LiFeP04 Batteries for Solar & BMS

[Living Large]

G'day, noticed some typing errors in my last post, our MH has 1000w of solar panels and 1000w psw inverter. If your charge controller is programmable and able to leave out the absorbion, float and temp compensation sections, then it will be fine if the upper and lower parameters you want can be set.

My BMS sends a 12v signal to a 150amp relay when the cell pack reaches 14.4v and that disconnects the charger controller from the pack. The same when it reaches 12v, the bms sends a signal to another relay which switches of power. Haven't got to that stage yet of disconnecting power, but have reached the upper limits a few times when we have been away and the cell pack is only being used for fridge, freezer and security systems. When at home, we connect the MH to the house system which gives us more power to use.

A lifepo4 solar charger should be one which can be programmed to suit lifepo4 battery pack requirements without temp, float and absorb sections. To me the best solar charger set up would be one which provides 150amp input and charges each cell line individually to a set parameter, then switches of each cell line when it reaches that number. Do the same with lower voltage paramaters and you have a charger controller that is also a bms. Not much different to current charger controllers which are all orientated to lead acid. This way there would be no need for cell balancers or to worry about anything, but enjoying your power supply.

Sadly there doesn't seem to be any 3.2v cell chargers that have high amperage inputs, no good trying to charge 700ah with 4x4.5 amp chargers coming from a 60-120 amp supply.

Thanks for the explanation. I believe the equipment I am considering is more compatible than yours, but that is a guess at this point. I won't be surprised if one or more issues arises when I look closer.

[tasman]

Good to hear your LFP bank is stable - that is what seems to be the prevailing wisdom - if balanced initially.

What do you mean by "yet to find a charge controller that does the job"? Can you not get the thresholds you want, the profile (CC, CV)? I haven't looked specifically yet at what the one I am considering offers, but was hoping I could program something acceptable. It is two of the Schneider 60A units.

G'day, noticed some typing errors in my last post, our MH has 1000w of solar panels and 1000w psw inverter. If your charge controller is programmable and able to leave out the absorbion, float and temp compensation sections, then it will be fine if the upper and lower parameters you want can be set.

My BMS sends a 12v signal to a 150amp relay when the cell pack reaches 14.4v and that disconnects the charger controller from the pack. The same when it reaches 12v, the bms sends a signal to another relay which switches of power. Haven't got to that stage yet of disconnecting power, but have reached the upper limits a few times when we have been away and the cell pack is only being used for fridge, freezer and security systems. When at home, we connect the MH to the house system which gives us more power to use.

A lifepo4 solar charger should be one which can be programmed to suit lifepo4 battery pack requirements without temp, float and absorb sections. To me the best solar charger set up would be one which provides 150amp input and charges each cell line individually to a set parameter, then switches of each cell line when it reaches that number. Do the same with lower voltage paramaters and you have a charger controller that is also a bms. Not much different to current charger controllers which are all orientated to lead acid. This way there would be no need for cell balancers or to worry about anything, but enjoying your power supply.

Sadly there doesn't seem to be any 3.2v cell chargers that have high amperage inputs, no good trying to charge 700ah with 4x4.5 amp chargers coming from a 60-120 amp supply.

[Living Large]

I have 780ah lifepo4 fed by 2000w of solar running a house (plus 250w wind generator), in the 8 months it has been operating never got to LVD, so top balance and use a dedicated BMS with active cells balancers and relay driven charge and LVD cut off.

Yet to find a solar charge controller that does the job for lifepo4. We have 24v cells and 12v storage, just about everything runs through a 3000-6000w PSW inverter and seems to run flawlessly. My LVD is 3v and top, 3.6v.

Also have 480ah lifepo4 and 100w solar running a 100w inverter and 204v and 12v appliances in a motorhome. I have yet to reach LVD in either our motorhome or house and in the MH, use the lifepo4 for hot water and A/C sometimes. This lifepo4 pack has been in use for more than 18 months and functions flawlessly. I have a small 2.3v charger for bringing up any cells that don't make the grade at the top, but only used that once when I installed them and since have changed the connection for input and output, so now every thing works fine.

