Monitoring lithium batteries

[Alan de Enfield]

19 minutes ago, Detling said:

Remember insulation just slows down heat transfer it does not provide heat, it also slows down the warming up heat coming in. Could you include a small loop of central heating pipe which provides a few watts of heat it only needs to keep the batteries above 0 degrees so a metre of 10mm copper pipe would probably do with insulated feed and return pipes they can all be micro bore as you don't need a massive flow rate..

If you have all this electricity 'going spare' then maybe consider heat trace cables - something like 10w per metre is what we use to stop out water pipes freezing - just keeps them above freezing temps.

 

Example :

https://www.traceheatinguk.co.uk/premade-trace-heating-kits-with-thermostat

 

 

[WotEver]

1 hour ago, Mike the Boilerman said:

As usually seems to happen when lithium batteries are discussed, this thread contains more heat than light (for me at least). 

Which bit is unclear to you Mike?  I’ll do my best to explain (assuming I know the answer to your question). 

Edited August 14, 2018 by WotEver
Clarity

[peterboat]

2 hours ago, Detling said:

Remember insulation just slows down heat transfer it does not provide heat, it also slows down the warming up heat coming in. Could you include a small loop of central heating pipe which provides a few watts of heat it only needs to keep the batteries above 0 degrees so a metre of 10mm copper pipe would probably do with insulated feed and return pipes they can all be micro bore as you don't need a massive flow rate..

There are heater pipes there already from the whispergen so will add some pipes into the battery boxes

[MtB]

2 hours ago, WotEver said:

Which bit is unclear to you Mike?  I’ll do my best to explain (assuming I know the answer to your question). 

 

Thanks! As per the OP, how to monitor the SoC of a set of LiFePO4 batteries. I don't think any consensus has been reached in this thread. 

 

I think we are agreed on the following points though:

 

1) The SoC cannot be monitored during charging.

2) All claims that one need not worry about it because "the BMS handles all that" are by definition, bobbins. 

3) THE SoC can be determined by measuring terminal voltage when the batteries are at rest, because LiFePO4 batteries do not exhibit surface charge.

 

I guess the bit I am unsure about is whether the SoC can be determined during discharge. And if not, why not. And if there are any devices which interpret the terminal voltage and display SoC directly, like a Smartgauge might if one was available for LiFePO4 battery banks.

 

I am also unsure what a 'BMS' does. I think it just disconnects the battery before it gets overcharged or undercharged. If this is true, how does it know the SoC during charging, given 1) ? AH counting? 

 

[WotEver]

Okay, let’s see where we can get to...

 

36 minutes ago, Mike the Boilerman said:

The SoC cannot be monitored during charging.

Well... if you’re using a CC (constant current) charge source then yes you can (I was wrong previously) because the slowly rising charge voltage directly correlates to SoC. If you’re using a CC/CV charger then no you can’t because you have to wait for the charge current at your chosen voltage to drop to a specified figure. 

 

39 minutes ago, Mike the Boilerman said:

All claims that one need not worry about it because "the BMS handles all that" are by definition, bobbins.

It depends. (Bet you love that answer). If the BMS is talking to the charge source like with a Victron system then the two combined “handle all that”. If you have a dumb charge source that can not communicate with the BMS then all the BMS is doing is acting as a failsafe. So yes, in those instances it’s bobbins. 

 

43 minutes ago, Mike the Boilerman said:

THE SoC can be determined by measuring terminal voltage when the batteries are at rest, because LiFePO4 batteries do not exhibit surface charge.

Yes. Or while discharging too with ‘reasonable’ loads, as they also suffer little voltage droop, having an internal resistance of close to 0. 

 

45 minutes ago, Mike the Boilerman said:

I guess the bit I am unsure about is whether the SoC can be determined during discharge. 

Yes it can, see above reply. 

 

45 minutes ago, Mike the Boilerman said:

if there are any devices which interpret the terminal voltage and display SoC directly, like a Smartgauge might if one was available for LiFePO4 battery banks.

