Solar
So I went solar. And as the process of learning everything I needed to know was somewhat involved, I thought I'd share some of my experience. By the way, my friend Rebecca says it is bad form to start a sentence with "So" and it is also bad form (in fact, aggressively looking for affirmation) to end a statement or diatribe with "Right?". But I do it anyway.
So solar. Basically there are a few choices - buy vs lease (I don't see a rational argument for leasing other than being green and not being able to afford solar). Then there is the panel choice. Then there is the installer choice. Then there is the consideration whether you will need a new roof any time soon (if so, do that at the same time). Finally, there is how much capacity to get.
There is a fantastic site that helps with all this, if you really want to geek out and know everything a lay person needs to know - SRoeCo Solar
Background and disclaimers
The only thing you need to know about electricity for all this is the difference between power (the rate at which energy is used or generated, expressed in Watts) and energy use over time (expressed in kWh, or kilo-Watt-hours. One kWh is the equivalent of using 1000 Watts of power for an hour.
An LED lightbulb uses about 14 Watts, a space heater can run at 1,500 Watts. So if you run a space heater for an hour, you use 1.5 kWh.
Your electric meter measures how many kWh you generate or use to calculate your bill. Typical residential retail cost for a kWh today is somewhere between $0.15 nationwide or $0.32 in California. At a supercharger, many charge $0.50 per kWh to charge a car.
A solar system of 8100 W generates (in theory, in full sunlight at a perfect angle) 8.1 kWh per hour. The efficiency of the system and the inverters, weather, amount of sunlight, etc. determine how much you actually end up generating.
There is a lot of talk about energy transition - e.g., getting rid of your gas furnace, water heater, stove, or car. I am staying out of making recommendations on that; there are calculators for the impact of any of those on you. Keep in mind that a lot of the financial impact of any of that, besides the immediate costs and subsidies or rebates, will depend on the future price of gas and electricity, which is very unpredictable. In the past ten years, electricity has largely become cheaper, while every other source of energy has more than doubled. But past performance is no guarantee of future success.
A lot of the discussion about solar is about whether you are green, "doing your part", climate change, sustainability, etc. Companies contribute to that (SolarEdge tells me that the 160.2 MWh I have generated correspond to 248,063.59 lbs CO2 emission saved and 1,874.34 Equivalent Trees Planted). I won't say I don't care, but that too I will stay away from other than pointing out that that tells only part of the story. It doesn't talk about the cradle-to-grave impact of the panels and inverters, the maintenance of the grid infrastructure, the heavy metals in batteries, the waste of throwing out perfectly good heaters and A/C units and gas furnaces and stoves to replace them with electric equivalents such as heat pumps, etc. These things are also subject to a certain amount of fashion. I grew up in the wake of Three Mile Island at the peak of the anti-nuclear movement, so I am bemused to see that the Diablo Canyon nuclear plant is now lauded as the largest source of "clean," zero-carbon electricity in California. My arguments here are simply about whether the substantial investment in solar makes sense for you in terms of your energy usage, cash flow, future use of electricity versus other sources of energy, and what to do and look out for if you decide solar is right for you.
The selection process
The most important choice (and the one that will guide all the following steps) is your installer. This is like picking a contractor for a remodel. Get a bad one, you end up being unhappy with your project and you get crappy recommendations for capacity, panels, inverters, getting city permits, etc. The trick is that some installers prefer some panels (or are not authorized for them) so if you have a preference (e.g. wanting Sunpower) then that guides you towards an installer. Angie's List is worth looking at for recommendations, or yelp, but definitely talk to a few before you pick one (and get multiple bids).
If you live in the South Bay, I highly recommend Cobalt Power as the installer.
As homework/prep, start by figuring out your electric usage -- take a few years of bills, think whether you will be growing (AC, hot tub, electric car, etc), and based on that come up with an estimate. You have to take into account that your usage will probably go up when you get solar because you will be less incented to save on AC, you may leave the lights on more, because it is "free". A typical household is in the 10s of kwh per day (my usage is around 30-40 kwh/day but that's a pretty small house in a temperate climate).
Then figure out what you want your solar to cover. For me, it was simple: I NEVER want to pay for electricity again (even if I end up getting an electric car) and I like the idea of generating more than I use. Some people just want to shave off their top-tier usage (the portion of your bill where they charge you like 35c/kwh) but again I don't see the point of going through all the trouble for that.
