The power of summer

Summer has a lot going for it—in a normal year. One of the many things I like about the season is that (in a normal year) my solar panels generate a lot of electricity, but was that dramatically different this year? Naturally, I had to look at the numbers to see what happened in this year's non-normal summer. That took a lot longer, and a lot more work, than I expected.

I’ve been working on this post for the past three weeks, each version different from the one before. It began because I wanted to see how the storms in January and February affected my power generation, but then I wanted to to know, precisely how well the system was helping me meet my only goal, namely, to tread a little more lightly on the planet and to live a little more self-sufficiently, as Nigel and always dreamed of doing. I think I finally figured out how to explain it.

It turned out to be difficult to work out the clearest and most accurate way to talk about how the solar panels perform. After several attempts, I finally worked out that I need to be able to answer these questions:

1. How much electricity is created by the panels?
2. How much of that electricity do I use myself? and so,
3. How much power is sent to the grid?
4. How does the production affect my power bills—how much electricity do I still have to buy?

The chart at the top of this post answers the first three questions for the first two summers the panels were in operation (the system was installed and switched on after Summer 2020/21 was over; in New Zealand, summer is from December 1 to March 1). The chart shows the results for each month of those summers, and the total amounts for the seasons.

Next, let’s look at how much electricity I purchased for all three summers I’ve lived in this house:

This chart shows how much electricity I was billed for in kilowatt hours (kWh). There’s an important caveat here: The billing period is from the third week to the third week of every month. That means that the bills for each summer month were actually for the period from the third week of November to the third week of February. Moreover, the meter reading wasn’t necessarily taken on the same day. Meanwhile, the data on my production is for calendar months. So, there isn’t a perfect alignment between the data for my generation and for what I was billed for. To get that, I’d have to put the daily data in a spreadsheet and then match it up more precisely with the billed periods.

Even if I did that, there’d still be one more problem: There’s a mismatch between how many kWh of electricity the power company credits for having been supplied to the grid by the panels, and how much the system records as having been sent to the grid, and that can be for more OR less electricity than my system says it sent. I have absolutely no idea why the numbers don’t match up.

All of that maters for the next chart, which ties to work out approximately how much total power was used at my house over the past two summers (with 2020/21 shown to compare months):

Putting aside the imperfect match between monthly periods of time, what this shows is that my approximate total power consumption—the solar power I used myself combined with what I bought—is actually higher than it was before the panels were installed. Despite that, what I actually paid less than I did before the panels were installed.

The lower bills are because of a simple fact: Every solar power kWh that I used myself is a kWh I’d have to buy if I didn’t have the panels. It’s not just that the power I used was free (thought it is), it’s that the price I pay to buy electricity is higher than the credit I receive for power I send to the grid. For both those reasons, it makes sense for me to use as much of the power I generate as I can. As I’ve talked about before, I do power-hungry things—laundry, running the dishwasher, even charging batteries—in the daytime. I also can run my heat pumps 24/7 to keep my house at a constant temperature and still pay less than I did during my first summer, before the system was installed.

Overall, the data shows that over summer I’m buying a bit more than half as much electricity as I did before the panels went online—in fact, in that first, pre-solar power summer, I bought nearly twice as much power as I did in the following two summers combined. So, yes, I’m treading a little more lightly on the planet (though I can do better still, of course).

Which brings me back to the question that started all this for me: How was the system’s production affected by all the storms? These two monthly charts show that:

This underscores what I said at the time: Cyclone Hale wasn’t too bad in Hamilton—roughly like any other rainy day—and the Anniversary Weekend storm was much worse. Cyclone Gabrielle, however, was the worst of all. There were still good days for solar production in those months, but bad days can be very bad.

In each of the past two summers, February was less productive than either December or January, and that makes sense: Bad weather is most likely in February, which is in cyclone season. Even so, production is always highest in summer, and lowest in winter, due to the longer days of summer (more sunlight). Winter is usually rainier/overcast more than in summer, too, further reducing the amount of power that can be generated.

That’s my swim through the pools of data about how the solar panels perform. The main thing I found is the most obvious thing: The panels are helping me to achieve my goal and to live a piece of the dream that Nigel and I had. That’s a huge win. The lower power bills are pretty good, too, though.