I don’t know. For a ‘proper’ system such as the Victron then it does all that for you. For a DIY setup I don’t know. 

 

47 minutes ago, Mike the Boilerman said:

I am also unsure what a 'BMS' does. I think it just disconnects the battery before it gets overcharged or undercharged.

BMS can mean two things: Battery Monitoring System or Battery Management System. We’ll assume the latter. It disconnects on voltage. Either too high or too low will cause it to disconnect. 

 

49 minutes ago, Mike the Boilerman said:

how does it know the SoC during charging, given 1) ? AH counting? 

As above, if using a CC source then it knows by voltage. If using a CC/CV source then it has to Ah count to know the instantaneous SoC, or wait until the current at the charging voltage reaches the specified figure (eg 0.05C for the Relion offering). 

[cuthound]

I don't know much about lifepo4 batteries and chargers,  but surely if off load battery voltage can be accurately tranlated into SoC and there is no surface charge, then charging SoC can be measured by momentarily interrupting the charge?

[Dr Bob]

2 hours ago, Mike the Boilerman said:

 

I think we are agreed on the following points though:

 

1) The SoC cannot be monitored during charging.

2) All claims that one need not worry about it because "the BMS handles all that" are by definition, bobbins. 

3) THE SoC can be determined by measuring terminal voltage when the batteries are at rest, because LiFePO4 batteries do not exhibit surface charge.

 

I guess the bit I am unsure about is whether the SoC can be determined during discharge. And if not, why not. And if there are any devices which interpret the terminal voltage and display SoC directly, like a Smartgauge might if one was available for LiFePO4 battery banks.

 

I am also unsure what a 'BMS' does. I think it just disconnects the battery before it gets overcharged or undercharged. If this is true, how does it know the SoC during charging, given 1) ? AH counting? 

 

This thread has been very useful to me but there are still some big questions.

MTB's point 1. Happy with Tony's answer and we can see voltage rise during charge so guesstimate SoC. Worst case is stop and look at the rested voltage....no surface charge. Can you do this minutes after turning off the charge?

MTB's point 2. Ok that the BMS handles all if you buy an itegrated system (ie Victron) but not if I get some second hand batteries.....the plan!!!..... Here the BMS is going to provide the emergency cut off at high or low voltage and I assume balance the batteries but I then need 'smart' charge devices that respond to stopping charging at set voltages.

MTB's point 3. Agreed. Also the battery monitor (BMV) will monitor Ahr out so that gives a clue as well. Also will work during discharge.

 

The questions I now have are about voltages. Peter is using 13.8v as his upper limit to give 80% charge. Why go higher? VIctron are talking about 14.2-14.4V. I think the Relion info was similar. Peter then said Jono was using a lower voltage (I assume 13.8v) yet this is on the victron integrated system. Why are victron pushing it up to 100% charge? I would operate between 25% and 80% and be happy if the batteries were down at 30-40% overnight.

 

It all sounds complex but the simplicity is that overnight I typically use 100Ahrs. With lithiums I can put 70A back in in an hour of running the engine (I assume from a 90A alternator) and then let solar do the rest. If no solar then after a few days I may need to run the engine for 2 hours. A big alarm clock would work with a BMS providing the emergency shut down.

 

Another more technical question is that the american link, should battery balancing at the low end of the voltage scale. I note that in Jono's video, the victron is balancing at the top end. Need to understand the pros and cons a bit better.

 

 

[WotEver]

36 minutes ago, cuthound said:

I don't know much about lifepo4 batteries and chargers,  but surely if off load battery voltage can be accurately tranlated into SoC and there is no surface charge, then charging SoC can be measured by momentarily interrupting the charge?

Then you’re back to manual intervention. 

[Dr Bob]

37 minutes ago, cuthound said:

I don't know much about lifepo4 batteries and chargers,  but surely if off load battery voltage can be accurately tranlated into SoC and there is no surface charge, then charging SoC can be measured by momentarily interrupting the charge?