Good to hear your LFP bank is stable - that is what seems to be the prevailing wisdom - if balanced initially.

What do you mean by "yet to find a charge controller that does the job"? Can you not get the thresholds you want, the profile (CC, CV)? I haven't looked specifically yet at what the one I am considering offers, but was hoping I could program something acceptable. It is two of the Schneider 60A units.

[tasman]

I have 780ah lifepo4 fed by 2000w of solar running a house (plus 250w wind generator), in the 8 months it has been operating never got to LVD, so top balance and use a dedicated BMS with active cells balancers and relay driven charge and LVD cut off.

Yet to find a solar charge controller that does the job for lifepo4. We have 24v cells and 12v storage, just about everything runs through a 3000-6000w PSW inverter and seems to run flawlessly. My LVD is 3v and top, 3.6v.

Also have 480ah lifepo4 and 100w solar running a 100w inverter and 204v and 12v appliances in a motorhome. I have yet to reach LVD in either our motorhome or house and in the MH, use the lifepo4 for hot water and A/C sometimes. This lifepo4 pack has been in use for more than 18 months and functions flawlessly. I have a small 2.3v charger for bringing up any cells that don't make the grade at the top, but only used that once when I installed them and since have changed the connection for input and output, so now every thing works fine.

[Living Large]

Correct the BMS is there only for the batteries. Having said that the BMS can send signals to control whatever you want. Example the Orion Jr, if used with a EV can tell the motor controller to go into Limp Home mode meaning reduce motor current. There are several outputs you can program to do whatever you want. What I cannot answer because I have no experience with them is the Xantrex Inverter. If it has control inputs I would guess they are logic level and should work. Sounds like a Home Work assignment for you.

Yeah, but the XW does not appear to be designed to work with any inputs other than the single REMOTE POWER OFF. Now that I understand BMSs a lot better, and returning to the conversation we were having 1 to 1-1/2 months ago, I better appreciate the thinking that solar equipment is not ready for LFP. But this may be all I need.

If the initial balancing is done, the batteries are monitored regularly to assess health (which the Orion does), and the BMS will prevent a meltdown by enabling me to create a POWER OFF signal, that appears to cover all the bases, unless I missed something.

It is described two ways in the Schneider documentation. Connect a switch between this pin and ground, normally open. Also, as a logic level input. Either way, this should do the trick.

EDIT I spoke with a Schneider rep via email today. He said there currently is no ability for BMS input (which we know), and that a third party disconnect would be needed. Schneider feels the built in functionality at the pack level should cover most situations. The rep did say that they are considering adding such additional capabilities, and that a firmware update is possible. I suppose that alludes to the XW possibly issuing a command for the MPPT to shut down if it receives a Remote Power Off, because currently I don't believe there are any inputs on the MPPT other than battery temp. What I could do is create a circuit that drives that battery temperature input, and shorts it to create what the MPPT sees as a fault condition, according to the manual. Might be able to just parallel a relay across the temperature sensor, driven by the BMS circuit.

[Sunking]

I need my memory refreshed Dereck. With the Xantrex equipment and programmable low and high voltages, the purpose of this BMS would only be to detect faults where I want to shut down?

Correct the BMS is there only for the batteries. Having said that the BMS can send signals to control whatever you want. Example the Orion Jr, if used with a EV can tell the motor controller to go into Limp Home mode meaning reduce motor current. There are several outputs you can program to do whatever you want. What I cannot answer because I have no experience with them is the Xantrex Inverter. If it has control inputs I would guess they are logic level and should work. Sounds like a Home Work assignment for you.

[Living Large]

On the discharge side you sit up the LVD to trigger when either statements are true: 1 Voltage = 2.6 to 2.9 vpc of less for more than 15 seconds. The time delay is used to prevent false trips from large starting loads causing excessive voltage sag. The second condition is a fail safe of Voltage = something like 2.0 to 2.5 trip immediately.

Now here is a legitimate argument one person could make and I touched on it earlier. If you are running everything off a standard battery Inverter they have built-in LVD and they are set for Pb batteries of 10.5, 21, and 42 volts. For example a 48 volt inverter the BMS if you set the time delayed trip point for say 2.6 volts x 16 = 41.6 volts, the inverter would trip before the BMS.

