How to tell after power outage if you are using PW Batteries or if the grid is back on?

[MorrisonHiker]

It depends on what state your PWs are in when the power is cut. If they are currently supplying power (like peak periods in Time Based Control), there will be no delay at all. But if they are in standby, then there could be a fraction of a second or more where they need to detect the loss and switch over to providing power.

This is why for critical device that are time sensitive (like servers), you still need to have UPS's. But you only need them to act for a few seconds at best, so a tiny one is all that is needed.

That's why we really liked using the Balanced option when it was available. With that option, we would run our household about >22 hours a day via the Powerwalls while still ensuring peak period was always covered. We used it more as a "scheduled self-powered" option. We never noticed any interruptions during outages. The current self-powered option wouldn't be ideal since it wouldn't ensure we had enough power for peak. We could probably tweak the time-based schedule/utility plan to get close to what the Balanced option allowed. Unfortunately, that would mess up the energy cost calculations but it would prevent most interruptions.

 

[jboy210]

It depends on what state your PWs are in when the power is cut. If they are currently supplying power (like peak periods in Time Based Control), there will be no delay at all. But if they are in standby, then there could be a fraction of a second or more where they need to detect the loss and switch over to providing power.

This is why for critical device that are time sensitive (like servers), you still need to have UPS's. But you only need them to act for a few seconds at best, so a tiny one is all that is needed.

Interesting point about operations modes. We are often in self-powered mode. However, this time of the year with we switch to Time-Based control because it maximizes savings since our system does not produce enough to cover our needs and charge the batteries to 100% before evening. I think I will do a test and see if the cutover to the batteries takes longer in Time-Based.

Also, good point about UPS's on computers and other critical gear. We have them on most of our critical gear since we develop software and training materials, and work from home. Loss of data equates to loss of money to us and makes the cost of a few small UPS units a minor business expense to cover potentially very large financial impact.

 

[aesculus]

I think I will do a test and see if the cutover to the batteries takes longer in Time-Based.

It's not only about being in Time Based, but what it's doing in Time Based, at the time.

It has to be discharging (supplying power to your house) to avoid that break. If it's charging or in standby, you will see the break.

I have not done an exhaustive study to determine if the switchover time while it's charging or in standby is different, but its really irrelevant anyway. There is still a break. But if it was different, some devices could be more susceptible than others.

So far for me only my NAS does not seem to ride over the disconnect (or seamlessly recover from it), so that is the only device I have on a UPS (well a few others are plugged into it like switches/printer, but that was a matter of plug convivence).

 

[jboy210]

It's not only about being in Time Based, but what it's doing in Time Based, at the time.

It has to be discharging (supplying power to your house) to avoid that break. If it's charging or in standby, you will see the break.

I have not done an exhaustive study to determine if the switchover time while it's charging or in standby is different, but its really irrelevant anyway. There is still a break. But if it was different, some devices could be more susceptible than others.

So far for me only my NAS does not seem to ride over the disconnect (or seamlessly recover from it), so that is the only device I have on a UPS (well a few others are plugged into it like switches/printer, but that was a matter of plug convivence).

Interesting. I will see if I can duplicate this behavior. So far, we never had to even reset the clock on the stove.

On the UPS we have had them on all of our computers without a battery, NASes, and networking racks for years. With the Powerwalls I suspect they are now quite oversized for even the worse-case transition. But a UPS usually adds some surge protection which I don't think we get from the Powerwall system.

 

[xWren]

It depends on what state your PWs are in when the power is cut. If they are currently supplying power (like peak periods in Time Based Control), there will be no delay at all. But if they are in standby, then there could be a fraction of a second or more where they need to detect the loss and switch over to providing power.

That would make sense, but I experience a noticeable delay even when the Powerwall is currently supplying power too. It is like the sudden absence of the grid causes it to briefly stop supplying while it figures things out. Since I'm in the UK while the grid is there we can export, but absence of the grid means the Powerwall must prevent that happening. Could isolating (even while not actually exporting) create the delay? Or maybe some property of my supply (I'm the single client on a transformer that takes power from the 11kV network, a little unusual)? Any junk on the supply as it cuts out has nowhere else to go and dissipate but our house.

