If you look on the app, did the solar shutdown when the Powerwall hit 100% during the power outage? If so, then the frequency increase was how the gateway shut down the solar production.
Powerwall 2 + UPS Connundrum - and solution
- Thread starter NinjaVece
- Start date
-
- Tags
- ups 65 hertz
If you look on the app, did the solar shutdown when the Powerwall hit 100% during the power outage? If so, then the frequency increase was how the gateway shut down the solar production.
Interesting. I see, looking back at my post, that I did say “full” when referring to the Powerwall. However, that was not quite true. It was getting close, somewhere in the 90’s when the power was restored to the neighborhood and the excess could be sent to the grid. But, I did wonder what would happen to the excess if the house took its share and the Powerwall was at 100% while the system was still in Gateway mode. I guess it has to shutdown since there would be no place for the excess to go.
In any case, sorry for the slight hyperbole. I was carried away by how well the system all worked.
In any case, sorry for the slight hyperbole. I was carried away by how well the system all worked.
It is easy to test this. When the PWs are getting close to fully charged turn off the grid. See if everything continues as expected as the PW reaches 100%, solar shuts down, the PW drains, and solar comes back online. At least that is the theory. As soon as Tesla gets back to me with confirmation they lowered the inverter cutoff frequency this is the test I am going to run.
Is this frequency shift to turn off the solar panel inverters the same on the Gateway 2 and Gateway 1?
I find it ridiculous that the gateway can't communicate in some other way with the inverters. All installations require a shut off breaker. Why can't it be controlled through relay like an automatic transfer switch for a whole house generator? What about IPcontrol, RS232, RS485, dry contact? Why resort to blasting the entire house with bad electricity?
I have an order in for 6 powerwalls. I was under the impression they would provide instantaneous, clean power. But it seems there is no data on how clean the power is (THD?), They purposely don't produce proper frequency, and they do not provide backup power instantaneously. And all these shortcomings are completely avoidable.
I find it ridiculous that the gateway can't communicate in some other way with the inverters. All installations require a shut off breaker. Why can't it be controlled through relay like an automatic transfer switch for a whole house generator? What about IPcontrol, RS232, RS485, dry contact? Why resort to blasting the entire house with bad electricity?
I have an order in for 6 powerwalls. I was under the impression they would provide instantaneous, clean power. But it seems there is no data on how clean the power is (THD?), They purposely don't produce proper frequency, and they do not provide backup power instantaneously. And all these shortcomings are completely avoidable.
There is no such thing as instantaneous switch over unless you are already running completely off solar/battery (and you can set it up that way) but if you are running off the grid, physically speaking it’s impossible to have an instantaneous switch. It’s on the order of 300-500ms.
If Tesla made an integrated inverter and battery solution I’m sure they could use some form of communication other than frequency to turn off the inverters. However there are dozens of inverter options and the easiest solution is to bump the frequency slightly. I’ve had mine changed to 62 Hz and everything works just fine...
If Tesla made an integrated inverter and battery solution I’m sure they could use some form of communication other than frequency to turn off the inverters. However there are dozens of inverter options and the easiest solution is to bump the frequency slightly. I’ve had mine changed to 62 Hz and everything works just fine...
Yes, the same for both
The reason is that there is no standard for this, other than voltage or frequency shift. The powerwall is designed to work with pretty much any inverter on the market and while some brands and models may support out of band control to shut off, not all of them do. And even the ones that do don’t have any standard. It would be a support nightmare if they tried to offer out of band support for only certain inverters.
Theoretically an automatic transfer switch might be an option, but that would involve extra hardware and extra wiring on every single job. And it would need to be a pretty beefy transfer switch too. There are a lot of big solar installs that have a lot of power flowing through those wires.
The other thing to remember is that the high frequency issue is not something that will happen normally during day to day operations. It will only happen when the grid is down *and* the powerwalls are fully charged. At all other times the frequency will be normal
There certainly are ways to avoid those issues, but the question is, at what cost? Powerwalls are already expensive and trying to work around these issues would likely add considerably to the cost. At some point there needs to be a compromise made in terms of ‘good enough’ to keep the price reasonable.
