62.8 Hz at 92%?????

[jjrandorin]

Disconnected grid at 82% today. Output went to 60.6 Hz and solar shut off so I'm on battery only and draining at 82% and falling.

I dont think any of this testing you are doing is going to change the fact that, if your powerwalls are full and there is a power outage, they are definitely going to put out a frequency higher than "60.5". In fact, unless you contacted tesla to get them to lower the frequency (which isnt clear to me if you did that or not), they will likely blast out 65 hz.

if you did contact tesla to lower the frequency, the only one I know of on these boards that got them to lower it lower than 62 hz is @gpez, who got them to do it before they somewhat "streamlined" the requests to their tier 2 support to lower the frequency. Most report tesla setting it (after requests to do so) at between 62 and 62.5 hz.

If you didnt make a request for them to lower the max frequency, you likely should start there, because of what @BrettS said. The scenarios you are testing of slowly rising up and having it slowly increase the frequency doesnt mean that is what will happen if the batteries are 97%+.

Of course, if you test when your batteries are full, you are running the risk of causing issues with all your home automation devices etc, as you already experienced, and possibly shorting out some.

 

[sorka]

Went down to the top of 81% Hz went to 59.7 and solar came back. Still at 59.7 at 83%.

 

[sorka]

I dont think any of this testing you are doing is going to change the fact that, if your powerwalls are full and there is a power outage, they are definitely going to put out a frequency higher than "60.5". In fact, unless you contacted tesla to get them to lower the frequency (which isnt clear to me if you did that or not), they will likely blast out 65 hz.

When I tested this last week, the frequency didn't start to rise until the SOC hit 98%. Went it hit 60.4 Hz half way through 98%, the inverter shut off.

This is the behavior I expect which works causes the inverter to shut off and doesn't ruin any of my equipment. 60.4 or 60.5 Hz (at the most) is more than sufficient to kill the inverter.

 

[sorka]

Of course, if you test when your batteries are full, you are running the risk of causing issues with all your home automation devices etc, as you already experienced, and possibly shorting out some.

...which means that a power outage can blow out my devices the same as if I'm testing it intentionally.

 

[MP3Mike]

What was your power flow showing when you shut the main power off? Were you exporting power?

 

[sorka]

What was your power flow showing when you shut the main power off? Were you exporting power?

None of the tests have been with exporting power to the grid.

It's overcast today with no visible sun and I'm only making 2.9 KW currently so it's a very slow climb.

 

[MP3Mike]

Hmm.. I was just thinking that maybe the Powerwalls were too cold to accept all of the excess energy at the time, and they started heating to accept more after you shut off the main power... But it probably isn't that cold where you are.

On second thought it looks like it was ~38 overnight, so if the Powerwalls are outside temperature could be an issue.

 

[sorka]

Hmm.. I was just thinking that maybe the Powerwalls were too cold to accept all of the excess energy at the time, and they started heating to accept more after you shut off the main power... But it probably isn't that cold where you are. On second thought it looks like it was ~38 overnight, so if the Powerwalls are outside temperature could be an issue.

That's possible although several days ago when it shot up to 62.8 Hz after shutting the grid at just after 1 pm the PWs had been charging at at 5 KW or so for the last few hours and should have been plenty warm by then.

I have no issue with the PW deciding it needs to protect itself for whatever reason (assuming it's short term) and kicking the inverter off, but my inverter shuts off at 60.4 to 60.5 Hz which doesn't really cause a problem with any of my equipment.

 

[jjrandorin]

When I tested this last week, the frequency didn't start to rise until the SOC hit 98%. Went it hit 60.4 Hz half way through 98%, the inverter shut off.

This is the behavior I expect which works causes the inverter to shut off and doesn't ruin any of my equipment. 60.4 or 60.5 Hz (at the most) is more than sufficient to kill the inverter.

That isnt the frequency that will be output (60/4 - 60.5 ) if your powerwalls are at full or close to full (97-100%) when the power goes out though. In any case, you will likely see it one way or another.

 

[sorka]

That isnt the frequency that will be output (60/4 - 60.5 ) if your powerwalls are at full or close to full (97-100%) when the power goes out though. In any case, you will likely see it one way or another.

Not sure I'm following but when I did this test last week, I disconnected the grid at 97% and the output frequency was 59.9 hz. It didn't start to climb until it passed 98% and then the inverter shutoff midway through 98% at 60.4 Hz (60.4 is what I first saw so 30 seconds earlier it might have been 60.5).

 

[BrettS]

Hmm.. I was just thinking that maybe the Powerwalls were too cold to accept all of the excess energy at the time, and they started heating to accept more after you shut off the main power... But it probably isn't that cold where you are.