Also look at how much surface area you have -- south/west facing, unobstructed by trees, etc. -- that's a real constraint for some people, not for me. If you are constrained, you may not be able to cover all your usage, even with the most efficient panels. Where you live makes a difference (number of sunny days and sunlight hours).
Given all that, you end up with how big a system you want to build - e.g. "I want a 9 kW system" -- any solar installer can help you with that calculation. My recommendation is go MUCH higher than the number you compute; once you become a power plant, you will start using more power, and the marginal cost of a few more panels is small. I have an 8.1 kW system (not sure how they compute that but the installer will tell you) and it generates more than 60 kWh/day on sunny days in June, about 40 on sunny days in October, but only about 10-20 on rainy winter days. But PGE computes your net usage over the whole year.
Now the panels. There is a lot of choice there, but in the end, for all the marketing hype, the difference between "cheap" and "expensive" is not a huge part of the cost - what you are partially buying is reputation and long-term warranty (if you go with the cheap-o panels, and one of them fails five years from now and the company no longer exists, where do you go?). The efficiency may seem like it does not really matter anymore (and people will try to tell you that), but where it comes into play is how much of your roof will be covered by panels. Especially if the panels are on the front of your house, looks matter.
Based on that, plus quality/ratings/efficiency/etc, I decided on Sunpower - still by far the most popular panel here in the Bay Area even though they are not the cheapest. At the time the most efficient panel, and they had a great warranty. Of course, unfortunately they also went Chapter 11 so we will see... Little did I know but the panels were manufactured by Maxeon, which has said they will continue to honor their 25-year warranty.
Your installer will also recommend an inverter. This is what turns the current that gets generated on your roof into the stuff that your house (and the grid) uses.
I ended up with SolarEdge inverters - very happy so far. Some are more efficient than others, etc.
There are some smaller choices as well - e.g., do you want the power line to go under the roof (prettier) or over the roof (cheaper).
And don't forget about the federal tax incentive (the California rebate has run out -- part of the fun of being a late adopter). Your installer will tell you all about it. It's real money - it ends up saving you about a third of the cost of the project.
Anyway, you can tell it's a pretty big project if you do it right, but when all is said and done, my PG&E bill went from $350/month to zero -- so it pays for itself. Seems like a good deal. Right?
Tracking
I would highly recommend making your own spreadsheet for tracking key data related to your solar investment and energy cost and consumption.
There are lots of gadgets and tools out there that help you get it down to the individual light bulb or when you turn on your dryer, but fundamentally that's in the noise, unless you are burning 1500W space heaters nonstop or you have some other frequently running mega-appliances (AC, hot tub, charging an electric car) and you want to know what they are costing you. And even then, the question is whether you are really willing to do something about it. That's why you have solar, right?
The real reason to capture data yourself is that most utilities only preserve two years of data for you. So if you want to have a long-term record of energy usage, cost, unit cost, and ROI, you have to do it yourself.
Here is what I've been tracking since about 2008, and for the solar-specific data, 2014:
Monthly and annual net kWh used or fed back to the grid (from PG&E)
E.g. -410 kWh for June 2024, meaning I generated 410 kWh more than I used that month. For all of 2024, my total was -3,526 kWh. If the annual total is negative, this becomes part of your annual rebate, but you are getting only part of the value back, so not direct part of ROI.
By comparison: For June 2013, before I had solar, my energy use was +1,141 kWh
Monthly and annual value of kWh used (from PG&E)
There are two different numbers you get every month:
"Current Monthly Charges" (e.g. $12.39 for June 2024 -- this is mostly the cost of having a connection, minimum delivery charges, etc. This a fixed monthly minimum that you pay as long as you are on the grid.
Estimated Net Energy Metering (NEM) charges (e.g. -$124.93 for June 2024). This is the actual cost or value of the electricity used from or fed to the grid, and the relevant number if you are interested in the cost or value of your power footprint and solar generation vs. use. By comparison: Before I had solar, in June 2013, my cost of electricity was +$309.
For all of 2024, this value (for the -3,526 kWh I net returned to the grid) was -$1,067.56 (again, this is not the actual amount you get back, but it's an annual true-up bank from which you can draw e.g. in the dark winter months).
