Lithium and Shorelines

[DShK]

I'm moored up for the winter in a basin, so I now have shoreline power, pretty much for the first time since installing my lithium setup. Now I am faced with keeping the batteries healthy while remaining plugged in - not as easy as lead-acids! I am wondering what others do for this?

 

We know that lithium doesn't like the be kept at 100% charge, so keeping the charger on full whack (float or not) is a no-no. So ideally the challenge is to keep the batteries around the 50% mark. My cruising setup disconnects all charging when the SoC hits a configurable value. And will re-enable solar charging when the batteries get to a second value. So this is my starting point for protection. However, needlessly cycling the batteries sounds like a waste of their life, especially as rapidly charging lithium batteries is the most harmful part of the cycle.

 

At the moment I am charging with a very low current (10 amps) which is slightly more than my average DC use (thus minimizing the harm from charging) The AC is passed through. When the battery SoC gets too high, the multiplus is set to "inverter only", thus turning off the charger. The problem with this is, this mode stops the AC pass-through. So the batteries could quickly discharge or the inverter overload. When the batteries get lower, the charger is re-enabled.

 

It works okay, but not perfectly. I have been screwing around with dynamically changing the charging current to try and match the DC power draw, but I don't think this will be perfect. I am also not sure if frequently changing the charge speed would somehow be bad for the unit.

 

The other method I've heard of is using a voltage that is unable to push the batteries beyond a certain SoC. Has anyone tried this? Does it work well? My aversion to this method is that it requires manually reconfiguring the multiplus (either using VEConfig with VRM or using a cable), as I don't think node-red and modbus can change this setting. In the end I would like a user-friendly system that I can share with people who don't want to screw around with this as much as I do!

 

What are your thoughts on keeping lithiums healthy when plugged in?

Edited November 9, 2023 by DShK

[rusty69]

I run the AC loads off the shoreline and leave the dc loads to the solar and lithium. 

 

That may not work for you. 

[MtB]

Unplug the LFPs completely and use your battery charger as a 12v power supply for the boat system.

 

[DShK]

6 minutes ago, rusty69 said:

I run the AC loads off the shoreline and leave the dc loads to the solar and lithium. 

 

That may not work for you. 

 

Ah, that might have worked but I removed all my solar panels! I plan to have them mounted in the butty hold (moored elsewhere)

 

4 minutes ago, MtB said:

Unplug the LFPs completely and use your battery charger as a 12v power supply for the boat system.

 

That's a great idea. Seems obvious now you mention it. And not over complicating the issue! The only downside is the lack of power backup, I'll just need to make sure I load up on electric credits so I don't lose any work!

Edited November 9, 2023 by DShK

[IanD]

15 minutes ago, DShK said:

I'm moored up for the winter in a basin, so I now have shoreline power, pretty much for the first time since installing my lithium setup. Now I am faced with keeping the batteries healthy while remaining plugged in - not as easy as lead-acids! I am wondering what others do for this?

 

We know that lithium doesn't like the be kept at 100% charge, so keeping the charger on full whack (float or not) is a no-no. So ideally the challenge is to keep the batteries around the 50% mark. My cruising setup disconnects all charging when the SoC hits a configurable value. And will re-enable solar charging when the batteries get to a second value. So this is my starting point for protection. However, needlessly cycling the batteries sounds like a waste of their life, especially as rapidly charging lithium batteries is the most harmful part of the cycle.

 

At the moment I am charging with a very low current (10 amps) which is slightly more than my average DC use (thus minimizing the harm from charging) The AC is passed through. When the battery SoC gets too high, the multiplus is set to "inverter only", thus turning off the charger. The problem with this is, this mode stops the AC pass-through. So the batteries could quickly discharge or the inverter overload. When the batteries get lower, the charger is re-enabled.

 

It works okay, but not perfectly. I have been screwing around with dynamically changing the charging current to try and match the DC power draw, but I don't think this will be perfect. I am also not sure if frequently changing the charge speed would somehow be bad for the unit.

 

The other method I've heard of is using a voltage that is unable to push the batteries beyond a certain SoC. Has anyone tried this? Does it work well? My aversion to this method is that it requires manually reconfiguring the multiplus (either using VEConfig with VRM or using a cable), as I don't think node-red and modbus can change this setting. In the end I would like a user-friendly system that I can share with people who don't want to screw around with this as much as I do!

