Has anyone been able to get REC to fully charge a battery?

[MurphyGuy]

SunnyBoy inverters w/Freq Shit + 2 Sunny Islands (AC coupling) + 25kW lithium battery + REC BMS.

Everything works great, right up until the last little bit of charging.

Settings:
Balance Voltage Start 3.95
Balance Voltage End 4.05
End of Charging 4.1
Vmax Cell 4.2

The system works great right up until the cells reach around 4 volts, at which time, the Frequency Shift Power control engages and starts limiting charge current into the batteries. As the battery bank gets close to 4.05 volts, the charge current drops to 2 or 3 amps and sometimes even goes negative for a while as house loads kick on and off.
So here I am at 4.023 volts per cell, charging at 1.28 amps, and Freq Shift sitting at 61.89 hz. There's probably 2.5kW of power between 4.02 and 4.1 volts.

I've tried several setting changes in the REC and nothing is making a difference. My SOC meter thinks my battery is at 60% charge while at 4.02 volts because I have never been able to get the battery to a full 4.1 for the SOC counter to reset. Not once in 3 years.

I've written to REC in the past, they suggested changes, it didn't help. I just wrote them again and told them the problem continues.
Lets see what happens. Anyone have any ideas?

If we don't get it fixed this time, I'm going to stop recommending REC to people.

 

[Maitake]

What is the whole bank voltage?

 

[Maitake]

Vomax on SI to the battery 63vdc.
63/16=3.94v

Edit: bad math

Edit edit:
If you are getting 4V per cell you are already above specified output voltage for the SI.
Probably not the BMS's fault.

 

[SilverbackMP]

Well shit, I have close to $5K of REC BMS on order.

 

[toms]

I have had no problems across many installs using LiFePO4 chemistry.

I’m using a 0.05V spread across balancing (start 3.40, finish 3.45)

The ramping of the SB you describe is typical behaviour.

What chemistry of cells are you using?

 

[400bird]

To he OP:
Are you running 16s lithium NMC (or some other non-Lifepo4)?

Common 14s lithium and 16s lifepo4 both are under 63 volts when fully charged.

Well shit, I have close to $5K of REC BMS on order.

Your probably fine, unless you're running an unusual battery config.

 

[SilverbackMP]

Your probably fine, unless you're running an unusual battery config.

Likely CALB 230s in 2p16s. Maybe just 16s. I haven't made up my mind. Edit: does sound like the OP is running lithium ion or something else that's not as common.

 

[400bird]

Likely CALB 230s in 2p16s. Maybe just 16s. I haven't made up my mind. Edit: does sound like the OP is running lithium ion or something else that's not as common.

Yeah, with Lifepo4 cells, your absolute max is 16*3.65=58.4
That should be fine. I haven't checked the spec sheet, but every 48 volt inverter I've seen hits at least 60 volts.

 

[toms]

Haven’t heard back from the OP, but for everyone else watching i have attached a pic of my REC display.

The number circled isn’t included in the documentation i can see - but appears to be a coulomb counter from the last time the system reached 100% SOC.

You can reset the percentage to 100% via the PC link, (i don’t use wifi). - but this doesn’t change the number i circled.

If your system is tapering current it means it is in the balancing phase. There are two reasons it won’t reach 100% from there. One is the cells are too far out of balance which will cause the REC to stop charging if it exceeds an internal temp.

The other is if you have the balancing voltage set too low, or the balancing delta too narrow.

I’m curious what changes you have actually made - and what difference they made.

 

[Baxter2120]

Set balance voltage end to the same value als end of charging.

Balance voltage end is missleading, because the rec is "balancing" any cell regardless of the charge current above this value. So you are blocking your cells from reaching the end of charge voltage.

Balancing Voltage END(BALV):

If errors 2, 4, 5, 8, 10, 12 are not present, any cell above balance END voltage is balanced regardless of the battery pack current.

Using LFP I am setting different values. End of charging is 3,45V, so balance voltage start is 3,45V and balance end 3,5V. In this balance window the rec balancer can do his job based on the cell impedance.

BTW, the cell impedance values are correct in your setup?

 

[toms]

Using LFP I am setting different values. End of charging is 3,45V, so balance voltage start is 3,45V and balance end 3,5V. In this balance window the rec balancer can do his job based on the cell impedance.

My settings are similar, except my end of charge is 3.5 (same as balance end)

 

[MurphyGuy]

What is the whole bank voltage?

I'm using new lithium cells from a Chevy BOLT (LG Chem, N2.1 60Ah). Configuration is 14s8p for a 480Ah battery pack.