I'm another new solar customer searching this site for information and help. Looks like I'm now a member of this problem group. I haven't been given PTO so I'm in a testing mode. Sun is out medium to strong. PWs at about 40% and I'm seeing home equipment flicker. For me it's my computer and even fish tank which has its own backup battery. Found this thread and on the phone holding for Tesla support. My added problem is my home office is fully loaded with computer, fish tank, etc which is probably already stressing the circuit as-is. This flicking of the UPS actually starts tripping the breaker.
The first tech I talked to tried to tell me that PWs are only intended for me to support critical load to lower my expectations. I quickly squashed that and said I have multiple PWs for that very reason and that answer is not acceptable. I asked him if he knows about this "frequency" issue .. which he did not. And said he needed to transfer to the PW side of the house. FYI here is if anyone calls for support..talk to the PW side. .not the Solar side.
The solar lady was awesome. Very technical, but did say that my home was unique in that my PWs were well under charge capacity (30%) so should not have been impacted by the scenario where PWs were full. She has been doing this for several years and couldn't figure it out and opened the tier 2 ticket after capturing my UPS models. The speculation is frequency, but she also saw spikes in usage which I'm attributing to either my ACs kicking in along with the power surges (soft start not installed yet). My soft starts come in in a few days which I'll run the same testing to see if that's related assuming they haven't addressed the frequency issue so I wouldn't know where the problem was...
The first tech I talked to tried to tell me that PWs are only intended for me to support critical load to lower my expectations. I quickly squashed that and said I have multiple PWs for that very reason and that answer is not acceptable. I asked him if he knows about this "frequency" issue .. which he did not. And said he needed to transfer to the PW side of the house. FYI here is if anyone calls for support..talk to the PW side. .not the Solar side.
The solar lady was awesome. Very technical, but did say that my home was unique in that my PWs were well under charge capacity (30%) so should not have been impacted by the scenario where PWs were full. She has been doing this for several years and couldn't figure it out and opened the tier 2 ticket after capturing my UPS models. The speculation is frequency, but she also saw spikes in usage which I'm attributing to either my ACs kicking in along with the power surges (soft start not installed yet). My soft starts come in in a few days which I'll run the same testing to see if that's related assuming they haven't addressed the frequency issue so I wouldn't know where the problem was...
Vines
Active Member
The PW lady is correct. Unless you are at 100% on the powerwalls you still have room in the PWs to store solar-generated electricity.
Strange problem. How many PWs do you have and what does your A/C draw on startup? PWs are 5kW (7 kW surge, but I hear you should not count on that) each so if you get close to the limit with baseload + A/C when the A/C starts things may not work as planned.
Last edited:
What exactly do you mean by “Seeing home equipment flicker”. I’m not sure how a computer can flicker.
And I’m not quite quite sure what you mean by this. A UPS doesn’t really flicker either.
I think it’s unlikely that a UPS would trip a breaker, unless maybe the circuit was nearly overloaded and after the UPS runs on battery for a bit when then power comes back and the UPS starts drawing extra power to recharge it’s battery maybe that’s enough to push it over the edge and overload the breaker.
Can you clarify about exactly what you are experiencing?
I do agree with the others that it doesn’t sound like the powerline frequency issue. That will only happen when the powerwalls are nearly charged (around 98% or higher), so you won’t be seeing that when they are around 40%.
