Powerwall 2 + UPS Connundrum - and solution

[JPP]

I think there are two "ramps" that could be confusing.

If the grid disconnects while the Powerwall is fully or close to fully charged then it will set the frequency to the max instantly. No "ramp". I'm not sure I understand the behavior @aesculus mentioned.

If the grid disconnects while the Powerwall has the ability to charge then the frequency will stay below the max and slowly increase to max as the Powerwall charges. Once the frequency is too high for the PV then solar production will stop, the Powerwall will start to drain and the frequency will start to decrease back towards 60hz. This is a "ramp".​

Makes sense--but still an issue. Do the inverters 'understand' the ramp and keep operating if the PWs can 'absorb' the power? Are the inverters supposed to support some kind of 'ramp down'? I thought the inverters just understood 'on' or 'off'.

 

[aesculus]

I was using a Kill A Watt to test but I doubt it has some sort of built in delay.

 

[mspohr]

Makes sense--but still an issue. Do the inverters 'understand' the ramp and keep operating if the PWs can 'absorb' the power? Are the inverters supposed to support some kind of 'ramp down'? I thought the inverters just understood 'on' or 'off'.

Why change the frequency at all as long as the PW needs to charge? Keep it at 60 until they are charged then raise it enough to turn the inverters off.

 

[aesculus]

Why change the frequency at all as long as the PW needs to charge? Keep it at 60 until they are charged then raise it enough to turn the inverters off.

Actually your idea of dynamic would be helpful to those that have different inverters cut off frequencies. It could keep some active for partial charging at a lower rate. But this is getting kind of complex.

 

[gpez]

Why change the frequency at all as long as the PW needs to charge? Keep it at 60 until they are charged then raise it enough to turn the inverters off.

Yes, I have my Enphase IQ 6 inverters configured to start curtailing production at 60.20hz and then will turn off completely at 61.4hz. This is independent of the over frequency limits, which are set to 61.8hz at 1000ms and 65hz at 100ms.

SolarEdge inverters all it "P(F) Power Frequency" (https://www.solaredge.com/sites/default/files/application_note_power_control_configuration.pdf, page 6) and my Enphase IQ6 microinverters call it "curtailment" or "ramp down" (https://enphase.com/sites/default/f...Considerations-AC-Coupling-Micros-Battery.pdf, page 6).

This allows the system to not "see-saw" turning the solar on and off during a grid outage and allows for the PV to dynamically ramp up as the load from the house increases.

 

[jjurroz]

I just had solar panels and 3 PW's installed 2 weeks ago. I have a SolarEdge 10000 inverter.

Today I did a test with the grid disconnected and the PW's at 100%. The PW's shut off the solar as expected but the frequency fluctuated between 60.0 and 60.5 for a few mins and eventually it stabilized at 60.0 while disconnected from grid AND solar off. At some point, the solar came back up and the cycled repeated once the batteries were full again.

I'm on firmware 1.44.4

None of my devices (routers, switches, pc's, UPS (both of my units did not go into battery mode and display showed 60.0-60.5)), internet gateway) complained.

Perhaps Tesla is catching on and setting up the PW's correctly from the get-go. In my case, throughout the whole process, it was known which inverter I was getting. They even provided documentation on the inverter with specs. It makes sense, that if they know specifically which inverter they will be installing that they could configure it beforehand rather than having to call into tier 2 support.

Am I missing something? did I do the test accurately?

 

[aesculus]

Today I did a test with the grid disconnected and the PW's at 100%. The PW's shut off the solar as expected but the frequency fluctuated between 60.0 and 60.5 for a few mins and eventually it stabilized at 60.0 while disconnected from grid AND solar off.

I am not sure how this would work unless there is another technique besides frequency shifting you can shut off the SolarEdge 10000 inverters and the PW system knows this (ie direct connect or wifi). 60.5 Hz could possibly turn them off but then they would constantly be turning on again if it dropped anything below that.

But the real question is what was your power draw at the time? Did the house consume more than you were generating? That would technically allow the inverter to stay on even though the PW was full (or what appeared to be full).

 

[jjurroz]

But the real question is what was your power draw at the time? Did the house consume more than you were generating? That would technically allow the inverter to stay on even though the PW was full (or what appeared to be full).

The house was only consuming 0.5 KW at the time and the solar was producing over 7kw. Because the grid was down and the batteries were full, the PW had to shutdown the inverters since the power had no where to go.

I still want to perform 1 more test:

- Battery full and in full backup mode so that no power from the solar goes to it.
- House operating from Solar
- Shut off the grid and see how quickly the batteries respond and how my equipment and UPS react.

 

[NathanielHrnblwr]

I called to have this adjusted on my newly installed system. I received an email two days later that they're going to schedule a service visit. Apparently the tier 1 agent did not comprehend the request.