A little off topic, but a practical question. The Schneider XW (48V) I am considering has a hard-wired 32V low battery threshhold, but also a programmable low battery cutoff voltage and a timer to account for sags. But this is on the overall battery pack. There is an auxiliary input that I assume is logic level, which is REMOTE POWER OFF, which will shut the XW down completely.

I need my memory refreshed Dereck. With the Xantrex equipment and programmable low and high voltages, the purpose of this BMS would only be to detect faults where I want to shut down? So I could "OR" a signal that indicates an individual cell has decreased below or exceeded a level outside the "normal" range that is taken care of by the overall pack voltage detection, and drive this REMOTE POWER OFF pin? I'd have to see if the XW can tell the MPPT via the Xanbus, or also via an input of the MPPT itself, to shut charging from the array off.

[northerner]

Yes, but maybe not for the reasons you are thinking. Remember that dirty word Internal Resistance? Lithium batteries have very low resistance, thus very little voltage sag on discharge, and not much rise in voltage on the charge side.

On the discharge side you sit up the LVD to trigger when either statements are true: 1 Voltage = 2.6 to 2.9 vpc of less for more than 15 seconds. The time delay is used to prevent false trips from large starting loads causing excessive voltage sag. The second condition is a fail safe of Voltage = something like 2.0 to 2.5 trip immediately.

Now here is a legitimate argument one person could make and I touched on it earlier. If you are running everything off a standard battery Inverter they have built-in LVD and they are set for Pb batteries of 10.5, 21, and 42 volts. For example a 48 volt inverter the BMS if you set the time delayed trip point for say 2.6 volts x 16 = 41.6 volts, the inverter would trip before the BMS.

On the charge side is where things get dangerous because to get to 100% SOC requires CC/CV method. You first apply Constant Current until the pack voltage = 3.65 vpc x Cell count. So for 16S would be 58.4 volts. Then hold 58.4 volts and limit current to the 1 amp or whatever value your Shunt Balance Board is rated for until current stops. When current stops all cells are at 100% aka TOP BALANCE.



You are on the right track, but sounds like you are over looking something. Regardless if you top balance or bottom balance you are still limited by the weakest cell. The danger with top balance is the stronger cells driving weaker cells into polarity reversal, thus destroying them. If you bottom balance you eliminate that threat. Second point and you hit on it is you will get much longer cycle life by only charging to 90% or less. One of the challenges of switching to LFP is getting out of the Pb box. With Pb batteries you want to keep them at 100%, but not Lithium as that does stress them and shorten cycle life. For long term storage you store them at 60% SOC and lithium are best operated and happy between 10 to 90% SOC as it does not weaken them in any way. That is very easy to to with a BMS because the charge/discharge is very flat in that range.

A long as one stays well enough away from the top and bottom, I take it factoring in current draw to the voltage state of the cell will not matter so much anyway. Perhaps more of a concern when charging and getting close to 90%.

I think one also has to take an operational viewpoint to determine whether to top or bottom balance. If one keeps their pack state on average at over 50 % state of charge for most of the time and also wants to get to 90% charge state on many days, particularly when the sun is shining (and using solar of course), then I would think that top balancing would make more sense. If you're not balanced at the top, then there is the risk that some cells are getting charged to over 90% on a regular basis, and will suffer capacity loss over time, which will compound. Going to 90% state of charge gives you only a 10% margin from the top, and the problem starts around that level. The top balancing I'm referring to would only be done very occasionally, thus having a minimal effect on the battery.

I'm aware what you are saying about going to low on the bottom, and reason you would stay well away from the cliff. But if you go down to say 20% state of charge, then you have a good safety margin, with no long term effects on capacity degradation, which you could have going close to the top. Do you see now?

[Sunking]

Would it be desirable to factor in current flow (ie through battery pack) as well as voltage to decide on the trigger point for the LVD and HVD? That way you could have trigger points based on the actual current flows, rather than doing approximations with timing, or does it work fine anyway?

Yes, but maybe not for the reasons you are thinking. Remember that dirty word Internal Resistance? Lithium batteries have very low resistance, thus very little voltage sag on discharge, and not much rise in voltage on the charge side.

