Sense vs Solar app: kWh differences

Technical Questions
1

We just had our solar system activated about a week ago and Tesla has there app that shows use and production. I also have a sense monitor on both the main panel and subpanel… grrrrrr thats another topic. I have the Solar CTs but haven’t installed them yet.

Has anyone experienced what Tesla shows for use vs what Sense shows doesn’t match? I do. Not sure yet what Tesla installed that can read usage but they are seeing things off each panel its just not the same as Sense. I can turn things on and off and Sense picks it up but the Tesla app doesn’t. As of right now Sense power meter reads 371W on one meter and 628W on the other. Tesla app shows 0.4kW. Not a match.

Right now we are producing but per Tesla not fully activated which means I don’t have daily usage vs production. So no history. Just log in and I see live data. They say things may be off a bit until fully activated. Honestly, I don’t buy that.

Any insight would be appreciated. Thanks!

2

We need to understand your system layout to be able to comment on how it’s working. Can you post a simple diagram showing how your panels are laid out, where the solar ties in, where the sense monitor(s) are tied in, and where the tesla monitor(s) are?

3

I don’t have Telsa Solar (lol, when I was solar shopping, they turned me down, I didn’t meet their requirements, utility bills were too low and my credit wasn’t good enough), but my SunPower Solar has an optional “consumption kit”… for $45 they add 2 CTs on the panel mains and the CTs plug directly into the SunPower Supervisor/Monitor… now, all it does is report what is exported and imported from the utility grid… simple consumption… but it doesn’t give any details other than overall import/export numbers per 15min… I knew I was going to get a Sense Monitor, so I didn’t opt for it…

now, I can give an opinion on running Sense without solar CTs installed and having solar already… I was in this situation when I first installed Sense… I had too many cloudy days and wasn’t getting enough solar production to complete the solar CT setup/calibration… this really screwed up all my numbers… production, consumption, live use… they were all off… things weren’t adding up right… so I wouldn’t advise trusting any numbers from that setup until you can setup your solar CTs…

now, with all that said, my experience of the numbers I get… I have production numbers from SunPower, production & consumption numbers from Sense, and consumption, export, & net numbers from my utility (Southern California Edison in my case)… so far, none of them match exactly… they get close, 1~3% or so… as I said above, I didn’t have that accuracy until I got my Sense setup/calibration completed successfully… to use a military analogy, my current accuracy is not as exact as a sniper rifle, and its not as loose as a hand grenade… but it can be compared to a shotgun… gets close enough to be effective… given that the Sense and SunPower numbers are not “meter” grade, it works for me… I have a few explanations of why the number are not exact (besides not “meter” grade):

  1. Sense Monitor takes readings every second (like my utility meter, but no average)
  2. SunPower only takes readings every 15min on wifi and only 1/hr on cellular backup when wifi is down (this is more like hand grenade style… only gets close… no chance of being accurate with average reading intervals like that)
  3. Utility Meter reads every second, but only sends average readings every 15min (now you could dispute this all day, but reading the fine print from my utility bills and utility rate sheets, the meter averages use for 15mins, then transmits that average to utility servers at the end of the 15min average)
  4. Utility Meter has 15min readings, but data download from utility is only 1/hr (so they are averaging again somewhere to turn 4x15min reads into 1/hr total)
  5. (forgot to mention) my utility bills monthly as most due… but the day and/or hour they close one month is not always the same… for instance, one month may be on the 11th, but the next might be on the 10th or 12th… but it gets even worse, when they close on a particular day, the hour of the day is random also, depending on work load and system status… so they are using hand grenades, but it all totals out right somewhere in the end, lol

so 3 different ways of calculating the same figures… for my situation (yours maybe different), these numbers will never be the same… just close… most here are seeing <3% differences, I can live with that…

for me, my solar PV system was sized to eliminate my utility costs with exception to the service charges… in 4-5yrs, my plans are to refi my house and get batteries… technology willing, I will remove myself from the grid… my purpose is not for savings, that’s the byproduct… here in CA, we are forcing PV installs on all new residential construction and additions over 1000 sqft begining on Jan 1 2020… all new non-residential construction (commercial) will be required to have solar beginning 2030… both of these have been publicized well… what isn’t as public is by 2047, all existing homes must be 50% below 1990 emission levels… and Dems want to accelerate that to 2040 I’ve heard… so for me, its all about self preservation… my goals are to make sure that my family can stay in our home without any chance of losing it to the high expenses coming down the pike…

Sense gives me:

  1. a way to monitor and reduce energy use
  2. insure I stay within my PV generation budget
  3. and continue to “energy net zero” so I have no one coming into my home, forcing me to do upgrades at high costs

and, as an Energy Consultant, I can use my setup as an example for others to follow… it puts more weight to the way I advise others…

4 Likes
4

I have an Enphase system and there is a difference in generation between what Enphase says and what sense says. My enphase reports more daily (generally around 1khw) in comparison to sense. I attribute some of that to sense running -5w from solar when the sun is down. Power flowing back to the enphase monitor. You can see my solar generation via the Enphase app below.

Now that I’m thinking about this, I can export the data out of Enphase and out of Sense. Might be interesting to do some visual comparisons between the two. I’m thinking that’s going on the to do list!

