Just completed a 16 hour grid outage simulation. I'm very pleased.

[abasile]

Is it possible your solar is connected outside the backup gateway, and had no power when the main breaker was open? Have you ever seen the solar work with the grid down?

No, our solar breakers are inside the backed-up subpanel. We’ve never seen the solar work with the grid down, and I’ve been on the phone with Tesla a number of times and had a technician come out. We were told to wait for a fix in the software, hence my interest in trying the new 1.35.2 release.

 

[NinjaVece]

No, our solar breakers are inside the backed-up subpanel. We’ve never seen the solar work with the grid down, and I’ve been on the phone with Tesla a number of times and had a technician come out. We were told to wait for a fix in the software, hence my interest in trying the new 1.35.2 release.

Weird. I wonder what Tesla could fix in the PW2 firmware to keep your inverter going. I would expect the inverter(s) to only be sensitive to Hertz levels. Maybe Tesla is working to clean up the sine wave coming from the PW2 inverter? Maybe that is inhibiting your inverters when the house circuit is islanded...

 

[abasile]

Weird. I wonder what Tesla could fix in the PW2 firmware to keep your inverter going. I would expect the inverter(s) to only be sensitive to Hertz levels. Maybe Tesla is working to clean up the sine wave coming from the PW2 inverter? Maybe that is inhibiting your inverters when the house circuit is islanded...

I'm assuming the frequency is still a little high. I may force the Gateway onto our Wifi (it's been using LTE) so that I can see what line frequency it's reporting.

Unless Tesla provides a fix, I'm going to try roughly the same test again when the weather is warm, since our Powerwalls are outdoors and there may be a temperature factor here. Of course, during outages, charging from solar needs to work in all temperatures. Winter is the worst time of year to lose power and the Powerwalls need to be able to recharge themselves.

 

[NinjaVece]

I'm assuming the frequency is still a little high. I may force the Gateway onto our Wifi (it's been using LTE) so that I can see what line frequency it's reporting.

Unless Tesla provides a fix, I'm going to try roughly the same test again when the weather is warm, since our Powerwalls are outdoors and there may be a temperature factor here. Of course, during outages, charging from solar needs to work in all temperatures. Winter is the worst time of year to lose power and the Powerwalls need to be able to recharge themselves.

Once you connect your PW2 to your WiFi, you can find the IP address on your router by looking for the MAC address of: 88: DA: 1A: XX: XX: XX - the first three bytes are the specification for a Redpine Systems communications chip which the Gateway uses for WiFi, etc. Simply in your router make that MAC address static to a specific IP address and you will always be able to find it no matter if you need to reboot the router over time.

Then do your API calls with confidence...

 

[wyldtek]

Wow!! Sounds like my (possible) system. They quoted me with 6.62 KW and one PW
although I asked for two. I have some morning shade.

Don't let them talk you out of getting two. The sales guy convinced me to just get one PW instead of two and now I regret it. Especially since the price pf PWs have gone up.

 

[gnumeric]

Don't let them talk you out of getting two. The sales guy convinced me to just get one PW instead of two and now I regret it. Especially since the price pf PWs have gone up.

Yes, I got 3 powerwalls and can’t imagine anything less. My sales guy was recommending 2 but I’m glad I pushed for 3.

My home is pretty efficient, we barely run the a/c, and I still feel like my powerwalls drain pretty quickly. Also remember that each Powerwall can only sustain 5kW output. So if you plan on running anything substantial (laundry dryer, a/c, electric oven, etc.) off of your powerwalls, 1 is not going to cut it.

 

[nyprepper]

oh hell yeah.... I got 2 and wish I got 4.

 

[aekemmerer]

I had a late morning power failure several months ago and my two PWs never kicked in despite being charged to 100% and ample solar production at the time. Tesla told me their data said there was too much demand and the PWs couldn’t handle this. No AC was running and my consumption was only about 1.3 kW when the outage happened. A few weeks later, they had me reboot the PWs and then test them with shutting off the main. Sure enough it worked then. To this day, I have no idea if this was a firmware issue or some other issue that required a reboot. Perhaps doing a monthly reboot (power off PWs, turn off breakers for the PWs, wait about 5 minutes and turn back on) is a good practice?

 

[Frankman60]

This is something I have been wondering about too. Do the PWs need to be periodically rebooted like the screens on my Model S? Do they automatically reboot periodically?

I had a late morning power failure several months ago and my two PWs never kicked in despite being charged to 100% and ample solar production at the time. Tesla told me their data said there was too much demand and the PWs couldn’t handle this. No AC was running and my consumption was only about 1.3 kW when the outage happened. A few weeks later, they had me reboot the PWs and then test them with shutting off the main. Sure enough it worked then. To this day, I have no idea if this was a firmware issue or some other issue that required a reboot. Perhaps doing a monthly reboot (power off PWs, turn off breakers for the PWs, wait about 5 minutes and turn back on) is a good practice?

 

[NinjaVece]

This is something I have been wondering about too. Do the PWs need to be periodically rebooted like the screens on my Model S? Do they automatically reboot periodically?