On the discharge side you sit up the LVD to trigger when either statements are true: 1 Voltage = 2.6 to 2.9 vpc of less for more than 15 seconds. The time delay is used to prevent false trips from large starting loads causing excessive voltage sag. The second condition is a fail safe of Voltage = something like 2.0 to 2.5 trip immediately.

Now here is a legitimate argument one person could make and I touched on it earlier. If you are running everything off a standard battery Inverter they have built-in LVD and they are set for Pb batteries of 10.5, 21, and 42 volts. For example a 48 volt inverter the BMS if you set the time delayed trip point for say 2.6 volts x 16 = 41.6 volts, the inverter would trip before the BMS.

On the charge side is where things get dangerous because to get to 100% SOC requires CC/CV method. You first apply Constant Current until the pack voltage = 3.65 vpc x Cell count. So for 16S would be 58.4 volts. Then hold 58.4 volts and limit current to the 1 amp or whatever value your Shunt Balance Board is rated for until current stops. When current stops all cells are at 100% aka TOP BALANCE.

My understanding is that some users prefer to top balance, if they find that they more consistently use the top end of the pack, than the bottom. And they may only occasionally go towards a lower charge state and likely stay well clear of the bottom if they do in fact top balance. Sounds to me like it could be a personal preference on whether to top or bottom balance, depending on how you operate. My understanding is that if you go over 90% state of charge on a regular basis with LFP, will increase capacity loss over time, and thus reduce cycle life of the pack. It not top balanced, you could have uneven capacity loss in that case, could you not? (ie the lowest capacity cell(s) becomes even lower over time than the higher capacity ones)

You are on the right track, but sounds like you are over looking something. Regardless if you top balance or bottom balance you are still limited by the weakest cell. The danger with top balance is the stronger cells driving weaker cells into polarity reversal, thus destroying them. If you bottom balance you eliminate that threat. Second point and you hit on it is you will get much longer cycle life by only charging to 90% or less. One of the challenges of switching to LFP is getting out of the Pb box. With Pb batteries you want to keep them at 100%, but not Lithium as that does stress them and shorten cycle life. For long term storage you store them at 60% SOC and lithium are best operated and happy between 10 to 90% SOC as it does not weaken them in any way. That is very easy to to with a BMS because the charge/discharge is very flat in that range.

[northerner]

Would it be desirable to factor in current flow (ie through battery pack) as well as voltage to decide on the trigger point for the LVD and HVD? That way you could have trigger points based on the actual current flows, rather than doing approximations with timing, or does it work fine anyway? I guess that would only apply to someone designing a BMS, unless a unit is available that has this factored in?

I have no experience with LiFePo4, but have done a little research due to my interest in them. My understanding is that some users prefer to top balance, if they find that they more consistently use the top end of the pack, than the bottom. And they may only occasionally go towards a lower charge state and likely stay well clear of the bottom if they do in fact top balance. Sounds to me like it could be a personal preference on whether to top or bottom balance, depending on how you operate. My understanding is that if you go over 90% state of charge on a regular basis with LFP, will increase capacity loss over time, and thus reduce cycle life of the pack. It not top balanced, you could have uneven capacity loss in that case, could you not? (ie the lowest capacity cell(s) becomes even lower over time than the higher capacity ones)

In either case, I agree that it's better to manually balance the cells and not use the current shunts.

[Willy T]

I don't follow you on this. I fly RC planes, and a common battery is 3S 3000 mah or 11.1 volts @ 3 AH. A 40 watt panel would generate 3.6 amps or a 1.2C charge rate which is normal for LiPo batteries because they have extremely low internal resistance and can be charged very fast. Solar is going to be around C/10 as with most EV's. The only real difference between an EV and RE application is going to be the discharge rate. RE applications are going to be very low rates of C/20 or less, where an EV will be 1C and higher.

Where am I going to see the quoted rates when I have 120 amps @ 48 v going into 800 amp hrs of batteries. It's bogus talk without any meaning to " Our Systems " and the way "We charge " . I guess I should have said like a 40 watt panel on a EV.

