Any insight to what the negative effect of controlling my lithium batteries via voltage instead of BMS comms?

I have a Sunsynk inverter and 2x Narada 48NPFC100 batteries. The batteries keep on falling out of sync with the SOC reported by the BMS.  If I have a few days of bad PV generation where the batteries are not charged to 100%, the batteries report 100% SOC when they are far from it (especially when comparing to the actual voltage).  This also happens when the batteries are kept at 100% SOC for a few days.

I am thinking it may be beneficial to let the inverter control the batteries via voltage instead for letting it communicate with the BMS but I was wondering what the negative effects would be (short or long term).

I will still be able to read the BMS reported SOC in my monitoring software, allowing me to be able to keep an extra eye on it

Thanks in advance

None what so ever, the BMS still takes care of charge and discharge, not the inverter. Just make sure the absorption voltage is set to the correct voltage according to the battery specs, set float voltage at the same and equalization voltage to 0 hours.

As you go along compare the BMS reported voltage to the voltage according to the inverter, especially at low SOC (between 10% and 15% SOC), and use this as a guide to set your low voltage cut off. You want the inverter to switch over to AC/ turn off (when no AC is available) before the BMS turns off the battery. 

1 hour ago, I84RiS said:

None what so ever, the BMS still takes care of charge and discharge, not the inverter. Just make sure the absorption voltage is set to the correct voltage according to the battery specs, set float voltage at the same and equalization voltage to 0 hours.

As you go along compare the BMS reported voltage to the voltage according to the inverter, especially at low SOC (between 10% and 15% SOC), and use this as a guide to set your low voltage cut off. You want the inverter to switch over to AC/ turn off (when no AC is available) before the BMS turns off the battery. 

Thank you. You have confirmed my exact thinking and plan... I have moved it across and plan to run the batteries down while monitoring the SOC one afternoon / evening to "calibrate" my settings properly

Thanks again

1 hour ago, I84RiS said:

compare the BMS reported voltage to the voltage according to the inverter

Also if you have a good Multimeter, check the battery voltage manually too to verify.
I found my BMS to be just about the same as my Fluke whilst the Inverter reported voltages are anywhere between 0,3 and 1 V more than the BMS depending on the SOC and whether the battery is under charge or discharging

10 hours ago, zsde said:

Also if you have a good Multimeter, check the battery voltage manually too to verify.
I found my BMS to be just about the same as my Fluke whilst the Inverter reported voltages are anywhere between 0,3 and 1 V more than the BMS depending on the SOC and whether the battery is under charge or discharging

I also find that running my system without BMS com's gives me various options to maximize the performance of my battery.

1-- I can control the charging current using the settings in my inverters, based on the weather I charge my battery at a slow pace if weather permits.

2 -- I can set the absorption and floating voltage, usually a bit bellow the battery OEM recommended values and I can still get the battery to 99% SOC

There is just one thing that really annoys me big time, the cycle count on the Seplos BMS, it counts 1 cycle every time SOC drops below 80%, I know how to fix it, but I will leave like it is in case some warranty issue and the OEM tells me that I've changed settings thus no more guaranty.

See last 24 hour of my system.

On 2022/10/21 at 9:30 PM, Jay-Dee said:

I am thinking it may be beneficial to let the inverter control the batteries via voltage instead for letting it communicate with the BMS but I was wondering what the negative effects would be (short or long term).

I will still be able to read the BMS reported SOC in my monitoring software, allowing me to be able to keep an extra eye on it

I agree. I run my batteries that way now for 2½ years and see no disadvantage. Important is to set the inverter to battery manufacturer specs with respect to bulk and floating charge voltages. Read also in my thread "Are BMS reliable?".

14 hours ago, zsde said:

Also if you have a good Multimeter, check the battery voltage manually too to verify.
I found my BMS to be just about the same as my Fluke whilst the Inverter reported voltages are anywhere between 0,3 and 1 V more than the BMS depending on the SOC and whether the battery is under charge or discharging

Very interesting... I will keep an eye on it

4 hours ago, Antonio de Sa said:

Is this Node Red?   I am just getting into Home Assistant but still trying to figure out what I would use it for besides a nice dash board.  I have my automation all sorted within it's own app as well as my solar in its own.  Thinking HA / Node Red may just add a level of complication for no benefit... Help me justify 🤣

23 minutes ago, Beat said:

I agree. I run my batteries that way now for 2½ years and see no disadvantage. Important is to set the inverter to battery manufacturer specs with respect to bulk and floating charge voltages. Read also in my thread "Are BMS reliable?".

