V cells a lot of the time and top-notch charge controllers, like nothing available in the "marine" range. They go back to first principles and cut no corners.
And yes, getting full output out of an alternator for long periods of time is a challenge. Some units, like some Electrodyne alternators I used many years ago, were rated "hot", but most are not. That's a different issue.
I appreciate that LiFePO4 batteries look very attractive, but I wouldn't go as far as advocating the generalisation of an empirical experiment without being aware of the underlying issues. The size of your alternator in relation with the size of the bank and running time you use all contribute to hold it together.
Replicate it 100 times with some variation, and it could look very different. Sell that principle and you could face very serious liabilities, you are outside all specifications etc. It is similar, but not as severe, if you bulk-charge your AGMs at 1C without the manufacturer's blessing.
I have seen people doing it, they added a switch over their alternator regulator! It made most of their battery problems go away. Should you do that? No. Should you build yourself a regulator that does it safely? Maybe if you know what you are doing, you are careful and keep an expert eye on what is happening.

[Lagoon4us]

Totally agree with post's 183/184/185, i'll also add i believe Terry has had his similar system nearly 2 years without an issue other than self made ones.

We have a lead acid battery on the charge circuit from the engine alternator (only on this circuit) we never see high voltages nor crazy charge amps the alternators regulator is sensing the lead acids resistance and it behaves beautifully.

Typically the 'theoretics' with a barrow to push hypothesise against those in the real world enjoying, trialling and experimenting with this 'different' system.

Past experience has shown they usually are linked to the other side ie LA's AGM's etc.......

[T1 Terry]

Hello Ocean Spray,
I don't know who has been feeding this nonsense but very little of it is correct. I think it probably comes from your confusion about different chemical make ups of different types of lithium ion batteries, they are not all the same. Tools and model planes don’t use lithium ferrous batteries so any knowledge you are attempting to transfer across is a waste of time and not applicable.
I now have around 50 systems I've had a hand in setting up 3 of them are into their 3rd yr 24/7 lived off house batteries, none of them have shown anything like what you are quoting as fact so some of your test data would be interesting to see how the problems you discovered occurred.
The highest I've been able to push a 90Ah battery charge to so far has been around 180 amps (3CA), the 2/0 welding cable started smoking so I stopped. The battery was down to 2.8v per cell, fully drained, after 15 mins I stopped the charging because of the cable smoking already mentioned, the 90Ah battery was holding 3.5v per cell while charging but dropped back to 3.35v per cell as soon as charging stopped, I finished charging with a 40 amp charger, at 3.45v per cell there was a 50mv difference between the highest and lowest. I wouldn't recommend anyone charging their system that hard, I just wanted to see what happened, the batteries or cells were not the weak link, the cable over heating was.
As far as charging direct from an alternator, it cost me a number of fan belts and a dead alternator ended the fan belt torture, I would not recommend direct charging from a standard automotive alternator. Unless they were especially built for 100% full output duty cycle with multiple fans and access to cool air, the alternator will fry, some form of current limiting is required if you wish to use a standard automotive alternator. A simple relay from the lead acid start batteries and a special purpose regulator with temp sensing to limit alternator output to control temperature is a good one, feeding an inverter from the start battery and charging with a mains charger rated at around 50% of the capacity of the alternator works well too. Most of our system that incorporate alternator charging use the inverter and a 40 amp charger, combine with the solar that seems to be plenty to recharge the house batteries.

As far as a BMS, the letters stand for battery management system, nothing more, if you check the battery voltage and stop charging when it’s full and stop discharging when it’s empty, you have a battery management system. All batteries have some form of battery management system, some mob thinking they can make a $$ selling BMS systems to those that don’t know they don’t need one are what is causing a lot of the angst wit these batteries. Most of these systems were designed on someone’s garage bench, a flash website and fancy packaging does not make a good product, a $14 Junsi cell logger will do the same job as long as you can hear the alarms and turn of either the charging or discharging as required, a smoke alarm attached makes a hard to ignore alarm. A battery monitor will make the end of charge and end of capacity a lot more predictable but cell voltage is what it’s all about, stop at 3.6v and 2.8v and you won’t damage the cells, it doesn’t get much easier than that.

T1 Terry

[noelex 77]

Originally Posted by T1 Terry

Tools and model planes don’t use lithium ferrous batteries so any knowledge you are attempting to transfer across is a waste of time and not applicable.

