Jump to content

Lithium battery abuse


TheBiscuits

Featured Posts

1 hour ago, nicknorman said:

Getting away from the annoyance of the lonnnnnggggg  sllloooowwwwww charge of the latter stages of LA charging is a blessing!

 

That was my precisely the thought that started this thread.  Have a lithium bank that's sized to be the final 10% or so of the existing LA bank and use that to finish off the charge on the LA's.  

 

Everyone goes on about overheating alternators with big banks, but the little 20Ah lithiums are limited to 20A charge current. The little batteries are ninety quid each - just abuse them.   

 

If that offends you, the Renogy B2B 20A is a hundred quid, add a changeover switch to the battery cabling and it could be used in both directions in a "compromise" battery chemistry setting.

Link to comment
Share on other sites

23 minutes ago, Dr Bob said:

 

The comments that some uniformed people make on here is mindblowing. Where do you get that information from? At best it is misleading and at worst DANGEROUS! If you had said 'mostly safe and happy under 90%' then I would agree.

 

See the following link to a thread posted 3 years ago. At that time there was certianly not enough data to support your claim. Our company has been working on the thermal runaway issues since then (and including LiFePO4 cells) helping develop a pre theremal runaway warning technique. I consider myself a technical expert on this type of battery chemistry, even if I cant design the electical system to control alternator charge.

 

 

I would recommend MarkH2159 read that thread and the associated links.

 

Anyone considering installing a home brew system of LiFePO4s needs to do a safety assessment on what they are installing and the starting position for that is to as much as possible keep away from 100% charging. I use 90% as my upper limit over which the auto disconnect works and my 'operating proceedure' says only go to 100% when I am personally in manual control of the voltage and current. Keeping LiFePO4s at 90% or below is very very unlikely to present an ignition hazard on well trusted cells (it is not going to happen on a canal boat). The risk is much higher at 99%.

There is very little point in taking these to 99%. The whole idea is that you dont need to fully charge. Why risk safety when thermal runaways are a risk?

Bob valence recommends that their batteries are taken to 100% and float is at that voltage. Me I don't because I want maximum life for them as I intend to live forever.

  • Greenie 2
Link to comment
Share on other sites

3 minutes ago, peterboat said:

Bob valence recommends that their batteries are taken to 100% and float is at that voltage. Me I don't because I want maximum life for them as I intend to live forever.

 

The Valence batteries self balance Peter, and Valence also say not use them without the Master BMS that they won't sell you.

 

It's completely different when using bare 3.2V cells - you really can do bad things to them.

  • Greenie 1
Link to comment
Share on other sites

38 minutes ago, peterboat said:

Bob valence recommends that their batteries are taken to 100% and float is at that voltage. Me I don't because I want maximum life for them as I intend to live forever.

 

32 minutes ago, TheBiscuits said:

 

The Valence batteries self balance Peter, and Valence also say not use them without the Master BMS that they won't sell you.

 

It's completely different when using bare 3.2V cells - you really can do bad things to them.

 

If you have a very good BMS system then you could go up to 100% but the issue is that once you get there, you are in the region where things can quickly go wrong. Manufacturing defects, denrites, contamination means that batteries are not infallable. Having a home brew system where you regularly get up to that level needs a very well thought out protection scheme. TheBiscuits is spot on when he talks about bare 3.2V cells. It's even worse when talking about cheap Chinese cells of unknown origin. In an out of balance homebrew system, there is a safety risk at 99% charge or even 95% if it is well out of balance.

.....and you say it yourself, why go to 100% if you want them work for a long time.

I am surprised that in the 3 1/2 years since putting mine in, there has been no degradation in the performance - and even better, the balance has not changed at all. I have not had to re-balance in the last 2 1/2 years - probably because I dont run them up to 100% normally (only twice a year to stretch them!).

 

The problem is that some manufacturers say - yes its fine to go to 100% - and people like MarkH2159 take that to mean it is fine for everyone else and try to convince everyone else of that fact.

 

Home brewers need to make sure they are doing things safely.

