Who should be in control?

[SaylorMade]

Originally Posted by sailorboy1

Guess is that right to the point. Nothing in my LFP battery manual suggests not following the charge parameters because the BMS will take care of it.

Just to clarify, nowhere in my question for debate have I argued not to follow the charge parameters from the manual. For LFP typically just a max charge voltage.

Not in any way related to letting the BMS disconnect and end the charging when the first cell reaches max voltage.
I am yet to see any good argument for not letting the individual cell voltage be the determining factor.
To go on pack voltage is dangerous in my current understanding, especially if you do NOT balance your cells, which by the way is the first time I have heard about.

I have no idea why anyone would NOT want to keep their cells balanced.

[sailorboy1]

Originally Posted by SaylorMade

To go on pack voltage is dangerous in my current understanding, especially if you do NOT balance your cells, which by the way is the first time I have heard about.

I have no idea why anyone would NOT want to keep their cells balanced.

Well just do what you want then, lots of lithium battery owners like the science project approach.

I have no idea why you would NOT what to make it easy instead of assuming there is a problem. I have had my solar at 14.4V absorption last couple of days, which is as much "balancing" I think about and just turned it back down to13.8V.

Going to leave this thread so you can direct your worry to someone else.

[lmxr]

Originally Posted by SaylorMade

So, in a few threads here a side-topic has arisen, about wether or not the charger should be in control of when to stop charging a LFP bank.

I think it's worth having a seperate thread about this, since it seems quite important to me, and I would like to hear the different arguments.

For the debate let us assume that the LFP bank has a BMS which will let the user set the different parameters.
We will also assume that the BMS is setup with appropriate values for OVP as per LifePo4 Cells specs, and also an OVP on pack voltage.

My own point of view is that the BMS should control when to stop charging, since only the BMS knows the voltage of each individual cells, and it is my understanding that a charger can only feel pack-voltage.

The arguments I have heard for NOT letting the BMS be in charge ("first-line-of-defence) is that BMS are not to be trusted, but this seem counter intuitive to me, since we will rely on the BMS anyway to keep the cells balanced anyway, so that pack-voltage can be any safe indicator of when max voltage has been reached. [...]

I'm also of the opinion that the BMS should control the chargers. A few comments:

- 'The BMS are not to be trusted' -> yes, there should be several lines of protection. For example, the BMS and the charger's internal voltage limit.

- The end of charge should be determined taking into account the battery current (see the charging instructions of the LFP cells, e.g. EVE). Without current, one cannot tell whether an LFP battery is anywhere near full. For example, 13.8 V can be in overcharge territory at very low currents, but is not anywhere close at high charge rates (e.g. large solar array + dual alternators).

- Using the highest cell voltage instead of pack voltage is better because it avoids cell overvoltage. Yes, the BMS tries to keep the cells synchronized, but in practice the cells don't tend to be synchronized well, especially on cruising boats where the battery may only reach full for a few minutes per month. The logic 'the cells are always synchronized' does not hold well, as by the same logic the BMS also wouldn't need a balancer (if they are synchronized, there's no balancing to be done, and if there's balancing to be done, they are not synchronized). Some BMS'es - that can control the chargers - extend the absorption duration to extend the balancing time.

- When the BMS senses overvoltage, it makes a lot of sense if the BMS first tries to alleviate the situation by turning of the chargers instead of immediately disconnecting the chargers. Turning off the chargers is far more graceful and less stressing. There can be many reasons for overvoltage, but many cases are covered by turning of the chargers instead of disconnection them. The disconnect relay still serves a function e.g. if a charger goes haywire.

- It makes sense take the SoC into account when charging, in particular to avoid keeping the battery at a high SoC causing fast degradation.

So, in a decent installation, the BMS cannot do much wrong (the chargers already limit their maximum voltage by themselves), has all the required information, and already controls the chargers to shut them down gracefully in approaching disconnect situation.

So, it makes a lot of sense to let the BMS control the charging a bit more smartly!

