PW Inverter shut off issues...

[Dom_t]

So, after 8 months of having the Powerwall 2 with 10.54kw (In NSW Australia) of Solar on my roof, we have finally had a blackout. In total, it lasted around an hour with the Powerwall going into "Backup Mode" the instant the grid shutdown as expected. The only problem was that both our 5kw Fronius Primo inverters were dead. For some context, our house is on 2 phase with our ancient, soon to be replaced hot water system wired to one of the phases. Each of our 2 inverters is on one phase each. We got everything installed at the same time, with the installers commenting on the complexity of the install and our home wiring.

When the blackout happened during the day (Around 10 am) the battery was sitting at around 30% charge. So I would have thought the solar to have powered the 1.5kW/h draw from the house whilst charging the battery. (at 10 am, we were making around 4.5kW/h) After 20 Mins or so, one of the inverters came back online with an error message which meant that the inbound frequency was too high. I presumed this was the Powerwall. But, am I correct in saying that as the battery comes further to a complete charge, it alters the frequency to tell the inverters to slow down or stop generation of power? (The inverters are post-2015 code, and have the ability in Australia to do this)

I would understand it the battery was at 100% to shut down the inverters, but at 30% it seems odd. Side note: During summer we generate around 55-60 kW/h per day from our solar, with the majority of it being exported back to the grid once the battery is charged during the day.

I have called Tesla about this issue, and they indeed say that it's odd. I am waiting for them to look further into the data provided by the PW2.

If anyone else is having these issues, I'd love to know! As it might just be an Aussie thing...

-Dom

 

[shs1]

I have seen very similar failure modes with a similar setup. I have a single PW2 and a total of 11 kW of solar coming from 7 and a 4-kw inverters. Everything works great when the grid is on, but during simulated power failures I have documented five different failure modes which includes the one you are seeing. My analysis and Tesla agrees, is that all of these failure modes would be mitigated with a second PW2 and I hope to get a second one early next year.

What happens in a power failure is that when the grid goes down, chances are that your inverter will also go down and the house will be running off of the PW2 with no help from solar. The inverter will see that the PW2 is supplying a voltage and try and sync after a waiting period, typically 5 minutes. Some of the failure modes I have seen have to do with the fact that when the inverter does try and come up, it generates more power than a single PW2 can handle, > 5 kW, so the PW2 raises it frequency to tell the inverter to curtail its output. Whether or not your inverter can respond proportionally to Frequency Shift Power Control (FSPC), the rate at which the PW2 raises it frequency will likely cause the inverter to go offline. Then the whole cycles starts again, often with the same result 5 minutes later. One can beat or test the system by having the solar come up early in the AM when less than 5 kW is being generated, or by carefully adjusting the house load so that the solar input to the PW2 falls under the max charging rate. The latter case can be tricky as the PW2 may still not like a sudden in rush of power from the inverter even if it is less than the max charging rate and raise the frequency.

There are also failure modes associated with that fact that when the PW2 is powering the house without the help of solar, e.g. as the inverter is waiting to sync, that intermittent house loads trying unsuccessfully to come on line can cause a brief voltage drops that will cause the inverter 5-minute waiting period to restart, and if the intermittent load keeps trying, the solar may never sync up.

So, either intermittent house loads or solar charging rates greater than a single PW2 can handle can cause a persistent failure mode where the solar may never come on line successfully to help power the house or charge the PW2. I am certainly looking forward to having a second PW2, but until then, hopefully my trusty generator, which is integrated into the system, will get me through any power failures that come this winter, here in California.

 

[miimura]

I have also had my solar shut down during grid outages. I got Tesla Powerwall support on the phone during one of the blackouts and they were able to verify that the frequency was at 63Hz when it should not be. I have a small solar system compared to Powerwall capacity, so it should not raise the frequency until the batteries are nearly full. My support case is still open, but they were able to get the solar back online during that outage. I feel this is likely a bug in 1.26.0.

 

[shs1]

I have seen 63 Hz when it should not be, and very very slowly coming down to 60 Hz. I have even seen 66 Hz on one occasion. Lots of stuff in my home that does not like 66 Hz. I just checked and I am running 1.28.0 right now, but my extensive testing of failure modes was with 1.26.

 

[Dom_t]

Here’s guys. So it’s not just me !

I’m also on 1.26. Hopefully Tesla gets cracking on fixing this issue.

Thanks again!

 

[shs1]

Hopefully Tesla can fix the problem of the frequency being too high when it should not need to curtail solar production, but I was trying point out that with 11 kW of solar there may be times when the solar will be generating more power than a single PW2 can handle, and under those circumstances, it will legitimately raise the frequency to protect itself from charging at over its maximum rate of 5 kW.

 

[Dom_t]

Hopefully Tesla can fix the problem of the frequency is too high when it should not need to curtail solar production, but I was trying point out that with 11 kW of solar there may be times when the solar will be generating more power than a single PW2 can handle, and under those circumstances, it will legitimately raise the frequency to protect itself from charging at over its maximum rate of 5 kW.

Yes, I totally agree.

One thing that does happen and my technicians agreed on, is that one of the 5kw inverters still remains on. But no power is being sent to the house or the Powerwall. I understand that if the solar is making more than the battery can charge and the house demands, then the remaining solar can't go back to the grid because it's down. My inverters do have the ability to curve the output with the frequency changes the Powerwall sends out. Because my solar system is oversized to what my house actually draws, I don't mind if one of the inverters shuts down, as they pretty much crank out full 5kw of power by 10-11 AM.

 

[shs1]

You obviously do understand then and it is great that your inverters can scale back output in response to FSPC. I am not sure mine can when configured for on-grid operation.