 

 

Link to comment
Share on other sites

5 minutes ago, Richard10002 said:

Not in my experience. It is so slow as to not be effective.

 

On Peters, where he is charging 24/7 with his huge solar array, its possible they may balance themselves.

 

Are they not active top balance? 

 

If so, they will only start to balance when they are approaching fully charged and it might be by trivial currents - probably milliamps between cells.  If you keep them below full the balancing circuit can't do anything unless you can force it to try via the software.  @peterboat will know if you can do this.

Link to comment
Share on other sites

18 minutes ago, TheBiscuits said:

 

Are they not active top balance? 

 

If so, they will only start to balance when they are approaching fully charged and it might be by trivial currents - probably milliamps between cells.  If you keep them below full the balancing circuit can't do anything unless you can force it to try via the software.  @peterboat will know if you can do this.

Yes, they start to balance at about 14v or so. The fact that the currents are trivial makes the "self balancing" ineffective.

 

I suppose if I could charge at 14v for days on end, it might have an effect.

Link to comment
Share on other sites

Just now, Richard10002 said:

Yes, they start to balance at about 14v or so. The fact that the currents are trivial makes the "self balancing" ineffective.

 

I suppose if I could charge at 14v for days on end, it might have an effect.

 

See if you can find out which way they work.  They might stop charging the higher cell(s) and continue charging the lower cell(s) to balance, in which case you need to take the voltage a bit further up, but obviously that has risks of damage without the master BMS box controlling it.

 

Do you have the PC lead and software?

Link to comment
Share on other sites

I do occasional take them up to more than 13.9 volts, I have set the equaliser button to do it, when I last had the puter on the batteries the cells were well balanced before it went active. When I first bought the batteries I placed them in parallel and charged them to 14.6 volts and held them there, I did it in a series of steps upping the voltage slightly with each stage. Whilst this takes time it pays dividends over the coming years. Last year I helped someone do a big install, the supplier assured Steve they were balanced, it soon became obvious they weren't when the REC BMS couldn't cope! Anyway we dismantled put them in parallel and charged to the max, dismantled back into 48 volts blocks attached BMS perfik. His system works very hard and stays balanced without interference from the BMS, this is because of the 20 to 90 rules he is using. 80% would be better but he needs that extra capacity as its for running freezers overnight. 

 

Link to comment
Share on other sites

1 hour ago, TheBiscuits said:

 

Are they not active top balance? 

 

If so, they will only start to balance when they are approaching fully charged and it might be by trivial currents - probably milliamps between cells.  If you keep them below full the balancing circuit can't do anything unless you can force it to try via the software.  @peterboat will know if you can do this.

It is milliamps but quality cells help keep things right. I couldn't run my boat on anything but LifePo4s as the speed of charging on solar is needed. For my Genny I have used the long cable system rather than any specialist equipment it works as my alternator cuts out quite low on my 24 volts system. The drive batteries are charged by a 72 volts lithium battery charger, its only a backup so isn't run much 

Link to comment
Share on other sites

6 minutes ago, TheBiscuits said:

 

The question was can you force them into active balancing mode below 14V?

They do it when you connect the computer at any voltage, now the battery balancing board is always live and whenever I connect I always find well balanced batteries so I suspect something must be happening but can't prove it as the only way to see the batteries is via the puter 

  • Greenie 1
Link to comment
Share on other sites

20 minutes ago, peterboat said:

They do it when you connect the computer at any voltage, now the battery balancing board is always live and whenever I connect I always find well balanced batteries so I suspect something must be happening but can't prove it as the only way to see the batteries is via the puter 

 

Do you have an extra connection to keep the balancing board live or do you just leave either a computer or just the computer lead permanently connected which might be doing the same thing?

Link to comment
Share on other sites

4 minutes ago, TheBiscuits said:

 

Do you have an extra connection to keep the balancing board live or do you just leave either a computer or just the computer lead permanently connected which might be doing the same thing?

They are always live they run off the battery, they  have either a flashing green light all is well, a flashing red they are below a certain voltage, solid red light they are nearly flat! Its the Battery balance board that flattens them  but it takes a long time.