At times my PW2 Gateway will lose WiFi connection and I can wait for many hours for it to find it again or simply reboot the Gateway and get it back in a couple of minutes...

 

[gpez]

Excellent! I have a 10.5kW system with 2 powerwalls. My problem is when we have experience an outage (we have had 3) all of the arc fault interrupter (AFI) circuit breakers trip and none of the other breakers trip. Tesla Power has been working on a solution since September 2018. They admit there is a problem. I cannot replace the AFI's with non-AFI's due to the local electrical codes. I suspect (being an engineer) that the the quality of electricity (amplitude, noise, etc) from the inverter (DC to AC) is not withing the requirements for the AFI's. If I was not retired I would get an oscilloscope form work and record the grid data and then simulate an outage and record the powerwall/inverter data.

I have the same problem and an open ticket with Tesla. My AFCI and combo AFCI+GFCI breakers as well as some GFCI outlets in my garage have tripped when the backup gateway switched over. A T2 rep reached out to me over the weekend asking what breakers I have in the panels, they're all Eatons. @APD1080 any chance you have Eaton breakers too?

 

[Jedi2155]

What would be better is if you left your powerwalls at 50-70% by the end of the night, so the solar can charge the powerwalls during the day.

At night your usage is pretty low, the real test is if it can handle the loads of your home during the day with the solar supporting.

 

[gpez]

@miimura @APD1080 - My AFCI/GFCI tripping issues were fixed when Tesla updated the frequency settings during off-grid operation. My understanding is some breakers are more sensitive to high frequencies (mine are Eatons). Once Tesla changed the off grid frequency to 62.5hz from 66hz my breakers stopped tripping.

My grid outage frequency issue is resolved!

 

[JES2]

• 8pm: PW's at 99% capacity
• 8pm: manually switched off the main (grid) to house, grid-down notification received
• 8pm-7:30am: house runs on PW's
• 7:30am: PW's at 80% capacity
• 7:30am: solar comes on-line and begins charging PW's while PW's power the house
• 9am: solar production increases to the point that it is both powering the house and charging PW's
• 10:30am: PW's reach 98% capacity
• 10:30am: solar shuts off
• 10:30am-11:30am: PW's power the house
• 11:30am: PW drops to 96% capacity
• 11:30am: Solar comes back on-line and begins powering the house and charging PW's
• 11:35am: manually switch main to the house back on
• 11:42am: solar comes back on-line and system resumes normal operations with excess power going to the grid
• 11:42am: grid-up notification received

Overall, I couldn't be happier. System performing exactly as expected. Running only essentials in the house, I think I could power the house indefinitely in an outage assuming nominal solar production. This is a 6.6kw array with two PW's.

I'm currently running a "real life" outage simulation.

Dallas, Texas had some pretty severe storms and over 400,000 are without power. Power went out back at 1 pm, but I was not aware. The solar panels (8.33 kw) and PowerWalls (2) took over without even a flicker. I was at 100% of Powerwall capacity when it hit and luckily the storm blew by quickly.

At 8 pm, the Powerwalls were still at 98%. The temps in Dallas are still hovering around 80 degrees, and at least once both of my A/C units were running while I was cooking on an electric stove. Total power being drained was about 8 kwh for five minutes or so.

Currently at 11:00 pm (9 hours down, so far) and sitting at 63% on the two Powerwalls. A/C running about every 15 minutes, three freezers, and the rest of the house runs at a stream of 500 wh.

This should carry me through the morning for when the solar should start shining again. Oncor is telling us that the estimate recovery of power will be around 5:30 pm tomorrow. (It won't surprise me to see that pushed out.)

Honestly, this is the situation that I actually got the solar panels for (along with the Powerwalls later on). Thank you, Tesla.

 

[JES2]

I'm currently running a "real life" outage simulation.

Dallas, Texas had some pretty severe storms and over 400,000 are without power. Power went out back at 1 pm, but I was not aware. The solar panels (8.33 kw) and PowerWalls (2) took over without even a flicker. I was at 100% of Powerwall capacity when it hit and luckily the storm blew by quickly.

At 8 pm, the Powerwalls were still at 98%. The temps in Dallas are still hovering around 80 degrees, and at least once both of my A/C units were running while I was cooking on an electric stove. Total power being drained was about 8 kwh for five minutes or so.

Currently at 11:00 pm (9 hours down, so far) and sitting at 63% on the two Powerwalls. A/C running about every 15 minutes, three freezers, and the rest of the house runs at a stream of 500 wh.

This should carry me through the morning for when the solar should start shining again. Oncor is telling us that the estimate recovery of power will be around 5:30 pm tomorrow. (It won't surprise me to see that pushed out.)

Honestly, this is the situation that I actually got the solar panels for (along with the Powerwalls later on). Thank you, Tesla.

Oncor (our power distributor) has pushed the time for recovery back to Tuesday afternoon now. Interestingly, the power has been restored in our neighborhood, but they are diverting it to critical locations. They can do this now with everyone having smart meters. So, businesses get power and public offices get power, but many homeowners do not. (I get that process, but you can see how it can easily be subverted at times.)

