Solar ROI

or: How I Learned to Liquidate Bonds Without Worrying and Love the Sun

When my family moved to PA we knew we wanted to generate our own energy with solar panels. We had a lot of evidence that made us think a solar array would be a good investment, even in cloudy central PA. Of course the broad findings of research don't always hold true for every individual. In this post I'll review how our solar system has performed over its first full year and what sort of return on investment (ROI) that gave us. Did our solar panels actually make us money, or were we bleeding out cash to be climate do-gooders?

UPDATE: From Jan to Dec in 2019 our panels generated 11,347 kWh. In 2020 they generated 11,358. Both of these numbers are a bit shy of the 12,700 estimate and it seems safe to say my installer over sold the panels output. SREC prices averaged $35.50 in 2019 and $20.57 in 2020 though. All told in 2019 and 2020 the panels earned $3,513 in tax free money, the equivalent of a 7.3% return on a taxed investment.

How much did we generate?

When you buy a solar system the installer always gives you an estimate of how much you'll generate in a year, but how does that hold up to reality? In our case we bought a solar system that was rated at 10.5 kilowatts (kW). That means that on a perfectly sunny hour the system would generate 10.5 kilowatt hours (kWh) of electricity. It is not perfectly sunny every hour of every day though, and so our installer used average solar radiance for our area to estimate that this system would generate 12,700 kWh each year. In reality though we only generated 10,800 kWh during the first 12 full months our system was operational. Why did this happen and how did it affect our ROI?

The why of it was actually pretty simple, the 2018/2019 year in PA was cloudier than normal. I talked with people who run solar companies and manage thousands of different arrays over the state and they said that they got 10-15% lower generation than normal over that timeframe. Cut 15% off 12,700 and you get almost exactly the 10800 that we generated. I have hope that future years will be less cloudy and we'll hit the full estimate. So far we're on pace to generate 60% more energy this September than we did last year, so I'm hopeful that the clouds of our first year were just outliers.

How much money did we make?

So how much money does 10,800 kWh of solar generation produce in PA? The short answer is $1382. Since this money is all tax free it's the equivalent of a long term investment that returned $1625 before taxes ($1727 if you make a lot of money and pay the higher 20% capital gains tax rate). To understand how these numbers came about you have to know a little about how electricity billing works and a few weird quirks about solar in PA.

Net Metering

First, PA has a net metering law that allows consumers to essentially avoid paying the full retail price of electricity for every kWh they generate. Most states have this, but there are small differences in how it works from state to state. In PA every month someone generates more power than they consume they bank the extra kWhs. Then, in a future month where they use more than they generate they draw the extra kWhs from this bank instead of having to pay for them. Once the bank is empty the consumer has to pay full price for the kWhs. What full price is depends on your utility, the time of year and how much total energy you use but where I live it averages out to about 10.5 cents/kWh. Because of Net Metering we get paid 10.5 cents for almost every kWh we generate, and only get paid at a lower rate if we generate more than we use.

PA's Solar Banking Quirk

The reason I said we only get paid for "almost every kWh we generate" is because of another little quirk in PA's Net Metering law. If a solar homeowner in PA has any extra kWhs banked at the end of May they don't carry over to June and are instead just paid for at a lower rate (we got 6.2 cents/kWh in 2019). This is different than most net metering states where you can just carry over your banked kWh forever. Forcing customers to sell their banked kWhs at the end of May adds a little twist to PA's solar market.

Generally, houses have pretty low electricity consumption in April and May when the weather is nice. At the same time, those months have pretty high solar generation because days are getting longer and the sun is shining a lot. This means that if a solar system is sized to generate 100% of the electricity that home uses then the bank of kWhs that that system builds up will be emptied out during the low generation winter months. The customer will then have to pay retail rate for whatever additional power they use before their panels start generating more than they use again. It took me a while to get my head around this. Maybe a graph will help you sort it out:

Generated, Used, and Banked kWh

Looking at this graph you can see how in the summer we generated more power than we used and built our bank up with this. In October our use and generation were almost the same, and then in the colder months our use was far higher than our generation and we started depleting our bank. In February our bank was gone and we had to pay an electricity bill that was higher than our usual $7 monthly service fee. In March we had our own version of March madness, seeing whether we could generate more than we'd use that month. We came up just short at the buzzer, but have high hopes for success next year. In April and May we started generating more than we used again and built up a bank of around 1200 kWhs. Our utility then paid us 6.2 cents for each of these and zeroed out our bank again the following June.

If our utility didn't force us to zero out our bank in May those 1200 kWhs would have carried over to this year and we would eventually use them to avoid paying the full market rate of 10.5 cents per kwh instead of being paid 6.2. That amounts to all of $50/year so it's not too concerning. But, if you're in PA and want to truly maximize the ROI of your system you should make it small enough that your generation during April and May matches your consumption during those months. As it stands this payout was stored as a credit on our account that can pay for our $7 monthly electric service fee for most of the year.

SRECs in PA

The other quirk of solar in PA is that we have a Solar Renewable Energy Credit (SREC) market. Only a few states and cities like DC, MA, and NJ have SREC markets. Every time you generate 1,000 kWh of solar electricity in one of these states you earn an SREC. Utility companies are required to own a number of SRECs at the end of each year or pay a penalty. Exactly how many SRECs a utility needs to own depends on the law, but in PA it's trying to get solar up to 0.5% of total generation so its not that many. A utility can build their own solar plants to generate the SRECs they need, or they can buy them from others.