 

What are your thoughts on keeping lithiums healthy when plugged in?

 

That "we-know" comes from the received wisdom for lithium-ion batteries using NMC and similar chemistries; LFP batteries are not the same, and there's good reason to believe that they're perfectly happy being kept at 100% SoC.

 

However if you do this when plugged-in there's nowhere for solar charge to go to reduce the amount of shoreline power. Victron have introduced a new feature called solar/wind priority which allows the battery voltage to drop to a set threshold (e.g. 50% SoC) before shoreline charging takes over, which allows solar to be used -- but unfortunately this only works for 7 days, after this the batteries are changed back up to 100% SoC from shoreline and stay there.

 

Discussions are ongoing with Victron to change this in a future software release, but no result so far... 😞

[GUMPY]

The joys of having a hybrid inverter. It just does it all for you 😉

 

[ditchcrawler]

56 minutes ago, DShK said:

 

 

At the moment I am charging with a very low current (10 amps) which is slightly more than my average DC use (thus minimizing the harm from charging) The AC is passed through. When the battery SoC gets too high, the multiplus is set to "inverter only", thus turning off the charger. The problem with this is, this mode stops the AC pass-through. So the batteries could quickly discharge or the inverter overload. When the batteries get lower, the charger is re-enabled.

 

I am just wondering how you manage to use over 240 Ah per day 

[DShK]

54 minutes ago, ditchcrawler said:

I am just wondering how you manage to use over 240 Ah per day 

It'll be a fair amount less than that because it is actually charging my batteries, it's just a value that is mostly above the highest typical draw during the day. But I do have both a fridge and freezer, a circulation pump running all the time and two routers running all the time. Plus im not as worried about turning lights off if I've got a shoreline 🙂 actually I just checked - I had some problems with my inverter the other day so did not use it. My daily use is more like 130ah on DC.

Edited November 9, 2023 by DShK

[nicknorman]

I am in a similar situation, all Li domestic bank and the boat is plugged into shore power when we are in the marina, via Mastervolt Combi. Trying to regulate all this by limiting shore or charging current is never going to work, the only way is by adjusting the charge voltage.

First of all, the Mastervolt has a "Fixed Float" setting. That means the voltage is fixed at the fixed float voltage setting. Then next step is to set this fixed float voltage to one that keeps the Li at around 50% SoC whilst providing any necessary current to run the boat's systems. This is around 13.2v

 

This means adjusting settings on the Combi, in my case this is done via my custom BMS but it can reasonably easily be done via a PC especially if you are not frequently leaving the marina and wanting to fully charge you batteries beforehand. Victron offer a number of gadgets to assist, such as the Venus OS gadgets that communicate with these sorts of devices (I think, no direct experience).

 

But the bottom line is that to achieve what you want to achieve, you need to precisely adjust the set charge voltage of the Combi. Works well for me.

 

Edited November 9, 2023 by nicknorman

[phantom_iv]

On 09/11/2023 at 19:37, IanD said:

 

That "we-know" comes from the received wisdom for lithium-ion batteries using NMC and similar chemistries; LFP batteries are not the same, and there's good reason to believe that they're perfectly happy being kept at 100% SoC.

😞

 

Yes this was my understanding as well, LFPs are quite happy to sit at 100%

[nicknorman]

On 09/11/2023 at 19:37, IanD said:

 

 LFP batteries are not the same, and there's good reason to believe that they're perfectly happy being kept at 100% SoC.

Do you know what this "good reason" is - and care to share? All the manufacturer's blurb seems to say that for long term storage, around 30-50% is best. And if you are in a marina on shore power not actually using the batteries to supply power, surely that is effectively the same as being in storage?

[IanD]

43 minutes ago, nicknorman said:

Do you know what this "good reason" is - and care to share? All the manufacturer's blurb seems to say that for long term storage, around 30-50% is best. And if you are in a marina on shore power not actually using the batteries to supply power, surely that is effectively the same as being in storage?