Thanks... Good to hear from a "long term" experience.

Re the spec's...  The funny thing is that the manufacturers spec on paper is vastly different to that of the inverter when communicating directly with the battery. Paper says charge at 54v where as BMS wanted 57.6 which to me is every high for a 15S config

Re your thread... I read it a while back and thought blah, it does not happen to me... Will go back an re-look at it again

8 hours ago, Jay-Dee said:

Thanks... Good to hear from a "long term" experience.

Re the spec's...  The funny thing is that the manufacturers spec on paper is vastly different to that of the inverter when communicating directly with the battery. Paper says charge at 54v where as BMS wanted 57.6 which to me is every high for a 15S config

Re your thread... I read it a while back and thought blah, it does not happen to me... Will go back an re-look at it again

It is highly unlikely that it is actually charging at 57.6V. 

What is the voltage according to the inverter display when it reached 99% SOC? Then also how far does the voltage of the battery climb from 99% to full?

If it is really charging at 57.6v then it will go all the way up to 57.6v as the last stage of a lithium charging profile would generally be constant voltage (57.6v in this case) while the Amps slowly reduce. 

Hi @Jay-Dee

The main reason for using an interface to the battery rather than just reading the battery voltage and working out the SOC from there is that it is not possible to get an accurate SOC in lithium batteries using the voltage because of the way that voltages rise in lithium batteries.  As you can see from this picture, the voltage difference between a 40% & 80% SOC on a lead acid is 6V, on a lithium is 0.5V. Voltages changes on lithium batteries is not linear the way it is with lead acids with big voltage changes only happening at the top and bottom of the SOC's.  

Most battery BMS's actually have a shunt in them that enables the BMS to know how much power has travelled through it in both directions and what the SOC is sitting at. Trying to get an accurate battery SOC from the voltage on a lithium battery is not possible. 

11 hours ago, I84RiS said:

It is highly unlikely that it is actually charging at 57.6V. 

What is the voltage according to the inverter display when it reached 99% SOC? Then also how far does the voltage of the battery climb from 99% to full?

If it is really charging at 57.6v then it will go all the way up to 57.6v as the last stage of a lithium charging profile would generally be constant voltage (57.6v in this case) while the Amps slowly reduce. 

So it would sit around 54 / 54.5v and then spike up to 57.6 for a very brief period and float somewhere between 54.5 and 55.5. So I am not sure where it was reading this voltage from / why but the BMS certainly would not let it float that high

The battery specs are charge and 54.0 and float between 54.0 to 54.5. I am currently charging at 54.0 and floating at 54.5. So far without any issues

39 minutes ago, Sc00bs said:

Hi @Jay-Dee

The main reason for using an interface to the battery rather than just reading the battery voltage and working out the SOC from there is that it is not possible to get an accurate SOC in lithium batteries using the voltage because of the way that voltages rise in lithium batteries.  As you can see from this picture, the voltage difference between a 40% & 80% SOC on a lead acid is 6V, on a lithium is 0.5V. Voltages changes on lithium batteries is not linear the way it is with lead acids with big voltage changes only happening at the top and bottom of the SOC's.  

Most battery BMS's actually have a shunt in them that enables the BMS to know how much power has travelled through it in both directions and what the SOC is sitting at. Trying to get an accurate battery SOC from the voltage on a lithium battery is not possible. 

Thank you for the feedback and while I cant argue with your graph comparison... I can tell you that in the real world with my batteries, it is a very different graph despite the manual suggestion it should look more like yours.  The load has a significant impact on the voltage too.

My issue is my BMS is not reliable... While the Ah in in vs out may be recorded and reported vs the SOC "correctly"... If I dont fully charge the batteries for a few days in a row, it will report 100% charge with full Ah / capacity showing but at 48 / 49v (confirmed with a multi-meter). When load shedding kicks in and the batteries are under load, they shut off quickly due to low voltage while SOC is still reporting 60 / 70+ %....     I then have the reverse issue in summer now when I dont use the batteries for a number of days in a row and the system floats them at 100% but they seem to "leak" power as when load shedding arrives, I drop from 100% to 10% SOC quickly and the system shuts off while the batteries are still actually at 49 / 50v.

Previously, I had the Narada on an Axpert inverter which does not communicate with them. Since the upgrade to my Sunsynk, I am now picking up the BMS issues and hence the move away from communication control

2 hours ago, Sc00bs said:

The main reason for using an interface to the battery rather than just reading the battery voltage and working out the SOC from there is that it is not possible to get an accurate SOC in lithium batteries using the voltage because of the way that voltages rise in lithium batteries.