Link to comment
Share on other sites

Of course the point of top balancing is to ensure that as the battery approaches 100% which is 3.65v per cell, no one cell is significantly higher than the others. However I’ve found that you can’t do this eg by looking at the cells when they are around 3.5v, because the highest in that region is not the highest cell at 3.65v. Not sure why, but I think MP found the same thing so it’s not just me. And of course this is why BMS designers don’t initiate balancing until the voltage gets to 14.4 (3.6v) or so.

 

I’ve more or less followed MP’s strategy in that when the first cell hits 3.6v, a snapshot of cell voltages is taken and a calculation done on the difference between the highest and the other three, then the balancing resistors are switched in on each of the 3 lower cells for a time proportionate to the voltage differences. It matters not whether the charging is continued or terminated at that point - in other words, the cells don’t have to be held high to balance, they just have to momentarily reach 3.6v.

 

But also, it seems that once top balancing is achieved, it persists. I suspect that if any significant subsequent balancing is needed it is because a cell is deteriorating, not because routine balancing is required.

Edited by nicknorman
  • Greenie 1
Link to comment
Share on other sites

37 minutes ago, nicknorman said:

But also, it seems that once top balancing is achieved, it persists. I suspect that if any significant subsequent balancing is needed it is because a cell is deteriorating, not because routine balancing is required.

My experience is similar, once top balancing is achieved, it stays that way.

For my first 6 months I did worry about balancing and after 9 months finally got them somewhere near balanced - a spread of 50mv over the cells, ie the bottom at 3.45V and the top at 3.5V typically at the end of the charge. It has not changed from that 2 1/2 years on. I dont think about balancing now. It's a non issue.

I am not sure if that would be the case of frequently taking the charge to 100%. A lot of the 'information' on the internet suggested high charging could lead to balance instability hence the general accepted norm of charging to 80-90%.

 

Given a number of us have now had Li's on board for 3-4 years and more, has everyone else given up looking at balancing as a non issue?.......but then we were all only charging to 80-90% so it wouldnt be an issue then anyway! (Tom, Simon?  .....ignoring Peter as his Valences self balance and cook the dinner as well).

  • Haha 1
Link to comment
Share on other sites

42 minutes ago, Dr Bob said:

My experience is similar, once top balancing is achieved, it stays that way.

For my first 6 months I did worry about balancing and after 9 months finally got them somewhere near balanced - a spread of 50mv over the cells, ie the bottom at 3.45V and the top at 3.5V typically at the end of the charge. It has not changed from that 2 1/2 years on. I dont think about balancing now. It's a non issue.

I am not sure if that would be the case of frequently taking the charge to 100%. A lot of the 'information' on the internet suggested high charging could lead to balance instability hence the general accepted norm of charging to 80-90%.

 

Given a number of us have now had Li's on board for 3-4 years and more, has everyone else given up looking at balancing as a non issue?.......but then we were all only charging to 80-90% so it wouldnt be an issue then anyway! (Tom, Simon?  .....ignoring Peter as his Valences self balance and cook the dinner as well).

I've left the balancing to BMS for the past year, at least and they're all quite happy. Sometime before Christmas 2020 I added a feature to the BMS to keep a running total of charge and discharge. The two values are now around 25,000Ah and pretty similar, as you'd hope. Given that someone mentioned manufacturers hopes for this metric in the MAh range upthread, I'm hoping my cells have plenty of life left. They certainly haven'y lost any noticable capacity after three years.

 

MP.

  • Greenie 1
Link to comment
Share on other sites

3 minutes ago, MoominPapa said:

I've left the balancing to BMS for the past year, at least and they're all quite happy. Sometime before Christmas 2020 I added a feature to the BMS to keep a running total of charge and discharge. The two values are now around 25,000Ah and pretty similar, as you'd hope. Given that someone mentioned manufacturers hopes for this metric in the MAh range upthread, I'm hoping my cells have plenty of life left. They certainly haven'y lost any noticable capacity after three years.

 

MP.

All good to hear MP.

