Energy

There are two angles to energy use in a household: production and consumption.

Consumption

With the addition of the heat pump added a significant amount of electrical load during the winter months. You can see this clearly by looking at this chart of our monthly consumption:

The above chart includes both energy drawn from the grid and self consumption from our solar system. For more information on that calculation, see the next section.

The system comes with a 10kW backup strip, but honestly we probably didn't use it much (if at all) based on the data I'm seeing. Unfortunately we aren't measuring regular consumption patterns but one day in January 2024 I got curious, went outside, and periodically checked the meter:

That works out to an average of about 1.91 kWh per hour on a very cold day. Between 9:24 and 12:46 (when the weather was coldest), we used only about 4 kWh. The specifications on the heat pump suggest that the maximum output of the heat pump itself is 3kW, leading me to suspect that we just didn't need the electric backup. It's probably good to have for insurance purposes (e.g. if the heat pump breaks down) but I don't think it was necessary otherwise.

Production

To offset the increased energy use of the heat pump, we installed a 6kW solar system. As mentioned, it's a 6kW system with a 5kW inverter. Its peak output is far more than what we would consume at any given time.

In Ontario, we use a system called "net metering", which allows you to resell excess solar power at the retail rate that you buy it (not including the delivery charge, which works out to a few cents per kWh).

We'll get more into the economics in the finance section, but the birds eye view of a system like this is that it produces a lot more in the summer than in the winter. If you're trying to fully offset your electricity use over a full year (not a possibility in our case) you'd size so as to produce enough in the summer to offset your winter.

The other variable here is of course self-consumption: by default, my solar system will first try to power the house. What's left over at any given time is what's sold to the grid. The utility only measures the latter, of course, what we use ourselves is only visible to us.

To get a handle on these variables, I sourced information from two different sources:

• We can get the kWhR values we send to the utility from our utility bill
• The SolarEdge inverter I have has telemetry values which tell you how much it produced over a given hour.

By subtracting the second from the first, we can get a rough idea of how much we used ourselves.

You can watch as our net consumption (total consumed minus total produced) goes down as the days get longer and our need for heating goes down:

It's also interesting to look at "self consumption" on its own. This is electricity directly consumed by our household, without being sent to the grid.

You'll note that only in September 2024 did we actually produce more than we consumed. This is due to two factors:

1. Our system is a little undersized. Unfortunately we ran out of useable south facing roof space! Such is life.
2. We were continuously running an inefficient dehumidifier in the basement until August 2024 which was using a lot of power. Oops 🤦. I wrote about this on mastodon last year.

Solar irradiance

Since the system runs by converting sunlight into electricity, the main factor influencing how much it produces is how much sunlight hits the panels. Obviously, this will vary a bit year over year. We can use the NASA POWER data to get a handle on what the values were for 2024, and then present the counterfactual of what they might have been for other years (where we had different amounts of solar irradiance).

Deriving an "irradiance" ratio

The NASA POWER data provides a total amount of insolation (in kWh/m^2/day) for a given location. We can compare this to the system's actual production to get an idea of how these two values relate.

As you can see, the ratio gets pretty messy (especially in the winter when there are other factors like snow cover at play), but overall things balance out. The mean value over this series was 4.7633. Multiplying this by the total irradiance value for year gives us an estimated value of 6343.67, pretty close to our actual generation figure of 6557.73.

Estimating production for previous years

To estimate the production for previous years, we can take the average of the ratio and multiply it by the total irradiance value for that year.

As you can see, there are notable differences between years.

Breaking that out into averages, we get 6,165.10 kWh as a minimum, 6,859.83 kWh as a maximum, and 6,447.55 kWh (or about 100 kWh below our estimated generation of 6557 kWh for 2024) as an average.

Conclusion

Irradiance can make a difference, but not an enormous one. We have sunny and cloudy days in Ontario, but averaged over a year, the end result is pretty similar. The results we got in 2024 are probably pretty close to what we'll get in subsequent years (subject to panel degradation, of course).