 

Not all manufacturer's blurb says this. Many BMS -- including inbuilt ones -- "float" the cells at close to 100% SoC for long periods of time with no reported ill effects. But for this to be done safely in a battery the cells must be well matched, and the BMS needs to monitor/limit the highest cell voltage not the battery voltage -- I'm sure you know this, but not all BMS do this or have effective cell balancing, and without this one or more cells can go above the upper voltage limit (100% SoC) without triggering a disconnect/shutdown.

 

For long-term storage where the charge gradually leaks away and may do so differently for different cells, keeping well away from the low and high voltage knees (around 10% and 90% SoC?) is recommended, and the further away the better since the voltage slope is so small. This isn't the same as being kept charged to 100% by a BMS with effective cell monitoring and balancing.

 

Another point is that at high charge/discharge rates like used in EVs the batteries don't like being pushed close to the SoC limits, for example charging usually slows down above maybe 80% and discharge may be limited below 20% -- but this isn't applicable to fractional-C applications like boats.

 

So maybe I should rephrase what I said -- in a fractional-C LFP system with effective cell monitoring and balancing there's no problem spending long periods at close to 100% SoC, which also ensures the cells *do* stay balanced.

Edited November 14, 2023 by IanD

[DShK]

Right, but my point isn't really which SoC they like best so to speak. It's that a LA battery can be put on charge and left on charge. If you do that with a LFP battery it's going to overcharge. Which is not the same as charging to 100% and cutting off charging.

 

The Solar priority stuff you describe Ian is stuff I can do in node-red, but I can add my own logic to it and customise it. Great fun! Can do stuff like cutting off alternator charging if the solar input is high etc.

[nicknorman]

41 minutes ago, IanD said:

 

Not all manufacturer's blurb says this. Many BMS -- including inbuilt ones -- "float" the cells at close to 100% SoC for long periods of time with no reported ill effects. But for this to be done safely in a battery the cells must be well matched, and the BMS needs to monitor/limit the highest cell voltage not the battery voltage -- I'm sure you know this, but not all BMS do this or have effective cell balancing, and without this one or more cells can go above the upper voltage limit (100% SoC) without triggering a disconnect/shutdown.

 

For long-term storage where the charge gradually leaks away and may do so differently for different cells, keeping well away from the low and high voltage knees (around 10% and 90% SoC?) is recommended, and the further away the better since the voltage slope is so small. This isn't the same as being kept charged to 100% by a BMS with effective cell monitoring and balancing.

 

Another point is that at high charge/discharge rates like used in EVs the batteries don't like being pushed close to the SoC limits, for example charging usually slows down above maybe 80% and discharge may be limited below 20% -- but this isn't applicable to fractional-C applications like boats.

 

So maybe I should rephrase what I said -- in a fractional-C LFP system with effective cell monitoring and balancing there's no problem spending long periods at close to 100% SoC, which also ensures the cells *do* stay balanced.

 

I am not sure I agree with some of this, but of course both your and my views are simply our opinions based on what we have read elsewhere.

 

As I understand it, having a Li cell at close to 100% or close to 0% puts mechanical stress on the relevant electrode  because the Li ions of course all go to one or other electrode at the extremes of SoC, cramming into the electrode, and so the battery is happiest at a mid state of charge where everything is evenly distributed. There is no difference in stress between 100% and 0% other than if at 0% (2.5v), any slight additional discharge is going to be bad. This seems plausible to me.

 

I have CALB cells and they recommend 30% SoC for long term storage. Not sure why not 50% but I suspect this is perhaps mingled with a desire to keep SoC at the lower end for transportation.

 

IMO what the BMS manufacturer recommends is not really relevant, it is what the cell manufacturer recommends that is important and since all LiFePO4 cells are pretty much the same, you would think that all cell manufacturer's recommendations would be the same in this sort of respect.

 

As to the BMS and cell monitoring, I would have though that any vaguely self-respecting BMS monitors individual cell voltages, although it is true that some on here have batteries from which they can't extract the BMS information. Those do at least have inbuilt balancing.

 

But on the subject of balancing, my experience after a couple of years is that the cells never go out of balance. I don't normally fully charge but when I do, every month or so, it is at 14.3v ie 3.575v/cell. My balancing algorithm kicks in if any cell exceeds 3.6v, and this is indicated on the display by an asterisk next to the cell voltage, but that just never happens. And as you know, the difference in Ah between a cell at 3.575v and one at 3.6v is minimal.