2 Likes
5

I have a Tesla (nee SolarCIty) system that was installed and activated in 2013, before my Sense. My 4.2kW system has a single SolarEdge inverter, that outputs energy production to the web and for download. After some analysis back in Aug, 18, I saw a few interesting things in comparison to Sense production data.

  • My SolarEdge/Solarcity results were systemically 3% optimistic compared to Sense results
  • Using the Sense solar data gave closer correlation to my revenue power meter, than the Solarcity data.
  • There are differences between the time interval accounting between Sense and Solarcity, so a scatter plot on an hourly basis gave sort of an ellipse, rather than a straight line. Easier to compare on a daily basis.
  • Both Sense and Solarcity data are affected by dropouts due to network outages and other issues, so a number of “bad” data points.

https://community.sense.com/t/does-anybody-have-any-cool-graphs-or-analytics-using-sense-data-please-share/3302/4

6

After about 100 days of operation, Sense is reporting about 2.1 percent higher solar production than my SolarEdge inverter. Hourly production numbers confirm Sense is higher. The inverter idles/consumes at about -4 watts at night.

Also, Sense is reporting about 1.7 percent higher kWh used than the utility since August 2018.

I confirmed the clamps are secure. I opened a case asking about these results.

7

I am seeing the same with the my setup, a 1Kwh difference between what enlighten (enphase) and sense reports, sense is lower by that amount.

8

1kWh difference over what period ? A day ?
You might try to plot daily production on Sense vs. your solar inverter output.

1 Like
9

Did you ever sort this out ? Did you ever plot daily solar output from Sense vs. your inverter output ?

10

I plotted the solar-generated values reported by my Utility, Solar Company, Inverter and Sense. The Inverter and the Utility Company were very close on daily values. Sense and the Solar Company consistently report higher with Sense reporting about 2.1 percent higher.

I stopped tracking the issue after Sense support says this discrepancy is typical of a “non-revenue grade” monitoring device.

I’ll revisit the production numbers after my first anniversary of solar production later this month.

1 Like
11

Interesting that you have different data from your Inverter and your Solar Company. I only have one set of data from my Inverter via my Tesla/SolarCity. I was seeing something like this:

Solar
Solar.png776×776 25.9 KB

After removing the few outliers, mostly due to connectivity and Sense monitor issues, I’m seeing the linear relationship below. Sense daily reading is 0.9668 * Tesla reading plus 141Wh.

Coefficients:
                (Intercept)  Sense3$TeslaSolar[keepers3]  
                     0.1412                       0.9668

The fit to that line is pretty good, 80% of the time within +/- 100Wh

Histogram
Histogram.png776×721 8.32 KB

That means I seeing maybe a .7kWh difference on a 25kWh production day.

2 Likes
12

Similar results here. I’m seeing ~.5 to 1.5kWh daily shortage on Sense compared to the inverter’s report. The inverter and the utility are very close each month.

1 Like
13

Yesterday, my Enphase inverter reported 47.73 kWh produced and Sense reported 46.10 kWh produced. Enphase reported 52.71 kWh consumed and Sense reported 52.70 kWh consumed. Consumption is spot on, but Solar is lagging 1-2 kWh each day. Haven’t gotten a bill from my power company since activating solar, so can’t compare that yet.

14

It doesn’t mean a whole lot, but from what I see in Enphase specs the I-THD of the AC output is roughly “2.6%”; V-THD is “0.03%” (for the IQ 7AS for example).

The Enphase IQ Envoy has ANSI C12.20 +/- 0.5% PV production metering and +/- 2.5% consumption monitoring.

Meaning, when you say Sense Solar is “lagging 1-2kWh” each day, to be fair I’d probably halve that figure and attribute some of the (to-be-expected) error to Enphase.

15

There are a couple issues that can cause this:

  1. An open CT sensor
  2. All solar production isn’t being monitored
  3. The monitor is falling offline

I would encourage anyone dealing with issues like this to reach out to Support and they can take a closer look.

1 Like
16

The AC from micro investors goes through a production (solar) CT in Enphase just like Sense solar CT. It’s the same wire. Good thought…

I’ll send it to support after they finish working on a “missing Other” usage issue I and some others have.

17

Part of mine is the Envoy monitoring system I am sure. Solar CTs run at a solid -5W when the sun is down and I’m sure that is powering the Envoy. Granted that is only .12kwh assuming that draw is the same all day.

18

I was hinting more at, say, a 2% difference between Sense and Envoy could be seen as “Sense is 2% off” in it’s measurement or that both Sense and Envoy are off by 1% … or something along those lines. Giving credit (& debit) where credit (& debit) is due!

19

I have my PVs online now and this morning. Half my panels aren’t operational due to a polarity wiring issue that has been corrected by they are not online at the moment, so I am running at half solar production. Currently, my solar is far exceeding my load right (3.6kW vs 1kW) now so I am feeding back into our power company (net metering). Sense isn’t identifying my Always On nor other while I am in this state even though the trend identifies that there is a load.

image
image.png1450×975 31.8 KB

image
image.png221×809 6.69 KB

Ideas?

20

By the look of those numbers your system is waiting to “re-calibrate”. Your main CT load is registering the net vs load.

Or you might be seeing this

Look through that thread. Either way, probably a support issue.

1 Like