ESS system shutting down high temp due to excessive charging when off grid

Question

question

daryl asked Sep 3, '20

ESS system shutting down high temp due to excessive charging when off grid

I have a 3 phase Multiplus II 5kVA ESS system with 28kWh of BYD battery, 15kW of AC coupled PV on the AC-OUT of the Multiplus II's and 2kW of DC coupled PV. I believe I am compliant to all of the Victron design guidelines for this system.

During a grid failure, the AC coupled PV flows into the batteries until the Multiplus II's decide that they are running too hot and shut down. This then causes a complete power failure.

How can I either:

a) Get the system to dial back the AC coupled PV through frequency shifting so that the Multiplus II's don't go into temperature overload

or b) limit the charge current to the batteries during a grid failure? I already have a "Maximum Charge Current" set under the DVCC settings but seems this is ignored during a grid failure.

EDIT:

To be clear, my assertion is that the Victron doesn't even try to use frequency shifting to avoid inverter overload when charging the batteries.

Thanks in advance

ESS

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Modifications Space

Node Red Space

Answer 1 · 2020-09-03T02:53:37Z

pracman answered · Sep 03, 2020 at 02:53 AM

I reduced the Charge voltage which calmed my two down. They had been allowing a bit much despite a few other settings like yours.

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Answer 2 · 2020-09-03T03:53:09Z

rrroonn answered · Sep 03, 2020 at 03:53 AM

Hi @daryl, just my personal experience ....

I have a similar setup (though smaller) with SMA 5000TL solar inverter. The SMA was installed a few years ago whilst the Victron gear just recently. During "off-grid" testing I discovered that the same problem. I did some research and discovered that the SMA has a frequency power-derating function that was not turned on. I had to get an electrician with a grid-code to come in and make the changes. Once this was done, the Victron was able to manage the SMA output in most situations.

I have Pylontech batteries rather than BYD, but I suspect that they have the same behavior. The BMS in these batteries will probably see high voltage (depending on max-charge rate and current State of Charge (SOC) and turn the battery off to protect themselves. I have had this happen to me even after the SMA changes.

The problem is that controlling an AC coupled solar inverter via frequency is not instantaneous - actually takes a few seconds for this to settle down. During this time, all the available solar energy has to go into loads and the battery. If the battery decides the voltage(caused by high current) is too high before the AC coupled solar output is dialled back then the lights go out.

The requisite conditions for me were:

1. Battery fully charged.

2. AC coupled inverter output high

3. Sudden grid loss - for me it was caused by sustained high grid voltage

My fix was to use my home automation system to monitor these variables and turn off the AC-coupled inverter when this situation arises. I describe my approach a bit more here:

https://community.victronenergy.com/questions/60636/sma-grid-tied-inverter-and-multiplus-ii-battery-ov.html

Hope this helps,

Ron

3

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

daryl commented · Sep 03, 2020 at 04:03 AM

Thanks for your reply @rrroonn. I've confirmed my AC solar is frequency shifting with Victron in general. My issue is that Victron is not even trying to shift in order to reduce the AC PV output and hence charging current to the battery in order to avoid overheating the multiplus inverters.

I don't think I'll see the problem you do with SMA during sudden grid loss . My Solaredge inverters shut down fairly instantly when the grid is lost, and start back up a couple of minutes later and ramp with load from then on.

0 ·

rrroonn daryl commented · Sep 03, 2020 at 04:35 AM

@daryl There is a conflict in what you said - if the Solaredge in on the AC output side then it will not see the initial grid loss. If it is on the AC input side then it doesn't matter what it does when the grid drops.

There is a setting in the ESS assistant to define the frequency changes ...

Ignore the actual values, and set them according to the configuration of the SolarEdge.

Monitor the frequency here to check that it does (or nor) ramp up the frequency:

0 ·

aa1.jpg (30.7 KiB)

aaa0.jpg (19.2 KiB)

aaa2.jpg (36.4 KiB)

daryl rrroonn commented · Sep 03, 2020 at 04:37 AM

@rrroonn

>here is a conflict in what you said - if the Solaredge in on the AC output side then it will not see the initial grid loss. If it is on the AC input side then it doesn't matter what it does when the grid drops.

