How accurate is sense

I’ve been watching my usage since i installed sense two weeks ago and i have two weeks till my meter is read again for me to compare how accurate sense is at reading power usage or how accurate my power meter on my house is. I’ve put in the correct cost per kWh and so far i am either going to have my best month with the lowest usage over the past two years OR sense is reading incorrectly OR my meter is reading incorrectly.

1 Like

oh and obviously i am keeping my fingers crossed that its my best month and everything else is accurate.

Technically, Sense is not a “revenue-grade” meter, i.e., it’s not meant as a replacement for your utility-provided meter. But it’s very accurate.

Keep in mind that billing cycle dates can shift, so your billing cycle usage might not match up perfectly. I was off <10kw last month because my cycle has some weird variances in it, but when I did the math for the actual days, Sense was spot on.

1 Like

The meter itself is pretty darn accurate (assuming no outages). Most people on the forums have reported between 1-5% diffrence on usage compared to the power company. My best month I was off by 3 kWh. My worst, 50kWh, but that was due to a network issue in the house.

Keep in mind that the price you enter per kWh may not be all that is needed to compare your bill. For accuracy, I would look at kWh reported by Sense vs the power company.
I’m in MA, so no tiered or TOU pricing, but there are about 8 different “per kWh” items on my bill plus flat monthly charges.

For my kWh number, I took the total bill divided by kWh to come up with a blended rate for what the cost is per kWh.

Check out this thread for some examples.

when i look at my bills they show the meter read dates between the 4th of the month and the 9th of the month so i know its going to change slightly each month and with longer and shorter months as well. my bill says my next meter read date is on the 8th of March but looking at what sense is telling me and knowing that my wife is finally starting to help me turn off things that we aren’t using I think i am on track to having my best month so far. I am also working on installing a radiant barrier in my attic to keep the temperature down up there in the hopes of lowering my usage during the summer as well.

I already calculated using the method that you posted as well as calling my electric company and they gave me the exact same rate of $0.11 per kWh that i calculated using my bill. I actually just read the thread that you linked and that’s what got me thinking about how accurate my meter is and why i posted this thread. this thread was basically me just thinking out loud on the forum.

1 Like

MA too! It’s a pain to get it all those lines represented as one fixed cost. Of course, one of those lines was higher last month for whatever reason, so I had to go back and adjust my rate. Yay :neutral_face:

I’ll dig into that thread. Must’ve missed it the first time around.

Yeah. It is a rough cost. I hadn’t updated in a while and just found that i was 4 cents off. I had been at 21c kWh but had to increase it to 26c per.

I went from 22c to 25c!

Also, since you’re in MA, keep watching the announcements forum. We try and do some local user meetups occasionally and I’m hoping to do one again in the Spring.

I get daily texts from my utility showing previous days usage. Sense is usually off by 2-8%.
If the utility says 78 KWH, sense will say 83.
Sense has been pretty consistent being a touch higher. It’s not enough to worry about for me and I like it better than the other way around. The way it’s going my bill is a little lower than sense predicts.

1 Like


To give you a quant answer, here is my distribution of differences vs. my utility company’s readings on an hourly basis, for 9360 hours (equivalent of 390 days). I have removed all hours (~700) with data gaps on the Sense side (some caused by network errors on my end, some likely by the mysterious data eating beast at Sense).

The mean of the difference between the two is 0.0281574 kWh (28Wh) per hour.
The standard deviation of the distribution is 0.2098271kWh (210Wh)

Not too bad when you consider my mean usage per hour is 1.751526 kWh. That’s an average error of 1.6%.

ps: The mean hourly difference goes to 0.07086072 (71Wh) and std. dev. goes to 0.5362477 (536Wh) if you include the hours with data gaps.


That is pretty accurate indeed

thank you