LiFePO4 issue?

[GeoffL]

Thanks again for all the help.
@wildebus: FWIW, the app that Alpha batteries had me install looks nothing like your screenshot and doesn't give any settings adjustment; it merely reports current SOC, whether the current mode is standby, charging or discharging, charge or discharge amperage, and also the individual cell voltages.

TBH, the only serious-ish concern remaining is that the BMS goes into standby mode too easily and the Victron SmartSolar goes into float -- today that happened after the 84% SOC from last night had increased to 86.something and so charging stopped even though the voltage from the solar panels was over 40V! I switched the MPPT to factory default, waited for it to rebulk, then switched it back to the previous setting. The MPPT switched to absorption fairly quickly and was supplying just under 50W -- energy that would have been lost if I hadn't intervened, but that lasted less than an hour. It would be great to be able to configure the BMS or MPPT to prevent this issue.

 

[wildebus]

Thanks again for all the help.
@wildebus: FWIW, the app that Alpha batteries had me install looks nothing like your screenshot and doesn't give any settings adjustment; it merely reports current SOC, whether the current mode is standby, charging or discharging, charge or discharge amperage, and also the individual cell voltages.

Mine doesn't give any adjustment either. It does exactly the same as yours, even if it doesn't look the same.
You don't tend to get these totally adjustable BMSes on batteries made by major battery makers, they tend to be on DIY ones and ones that are 'white-box' importers.

TBH, the only serious-ish concern remaining is that the BMS goes into standby mode too easily and the Victron SmartSolar goes into float -- today that happened after the 84% SOC from last night had increased to 86.something and so charging stopped even though the voltage from the solar panels was over 40V! I switched the MPPT to factory default, waited for it to rebulk, then switched it back to the previous setting. The MPPT switched to absorption fairly quickly and was supplying just under 50W -- energy that would have been lost if I hadn't intervened, but that lasted less than an hour. It would be great to be able to configure the BMS or MPPT to prevent this issue.

 

[GeoffL]

TBH, I didn't know that the "standby" state even existed! I (wrongly) thought that the BMS would grab any energy offered until at least one of the cells was at 100% SOC or a lower (configurable?) value. The SmartSolar profile "expert mode" lets me set a fixed absorption stage, albeit with a seemingly stern warning that it could damage my battery -- but it's possibly the best way I've found to work around the issue. I guess that, with adaptive absorption, cloud or shade temporarily reduces the available solar power, which causes the charge current to drop below a threshold, which causes the BMS to switch to standby mode [i.e. stop accepting charge], which causes the MPPT to switch to float! From that point, the MPPT needs to send a higher charge current than float provides to 'wake up' the BMS to resume accepting charge. By setting a fixed absorption period (6 hours in my case), the MPPT is prevented from switching to float and so provides enough charge to 'wake up' the BMS when the solar power level returns?

 

[merl]

I *think*, Rob, that the BMS takes an overall view, rather than looking at the individual cell levels;

Not quite right there Steve I'm afraid. The BMS monitors all 4 cells individually.
In an attempt to keep all 4 cells at exactly the same voltage it places a load on the highest cell/cells. In the simple/ cheapest BMSs this load is simply a resistor and the excess energy in a cell is simply wasted as heat, depending on how imbalanced the cell is, it could take hours to remove the excess.
In an battery with "Active Balance" the excess voltage/energy isn't wasted as above but rather is moved from the high cell/cells and into the lower cells. Because this method doesn't generate as much heat it can be done at a higher rate(current) and hence it's much faster and obviously more efficient. Unfortunately it's more expensive.

 

[merl]

I guess that, with adaptive absorption, cloud or shade temporarily reduces the available solar power, which causes the charge current to drop below a threshold, which causes the BMS to switch to standby mode [i.e. stop accepting charge],

I don't think the BMS is switching to standby Geoff. Rather the SCC is switching to standby/float/off because there's not enough solar to provide a MPPT bulk charge.
My non Victron controller doesn't have a lithium setting but I've configured the settings to float at 13.2 volts and behaves the same, and switches to float under low light and back to bulk when the sun increases again, if I had the option to disable float completely (which I don't) then the controller would simply switch on and off in and out of bulk like yours is doing.

