Possible memory effect in Freedom Won.

[Jaco De Jongh]

Recently I got involved in making a few changes to a system that has been running for almost 18 months. For the first 18 months the battery's were  only allowed to discharge to 50% (Partial discharge) recently the client requested that the batteries be discharged to 10%, yesterday I made the config changes to allow it and when I checked the system this morning I found this:

SOC came down to its normal 50 percent and after that drops to 10 % immediately....... 

Upon reading up a bit I found this:

"Here we report a memory effect in LiFePO4-one of the materials used for the positive electrode in Li-ion batteries-that appears already after only one cycle of partial charge and discharge. We characterize this memory effect of LiFePO4 and explain its connection to the particle-by-particle charge/discharge model. This effect is important for most battery uses, as the slight voltage change it causes can lead to substantial miscalculations in estimating the state of charge of batteries." 

Can this be related? Anyone  out there whit a similar experience or some knowledge on this topic?

[___]

2 hours ago, Jaco de Jongh said:

SOC came down to its normal 50 percent and after that drops to 10 % immediately....... 

The BMS got out of sync, or that would be my guess. Rather a hectic amount though!

We always say that voltage is not a good measure of SOC, but that's a bit too general and only partially true. Voltage is a good indication of when you reach the extremes. When an LFP cell drops below 2.8V.... it is completely empty. When the voltage spikes up past 3.6V.... it is completely full. So a BMS will reset/recalibrate at the extremes, and estimate in the middle. We already covered the inaccuracy of low current draws in another thread, though I can't say how much of a role it would have played here. I'm guessing it played almost no role, the Orion BMS is one of the better ones in that regard.

So what happened is the voltage on one of the cells dropped of rather sooner than expected and the BMS went "whoa! I'm really far out, let me reset!".

Chances are the SOC will be more accurate now. Let's just hope the capacity isn't significantly diminished.

[Youda]

Hi @Jaco de Jongh

How do you measure SOC, Volts and Amps here?

• BMV700?
• Via internal BMS of the battery?

To me it sounds like a problem with the BMS/BMV calibration: In order to reset the SOC counter of the BMS, each LFP cell should be charged to 3.5V from time to time. Otherwise, the SOC will be calculated based on the charge/discharge Amps counter only. As a result, SOC cummulative error will grow over time.

Example:
If charging to 54V only, the internal balancers won't be able to engage and the cummulative SOC error will grow. After couple of weeks, you will end-up in a situation where the SOC counter is indicating a full charge, but in reality, the battery is charged to 80% only. Then, as you start discharging, the cells will hit the minimum voltage too soon. Almost every modern BMS has a low-level cell voltage protection, so it will immediatelly disconnect the battery from the inverter, or will declare a low SOC and assumes that the inverter will changeover to GRID and starts charging the battery ASAP.

 

[Jaco De Jongh]

Just now, Youda said:

Via internal BMS of the battery?

Getting it from the BMS.

[Jaco De Jongh]

8 minutes ago, plonkster said:

BMS went "whoa! I'm really far out, let me reset!".

 

2 minutes ago, Youda said:

To me it sounds like a problem with the BMS/BMV calibration:

Thanks guys, so what I gather from this is that theoretically  over time it might correct itself as well. It just needs new "Reference" points to work from. 

[Youda]

Okay, then, to me it looks like you have too low C.C. voltage -> balancers are not able to engage.

Please, take a look at BMS SOC chart when the battery is charging. Does OSC increase slowly and nicely up to 100%, or do you see a bump around 80%?

Like this one:

[Jaco De Jongh]

12 minutes ago, Youda said:

Please, take a look at BMS SOC chart when the battery is charging. Does OSC increase slowly and nicely up to 100%, or do you see a bump around 80%?

So far from the bottom up it looks good, a gradual increase from 10%.

Ill have to see what it will do at 80%, in the past it was a smooth gradual increase,

but ill need to see what it will do later today. 

[Youda]

It may sound dumb, but the smooth SOC curve during charging is actually BAD.
Basically, it means that the SOC has been calculated purely using Amps counter. And it's higly unlikely that BMS is perfectly synchronised with the characteristics of all the cells in a pack.

Therefore, you want to see the bump from time to time. That bump indicates that cells reached their top voltage, balancers were engaged and SOC counter was reset.

I've just checked the FreedomWon Lite manual and it says that:

• nominal voltage is 52V (that means 16-cell LFP x 3.2V)
• you should be charging to 55.5V
• do not discharge below 10%, as BMS protection will disconnect the battery

Based on your observation, I would suggest you to check the C.C. and C.V. settings. It may be too low.

https://www.freedomwon.co.za/wp-content/uploads/2019/08/freedom-lite-installation-manual-26-07-2019-_1_.pdf

 

Edited August 13, 2019 by Youda

[Jaco De Jongh]

9 minutes ago, Youda said:

you should be charging to 55.5V

Charging is set to 55.8V

 

10 minutes ago, Youda said:

do not discharge below 10%,

Thanks, this was the agreed minimum. 

[Guest]

Well blow me down. First I learn here on PF that BMS'es are not that accurate ito SOC. Understandably so. 

Now this ... 18 month old bank, very interesting to see where this goes i.e ito the tweaks and quirks of different makes and BMS'es of lithium banks.

