The Energy Information Administration dropped its latest monthly estimate of solar module shipments a few days ago, so this seems like a good time take to take stock. Here's an extrapolation of where we'll be by 2030 if solar keeps rising at its current rate:
At the current rate of growth, solar capacity will reach about a thousand gigawatts by 2030, which would probably be about half of total demand. Raw cost will drop from 30¢ per watt to 15¢ per watt, producing a levelized cost per kWh less than any other source.
This, of course, assumes that both shipments and cost of solar modules can stay on their exponential paths for the rest of the decade. Fingers crossed.
This, of course, assumes that both shipments and cost of solar modules can stay on their exponential paths for the rest of the decade.
But, China tariffs.
That would affect the shipment and cost of panels coming into the U.S. The rest of the world would benefit.
Yeah I've been wondering about that. How much of the demand for Chinese solar panels is coming from the U.S.?
And if we slow significantly (b/c we have to rely on more expensive panels) how much would that slow down the world overall?
> This, of course, assumes that both shipments and cost of solar modules can stay on their exponential paths for the rest of the decade.
Really, it just assumes that shipments increase exponentially. Prices drop pretty much automatically from both economies of scale, and the fact that the increased revenue funds more research. There was an exception to this rule when the PV industry got large enough that it needed more pure silicon than was being provided from the chip industry's waste. But we don't see any additional bottlenecks on the horizon.
And the growth is pretty much baked in. There are a lot of places in the world where Solar is very cost effective and electricity is very much under-supplied. And as long as batteries are growing exponentially, solar is the cheap way to charge those batteries.
And yet, it seems like solar is still being subsidized. At what point does the need for subsidies go away?
Look, I understand that coal and gas are subsidized indirectly because their negative externalities aren't priced in, but that isn't going to change soon, so solar still has to compete with alternatives on the ledger.
We had rooftop solar for nine years before we moved. Even though half the cost was paid by the utility and we got a 30% tax rebate on the rest, we never made back our investment, even in nominal dollars.
A large part of the US military exists to protect the flow of oil.
"We had rooftop solar for nine years before we moved. Even though half the cost was paid by the utility and we got a 30% tax rebate on the rest, we never made back our investment, even in nominal dollars."
Rooftop solar is much more expensive than utility scale, FWIW (that whole economies of scale thing). But, even presuming you moved in 2024, nine years ago would have been 2015, when solar would have cost more than double what it costs now.
". . . coal and gas are subsidized indirectly because their negative externalities aren't priced in, but that isn't going to change soon, so solar still has to compete with alternatives on the ledger."
Why, exactly? Cars without catalytic converters are "subsidized indirectly because their negative externalities aren't priced in," but we don't stroke our beards and muse that someday the effective subsidies inherent in requiring cat converters must someday be phased out so as to allow "compet[ition] with alternatives on the ledger."
We subsidize all sorts of things (public school, roads, etc.) precisely because their benefits can't compete in the market for individual products. Solar ain't different.
Generally a payback time is calculated when individuals install rooftop solar. What was the result given to you when you purchased?
We were told seven years. Not even close.
Did you never make back your investment, measured by the return on excess power sold back to the grid?
But the huge gain with home solar is the power you don't buy from the grid: power that is not recorded in, or out, through your meter. To work that out even approximately, I have to look at metered power decline, season by season, compared to pre-solar.
Allowing for that, my payback from about the same time as you was in just over two years. And home solar has saved me a lot of money since.
Ameren bought back excess generation at a tenth of the market rate. And we only generated excess a couple months of the year.
I got real-time and cumulative output reports online on our rooftop solar in kWh. Ameren's rate in kWh is published. It is trivial arithmetic to calculate our savings. That's how I know we came nowhere close to paying off our put even after nine years.
Are you calculating your return from rooftop solar as the excess generation, and the (low) rate paid for that?
The real return from rooftop solar, as with any self generation, is the power you don't buy because you generated it.
That return requires you to work out how much less you bought because of rooftop solar. If you've done that I don't see how you could be so far from recovery as you say.
Wow. Extrapolate that line to the end of the century and we'll be at a million gigawats at zero cost!! Give it up with extrapolating quadratic fits, already. They bear no relation whatsoever with reality.