 

[GeoffL]

I don't think the BMS is switching to standby Geoff. Rather the SCC is switching to standby/float/off because there's not enough solar to provide a MPPT bulk charge.
My non Victron controller doesn't have a lithium setting but I've configured the settings to float at 13.2 volts and behaves the same, and switches to float under low light and back to bulk when the sun increases again, if I had the option to disable float completely (which I don't) then the controller would simply switch on and off in and out of bulk like yours is doing.

Prior to the upgrade, that Victron controller would supply just a few watts in absorption mode during shady conditions and remain in absorption until the FLA batteries were fully charged. Now, the FLA is pretty much maintained at full charge and the LFP does all the 'donkey work'. So, if the LFP goes into standby mode (effectively taking it out of the system), it seems reasonable that the SCC would sense just the FLA and respond by switching to float. One day I might just get both the Victron and LFP BT apps running at the same time and watch like a hawk to ascertain the order in which the controller and LFP go into float and standby respectively. What I can say is that while the charge going into the LFP is >1A the BMS seems to remain in charging mode. However, when I check the battery via the app during daylight hours and note the charge has dropped to zero, the BMS has always switched to standby mode. On one occasion, the SCC was still in bulk mode -- albeit with just a trickle being supplied. From this, I suspect that the charge current drops below a threshold (somewhere below 1A) and the BMS switches to standby mode in response with the SCC switching to float shortly after. What's needed is a way to disable (or to switch the BMS from standby mode) -- there is a physical switch on the battery casing to wake the battery up, but my battery box is underslung and I have a heap of stuff stowed on top of the access cover, making it difficult to access.

 

[merl]

Prior to the upgrade, that Victron controller would supply just a few watts in absorption mode during shady conditions and remain in absorption until the FLA batteries were fully charged. Now, the FLA is pretty much maintained at full charge and the LFP does all the 'donkey work'. So, if the LFP goes into standby mode (effectively taking it out of the system), it seems reasonable that the SCC would sense just the FLA and respond by switching to float. One day I might just get both the Victron and LFP BT apps running at the same time and watch like a hawk to ascertain the order in which the controller and LFP go into float and standby respectively. What I can say is that while the charge going into the LFP is >1A the BMS seems to remain in charging mode. However, when I check the battery via the app during daylight hours and note the charge has dropped to zero, the BMS has always switched to standby mode. On one occasion, the SCC was still in bulk mode -- albeit with just a trickle being supplied. From this, I suspect that the charge current drops below a threshold (somewhere below 1A) and the BMS switches to standby mode in response with the SCC switching to float shortly after. What's needed is a way to disable (or to switch the BMS from standby mode) -- there is a physical switch on the battery casing to wake the battery up, but my battery box is underslung and I have a heap of stuff stowed on top of the access cover, making it difficult to access.

Without an intimate knowledge of how everything is configured in your hybrid system it's going to be hard to diagnose issues Geoff, having a battery with a sleep/wake up button further complicates things but from everything you've posted I personally reckon you've got a very mismatched/bad cell in the pack and/or a very poor BMS or it isn't set on up well.
Have you chosen a lithium charge profile? Ie is the float component active or disabled at the SCC?
Do you have a shunt? If you do then I'd remove the LA, discharge the lithium, reset the shunt to zero and fully recharge the lithium and get an accurate capacity value for it.
What are the typical individual cell voltages after the battery has entered stand by? Are these voltages consistent if you discharge and recharge? More importantly, if the battery is left idle overnight do the cell voltages change and equalise at all?

