aesculus
Still Trying to Figure This All Out
Well I called Fronius today, an Austrian company, and other than the big hold time (US Tech Support) they answered the question quickly for me.
Both the Fronius Galvo and Primos trip at 59.5 to 60.5 Hz.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Off-Grid Test with fully charged PowerWalls
- Thread starter bob_p
- Start date
Both the Fronius Galvo and Primos trip at 59.5 to 60.5 Hz.
Tesla changed our microinverter shutdown frequency to 61.5 Hz.
I ran an off-grid/full PW test and verified the lower frequency correctly shutdown the microinverters.
All but one of our UPS systems operated through the grid/off-grid/grid switchover.
We have a CyberPower 1500AVR that turns off whenever it sees a change between grid and PowerWall power. I've made some setting changes to that UPS, and if it doesn't work correctly, will likely take it out of service.
While operating off-grid, I found that only 2 of our 3 air conditioners would spin up. The 3rd air conditioner would start to run and immediately turn off. Even with 4 PowerWalls fully charged, there evidently isn't enough power to handle the surge for a 3rd air conditioner - and will work with our installers or HVAC service to add "sure start" to all 3 units. We had hoped that with 4 PowerWalls able to provide a sustained 20 KW, we could avoid that.
Yesterday, we had two brief unplanned power outages. And didn't realize we were having grid power issues until the second one - because the system worked as expected, and switched over to PowerWall power automatically.
I ran an off-grid/full PW test and verified the lower frequency correctly shutdown the microinverters.
All but one of our UPS systems operated through the grid/off-grid/grid switchover.
We have a CyberPower 1500AVR that turns off whenever it sees a change between grid and PowerWall power. I've made some setting changes to that UPS, and if it doesn't work correctly, will likely take it out of service.
While operating off-grid, I found that only 2 of our 3 air conditioners would spin up. The 3rd air conditioner would start to run and immediately turn off. Even with 4 PowerWalls fully charged, there evidently isn't enough power to handle the surge for a 3rd air conditioner - and will work with our installers or HVAC service to add "sure start" to all 3 units. We had hoped that with 4 PowerWalls able to provide a sustained 20 KW, we could avoid that.
Yesterday, we had two brief unplanned power outages. And didn't realize we were having grid power issues until the second one - because the system worked as expected, and switched over to PowerWall power automatically.
I had been waiting to call tesla to get the frequency adjusted till I was home during a weekday, when the batteries were 100% charged, and I was making good solar. I know tesla support is there on the weekend, but I figured there would be a better chance of getting a person who understood the issue during the weekday.
Anyway, I was finally home yesterday (wednesday) and my batteries are now filling by about 12:30 - 1pm when the sun is out, so I called tesla to ask them to lower the frequency. A nice support agent answered the phone, and asked if they could help me, and I said:
"Hello, I am hoping you can help me with a very specific issue. I am hoping you have heard of the issue. When my batteries are full, and I am disconnected from the grid, the powerwalls put out a frequency that is too high for my battery backup devices and some other devices in my home". I would like to get the frequency lowered to a lower frequency that still shuts off my solar. I believe my solar will shut off at 61 Hz.
The agent said "Oh do you have enphase inverters?" I said no, but it was clear the agent had heard of the issue, at least as it relates to those specific inverters. Then the agent asked me "what happens, exactly, in your home, in the situation you describe?"
I answered " My APC ups devices go into backup mode and and drain their batteries instead of being fed from powerwall power, because the frequency is too high". She then asked me "do your lights flicker, too"? I said "yes they do".
Then she asked me "when was the last time this happened?" I told her the last time I tested it, and she put me on hold to look in the logs for it. She couldnt find the frequency in the logs, so I asked her if we could test it again right now. I went and flipped the main breaker, taking myself off grid. Powerwalls kicked in, and shut the solar down with the frequency. It took a few minutes for it to show up on her end while she was watching. When she saw it, she said:
"Ok, I see it now. I am grabbing the logs and going to escalate the request for you"
I mention all this for a couple of reasons. One, to thank everyone for posting all this information, and specifically for helping me with my request for information earlier in this thread. Second, I wanted to let anyone else know who wants to call tesla on this, that it appears tesla is requiring logs from the first level support agents before they can escalate this issue to the tier 2 support who apparently performs this work.
If someone is planning on calling tesla to get this adjusted you will need to have actually tested it before hand, by having full batteries, and taking yourself off grid long enough for the powerwall to send the signal to shut down solar. You will also want to note down date and time you performed this test because they will need to go review logs and attach them to your request on their side.
The good news is, the agent on the phone was familiar with the request, and seemed to have a defined path forward with resolving it. Having managed people in a call center before, this tells me that either I got a somewhat more experienced person, and / or, they have gotten enough calls that the knowledge on this is spreading there, and its either somewhat common knowledge there.
