Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Powerwall 2 + UPS Connundrum - and solution

This site may earn commission on affiliate links.
So it sounds like it's all based on SoC of the PWs. If it truly was measuring the output of the solar it would even bother adjusting the frequency if there was no output. And the amount of time it would take to respond if the inverters miraculously came on would seem to be negliable.

It's going to be a rare occasion when this happens but nice to be aware of it none the less.

Last fall at the start of an outage, my Powerwall system raised the frequency to shut down my inverters because the batteries were full (PG&E PSPS event, Stormwatch enabled, batteries full). But there was no solar output to curtail because it was the middle of the night! :mad:

Bruce.
 
So it sounds like it's all based on SoC of the PWs. If it truly was measuring the output of the solar it would even bother adjusting the frequency if there was no output. And the amount of time it would take to respond if the inverters miraculously came on would seem to be negliable.

It's going to be a rare occasion when this happens but nice to be aware of it none the less.

As I mentioned earlier (page 15, ancient history on a thread like this!) on this thread from my testing back in June of 2019 the only two factors I could tell that affected the Powerwall frequency when the grid was down was primarily a) SoC and secondarily b) home consumption (load). Solar did not factor in and in fact around the 1740s mark of the test I did you can see that the PV turned back on but there was no change in frequency.

For my system I found that:

High SoC, low load = 62hz
Low SoC, low load = 60hz
High SoC, high load = 61hz
Low SoC, high load = 60hz

(Again the solar output did not play as a factor).

Utility outage simulation data dump

I'll caveat this all with this was a while ago and it's possible later firmware updates have changed the behavior and/or different systems' configurations cause different behavior.
 
Last edited:
We had Tesla adjust our settings a few weeks ago. Now we're dealing with Tesla and our electric company and a potentially bad transformer. We've had over 200 "outages" in the past few weeks. We didn't notice anything in the house since the Powerwalls handle everything and the UPS devices no longer complain. However, when I watch the solar production in the app, I can see over 200 Backup History events due to voltage issues on our line since mid-May.

Ironically, we had a big storm on Saturday which knocked out 4 power poles and multiple transformers in our neighborhood and we were off grid for nearly 40 hours. The utility came out and replaced those four poles and transformers overnight...but we're still stuck with the bad transformer on the one pole that didn't blow down. <sigh>
 
We had Tesla adjust our settings a few weeks ago. Now we're dealing with Tesla and our electric company and a potentially bad transformer. We've had over 200 "outages" in the past few weeks. We didn't notice anything in the house since the Powerwalls handle everything and the UPS devices no longer complain. However, when I watch the solar production in the app, I can see over 200 Backup History events due to voltage issues on our line since mid-May.

Ironically, we had a big storm on Saturday which knocked out 4 power poles and multiple transformers in our neighborhood and we were off grid for nearly 40 hours. The utility came out and replaced those four poles and transformers overnight...but we're still stuck with the bad transformer on the one pole that didn't blow down. <sigh>

I've found that the Powerwall can be pretty sensitive to grid events. Just a few minutes ago the power blipped very briefly and while my oven (which is not backed up) was still on and the clock didn't reset the Powerwall switched to backup mode for 5 minutes.

Pertinent to this thread my Powerwall was at 99% charge when the blip happened. As expected the solar shut down and I was surprised to see that after a few moment it actually started producing again at 1kw. House demand at the time was 1.2kw so there was a small draw from the Powerwall which was still reading 99%. It's pretty cloudy out this morning but since my system is capable of 8.55kw and I have the IQ6 set to use frequency based ramp down I expected the Powerwall to have to dip down a few more percentage points before the PV started producing again.

Unfortunately I didn't have my data collection tool running at the time so I couldn't see exactly what the frequency was. I'll look to do a test here again sometime soon but thought it was interesting to share that my PV was producing (a little) even though the Powerwall SoC was 99%. I wonder if they changed some of the behavior on recent firmware updates?
 
I've found that the Powerwall can be pretty sensitive to grid events. Just a few minutes ago the power blipped very briefly and while my oven (which is not backed up) was still on and the clock didn't reset the Powerwall switched to backup mode for 5 minutes.

Pertinent to this thread my Powerwall was at 99% charge when the blip happened. As expected the solar shut down and I was surprised to see that after a few moment it actually started producing again at 1kw. House demand at the time was 1.2kw so there was a small draw from the Powerwall which was still reading 99%. It's pretty cloudy out this morning but since my system is capable of 8.55kw and I have the IQ6 set to use frequency based ramp down I expected the Powerwall to have to dip down a few more percentage points before the PV started producing again.

