Featured Post

This is the Kodak Moment for the Auto Industry

Plug-In Drivers Not Missin' the Piston Electric vehicles are here to stay. Their market acceptance is currently small but growing...

Monday, February 15, 2021

Powerwalls During A Power Outage (Valentine's Day Blackout)

Downed trees in Oregon
Photo Credit: Wade Radcliffe

Our Powerwalls were installed on December 31st, 2020. Just six weeks later, we've had our first power outage. This was our chance to see how well they perform. 

There's an ice storm in the area. Freezing rain has glazed over the bare winter branches in our yard. Storm Watch is enabled and the Powerwalls are charged to 100%. Over 250,000 homes in the state have been hit with power outages. Some have been without power for 48 hours as I write this.

On the evening of Valentine's Day, our neighborhood joined many others in the darkness. The streetlights went out. Looking up and down the block, the houses were dark, with the exception of ours. 

The power had been out for about 5 minutes before we even noticed. We didn't notice because, in our house, things had continued normally. The Powerwalls had taken over and they were running everything. The lights were still on, TV was on, the internet was even still working, and the washing machine was running.

We had no idea how long the power would be out. The sun had set for the day; we were not going to have any solar support. With grid out and no solar, it was all up to the Powerwalls. 

The first thing I did was run around the house turning off and unplugging nonessentials. As I mentioned above, the washing machine was running. It was in the final spin cycle, so we opted to let it finish the run. This was one nice benefit to having Powerwalls. If we didn't have them, the washing machine would have been stopped mid-cycle. This could have left us with soaking wet soapy clothes. Once the washing machine finished, our consumption rate dropped.

I collected data periodically for the charge level of the Powerwalls during the outage. You can see it in the graph below:

After 2 hours and 35 minutes, our power was restored. We were lucky that our outage was just a short one. A few minutes after the power was restored, the Powerwalls handed the load of our home back to the grid. Then as if to test the newly restored grid, the Powerwall started to recharge. The Powerwalls were recharging at a rate of 10 kW, with the other loads in our house, we peaked at a 16kW grid load. After ~20 minutes the Powerwalls were recharged and our grid load dropped to a normal rate. 

How Long Will The Powerwalls Keep Lights On?

Now that the outage is over, it's time to see how long the Powerwalls would have lasted. Luckily, our outage was only a few hours and the Powerwalls were never below 80%, so to determine how long they'd last in total, we'll have to do some extrapolation. To do this, we'll use a few methods. We'll add a trendline to the above graph and see where it lands. We'll look at our consumption data (percentage-wise and kWh-wise) and do the math based on the Powerwall capacity. 

Method One: Lineraly 

This one is a simple analysis. We used 19% of our charge in 2 hours 35 minutes. Taking this linearly, the total charge would have lasted about 5 times that duration or 13 Hours 35 Minutes. 

This, however, includes the time that the washing machine was running and so is not a very accurate estimate.

Method 1 Result: 13 Hours 35 Minutes 

Method Two: Trendline

For this trendline, I selected a 3-period trailing moving average. This allowed the washing machine period to be ignored and projected forward based on the latter portion of the sample. 

This trendline predicts that we'd have drained the Powerwalls by 11AM the next morning. This is just ~18 hours. It's longer than Method 1, but not as long as my pre-purchase calculation which led to our decision to buy three Powerwalls. The goal was 24 hours of backup.

Method 2 Result: 18 Hours

Method Three: Consumption

The other methods were tops-down. This one is bottom-up. It usually a good idea to try to look at things from a few angles to see the bigger picture. For this one, we need to know how much capacity we have. We have 3 Powerwalls. Each Powerwall has 13.5 kWh of usable energy, so that's 40.5 kWh. 

Looking at our usage data from the Tesla app during the outage, it says that we used 3.375 kWh during the outage. That's an average rate of 1.3kW. At that rate, 40.5 kWh would last 31 hours. 

Method 3 Result: 31 Hours 


I was very happy that we had Powerwalls when a blackout hit our neighborhood. It allowed us to keep the lights (and heat) on during the outage. Our outage was less than 3 hours long and the Powerwalls covered it completely with more than 80% charge remaining when the grid came back online. 

We used 3 methods to estimate the total duration the Powerwalls would have powered our home. The most pessimistic estimate was 13 hours 35 minutes. The most optimistic estimate was 31 hours. To determine which one is more accurate, we'll need a larger sample set. The truth is likely in the middle and if we needed to stretch things further, we could have reduced our home load to just the bare essentials, until the sun comes out and gives us a little solar power support.

Disclosure: I am long Tesla


  1. I have a Tesla 3 and 9.6 kW Solar system. I would think you would lose about 30+% of your total kWh battery capacity if the powerwall is in an unheated garage. I've been testing out my Tesla capacity in this cold weather in Colorado.

    1. kd, that's a good point. I don't know how much energy was being used to heat the Powerwalls themselves on this cold night. They certainly would need it. That could explain some of the difference between methods 1/2 and 3. Thanks.