 

[aesculus]

I called to have this adjusted on my newly installed system. I received an email two days later that they're going to schedule a service visit. Apparently the tier 1 agent did not comprehend the request.

I think this is normal. I was contact by Tier 2 with the same email and am awaiting the "service" too for a week now.

 

[NathanielHrnblwr]

I think this is normal. I was contact by Tier 2 with the same email and am awaiting the "service" too for a week now.

Strange. This thread suggested it was a remote resolution. I guess if they want to send someone out, great.

 

[jjrandorin]

Strange. This thread suggested it was a remote resolution. I guess if they want to send someone out, great.

It was definitely remote in my case, no truck roll. I said up thread what I specifically told them. I have managed people in a call center before, so I know that what you tell the first line call center employee matters in getting the result you want. Sometimes luck of the draw on employees knowledge, but nothing you can do about that other than call in a second time if you get a result you dont want from the first call.

 

[bob_p]

We had a 3rd party Tesla-approved installer; I contacted Tesla directly by phone to report the 65Hz problem.

Prior to the call, I performed a test to verify the frequency was too high (fully charging PowerWalls, throwing breaking the Tesla Gateway panel to go off grid, and then observing several UPS boxes going offline, plus issues with pumps/air conditioning).

I believe the Tesla rep was able to pull up the log from our TEG and verify the 65Hz cutoff frequency and then referred the issue to "tier 2" support, who resolved it remotely. Before doing that, I was asked to provide the model numbers of our solar panels and microinverters and confirm the specification of their cutoff frequency.

If you can't provide this information, that could require an onsite visit to verify the specific configuration you have so Tesla is confident they are setting a safe cutoff frequency, since they want to be sure the solar power is cutoff when operating off grid and the PowerWalls are full.

 

[aesculus]

Strange. This thread suggested it was a remote resolution. I guess if they want to send someone out, great.

I think the messaging is sort of standard that can apply for a real visit or a virtual one. I really think it was more "we will get back to you when we want to deal with this".

 

[NathanielHrnblwr]

I think the messaging is sort of standard that can apply for a real visit or a virtual one. I really think it was more "we will get back to you when we want to deal with this".

"Our team will want to be on site ..."

Again, I don't care if they want to waste time and come onsite, but I'm sure they have better things that resource could be doing.

 

[jjrandorin]

"Our team will want to be on site ..."

Again, I don't care if they want to waste time and come onsite, but I'm sure they have better things that resource could be doing.

If they think the "team needs to be onsite" then they didnt understand the question / issue. There is nothing that an "onsite" team can do other than read the labels of your solar, at least as I understand it, regarding this issue.

 

[bob_p]

While I doubt they would do this, the other thing a team could do onsite would be to verify that the revised cutoff frequency actually turns off the solar panel power when the PowerWalls are fully charged and that the lower frequency resolves problems with UPS and other devices - something I did with our system after Tesla made the setting change.

 

[jjrandorin]

While I doubt they would do this, the other thing a team could do onsite would be to verify that the revised cutoff frequency actually turns off the solar panel power when the PowerWalls are fully charged and that the lower frequency resolves problems with UPS and other devices - something I did with our system after Tesla made the setting change.

That would make sense if they did it for everyone, or as a requirement for making this change, but in your case (with third party installer) and my case (tesla installed powerwalls and solar city installed solar from 2015), they did not need to "roll a truck" to verify the solar turns off.

I performed the test myself like you did, because they never said what frequency they set mine at in the email when they said "we have completed this work" (mine was set to 62).

 

[bmah]

That would make sense if they did it for everyone, or as a requirement for making this change, but in your case (with third party installer) and my case (tesla installed powerwalls and solar city installed solar from 2015), they did not need to "roll a truck" to verify the solar turns off.

I performed the test myself like you did, because they never said what frequency they set mine at in the email when they said "we have completed this work" (mine was set to 62).

Right. (Also, given current events, they may want to think harder about when they really need to send service people out.)

So I had not one but two service visits in my case (Tesla installed Powerwalls and Solar City PV). In the first one, we verified the symptoms of the problem and consulted with the inverter manufacturer (Fronius) to find the cutoff frequency. This could easily have been done by a back-end person, but I think this problem wasn't well understood at the time. The frequency change was made completely on-line, and I traded a few emails with Tesla Powerwall support over this. They then sent the same service technician out again, but by the time of the service appointment, I had already verified that the frequency parameters were working as desired, so when he got to my house, I basically told him "it works", we chatted a bit, and he was on his way.

I suppose if the homeowner didn't know how to do their own off-grid test under the right circumstances to verify the solution, Tesla might want to have someone on-site.

Bruce.

 

[NathanielHrnblwr]

Good news! They’ll be here in a month.

Not that this is an urgent issue, just a bit silly.