Unfortunately I didn't have my data collection tool running at the time so I couldn't see exactly what the frequency was. I'll look to do a test here again sometime soon but thought it was interesting to share that my PV was producing (a little) even though the Powerwall SoC was 99%. I wonder if they changed some of the behavior on recent firmware updates?
We've had a few "outages" per month over the past year and a half but with the most recent change, we're having several dozen outages a day. These "outages" have caused our solar system to drop from producing over 140 kWh a day down to less than 90 kWh a day.

We're thinking Tesla must've adjusted something besides the frequency, perhaps narrowing the limits on allowed voltages. We've known since 2018 that the voltage drops when our neighbor's A/C or pool pump is on. We complained to our utility and they wanted something like $10k for a study to be done so we opted to live with low voltages at certain times of the day. Originally, we found that only 100 outages could be shown in the app. Tesla made a change and now we see 280 outages since July 2018. Over 230 of them have been in the past 3 weeks.
 
We've had a few "outages" per month over the past year and a half but with the most recent change, we're having several dozen outages a day. These "outages" have caused our solar system to drop from producing over 140 kWh a day down to less than 90 kWh a day.

We're thinking Tesla must've adjusted something besides the frequency, perhaps narrowing the limits on allowed voltages. We've known since 2018 that the voltage drops when our neighbor's A/C or pool pump is on. We complained to our utility and they wanted something like $10k for a study to be done so we opted to live with low voltages at certain times of the day. Originally, we found that only 100 outages could be shown in the app. Tesla made a change and now we see 280 outages since July 2018. Over 230 of them have been in the past 3 weeks.

That's pretty crazy.

Are you in self powered or backup mode? Just thinking if you were able to set your PWs to self powered with 90% reserve then that would give you some headroom during the short outages to charge the PW back up and limit the amount of lost solar.
 
That's pretty crazy.

Are you in self powered or backup mode? Just thinking if you were able to set your PWs to self powered with 90% reserve then that would give you some headroom during the short outages to charge the PW back up and limit the amount of lost solar.
Usually we are in Advanced TBC since we have 4 Powerwalls and try to cover our usage as much as possible vs. taking it from the grid. If there was a scheduled self-powered mode, we'd use that but for now, we stick to TBC-Balanced and set the periods so that we should really only draw from the grid for a couple hours a day. We've tried setting it to self-powered but it still takes a trickle from the grid which is enough to cause the Powerwalls to take over when they sense an issue. The system worked perfectly during the recent 40 hour outage. We actually ran an extension cord to a neighbor's house to help power her refrigerator. While the 40 hour outage is over, she's still experiencing problems with her power and asked if she could keep it plugged in a bit longer. Funny/sad that we can provide more reliable power than the electric company.
 
I've found that the Powerwall can be pretty sensitive to grid events. Just a few minutes ago the power blipped very briefly and while my oven (which is not backed up) was still on and the clock didn't reset the Powerwall switched to backup mode for 5 minutes.
Just note that this "grid only blinked, but Powerwall took over for 5 minutes" experience is normal and expected behavior. In case the grid is in a degraded or unstable up/down state, the Powerwall will monitor for grid stability for 5 minutes after the initial disruption before returning to grid power. If there is further disruption during that window, it resets the timer until things have stabilized for 5 minutes.
 
  • Informative
Reactions: aesculus
Just note that this "grid only blinked, but Powerwall took over for 5 minutes" experience is normal and expected behavior. In case the grid is in a degraded or unstable up/down state, the Powerwall will monitor for grid stability for 5 minutes after the initial disruption before returning to grid power. If there is further disruption during that window, it resets the timer until things have stabilized for 5 minutes.
I'm definitely glad the Powerwalls are taking over are providing consistent power. There's obviously an issue with the power from the grid since the Powerwalls are having to take over 30 times a day.
 
  • Like
Reactions: Cirrus MS100D
I'm definitely glad the Powerwalls are taking over are providing consistent power. There's obviously an issue with the power from the grid since the Powerwalls are having to take over 30 times a day.


Totally justifies your setup, though (not like you needed justification before though.... lol). You could probably sell your entire neighborhood on powerwalls + solar if you havent already, because its highly likely your neighbors on the same transformer are impacted by all these power outages, while you merrily carry on.
 
Last fall at the start of an outage, my Powerwall system raised the frequency to shut down my inverters because the batteries were full (PG&E PSPS event, Stormwatch enabled, batteries full). But there was no solar output to curtail because it was the middle of the night! :mad:

Bruce.
to raise the frequency, the gateway would have had to cut the grid after charging. that would be strange.
 
