Power glitch when exporting to the grid?

[Boatguy]

I have Tesla solar and two PWs. My PG&E service is notoriously "glitchy". PGE does some sort of switching which will cause a momentary drop or spike in my service. Over a period of five years this caused a series of failures in audio equipment power supplies. Finally I installed line interactive UPS's on each unit. The line interactive UPS switch much faster than stand-by UPS's and now I have no problems. That's the background.

Today there was a power outage in part of my city, but it did not affect my home. But apparently it caused PGE to do some switching as referenced above. The odd thing is that they did this while the Tesla app showed my house was running on solar and also exporting to the grid. So as reported, I was not drawing any power from the grid. Yet several times my UPSs triggered to backup my audio and computer equipment. These were transients, less than a second. The drops did not trigger the PWs. Just long enough to trigger the UPSs.

So the question is, if my house was being powered by solar, with excess power going to the grid, why would a PGE glitch trigger my UPSs? If my house is 100% solar powered at that moment, how could anything PGE does cause a change in my house power?

 

[charlesj]

Sheer speculation on my part but these short duration glitches may have cause voltage, frequency spikes one way or another, short duration, that system was unable to properly respond in a meaningful manner? Your Ups batter could as it seems to be instantaneous.

When I did my grid down simulation, PW was fast enough so that microwave and oven clocks didn't reset to 0:00 but it was a smoother even than who know what their switching has generated.

 

[cwied]

Remember that when the grid is up, the Powerwall and solar inverters are connected to the grid, they have to follow what the grid is doing. The Powerwalls are not a pass-through system, they're just attached to the same wiring that the grid is. The solar inverter and Powerwall inverters just synchronize to the grid then adjust slightly to provide power into the wiring. It's only when the grid is down that the switch in the gateway disconnects the house and the Powerwall inverters take over as the primary power source that they can fully control the generated power.

 

[aesculus]

So the question is, if my house was being powered by solar, with excess power going to the grid, why would a PGE glitch trigger my UPSs? If my house is 100% solar powered at that moment, how could anything PGE does cause a change in my house power?

That is an interesting question. Do you know if the solar inverters recycled during any of this?

Not sure of PG&Es equipment in your case. They may have very fast acting transfer switches but obviously not instantaneous (ie 0 time). The best industry practice here is a "make before break" where power if fed from two sources at one time and the only thing you are actually changing is the paths to the load from the source. In this case you always have a constant power source because you add one before you remove the other.

 

[Boatguy]

That is an interesting question. Do you know if the solar inverters recycled during any of this?

Not sure of PG&Es equipment in your case. They may have very fast acting transfer switches but obviously not instantaneous (ie 0 time). The best industry practice here is a "make before break" where power if fed from two sources at one time and the only thing you are actually changing is the paths to the load from the source. In this case you always have a constant power source because you add one before you remove the other.

@cwied: I'm not sure what "synchronize to the grid" means. I think the presence of the PWs was probably irrelevant since they were fully charged and their backup switch was not triggered. But the solar panels were feeding the house and the grid which I assume they do by having a slightly higher voltage than the grid, but I've reached the border of my EE knowledge!

@aesculus: When you say "recycled", do you mean like a boot cycle or something that interrupted their behavior? If so, I saw no anomaly in the app. It showed the flow from the solar panels split between the house and the grid.

Having experienced this for years I think it's safe to assume PGE does not have "make before break" switching.

This happened a couple of times that day and in one case I was at my computer and the screen dimmed slightly (per the UPS settings) and a message came on the screen saying that the computer was now powered by the UPS. That lasted maybe a second. That suggests to me that UPS had lost it's power from the wall. So my question is if the solar panels were powering the house, how can the wall outlet lose power? I guess I do not understand the power path under various scenarios.

 

[jjrandorin]

I think, in the terms that @cwied ment when he said "syncronize to the grid" he is talking about the grids frequency / waveform.

 

[BrettS]

This happened a couple of times that day and in one case I was at my computer and the screen dimmed slightly (per the UPS settings) and a message came on the screen saying that the computer was now powered by the UPS. That lasted maybe a second. That suggests to me that UPS had lost it's power from the wall. So my question is if the solar panels were powering the house, how can the wall outlet lose power? I guess I do not understand the power path under various scenarios.

The issue is this: AC power in the US should operate at exactly 60hz. However, due to a number of possibilities it often isn’t exactly 60hz. Sometimes it might very by a little bit, maybe 59.5hz or 61hz, occasionally it might vary by a lot, it might drop to say, 50hz or 55hz for a fraction of a second if there is an issue at the power company.

