Sense vs. Billing- Big discrepancy I'm having trouble understanding

I’ve slowly been discovering facts about a discrepancy over the past few days. I started discussing it in the “how have you kept your energy bill manageable” thread, but I’ve discovered enough that it probably deserves it’s own thread.

I’ve been using Sense for a bit over two months, and just tried to reconcile it to my bill. I expect some minor discrepancy as my billing cycle is based on whichever day the meter reader comes around, at whatever time they show up. Sometimes it’s a 32 day billing cycle and sometimes it’s a 28 day billing cycle. We don’t have smart meters yet, so I just have a total monthly number.

My July/August bill reads 726kWh of usage.

Sense tells me that over that same timeframe, my usage was a total of 634kWh, and my solar system generated 341.4kWh. So I would expect a net usage of ~293kWh.

Things I’ve checked/done so far:

  1. I keep a close watch on my usage, both at the total level and the individual device level. The total is consistently the sum of the parts. Individual devices also seem to measure roughly in line with what I expect based on their labels (a coffee maker uses ~900W while on, a string of 7 LED lights uses ~60W, etc.

  2. I’ve had a few devices on smartplugs that are detected. Sense and the smartplug consistently report back the same value, to within 1W.

  3. I’ve compared my Enphase reporting to Sense’s solar reporting. I’ve found Enphase to report ~10%ish more power generation than Sense does. The outcomes were very similar to the other recent thread on Enphase Enlighten. This could explain a little bit of total variance, but not major amount.

  4. The one exception to point #1 above is my Air Conditioner. We’ve been using the AC a lot in the last month. When I look at the power meter, I only see ~700W jump in usage when the AC kicks on, and no other unexplained major loads kick on while it’s running. Based on the energy rating on the AC, I would expect a ~5,000W load. The ~700W load (which hasn’t been detected yet) can be explained as being the air handler. The AC is clearly working, as we have a cool house.

Maybe I’m missing something else, but the only explanation I can think of is that somehow the AC compressor load is bypassing the mains where I clamped Sense.

Our AC does have a “Load Control Switch” on it. We don’t subscribe to the load control program, but I assume the prior homeowner did. I don’t know how this could impact the load on the main, but maybe it’s relevant?

Any ideas?

Hi @brian.chatagnier, I put together a checklist of ways to check accuracy here. The challenge is that most of the latter, more accurate techniques, are not possible without a smart meter. You still have a human reading the meter so you have to hew to the periods and readings they deliver if you want to do a comparison.

One question for you since you don’t have a smart meter - does your meter even run backwards when your solar is producing ? Many meters don’t run backwards - we had to get a special net meter when solar was installed.

1 Like

Thanks Kevin! As someone who professionally dabbles in numbers and statistics, i am impressed with your thoroughness.

I am less impressed with the thoroughness of my utility (Xcel). I have verified that my meter does run backwards, and it has “Net” labelled on it, and has a highlighted “Net” in the display. I think that confirms my solar is working and is being measured.

Unfortunately, my meter display doesn’t have a space for the current load, so I can’t do a quick check. It just displays total kWh to read, with no decimal places.

I think the best I can do is to look at full day measurements of kWh used, and compare days that I use the AC with days I don’t.

On the brights side, I’m one of those weird people that enjoys digging into these types of problems and figuring this out. Now I’m curious, although I would like to get it resolved. I’m also scheduled to get a smart meter in late 2021 or early 2022, so I’ll have some better data sources in a few months.


I have one other hypothesis, but this might be based on a novice/flawed understanding of how electric panels & Sense work.

My main panel only has two breakers in it; one for Solar and the other for the AC. The entire rest of my house is on a subpanel.

Is it possible that all of my house except for the AC is on one “side” of the 240V mains, and the AC is on the other? Could there be an issue with the Sense monitor connected to the AC side?

I’ve looked in the panel and both clamps look fully closed. I moved one and re-clamped it just to be sure. It looks good to my novice eye.

I’ve looked at the “Power Quality” graph, and I can see good readings across L1 and L2, so this implies that there are some readings from both sides.

Any thoughts?

A picture of your main box and the CTs might be helpful. There’s a chance, based on how your main panel branches to the subpanel that your “mains” CTs might only be seeing your AC. If the split between your main panel and the subpanel with al the other breakers is upstream from where you placed your CTs, you likely have parallel subpanels. That’s what I have, I had to place my mains CTs before my meter in order to catch the house load before it branched off into the two parallel subpanels. For me, there’s a busbar that goes between the meter and the split to the two parallel sub panels that I couldn’t fit the CTs to.

Problem identified! So I opened up the panel to get a good picture, and took it in with a fresh set of eyes.

Those big massive mains I clamped? Those were the cables to the sub-panel. The smallish cables in the back I didn’t even notice before? Those are the mains. I’ve just been measuring everything wired to the sub-panel (which is everything minus the AC and solar).

I still wasn’t able to fix it, as the mains are physically too close together to get the clamps around. I’m not comfortable with the safety aspect of trying to adjust those mains myself.

Thank you Kevin for your assistance! I wouldn’t have figured it out without you.


Glad you solved the mystery ! It will be interesting to hear how the comparison works out once you move your CTs and get recalibrated. I would check in with to see if you need to go through setup again, even if you transfer the CTs with the exact same orientation.

1 Like