Totally justifies your setup, though (not like you needed justification before though.... lol). You could probably sell your entire neighborhood on powerwalls + solar if you havent already, because its highly likely your neighbors on the same transformer are impacted by all these power outages, while you merrily carry on.
Exactly! When we had our system expanded at the end of 2019, Tesla pointed out that we'd almost be powering the entire neighborhood (13 houses) on sunny days. My neighbor (that we ran the extension cord to) told me she wants to learn more about solar. She's tired of frequent outages but I don't know how often she is experiencing them. They had mentioned getting a generator and we're trying to steer them away from that as our other neighbor has a generator that runs nonstop and is loud. Another neighbor a few houses down has 10 kW of solar but doesn't have any battery backup so it's worthless when the grid is down. I don't think most of the neighbors notice the 30 "outages" a day as they probably don't have monitoring equipment and things just work for the most part.

Anyhow, we're pleased Tesla was able to fix our UPS frequency issue but are annoyed that it revealed our utility's "dirty power" issue which has yet to be fixed.
 
A few pages back I documented the problem I was having during an early morning power failure. Although there was minimal solar output, the PW2 frequency went to 65 Hz and stayied there for a long time even when the SOC was at only 60% SOC. Beeping UPSs and blinking LEDs were the symptom that was objectionable. We (this forum) attributed it to the fact that it was cold, 34° F, and the battery wanted to warm up before it would accept additional charge. That prompted me to call Tesla Support and have them lower the cutoff frequency to 62 Hz. Level 1 support would not connect me to Level 2 and insisted that my installer send them details about my inverters and they would deal with him, not me. I offered to send that material, but no, they had to hear from my installer before they would take this to Level 2 support.

A few weeks later, my installed called and said the Level 2 needed to know why I wanted the upper cutoff frequency lowered. What were the symptoms I was seeing? I was a bit dumbfounded as I thought that this problem was well understood, but I gave him the beeping, blinking and motors not running story and he relayed that to them. He called back a bit later and said that they would lower to 62.5 Hz, but not below.

Today I got around to testing the system starting in the morning with 49% SOC and the solar putting out about 5.5 kW which should be no problem for my 2 PW2s to accept. I threw the main breaker to the grid and was happy to see that both solar inverters remained on and the PW2 continued charging. The line frequency at that point was 59.7. The frequency slowly increased with the SOC and at 94% SOC the frequency was 60.2. At 96% SOC the frequency was at 60.4 and both inverters shut down. So far so good, but although the solar was shut down the frequency continued to increase until it reached 62.4 where it held. Unfortunately, at this point my LED lights were flashing. The UPSs were OK, but the LED lights (or their Insteon dimmers) were not. When the SOC dropped to 95% the frequency started falling, but it was not until 94% SOC and 60.8Hz that the lights stopped blinking.

Is there any point is asking my installer to ask Tesla to lower my frequency to 60.5 or 61 Hz? Will they do it? Or should I hope for a more intelligent frequency control in the future that either learns the inverters, or simply notices that there is no solar and no need for additional increases in frequency? Obviously given the posts above, I was not surprised the PW2 ignored the solar output, or lack of it, and raised the frequency to well above what was needed to shut down the inverters. The disappointment was that my LED lights were not OK with my new 62.5 Hz cutoff frequency.
 
  • Informative
Reactions: Cirrus MS100D
A few pages back I documented the problem I was having during an early morning power failure. Although there was minimal solar output, the PW2 frequency went to 65 Hz and stayied there for a long time even when the SOC was at only 60% SOC. Beeping UPSs and blinking LEDs were the symptom that was objectionable. We (this forum) attributed it to the fact that it was cold, 34° F, and the battery wanted to warm up before it would accept additional charge. That prompted me to call Tesla Support and have them lower the cutoff frequency to 62 Hz. Level 1 support would not connect me to Level 2 and insisted that my installer send them details about my inverters and they would deal with him, not me. I offered to send that material, but no, they had to hear from my installer before they would take this to Level 2 support.

A few weeks later, my installed called and said the Level 2 needed to know why I wanted the upper cutoff frequency lowered. What were the symptoms I was seeing? I was a bit dumbfounded as I thought that this problem was well understood, but I gave him the beeping, blinking and motors not running story and he relayed that to them. He called back a bit later and said that they would lower to 62.5 Hz, but not below.