So I can't see any advantage in keeping the cells at 100% just so some balancing could take place. It will never be needed!

 

Of course these cells have very long lives and so whether or not keeping cells at 100% for long periods is going to reduce life, firstly is very hard to determine, and secondly may not matter too much.

 

Nordkyn Design chappie seemed to think that holding cells at 100% was bad. He probably knows more about it than I do.

 

So all that is why I keep the batteries at a mid state of charge, around 50%, when we are away from the boat for long periods.

1 minute ago, DShK said:

Right, but my point isn't really which SoC they like best so to speak. It's that a LA battery can be put on charge and left on charge. If you do that with a LFP battery it's going to overcharge. Which is not the same as charging to 100% and cutting off charging.

 

The Solar priority stuff you describe Ian is stuff I can do in node-red, but I can add my own logic to it and customise it. Great fun! Can do stuff like cutting off alternator charging if the solar input is high etc.

Well IMO it is all about the charge voltage. Keep it down at around 13.3v if on shore power for a long time, and everything will be fine.

[DShK]

Yep, problem is I can't algorithmically change the charge voltage. Just my preference though that I'd like to not have to update firmware settings as a regular thing. I'd like to design a system which my mum could use, for example. (I will be doing an electrical install for her soon, likely going to put in LFP batteries there too). Disconnecting the batteries seems like a simple fool proof solution.

[Col_T]

Genuine question - given that LiFe batteries have life-spans quoted in 1,000’s of cycles, is what constitutes ‘long term storage’? I kind of imagine a year or two, rather than a month or four?

[nicknorman]

3 minutes ago, Col_T said:

Genuine question - given that LiFe batteries have life-spans quoted in 1,000’s of cycles, is what constitutes ‘long term storage’? I kind of imagine a year or two, rather than a month or four?

Who knows? They don’t say! But when we leave our boat for a couple of months, I tend to consider that as “storage”.

 

You make a good point though because leisure boaters like us are never going to get close to the cycle life, however there is also the calendar life to consider and in our case, that is probably the predominant limitation. Hence my desire to give them a “nice time” when they are not doing anything useful, to hopefully maximise the calendar life.

[IanD]

32 minutes ago, nicknorman said:

 

I am not sure I agree with some of this, but of course both your and my views are simply our opinions based on what we have read elsewhere.

 

As I understand it, having a Li cell at close to 100% or close to 0% puts mechanical stress on the relevant electrode  because the Li ions of course all go to one or other electrode at the extremes of SoC, cramming into the electrode, and so the battery is happiest at a mid state of charge where everything is evenly distributed. There is no difference in stress between 100% and 0% other than if at 0% (2.5v), any slight additional discharge is going to be bad. This seems plausible to me.

 

I have CALB cells and they recommend 30% SoC for long term storage. Not sure why not 50% but I suspect this is perhaps mingled with a desire to keep SoC at the lower end for transportation.

 

IMO what the BMS manufacturer recommends is not really relevant, it is what the cell manufacturer recommends that is important and since all LiFePO4 cells are pretty much the same, you would think that all cell manufacturer's recommendations would be the same in this sort of respect.

 

As to the BMS and cell monitoring, I would have though that any vaguely self-respecting BMS monitors individual cell voltages, although it is true that some on here have batteries from which they can't extract the BMS information. Those do at least have inbuilt balancing.

 

But on the subject of balancing, my experience after a couple of years is that the cells never go out of balance. I don't normally fully charge but when I do, every month or so, it is at 14.3v ie 3.575v/cell. My balancing algorithm kicks in if any cell exceeds 3.6v, and this is indicated on the display by an asterisk next to the cell voltage, but that just never happens. And as you know, the difference in Ah between a cell at 3.575v and one at 3.6v is minimal.

So I can't see any advantage in keeping the cells at 100% just so some balancing could take place. It will never be needed!

 

Of course these cells have very long lives and so whether or not keeping cells at 100% for long periods is going to reduce life, firstly is very hard to determine, and secondly may not matter too much.