As above, I have just forgotten about balancing. My graph of rested voltage vs Ahrs out is exactly the same now as in summer 2019 although I do need to reset the Ahrs out frequently as there are a few drains not wired into the shunt. To be honest, I just watch the voltage and dont bother with the Ahrs, and never the SOC.

 

Link to comment
Share on other sites

2 hours ago, Dr Bob said:

All good to hear MP.

As above, I have just forgotten about balancing. My graph of rested voltage vs Ahrs out is exactly the same now as in summer 2019 although I do need to reset the Ahrs out frequently as there are a few drains not wired into the shunt. To be honest, I just watch the voltage and dont bother with the Ahrs, and never the SOC.

 

 

I trust the BMS's idea of SOC implicitly now. When we're not using the engine I can know with confidence how many days are left before we need to move or run the genny.

 

MP.

Link to comment
Share on other sites

4 hours ago, Dr Bob said:

My experience is similar, once top balancing is achieved, it stays that way.

For my first 6 months I did worry about balancing and after 9 months finally got them somewhere near balanced - a spread of 50mv over the cells, ie the bottom at 3.45V and the top at 3.5V typically at the end of the charge. It has not changed from that 2 1/2 years on. I dont think about balancing now. It's a non issue.

I am not sure if that would be the case of frequently taking the charge to 100%. A lot of the 'information' on the internet suggested high charging could lead to balance instability hence the general accepted norm of charging to 80-90%.

 

Given a number of us have now had Li's on board for 3-4 years and more, has everyone else given up looking at balancing as a non issue?.......but then we were all only charging to 80-90% so it wouldnt be an issue then anyway! (Tom, Simon?  .....ignoring Peter as his Valences self balance and cook the dinner as well).

Remember Bob, I have said from the beginning that balancing is a non event as long as top balance was done correctly, the mantra of 20 to 80% is key to happy LifePo4 batteries. Mine and Richards batteries have a balance board which is powered, so because on checking with the puter they are always balanced, the assumption is the balance board works. My drive batteries are the same green flashing lights means that they are ok, and I regularly drag 400 amps out of them mind there are 30 of them. 

  • Greenie 1
Link to comment
Share on other sites

3 hours ago, Dr Bob said:

All good to hear MP.

As above, I have just forgotten about balancing. My graph of rested voltage vs Ahrs out is exactly the same now as in summer 2019 although I do need to reset the Ahrs out frequently as there are a few drains not wired into the shunt. To be honest, I just watch the voltage and dont bother with the Ahrs, and never the SOC.

 

 

Can I ask, how do you know when the voltage reading is rested? 

My voltage never seems to rest- there is always a load of some sort going on, whether its a couple of amps for the inverter, fridge, or whatever- there is always a load. Mostly small, but occasionally up to 80 amps for the kettle or immersion heater. 

To add to the challenge, the solar is usually putting something in during daylight. In the summer this might be 70 amps or so in the peak days, but even this afternoon I was getting 10 amps, but intermittently. 

So none of the charging or discharging is at steady levels, and it never totally stops. It seems like one minute I have a net load of 5 or 10 amps, and a few minutes later the sun comes out in between clouds, and I have a net charge of 10 amps.

So I never really feel like I know what the rested voltage is. 

 

 

Edited by Tony1
Link to comment
Share on other sites

52 minutes ago, Tony1 said:

 

Can I ask, how do you know when the voltage reading is rested? 

My voltage never seems to rest- there is always a load of some sort going on, whether its a couple of amps for the inverter, fridge, or whatever- there is always a load. Mostly small, but occasionally up to 80 amps for the kettle or immersion heater. 

To add to the challenge, the solar is usually putting something in during daylight. In the summer this might be 70 amps or so in the peak days, but even this afternoon I was getting 10 amps, but intermittently. 

So none of the charging or discharging is at steady levels, and it never totally stops. It seems like one minute I have a net load of 5 or 10 amps, and a few minutes later the sun comes out in between clouds, and I have a net charge of 10 amps.

So I never really feel like I know what the rested voltage is. 

