Sizing Our Solar Panels

We got solar panels last year. This was before we were on the whole #ElectrifyEverything kick, but we felt some time pressure:

• Tax Credit: We wanted to lock in the 26% tax credit before it decreased. The credit has since been extended, but we didn't know that at the time
• Lock in Utility Rate: Our utility, PG&E, was looking to make pricing changes, most notably around connectivity fees ($10 => ~$60), time-of-use rates (more expensive in the evenings), and the price they purchase excess solar back (it's dropping). We wanted to lock in the current rates/policies (Net Metering 2.0) for 15 years, which meant installing our system before Net Metering 3.0 gets passed. (as of this writing, the new regulations are still being decided. If you live in CA, call Gov Newsom and urge him to keep solar affordable)

That's the boring part. The more interesting part was how to size our solar panels AKA how many panels to get. (a note on sizing things being "the more interesting part": I personal like combinatorial optimization like the fixed-size knapsack problems. YMMV)

We went with SunPower (a very reputable CA solar installer. hit me up if you want an introduction). Our sales rep looked at our existing utility bills to help us size our system. The problem is, our life is almost entirely natural gas, so historical data was unhelpful.

Rather, we had to size our system based on probability that we would switch to electric, and when. 

There's a vast difference in the number of panels needed for our life today, and our future with 2 EVs, an electric jacuzzi heater, and more. But we didn't know when we would be switching everything over... it could be 10 years! So our payoff calculations change drastically if we oversize now but don't use it, or undersize now but transition to electric quickly. Upgrading later is possible, but without the same tax credits and paying for labor twice. Oh, and there are different manufacturers and photovoltaic density on panels, each with different costs. Another variable.

Aurora Solar, one of the really cool tools that installers use to estimate projects, just doesn't handle this much ambiguity from a customer. 

I quickly realized this was going to lead to an overconstrained series of equations, so I got all the ratings, metrics, and costs and plugged it into Google Sheets. 

In the end, we sized for 2 EVs and a hot tub. This was definitely overkill for 2021 but will likely serve us will in 2022 and 2023. It's likely too small for an all electric future, so we may want to revisit later as we upgrade appliances too. (Note: utilities buy power back at pennies on the dollar. It's not like you make a profit there. So an over-sized system is generally just wasteful)

Oh, and one interesting hack we found... newer panels are denser, which gives more energy per square foot. We have a very large roof, so the number of panels wasn't a concern for us. Turns out we could purchase more of the older, less dense panels for less money than fewer high density ones. It takes up more space on the roof, but we have plenty of room, so that seemed like a nice money saving trick. 

Conclusion: this stuff is remarkably hard to get right! Maybe I was overthinking some of this by trying to optimize carbon footprint with economic payback time, but sizing a system and laying out tens of thousands of dollars (or signing a financing contract) BEFORE you've electrified everything and your electricity bill is small (like ours today) is a tough leap of faith to make.