When your solar panels and/or power walls are generating power they need to match the frequency of the utility exactly or bad things while happen. If, for example, the power company is at 60.5hz and the solar inverters are operating at 60.0hz they will get out sync and this can cause damage to the inverters as well as the electronics in your home. So the solar inverters are designed to sense the frequency of the power being provided by your utility and match it.

Normally this is invisible and you don’t really think about it. If the utility power varies between 60.0hz and 60.5hz the inverters just change to match it and everything carries on. However, if the utility starts providing power that is way out of spec, say 50hz or 55hz, the solar inverters still need to match that (or just drop offline entirely) to prevent damage. However, at that point the power is bad enough that your UPS’s will probably switch over and you may notice your lights flicker and such. If the power continues to be poor then the gateway will disconnect from the utility and switch to the powerwalls, but if it recovers quickly then that likely won’t happen.

The UPS’s are kind of in a tough spot. If the power just flickers or goes out of spec for a fraction of a second, then they probably wouldn’t need to come online. But the problem is that they don’t know if the problem is just going to last for a fraction of a second or if the power is going to be out for minutes or hours. If they take a wait and see approach and it is an extended failure, then they might not be able to come online fast enough and your computer might reboot. So UPS’s will switch over for even the tiniest of problems. In that case the equipment will run on the UPS for a few seconds and the UPS will see that the utility power is clean again and then it will switch back. But if the UPS’s didn’t do this then when there really is a power failure they wouldn’t be able to switch over fast enough.

So briefly what is happening is that your solar/powerwall system needs to match the frequency of the provided utility power, and when something happens at the utility that causes a small interruption or frequency change the solar/powerwall system needs to match that as well. Your UPS will go online as soon as it sees any small interruption. If it is more than just a flicker then the gateway will disconnect your house from the grid and the powerwalls will come online and their inverters will be the primary frequency that the solar inverters match. But if it’s just a small flicker then that won’t happen, but it will likely still be enough for your UPS to kick in. Just be cause the UPS goes online doesn’t mean that it lost power. It could mean that it wasn’t happy with the power that was being provided.

 

[DrSmile]

No matter how much solar you're producing, as long as the grid is connected your solar inverter needs to match the grid voltage. I found this out when my power company was supplying low voltage, the inverter fried the breaker because it was feeding more amps than the breaker rating. I had to run the system completely off-grid for a while until they fixed the low voltage problem.

 

[Boatguy]

OK. That's all interesting information. Can someone connect the dots and tell me why, when the solar is powering the house, a power company glitch shows up at the wall outlet?

Or maybe we don't know because we don't know the nature of the glitch (frequency, voltage, etc.)?

 

[aesculus]

I still think it's odd that your inverters did not disconnect and reboot. This leads me to believe there must be some sort of delay before they drop off line when they sense a voltage and or frequency anomaly. What that delay is I have no clue. You might have to pose this question to your inverter company.

It sounds like they were willing to keep going for that second or so but the UPS was having none of it.

 

[power.saver]

The UL 1741-SA standard has a requirement for "ride through" which means inverters are to continue to operate during minor or brief disturbances, so as to help stabilize the grid. If the disturbance is large or prolonged, they will then disconnect. The UPS, however, does an immediate switchover as soon as it detects any anomalies.

 

[charlesj]

No matter how much solar you're producing, as long as the grid is connected your solar inverter needs to match the grid voltage. I found this out when my power company was supplying low voltage, the inverter fried the breaker because it was feeding more amps than the breaker rating. I had to run the system completely off-grid for a while until they fixed the low voltage problem.

I really question this that voltage needs to match, especially if the panels are to send power to the grid. At same voltage it cannot send power, it has to have more force to get to the grid.
My Enphase page shows 245.8V and grid meter shows 243. That difference will transfer power.
Inverter controls the voltage parameters and the breaker will shut down over current conditions. That is the reason for breakers, current breakers. And, solar panels are also limited by design what they will deliver, both DC voltage and current. A short will generate the max rated solar panel current that gets converted to AC volts and currents. I seriously doubt solar can fry a circuit breaker before the breaker tips.

 

[getakey]

OK. That's all interesting information. Can someone connect the dots and tell me why, when the solar is powering the house, a power company glitch shows up at the wall outlet?

Or maybe we don't know because we don't know the nature of the glitch (frequency, voltage, etc.)?

Excellent explanation by Brett. It is because the Inverter must match the "glitch" in frequency and it shows up at the outlet

 

[cwied]

It's probably easier to think of it in terms of DC than AC (at least it is for me). So, ignoring frequency for a moment, if the grid voltage drops, the solar inverter has to drop it's voltage too. If it doesn't, a current flow will result out to the grid. You can't power the house at a different voltage than the grid unless you disconnect from it. This is unlike a double-conversion UPS where the grid power is not directly connected to the output, so the output can differ from the grid power.