 

 

 

One of the advantages of having the hybrid system of 400Ahr Li's in parallel with 400Ahrs of LAs is that you can isolate the Li's for 30 mins to get a rested voltage. I used to do this regularly to be able to draw the graph of Voltage vs Ahrs out (after resetting the Ahr counter on the last charge).

Once you understand this relationship, you can then look at the rested voltage vs actual voltage under load. My system seems to register a drop of 0.010V per 1A of current going out, so if I am running off the Li bank and the voltage is 13.05V and there is 5A load, then I know the rested voltage will be circa 13.10V. This relationship seems to work over the 5A to 20A out range and doesnt seem to be affected by temperature. I have not looked to see if it works at higher loads. This of course doesnt work if the solar is active. We have run our new boat without solar for the first year but now have just fitted 500W so now I can only do the 'rested voltage' estimation at night!

There is another way of doing it without the hassle of doing any work. If you have a BMV 712 (or similar), then hook it up to one of the victron devices that save all your data to the victron site - ie a venus GX or the new one, or if you are skint, download the victron firmware onto a Rasp Pie and be able to store all your data for less than £90 (the cost of the Pie + cables). Then at the end of a months cruising, download the log that has the actual Volts, the Ahrs used and the current draw/charge as an excel file. Then order the file first on Ahr used, and secondly on Amps. Reject any line that has an Amperage > or < than 0. ie you only have data points where no current was flowing - ie when there was no draw (likely never) or when the draw was equal to the solar going in. Ok, the voltage at A=0 will not be rested as it is likely it was higher just before that reading but you will get a scatter plot of these values if you plot V against Ahrs. Look at the edge of the scatter plot where the voltage for any given Ahr  is lower and that is likely to be the most rested point. You will be surprised in a months worth of data how many readings you will get with the Amps very near 0. Checked the results from the first method vs second method and they match exactly. Siiiimples.

I use the data output from the victron service (which is totally free) all the time.  The Rasp Pie is brilliant.

 

 

  • Greenie 1
Link to comment
Share on other sites

7 minutes ago, Dr Bob said:

 

One of the advantages of having the hybrid system of 400Ahr Li's in parallel with 400Ahrs of LAs is that you can isolate the Li's for 30 mins to get a rested voltage. I used to do this regularly to be able to draw the graph of Voltage vs Ahrs out (after resetting the Ahr counter on the last charge).

Once you understand this relationship, you can then look at the rested voltage vs actual voltage under load. My system seems to register a drop of 0.010V per 1A of current going out, so if I am running off the Li bank and the voltage is 13.05V and there is 5A load, then I know the rested voltage will be circa 13.10V. This relationship seems to work over the 5A to 20A out range and doesnt seem to be affected by temperature. I have not looked to see if it works at higher loads. This of course doesnt work if the solar is active. We have run our new boat without solar for the first year but now have just fitted 500W so now I can only do the 'rested voltage' estimation at night!

There is another way of doing it without the hassle of doing any work. If you have a BMV 712 (or similar), then hook it up to one of the victron devices that save all your data to the victron site - ie a venus GX or the new one, or if you are skint, download the victron firmware onto a Rasp Pie and be able to store all your data for less than £90 (the cost of the Pie + cables). Then at the end of a months cruising, download the log that has the actual Volts, the Ahrs used and the current draw/charge as an excel file. Then order the file first on Ahr used, and secondly on Amps. Reject any line that has an Amperage > or < than 0. ie you only have data points where no current was flowing - ie when there was no draw (likely never) or when the draw was equal to the solar going in. Ok, the voltage at A=0 will not be rested as it is likely it was higher just before that reading but you will get a scatter plot of these values if you plot V against Ahrs. Look at the edge of the scatter plot where the voltage for any given Ahr  is lower and that is likely to be the most rested point. You will be surprised in a months worth of data how many readings you will get with the Amps very near 0. Checked the results from the first method vs second method and they match exactly. Siiiimples.

I use the data output from the victron service (which is totally free) all the time.  The Rasp Pie is brilliant.