 

[GeoffL]

Without an intimate knowledge of how everything is configured in your hybrid system it's going to be hard to diagnose issues Geoff, having a battery with a sleep/wake up button further complicates things but from everything you've posted I personally reckon you've got a very mismatched/bad cell in the pack and/or a very poor BMS or it isn't set on up well.
Have you chosen a lithium charge profile? Ie is the float component active or disabled at the SCC?
Do you have a shunt? If you do then I'd remove the LA, discharge the lithium, reset the shunt to zero and fully recharge the lithium and get an accurate capacity value for it.
What are the typical individual cell voltages after the battery has entered stand by? Are these voltages consistent if you discharge and recharge? More importantly, if the battery is left idle overnight do the cell voltages change and equalise at all?

The default SCC setting is AGM and this is the setting I momentarily set the system to so that the SCC comes off float. I could only find one lithium profile, and that includes a float (13.5V by default IIRC) with no apparent way to disable the float stage. However, that's the profile I've based the SCC config on. The battery product sheet is at https://www.alpha-batteries.co.uk/x...ium-battery-with-bluetooth/#product-datasheet if that's any use.

I'm off to the van (I store it over half an hour from home) this afternoon and will try to remember to take screenshots from the BMS app to post later today.

 

[marchie]

That seems to tie in with the AI summary Steve. Although their definition of the abbreviation COV is different the cell balancing recommendation seems to agree with your solution so sounds a good path to follow.

You would think the BMS would handle this though?

I *think*, Rob, that the BMS takes an overall view, rather than looking at the individual cell levels; however, the Roamer Smart 4th [?] Generation BMS has active Cell Balancing built in [mine is the previous generation, so lacks this sophisticaton, and at only 16 months old, I can't justify binning it, having paid just north of £1,000 in the January 2024 sale! The irony is that the Roamer reports cell levels to 3 decimal places, whereas the KS Energy predecessor only reported to 2 decimal places, so, whilst Roamer made me panic with 3.311 to 3.328v, KS would have have shown 3 cells at 3.31v and one at 3.32/3.33v, and I would have lived in blissful ignorance until the much cruder BMS copped a strop and the battery went into a coma that may or may not have been possible to wake from ... 'See a problem' spend £££ to solve it, create another problem that causes angst and costs ££££ to resolve' seems the way of the world at present; but then I had the DPF Sensor failure to distract me from BMS problems ...

Steve

Not quite right there Steve I'm afraid. The BMS monitors all 4 cells individually.
In an attempt to keep all 4 cells at exactly the same voltage it places a load on the highest cell/cells. In the simple/ cheapest BMSs this load is simply a resistor and the excess energy in a cell is simply wasted as heat, depending on how imbalanced the cell is, it could take hours to remove the excess.
In an battery with "Active Balance" the excess voltage/energy isn't wasted as above but rather is moved from the high cell/cells and into the lower cells. Because this method doesn't generate as much heat it can be done at a higher rate(current) and hence it's much faster and obviously more efficient. Unfortunately it's more expensive.

I could have explained myself better [not for the first time!]. What I meant was, the BMS allows 1 cell [Cell 4 in my case] to roar away into excess voltage, leaving 3 others at way less than maximum voltage , but cuts off power intake. This meant that Cell 4 + [lower voltage] Cells 1-3 in total were 13.4v 'ish, instead of the BMS realising that Cell 4 needed to be brought back into line, and allowing Cells 1-3 to continue charging. Apparently Roamer Smart 4 BMS has the Active Balancing BMS to solve the undercharging I was experiencing [that took weeks, not hours, to adjust!]

Steve

 

[GeoffL]

What are the typical individual cell voltages after the battery has entered stand by? Are these voltages consistent if you discharge and recharge? More importantly, if the battery is left idle overnight do the cell voltages change and equalise at all?

As promised in an earlier post, the screenshots are attached. To answer your questions:

1. Typical cell voltages depends upon SOC. At about 80%, the cells are within 0.003V of each other. Close to 100%, the cell voltages have a spread of just over 0.1V.
2. These voltages have not significantly changed (i.e. they are consistent) since I started using the van in earnest after installing the battery. However, that's only a couple of weeks!
3. If the battery is left overnight, the voltages seem to move slightly towards equalisation -- but I haven't had the battery long enough to give a definitive answer.