When we moved to PA SRECs were selling for something like $9 each and we sold our first batch for $15 each. This was because PA utilities could buy SRECs from pretty much anywhere. This meant that big solar farms in North Carolina were getting most of the SREC payouts and SREC prices were low. During the last year the PA Public Utilities Commission (PUC) clarified a rule on SRECs to state that they had to be generated inside the state of PA to be sold here. This ensures that SREC payouts are driving solar development inside our state and has provided a large bump to SREC prices. After the change we sold one batch for over $40/SREC and averaged over $35 for all of them. Here's a graph of PA SREC prices if you're interested. This amounts to a boost of 3.5-4 cents per kWh, so it really helps ROI. I expect that SREC prices will fall as more solar generation comes online in PA, but I'll enjoy the high prices while I can get them.

What was our total ROI for year 1?

So lets do some quick math and figure out our ROI. We generated 10,800 kWh and avoided paying utility bills for about 10.5 cents for each, so $1134. We lost about $50 of this due to the quirk of our solar bank zeroing out in May, so drop it to $1074. We sold our SRECs for almost $310 total, but only $280 after fees, so bump this up to $1354. We generated 80% of another SREC so we should really add another $28 to this to be fair, bringing it up to $1382. Also note that there are no taxes on SREC income or lowered utility bills. A regular investment would likely pay 15% long term capital gains taxes on its return, so you'd have to earn $1625 in it to take home $1382 after taxes (1625-1625*.15=1381).

PROTIP: Fully max out your ROI by lowering your tax withholding in January and installing near the end of the year. This builds up the tax credit in cash before laying out money to pay for the system, like a free loan. Be sure your installer commits to install the system in a cold month so you don't fail to install that year and run afoul of the tax man.

What about costs? Well, our 10.5 kW system cost $34k to install. The 30% federal tax credit for new solar installations brought this down to $24k. It's important to note that prices vary wildly by location and have been dropping pretty consistently over time. A friend of mine in Virginia got a system installed for over 30% less per watt, while another in CA paid almost 20% more. Rather than try and predict a price for your location I'll just say get a few quotes from a few local installers in your area to see what an array will cost for you.

How Solar Crushes Bonds 

If we had invested this $24k in a bond fund and it gained $1625 before taxes that would be a 6.7% return on investment. This may seem low when comparing to stocks, but solar has next to no risk to this return so it is fairer to compare it to bonds and other "safe" investments. The Vanguard Total Bond ETF (BND) that I use as a safety hedge for my stock investments has only returned more than this ONCE in the last decade. And remember this year's solar radiation was down 15% from expected for us. This ROI could have gone up as high as 7.8% and even higher if SREC prices continue to rise or if we’d found a cheaper installer. In the worst case scenario where it is cloudy forever, and utilities build enough solar plants to drop SREC prices to $0 this would still equate to 5.2% ROI (1074/(1-.15)/24000=0.052) which trounces the 3.8% average annual return of BND.

There is over $200 billion invested in BND right now. There are trillions more invested in other low returning "safe" funds. How many of those investors do you think live in states with net metering and own houses with nice sunny roofs? Are you one of them? Why aren't you transferring as much money from those bonds into solar on your roof as possible? Do you hate money?!? Are you actively trying to drive climate change to flood people out of their homes!?!? WHYYYYYYYYYYYYY?!?!?!?!?!?

What about depreciation?

Okay, things started to get a little out of control there, but I'll catch my breath and talk about the main reason why solar may not be quite as good as the above yelling makes it out to be. The big problem with solar is that the technology just keeps getting better. The efficiency of solar panels has skyrocketed over the years, while the prices to build and install them have fallen. At some point this continuous improvement has to reach a limit, but it's been continuing apace for decades now. It seems possible that in 10 years someone could install a system that generates as much power as my family's does for 1/2 the price.

Does the falling price of solar panels mean that your panels depreciate 50% every 5 years? Not exactly. Even though solar panels are dropping in price, it seems unlikely that utilities are going to drop the rates they charge for power any time soon. Utilities get to basically pass all of their costs onto their customers so while they'll be getting cheaper solar power in the future their customers will still be on the hook for all the coal and gas plants that they'll be decommissioning early because it'll be cheaper to build new solar than to run them.

Yes, if you try to sell your house a savvy buyer will price the panels based on what a new system would cost, but if you just keep owning it the annual return you'll see will likely stay fairly steady. Most solar panels are warrantied to produce 80% of their initial output after 25 years (fancy ones from Sunpower increase this 92%). This means that your annual return will only drop about 0.8% per year (or 0.32%/year for Sunpower). It is also likely that panels you install today will keep producing power decades after their warranties are up. With this in mind I think it's a tough question as to what rate of depreciation should be applied to panels.

Finally, there's the fact that the federal solar tax credit is starting to drop and this will likely more than offset any reductions in panel prices over the next few years. If you get a system installed before the tax credit expires and then sell it shortly afterwards you may even see appreciation from your after tax cost because the buyers will be comparing it against new solar arrays with no tax credit. As time goes on though it is likely that the improvements in solar technology will lead to depreciation for your panels. How much is a tough question. All I can really say is that a solar array will keep producing valuable electricity for a very long time, and that the value of these electricity is what I focus on with my ROI calculations.

Wrap up

After a year's experience with solar panels I am very bullish on them. How bullish? Well, I went and created this website almost explicitly to share this with the world. My family's panels have significantly outperformed the bond fund where we had our money invested in beforehand with less risk. Part of me is sad that I didn't wait for solar technology to improve further so I could get an even higher ROI, but another part is happy that our entire house is being powered by clean solar energy and I'm beating our previous investment already. If you've got a few thousand dollars in some low return investment like Bonds, CDs, or [shudder] a savings account and own a home with a relatively sunny roof you'd be a fool not to look into installing solar. Put some panels on your roof and you too can profit greenly.

Further Reading: Heat pumps pair with solar like peanut butter and jelly. Check out how a heat pump can save you money, how to combine that with solar to go net zero, and why you probably don't need a home battery yet.