LiFePO4 Battery are used in some model applications and in tools. The chemistry of of the various manufactures is slightly different, but there is a lot that can be learned from the modelers experiences who have been using LiFePO4 for the very early days.

Mostly its very good news. These batteries are very robust and safe coping well with high charge and discharge rates.

There is no doubt in my mind that is the future of marine battery systems. I think it will take another couple of years before they are suitable for the electrically illiterate. It will also take longer before we see the full advantages engine alternators with an output that is appropriate for lithium batteries for example, but there are still many benefits without these developments.

The one caution I would add is the overall life of these batteries. They are significantly better than LA, but I think many advocates are getting carried away.
My own experience with small LiFePO4 cells is that the lifespan is reasonably age dependent. I think many people are extracting the cycle life and not taking the ageing into account. I have my doubts that some of the optimistic predictions of battery life will be close to reality, especially for the majority of yachts cruising the warmer parts of world.

I also feel many users are still applying lessons learnt from LA batteries that are not applicable to these new chemistries. In particular keeping the batteries at 100% charge. This is good for LA, but lithium batteries are less happy at this high state of charge.

[OceanSeaSpray]

No guys, I have nothing to do with the LA or SLA industry and I don't like lead. As an engineer I am just a little worried when I see lithium cells charged at high current in series basically. Some success is not a demonstration of adequacy, it only shows the potential of the technology.
Someone was referring to tools and toys in relation with charge rates and LiFe technology is indeed getting into that sector. Regardless, it is not the point.

In all your voltage figures, you are referring to V/cell, once you have put elements in series, you are relying on the health of the cells for the voltages to distribute properly. Once the cells start getting old, there is no guarantee that it will still happen nicely and the outcome could be an accident. A lithium fire on board would be pretty hot and very hard to put out.
It is interesting to see that you are one of the people who bypassed regulators. Nothing wrong with experimenting, I have experimented quite a lot with my own solar regulators while living aboard, but there is a step between an experiment and a solution. You could have added a resistor in the field circuit and that would have kept the alternator within its continuous rating.
All I am saying is "be careful" and - if you are prepared to step outside of the perimeter anyway - "you could get much improved results just out of AGMs at lesser risks".

When you refer to using an alternator to feed an inverter to feed a charger to charge the battery... well that's a lot of conversion to get back to the starting voltage isn't it? I don't like seeing things like that, not to mention the complexity, failure points and losses. But fair enough, thanks for sharing it.

I am working on a controller to optimise charging efficiency over 3 banks from a solar source, which means MPPT followed by 3 independent regulation algorithms. I plan to use a strategy where I track the state of charge by calculating the internal resistance and bulk charge beyond the traditional voltage limits as long as the batteries can absorb the current. I have experimented in the past and I got much faster charge rates and batteries that did actually go back to near 100% early every day.

Anyway, one thing you could do playing with LiFe cells is monitoring individual cell voltages and tripping the alternator if any of them gets past the safe limit. Charging LiFe cells with no form of charge termination and no protections is not a good idea because it can go very wrong. The industry won't do anything with lithium cells without protective circuitry - there are reasons for that.
Now, I would love having a LiFeSO4 bank on board - safely - and I didn't mean to ignite a raging controversy.
If you have AGMs and you are not happy, you could think about scheming a way to bulk charge at higher voltage, because the technology can absorb the current. If you are experimenting with LiFe cells, keep doing it carefully, but think about building some protections too, even very simple ones.

If someone can post a link or document with the full specs for a LiFeSO4 cell they have used for a marine bank, I would love to have a look and a think about it.

[CharlieJ]

As an engineer I am just a little worried when I see lithium cells charged at high current in series basically.

Most reputable LFP manufacturers recommend a max charge rate of C/3 and for most large (>600Ahr) house banks achieving a C/3 rate is a challenge so this is not an issue.

In all your voltage figures, you are referring to V/cell, once you have put elements in series, you are relying on the health of the cells for the voltages to distribute properly.

For those of us that advocate a BMS, the paralleled cells voltages are monitored with protective circuitry that operates should one of the parallel strings go out of spec.

You could have added a resistor in the field circuit and that would have kept the alternator within its continuous rating.

An excellent "get home, McGyver-esq" solution. I would prefer to use the full programming ability of, for example, a Balmar MC-614 alternator regulator with alternator temperature sensing, field limiting, soft start, etc., etc. in my designs.