I have 4 PWs (the largest UK domestic installation I believe) and about 14 KW solar with 3 inverters (2 * 4KW and 1 * 6KW).
i see strange behaviour at two different times:
1. If the grid is connected, the sun is strong and the PWs are full, the solar generates a lot of power to export. This is achieved by the inverters raising their voltage to higher than the grid voltage by 1-2 V and the current then flows into the grid. As we are quite a way from the local last stage transformer, the at maximum export the inverters have to raise the voltage too high to force the current into the grid, and at least one of the inverters will then detect the over voltage and shut down for a period before trying again. This causes voltage variations of what can be as much as 15V from no export to max export before an inverter shuts down;
2. When the PWs are full, and the grid has failed ( I simulated this by removing the main 100A fuse), there is nowhere for excess generation to go. I have a frequency meter permanently installed for nerd interest, and watched the usual 50 Hz (remember this is non US) rise to nearly 52 Hz, and the inverters again shutting down.
So in one case there are voltage fluctuations, and in the other frequency fluctuations. They are both of material amounts. attached equipment might or might not be sensitive to these.
There is one other condition where I see voltage jumps and that is when the house moves into the off-peak charging time and the PWs start to charge in case and my EVs start at the same time to charge. The voltage then can drop by about 20-25V (going from say 2KW draw to 20 KW total). My Model S sometimes decides that voltage drop could only occur if I was using an extension lead and protects itself by going red and stopping charging. Solution: don't start PW and car charging at exactly the same moment!
i see strange behaviour at two different times:
1. If the grid is connected, the sun is strong and the PWs are full, the solar generates a lot of power to export. This is achieved by the inverters raising their voltage to higher than the grid voltage by 1-2 V and the current then flows into the grid. As we are quite a way from the local last stage transformer, the at maximum export the inverters have to raise the voltage too high to force the current into the grid, and at least one of the inverters will then detect the over voltage and shut down for a period before trying again. This causes voltage variations of what can be as much as 15V from no export to max export before an inverter shuts down;
2. When the PWs are full, and the grid has failed ( I simulated this by removing the main 100A fuse), there is nowhere for excess generation to go. I have a frequency meter permanently installed for nerd interest, and watched the usual 50 Hz (remember this is non US) rise to nearly 52 Hz, and the inverters again shutting down.
So in one case there are voltage fluctuations, and in the other frequency fluctuations. They are both of material amounts. attached equipment might or might not be sensitive to these.
There is one other condition where I see voltage jumps and that is when the house moves into the off-peak charging time and the PWs start to charge in case and my EVs start at the same time to charge. The voltage then can drop by about 20-25V (going from say 2KW draw to 20 KW total). My Model S sometimes decides that voltage drop could only occur if I was using an extension lead and protects itself by going red and stopping charging. Solution: don't start PW and car charging at exactly the same moment!
This definitely sounds unusual and I’ve never seen this behavior before. I would suggest talking to your installer to see what can be done about this.
This is normal behavior. When you are off grid and the powerwalls are full the inverters need to shut down otherwise there will be too much power with nowhere for it to go. In the US the system will raise the frequency from 60Hz to 65Hz, so in your case going from 50Hz to 52Hz seems normal. When you are operating off grid and the powerwalls fill up you will see a cycle. The system will raise the frequency which shuts off the inverters, then the house runs on the powerwalls for a short time. When the powerwalls get down to about 97% the frequency returns to normal and the inverters come back online. If the sun is still strong the powerwalls will start charging again and once they get to about 98% they will raise the frequency again to shut off the inverters.
@BrettS: thank you - yes, I was actually aware of the intended behaviour in case 2, which is why I installed the frequency meter - so I could test it.
case 1 is the normal risk if someone has too much solar and the resistance in the line back to the final transformer is a little high - as it tends to be if you’re a long way from it. Not a city problem, definitely a rural one. Sadly there is no easy solution. Of course what I really want is to dynamically bleed any excess of generated power off into a Tesla car. Not easy to do right now - I wish Tesla had created the mechanism to do this!
case 1 is the normal risk if someone has too much solar and the resistance in the line back to the final transformer is a little high - as it tends to be if you’re a long way from it. Not a city problem, definitely a rural one. Sadly there is no easy solution. Of course what I really want is to dynamically bleed any excess of generated power off into a Tesla car. Not easy to do right now - I wish Tesla had created the mechanism to do this!