 

[MurphyGuy]

To all posters above.

Sorry I didn't get back earlier, for some reason, the forum isn't notifying me I have new posts here. I'll pay more attention now.

Here's a screenshot of what's going on. REC BMS interface in the foreground, and in the back ground is the interface for the SMA SunnyBoy solar inverter in Island Mode.
Notice the frequency is in Frequency Shift mode (circled with arrow) We have full sun and if I was pumping energy into the grid, the array would probably be generating around 3000 watts right now. We have plenty of solar power, but the REC BMS is throttling it back WAY WAY too much.

I have tried playing with the balance voltages and no luck.

I've written to REC about this a couple times and no solution yet. This thing could charge at 5 amps all day long and never finish off the last 70mV.
Worse yet, if a house load kicks on and draws 20 or 30 amps for a few minutes, I loose that much juice and the REC will continue at that insignificant 5 amps.

It will actually start to discharge the battery as loads come on and off, and the REC response time is so slow at this point as to be irrelevant.

I think their BMS's are great, but they are designed for charging from a steady source of power and one that doesn't encounter loads while charging, like a vehicle.

There are no user adjustable settings for this problem and I've dug into it quite extensively now. I suspect everyone using a REC BMS in a solar storage application is going to run into this issue.

 

[MurphyGuy]

Set balance voltage end to the same value als end of charging.

Balance voltage end is missleading, because the rec is "balancing" any cell regardless of the charge current above this value. So you are blocking your cells from reaching the end of charge voltage.

Balancing Voltage END(BALV):

If errors 2, 4, 5, 8, 10, 12 are not present, any cell above balance END voltage is balanced regardless of the battery pack current.

Using LFP I am setting different values. End of charging is 3,45V, so balance voltage start is 3,45V and balance end 3,5V. In this balance window the rec balancer can do his job based on the cell impedance.

BTW, the cell impedance values are correct in your setup?

Very interesting. I just updated my settings to:
Balance Voltage Start 4.1
Balance Voltage End 4.15
Max Cell 4.2
End of Charge 4.1

I just flipped back to Grid Tied mode so my system is idle right now and not doing anything. I'll turn it back on when the sun goes down, let it power the house overnight to drain some energy away, and see how it behaves again tomorrow.

Your post surprised the hell out of me in that your balance voltages don't do anything until you've already charged the system. That's interesting, and I'm going to try it.

This is a mindset change from "get those cells balanced" to "allow chaos until full charge".

Thank you...

 

[OGSS]

One of the big challenges with balancing Lithium cells is that the voltage curve is so flat across its operating cycle it is not practical to try and balance at mid state. You need to get them up into a near full condition before you start balancing so as to give you a good indication of their state as they come out of the flat part of the voltage/capacity curve

 

[MurphyGuy]

Well balanced.

 

[SilverbackMP]

One of the big challenges with balancing Lithium cells is that the voltage curve is so flat across its operating cycle it is not practical to try and balance at mid state. You need to get them up into a near full condition before you start balancing so as to give you a good indication of their state as they come out of the flat part of the voltage/capacity curve

@MurphyGuy this is Off Grid & Marine Energy Systems, the main distributor/retailer for REC BMS in the states. I pinged them in an email.

 

[MurphyGuy]

@MurphyGuy this is Off Grid & Marine Energy Systems, the main distributor/retailer for REC BMS in the states. I pinged them in an email.

I bought mine straight from REC and have been using their tech support guy "Tine". Wrote him an email two days ago but no answer yet.

 

[toms]

That is actually the best feature of the REC. As soon as the first cell needs balancing, the current is reduced to keep the heat of the balancing resistors under control.

I’m surprised Tine couldn’t help you with this straight away - he is well aware of the way the end of charge algorithm works.

The problem you have encountered is definitely not a limitation of REC. I have installed dozens of these in off-grid systems with no issues.

Hopefully now you are using the correct settings you will be satisfied with your system performance.

 

[MurphyGuy]

That is actually the best feature of the REC. As soon as the first cell needs balancing, the current is reduced to keep the heat of the balancing resistors under control.

I’m surprised Tine couldn’t help you with this straight away - he is well aware of the way the end of charge algorithm works.

The problem you have encountered is definitely not a limitation of REC. I have installed dozens of these in off-grid systems with no issues.

Hopefully now you are using the correct settings you will be satisfied with your system performance.