Once the cells start getting old, there is no guarantee that it will still happen nicely and the outcome could be an accident.

This is fear mongering. The same sort of comment was said about the internal combustion engine and can be said today about large battery banks comprised of LA batteries.

A lithium fire on board would be pretty hot and very hard to put out.

Are you aware of just how little Li is in a 180Ahr LFP cell? How does this miniscule quantity of Li contribute to the fuel load?

Anyway, one thing you could do playing with LiFe cells is monitoring individual cell voltages and tripping the alternator if any of them gets past the safe limit.

Thanks for the advice; however, please read #184 where I discuss a "prudent designer". This feature is already incorporated.

Charging LiFe cells with no form of charge termination and no protections is not a good idea because it can go very wrong.

Most, if not all, of the early adapters have at least a minimal BMS; e.g., Junsi Cellogger to protect against a HVE or a LVE.

Now, I would love having a LiFeSO4 bank on board

Why? How are they superior to LFPs? What is the knowledge base for this chemistry? Who is using them now? Who manufacturers them? What is their cost per Ahr? Etc., etc., etc.

and I didn't mean to ignite a raging controversy.

Believe me, you didn't. Those of us, with real world LFP experience, are simply attempting to keep the record straight.

[Saltyhog]

[OceanSeaSpray]

Charlie,

Somehow I had picked up a reference to LiFeSO4 somewhere here, which left me a bit puzzled because I couldn't find any specs for large cells, but I also see that someone managed to edit it back correctly to LiFePO4 in earlier posts including one of mine. Nice move.

This I could find decent data sheets for (http://liionbms.com/pdf/huanyu/HYP-3.2V-100Ah.pdf and others) and it is the chemistry that has been used with many electric vehicles for a few years now.
There I can agree a lot more readily. People who got into strife with charging those in series often destroyed a cell through undervoltage and then things didn't go well at all when they recharged.
I must say that 4 of those cells marry up amazingly readily in a conventional 12V system with their end of charge voltage at or slightly above the range of a standard alternator.
If you have HV charge cut-off and LV load disconnect based on the voltage of each cell or parallel group as some have done, as far as safety goes, I think you are there indeed.

Automotive alternator regulators are not ideal for charging them, instead of achieving constant current/constant voltage, you get a high current (limited by the alternator capacity worst case) tapering down and then constant voltage. That constant voltage can be too low to achieve a full charge etc, there are niggles, but reaching full charge is not the bottom line and there is no harm there. It just comes down to how much capacity is actually made available.
Ideally, these batteries still shouldn't be charged with tremendous currents if long life is the objective so it tends to go in the right direction. The credible data I found ranged from 0.1 to 0.3C (maximum allowed always much higher so a lot of margin). And a solar system is easier to set up to look after the end of charge like Lagoon4Us was saying.

If you "fixed" the charging aspects for the LiFePO4 in the alternator control, then you would need to dedicate it to that technology and you couldn't share with lead-acid starting batteries anymore. When looking into it closely, splitting over diodes between lead-acid and lithium is not perfect, but surprisingly free of critical issues. Sensing the lead-acid side is the conservative option, especially if it is the starting bank, never normally low.
Even in a very basic installation with just a stock alternator and a lithium bank, maximum charge current can be adjusted down by simply adding a resistor in the field circuit as I was alluding to.

What I am thinking now is that it would be near trivial to create a standard little board that would allow taking 4 such cells (or groups) and make a perfectly workable drop-in replacement for lead-acid that would take care of the safety aspects without any engineering and configuration. Performance considerations, like effective available capacity typically, would still depend on external factors, but it is so easy to achieve a high rate of charge, not a huge deal.

Initially I had said "why bother" because of the prospect of having to charge individual elements, but unlike AGMs, the problem of "keeping them up to keep them good" is gone and that is the biggest issue when living aboard.

Keeping balanced cells on the long run with aging etc (and I am not referring to any overcharging at the top) could still cause a few issues, i.e. tripping by the monitoring circuit, but probably nothing that can't be fixed manually. The EV guys say that cells vary in capacity right out of the box - sensible - and it gave them grief. What some did is not silly and they say that so far they didn't have to go back to it.

Has anyone looked for a difference in available capacity from new after 2-3 years, based on voltages before recharging maybe? At the far end, if no cell packs up, the bank should cut out earlier and earlier until it has to be thrown out. I tried to find lifetime figures, but the technology is still quite new for that. 6-7 years is something I dug up.