 

Nordkyn Design chappie seemed to think that holding cells at 100% was bad. He probably knows more about it than I do.

 

So all that is why I keep the batteries at a mid state of charge, around 50%, when we are away from the boat for long periods.

Well IMO it is all about the charge voltage. Keep it down at around 13.3v if on shore power for a long time, and everything will be fine.

I don't think the cells need to be kept at 100% for long periods for balancing -- even if this is needed (and it often isn't, as you say) it only needs doing occasionally for a relatively short time.

 

But my point was that with a good BMS and balancing, keeping them close to 100% isn't going to do any harm if that's how the system is configured (and many are when plugged into shoreline, including Victron) -- but it would be better for them to spend most of the time at a lower SoC with occasional bursts up to 100%.

4 minutes ago, nicknorman said:

Who knows? They don’t say! But when we leave our boat for a couple of months, I tend to consider that as “storage”.

 

You make a good point though because leisure boaters like us are never going to get close to the cycle life, however there is also the calendar life to consider and in our case, that is probably the predominant limitation. Hence my desire to give them a “nice time” when they are not doing anything useful, to hopefully maximise the calendar life.

Agreed, I worked out that even for an intensive CCer a decent-sized LFP bank will outlast both the hull and probably the boater... 😉

37 minutes ago, DShK said:

Right, but my point isn't really which SoC they like best so to speak. It's that a LA battery can be put on charge and left on charge. If you do that with a LFP battery it's going to overcharge. Which is not the same as charging to 100% and cutting off charging.

 

The Solar priority stuff you describe Ian is stuff I can do in node-red, but I can add my own logic to it and customise it. Great fun! Can do stuff like cutting off alternator charging if the solar input is high etc.

With an LFP it's absolutely necessary to terminate charging at 100% SoC.

[nicknorman]

9 minutes ago, IanD said:

I don't think the cells need to be kept at 100% for long periods for balancing -- even if this is needed (and it often isn't, as you say) it only needs doing occasionally for a relatively short time.

 

I know that some BMSs only balance when held close to fully charged voltage, however this is not the only way of doing it and is not the way I do it. With acknowledgement to Moominpapa, my algorithm is triggered when the first cell goes over 3.6v and notes the voltage of the cells. The voltage differences between the lowest cell and the other 3 are then multiplied by a factor and that determines how long each balancing resistor is turned on for. So as soon as a cell hits 3.6v, charging can be stopped but the balancing process continues.

 

If there is a lot of imbalance, it may take several iterations (because it is better to underestimate the balance time than to overshoot) but as I said, once they are balanced they don’t seem to go out at all.

[IanD]

16 hours ago, nicknorman said:

I know that some BMSs only balance when held close to fully charged voltage, however this is not the only way of doing it and is not the way I do it. With acknowledgement to Moominpapa, my algorithm is triggered when the first cell goes over 3.6v and notes the voltage of the cells. The voltage differences between the lowest cell and the other 3 are then multiplied by a factor and that determines how long each balancing resistor is turned on for. So as soon as a cell hits 3.6v, charging can be stopped but the balancing process continues.

 

If there is a lot of imbalance, it may take several iterations (because it is better to underestimate the balance time than to overshoot) but as I said, once they are balanced they don’t seem to go out at all.

 

I know you know this, but for the benefit of others you can't do any cell balancing until the cells -- or at least one of them -- start to go up the "voltage knee", which on mine starts around 3.45V and 90% SoC (100% SoC is around 3.55V according to the BMS), because below this there simply isn't enough voltage slope to spot any SoC differences between cells -- I believe that cell balancing starts about 3.5V/cell (95% SoC). 3.6V/cell would be too high to start balancing, but this will depend on the exact battery chemistry used, mine are Winston LiFeYPO4. Bigger cells (mine are 700Ah) need either higher balance currents or longer balance times, also voltage is less dependent on charging current. All these numbers are for one particular installation and may well be different for another one using different cells and BMS.

