Usage discrepancy


#1

I’ve had the Sense monitor installed for the last couple of months and now noticed that the overall usage daily/weekly/monthly reported by Sense is not the same as reported by my utility. The sense monitor is showing about 25% less usage than my utility is reporting.

How accurate should the Sense monitor be compared to my utility reporting? A 25% discrepancy is a large difference.

I only have a single main electrical panel and my Internet is very stable. The sense monitor is installed on it’s own 15A DP breaker in the main panel. Is there something wrong with my unit?


#2

does your utility bill on calendar month or based on account open date? that could throw things off… I think most here have reported that Sense is within 1-3% of actual bill… mine was like that for 1 month… but I had another issue and reset my data… I don’t have a complete month yet… you can set your billing date in the Sense settings… I don’t know if that would change what the “month” is on the timeline… haven’t played with that yet…

another thought, my utility doesn’t start/end the bill on an even day either… my numbers are always off from my solar monitor… after questions customer support at my utility, found out they start/end the bill in the middle of the day, not always the same time… so based on the shotgun/hand grenade theories, it all washes in the end… just be aware that some utilities do what they want, when they want and how they want…


#3

Sense technically isn’t a “revenue-grade” meter, but it’s very accurate. A 25% difference is highly abnormal. Have you reached out to Support? There could be an installation issue that’s causing this.

https://www.sense.com/contact


#4

If your utility offers hour-by-hour downloads, try downloading/exporting from them and from Sense, then do a scatter plot of utility (x axis) vs. Sense (y-axis) for the same hour. That plot should inform you of your issue.

  1. It should be a 45 degree (slope=1) line with minor variations (maybe 2%).
  2. If you see a 45 degree pattern, with a bunch of points that are lower on y-axis from the 45 degree unity slope line, you are likely seeing data dropout from Sense, during a number of hours.
  3. If you have a nice line, but it is less steep than 45 degrees, then you have a systemic calibration problem.
  4. If you have an 45 degree nice line, but it appears offset on the x or y axis, then you have some kind of additional feed or drain problem, where Sense is accounting for incomplete (or extra in the case of some kind of feed-in) usage.
  5. Some combination of above…

Here’s what I discovered when I did the comparisons.