But the issue is not to accurately know the SOC but to perform the charging procedure that manages the battery best. In that respect the battery voltage is the most reliable parameter that never lies. I use the BMS SOC reading only for my information. Read my thread "Are BMS reliable?".

4 hours ago, Sc00bs said:

The main reason for using an interface to the battery rather than just reading the battery voltage and working out the SOC from there is that it is not possible to get an accurate SOC in lithium batteries using the voltage because of the way that voltages rise in lithium batteries.  As you can see from this picture, the voltage difference between a 40% & 80% SOC on a lead acid is 6V, on a lithium is 0.5V. Voltages changes on lithium batteries is not linear the way it is with lead acids with big voltage changes only happening at the top and bottom of the SOC's.  

Most battery BMS's actually have a shunt in them that enables the BMS to know how much power has travelled through it in both directions and what the SOC is sitting at. Trying to get an accurate battery SOC from the voltage on a lithium battery is not possible. 

If this is the case, that would make setting a cutoff voltage pointless. With Gel batteries, I tend to set the cutoff at 44V(in closed circuit). This makes sure the DoD of 50% isn't exceeded but by the time a lithium battery reached 44V the DoD would have been exceeded damaging the battery. This is why I am super sceptic about lithium's running without BMS, you can make it work but there is nothing protecting the batteries from deep-discharge. This is the point of lithium's, they have a much longer life cycle but constantly exceeding the DoD will wreck the battery in no time. 

Just make sure you check the DoD and other specs like that and you may be able to find a voltage that would work for this. It'll be something weird like 56,5V or something like that, just make sure your batteries aren't run into the ground.

29 minutes ago, Psy said:

If this is the case, that would make setting a cutoff voltage pointless. With Gel batteries, I tend to set the cutoff at 44V(in closed circuit). This makes sure the DoD of 50% isn't exceeded but by the time a lithium battery reached 44V the DoD would have been exceeded damaging the battery. This is why I am super sceptic about lithium's running without BMS, you can make it work but there is nothing protecting the batteries from deep-discharge. This is the point of lithium's, they have a much longer life cycle but constantly exceeding the DoD will wreck the battery in no time. 

Just make sure you check the DoD and other specs like that and you may be able to find a voltage that would work for this. It'll be something weird like 56,5V or something like that, just make sure your batteries aren't run into the ground.

@Psy   Depending on the BMS your battery is equipped with, in my case running without comm's one can clearly see that once battery voltage get's to 48 V it asks the inverter for a bit of charge to reach 50 V, so my battery never reaches the low voltage cut out recommended by the OEM of 44 V see attached dashboard.

26 minutes ago, Antonio de Sa said:

Depending on the BMS your battery is equipped with, in my case running without comm's one can clearly see that once battery voltage get's to 48 V it asks the inverter for a bit of charge to reach 50 V, so my battery never reaches the low voltage cut out recommended by the OEM of 44 V see attached dashboard.

The BMS asks? Or the inverter, which has the various voltages loaded, says "oooh... I better charge this a bit"? The graph tells us what is happening, but not who asked for that to happen.

Not trying to get into a dispute with you, just trying to understand. Purely academic. 

28 minutes ago, Bobster. said:

The BMS asks? Or the inverter, which has the various voltages loaded, says "oooh... I better charge this a bit"? The graph tells us what is happening, but not who asked for that to happen.

Not trying to get into a dispute with you, just trying to understand. Purely academic. 

@Bobster. Good question indeed, however, looking at the graph the inverter does what it suppose to do, the settings is, back to utility at 50.2 Volts. According to Seplos BMS manual defaults settings start charge at > 1 Amps once it reaches 48 V that is why I believe the BMS allows the inverter to charge back to 50 Volts. Also my understanding is purely academic. 

1 hour ago, Antonio de Sa said:

@Psy   Depending on the BMS your battery is equipped with, in my case running without comm's one can clearly see that once battery voltage get's to 48 V it asks the inverter for a bit of charge to reach 50 V, so my battery never reaches the low voltage cut out recommended by the OEM of 44 V see attached dashboard.

 

1 hour ago, Antonio de Sa said:

@Psy   Depending on the BMS your battery is equipped with, in my case running without comm's one can clearly see that once battery voltage get's to 48 V it asks the inverter for a bit of charge to reach 50 V, so my battery never reaches the low voltage cut out recommended by the OEM of 44 V see attached dashboard.

@Psy As you can see from the graph bellow my battery never gone bellow 48 V and that is without BMS comm's, just using inverter voltage settings. I'm pretty confident that I'm looking well after the health of my battery. 