On second thought it looks like it was ~38 overnight, so if the Powerwalls are outside temperature could be an issue.

You might be on to something here. If it is temperature related then it may immediately jump to the max frequency instead of ramping up like it does with a high SOC. I’m thinking that temperature might be the factor that we were missing here.

 

[jjrandorin]

Not sure I'm following but when I did this test last week, I disconnected the grid at 97% and the output frequency was 59.9 hz. It didn't start to climb until it passed 98% and then the inverter shutoff midway through 98% at 60.4 Hz (60.4 is what I first saw so 30 seconds earlier it might have been 60.5).

Try your test with your powerwalls at 100% and in standby mode, and see what frequency they send to shut off your solar, is what I am saying... but be prepared for it not to be 60.5 or anything like that.

 

[sorka]

Try your test with your powerwalls at 100% and in standby mode, and see what frequency they send to shut off your solar, is what I am saying... but be prepared for it not to be 60.5 or anything like that.

I have no doubt it will be way higher than that at 100%. When I leave the system on full time, is it going to charge my powerwall to 100% if there's enough excess off peak? If it does, I may have to get into micromanagement and continually modify the TOU settings to convince it to dump excess to the grid after it gets to say 95%.

If that becomes a thing, then I'll have to write a watchdog running on my RPI to monitor the SOC and change the TOU schedule when it gets above a certain threshold. This would include turning stormwatch off any time the charge exceeds 95%.

 

[jjrandorin]

I have no doubt it will be way higher than that at 100%. When I leave the system on full time, is it going to charge my powerwall to 100% if there's enough excess off peak? If it does, I may have to get into micromanagement and continually modify the TOU settings to convince it to dump excess to the grid after it gets to say 95%.

If that becomes a thing, then I'll have to write a watchdog running on my RPI to monitor the SOC and change the TOU schedule when it gets above a certain threshold. This would include turning stormwatch off any time the charge exceeds 95%.

I dont use the cost based savings modes with peak and off peak, etc so I am not sure exactly what that mode does with excess solar.

What I think, however, Is that, at least during the spring / summer when you are making a lot of power, your powerwalls will get charged to 100% before peak. At least in self powered mode, the powerwall will charge to 100% if there is excess power. In the TOU modes, I think it charges to 100% from excess solar as long as that is prior to peak time.

In any case, you are going to see your powerwalls at 100% in the spring / summer, because of the amount of power your PV is going to generate, even if you dont get to 100% from excess solar now.

 

[sorka]

I dont use the cost based savings modes with peak and off peak, etc so I am not sure exactly what that mode does with excess solar.

What I think, however, Is that, at least during the spring / summer when you are making a lot of power, your powerwalls will get charged to 100% before peak. At least in self powered mode, the powerwall will charge to 100% if there is excess power. In the TOU modes, I think it charges to 100% from excess solar as long as that is prior to peak time.

In any case, you are going to see your powerwalls at 100% in the spring / summer, because of the amount of power your PV is going to generate, even if you dont get to 100% from excess solar now.

If the solar was on every day (Don't yet have PTO) I'd still be hitting full charge on most days right now since my house is not using drawing from the PWs during off peak so nearly all production is going to the powerwalls since peak doesn't even start until there's almost no sun left currently.

 

[sorka]

So Tesla is here to diagnose the problem first before making a frequency adjustment. Unfortunately I can't reproduce the problem for them at this moment because the fog only just cleared and it will take two hours to charge the PWs up high enough to demonstrate.

They're about to leave without any changes because I couldn't show the problem while they were here. Can't they see all this from the logs from the dates and times I gave them when I did the tests?

 

[sorka]

I asked him before he left if they could just engage stormwatch manually and force the powerwalls to charge up and they said yes, so the tech is waiting around for another 35 minutes which is how long it will take to get them to an SOC where I can prove the problem.

 

[sorka]

He shut the grid off at 87%. Powerwall frequency shot up to 64Hz instantly and the solar inverter shut off. After 5 minutes, the powerwall frequency was STILL at 63Hz. Totally unacceptable. At the very least it should have gone back down to 60 Hz as soon as the solar inverter shut off.

 

[sorka]

Got them to lower maximum frequency to 61% which is an improvement. At least my switches won't go haywire and burn out but communication fails at anything above 60.4Hz. The bigger issue is that at a high SOC above 90%, the output will STAY at 61Hz even when the inverter is off. They said this is intentional to keep the solar inverter from cycling above 90%

 

[sorka]

They left stormwatch charging on so it looks like they're going to charge to 100%