I had a similar issue, but after the sun was already down for a while. The Grid then went out, and I had electronics "humming" and a UPS failure - so it seems like the Powerwalls raised to 65 Hz, even though there was no need to inhibit solar panels? Is that normal?
I have not had this issue when in Self-Generation mode every day and pulling from the Powerwalls after the sun went down. So it seems like Grid Outage = Powerwalls changing to 65Hz regardless of whether the panes are active or not?
I have not had this issue when in Self-Generation mode every day and pulling from the Powerwalls after the sun went down. So it seems like Grid Outage = Powerwalls changing to 65Hz regardless of whether the panes are active or not?
carteriii
Member
I've seen similar behavior to what you describe and I believe it is primarily or entirely due to the charge state of the Powerwalls. I have a plug-in Kill-a-watt meter which showed 65Hz at night, when no sun was out, when my Powerwalls were 99% charged. After my Powerwalls had discharged by 10-15%, I saw about 62Hz (still at night). I went to sleep and didn't get to see if it lowered further before I woke up and the grid was connected again.
I'm a bit puzzled...
Please allow me to explain: I have a 14.4kW solar tracker system with 5 PW2 and latest gateway.
I never see my 5 PW charged 100%, max is at 99%. Last year I had 4 PW installed and a gw. Recently I got my 5th PW coming with the latest gateway. So because of this recent gw swap, I think I lost the initial customized firmware that Tesla put in last year to address the freq change.
After trying to address this issues repeatedly with Tesla via 800 number and local manager, I gave up, because of absolutely no desire from Tesla to help.
Besides the above issue, I also ended up with 2 gateways in my app, and I cannot get the old one replaced because of the above reasons.
Any suggestions?
Thanks!
Please allow me to explain: I have a 14.4kW solar tracker system with 5 PW2 and latest gateway.
I never see my 5 PW charged 100%, max is at 99%. Last year I had 4 PW installed and a gw. Recently I got my 5th PW coming with the latest gateway. So because of this recent gw swap, I think I lost the initial customized firmware that Tesla put in last year to address the freq change.
After trying to address this issues repeatedly with Tesla via 800 number and local manager, I gave up, because of absolutely no desire from Tesla to help.
Besides the above issue, I also ended up with 2 gateways in my app, and I cannot get the old one replaced because of the above reasons.
Any suggestions?
Thanks!
If the grid is down and the Powerwalls are full there is no where for excess power from solar generation to go. Electricity has to go somewhere and with no grid and a full battery the options from there are limited to very bad things (read: fire). To prevent PV from producing when the grid is down and Powerwalls are full regardless of when the system increases the microgrid frequency. This is all very expected and normal.
What's odd is if you SoC dips below 95% and you're still seeing high frequencies (62hz or more). All of this is configurable by Tesla. I was able to get them to match my off grid profile to how my IQ6 inverters are configured to maximize the Powerwall while keeping the frequency as low as possible during off grid operations.
What's odd is if you SoC dips below 95% and you're still seeing high frequencies (62hz or more). All of this is configurable by Tesla. I was able to get them to match my off grid profile to how my IQ6 inverters are configured to maximize the Powerwall while keeping the frequency as low as possible during off grid operations.
MorrisonHiker
S 100D 2021.4.12
What setting do you use? Backup-only, Self-powered, Advanced Time-based control Balanced or Advanced Time-based control Cost Saving? We use Balanced and ours charge to 100% every day there's enough sun.
carteriii
Member
I believe @MorrisonHiker likely hit on the core issue. When backup-only, I don't see a full 100% which to me makes sense because you don't want to keep the batteries at 100% charge for an extended period of time. By contrast, when I run with time-based control, I always see 100% charge which I assume is because the system knows the batteries will be used some every day.
Thanks, yes I only use backup only...However knowing this now, I should not hit on the freq increase issue? Because there is always room to charge in that case, correct?
Also how do you guys get in touch with Tesla Energy, any tricks? I always hit long on hold times, people never follow up etc
Thanks!