That's what I thought. Is this how the BMS tests them, ie tests each cell during charge?

[Detling]

1 hour ago, WotEver said:

 a CC/CV source

I am confused

 

CC = Constant current I understand this, constant current so voltage depends on resistance of the battery being charged.

 

CV I understand this constant voltage, so charge current depends on the resistance of the battery being charged.

 

but CC/CV ????? this is impossible at anything other than 0 amps.

[cuthound]

2 minutes ago, WotEver said:

Then you’re back to manual intervention. 

Not necessarily, the BMS could automatically interrupt charge for a millisecond every 10 seconds or so to calculate SoC.

[WotEver]

36 minutes ago, Detling said:

I am confused

 

CC = Constant current I understand this, constant current so voltage depends on resistance of the battery being charged.

 

CV I understand this constant voltage, so charge current depends on the resistance of the battery being charged.

 

but CC/CV ????? this is impossible at anything other than 0 amps.

CC/CV is as I explained previously - CC followed by CV. Think Bulk followed by Absorption. 

37 minutes ago, cuthound said:

Not necessarily, the BMS could automatically interrupt charge for a millisecond every 10 seconds or so to calculate SoC.

And thereby destroy your alternator?

[WotEver]

And... is a millisecond long enough for the battery voltage to stabilise? I have no idea. 

[MtB]

1 hour ago, Dr Bob said:

but not if I get some second hand batteries.....the plan!!!.....

 

Dr Bob, which type of second hand LiFePO4 batteries are you thinking of? 

 

1) Four whooping great bare single cells, and design your own charging and disconnection systems?

 

Or

2) A nominally 12v LiFoPO4 battery comprising several hundred tiny cells in a series/parallel grid inside (EG the Valence brand batteries) with what is claimed to be a BMS PCB inside?

Edited August 14, 2018 by Mike the Boilerman

[mrsmelly]

8 minutes ago, Mike the Boilerman said:

 

Dr Bob, which type of second hand LiFePO4 batteries are you thinking of? 

 

1) Four whooping great bare single cells, and design your own charging and disconnection systems?

 

Or

2) A nominally 12v LiFoPO4 battery comprising several hundred tiny cells in a series/parallel grid inside (EG the Valence brand batteries) with what is claimed to be a BMS PCB inside?

Jeeeeesus, thank God for flooded lead acid batteries!!!!

[Robbo]

1 minute ago, mrsmelly said:

Jeeeeesus, thank God for flooded lead acid batteries!!!!

Because they or so much easier to charge - give us a break.

[WotEver]

1 hour ago, Dr Bob said:

Another more technical question is that the american link, should battery balancing at the low end of the voltage scale. I note that in Jono's video, the victron is balancing at the top end. Need to understand the pros and cons a bit better.

If I’ve understood the case correctly, high balancing requires a charge to 100%. The internal BMS monitors each cell in series and as one reaches ‘full’ so it shunts the charge away from that cell. Once all the cells in series are ‘full’ the charge current is cut off completely. The cells are now balanced. Low balancing occurs on discharge, I think. So as each cell voltage drops, the circuitry uses FETs to relieve the stress on the lower voltage cells, allowing the higher ones to drop down into sync. 

 

I may may have misunderstood some of the above and I’ll be happy to be corrected if I have. 

[MtB]

1 minute ago, Robbo said:

Because they or so much easier to charge - give us a break.

 

Yes they are FAR easier to charge. Just start your engine and run it for ten hours. What could be easier?

 

With Lithiums we have to design our own upper and lower voltage cut-offs and figure out how to avoid charging to 100% or FIZZZZ, £3k worth of battery up in smoke. All whilst being equally careful never to allow to fall below 0C or the same thing happens 

[Robbo]

Just now, Mike the Boilerman said:

 

Yes they are FAR easier to charge. Just start your engine and run it for ten hours. What could be easier?