 

Ricky told me that none of the boats he's remotely monitoring ever seem to go significantly out of balance, which confirms what you found 🙂

Edited November 15, 2023 by IanD

[nicknorman]

25 minutes ago, IanD said:

 

I know you know this, but for the benefit of others you can't do any cell balancing until the cells -- or at least one of them -- start to go up the "voltage knee", which on mine starts around 3.45V and 90% SoC (100% SoC is around 3.55V according to the BMS), because below this there simply isn't enough voltage slope to spot any SoC differences between cells -- I believe that cell balancing starts about 3.5V/cell (95% SoC). 3.6V/cell would be too high to start balancing, but this will depend on the exact battery chemistry used, mine are Winston LiFeYPO4. Bigger cells (mine are 700Ah) need either higher balance currents or longer balance times, also voltage is less dependent on charging current. All these numbers are for one particular installation and may well be different for another one using different cells and BMS.

 

Ricky told me that none of the boats he's remotely monitoring ever seem to go significantly out of balance, which confirms what you found 🙂

The other point re. balancing is that I have noticed that at lower voltages, just at the start of the knee, the cell balancing "order" (ie which are high and which are low) can be quite different from that when the cells approach 3.6v. In part this may be due to the higher current at lower cell voltages and the fact that I have 3 x 200Ah cells in parallel, and thus there might be some slight differences in the interconnect resistance and hence some small voltage drops somewhere (although I have never found any), but even despite that I think that the cells do not show their "true colours" until max cell voltage is approached.

[blackrose]

Looking at this thread it seems that there are still a lot of "known unknowns" as far as lithium batteries are concerned and opinion varies according to what one has read. It's very different from say lead acid batteries where everyone's agreed on how best to charge, discharge and store them when on shore power. 

 

Wrecking a set of lead acid batteries in a short time because you didn't maintain them properly as I have done is bad enough. I topped up a £600 set of Trojans with a cheap unknown brand of deionised water, which I later suspected was just tap water. But imagine wrecking a set of very expensive lithium batteries because you get things wrong? It doesn't bear thinking about.

Edited November 15, 2023 by blackrose

[IanD]

5 minutes ago, blackrose said:

Looking at this thread it seems that there are still a lot of "known unknowns" as far as lithium batteries are concerned and opinion varies according to what one has read. It's very different from say lead acid batteries where everyone's agreed on how best to charge, discharge and store them when on shore power. 

 

Wrecking a set of lead acid batteries in a short time because you didn't maintain them properly as I have done is bad enough. I topped up a £600 set of Trojans with a cheap unknown brand of deionised water, which I later suspected was just tap water. But imagine wrecking a set of very expensive lithium batteries because you get things wrong? It doesn't bear thinking about.

 

I don't think there are really more unknowns about LFP than LA -- they're just as easy to wreck, but in different ways, and need a different charging system which is more complex than LA. If you do wreck them then they cost more -- but less than wrecking multiple sets of LA batteries because you don't know how to charge them. You could even say that when properly sorted out LFP are easier to use than LA because they just act as a "charge bucket" regardless of how you use them.

 

And at least you can't wreck LFPs by topping them up with the wrong kind of water... 😉

[blackrose]

3 minutes ago, IanD said:

 

I don't think there are really more unknowns about LFP than LA -- they're just as easy to wreck, but in different ways, and need a different charging system which is more complex than LA

... 😉

 

Yes I'm sure they're just as easy to wreck, but it certainly sounds like there are more unknowns about LFP from this discussion. I never see different views from different experts on how best to maintain LA batteries because as I said, everyone's agreed. The technology is mature and the only people asking questions about LAs are the newbies, everyone else knows exactly how to use and maintain them and there is no debate.

 

 

[nicknorman]

26 minutes ago, blackrose said:

Yes I'm sure they're just as easy to wreck, but it certainly sounds like there are more unknowns about LFP from this discussion. I never see different views from different experts on how best to maintain LA batteries because as I said, everyone's agreed. The technology is mature and the only people asking questions about LAs are the newbies, everyone else knows exactly how to use and maintain them and there is no debate.

 

It probably does seem like we are making a bit of a meal of it, but in reality this is just the fine details which may or may not be significant. It might make the difference between a life of 5000 cycles and 3000 cycles. Or maybe not.
Overall I think LFP is less easy to damage provided you stick with reasonable charging voltages, and the main reason for killing lead acid - failing to fully charge frequently - has vanished.