[Sunking]

Sounds like a 40w panel on a RC battery pack.

I don't follow you on this. I fly RC planes, and a common battery is 3S 3000 mah or 11.1 volts @ 3 AH. A 40 watt panel would generate 3.6 amps or a 1.2C charge rate which is normal for LiPo batteries because they have extremely low internal resistance and can be charged very fast. Solar is going to be around C/10 as with most EV's. The only real difference between an EV and RE application is going to be the discharge rate. RE applications are going to be very low rates of C/20 or less, where an EV will be 1C and higher.

[Willy T]

Well guess what? C/12 is exactly what would be used in a solar application.

Exactly !!!

This might be a good reason for people to talk to someone that actually has a Solar bank of significant size that charges with PV and discharges the bank on a daily basis.

if you charge your cells at 0.05C to maybe 0.1C (which could be the case with a regular solar setup )

Sounds like a 40w panel on a RC battery pack.

[Sunking]

I know little about this subject, other than appreciating what you are saying. So I am hoping to see some discussion by the people who are knowledgeable.

What PN is saying is an EV charges at a much higher rate. Well I should say some do if they can afford a large format charger. On the DIY side most guys use 144 volt 100 AH LFP. They use a C/10 charge rate of 10 amps which is the maximum a 120 Vac receptacle can handle at 1500 watts. Well guess what? C/12 is exactly what would be used in a solar application.

The jest of what PN is saying if you charge slowly, then the cut off voltage point should be lower. A fully rested LFP battery at room temp at 100% SOC is 3.43 volts. So the conclusion is with a bottom balanced pack at a slow charge is to lower the cut-off voltage which I agree with. Its a tweak to match the application.

Myself on the other hand I have a 48 volt 50 amp charger operating at 240 Vac circuit, so I am charging at C/2, and thus set my cut-off at 3.6 volts. When the batteries rest the weakest cell has the highest resting voltage of around 3.35 volts which is about 90 to 95% SOC which is perfect. The stronger cells are at a slightly lower SOC, but have the exact same capacity because I referenced at the 0% of bottom of the scale rather than the top. The weakest cell is around 102 AH when fully charged. My strongest cell is up around 115 AH. So when I fully charge the weakest cell it has the rated 100 AH along with all the other cells. So when I discharge when I get to 0% all cells arive at 0 at the same time and cannot damage any weaker cells. The cliff is 2.0 volts. So we set our LVD for something higher than 2.0. In an EV with much higher discharge rates I can set for a lower voltage say 2.5 for 15 seconds, or 2.1 instantly. For a slow discharge in solar you can set them a bit higher with longer time delay.

The beauty of bottom Ballance is even if your LVD fails all cells reach 0% at the same time, and if that happens no stronger cell can destroy the rest by reverse polarity. If working with an Inverter you have two fail safes. One is you LVD will operate at roughly 2.9 vpc x 16 = 46.4 volts. If that fails your inverter also has a cut-off voltage at 42 volts or 2.6 volts per cell making it almost impossible to over discharge your LFP battery.

[PNjunction]

I agree wholeheartedly with Sunking for EV use, and a GREAT writeup on that situation.

Generally, although it is not related to Peukert, but only seems similar at the edge-cases, the higher the current you charge at, the higher your cutoff voltage can be. Accordingly, the lower the current, the lower the voltage setting should be. At very low currents such as we do with solar of an appropriately sized bank, you are fully lithiated at a lower voltage.

At very low current, but high cutoff voltages, it is interesting to witness the stall in voltage at zero amps (not something you strive for actually), and if allowed to sit long enough, watch the voltage rapidly rise with NO current flowing! At this stage, one has transitioned beyond the charging stage, and into the parasitic reaction stage causing the rise in voltage with no current.

Having smaller cells allows me to watch this phenomenon in a reasonable amount of time. Aside from a negligible increase in cell capacity, seeing this with my own eyes is what convinced me to be conservative - in this low current application.

Because of this low-current app, I have no need to go beyond the norms one would use with any battery such as an LVD and HVD set conservatively and just making sure I start out sanely. I do an initial top-balance, even though that really means nothing to ensure I'm reasonably sane. However many would argue that last point.