Going on about 1.5 years now with Sense and 75% of my electric load is still not determined. Well - I guess 28.5 - always on, is a determination of sorts but of the 50 devices that Sense has detected, most of them (80%) of them make up <10% of my overall usage.
I’m still running about 47% unknown.
I did some tuning based on the data I have been getting from Sense but now it’s stagnated.
Andrew’s report is disappointing. I’m a bit over two weeks in and Sense has identified only 7 devices. One of them is the dryer’s motor but it has not yet identified the dryer’s heat element. No new devices in the past week. I hope that Sense will continue to progress.
Sense has utility to me as an instantaneous solar output monitor and a total consumption monitor.
I’m roughly the same time in. 19* devices, including heavy hitters like the heat pump, water heater, and stove. During the summer, that represented 75% of our usage. Now? 20-30%.
Sense has helped us recognize a failing water heater and a fridge with blocked radiator fins, saving us time, money, and headache.
The detection curve was logistic - slow at the start, ramped up in the middle, and now it’s plateaued. Because there’s not much left - dozens of low wattage leds, a few ceiling fans, and the garage door opener.
*Right now my hope is that detection becomes more consistent - we still have hit or miss “on” detection.
I’m very happy because, as a solar customer, I now have a very good idea what our yearly ins and outs are, and what to expect on a month by month basis. I’m happy support has improved significantly, and the community here is great - up there with other new tech communities.
You are doing great. I’m a year in, and Sense has only identified about 15 of our 57 significant electrical devices, and only reports about four of those consistently.
Clearly, their technology has improved over the year. We took almost two months to find our first two devices, and even then not consistantly
My disappointment is in Andrew’s 1.5 year experience report, not in my own short period which I know is just a beginning. I am hopeful. And the solar aspect is useful even though I have Solaredge’s own app as a monitor. Sense provides instantaneous data. [disclosure - note edits to my earlier post]
I already had Enphase micro-inverter monitoring and WELserver whole house monitoring (both of which gave me solar data), so when I saw Sense on a This Old House episode a year ago, I immediately purchased it to provide device by device monitoring, at least for all the biggest consumers. When the solar came out a few months later, I immediately added that.
The Sense total consumption and solar generation are close enough to the revenue grade metering I already had. It’s the device by device monitoring that I’d hoped for, but after a year am losing hope.
So, since Sense still doesn’t actually understand most of my devices, I now have three whole house consumption and solar generation monitors, and still not much of any detail. Only three devices consistently (and accurately as far as I can tell), geothermal heat pump, hot water heater, and microwave are being reported…a few others sporadically, mostly not at all.
Sigh…great promises, execution falls WAY short.
For what it’s worth, here’s a plot of my Sense learning history. I’ve reset the unit several times and each of those learning curves is a different color. The legend explains why I reset it each time except for the most recent entry which describes an experiment with my TV. The last reset was 8 months ago. I deleted two devices along the way and the Sense “deleted” one by consolidating it with another device.
Obviously I’m not doing as well as some folks. I think one problem is I don’t have many of the “normal” devices the Sense is good at detecting. For example, I have an electric range but I’ve used it maybe three times since I’ve had the Sense and then just to reheat pizza. My water heater is a gas, tankless model. Furnace is gas as well. My TV is a plasma model which appears to make all detections slightly harder as seen by the last entry in the legend.
Stuart1 - even with the resets, you’re still running at >80% in the always on and other category. Not helpful when trying to tune usage.
Further to my point is that I’m a bit confused how Sense can’t identify some of the big hitters in my usage? Big things like:
EV (Leaf)
Electric Oven (GE)
Whole House Fan
Large groupings of lights (even LEDs)
Really great chart, thanks for sharing. I have the opposite situation, with a 100% electric home and 57 electric consumers not including lighting. In a year I have about the same success you do, single digit.
So far for February after installing late November last year:
13 significant devices detected; a couple others that are unidentified and never seem to run.
33% water heater
14% dryer
8% always on
5% refrigerator
4% range
3% chest freezer
3% oven
1.5% well pump
1.3% electric kettle
1.1% washing machine
0.5% microwave
26% other (unknown)
Of that unknown, a significant fraction seems to be parts of the dryer cycle not fully detected. The rest is a PC, lights, fans, phone/laptop chargers, and small electronics.
Overall, I have to say I’m impressed.
Sure wish we were seen anything like that, what’s your secret?