[mcarthur]

I'd be happy with the BMS is the central "controller" - but only if it communicates to others (Cerbo, chargers, mppt etc.).
Also, the BMS OVP for cell and pack have to work well: there's some nice people testing some BMS online where they don't always function the most obvious way - or at all. OffGridGarage is decent, although not covering boat situations as much as home.
But a BMS that has just a single OVP and shuts down the entire bank is pretty useless in comparison to one that communicates an OV alarm to other systems so that they can nicely stop charging (MMPT's, chargers). This continues having the battery online for use rather than disconnecting it entirely.
Unfortunately, BMS don't seem to a) have the right functions, and b) they don't always work. However, they are probably still better than a fully distributed "stop charging" regime (noting that this is different to an "oh sh*t, something is badly wrong regime").

[wholybee]

Originally Posted by SaylorMade

I am yet to see any good argument for not letting the individual cell voltage be the determining factor.

It just doesn't matter. Using cell voltage is fine. Using pack voltage is fine. It's just that using cell voltage requires a more complex and expensive charging system.

I don't even want to venture a guess as to how many installations, not only boats, but RV's, off grid solar, etc. charge by pack voltage. 100's of thousands? Dominantly, they charge by pack voltage. It only seems to be overzealous boat owners that promote overly sophisticated charging and BMS systems.

Healthy cells will stay in balance several months with zero balancing. With minimal passive balancing, they will never go out of balance, and if they do, something needs immediate attention. That is when the BMS steps in, disconnects, and alerts the user.

If cells are NOT healthy, then an active balancer will keep them in balance, again, they will never be out of balance unless something is wrong, in which case you need to know about it.

Real world. My bank is 3 years old. I have an inexpensive Overkill/JBD BMS. Primary charging is from 2 Victron MPPT controllers. They connect to a BMV-712 to directly read battery voltage and current. Rarely, I can top off with my alternator, but I have a switch installed to turn that off, and only charge with the alternator during exceptional situations. I have about 100 cycles during the past 3 years, according to the BMV.

I mostly ignore my battery. It just works and I don't need to monitor it at all. About monthly I will take a closer look. My cells are still in perfect balance. I am still getting more than rated capacity. (Rated is 300Ah, I get about 310Ah.) I have had one instance of LVC, I was out of town for 2 weeks, and we had a rare 2 weeks of solid rain (in California). With no sun for 2 weeks the battery was fully discharged, and the BMS did its job.

And that is all the BMS needs to do. The Victron MPPTs are perfectly capable of charging the battery as well as any BMS could, even using the actual charge current going into the battery and stopping charging based on tail current. So, even that argument for BMS charging is nonsense.

There is nothing WRONG with a BMS controlling charging or using cell voltages for charging. It is just that there isn't any reason for it. It's just a waste of money.

[SaylorMade]

Originally Posted by mcarthur

But a BMS that has just a single OVP and shuts down the entire bank is pretty useless in comparison to one that communicates an OV alarm to other systems so that they can nicely stop charging (MMPT's, chargers). This continues having the battery online for use rather than disconnecting it entirely.
Unfortunately, BMS don't seem to a) have the right functions, and b) they don't always work. However, they are probably still better than a fully distributed "stop charging" regime (noting that this is different to an "oh sh*t, something is badly wrong regime").

An OVP disconnect on the two different BMS I use (what you call "cheap chinese"), only disconnects the charging, not discharging. Charge and discharge is individually monitored, so while the BMS does cut the charge-connection, discharge still continues with no interruption.

This is also why you can selectivly enable or disable charging and discharging individually. Another neat feature of LifePo4 with BMS

[wholybee]

Originally Posted by mcarthur

Unfortunately, BMS don't seem to a) have the right functions, and b) they don't always work. However, they are probably still better than a fully distributed "stop charging" regime (noting that this is different to an "oh sh*t, something is badly wrong regime").