Yeah, I think I was being too conservative with my numbers. I switched back over to the off-grid system for tonight to let the pack drain down to 3.7-ish. Then I'll fire up the solar again in the morning and it will charge at 100+ amps. Lets see if it will race up to the 4.1 (End of Charging) setting and then disable the charge. I expect it will taper off some, but that (1.1 amps x number of devices) would take forever to move a 480ah pack.

Baxter kind of opened my eyes as to how I should interpret the balance settings.

 

[toms]

Yeah, I think I was being too conservative with my numbers. I switched back over to the off-grid system for tonight to let the pack drain down to 3.7-ish. Then I'll fire up the solar again in the morning and it will charge at 100+ amps. Lets see if it will race up to the 4.1 (End of Charging) setting and then disable the charge. I expect it will taper off some, but that (1.1 amps x number of devices) would take forever to move a 480ah pack.

Baxter kind of opened my eyes as to how I should interpret the balance settings.

I think you’ll be surprised. I have installed a REC on a 600ah pack and it’s doing fine.

The current needs to taper when balancing, it doesn’t take much to reach 55°C then it cuts charging.

If your cells need more balancing, install cooling (a fan) on the REC module.

 

[MurphyGuy]

I think you’ll be surprised. I have installed a REC on a 600ah pack and it’s doing fine.

The current needs to taper when balancing, it doesn’t take much to reach 55°C then it cuts charging.

If your cells need more balancing, install cooling (a fan) on the REC module.

Doesn't seem to be working.

No cell has gone over 4.055 when charging at high amps, and the BMS is already ramping down at 61.60hz. It is only allowing 40% of the available solar power to charge the battery.

Per the instructions:
When any cell reaches the voltage interval between Balance Voltages Start and Balance Voltage End, the charging current starts to ramp down to 1.1 A x Number of Devices until the last cell rises to the End of Charge Voltage

But that's NOT what's happening. It is ramping down as soon as it gets within 100 mV of the END OF CHARGING setting. No cell has reached the 4.1 volt (Balance Voltage Start) setting, or has come closer than 4.055. I know because I'm sitting here staring at it.

As I'm typing this, it has now increased to 61.65hz, charging amps reduced to 19.5 amps, highest cell at 4.032
UPDATE: Now at 61.69hz with the highest cell still at 4.032.

The BMS is not waiting to reach BALANCE VOLTAGE START to begin its ramping down of the charge current. And I can predict, by past behavior, that by the time the first cells reaches 4.08 volts, the frequency will be up in the 61.90hz range, and it will only allow 3 to 5 amps of charge. By the time the highest cell reaches 4.1 volts, which it probably never will, the sun will be going down.

And this all assumes house loads stay steady. If someone turns on a vacuum cleaner or something, it will get even worse and possibly even go backwards.

I'm going to let it continue and see what happens, but it seem to be behaving exactly the same as yesterday. Gets within 100mv of Balance Voltage Start, then starts ramping down the current. It almost seems to follow the Inverse Square Law where twice as close = four times the attenuation.

Currently at 61.76 hz and trickling 9.07 amps into the battery with the highest cell being 4.032 volts. At that rate, the highest cell will never get to 4.1 before the sun goes down, and the lowest cell will be 10 to 15mV behind that one. My SOC and Coulomb counters will never reset.

 

[Baxter2120]

I am using REC-BMS together with Victron DVCC and can definitly confirm that it only starts to ramp down the charge current when the first cell hits the end of charge voltage.

Are you sure that the BMS tells the chargers to reduce the charge current and there is no internal charge control of your inverter/chargers active?

Can you see what the REC-BMS is communicating to the SMA SI Master via CAN?

 

[MurphyGuy]

Here's an update:

I am using REC-BMS together with Victron DVCC and can definitly confirm that it only starts to ramp down the charge current when the first cell hits the end of charge voltage.

Are you sure that the BMS tells the chargers to reduce the charge current and there is no internal charge control of your inverter/chargers active?

The Sunny Island's have no problem pumping 120 amps into the battery. I have the max amps set at 120 within the Sunny Islands.

Can you see what the REC-BMS is communicating to the SMA SI Master via CAN?

Can I see it? How would I do that?

The only thing I've done is to communicate with the BMS to verify it reports back the correct settings. For a while, I thought maybe the changes I was making in the user interface software were not being recorded to the BMS, but when using the Communicate function and having to type in the codes MAXC? SISN? etc, it returns all the correct values.

I have a 2nd brand new never used REC BMS unit here.. Going to plug it in an see what happens tomorrow.

And here we are where it always ends up.. It will sit like this for the rest of the day.

 

[Baxter2120]

Just to be sure, have you established the CAN communication between the REC-BMS any our SMA?