 

[power.saver]

I really question this that voltage needs to match, especially if the panels are to send power to the grid. At same voltage it cannot send power, it has to have more force to get to the grid.
My Enphase page shows 245.8V and grid meter shows 243. That difference will transfer power.
Inverter controls the voltage parameters and the breaker will shut down over current conditions. That is the reason for breakers, current breakers. And, solar panels are also limited by design what they will deliver, both DC voltage and current. A short will generate the max rated solar panel current that gets converted to AC volts and currents. I seriously doubt solar can fry a circuit breaker before the breaker tips.

No, inverters are current sources and match the voltage of the grid. They are not voltage sources like a battery that require a difference in potential to move current. The reason you see a nearly 6 volt difference is because your inverters are far from your meter and the wiring is undersized. You are creating a voltage difference between the inverters and your meter when they are sourcing current. Watch as solar production drops and you will see this voltage difference drop as well. Six volts is a lot of resistive loss in your wiring.

 

[cali8484]

So the question is, if my house was being powered by solar, with excess power going to the grid, why would a PGE glitch trigger my UPSs? If my house is 100% solar powered at that moment, how could anything PGE does cause a change in my house power?

It's because your house never disconnected from the grid so so the "grid forming" voltage waveform required by your PV inverters is still coming from the grid. Your PV inverters 100% powering your house just means current is flowing towards the grid instead of from the grid but it won't stop any voltage waveform glitches from the grid from getting to your house/UPS.

 

[charlesj]

No, inverters are current sources and match the voltage of the grid. They are not voltage sources like a battery that require a difference in potential to move current. The reason you see a nearly 6 volt difference is because your inverters are far from your meter and the wiring is undersized. You are creating a voltage difference between the inverters and your meter when they are sourcing current. Watch as solar production drops and you will see this voltage difference drop as well. Six volts is a lot of resistive loss in your wiring.

That voltage difference is 2.8V, 243 and 245.8
If it is current driven the resistance needs to change someplace. And, if the load in the house doesn't change and panel produces more than needed, how does the grid drop resistance to push the power into the grid?
Something has to change for current to change directions; load changes or voltage changes.

 

[power.saver]

If it is current driven the resistance needs to change someplace. And, if the load in the house doesn't change and panel produces more than needed, how does the grid drop resistance to push the power into the grid?

Grid impedance is very low, typically 0.2 ohms or less. This doesn't change when you are producing power from your micro inverters.

Something has to change for current to change directions; load changes or voltage changes.

The current is AC and changes direction 60 times a second. You are confusing this with DC where it flows in only one direction. The voltage difference you see is due to the current flowing in your wiring, not a change in the grid. If your inverters were connected directly to the service point at your meter, there would be no difference in the voltages, yet power would still flow from them to the grid. But since your inverters are far from your meter, there is a voltage drop, due to the current flowing in the wires between them. This voltage drop appears as a higher voltage at the inverters.

The inverter is not increasing the voltage to get current to flow, which was your original premise. The voltage difference is the result of current flow in your wiring.

 

[charlesj]

Grid impedance is very low, typically 0.2 ohms or less. This doesn't change when you are producing power from your micro inverters.


The current is AC and changes direction 60 times a second. You are confusing this with DC where it flows in only one direction. The voltage difference you see is due to the current flowing in your wiring, not a change in the grid. If your inverters were connected directly to the service point at your meter, there would be no difference in the voltages, yet power would still flow from them to the grid. But since your inverters are far from your meter, there is a voltage drop, due to the current flowing in the wires between them. This voltage drop appears as a higher voltage at the inverters.

The inverter is not increasing the voltage to get current to flow, which was your original premise. The voltage difference is the result of current flow in your wiring.

Well, I am not confusing DC and AC current.
If grid voltage is same as inverter, why would current flow into the grid. Current doesn't flow without voltage into a load, the grid or house.
Why don't you need a higher voltage to overcome the grid voltage for current to flow in that direction. Yes, AC current still flows to loads because of the voltages. Just not seeing why such a split in the loads, house and or grid when the voltages are the same.

The current from the micro inverters to the meter goes through huge wire, 100A capacity for the 100A breaker. Voltage drop is a small fraction of a volt, even at max 15A current.
ps.
perhaps the current/power flows to the grid because of the differences in the house load and grid load, 24 Ohms to .2 Ohms?

 

[power.saver]

If grid voltage is same as inverter, why would current flow into the grid.

Because the inverter is a current source, and it pushes current into the circuit irrespective of voltage.

Current doesn't flow without voltage into a load, the grid or house.

That is true for a voltage source, but not a current source. Your inverters are a current source.

Why don't you need a higher voltage to overcome the grid voltage for current to flow in that direction.

Perhaps reading up on current sources will help you understand this concept.