Note that the screenshots are after the van had spent most of the day soaking up sunlight. It was at low 80s SOC yesterday morning and stopped charging at about 08:30. I 'reset' the SCC and got about another hour of absorption. I then left Stourport at midday and spent about six hours on the road. Add that to today's solar 'harvest' and you have the reason why the current SOC is 99.9%!

 

[merl]

Well, the fact that you can actually get the battery to 100% show's that there's not a lot wrong.
It's a little baffling TBF because the absorption stage timer shouldn't start counting until you reach your 14.6v set voltage, by then the lifepo4 will be chock full because anything over 13.9v basically fully charges them.
What's the cable length (there and back) and gauge between SCC and battery?

 

[wildebus]

You mentioned about the different modes in the MPPT - fixed and adaptive? I don't know if you know how the adaptive mode works, but is ties in with what you are seeing. In essense, the higher the battery voltage on MPPT wakeup, the shorter the time the controller is in charge mode (that is just a rough summary - for full info and description, check the manual).
That adaptive mode is, IMO, pretty annoying and rubbish and using the fixed mode gives better results (IMO again).

I don't know what condition your Lead Acid battery is in, but I think I would be inclined to take it out the picture while you are investigating what is happening and to optimize the setup, especially if you suspect there is a problem with the battery (I don't know if you do or not?). Alpha would need the battery on a 'normal' setup first to check how it is operating.

A Lead/Lithium hybrid combination can work very well and really is not that complex to setup, but you can't just add a Lithium battery in without taking account of its requirements in the configuration. You have to have either Lithium compatible (configured) chargers, which will mean the Lead batteries will not be able to optimally charge, or you need a way to disconnect the Lithium from the chargers to allow the Lead to charge at the higher voltage they like.
On my setup, I have all the chargers and solar controllers configured for a Lead Acid profile, and have the Lithium Battery Bank disconnecting automatically at 14.19V (which is a full charge for Lithium) and the chargers continue with the Leads, by using the VSDR Lithium Controller device I developed specially for Hybrid setups. Been running like this for over 3 years without a hitch.

 

[GeoffL]

@merl @wildebus
Thanks for the continued help. Re. cable lengths: the SCC is connected to a busbar via the largest size cable that would fit the terminals. Cable run between these is about 1.5m in each direction. The batteries are connected to the busbar with cable the same size as is normally fitted to car batteries. These are wired to the busbar as per the illustration below and the connection to the positive busbar is protected with a 30A circuit breaker. The longest cable in this setup is just over a metre long.

AIUI, the adaptive absorption mode is time based (at least, that's what an Internet search told me) and this time is calculated in the morning from the SOC at that time. So, I'm guessing that the calculation suggests the battery is almost fully charged and establishes an absorption time that's way too short. However, setting a fixed absorption time worries me unless I can set the voltage to a value that isn't going to 'cook' the FLA cells! If 14.4V is safe in this respect, I can configure fixed absorption for (say) six hours per day, and this would solve the problem. Perhaps better to reduce the absorption voltage to 14.2V and accept that the FLA won't get a 100% charge if that won't damage it. Perhaps even lower to avoid the LFP ever getting into a over-voltage state?

 

[wildebus]

@merl @wildebus
Thanks for the continued help. Re. cable lengths: the SCC is connected to a busbar via the largest size cable that would fit the terminals. Cable run between these is about 1.5m in each direction. The batteries are connected to the busbar with cable the same size as is normally fitted to car batteries. These are wired to the busbar as per the illustration below and the connection to the positive busbar is protected with a 30A circuit breaker. The longest cable in this setup is just over a metre long.

AIUI, the adaptive absorption mode is time based (at least, that's what an Internet search told me) and this time is calculated in the morning from the SOC at that time. So, I'm guessing that the calculation suggests the battery is almost fully charged and establishes an absorption time that's way too short.

The Adaptive mode is described in the manual as being worked on the state of charge of the battery, but it really is not - and as the manual goes on to more correctly describe, it is based on the battery voltage in the morning.
The trouble is the battery voltages used are only relevant to Lead Acid batteries and absolutely not Lithium. See the table below from the manual ....