To be fair, an extrapolated quadratic fit for solar prices is probably more reasonable than in a lot of other cases.
After all, nature has already demonstrated that self-replicating nano-scale solar-plus-storage devices can be deployed globally at less-than-zero cost to humans.
The cost curve is obviously hyperbolic. It will approach zero but not reach it.
Storage…
"At the current rate of growth, solar capacity will reach about a thousand gigawatts by 2030, which would probably be about half of total demand."
Not if Trump can help it. He's going to drill and flood us with oil whether we "like it or not."
"Not if Trump can help it. He's going to drill and flood us with oil"
Periodic reminder: oil and solar do not compete*--oil is a transportation fuel, and not used for electricity; solar is used for electricity. All the oil in the world doesn't displace any demand for solar generation.
______
*With EVs, of course, that picture changes. There aren't enough of them--yet!--though, that oil and solar are replacement goods.
Solar panels come from other countries besides China.
Is this just the cost of the panels? As the panels become less expensive other costs start to dominate: inverters, cables, installation, batteries.
We put in solar panels and a battery. I did not expect to recover the cost soon. My goal was to not rely on PG&E and not use any natural gas so I put in much more capacity than needed. Still, I expect to recover the costs in less than 8 years as the cost of electricity in California has risen so much. Some of this may be mitigated because PG&E wants to charge you even if you don’t use much power. This is a charge on solar power.
The other new(ish) fee in CA is a delivery charge. I elected to get my power from a green supplier, PG&E tacks on a per kW charge to deliver that power.
That curve is before the Fossil Equalization Tax mandated on page XXX of Project 2025.
Looking at the EIA data, about 3.1 GW of solar was shipped per month in the US, and the current price is $0.29/W. The leading US supplier is First Solar, and for 2024 they project to ship 14.2 GW for $4.1B, which translates to 1.2 GW/month and also $0.29/W. So this one US supplier can handle a good chunk of the market!
They use a different tech from the Chinese vendors - thin films of cadmium-tellurium instead of bulk poly-crystalline silicon. CdTe is slightly less efficient than silicon, and so needs more area, but also uses less material and is better at high desert temps. They're hard at work on the next generation, a tech that has taken forever to settle called peroskovite. That can be much more efficient and much cheaper still, but is nowhere near as far down the learning curve.
Anyway, the US can compete if the Chinese don't dump to seize the market, and the fossileers don't pay off Trump to put big regulatory roadblocks in the way.
Perovskites will be very nice, if they ever prove practical. They work well but for short times. The research is in finding stable materials. On the other hand, single crystal silicon is guaranteed to for about 40 years. High quality panels like those from SunPower (now Maxeon) last a really long time and degrade slowly.
Another new technology in the wings is a way to add second layer over the silicon to make electricity from light wavelengths not absorbed by silicon. This could add another 5 to 10 percent efficiency. No idea about costs or timeline.
Another technology in the offing is better battery tech for home or grid scale storage. Right now they tend to use Lithium. This is great for portable uses, but for sites that will never move weight doesn't matter and lifetime, cost and discharge rates are more important. Several chemistries are competing for this area.
... producing a levelized cost per kWh less than any other source ..."
Apparently levelized cost doesn't account for the batteries (or other storage system) required to use solar power to satisfy demand when the sun isn't shining. So it isn't a fair way to compare solar to a source like natural gas turbines that can follow demand.
Well, EIA, Lazard, and others produce LCOE comparisons of different systems including PV + storage, wind + storage, etc.
Regardless of their assumptions under different scenarios, the EIA 2023 AEO projects a baseline US crude oil production at 12.5mmbd in 2050, up from 11.39mmbd in 2022.
Curious, given the opinion of the interwebs that peak oil is coming very soon, doncha think? Some people still believe we can hold climate change to +1.5°C, yeah? Maybe even a little overshoot. Just about no one on the interwebs believes the experts who've informed us that we're actually on track to +3.0°C. ¯\_(ツ)_/¯
But hey, y'all server farms prefer to invest billions into the future of SMRs rather than whatever PV and battery storage you can install right now? And when they're fully amortized, y'all could just give them away when (if) SMRs came online, yeah?
Nah, we've got time. Lots of it. ᕕ(ᐛ)ᕗ