Inaccurate power consumption (correlated with solar generation)



I have a relatively new Sense install. I’ve been running a few weeks. A breaker got cycled for some unrelated electrical work a week back (July 9) Since that time, my power consumption has been inaccurate during the day when solar is being generated. See the attached image. Now, my consumption during the day seems to be correlated w/ the solar generation. My house is empty is during the day and I’ve confirmed with my utility that the consumption is much higher than they are reporting (by ~70%+). Eyeballing it, the consumption prior to the cycle looks +/- correct.

So my questions:

  1. Given its the breaker was cycled, what is the right process for re-calibration (assuming that is the issue)?
  2. If recalibration is required, is there a way to do that without a factory reset?
  3. If I have to do a factory reset will I lose my device learnings? [that’s probably the only reason I care]
  4. I read something about Sense also having to do something on their end, is that true?
  5. In the future, what is the right process for cycling the breaker? (I am pretty sure it was the breaker where Sense is powered)

Thanks for any perspectives here.


I would contact tech support, and send them an image of what you have here to speed things along ( This is not an issue that you can (permanently) fix with a restart and recalibration, at least that’s my experience. This is something that tech support needs to have a look at and fix on their end. I have had this issue as well months ago, and with some help from Alex or Rayshawn it was corrected.

note: it looks like this problem started well before the breaker issue - I can see problems in the usage graph already on July 7th - got any data further back than that? This may additionally be an installation error.

Cycling a breaker - not sure if there’s a correct way, but perhaps there’s a processing-side fix Sense can work in so that it doesn’t happen to begin with?


Thanks. Yes, I started a support ticket in parallel to this thread - wasn’t sure of proper etiquette. I had been reading the forum to see if others had similar issues. Calibration issues seemed like the closest fit.

The issue on the 7th was a Sense-induced reboot of my system based on “improper configuration” during install. Sense contacted me so I’m not sure what they saw or what it fixed. Prior to the 7th I was getting readings for a day or so, but it was basically a new install at that point.


Here’s what I’m seeing:

The areas circled are, well, bad data (as are the rest of the daylight periods as you pointed out). They show your usage (the red line) to be at or below zero, implying that one or both of your main line (incoming from the grid) CTs was reading incorrectly. The red line should be recording usage no matter whether you’re producing or not. Probably just one CT is misreading, since I can see on the 8th and 9th positive activity during daylight hours. At around noon on July 10th, polarity flips and they start reading in parallel with the solar production.

I can’t help much beyond that, but is your solar system tied into the panel through a dedicated breaker (line-side solar), or is it tied into the main lines (load-side solar)? That information might be helpful for support when they try to troubleshoot the problem, but it’s apparent that whatever configuration fix was applied did not properly correct the problem. Hopefully they’ll get it taken care of soon!


You’re right - that’s a great observation - I hadn’t picked up on that.

The solar is tied through a dedicated breaker.

Thanks for all your thoughts. Much appreciated.


Same here, so I know it should work. I know it’s difficult to wait, but support will have you righted soon enough :slight_smile:


I recognize that this is an old thread but:

  1. If solar is tied into the panel through a dedicated (back-fed) breaker this is a load side connection.

  2. Line side is when the mains feed to the panel is tapped before the main breaker.

These conventions are reversed in the post. Line side connections are typical when a whole house generator and transfer switch are present. The load side connection is typical when a generator is not present; it accomplishes a safety manual disconnect and overcurrent protection for the cost of just a circuit breaker.


I’ve noticed, on my install, that my usage is underreported by a small amount (15-25 watts or so) when solar production is high. I imagine this is simply a matter of CT accuracy. If the clamps are accurate to within 1% or so, 1% of 6000 watts is 60 watts. Usage is calculated as main current minus solar current, so if the solar CTs are slightly overreporting and/or the main CTs are slightly underreporting - even by 1% or less - and usage is a few hundred watts or less, then usage can be off significantly percentagewise.

This isn’t really a concern of mine given its magnitude; just wondering if others have observed a similar effect.