What functions are missing and how do they not always work? Even cheap BMS's (JBD/Daly) have every function required to safely manage a battery and have it last for the maximum life. And they are quite reliable and work very well.

[fxykty]

Originally Posted by wholybee

…

And that is all the BMS needs to do. The Victron MPPTs are perfectly capable of charging the battery as well as any BMS could, even using the actual charge current going into the battery and stopping charging based on tail current. So, even that argument for BMS charging is nonsense.



There is nothing WRONG with a BMS controlling charging or using cell voltages for charging. It is just that there isn't any reason for it. It's just a waste of money.

I believe that you don’t live on your boat full time, and if that’s the case then you aren’t exercising the batteries much (you say ~100 cycles in three years; we live aboard and cruise full time and have just over 300 cycles in three years). You have a very simple system that functions well, so you do you.

We have a much more complex system with 4 solar controllers, wind generator, 2 alternators, plus an inverter/charger (and soon we’re adding a hydro generator). Except for the wind generator, each of our charge sources can be programmed for custom profiles. The wind generator controller only allows a charge voltage setting. In the absence of a BMS, all the chargers would, except the wind generator, charge according to their profiles to a reasonable, but not optimal, level for our needs.

That’s why we have our BMS control all the chargers, the Victron Energy chargers to a greater extent than the others. The BMS controls charge voltage and maximum charge current based on and varies them according to a whole bunch of parameters that we set. Examples are high temperature, low temperature, cell voltages out of balance, cell temperatures out of balance, high SOC, etc. We want the BMS to control the chargers and to stop charging according to parameters that we set. Those parameters are not necessarily part of any individual component’ charging settings. And that’s OK.

For certain chargers, such as the wind generator and our two alternators (as we have Balmar regulators that can’t be externally controlled, unlike the Wakespeed regulators), we can only stop charging. With chargers that can be externally controlled via CAN Bus (such as all of our Victron Energy components), the BMS controls the charging algorithms and also stops charging. This satisfies our relatively conservative rules for using our battery (refer to Nordkyn Design and MarineHowTo for those).

The BMS also stops our inverter based on a variety of parameters. Again, the inverter has its own parameters, but we don’t consider those sufficient for our needs.

Note that our BMS is also responsible for disconnects of either or both charge or load buses, for the standard OVP, LVP, OTP, LTP quartet of conditions. Never seen any of these.

For us, with a larger LFP bank and needs for 24/7/365 operation, a BMS that controls (and automates) charging is a necessity. Otherwise, it’s just too much manual interventions and memory tasks.

We do we. I’m only being descriptive and I don’t care what you do; I’m certainly not being proscriptive here.

[s/v Jedi]

Well, no.

The LFP cell manufacturers state you must use a certain charge algorithm, which is a CC algorithm with a termination voltage. That is it. All the rest, as made up by battery manufacturers, and many of us are among them, is just an opinion, hopefully based on that cell manufacturers specification.

When you connect cells in series to form a battery of a higher voltage, you do what has been done with every chemistry ever. And if your cells are of high quality you can still do that today, see the Winston 12V batteries that come just like that and simply work.

The problem is that the Lithium-cobalt chemistries, popular for high energy density, can have nasty thermal runaways which can have potentially lethal outcomes like we have seen with burning airplanes, cars etc. To avoid this, a BMS was added which monitors cells and disconnects the battery before something bad happens. That’s it.

And I believe that for sailboats that want hardened systems that are near impossible to bring down, this is where the BMS needs to stay. If you like your boat to be a showcase of technological innovation then sure go ahead and add new or even experimental BMS features, but know that you bring reliability down in a major way.

Same for balancing, this shouldn’t be in the BMS, it is a job for a cell balancer.

Now we get to the counter arguments: a 12V charger, programmed for LFP doesn’t know individual cell status! Correct, it doesn’t because it doesn’t need to. The charger is charging the whole battery at 12V and only needs to know the voltage at the 12V battery terminals and will do a perfect charge because it is programmed to do so. That’s it, that’s all there is to it.