[T1 Terry]

I think the longest continuous use battery is in the Mars Rover, launched in 2003, had a hat load of issues with dodgy electrical equipment running continuously like heaters in arm joints, a few dust storms and no one there to clean the solar panels, but it's still going, you couldn't get a much harsher environment that that :lol:

LiFeP04 cells don't catch fire if they are over charged and believe me, I've tested that one out, a continuous 40 amps into 4 fully charged cells in parallel for I'm guessing 5 hrs. It turned them into a kind of bulged out box, flattened out the ribs in the casing and even the top and bottom bulged, and it was firmly strapped, till it snapped a weld and alerted me to the fact there was a problem. I'd turned the wrong switch off and the charger was still running, from brand new to rubbish in around 10 hrs, but no fire, no explosion, didn't even rupture the plastic case. the cells didn't go instantly flat or open circuit either, in fact 2 of the cells are still in use, the 2 centre cells that over heated the most slowly discharged over the following week or so, there was obviously an internal short, but not a dead short, so the discharge was gradual. If it had need to remain in service charging cycles would have required a lot more attention, but they would have remained serviceable till replacements could be obtained.

If you charge these batteries to the end voltage of the average lead acid designed alternator you will not get a long service life, Li cells charge at a lower voltage than lead acid cells

All Li systems should have cell voltage monitoring, it would be crazy not to, especially as the units as so cheap, $14 isn't a lot out of pocket for a Junsi Cell log 8

The manufacturers data sheets state, all their life cycle testing is done at 0.5C, both charging and discharging, that's 300 amps into a 600Ah battery, that's getting up there for any on board charging system don't you think? They recommend a max of 3C charging and discharging, that is way outside the capabilities of a house battery sized charging system, 1800 amps into or out of a 600Ah battery, I don't think that is likely to happen do you?

I think the problem is you have based your knowledge on using these cells as EV power supplies, a whole different world to house battery use, the same cells are about the only thing in common.

As far as LiFeS04, check the Winston battery website for info, they use them over there i the bus fleets, a bit big and bulky for boat house batteries and the cycle life isn't as good but the charge and discharge standard rate is 1C, so quite suitable for what they are using them for, but not really suitable as house batteries.

I hope that cleared up a few misconceptions you may have had about these batteries and those of us that are using them, and have been for some time.

T1 Terry
Apologies, my typing and spelling are getting worse the more I type, I thought it was supposed to go the other way.

[T1 Terry]

Originally Posted by noelex 77

LiFePO4 Battery are used in some model applications and in tools. The chemistry of of the various manufactures is slightly different, but there is a lot that can be learned from the modelers experiences who have been using LiFePO4 for the very early days.

Mostly its very good news. These batteries are very robust and safe coping well with high charge and discharge rates.

There is no doubt in my mind that is the future of marine battery systems. I think it will take another couple of years before they are suitable for the electrically illiterate. It will also take longer before we see the full advantages engine alternators with an output that is appropriate for lithium batteries for example, but there are still many benefits without these developments.

The one caution I would add is the overall life of these batteries. They are significantly better than LA, but I think many advocates are getting carried away.
My own experience with small LiFePO4 cells is that the lifespan is reasonably age dependent. I think many people are extracting the cycle life and not taking the ageing into account. I have my doubts that some of the optimistic predictions of battery life will be close to reality, especially for the majority of yachts cruising the warmer parts of world.

I also feel many users are still applying lessons learnt from LA batteries that are not applicable to these new chemistries. In particular keeping the batteries at 100% charge. This is good for LA, but lithium batteries are less happy at this high state of charge.

Granted, some of the cells used for tools and models were LiFep04 chemistry, A123 cells come to mind, but they were cylindrical cells that used much thinner coating and this enabled them to provide very high discharge and recharge rates, but the pay back is a shorten cycle life. Part of the problem is the cylindrical design, one continuous pair of plates wound in a spiral, this leads to poor cooling towards the centre of the cell and the electrolyte starts to break down.
The cells used for house batteries are prismatic cells, a different design, not designed for the same high discharge and charge rates that the cylindrical cells can achieve, but a lot more capacity in each cell. This format suits house batteries much better, they can dissipate any heat build up reasonably quickly but during normal operation as house batteries these cells generate very little heat any way. Abuse is a different story though, they get hot then.

T1 Terry