I have had Sense for about a month and a half. This month was the first full month that I can look at my utility bill and compare to Sense. Basic question is how accurate is Sense. My utility bill says I used 1817 kWh. Sense says 1772. Is that within the accuracy range…97-98%? Just curious…Thanks!
Your accuracy vs utility looks a little further away than most. Usually within +/- 1% Other peoples’ experiences and measurement techniques here.
It shouldn’t be that far off. Are the clamp meters connected securely and in the right direction? No chance they’re loose or you have a cramped panel and it’s picking up interference?
My utility portal has daily and weekly usage data rounded to 1kwh. The measurements reported by my Sense almost always round up or down to the exact kwh my utility reported. (unfortunately, I cannot compare it to my bill because my utility has variable billing periods)
It depends on your utility.
My billing cycle runs anywhere between 27 and 35 days. My meter is read mostly bi-monthly so most every other month it is estimated. There is no time of use zones. My bill is based off the average daily rate for the billing period.
It is near impossible for me to compare to my utility bill with any sense of accuracy.
I could create a spreadsheet, read my own meter and compare it to the cycle date I set in Sense to verify accuracy of the meter. Last reading was 3/22 Next readying will be 5/19. No idea when this months estimated date is.
I’m not sure I understand what is going on. I am attaching what my utility is saying vs Sense. Doesn’t seem very accurate. I was told I have a 300 amp split circuit. The panels are next to each other in the mechanical room. I have 2 additional sub panels that these main panels feed. I must be missing some electricity somewhere right? Thanks
If it weren’t for the last reading, I would opt for the missing panel or open CT hypothesis. But Sense greater than utility, even once, kind of quashes that. A couple questions for you:
- Do you have solar energy and Sense solar ?
- Do you have a smart meter ? Are all your reads actual or are some estimated ?
I would also check in with firstname.lastname@example.org. Sometimes they can see issues that aren’t obvious.
Thanks for the suggestion about sense support. I do not have solar energy. I have a smart meter (FlexNet). I can get on the utility website and get hourly/daily/monthly readings. I checked my 2 sub panels by switching them off and the sense wattage dropped so I know they are being powered by the 2 main panels. Unless I have another panel somewhere that I don’t know about I’m not sure what the problem could be. Thanks
If you want to do some more serious sleuthing, I would suggest comparing hourly data, just to have more data and to potentially narrow issues by time. It would also be useful to compare data for the same day/hour in a x-y scatter plot (no time axis - just utility power vs. Sense power).
Lots of pointers here.
Do you have the Sense solar auxiliary CT’s (used for 2nd panel)? If you have a class 320 service, (2 - 200 amp panels), then you need the additional CT’s to meter the 2nd panel since it’s independent from the main panel.
That’s a really good question @phil435 ! I didn’t give it a second thought because I have a similar CL320 service, but put my Sense CTs on the pre-meter mains, because there was no other way to do it at the time (and still isn’t for me since I have solar).
@whalleye, how are your CTs configured to obtain the data for both parallel panels ?
Yes. I have CT sensors for both panels
@whalleye Have you verified that the CT’s are oriented the same way on both panels in relation to the incoming source for all CT’s? Also, I assume you set the aux CT’s set to 400 amp split service?
My app says that 400A split service is enabled. Have not checked the CT orientation. An electrician put them in. I thought if they were oriented incorrectly it would be off more than it appears to be.
I’m not sure what sense does on the backend if one CT out of the 4 is 180 degrees out of phase. They may automatically correct this with software on the backend. Typically to be in phase with the voltage, current flows in the dot on the CT from the source. I can’t remember if the Sense CT’s have a dot or just a sticker. You could also be out of phase with the voltage but I assume Sense detects this on the backend and corrects this as well.
FYI - I did some early tests with 2x200A accuracy. I did find that accuracy was sensitive to CT closure. But I don’t think that Sense will come in higher that your utility with one or more partially open CTs, only lower.
When I download hourly data, do I sum up all the kWh or just look at the total usage line each hour. Thanks!
I total the mains line to compare against the electric meter. I don’t have the 400 amp dedicated so I don’t know what the export looks like for this but I assume it’s separated out. I assume there is a line for the total for both? Maybe @kevin1 can chime in.
No need to sum up each device.
Just Total Usage for comparison against the utility.
I just ran a report and looked at my utility bill. My bill says I used 1817 kWh and Sense said 1603 kWh. Pretty big difference. I’m uploading 2 graphs. One shows Sense vs Utility and the other I broke down the number of hours in the month that fell into different “accuracy” buckets. Not sure if this is useful or not but right now Sense is significantly off during many hours. I checked the CT’s and the sticker is facing up on all four of them and they seem fully closed. They are loose and laying on other devices. Is it ok if they touch other wires as long as they are around the mains? I would think if it were an installation issue it would be more consistently off. Starting to wonder now if there is another panel in my house. Also interesting that 3 hours during the month sense showed over the utility by 150-190%.
Thanks again for any input/advice. I really appreciate this community.
Cool and weird. Try the utility vs Sense x-y scatter plot with the data points colored by time of day, and maybe day of week. You might see a pattern that gives more hints. You might also want to try binning the difference by kWh difference vs. percentage. That would help see if the actual error is balanced in kWh between above and below 0.