If the battery is above 12.6V (which is virtually full for a Lead, but would be very very low for Lithium), the time is reduced to One Hour (1/6th of normal). Use Adaptive with Lithium and your controller will run for until around 7AM in the summer and then do nothing for the rest of the day.
This is why one reason why virtually no one uses the Adaptive setting and there was a collective cheer when Victron finally added an alternative to that setting some years back.

I think you are overworrying about higher absorption times.

However, setting a fixed absorption time worries me unless I can set the voltage to a value that isn't going to 'cook' the FLA cells! If 14.4V is safe in this respect, I can configure fixed absorption for (say) six hours per day, and this would solve the problem. Perhaps better to reduce the absorption voltage to 14.2V and accept that the FLA won't get a 100% charge if that won't damage it. Perhaps even lower to avoid the LFP ever getting into a over-voltage state?

View attachment 143566

 

[merl]

However, setting a fixed absorption time worries me unless I can set the voltage to a value that isn't going to 'cook' the FLA cells! If 14.4V is safe in this respect,

14.2 to 14.4v will be fine for a lead acid battery, after all that's the sort of typical alternator charge voltage seen for hours on end in millions of cars world wide.
Nothing wrong with the wiring either. Everything considered it certainly looks like the adaptive mode feature is causing the issue and TBF it seems like a daft feature anyway because it's only considering time, to be really useful and accurate it needs to consider Ah, ie time AND current.

 

[GeoffL]

14.2 to 14.4v will be fine for a lead acid battery, after all that's the sort of typical alternator charge voltage seen for hours on end in millions of cars world wide.
Nothing wrong with the wiring either. Everything considered it certainly looks like the adaptive mode feature is causing the issue and TBF it seems like a daft feature anyway because it's only considering time, to be really useful and accurate it needs to consider Ah, ie time AND current.

Thanks for that. You're correct re. alternator charge voltage. My instinct told me that those batteries in millions of cars world wide are of different design as they provide a large starting current and then are immediately fully recharged and kept topped up whereas leisure batteries are designed for deep cycle. Victron further reinforced that instinct by providing different profiles for AGM, Gel, etc. That said, they are the same basic chemistry...

I think you are overworrying about higher absorption times.

I strongly suspect that you're correct -- particularly if everyone else isn't using the default adaptive absorption. TBH, I don't think I should be concerned about the FLA battery only being charged to (say) 75% as it's there primarily as 'ballast' and for a little resilience should the LFP get run down (which I suspect won't happen April to September given experience this far!) In any case, I still have the FLA's brother as a spare should the one in the van become 'toast'.

All in all, I suspect that setting the absorption voltage to 14.2 and fixed absorption time of six to eight hours might give best results. The van's alternator produces 14.3V measured at a cab cigar lighter socket, so that's probably 14.2V at the LB (which is at the other end of the moho)...

Thanks to all for the help.

 

[wildebus]

If you want to see some information about how the Lithium and Lead interact in use, you may want to have a look at this series on my Website - https://wildebus.com/hybrid-battery/.
This was written before I had my VSDR Lithium Controller developed so it directly relates to your setup in terms of connections (The VSDR is all about optimising the charge, it doesn't really impact the discharge/use).

Just for some extra info, my Hybrid Bank has changed over time as I have tried different batteries, added and removed batteries. Always had 300Ah of Lead-Carbon, but varied between 100Ah, and 200Ah of Lithium and currently have had 320Ah of Lithium for the last year or so? (I introduced Steves old KS Energy Lithium to the pair of Polarmax Lithiums), but the actual operation and results have never varied and has workesd without fail). Current Charge State and Current distribution from solar is:

 

[merl]

Ignoring the parrot fashion 'knowledge' for a moment re not floating LFP, has anyone seen any actual scientific data saying that holding a LFP battery at 14.3-14.4 volts, as would happen WRT direct coupling to an alternator, is actually harmful and if so HOW harmful? I scoured the internet for science based data on this a while ago and there wasn't any conclusive data either way but time has moved on and I haven't searched again recently.