(computed) Annual cost per kWh
I use annual average $/kWh (i.e., the sum of the monthly nets divided by the sum of the monthly values. In my case, you divide two negative numbers to get a price per kWh (e.g. for 2024, -3526 kWh / -$1,067.56 = $0.30/kWh. This number is slightly different from the national statistics provided for the Bay Area (e.g. from the Bureau of Labor Statistics or the US Energy Information Administration), which were $0.32-$0.37 for 2024. Whatever.
Monthly and annual kWh generated (from SolarEdge)
E.g. 1614 kWh for June 2024, or 14,298 kWh for the whole year. Of that 1614 kWh, I ended up not using 410 kWh (as mentioned before).
(computed) Monthly and annual value of generated power at retail
E.g. 1614 * $0.30 = $488.67 for June 2024, or $4,328.98 for all of 2024. This is a theoretical number, as you neither pay nor get paid this number. Not used elsewhere, but o theoretical interest. Not the number to use in ROI calculations.
(computed) Annual kWh generated and used
This is part of your ROI. This is where net metering makes a big difference, because you can time shift between when you generate the power and when you used it (you can still do that if you don't have a net metering rate plan, but it requires investing in battery storage).
This is an interesting number, because you need to put together your SolarEdge data and your PG&E data to get it, but it is really the number that matters to most in terms of ROI: how many kWh that I would otherwise have had to pay for did I get "for free" as part of my investment in solar panels?
For 2024, I generated a total of 14,298 kWh (from SolarEdge). I returned a net total of 3,526 kWh to the grid (again, with time shifting etc. - I did not instantly use that power), so my net savings were 14,298 - 3,526 = 10,772 kWh. When you multiply this by the actual cost per kWh ($0.30 for 2024) you get the number that is your actual energy savings -- $3.261.42 for 2024. THIS is the number you can use in your ROI calculations.
Caveat: Once you have solar, it is tempting to use more power given that it is already "paid for." Net energy metering discourages this, especially if the returned kWh are valued the same as what they would have cost if you used them and even more so if that value, should it end up being negative, gets fully reimbursed at the end of the year. In my current rate plan, Silicon Valley Clean Energy fully credits the kWh towards future use (within a net metering year) but gives partial credit for unused kWh.
Annual rebate (from SVCE cash-out statement, which shows up on your May PG&E bill)
For me, this is a few hundred dollars a year, based on wholesale value of kWh generated and not used - small part of ROI)
My SVCE rebate is a few hundred dollars a year (up from a few tens under PG&E). This is the compromise -- SVCE pays about $0.08/kWh for energy that is generated and not used by the end of the year. PG&E pays much less -- in the years they administered it, they rebated about $0.03/kWh unused. SVCE used to pay more ($0.24 in 2019, $0.19 in 2020, etc.) but it seems reasonable that it is somewhere around the wholesale marginal cost of generation and delivery per kWh.
Return on Investment
When you put all these number together, you can see how you are doing on ROI on your solar investment.
Payback: Of course, all you finance people know that payback is a terrible metric. But the reality is that we expected electricity costs to keep rising, and they didn't. My cost per kWh had gone from $0.15 in 2009 to $0.25 in 2013 to $0.37 in 2014!! The net-net of it is that when I put it in, we expected about a six-year payback, and it actually took 10 years.
ROI: A more realistic metric is the dollar return of the system relative to the amount invested in the system. The return is the value of energy generated and used, plus annual rebate checks. This annual return has been consistent around 10% of the system out-of-pocked cost. So, adjusted for inflation, about the same as the stock market. And in the meantime, you have eliminated a good chunk of cash flow from your monthly expenses, and saved a few trees.
In the ten years since 2014, most years the cost per kWh has been around $0.20, which largely accounts for the fact that my payback was 10 rather than 6 years. If it would have stayed at $0.37 per kWh or kept going up, like all other utility, the return would have been a lot faster/greater. For comparison: in the same period, the cost per therm of gas has gone from $1.11 in 2014 to $2.63 today, an increase of 136%. In the same period, monthly sewer cost has doubled (gone up by 101%), and water per CCF has gone up by 172% (almost tripled).
Yes, it is somewhat upside down but logical that the payback period of an investment in solar becomes longer if electricity becomes cheaper. I still saved the same number of kWh, but those kWh were worth less than expected.
