Strange SOC graph

[BigDean]

I have plenty of solar on the van which reaches the 14.4v most days quite easily for a couple of hours then steps down to 13.8v until the sun goes down. All as expected (I think)

But my SOC doesn’t go to 100%. It is reducing every day. See here:

BMV settings:

Seems strange.

Any ideas?

Thanks

 

[Pausim]

It looks to me that your solar panel is no longer making sufficient power to achieve a full charge. The trend is falling slowly which is entirely consistent with the declining sunlight and the worsening angle of a horizontal panel. I am no expert but I don’t think the voltage tells the whole story, the voltage may held at 14.4V for a while but there might be very little power behind it i.e. not enough amps. Perhaps the solar charger is dropping back to a maintenance voltage because the output (Amps) falls below the set threshold (tail current). In other words it is reacting to the lack of amps not because the battery is full but because the solar panel is not producing enough.

 

[Raul]

What’s the settings in the mppt? As Pausim says, you don’t have enough charge to bring it to 100%. If you have a tail current set in the solar controller, you need to disable it at this time. It falsely drop it into float before reaches full charge. Is this lead or Li?

 

[BigDean]

What’s the settings in the mppt? As Pausim says, you don’t have enough charge to bring it to 100%. If you have a tail current set in the solar controller, you need to disable it at this time. It falsely drop it into float before reaches full charge. Is this lead or Li?

It looks to me that your solar panel is no longer making sufficient power to achieve a full charge. The trend is falling slowly which is entirely consistent with the declining sunlight and the worsening angle of a horizontal panel. I am no expert but I don’t think the voltage tells the whole story, the voltage may held at 14.4V for a while but there might be very little power behind it i.e. not enough amps. Perhaps the solar charger is dropping back to a maintenance voltage because the output (Amps) falls below the set threshold (tail current). In other words it is reacting to the lack of amps not because the battery is full but because the solar panel is not producing enough.

It’s li. Looking in the MPPT (Victron) it is set to 2 hours fixed absorption, so will change that to adaptive which I guess will sort it. There is no tail current I can see.

Thanks chaps.

 

[gpat]

Are you sure it's not just the days getting shorter and the sun being lower in the sky?

^{_{Subscribers  do not see these advertisements}}

 

[autorouter]

Another possibility is that the 'charge efficiency' is a trifle overoptimistic at 99%. The peak SOC is only dropping by less than 1% per day. You could try 98% to see if that corrects it.

 

[Raul]

You will not be wrong by selecting the victron preset of LiFePo4, then just modify the absorption voltage to 14,4v if you like. I’m confident it will fully charge on default 14,2v if enough charge is available. You just discovered that solar is not enough for this time of year.

 

[BigDean]

Are you sure it's not just the days getting shorter and the sun being lower in the sky?

You will not be wrong by selecting the victron preset of LiFePo4, then just modify the absorption voltage to 14,4v if you like. I’m confident it will fully charge on default 14,2v if enough charge is available. You just discovered that solar is not enough for this time of year.

Will try that.

There is definitely more solar available than is going in. The curve is too uniform for it to be a shortage of sun.

 

[BigDean]

Will try that.

There is definitely more solar available than is going in. The curve is too uniform for it to be a shortage of sun.

Just selected the Victron preset. I think that’s what I did originally. In that preset, it sets the absorption to fixed for 2 hours. Looks like that isn’t enough.

 

[Hoovie]

Will try that.

There is definitely more solar available than is going in. The curve is too uniform for it to be a shortage of sun.

I would suggest you spotted the problem when you said it was set to 2 hours fixed absorption. You could change to adaptive, but why have you set to just 2 hours? That is a very low value. The Victron default is 8 hours IIRC (2 hours default seems very strange) and there are very few batteries that would have a recommended absorption/CV time of less than 5 hours. I would keep at fixed, but bump the duration upto 6 or 8 (unless you know what your specific battery recommends and set to that).

^{_{Subscribers  do not see these advertisements}}

 

[BigDean]

I would suggest you spotted the problem when you said it was set to 2 hours fixed absorption. You could change to adaptive, but why have you set to just 2 hours? That is a very low value. The Victron default is 8 hours IIRC (2 hours default seems very strange) and there are very few batteries that would have a recommended absorption/CV time of less than 5 hours. I would keep at fixed, but bump the duration upto 6 or 8 (unless you know what your specific battery recommends and set to that).

Victron sets it to 2 hours

 

[Hoovie]

Victron sets it to 2 hours

Up it

 

[BigDean]

Up it

 

[autorouter]

Looking in the MPPT (Victron) it is set to 2 hours fixed absorption, so will change that to adaptive which I guess will sort it.

I thought that Lithiums don't need a timed absorption stage, that stage ends when the tail current drops below the preset threshold. At that point the battery should be at 100% charge. If that is happening before the sun drops too much, then it should be reaching 100% every day.