 

[miimura]

Yes, I totally agree.

One thing that does happen and my technicians agreed on, is that one of the 5kw inverters still remains on. But no power is being sent to the house or the Powerwall. I understand that if the solar is making more than the battery can charge and the house demands, then the remaining solar can't go back to the grid because it's down. My inverters do have the ability to curve the output with the frequency changes the Powerwall sends out. Because my solar system is oversized to what my house actually draws, I don't mind if one of the inverters shuts down, as they pretty much crank out full 5kw of power by 10-11 AM.

So, you have one Powerwall connected to only one phase and that phase only has one solar inverter. Is that right? What is the kW size of the solar inverter that should be working with the Powerwall during a backup event?

 

[Dom_t]

You obviously do understand then and it is great that your inverters can scale back output in response to FSPC. I am not sure mine can.

check the certification codes for your inverters. Not sure about Cali, but in Aus, you can do a quick google search of the certification number to double check with it's ability to respond to freq changes

 

[Dom_t]

So, you have one Powerwall connected to only one phase and that phase only has one solar inverter. Is that right? What is the kW size of the solar inverter that should be working with the Powerwall during a backup event?

Correct. I have 2, 5KW Inverters, each on their own phase. (Explains why one goes dead during an outage) Since the battery and the inverter are on the same phase.

 

[miimura]

Correct. I have 2, 5KW Inverters, each on their own phase. (Explains why one goes dead during an outage) Since the battery and the inverter are on the same phase.

In that case, it's likely you're experiencing the same bug that I did. If I don't hear back from Tesla soon, I'm going to simulate another outage by turning off the main breaker and watch the frequency on my Kill-A-Watt meter.

 

[shs1]

My SNA inverters are good with FSPC when configured for off-grid operation, but I have been told by others on this forum that they are not OK with FSPC when configured for on-grid operation as that would violate UL1741 and my agreement with the power company. As configured, they do shut down via FSPC but I have not seen them curve their power to use your term.

 

[PBBear]

I presumed this was the Powerwall. But, am I correct in saying that as the battery comes further to a complete charge, it alters the frequency to tell the inverters to slow down or stop generation of power? (The inverters are post-2015 code, and have the ability in Australia to do this)

If it happens again, you can check the frequency from the PW2's API by browsing to the PW2's IP address with the following URL:
https://192.168.0.63/api/meters/aggregates (replace this IP address with your PW2's address)

You get these stats separately for the grid, battery, solar and load (house) - the frequency should be identical in all categories.

 

[Dom_t]

yeah, I'll be switching the main breaker off today when we are making some solar, since the

My SNA inverters are good with FSPC when configured for off-grid operation, but I have been told by others on this forum that they are not OK with FSPC when configured for on-grid operation as that would violate UL1741 and my agreement with the power company. As configured, they do shut down via FSPC but I have not seen them curve their power to use your term.

Interesting.

I'm not well versed in Californian Electric code, But in Australia, there are stupid amounts of red tape. Our electricity prices are some of the highest in the developed world. Our hot water system is soon to be replaced, being routed to the 2nd phase of our power. (We use it on off- peak times) Once we get a new one, we'd hope to have everything on the single phase, along with our dual solar inverters. However, we then need to resubmit an application to the government to have a 10kw system on the single phase. Aus grid (Australia's state grid operator) can allow up to 10KW systems connected to a single phase.

Am I right in saying that this would allow both inverters to power the battery/house during a power outage? More specifically, could this then solve the issue of the solar and battery not being able to power the hot water system? (For context, our hot water is the main item that reflects our bill) Maybe this is a discussion for a different thread.

 

[Shygar]

Would this only happen when solar generation is high? We had an outage for 2.5 hours a week ago and powerwall performed well, solar was just starting to come on so it wasn't generating much more than 1kW at the time. I'm also on 1.26

 

[Dom_t]

Would this only happen when solar generation is high? We had an outage for 2.5 hours a week ago and powerwall performed well, solar was just starting to come on so it wasn't generating much more than 1kW at the time. I'm also on 1.26

I believe so. I'll be testing this in the morning when our solar is only making around 1-1.5kw/h

 

[Dom_t]

If it happens again, you can check the frequency from the PW2's API by browsing to the PW2's IP address with the following URL:
https://192.168.0.63/api/meters/aggregates (replace this IP address with your PW2's address)

You get these stats separately for the grid, battery, solar and load (house) - the frequency should be identical in all categories.
View attachment 355935

Thanks for the link!

seeing 49 Hz all around. Will see what happens when I switch off the grid supply soon.

 

[miimura]

Interesting.

I'm not well versed in Californian Electric code, But in Australia, there are stupid amounts of red tape. Our electricity prices are some of the highest in the developed world. Our hot water system is soon to be replaced, being routed to the 2nd phase of our power. (We use it on off- peak times) Once we get a new one, we'd hope to have everything on the single phase, along with our dual solar inverters. However, we then need to resubmit an application to the government to have a 10kw system on the single phase. Aus grid (Australia's state grid operator) can allow up to 10KW systems connected to a single phase.

Am I right in saying that this would allow both inverters to power the battery/house during a power outage? More specifically, could this then solve the issue of the solar and battery not being able to power the hot water system? (For context, our hot water is the main item that reflects our bill) Maybe this is a discussion for a different thread.

If you want to combine everything onto one phase, you should get a second Powerwall, both to supply the added load from the water heater and also to absorb the added solar.

 

[shs1]

A Kill A Watt meter is also a low cost way to continuously monitoring, but not recording, the line frequency.