I installed my system on the heels of California's energy crisis (anyone remember Enron?) when we had grid, supply, and generation capacity issues. We all expected electricity prices to keep going up. Since then, an enormous amount of generation capacity has come online, including huge amounts of solar. This has kept costs in check, and made them come down relative to inflation.
What is different today?
When I put in my system ten years ago, 8.1 kW was still large, so the out-of-pocket cost (even after incentives and rebates) was substantial, in the tens of thousands. Since then, a few things have changed:
Sunpower is in bankruptcy. So even though they were great panels, installers probably would not recommend them (but you can buy them very cheaply in surplus auctions).
The cost of panels and of getting a system has come down by a lot. The actual dollar cost of a comparable system (before incentives or rebates) is probably half of what it used to be (in actual dollars), or (using my inflation calculator) 0.38% of what it cost then.
There are still tax credits and incentives, but they vary depending on where you live. And I would not count on federal credits in the future.
Net energy metering is probably on its way out. That makes some sense, given that it was originally intended as an incentive to get people to adopt solar and help reduce reliance on fossil fuels for electricity generation, a mission largely accomplished in my part of the world. Even in California, under NEM 3.0 new customers get substantially less credit value for excess power sent to the grid. If you are putting in solar today, you should probably put in battery storage as well, so you can store the power you generate during the day and use it at night, or during power failures. Otherwise you risk generating at off-peak wholesale prices ($0.03/kWh) and using later in the evening at peak retail ($0.50/kWh).
Discussion and speculation
The arguments against net metering are, well, arguments. The utilities have made successful inroads as more and more people have adopted solar to make it financially less attractive. Their arguments are roughly as follows:
Energy infrastructure has a fixed element (things that everyone who ever uses a single kWh not generated by themselves uses, ranging from power plants with peak capacity to power lines and emergency crews that repair storm damage) and a variable element (energy you consume right this moment from the grid).
Traditional rate structures, including net energy metering, have amortized the fixed costs over the variable usage, into a rate structure that charges some price per kWh, no fixed rate for all the infrastructure you are connected to and dependent on.
Unless you are totally off the grid (you, your panels, maybe your generator are connected only to your own batteries), even if you generate more power than you use, you use the grid to get rid of excess capacity and to grab power you didn't generate at other times. So you are using the grid as a big battery, even if your net use is zero. And you aren't paying for that, other than perhaps the wholesale value at time of generation of the extra kWh you dump on the grid.
The utilities argue that this is not fair to people who don't have solar, who subsidize the infrastructure for everyone else. There is some truth to that, because these companies are regulated monopolies (public utilities) and typically their rates are set based on their cost and a fair rate of return. I suppose the same argument can be made about people who drive electric cars and who don't pay gasoline taxes, which pay for our roads. Until we can think of a better model, gasoline consumers subsidize the owners of electric cars
By the way, this also means that when you "fine" PG&E, say for the fact that some downed power line might possibly have contributed to some hypothetical wildfire in the past, you are simply fining their customers, in a manner entirely proportional to their net energy usage. It's basically just a tax. But I diverge.
For now, the rates are still mostly per kWh. There are proposals to change that radically, and to charge per connection. Regardless of whether you have battery storage, there is likely to be some sort of future substantial "grid use charge" for anyone who is connected, regardless of your net energy usage. Oh well, live with it - the monthly sewer bill also doesn't charge you per toilet flush.
Most municipalities also don't allow you to disconnect completely, even if you can afford to build a closed system. The arguments for this are less obvious, but probably originated in wanting to avoid a bunch of people with stinky generators and are now based more in fear that the costs of the overall system would become unsustainable if there is the equivalent of a bank run of people leaving the grid. There is shared infrastructure, and part of being a community is that we, well, share that infrastructure as well as its cost.
Regardless of whether you or I think it is fair to charge every grid user for the actual proportional cost of being connected to a well-maintained and functional grid, these changes are adding to the strong and growing resentment against PG&E and Souther California Edison, which have led the efforts against net metering and which still have monopolies in large parts of the state. I would suspect that many municipalities, especially in "green" political constituencies, would disconnect themselves from PG&E and just pay wholesale cost for energy used. A few, like Palo Alto, already have (with rates 50-60% lower than PG&E). So the net effect may be that PG&E becomes a generator/wholesaler and local municipalities will be responsible for the last mile, just as they are for roads, water, and sewer.