At some point in the year there won't be enough solar to get it to 100%. But that depends not only on the sun but on how much you are taking out of the batteries. If the MH is in storage it should reach 100% nearly every day.

Then there is the question - do you really want your LiFePO4 batteries to be charging to 100% every day?

 

[BigDean]

I thought that Lithiums don't need a timed absorption stage, that stage ends when the tail current drops below the preset threshold. At that point the battery should be at 100% charge. If that is happening before the sun drops too much, then it should be reaching 100% every day.

At some point in the year there won't be enough solar to get it to 100%. But that depends not only on the sun but on how much you are taking out of the batteries. If the MH is in storage it should reach 100% nearly every day.

Then there is the question - do you really want your LiFePO4 batteries to be charging to 100% every day?

It on the driveway, with only the essential services running.

Tail current is disabled as default on the Victron LiFoPo preset.

I expect the battery to outlive the MoHo, so happy for it to be topped up to 100% every day.

Thanks

^{_{Subscribers  do not see these advertisements}}

 

[autorouter]

You have to be a bit careful here. There is a mismatch between the SOC calculation of 100% and the absorption voltage/tail current readings that decide it's reached 100%. You say there's plenty of sun, so if the SOC calculation is a bit wrong and you change the charge settings to force extra charge in, then you are in danger of overcharging, which lithium batteries don't like at all.

If that usage graph is typical, it never drops below 70%, so you'd be better cycling between 60 to 90% rather than 70 to 100%. How to do that is, as they say, another story.

 

[BigDean]

You have to be a bit careful here. There is a mismatch between the SOC calculation of 100% and the absorption voltage/tail current readings that decide it's reached 100%. You say there's plenty of sun, so if the SOC calculation is a bit wrong and you change the charge settings to force extra charge in, then you are in danger of overcharging, which lithium batteries don't like at all.

If that usage graph is typical, it never drops below 70%, so you'd be better cycling between 60 to 90% rather than 70 to 100%. How to do that is, as they say, another story.

I have set the absorption to be dynamic, extended to 4 hours and reset the SOC to 100% (I charged on mains this morning to ensure it was full).

I will monitor and report back in a few days.

Thanks all.

 

[Pausim]

LiFePO4 batteries do not need an absorption phase. An Li charging profile should have a single bulk charge phase where the voltage rises as the battery fills, this bulk phase ends when a) the pre set Voltage is achieved and b) the Amps fall below the current threshold. The charger may then revert to a maintenance charge of say 13.8V.

In autumn and winter the output of solar panels is much reduced. The mppt will charge at 14.4V and relies on the Amps falling below the current threshold to know when the battery is full. Because there is very little power coming from the panels the Amps fall below the current threshold before the battery is full and the mppt goes into it’s maintenance charge phase of 13.8V. Adding an absorption phase may keep the voltage higher for longer but it will not make much difference if the panels are not producing enough power to fully charge the battery.

Your SOC graph is entirely consistent with the sun shining from an increasingly lower angle and for less time each day. It is to be expected at this time of year and I don’t think you have anything to worry about.

 

[BigDean]

LiFePO4 batteries do not need an absorption phase. An Li charging profile should have a single bulk charge phase where the voltage rises as the battery fills, this bulk phase ends when a) the pre set Voltage is achieved and b) the Amps fall below the current threshold. The charger may then revert to a maintenance charge of say 13.8V.

In autumn and winter the output of solar panels is much reduced. The mppt will charge at 14.4V and relies on the Amps falling below the current threshold to know when the battery is full. Because there is very little power coming from the panels the Amps fall below the current threshold before the battery is full and the mppt goes into it’s maintenance charge phase of 13.8V. Adding an absorption phase may keep the voltage higher for longer but it will not make much difference if the panels are not producing enough power to fully charge the battery.

Your SOC graph is entirely consistent with the sun shining from an increasingly lower angle and for less time each day. It is to be expected at this time of year and I don’t think you have anything to worry about.

Thanks again, but what I don't understand is why it does not achieve 100% SOC. Even on sunny days, it never gets to 100% SOC.

The reduction in SOC on the graph would not be so uniform. You can see a couple of days where it was bad weather, but on the other days there was enough sunshine to achieve 100%.

 

[autorouter]

LiFePO4 batteries do not need an absorption phase. An Li charging profile should have a single bulk charge phase where the voltage rises as the battery fills, this bulk phase ends when a) the pre set Voltage is achieved and b) the Amps fall below the current threshold. The charger may then revert to a maintenance charge of say 13.8V.

I thought that Lithiums don't need a timed absorption stage, that stage ends when the tail current drops below the preset threshold. At that point the battery should be at 100% charge. If that is happening before the sun drops too much, then it should be reaching 100% every day.

Pausim is right that no absorption phase is needed for lithiums. But I still think that the batteries are actually reaching 100% most days, and the SOC calculation is not accurate enough. Because the MH is not being used, the demand for battery power is small, and the solar is quite enough to top up the battery nearly every day.