From the Powerwall's perspective, the grid is now operational. (There is no longer an "X" across the grid.) However, no energy would flow to our house, since it is diverted. I do believe that once I fill up my Powerwall, then the excess electricity will be allowed to flow to the grid. I'll know that in a few hours or so. It is a clear day, so the PWs should fill up relatively quickly.

This will also give me a chance to check out charging the Tesla Model 3 with the two PWs and the solar, (but no grid power). I'll make sure the A/C units are off. Curious to see what the Tesla Model 3 pulls in that condition.

I ended up with 32% of the Powerwalls at 7 am. Pretty easy making it through the night. Over 18 hours without power from the grid.

 

[abasile]

Unless Tesla provides a fix, I'm going to try roughly the same test again when the weather is warm, since our Powerwalls are outdoors and there may be a temperature factor here. Of course, during outages, charging from solar needs to work in all temperatures. Winter is the worst time of year to lose power and the Powerwalls need to be able to recharge themselves.

On June 4 in mid-morning, with an ambient temperature around 20C (68F) and strong sunshine, I tried a "warm weather" outage test by flipping the main breaker. This time, within a couple of minutes or so, our solar array went right back to producing! I believe the software version was still 1.35.2.

Today, in the cool of the morning just past 7AM, with the temperature at about 13C (55F) and some sun, on software version 1.37.1, I tried another outage test. Again, successful! This is looking promising. I hope to try another test when the morning temperature is below 10C (50F), which is not at all unusual for us in June before the summer monsoonal moisture arrives.

EDIT: In my June 4 test, the SOC (state of charge) was about 83%, and it was 62% in this morning's test.

 

[Frankman60]

I love reading how well your PWs are working and am interested to see what the PWs do during the time that the power is “diverted”. I did not know that “diversion” of power is a thing.

Oncor (our power distributor) has pushed the time for recovery back to Tuesday afternoon now. Interestingly, the power has been restored in our neighborhood, but they are diverting it to critical locations. They can do this now with everyone having smart meters. So, businesses get power and public offices get power, but many homeowners do not. (I get that process, but you can see how it can easily be subverted at times.)

From the Powerwall's perspective, the grid is now operational. (There is no longer an "X" across the grid.) However, no energy would flow to our house, since it is diverted. I do believe that once I fill up my Powerwall, then the excess electricity will be allowed to flow to the grid. I'll know that in a few hours or so. It is a clear day, so the PWs should fill up relatively quickly.

This will also give me a chance to check out charging the Tesla Model 3 with the two PWs and the solar, (but no grid power). I'll make sure the A/C units are off. Curious to see what the Tesla Model 3 pulls in that condition.

I ended up with 32% of the Powerwalls at 7 am. Pretty easy making it through the night. Over 18 hours without power from the grid.

 

[APD1080]

The Arc Fault breakers should only trip at the moment the power goes out and the Backup Gateway's switch opens. I have a similar problem with my APC ATS when it switches between sources.

If you really want to take power out of an EV, you can use a grid-tied inverter connected to the 12V system to inject about 1,000W into your house to support the Powerwall system. This will act as a runtime extender. I did a proof of concept with a cheap 600W inverter and I still have it for disaster use.

Powerwall 2.0 Backup Runtime Extender

Thanks for the links and info. It would seem that Tesla could package this together as a part of a COMPLETE solar/EV home solution. I'm not sure I want to tackle this at this time but I will definitely keep the info for future planning.

 

[JES2]

I love reading how well your PWs are working and am interested to see what the PWs do during the time that the power is “diverted”. I did not know that “diversion” of power is a thing.

Oncor was able to turn on power in the late a.m. to our neighborhood, so my PWs were not fully at 100%. I was not able to see if they would accept my excess power (even while I was not receiving power.) I'm also curious what happens when the grid is "down", but my solar panels are producing excess energy while my PWs are at 100%. How do they eliminate the excess solar energy?

 

[gpez]

I love reading how well your PWs are working and am interested to see what the PWs do during the time that the power is “diverted”. I did not know that “diversion” of power is a thing.

I believe @JES2 meant that when the grid came back online the excess solar power being produced by the panels was being stored ("diverted") by the Powerwall and not out to the grid. The grid does not do any "diversion" - either it's on or off.

Oncor was able to turn on power in the late a.m. to our neighborhood, so my PWs were not fully at 100%. I was not able to see if they would accept my excess power (even while I was not receiving power.) I'm also curious what happens when the grid is "down", but my solar panels are producing excess energy while my PWs are at 100%. How do they eliminate the excess solar energy?

From my experience during grid outage simulation tests is that the Powerwall triggers the solar inverters to turn off when the battery is sufficiently full (>95%) by increasing the frequency of the power. What I saw with my microinverter set up is that this behavior can actually turn some, but not all, of the inverters off. While not perfect it does seem to keep the solar production close to in line with the house consumption when the grid is unavailable to sell power through, keeping the battery use minimal while the sun is shining. I have some details in my write up here My grid outage frequency issue is resolved!.