 

 

We can do all the above Bob but why? A quick glance I see 13.4 volts with. 5 of an amp going out, batteries are fully charged simples. Remember my float is 13.4 volts

Link to comment
Share on other sites

38 minutes ago, Dr Bob said:

 

One of the advantages of having the hybrid system of 400Ahr Li's in parallel with 400Ahrs of LAs is that you can isolate the Li's for 30 mins to get a rested voltage. I used to do this regularly to be able to draw the graph of Voltage vs Ahrs out (after resetting the Ahr counter on the last charge).

Once you understand this relationship, you can then look at the rested voltage vs actual voltage under load. My system seems to register a drop of 0.010V per 1A of current going out, so if I am running off the Li bank and the voltage is 13.05V and there is 5A load, then I know the rested voltage will be circa 13.10V. This relationship seems to work over the 5A to 20A out range and doesnt seem to be affected by temperature. I have not looked to see if it works at higher loads. This of course doesnt work if the solar is active. We have run our new boat without solar for the first year but now have just fitted 500W so now I can only do the 'rested voltage' estimation at night!

There is another way of doing it without the hassle of doing any work. If you have a BMV 712 (or similar), then hook it up to one of the victron devices that save all your data to the victron site - ie a venus GX or the new one, or if you are skint, download the victron firmware onto a Rasp Pie and be able to store all your data for less than £90 (the cost of the Pie + cables). Then at the end of a months cruising, download the log that has the actual Volts, the Ahrs used and the current draw/charge as an excel file. Then order the file first on Ahr used, and secondly on Amps. Reject any line that has an Amperage > or < than 0. ie you only have data points where no current was flowing - ie when there was no draw (likely never) or when the draw was equal to the solar going in. Ok, the voltage at A=0 will not be rested as it is likely it was higher just before that reading but you will get a scatter plot of these values if you plot V against Ahrs. Look at the edge of the scatter plot where the voltage for any given Ahr  is lower and that is likely to be the most rested point. You will be surprised in a months worth of data how many readings you will get with the Amps very near 0. Checked the results from the first method vs second method and they match exactly. Siiiimples.

I use the data output from the victron service (which is totally free) all the time.  The Rasp Pie is brilliant.

 

 

 

Thanks, I'll see if I can download the data. I did use a rasp pi a few years ago for dodgy tv, but I've long forgotten how to use one now. 

But I'd be very grateful for your thoughts on the length of time needed before the batteries lose their 'surface charge'?

I know it cant be the same as lead acid surface charge, but I have noticed that for some hours following a high-current engine charge at 90 amps (or maybe 130 amps if there is some solar), the voltage seems to stay artificially high, in the sense that it doesnt follow the voltage/SoC graph I tend to use. 

 

So this week, I've charged the batteries from maybe 35% up to say 55-60% with a quick 30 min charge, and that boost will give me enough charge for the next 24 hours (when combined with the 800-1000Wh of daily solar we're getting at the moment).   

 

But the problem is that even though the batteries are only at say 60%, their voltage (for a few hours) seems to reflect a SoC that is higher.

I havent recorded these values so this is a bit vague, but as an example:

I have an old lithium graph of resting voltage vs SoC, and it indicates that if my batteries are at 13.20v, they should be about 70% SoC. 

But after a charge, the voltage seems to stay at a level that appears to be too high for the SoC, and this discrepancy can be present for up to 10 hours. It could just be a characteristic of these batteries, of course. 

 

But I know from the BMV that my batteries are at say 60% (and I trust that figure, because I synchronised to 100% SoC a few days before), so at 60% my graph suggests that the battery voltage should be about 13.16v.

But what I'm seeing is the voltage is staying at 13.20v and sometimes above that, for ages.

It seems to be some kind of lithium equivalent of the surface charge that I used to see with the lead acids, but its more persistent, and it makes reading voltage a bit tricky. 

 

But on the load/voltage issue, I have learned to add a bit of compensation for the voltage being pulled down by a load on the batteries (even a small one), so I'm able to reasonably predict what the resting voltage would be at any given time (unless the load or charge is more than 10 amps), and my estimations of the voltage drop caused by a load on the batteries are very similar to yours. 

 

Edited by Tony1
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.