^{_{Subscribers  do not see these advertisements}}

 

[Gromett]

Is there a setting for Peukert's?

Peukert's law - Wikipedia

en.wikipedia.org

If you have this set wrong and pull higher current from the battery than expected then your SOC measurement based on coulomb counting will drift.

I can explain in more detail if necessary?

 

[BigDean]

Is there a setting for Peukert's?

Peukert's law - Wikipedia

en.wikipedia.org

If you have this set wrong and pull higher current from the battery than expected then your SOC measurement based on coulomb counting will drift.

Yep, set in the BMV:

 

[Gromett]

Yep, set in the BMV:

View attachment 550557

OK I see it. check with your battery supplier for the correct setting for your battery. This setting completely explains your problem if this is wrong.

Unless you have lithium batteries this looks a bit low.

 

[Gromett]

8. Battery capacity and Peukert exponent

www.victronenergy.com

 

[autorouter]

I don't think the Peukert coefficient is a problem here. The value of 1.05 is about right for lithiums. It is relevant if the amps in/out of the battery is a high value, running an inverter for example. It accounts for higher amps taking more than it's fair share of the battery capacity compared to low amps values.

Looking at the graph in the first post, on most days the SOC drops by about 10% over 24 hours, so that's an average amps value of 10/24 or about half an amp. And since the graph is a steady fall over 24 hours, not big jumps, it's probably a fairly constant half an amp all the time. The Peukert factor is not going to affect that small value very much.

By the way, if you don't follow the 'Tip' in Gromett's link, you've not gone stupid, it's wrong. 5C is the same as C0.2, and .2C is the same as C5. I've not checked the other calculations in detail, but they look correct.

^{_{Subscribers  do not see these advertisements}}

 

[Gromett]

I don't think the Peukert coefficient is a problem here. The value of 1.05 is about right for lithiums. It is relevant if the amps in/out of the battery is a high value, running an inverter for example. It accounts for higher amps taking more than it's fair share of the battery capacity compared to low amps values.

Looking at the graph in the first post, on most days the SOC drops by about 10% over 24 hours, so that's an average amps value of 10/24 or about half an amp. And since the graph is a steady fall over 24 hours, not big jumps, it's probably a fairly constant half an amp all the time. The Peukert factor is not going to affect that small value very much.

Just re-read some earlier posts, I missed the lithium mention. 1.05 is about right for lithium on average, but may still be out a little and could still be the cause.
If the battery IS reaching 100% SOC but the monitor is out due to coulomb counting then the Peukert setting could well be the cause.

I am not convinced this is the cause, but it is certainly the first thing I would look at if the battery is getting to 100% SOC.

Bit hard to tell from that graph what the usage is to be honest. It covers a number of days and it could be he pulls big'ish chunks that get averaged out on the graph.

 

[Pausim]

The settings shown in the pictures appear to be the ones for the battery monitor. What are the mppt settings? The current threshold of the mppt will be the one that will be dropping the charge voltage back to 13.8, not the BMV settings.

 

[Gromett]

The settings shown in the pictures appear to be the ones for the battery monitor. What are the mppt settings? The current threshold of the mppt will be the one that will be dropping the charge voltage back to 13.8, not the BMV settings.

The voltage is not relevant to the graph, it is a coulomb counting SOC graph. The SOC graph does not use current voltage to calculate the SOC as Li batteries especially LiFePO4 ones have a pretty flat voltage across much of the charge/discharge range.

 

[autorouter]

Looking in detail at the graph, I notice that on most days the peaks reach an imaginary curve that gradually falls with time. I also notice that some of the peaks have flat tops, which to me indicates that the charger has stopped charging, presumably because it thinks the battery is full. That's why I think the battery is reaching 100% SOC. What I can't explain is that the flat tops seem to be sloping, with the same general slope as the imaginary curve I mentioned. I think the imaginary curve a the drift of the SOC calculation, for whatever reason. Peukert coefficient, charging efficiency, or even an inaccurate value for the nominal capacity, which might be 95Ah rather than 100 Ah.

 

[Pausim]

The voltage is not relevant to the graph, it is a coulomb counting SOC graph. The SOC graph does not use current voltage to calculate the SOC as Li batteries especially LiFePO4 ones have a pretty flat voltage across much of the charge/discharge range.

I agree but it is relevant to why the charging is ceasing before the battery is full. The OP was thinking that because the mppt charging voltage dropped from 14.4V to 13.8V the battery must be 100%. I think the battery monitor is being accurate and the mppt is getting it wrong. The mppt is probably dropping the output voltage not because the battery is full but because the current passing through it has dropped off below it’s own threshold. In other words the the mppt is fooling itself that the battery is full but not fooling the battery monitor, which accurately reflects the coulomb count.

^{_{Subscribers  do not see these advertisements}}