Powerwall: "You will not even notice that the power went out."

[Yonki]

Tesla makes this claim at the end of the answer to the second question here.

Has anyone (conscious, and with a visible light on) found this to be true?

Hilariously, as I was typing the word "found", our power went out. The TV (that I use as a monitor) clicked and went black, the A/V receiver clicked, but both came back after 1-2 seconds (PC and cablemodem/router are on UPS so no glitches there).

I'm loving having power during outages, but it certainly is noticeable. Wondering if that's universal or I need maintenance.

 

[commasign]

Yes, definitely notice a momentary loss of power. TV, computer, etc will restart. Lights dim or flicker.

 

[tjka4231]

It was definitely noticeable. Had an one minute loss of grid power and my computer restarted, lights flickered.

 

[DrSmile]

I was surprised that Tesla doesn't use capacitors to bridge power, I found this out when I was told that the Powerwalls would not be able to power my A/C units without soft starts. Does anyone why this isn't done? Is it a safety issue?

 

[Vines]

When testing with a customer I have had reports from the Lead Installers of both experiences. Sometimes its nearly seamless with just a dimming of lights, other times its a long enough delay to cause crashing of electronics.

I'm not really sure what the difference is between the cases. Could be related to the current state of the battery, whether its already discharging, charging, or just waiting. Could be related to size of the connected load, its just not yet clear. Maybe someone else can enlighten me.

Not sure about the technical aspects of why, but also know AC units have a relatively large surge current or Lock Rotor Amps (LRA) and don't play nice with Powerwalls unless they have soft start. The Powerwalls don't respond well to being worked too hard, they just disconnect and fault instead.

 

[mongo]

I was surprised that Tesla doesn't use capacitors to bridge power, I found this out when I was told that the Powerwalls would not be able to power my A/C units without soft starts. Does anyone why this isn't done? Is it a safety issue?

Capacitors are DC, what you need is AC, and that's the inverter's job, plenty of juice available in the batteries. For air conditioning, the issue is the near instantaneous high load the compressor motor places on the circuit. The control loop of the powerwall is fast, but that startup load draw is more than the unit is rated or designed for.

For loss of power, the issues are

• the waveform that the grid makes before reaching the threshold of the gateway isolating the house (sharp cut off or brownout)
• the time for the gateway to reliably isolate,
• the time for the Powerwall inverter to ramp up to the load level of the house. Probably takes a little longer if the batteries were charging at the moment of power loss
• The item being backed up. Cost optimized consumer electronics have minimal ability to handle power drops, some things react to the loss of one cycle, 16.7 milliseconds

For critical applications, one form of power backup and conditioning a a large 3 phase motor feeding a large 3 phase alternator. It provides power isolation and the large mass evens out any supply issues. The inertia keeps things running long enough for the backup generators to kick in.

 

[Yonki]

When testing with a customer I have had reports from the Lead Installers of both experiences. Sometimes its nearly seamless with just a dimming of lights, other times its a long enough delay to cause crashing of electronics.

I'm not really sure what the difference is between the cases. Could be related to the current state of the battery, whether its already discharging, charging, or just waiting. Could be related to size of the connected load, its just not yet clear. Maybe someone else can enlighten me.

I have solar, and I think they need to disconnect the solar before they switch over to the battery (which was charging from solar at the time), so that could certainly be some of it. If we're lucky enough to lose power at night (no solar, powerwall idle) it might switch over faster.

 

[BJReplay]

I have found it entirely depends on whether the powerwall is discharging when the grid drops or not. If it is discharging, most devices seem to ride through the voltage dip before it disconnects from the grid. If it is not discharging, it doesn't react quickly enough to start the inverter and sensitive equipment will trip.

There have been posts that suggest that later firmware reacts more slowly, so it is more noticable. I certainly believe that I can see slower response to changing load with the latest firmware - the local GUI and app both show draw from grid when load spikes up before the inverter responds (and feed into grid when load drops).

 

[turnersteve]

I have a 10.5 kWh system with 2 power wall 2's. I have had the grid go down several times, if it happens during the time the PV cells are producing, no flicker, nothing turns off, AC keeps running. If the grid goes down at night there is a flicker of lights, but nothing shuts down.
I do have my computer, printer, cable modem and wifi router on a APC ups, so I do not know if they would be affected. I have my Powerwalls set for backup only.

 

[arnolddeleon]

I have found it entirely depends on whether the powerwall is discharging when the grid drops or not. If it is discharging, most devices seem to ride through the voltage dip before it disconnects from the grid. If it is not discharging, it doesn't react quickly enough to start the inverter and sensitive equipment will trip.
...

I've experienced the "you won't notice it" as well the as several seconds of no power before it kicks in.

It does seem if the inverter is already running and the grid drops out it is practically seamless. The only thing that needs to happen in this is for the system to detect the grid outage and have the transfer switch open.

When it is on standby like it when it is in "storm mode" I assume the sequence of events is, detect grid outage, open the transfer switch (in the gateway) and start the inverters.