Today I got around to testing the system starting in the morning with 49% SOC and the solar putting out about 5.5 kW which should be no problem for my 2 PW2s to accept. I threw the main breaker to the grid and was happy to see that both solar inverters remained on and the PW2 continued charging. The line frequency at that point was 59.7. The frequency slowly increased with the SOC and at 94% SOC the frequency was 60.2. At 96% SOC the frequency was at 60.4 and both inverters shut down. So far so good, but although the solar was shut down the frequency continued to increase until it reached 62.4 where it held. Unfortunately, at this point my LED lights were flashing. The UPSs were OK, but the LED lights (or their Insteon dimmers) were not. When the SOC dropped to 95% the frequency started falling, but it was not until 94% SOC and 60.8Hz that the lights stopped blinking.

Is there any point is asking my installer to ask Tesla to lower my frequency to 60.5 or 61 Hz? Will they do it? Or should I hope for a more intelligent frequency control in the future that either learns the inverters, or simply notices that there is no solar and no need for additional increases in frequency? Obviously given the posts above, I was not surprised the PW2 ignored the solar output, or lack of it, and raised the frequency to well above what was needed to shut down the inverters. The disappointment was that my LED lights were not OK with my new 62.5 Hz cutoff frequency.

They seem very hesitant to lower it below 62. For example, my system is a solar city PV ( so now tesla ), as well as tesla powerwalls, so they have ALL specs of my inverters etc. Through this forum, some people helped me decipher the manual for my inverter and we were able to see that the cutoff is 60.5 for mine. It doesnt ramp either. Even though they installed it, and know the specs, they set my cutoff to 62hz.

I think you have a leg to stand on to ask to be set to 62hz instead of 62.5, because your LED lights are not functioning properly at 62.5. I doubt they go lower though.

I know some have gotten them to set it lower back when tesla was just figuring this out, but right now It appears they are holding the line on 62, so thats what I would ask for if I was in your shoes.
 
They seem very hesitant to lower it below 62. For example, my system is a solar city PV ( so now tesla ), as well as tesla powerwalls, so they have ALL specs of my inverters etc. Through this forum, some people helped me decipher the manual for my inverter and we were able to see that the cutoff is 60.5 for mine. It doesnt ramp either. Even though they installed it, and know the specs, they set my cutoff to 62hz.

I think you have a leg to stand on to ask to be set to 62hz instead of 62.5, because your LED lights are not functioning properly at 62.5. I doubt they go lower though.

I know some have gotten them to set it lower back when tesla was just figuring this out, but right now It appears they are holding the line on 62, so thats what I would ask for if I was in your shoes.

Thanks for the quick reply and sharing your experience. Unfortunately based on today's testing the LEDs blinked if the frequency was above 61 Hz, so I am not sure that 62 Hz is much better than 62.5 Hz. I am not sure whether the motors on my refrigerator might do better, a subject perhaps of more testing. Of course, if my Model 3 was at home during the test, it would have taken the excess solar production and used it for charging – a really cool feature, but not one that works if your wife has it at work.
 
Thanks for the quick reply and sharing your experience. Unfortunately based on today's testing the LEDs blinked if the frequency was above 61 Hz, so I am not sure that 62 Hz is much better than 62.5 Hz. I am not sure whether the motors on my refrigerator might do better, a subject perhaps of more testing. Of course, if my Model 3 was at home during the test, it would have taken the excess solar production and used it for charging – a really cool feature, but not one that works if your wife has it at work.

Are these a specific brand of LED light? And do they all have issues with 62 Hz? All of the recessed lights (40 or so) in my house are LED.
 
Are these a specific brand of LED light? And do they all have issues with 62 Hz? All of the recessed lights (40 or so) in my house are LED.

I have 37 recessed lights in my home that are also all LED (led floodlight style) that I bought from costco).

OP, you touched on this but since the LED lights seem to be the thing thats left that is troubling you, any chance its the dimmers and not the bulb? Have you tried the lights with the dimmers turned all the way to the ON position, or (harder) take one that is flickering at the 62.5hz and remove the dimmer and put on a regular switch and see if that same light now is solid?

More work, but replacing a light switch is something that is not too bad, and within the realm most of us are able to do, I would think.

I would still pursue getting reduced to 62 hz though, even if you dont think it makes much difference. Lower is better imo (with "lower" meaning as close to the cutoff of the inverter they will let you go).
 
My LED lights are a variety of brands. I agree that it might be my Insteon dimmers not the lights, as I mentioned above, but replacing those would be a real pain as they are part of a home automation system, etc. The "bulbs" range from quantum dot LEDs that I bought 10 years ago to some Phillips under counter lights fixtures and a rope light for inside the cabinets that are much more recent.