I guess I am lucky, and I have no complaints, even though there are a few things I wish Sense would find. But I have had mine almost 5 months and so far it has found:
Also, my kitchen fridge was detected in several parts: Appliance lights in both fridge and freezer, Fridge side, Freezer side, and Ice maker.
I wish it would find other things that I suspect are high usage devices, like my other AC/Furnace, the towel warmer, and the infrared heater on the porch, which uses 5000 watts when on (seems like that would be an easy one). However, overall I’m pretty happy with it. If nothing else, it reports my current bill cycle usage within 10 dollars of what the bill will be, and it also tells me when someone leaves at least some of the lights on.
Could all these differences be in the unit itself? Did the actual box change much from when it was released to when I got mine in September? Or is it more related to what’s in each house and what kind of wiring they have maybe?
Excellent question and I’ve suspected hardware improvements (they claim to upgrade software), though many recent purchasers have the same awful experience that so many of us keep reporting…so who knows, and Sense certainly isn’t being forthcoming.
@Mcraeh: I was hoping for results like yours. Very impressive!
Mine’s been in over a year and it’s correctly identified only a handful of minor loads. Still hasn’t identified my two EV chargers nor my pool pump. Granted, I have an “electrically noisy” home with many lights & electronics, some of which are uncommon. But I don’t think this situation is rare for Sense customers. After a year reading about similar results in similar homes my hopes for this technology are greatly diminished and I’m back to recommending killawatt meters and socket monitors to anyone who asks. Sense is extremely accurate as a whole-home energy monitor, but my smart meter is too, and it was free.
Beyond the challenge of device identification, I also see issues with device-level energy consumption measurements. The few times I’ve been able to compare Sense energy measurements to a killawatt reading or a plug load monitor, they are quite different. I fear Sense often misses the “off” signature and then somehow guesses.
Have others done these comparisons and found matching results? Not for a toaster or garbage disposal, but for something with a more complex energy profiles…?
Maybe I should add that I do not have any EV’s or anything like that. However, since my Always On is so high (averages over 700 watts) I have plenty of other things to clutter up the landscape. I wonder if it’s easier to find the little things if you don’t have a big one, like an electric car.
As for comparing individual device usage to what Sense reports, I could give that a try, but how do you do that? I have a Killowatt, but will it keep track of usage or is that strictly real time?
My Kill-a-Watt tracks cumulative usage, and is fairly accurate
My typical week after 1 1/4 years:
Other - 29.9%
EV1 - 26.2%
EV2 - 20.3%
Always On - 11.8%
Furnace Fan Down - 3.5%
Dryer Heating Element - 2.8%
Backyard Halogens - 1.2%
Wine Fridge - 1.1%
Instant Water Heater - 0.7%
AC Down - 0.5%
Central Vacuum - 0.3%
Microwave - 0.3%
Main Fridge - 0.3%
Motor 3 - 0.3%
Toaster - 0.2%
Toto Toilets - 0.1%
Garage doors, Trash Compactor and Garbage Disposal stetted but below 0.1%
OK I cheated a little on EV2 - I used a Tesla app to look at the power used for changing on that one, but EV1 was detected. Still missing washer, dishwasher, the rest of the dryer, plus a couple missing fridges (plus I thin they are underreporting).
I’m only 5 wks in but still wish I had more BIG users ID’ed. I’m running
Other 69.3%
Always On 20.4%
Mini Fridge 4.9%
Dryer 4.3%
Stove 1 Burner 0.4%
Microwave 0.2%, Coffee Maker 0.2%, Washer 0.1%
This means only 10% of my load is ID’d. Sense has ID’d 25% (11/44) of my listed appliances but they tend to be the smaller loads. None of my BIG users; EV, Stove burners & Ovens are tagged, even though these are all 240V appliances. I’m glad to hear that it has ID’d one of your EV’s. Maybe mine will follow shortly. Curious, as mentioned in other posts, the Sense is unable to ID the start-up load of my Kuerig coffee maker, used twice daily. It only tags the maintenance heater cycle; 475w/5 secs every 2 mins. I expect the total load is 10x the maintenance load.
I expect to work on my “Always On” by putting them on timers to remove them from the nighttime load, as done by @kevin1 and possibly move their use to “Other”. It’s my hope that that may move them from all the background noise & other devices will be found.
@andy Possibly having a relatively noise-free electrical system: small house, no TVs, few always-on electronics.
Be glad it’s doing as well. After a year, our Sense has only identified one of the items (microwave) you are tracking and only about three more devices reliably.
