Yeah, but give us a guideline. Give us some damn graphs and charts so we can figure out a ballpark.

Outstanding, thanks

I feel you on not knowing where the break even point of heat pumps is! It's so frustrating that that isn't published.

Well it would vary for every house, season, amount of sunlight, etc. It'd require extensive testing and data collection to determine.

I believe this is it: https://emporia-vue-local.github.io/docs/tutorial/intro/

I saw reviews where the clamps measured as much as 40% high. If my heat pump is 75% of my usage and it read that high, then when I look as a percentage of total use, it will be over 100% and my “untracked” usage will be invisible

I agree it’s expensive, but the emporia one someone else mentioned has more meters on one device. I do have smart switches which directly measure power consumption so lights are covered. I know there are outlets so I could probably cover things like a freezer if I wanted to. But there’s really only a handful of devices which use noticeable power

Yeah, I have mine run if I'm home via location of my phone, then it adjusts room-by-room based on where people (or the Cat) are. It's pretty much completely automatic based on a bunch of variables, I rarely have to adjust it manually.

I feel you on not knowing where the break even point of heat pumps is! It's so frustrating that that isn't published.

Our Ecobee thermostat does detect motion and can, in theory, automatically let the temperature drop if no one is home, for example. I tried it when we first got it, and I found that it must have been detecting our cats, so it was activating the heat when no one was home. I turned that feature off and went with the schedule instead.

Also, we have a heat pump with electric heat backup, so it may actually be cheaper to keep the heat constant, if it avoids using the electric heat. Unfortunately, I have no idea where that break-even point is (i.e., at some point, it's likely cheaper to kick on the electric heat to bring the house up to temperature, versus running the heat pump for many hours - but I don't know where that point is - and it certainly varies based on outside temperature, sun hitting the house, etc...).

These days, my schedule is extremely variable, so I just leave the normal schedule in place most of the time. I wish Ecobee made it easier to go, "Hey, today, I'm going to be out from 10-3, so let the temperature drop during that time." (As it is, I have to adjust the permanent schedule each time, or I can change it as I'm leaving the house and tell it to resume the normal schedule in a few hours. But I forget to do it.) Home Assistant isn't great with that kind of one-off thing either, unfortunately, as far as I'm aware - you pretty much have to set up an automation to run.

Shelly EM Gen 3

It's been a little while but I did follow a guide for the emporia vue.

There are newer versions and other similar brands out there if I find the guide I used it will be fairly specific to this hardware (certainly the base code setup)

I will look tomorrow

Ahh, yeah, I guess I haven't compared it to my actual bill yet. I should do that. Have to wait a few more months for data to pile up as I just moved in. I will say, I don't understand why it matters if it perfectly matches the utility company, what are you exactly trying to do?

And yeah, two sensors. I put it on the lines into my house as my main concern was going over 100a in the whole house, as my old house doesn't have a main breaker. I do see now that you wanted to do a device-by-device breakout so that would require buying a fair few of these, and there's gotta be a cheaper way.

That’s an interesting idea - is there a guide for newbies? I’ve never done anything with esphome.

I had been expecting to use the total power from my utility but perhaps that’s a better idea: measure total power with the same tool as individual.

Right, a big part of the goal is to identify whether untracked is significant.

Maybe accuracy is not the right word: I see your point that it’s likely accurate enough. My concern was consistency.

The total power usage as reported by my utility is measured one way and these measure a different way. Any error compounds, and the goal is to graph them against each other. Is it enough to make an impact?

That specific product has two sensors. That means only one 240v circuit, right?

Yeah, my thermostat automations and scripts are crazy complicated. They also involve location and motion of course.

I have a ludicrously complicated script that sets the temperature for my pellet stove, using a generic thermostat (using a zigbee temperature sensor, and a Shelly switch to control the stove).

For example, if the temperature outside is above 50 (F), set the pellet stove temperature to something really low to turn it off. Turn on below 45.

If the electric heat kicks on (from the main hvac thermostat), set the pellet stove temperature high - I want that running if the electric heat is on.

I think it aims for 72 during the day and 70 at night, or something like that.

Oh and the ceiling fan nearby kicks on if the pellet stove is running. As does a vent fan in our bedroom.

https://a.co/d/0gDkYwiB

I have this thing, if you didn't find it in your searching. It seems accurate enough to me, but I'm not sure how I (or any other reviewers) would test that for certain. Guess I could turn off all but one circuit and put a kill a watt on a microwave or something. Whenever I've checked the graphs the readings make sense to me based on what was running at the time.

I have succesfully used a cheap tuya meter from aliexpress (white box with two clamps and two wires for voltage). With the official tuya integration you can monitor one device (or whole hose) but if you are willing to jump through some hoops and using local tuya you can get both measurements separetly. I also have a cheap aliexpress DIN style tuya smart breaker that works fine with tuya integration. And finally I have used a tplink kasa smart plug hs110 for 240V measuring (it is advertised for 120V in north america and it has 120V outlet. All of this in 240Vac 60Hz.

Accuracy for monitoring purposes is fine.

I have tons of window sensors tied into my automation and it works mostly as expected. Sometimes it doesn't trigger on close and reports a window/door open that isn't or a battery needs swapping. Saved me a few times from heating the outdoors.

the minisplits I have don't have an actual thermostat. I use esphome controllers to control each one via wifi

There’s a “Generic Thermostat” helper entity that lets you point to other sensors and switches and then you can use all the thermostat features in HA.

Unfortunately, the minisplits I have don't have an actual thermostat. I use esphome controllers to control each one via wifi

I actually already had that blueprint. I'll take a look thanks!

I just use the built-in scheduler on my thermostat, it lets me set different temps for different times of day and days of the week, and it's easy to adjust when the weather changes.

Home assistant could be another option, but theres a decent amount of setup involved

I mostly just have automations based on open windows and what mode it's already in (cool for summer, heat for winter), and it's been fine. However after seeing the other comment I'm going to check that out

I use advanced heating control and a cheap zigbee thermostat. It's a pretty good budget solution with a familiar interface for non-nerds in the house.

https://community.home-assistant.io/t/advanced-heating-control/469873

Shelly pro EM50 breaker box monitor, or similar monitor system. Sometimes cheap, sometimes a few hundred USD. They're are many other similar models. Can be even installed before the breaker box to monitor the entire home to check for independent electric usage against your power company bill.

I have two 150a panels in my house and I have 2 emporia vue 2's one on each panel. I reflashed them with esphome so I own the data, it's not going out to emporia. but they work as individual devices collecting data from the clamps.

For non 240 loads you could also use smart plug with power consumption if you're just wanting to track that device. In my use case I am capturing all usage from the two big clamp at the top of each panel then each clamps usage which gives me an 'untracked' value also since I have more than 16 circuits in each panel I cannot track every circuit.