Counter argument: yes, but when the cells go out of balance then the charger doesn’t know! Well, no, it doesn’t know. But the BMS knows. It can do many things, here is the list of things in order of priority:

1. If a cell enters high voltage area then it must do a HVC event (high voltage cutoff i.e. disconnect the battery). In my opinion all BMS’s do this. Also IMO, this is enough if you were wise enough to create redundancy by having multiple batteries. This one battery with unbalanced cells has a problem but the other battery/batteries are okay.

2. Needs balancing signal. Visual indication as simple as a LED and either activate an internal balancer (meh…) or preferably turn on an external balancer (yay!)
This feature is very nice to have. IMO all BMS’s have it.

3. Needs to stop charging in order to prevent HVC event. This is a warning that can be used to turn off a charging source. It means something is wrong with either the charger or the battery. If this signal is used to automatically turn off charging sources, then that automation MUST be able to be manually overruled because it may very well be a bad battery while other batteries need charging. This is the big negative for this feature, and why it requires an audible alarm so that the user can investigate. This is simple as the user can view cell voltages and compare with other batteries that are online, immediately revealing if the battery has a problem or the charger. If it’s the battery then take it offline and overrule it’s BMS order to stoo charging so that other batteries can do their job. Now you can see if the battery can be fixed or needs replacement, while your boat keeps going.

So, without 3 you would be just fine unless the charging source is defective. But this can be monitored with a simple battery monitor.

The reason I support #3 is for cases where people want to use their alternator to directly charge the LFP batteries. I don’t like that because IMO you should make solar your primary charge source instead of ICE (using a dc-dc converter for the alternator as secondary charge source) but hey others love their fossil fuels and tend to be the ones with only one LFP battery which can lead to damage during an HVC event from a transient surge from the alternator when the only battery is taken offline. I mean damage other than loosing all power, as it’s likely that alternator rectifier diodes are fried and possibly electronics are damaged too.

So that’s where I leave this. There are many other required features like for low voltage/LVC, cell based alarm (which is #3), temperature monitoring etc. but thise are outside the scooe of this thread.

So the big reason against letting the BMS decide these things: a good boat must have multiple autonomous battery systems, each with it’s own BMS and must be able to operate at least after one battery has failed. Even 1980’s boats have had this, it is required to be safe.

When you have multiple BMS systems, which one would decide how to charge? Would they argue and who would win? The only way to do that right is with separate charge busses and each battery it’s own dedicated chargers that it controls, yet nobody is willing to do that unless you’re a battleship, spacecraft etc.

[fxykty]

Originally Posted by s/v Jedi

…

The reason I support #3 is for cases where people want to use their alternator to directly charge the LFP batteries. I don’t like that because IMO you should make solar your primary charge source instead of ICE (using a dc-dc converter for the alternator as secondary charge source) but hey others love their fossil fuels and tend to be the ones with only one LFP battery which can lead to damage during an HVC event from a transient surge from the alternator when the only battery is taken offline. I mean damage other than loosing all power, as it’s likely that alternator rectifier diodes are fried and possibly electronics are damaged too.



...

Lots to consider, thanks for posting. I’m only going to quibble with your assertion that alternator(s) should not directly charge an LFP battery or bank of LFP batteries.

Why not? It’s got nothing to do with making ICE the primary charge source, nor of loving fossil fuels. Seriously, where did that come from?

It’s simply about using LFP charge acceptance to make ICE charging as efficient as possible. Solar (and any other source such as wind or hydro) are still the primary charge sources, with ICE charging taking place opportunistically, if needed, when the engine(s) is(are) running for propulsion. What’s wrong with that? Or if the primary charge sources are not keeping up with demand due to lower generation.

Regarding alternator damage due to BMS disconnect of the battery(ies), that’s an old trope that has several easy solutions that have been discussed on this forum multiple times. To reiterate, it’s simple to have a relay cut the ignition or field of the alternator regulator to gracefully shut down the alternator prior to a disconnect command.