10 minutes ago, Antonio de Sa said:

@Psy As you can see from the graph bellow my battery never gone bellow 48 V and that is without BMS comm's, just using inverter voltage settings. I'm pretty confident that I'm looking well after the health of my battery. 

That's not bad, the resolution on the Voltmeter is pretty decent. It does depend on what battery you have and what the manufacturer says the safe DoD is. For example, lets say your lithium drains according to this graph and has a safe DoD of 80%:

That would mean that when the battery is drained below 54V, you are technically way out of the DoD and are chipping away at the number of cycles you are going to get from the unit.

This is all hypothetical, you should be able to get the safe DoD range from the battery manufacturers and the specific values will change from one to another. I am just saying that is something you need to take into account when setting up a lithium without BMS coms

3 minutes ago, Psy said:

I am just saying that is something you need to take into account when setting up a lithium without BMS coms

This is 100% correct and is very definatately taken into account. Narada has a reasonable amount of information on this

 

Float = 54.0v
Charge current = 54.0 to 54.5v
LLVD (low voltage non essential load disconnect) = 47v
BLVD (low voltage battery disconnect) = 43.2v
Full battery discharge (0% SOC) = 40.5v

Implementing the above settings is not an issue nor is maintaining them. My question was raised that if I stick to these parameters, is there any long term negative effect.

Currently in bad weather the batteries BMS is letting it go down into the very low 40s (sub 42v) as it things the SOC is still fine which is obviously a major concern for me. In good weather, the BMS things the battery is depleted at 49v... So my mind says it is safer to run on voltage than rely on the BMS and SOC.  

Also without a firmware update or the likes of from Narada, the batteries are not very useful when using the comms as I am lucky to get around 40Ah out of the 200

8 minutes ago, Psy said:

That's not bad, the resolution on the Voltmeter is pretty decent. It does depend on what battery you have and what the manufacturer says the safe DoD is. For example, lets say your lithium drains according to this graph and has a safe DoD of 80%:

That would mean that when the battery is drained below 54V, you are technically way out of the DoD and are chipping away at the number of cycles you are going to get from the unit.

This is all hypothetical, you should be able to get the safe DoD range from the battery manufacturers and the specific values will change from one to another. I am just saying that is something you need to take into account when setting up a lithium without BMS coms

As you can see, I'm well above the recommended OEM values. I don't need to discharge the battery to 100% DOD, the battery is there to cope with load shedding.

Presently it goes down to about 40% Soc. daily, I'm not worried about cycle count, Seplos has a stupid cycle count, it adds one cycle every time the SOC goes bellow 80% SOC

2 minutes ago, Antonio de Sa said:

As you can see, I'm well above the recommended OEM values. I don't need to discharge the battery to 100% DOD, the battery is there to cope with load shedding.

Presently it goes down to about 40% Soc. daily, I'm not worried about cycle count, Seplos has a stupid cycle count, it adds one cycle every time the SOC goes bellow 80% SOC

100% and totally on the same page... 

My batteries are 4 years old already... The first 3.5 years were controlled via voltage and the SOH when reading the BMS is sitting at 96%. I dont know why I did not think about this before posting the topic as it answer my question 🤣

I use my for both load shedding and savings over night / charge during the day but very rarely go below 30% DOD when they are working as they are supposed to.

The above SOH has answered it for me... I am staying on voltage control for now... Perhaps Narada will answer their emails and send me a firmware update one day

33 minutes ago, Jay-Dee said:

Implementing the above settings is not an issue nor is maintaining them. My question was raised that if I stick to these parameters, is there any long term negative effect.

OK great, well as long as you are being meticulous about the parameters it should be fine. At that point the only additional thing you get from a BMS is battery balancing. Take a bunch of cells in series, put them on charge and you will see that some of the cells will be full before the others. The BMS would then try balancing out the load so you don't end up over charging. Even then, the BMS should be able to do that without needing to talk to your inverter (As long as its a decent BMS) so you should be fine using "User Defined" battery parameters based on voltage.

25 minutes ago, Psy said:

OK great, well as long as you are being meticulous about the parameters it should be fine. At that point the only additional thing you get from a BMS is battery balancing. Take a bunch of cells in series, put them on charge and you will see that some of the cells will be full before the others. The BMS would then try balancing out the load so you don't end up over charging. Even then, the BMS should be able to do that without needing to talk to your inverter (As long as its a decent BMS) so you should be fine using "User Defined" battery parameters based on voltage.

@Psy Yes, the BMS takes care of that every time it reaches 100% soc.