I understand what you are saying. They are on the AC output side, there is a relay on the multiplus that triggers an input on them to tell them that they are going off grid and to change the grid code. I guess this causes them to temporarily drop out, which is probably a good thing. On your SMA solution, another option may be to use the ac-out2 and have that drop out for a couple of minutes using assistants when grid drops .

I will double check that frequency ramping is working correctly shortly - the charging/overheating problem wont occur now that the battery has reached 100% SOC for the day.

0 ·

Answer 3 · 2020-09-03T06:35:16Z

Guy Stewart (Victron Community Manager) answered · Sep 03, 2020 at 06:35 AM

Hi @daryl,

There was a bug fixed in the version 0173 of ESS assistant that would cause this issue in multiphase systems.

Please get in touch with your installer, and explain the situation. The next step is to make sure that the ESS assistant is reloaded from the latest version of VE.Configure to your Multis. Remember it must be updated on all units seperately.

There is more information here that is available to Victron Professionals - Assistants Change log

1

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

daryl commented · Sep 03, 2020 at 07:27 AM

Thanks Guy, I am 99% sure I am running something newer than ESS 0173. Reading the changelog on Victron Professional, 0173 was released in Jan 2019 and this system was installed only about 6 weeks ago. To be sure, I checked in VEConfigure if I had the latest assistants (I did), downloaded the newest version of VEConfigure from the website, checked for latest assistants again (no updates) and pushed all config (including the assistants) to all 3 inverters just now. I really don't expect this to help since I pushed assistants to all the inverters a week or two ago as well.

EDIT: I didn't realise you could right click assistants to check the version - I have 017A. I've not right clicked and upgraded at all, so I beleive I've been on 017A the whole time since system install.

I'll re-test the off-grid charging in the morning once the SOC is low and the solar is high again to be double sure.

I'd welcome any other suggestions you may have.

Daryl

0 ·

Answer 4 · 2020-09-03T09:00:12Z

Guy Stewart (Victron Community Manager) answered · Sep 03, 2020 at 09:00 AM

Hi @daryl,

With all your firmware up to date, I've just had a closer look at the system and the errors that occurring.

Another thing I've noticed, the system does not meet the Victron guidelines for AC PV to Battery capacity.

This is documented here - https://www.victronenergy.com/live/ac_coupling:start#lithium_batteries

100Ah at 48V is required per 1.5 kWp of AC PV.

For your system, that is ~48 kWh of lithium for your 15 kWp, but your system only has 28 kWh.

I missed it the first time around, but when looking at your system on VRM, I was seeing DC ripple warnings, and very high spikes of battery voltage.

dc-ripple.jpg (843.1 KiB)

4

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

daryl commented · Sep 03, 2020 at 09:32 AM

Thanks for the observation Guy. I will look into that further. I can fairly easily move half of that solar from AC-OUT to AC-IN worst case.

I do see this (maybe incorrectly?) as unrelated to the high temperature alarms/shutdowns caused by high AC just before 12:00 when I turned off the grid to test. I would still like to get to the bottom of how to avoid that. My view is the Victron should lower the charging rate to stop the Multi's overheating. There seems to be no user definable configuration related to this, and I've not been able to find any documentation about the expected behavior as a result of the changes on 0173:

0 ·

1599125288199.png (13.9 KiB)

1599125508428.png (10.1 KiB)

Guy Stewart (Victron Community Manager) ♦♦ daryl commented · Sep 03, 2020 at 10:01 AM

Are you sure that the solar edge is correctly configured?

See here - the system is ramping up the frequency to 51.3 hz, that should reduce the Solar Edge to 0%, but the solar edge is still producing 18 kW.

0 ·

no-freq.jpg (782.3 KiB)

daryl Guy Stewart (Victron Community Manager) ♦♦ commented · Sep 03, 2020 at 10:14 AM

Perfect, thanks - I feel like an idiot for missing that :-)

0 ·

Guy Stewart (Victron Community Manager) ♦♦ daryl commented · Sep 03, 2020 at 12:14 PM

You're welcome @daryl, I hope I didn't say anything in a way to make you feel that way. I want the process of getting stuck and asking for help here to feel empowering.

I am very grateful for the VRM, it's a very powerful troubleshooting tool.

Anyway you should be feeling feeling pretty good once you get all sorted and working smoothly, and understand even more about how your power system works.

0 ·

question