The agent I was talking to did not "look up my question in a database" because I know what that sounds like when you are on the phone with a support agent. You ask a question, you get a pause from the person on the other side, and then hear typing while they look it up. There was none of that in this case, she knew enough to know what the symptoms would be.
First, thanks for the information. Very useful and detailed.
Second, did they give you a time frame to implement that frequency change?
Merrill
Merrill
No they didnt give me a timeframe to implement it, she just told me she would escalate the request. I doubt she could actually give me a timeframe because she is not the responsible party for fixing it. Maybe it doesnt bother me, because I know how these things work, since I am an "escalated to" support agent myself.
@Merrill
No idea, but my guess is, a couple of reasons. First, it would be very difficult to actually test this during install, because of the parameters needed. You need to both be making good sun at the time (fairly simple / likely) and also have the powerwalls at 100% (highly unlikely during install). I am sure they can go into some service mode to trigger the powerwalls to see what setting the inverters turn off at, but lots of inverters have different frequencies for this. You also might have some installers who dont understand, etc.
I am not making an excuse, just thinking about why they do it that way. I happen to agree it should be done on install, but there are so many variables etc that I can see why they dont.
charlesj
Active Member
Now I am wondering if this could not be checked at night and grid down, manually, what frequency would the PW supply to the house?
Also, since I have Enphase inverters would they adjust without any testing?
I guess I'll find out after April 7 when my install happens.
Also, since I have Enphase inverters would they adjust without any testing?
I guess I'll find out after April 7 when my install happens.
jrweiss98020
Tessa's Tesla
I don't really understand all this discussion about frequency limits. All I know is that my system switches seamlessly between solar, PW, and grid. For that to happen, the PW and Enphase microinverters must be in phase with the grid at 60Hz. The only time I hear my UPSes switch over is momentarily on occasion when the grid goes down. For all I know, they may be responding to the low voltage before the PWs kick in...
aesculus
Still Trying to Figure This All Out
It's the rare case when your powerwalls are charged to 100% and you are still able to produce solar energy but the grid is down. The solar panels will be outputting power and there will be nowhere for it to go if the load in the house is less than the outputted solar power. This could cause an issue.
So the PWs will put out a frequency that the solar inverter does not like and this will cause them to shut down. Some UPS units are affected by the default frequency Tesla sets (65Hz) for this function and many have found a more compatible frequency that works with both their inverters and their UPSs.
It took Tesla about a week between the first call to energy support and when they pushed the frequency change to our Gateway.
A few days after the first call, they called me back and asked for the manufacturer/model number of our microinverters. Since the manufacturer had already provided me with the cut-off frequency (60. 5 Hz), I was also able to give this to Tesla.
And once the change was made, I did another off-grid test and verified the power was running at 61.5 Hz, with the solar panels cut-off when the PowerWalls were full. [One of our UPS systems can display the power frequency.]
Some equipment doesn't like running at the default 65 Hz frequency, and wouldn't be surprised to see Tesla change the default to a lower number (63 Hz?). Some of our UPS systems wouldn't run at all when the power was at 65 Hz. And equipment with motors (refrigerators, pool pumps, air conditioners) may not work well at the higher frequency.
This really should be a configurable setting for the installers - who should know the cut-off frequency of the solar inverters/microinverters, and set the frequency only high enough to trigger the solar power cutoff.
A few days after the first call, they called me back and asked for the manufacturer/model number of our microinverters. Since the manufacturer had already provided me with the cut-off frequency (60. 5 Hz), I was also able to give this to Tesla.
And once the change was made, I did another off-grid test and verified the power was running at 61.5 Hz, with the solar panels cut-off when the PowerWalls were full. [One of our UPS systems can display the power frequency.]
Some equipment doesn't like running at the default 65 Hz frequency, and wouldn't be surprised to see Tesla change the default to a lower number (63 Hz?). Some of our UPS systems wouldn't run at all when the power was at 65 Hz. And equipment with motors (refrigerators, pool pumps, air conditioners) may not work well at the higher frequency.
This really should be a configurable setting for the installers - who should know the cut-off frequency of the solar inverters/microinverters, and set the frequency only high enough to trigger the solar power cutoff.
I checked on the spec sheet for my UPSes (CyberPower) and it says 60 Hz +/- 3 Hz, i.e 57 to 63 Hz.
BTW, how do you know when Tesla completed this change and you could run you test. Did they send you a message?
Speaking of testing after, what devices is everyone using to measure power frequency? I dont have a voltmeter, but would be looking for a simple device that would measure what frequency the power is on, to verify that this is indeed corrected when they correct it.
