Frequency Shift Power Control- gradual ramp?

Q&A and troubleshooting
, , , , , ,
1

Kia ora all, We have recently converted our off grid Easysolar II GX 3000 from using AGM to LiFePO4 with a (non-ve.direct) BMS. We also added a second set of panels running into a Sunny Boy SB3000HF grid tied inverter. The grid tied inverter connects to the AC Out of the Easysolar, standard AC coupled usage. We use the PV Inverter Assistant to tell the EasySolar to use Frequency shift Power Control to manager the power supplied by the grid tied inverter. The grid tied inverter and its panels are both under the power capacity of the EasySolar inverter (so meet the Factor 1.0 rule). It works nicely to power AC loads directly during the day.
When the LiFePO4 batteries are exactly full, they signal 100% to the the EasySolar by BMS-CAN but also at almost exactly the same time, the LiFePO4 BMS shuts off the charging relay so the Easysolar can immediately not sink power into the LiFePO4. The EasySolar then uses Frequency Shift Power Control to signal to the grid tied inverter which sees the frequency rise very suddenly and shuts off, calling it a Grid Fault. It then backs off and does not try to deliver power for another 5-10 minutes because of the fault condition.
Unfortunately the grid tied inverter might be carrying most of the AC load at the time of shutoff, for example, charging an electric vehicle. Suddenly the AC load has to be picked up by the Multiplus, until both the LiFePO4 battery has dropped to less than 100% charge and the grid-tied inverter has been backed off long enough to want to try again.
My question is- how can I get the EasySolar to recognise (via the BMS-CAN) that the LiFePO4 battery is approaching full (ie it is at 98% or so) and to gradually increase the AC frequency so that the grid tied inverter gradually reduces the power it supplies, rather than the instantaneous switch that happens now.
Thanks

2

You need to change the set points in the PV inverter assistant to change the frequency at a slightly lower voltage than the 100% shut off voltage. Likewise the Charger function of the inverter also needs to be set to this lower voltage. Then the battery will get to 99%, and not shut off. It would also be possible to change the P(f) function in the sunny boy so as the output drops off sooner - say 0% at 51.8Hz, rather than 52Hz.
If you are relying on the BMS-CAN interface, then the DVCC function needs to be enabled, in order to allow the GX to control the battery charge current dynamically.

1 Like
3

Thanks @Mike. re the voltage, I am on the horns of a dilemma:

  • the LiFePO4 batteries are brand new but showing SoH only about 90% of the advertised capacity. The vendor tells me to keep the voltage high and over time the batteries will report their advertised capacity. Keeping the voltage high has resulted in a bump in the SoH so the vendor may be right;
  • otoh, as you point out keeping the voltage high results in the battery ‘slamming on the brakes’ at 100% SoC (even though still 90% SoH).

I am locked out of most changes to the Sunny Boy’s parameters; most of them are hidden behind a “Grid Guard Code” SMA gives out only to authorised installers (I am not one, and it’s a second hand inverter) and is specific to the installation and expires after a few hours.

I have only read about DVCC in context of lead-acid batteries with a BMV monitor and multiple DC-coupled MPPTs. Is there a downside to turning it on when these things are not all true?
Thanks, Graham

4

I have just found the DVCC pages in the Venus GX manual, having previously not read them.
It lists some settings within DVCC and how they should be set for different BMS’, but it does not mention mine.
My BMS is CANbus connected and I presume it to be Pace BMS based on the physical arrangement of the front panel, look and arrangment of the LCD menus and look of the PC-based software. The batteries have the same front panel arrangement (specifically, number, type and arrangement of RJ45 and RJ11 ports, dry contact relays, charge lightsetc) as the SOK batteries sold in the US by CurrentConnected.com.
The SOK battery manual SK48V100 Server Rack Battery Owner & Operator's Manual | Current Connected has a description of using those batteries with DVCC. But it doesn’t cover the parameters listed in the GX manual’s table of BMS types- VE.Bus, Freedom-won, BYD, Pylontech, BMZ, MG Electronics.

My Battery reports this information to the GX device which shows it on the console:

  • Charge Voltage Limit (55.7 V)
  • Charge Current Limit (189.0 A)
  • Discharge Current Limit (189.0 A)

Does anyone know the correct DVCC parameters for a Pace BMS? Is it enough to let DVCC Auto-config? There is no obvious Auto-config setting in DVCC, does that just mean I should enable it and accept the defaults presented?

Thanks!
Graham

5

Normaly the BMS safety cut off voltage should not be reached. Often this is 3.65V. Default charging by Victron only goes to 3.55V. The is wel below the safety cut off. Problem is that is mesured by Victron over the whole battery pack and not over induvidual cells like the BMS does. This means induvidal cells can reach 3.65v tripping the BMS if the pack is not well balanced. I suggest looking into voltages of all cells just before the BMS trips and see if there is unbalance. And maybe rebalance the battery pack

6

You can enable it, and check through the defaults. Most of these should be good. Some Lithium batteries need to cycle a few times to get to full capacity. The ESS should pick up the correct limiting values from the BMS ok, these are normally broadcast on the CAN data

7

Thanks @MikeD and @Warzel
I am using it with the defaults and so far so good. The BMS relatively often reports High Voltage Alarm, but according to the battery & BMS supplier that should help the battery reach its designed capacity- so far it is showing only ~90% SoH.