[fxykty]

BMS is useful for controlling chargers when you don’t want to routinely charge to 100% SOC based on CV and trailing current. And in cases where the charger isn’t capable of making the decision of when to stop charging for itself.

If you do want to always charge to 100%, OK, let the charger’s(s’) charge algorithm(s) make the decision(s) instead.

Nothing is gained labelling one method as silly and the other as righteous. Each system owner needs to decide for themselves which method is right for them based on evidence, not emotion.

[team karst]

Originally Posted by SaylorMade

...

I realize that the sudden cut-off from a BMS may have implications for an alternator so that should be adressed seperately.

or solar controller(s), or mains charger. Depending on design quality, if they suddenly see an open circuit and are in constant current mode, that might be outside the design window.

[wholybee]

Originally Posted by fxykty

Lots to consider, thanks for posting. I’m only going to quibble with your assertion that alternator(s) should not directly charge an LFP battery or bank of LFP batteries.

Why not? It’s got nothing to do with making ICE the primary charge source, nor of loving fossil fuels. Seriously, where did that come from?

It’s simply about using LFP charge acceptance to make ICE charging as efficient as possible. Solar (and any other source such as wind or hydro) are still the primary charge sources, with ICE charging taking place opportunistically, if needed, when the engine(s) is(are) running for propulsion. What’s wrong with that? Or if the primary charge sources are not keeping up with demand due to lower generation.

Regarding alternator damage due to BMS disconnect of the battery(ies), that’s an old trope that has several easy solutions that have been discussed on this forum multiple times. To reiterate, it’s simple to have a relay cut the ignition or field of the alternator regulator to gracefully shut down the alternator prior to a disconnect command.

My answer to this is that because solar is my primary charge source, I chose not to spend huge money on a big alternator, wakespeed regulator, etc. I spent that money on more solar instead. And with a 70A alternator, you aren't really gaining much if any efficiency by direct charging. With a 60A DCDC I can run it all day long without overheating. It just isn't worth the expense and complexity for another 10A, that would only last until the alternator heats up and is regulated down to 60A.

If you have deep pockets and have a large frame 200A alternator with wakespeed regulator, then yeah, direct charging is going to be much better. But why spend that money on something you won't hardly ever use?

[wholybee]

Originally Posted by fxykty

I believe that you don’t live on your boat full time, and if that’s the case then you aren’t exercising the batteries much (you say ~100 cycles in three years; we live aboard and cruise full time and have just over 300 cycles in three years). You have a very simple system that functions well, so you do you.

We have a much more complex system with 4 solar controllers, wind generator, 2 alternators, plus an inverter/charger (and soon we’re adding a hydro generator). Except for the wind generator, each of our charge sources can be programmed for custom profiles. The wind generator controller only allows a charge voltage setting. In the absence of a BMS, all the chargers would, except the wind generator, charge according to their profiles to a reasonable, but not optimal, level for our needs.

That’s why we have our BMS control all the chargers, the Victron Energy chargers to a greater extent than the others. The BMS controls charge voltage and maximum charge current based on and varies them according to a whole bunch of parameters that we set. Examples are high temperature, low temperature, cell voltages out of balance, cell temperatures out of balance, high SOC, etc. We want the BMS to control the chargers and to stop charging according to parameters that we set. Those parameters are not necessarily part of any individual component’ charging settings. And that’s OK.

For certain chargers, such as the wind generator and our two alternators (as we have Balmar regulators that can’t be externally controlled, unlike the Wakespeed regulators), we can only stop charging. With chargers that can be externally controlled via CAN Bus (such as all of our Victron Energy components), the BMS controls the charging algorithms and also stops charging. This satisfies our relatively conservative rules for using our battery (refer to Nordkyn Design and MarineHowTo for those).

The BMS also stops our inverter based on a variety of parameters. Again, the inverter has its own parameters, but we don’t consider those sufficient for our needs.