1. I and several others use a Kill-a-Watt:
Kill A Watt Meter - Electricity Usage Monitor | P3
2. Some UPSs with front panel displays can show the power line frequency, although the APCs I have will only display frequencies within the limit of their input tolerance (60+/-3 Hz).
3. One can also query the Powerwall gateway API to get the frequency, though that probably doesn't qualify as "simple". Something like this from a UNIX-like box:
% curl --insecure --output - https://teg/api/meters/aggregates
I have used all of these at one time or another.
Bruce.
1. I and several others use a Kill-a-Watt:
Kill A Watt Meter - Electricity Usage Monitor | P3
2. Some UPSs with front panel displays can show the power line frequency, although the APCs I have will only display frequencies within the limit of their input tolerance (60+/-3 Hz).
3. One can also query the Powerwall gateway API to get the frequency, though that probably doesn't qualify as "simple". Something like this from a UNIX-like box:
% curl --insecure --output - https://teg/api/meters/aggregates
I have used all of these at one time or another.
Bruce.
Thanks very much for the quick answer!
A simpler version would be to use a web browser to browse to http://xxx/api/meters/aggregates where xxx = IP address of the Tesla Energy Gateway.
aesculus
Still Trying to Figure This All Out
Well Tesla and or PG&E screwed up my PTO and it's been in the works for over a month. Multiple calls and emails to Tesla has gotten me nowhere.
Today I called to have the frequency set and the customer support rep would not transfer me to tech support because my PTO was not complete and claimed the system was not operating completely and therefore they would not look into any issues.
Today I called to have the frequency set and the customer support rep would not transfer me to tech support because my PTO was not complete and claimed the system was not operating completely and therefore they would not look into any issues.
JPP
Active Member
I have now had 3 calls over 2 weeks with Tesla Powerwall Support on this issue. 2 different tier 1 reps asking for the make & model of my UPS (can't they read the notes they have supposedly entered from the last call?). Even though I ran by own off-grid test, I had to repeat this while on the phone with the tier 1 tech so that they could monitor it (??? what, no logs??). They now have to bump the request to tier 2. More delays. This should really not be so difficult. And FWIW why do all PWs ship out of the box with a default 65 Hz frequency bump? Unnecessary. Complete waste of time. And I continue to hold payment for my project (Tesla Fin Ops knows) until they get this fixed plus another outstanding issue (Neurio not talking to Tesla Servers). Frustrating.
I've been a software developer, development manager, architect, and product manager for many years at Fortune 500 companies you've absolutely heard of. Companies that their bread-and-butter business is building smart products. The number of dumb designs I've seen is astounding. I've been in so many conversations where a simple change ask turns in to months of work because what an outsider thought of as simple was actually way more difficult because it was designed in a totally different way than what the outsider expected to have been.
This experience has lead me to one of the universal truths I (try really hard to) acknowledge which is: "never assume something is designed the way you would have designed it."
Tesla has really smart people and the Powerwall is an incredibly smart product. That does not mean that there haven't been poor design decisions or that those poor design decisions were made accidently. Doesn't excuse crappy product behavior, just explains it.
Circling back to report that I received an email today from Tesla saying that "tier 2 had completed their configuration" and to let them know if I had any more questions. I happened to be working from home today, and its sunny, and my powerwalls were full, and I was generating good solar, so conditions optimal for testing.
Flipped main breaker.
Since powerwalls were at 100% and on standby, I had a noticeable blip. My microwave and electric double wall oven had their clocks reset, and my TV rebooted. Plugged in the device I got to measure Hz (thanks @bmah for pointing me in the correct direction for that). I got this one: https://www.amazon.com/gp/product/B0777H8MS8/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1
It has a lighted display and has several modes, one of which is displaying frequency. Before throwing main breaker, frequency was 59.98 to 60.01. After throwing main breaker, frequency was 61.96 - 62.02 (meaning it was fluctuating between those two values).
Verified that solar was off at the inverters, which displayed "out of sync," Verified with tesla app that grid had an X and solar was off, and powerwalls powering home.
Ran this way for about 15-20 minutes, checked internet, etc.. all good.
Returned to grid by throwing main breaker back. Hz returned to 59.98 to 60.02 by the measuring device above. Took inverters about 5 minutes to return to providing solar which is normal and expected.
Again, want to thank the trail blazers here for determining the issue, the resolution, and how to approach tesla to get this resolved. Really appreciate all of the knowledge sharing, and the people who managed to get tesla to take a look at this in the past.
Time from reporting to resolution for me was Feb 20th (my quoted post above) to Feb 28th. I made only one contact to tesla on this topic, with no follow ups to check on it. As a data point, my solar is solar city (now tesla) so they have my specific equipment list.