 

With Lithiums we have to design our own upper and lower voltage cut-offs and figure out how to avoid charging to 100% or FIZZZZ, £3k worth of battery up in smoke. All whilst being equally careful never to allow to fall below 0C or the same thing happens 

Well I can do that with LFP’s as well?   You don’t have to design your own - these things are available off the shelf just like LA’s!

 

With decent LA’s I need to charge at a high voltage which probably too high for the starter battery.   I need to fully charge regularly to get the best out of them as well an equalise every month at a even higher voltage that may potentially damaging my electronics.   Once LFPs are wired up correctly you just use them. 

[MtB]

3 minutes ago, Robbo said:

Well I can do that with LFP’s as well?   You don’t have to design your own - these things are available off the shelf just like LA’s!

 

Generic devices the user can configure? Do you have some links please?

 

I suspect you are meaning the proprietary integrated battery/BMS/charger systems marketed by Victron, Valence etc 

[Robbo]

2 minutes ago, Mike the Boilerman said:

 

Generic devices the user can configure? Do you have some links please?

 

I suspect you are meaning the proprietary integrated battery/BMS/charger systems marketed by Victron, Valence etc 

Yes like from the likes of Victron, if you want a turn key solution they make very good solutions especially if you already have Victron equipment.    I’m personally looking at the BMS12/200 and around 300ah of LFP’s as my combi is too old to be directly controlled via the other BMS they do and my combo is only 2000watt (the BMS can only handle upto 200amp).   The other BMS Victron do is more a remote controller telling things to be switched off where the 12/200 the batteries, load and charging devices are connected directly to the BMS.

[MtB]

7 minutes ago, Robbo said:

Yes like from the likes of Victron, if you want a turn key solution they make very good solutions especially if you already have Victron equipment.    I’m personally looking at the BMS12/200 and around 300ah of LFP’s as my combi is too old to be directly controlled via the other BMS they do and my combo is only 2000watt (the BMS can only handle upto 200amp).   The other BMS Victron do is more a remote controller telling things to be switched off where the 12/200 the batteries, load and charging devices are connected directly to the BMS.

 

A turnkey solution is the last thing I want. 

 

I want a system where I understand what it is happening every step of the way, not a black box 'we'll handle all that for you' type of thing. Starting with four bare cells which I will connect in series  

[Robbo]

3 minutes ago, Mike the Boilerman said:

 

A turnkey solution is the last thing I want. 

 

I want a system where I understand what it is happening every step of the way, not a black box 'we'll handle all that for you' type of thing. Starting with four bare cells which I will connect in series  

Victron’s Smart batteries connect via Bluetooth to an app on ya phone/pc and show you live data, including individual cell voltage.

[cuthound]

2 hours ago, WotEver said:

And thereby destroy your alternator?

 

Not if the BMS was programmed to switch in a dummy load at the same time.

 

2 hours ago, WotEver said:

And... is a millisecond long enough for the battery voltage to stabilise? I have no idea. 

 

Neither have I,  but if there is no surface charge the cell voltage should drop to off load voltage pretty quickly.

 

I'm just thinking aloud, as I said I have no practical experience of any Lion batteries beyond iPod,  mobile phones and power tools.

[Dr Bob]

3 hours ago, Mike the Boilerman said:

 

Dr Bob, which type of second hand LiFePO4 batteries are you thinking of? 

 

1) Four whooping great bare single cells, and design your own charging and disconnection systems?

 

Or

2) A nominally 12v LiFoPO4 battery comprising several hundred tiny cells in a series/parallel grid inside (EG the Valence brand batteries) with what is claimed to be a BMS PCB inside?

I want to see what's available. I like Peter's solution. My LA's are now 18 months old and still over 90% capacity so maybe 2 to 3 years time. Currently looking for a source of used batteries. I would think it would be more likely option 2 but using the BMS as the emergency shut off and a DIY solution to monitoring voltage/charging on/off. I do like the idea of only an hour a day fo running the engine.