Note that our BMS is also responsible for disconnects of either or both charge or load buses, for the standard OVP, LVP, OTP, LTP quartet of conditions. Never seen any of these.

For us, with a larger LFP bank and needs for 24/7/365 operation, a BMS that controls (and automates) charging is a necessity. Otherwise, it’s just too much manual interventions and memory tasks.

We do we. I’m only being descriptive and I don’t care what you do; I’m certainly not being proscriptive here.

I was full time liveaboard for the first 2 years. Now I am replenishing the kitty. I do leave a few things on when I'm not on board, so I get some minimal cycling but not much. Even on board, I have enough solar for my minimal needs, so that battery doesn't cycle much.

That really doesn't change anything. All of your charge sources, including the Balmar regulator and the wind generator, are able to shut down at full charge without a BMS. (some wind generators might have an issue with this, i'm not sure. In that case I might add a DCDC charger)

In particular, the Victron stuff is really good. I'm pretty sure letting the Victron do it's thing is going to work better than the BMS.

So, the most I would consider the BMS controlling is shutting down the alternator and wind generator prior to a disconnect. But even with a dozen charge sources, I would let them all do there thing independently of the BMS.

With many charge sources like you have, I would probably go with a Cerbo GX route, and have that connect to all the charge sources, but not the BMS.

[s/v Jedi]

Originally Posted by fxykty

Lots to consider, thanks for posting. I’m only going to quibble with your assertion that alternator(s) should not directly charge an LFP battery or bank of LFP batteries.

Why not? It’s got nothing to do with making ICE the primary charge source, nor of loving fossil fuels. Seriously, where did that come from?

It’s simply about using LFP charge acceptance to make ICE charging as efficient as possible. Solar (and any other source such as wind or hydro) are still the primary charge sources, with ICE charging taking place opportunistically, if needed, when the engine(s) is(are) running for propulsion. What’s wrong with that? Or if the primary charge sources are not keeping up with demand due to lower generation.

Regarding alternator damage due to BMS disconnect of the battery(ies), that’s an old trope that has several easy solutions that have been discussed on this forum multiple times. To reiterate, it’s simple to have a relay cut the ignition or field of the alternator regulator to gracefully shut down the alternator prior to a disconnect command.

Well, I support #3 which is the way to prevent damage when using ICE to charge when a HVC of the only battery that is online occurs. But that doesn’t mean I like it. Others don’t like solar arrays, think they are ugly, a waste of money and swear by using ICE for charging. But the serious ones use a genset, not an auxiliary diesel with alternator.

I can understand your position for motorboats but not for sailboats who normally not run the engine other than for close quarters maneuvering or a calm, in which case a simple, cheap dc-dc converter is all that is needed to run electronics, refrigeration etc. even during extended motoring because you still have solar as primary charge source.

I have an updated large alternator with external regulator that supports LFP and can shutdown when a signal to do so from a BMS comes. But I just charge the LA start battery with it, plus 60A worth of dc-dc converters.

For damage during HVC, again, that is handled by task 3 of the BMS. But remember, a boat with a solid, hardened system doesn’t need it because when a BMS does HVC, you have at least one other battery online and there is no transient surge at all.

Personally, we first replaced one battery bank with LFP and went for 10.25kWh of the best available and cruised for two seasons with only that. We found that it was more than what is the minimum required for us, so we replaced our second battery bank also with 10.25kWh of LFP, this time some of the cheapest available. We know we can get away with that because we could do just with one battery as we proved for two seasons, but I also think they will perform the same… mostly. I mean, my Winston based battery never required balancing other than the initial balance and the new battery may be balancing every day, but I simply don’t know because it’s BMS is silently doing it’s work inside it.

Now I’m left with your statement about finding it silly that I talk of love for fossil fuels and using ICE as primary charge source.

Why would someone who uses solar as primary charge source and is actually sailing instead of motoring, spend a lot of money on creating a safe way to massively charge LFP from the auxiliary engine(s)? It would be much better to put the money in more solar or an extra LFP battery.