Anyway not a huge problem, just something I'm curious about especially when others experience uninterrupted supply.

 

[aesculus]

That would make sense, but I experience a noticeable delay even when the Powerwall is currently supplying power too. It is like the sudden absence of the grid causes it to briefly stop supplying while it figures things out. Since I'm in the UK while the grid is there we can export, but absence of the grid means the Powerwall must prevent that happening. Could isolating (even while not actually exporting) create the delay? Or maybe some property of my supply (I'm the single client on a transformer that takes power from the 11kV network, a little unusual)? Any junk on the supply as it cuts out has nowhere else to go and dissipate but our house.

Anyway not a huge problem, just something I'm curious about especially when others experience uninterrupted supply.

Interesting. Maybe a UK (EU?) thing? In the US my experience has been no delay at all when the PWs are feeding the house. I don't have a decent understanding of how the 3 phase systems work to comment technically.

 

[BGbreeder]

That would make sense, but I experience a noticeable delay even when the Powerwall is currently supplying power too. It is like the sudden absence of the grid causes it to briefly stop supplying while it figures things out. Since I'm in the UK while the grid is there we can export, but absence of the grid means the Powerwall must prevent that happening. Could isolating (even while not actually exporting) create the delay? Or maybe some property of my supply (I'm the single client on a transformer that takes power from the 11kV network, a little unusual)? Any junk on the supply as it cuts out has nowhere else to go and dissipate but our house.

Anyway not a huge problem, just something I'm curious about especially when others experience uninterrupted supply.

Do you have a three phase installation?

Being a single client should not be a generally a particular issue; locally, almost everyone on our 22kV branch has their own transformer(s). What is upstream of your transformer might be an issue, as the other loads may cause the voltage to be erratic as they disconnect/power down. (But that isn't exactly a readily fixable problem...)

I would be open to the possibility that your outages are not just 240V to zero instantly, and that there may be voltage irregularities on the way to failure that make it difficult for your Gateway to "know" when to disconnect. If you log into your Gateway, you can check to see what grid function it is programmed for. The particular grid specifications may shed some light on the switchover behavior...

I am afraid that the best way to understand your power failure switchover is to install a power monitor to log what your power actually is; voltage variation, frequency shift(s), and the exact nature of the disruption. However, I would rather bottom line the issue; what would you change, and what would you expect Tesla and your power utility to change? If the answer is nothing, I wouldn't bother.

On a personal level, we have a UPS on critical equipment like routers and certain electronic devices that do not handle power disruptions well. We occasionally see the lights flicker when an outage starts, and can hear the Gateway relay closing (it is a hard to miss "clunk" if the house has significant load...), but that is it. Mostly, I get a text saying "you are experiencing an outage" as the first sign that there is an issue. I haven't personally noticed a difference to the source of power, grid/solar/Powerwall, in terms of switchover behavior. Our outages tend to be a) a fuse on the 22kV line blowing due to a local short, or b) more frequently, a tree branch touches the 22kV line somewhere, and the utility "sectionalizer" opens the circuit, i.e. a transiently lower voltage leads to zero voltage in 1/120th of a second. The latter is a bit of an issue as all forty or fifty miles of power lines have to be manually inspected before the utility re-energizes the lines.

All the best,

BG

 

[WannabeOwner]

Anyone know why I'm not getting the smooth switchover

I have this problem (I'm in UK too [Single phase here])

Original install (quite some time ago) gateway didn't have this ability. Then an OTA upgrade enabled it, and after that I had sufficiently "instant" switch over that nothing reset. Some time after that it started taking a second or so to switch over, and now I get a break sufficient to trigger everything to reset. Annoying - Particularly given I had a period where it worked just fine.

I'm in a rural location and have had 22 power cuts YTD, only one of them for a significant amount of time (and that was for scheduled maintenance). Power Company have been telling me for years that they have an old / flaky piece of kit somewhere and they are imminently replacing it ... clearly NOT!