Question:
With lead acid banks we all know you just don't go below 20% DOD / 80% SOC unless there are failures, and we did that from day the bank was installed.
Would I be wrong to think that if you have a lithium bank, no looming power failures, you set the SOC to the min the bank allows from day one? Use it.

If it is not a good idea, why?

[___]

18 minutes ago, The Terrible Triplett said:

Would I be wrong to think that if you have a lithium bank, no looming power failures, you set the SOC to the min the bank allows from day one? Use it.

Yes and No. Most LFP batteries only balance at the top. So you have to recharge them fully every few days (once a week or so). They don't get damaged if they are left in a partially discharged state, but they do get unbalanced. If you simply set the MinSOC to 10%, and you don't have enough sun to get it all the way to 100%, then that might be a problem in the long run.

Ideally you want to use as much of the battery as possible, but not so much that you can't get it recharged in a reasonable time. It seems to me that the best thing is to occasionally discharge it to a nice DoD. Another thing you can do instead of this, is to look at individual cell voltages when it's at a lower SOC (this might need additional cabling and software, it's unfortunately not included in the CAN-bus comms). With my batteries I can use the bluetooth feature and VictronConnect to inspect cell voltages. If you don't see variations in the voltages at lower levels of charge, then there is probably no reason to be too concerned.

[___]

This is really useful:

[Guest]

7 minutes ago, plonkster said:

Ideally you want to use as much of the battery as possible, but not so much that you can't get it recharged in a reasonable time.

Good point.

I asked based on the premise that one has enough panels on the roof to do that with ease daily, and with bad weather, weekly, or one can easily change the settings to say i.e. Keep the batteries Charged, we need this rain.

[___]

Just now, The Terrible Triplett said:

on the premise that one has enough panels on the roof

Yeah, I'd take them down to 20% SoC. I would not take them down to zero, that does have a significant effect on cycle life. In general I agree. You paid for the batteries. Use them. Unlike lead acids, there is not a significant increase in cycle life if you cycle them shallowly.

I cycle mine only to 50%... but that is because I don't have enough PV to recharge them fully... not with all the day time loads

[Guest]

11 minutes ago, plonkster said:

... but that is because I don't have enough PV to recharge them fully...

3.5kw array seems to do the thing in winter ... do it, you know you want to.

Still tweaking the loads and times with the weather in-between ... we need the rain, we need the rain ...   
Batts are full every day. With a 450ah bank with SOC of 80% / DOD of 20%. 

[Jaco De Jongh]

19 hours ago, Youda said:

Therefore, you want to see the bump from time to time.

Here is the bump now, but I suspect its a bit to big 

[Youda]

This one is way too big!

Either the shunt of the internal BMS is very inaccurate, OR the cells have already lost roughly 40% of their capacity. Check how the discharging curve will look overnight. If there will be the same huge drop from 50% to 10%  like yesterday, you have to reach the manufacturer.

Also, if you have a spare BMV700, you can use it to measure the remaining capacity of the battery. Two findings might occur:

- BMV will show that the battery still has it's nominal capacity -> internal BMS of the battery is bad
- BMV will show that the battery has much lower than nominal capacity -> damaged cell(s)

[Guest]

1 minute ago, Youda said:

... damaged cell(s) ...

Interesting. We may then hear of how they deal with under warranty claims.

[Jaco De Jongh]

33 minutes ago, Youda said:

Also, if you have a spare BMV700, you can use it to measure the remaining capacity of the battery.

I do have one, I will ship it down to the installer. I feel bad for getting involved in this so late, The installer asked me for some help about a month ago... I hope the batteries are still fine.. 

[root]

Very interesting. Now I also know that the bump is indeed normal.

Below is a shot of our pylon 6x US2000B (3S2P) bank from 10 Aug. Due to rain, cloud and crazy misty days the bank hasn't top balanced since 12 Jul. So close to a month. The bank cuts over to grid at 25% SOC. But due to internal draw it can go down to 18% ish overnight since it only gets solar to recharge.

Bump is from 89% SOC to 100% SOC. In the past its been from 90/91/92. So there is some drift. But not too bad for 30 days no top balance.

[JacquesVDM]

My 4x US3000's does the same. Charges to 89%, stays there for about 30 minutes and then goes to 100%. Been doing it since day. Cycling it down to about 55% each day.

[Youda]

That's okay.

But the jump from 55%, as described above, is not 

[Jaco De Jongh]

16 hours ago, Youda said:

Check how the discharging curve will look overnight. If there will be the same huge drop from 50% to 10%  like yesterday, you have to reach the manufacturer.

No improvement . Now to inform everybody involved that there might be a problem. 

[Jaco De Jongh]

On 2019/08/13 at 12:10 PM, The Terrible Triplett said:

Interesting. We may then hear of how they deal with under warranty claims.

Interesting story so far. These Freedom Won's are build locally by a manufacturer in the West rand using imported cells ext. They only offer support via WhatsApp and the installer got queued early in the morning, He is number 3 in the queue and they will come back to him once they have sorted the other 2 tickets. He asked for a number to call because he has to much info to share, and they told him that its not possible, "Support by WhatsApp only"    

In the meantime the trend repeats itself. 

 

[___]

10 hours ago, Jaco de Jongh said:

These Freedom Won's are build locally by a manufacturer in the West rand using imported cells ext

I know the guys well. It's Sinopoly cells, with an Orion BMS. On the face of it, I think there is a bad cell and this should be covered by the performance guarantee.