US solar power

The falling costs of US solar power

by David Roberts

The fate of the world depends on driving down the cost of solar power.

Yes, that’s a melodramatic way of putting it. But it’s not wrong. Any scenario that has humanity avoiding the worst ravages of climate change involves explosive global growth in solar power.

That’s why the US Department of Energy has a program, the SunShot Initiative, devoted entirely to driving down the cost of electricity generated by solar panels — the target is solar power with $1 per watt installed costs by 2020, a 75 percent reduction in costs from 2010.

So how’s that going?

Happily, Lawrence Berkeley National Laboratory (LBNL) releases a set of reports each year devoted to tracking solar prices; they’ve just released the latest editions. Long story short: Prices are steadily falling, more or less on schedule

There are two reports, one for each type of solar power. One is on "utility-scale solar," which means solar systems larger than 5 MW. The other report is on solar photovoltaic (PV) systems under 5 MW.

Those are two very different markets, but I’m going to squish them together in this post.

Solar is growing, growing, growing

A few things to notice. First, there’s about 29 GW of solar installed in the US now; LBNL expects that to clear 100 GW sometime around 2020. That’s crazy-fast growth (from almost nothing in 2007!), but it will still only put solar at around 3 percent of the US electricity mix in 2020.

Second, on a total installed capacity basis, there’s been more utility-scale solar than small solar since 2012, and that is expected to continue for at least the next five years. For all the hype around rooftop solar, big power plants remain the solar workhorse.

Third, the giant spike in utility-scale PV happening this year is an artifact that reveals how much solar still depends on policy. Everyone thought the 30 percent federal investment tax credit (ITC) for solar was going to expire this year. Contracts signed in 2016 would have been the last to qualify. So there was a huge rush to get projects on the books.

As it happens, the ITC was unexpectedly extended late last year (it will phase out over the next five years), or else the spike would have been even bigger. As it is, more than twice as much utility-scale PV capacity will be added in 2016 than in any previous year.

Prices for utility-scale solar are falling

Prices are falling for both big and small solar, though at different rates and for different reasons.

Let’s start with utility-scale solar. Recall, the vast bulk of utility-scale solar plants are ground-mounted PV farms.

There are also a few concentrating photovoltaic (CPV) plants, which use lenses to concentrate light on PV panels, and a few concentrating solar power (CSP) projects, which use mirrors to concentrate sunlight on towers with steam generators. For our purposes here, we can mostly ignore them, since their growth has been dwarfed by PV’s and PV is expected to dominate growth for the next several years.

So, what about utility-scale costs? They are falling. One way to measure that is "installed costs," i.e., the cost to build and install a project.

This chart is unnecessarily complicated — I’m not going to get into the debate over whether to measure costs via DC or AC — but the thing to notice is that the lines go down. Installed costs are falling each year.

That’s the cost of building these plants. What about the power they produce? Those costs are revealed in power purchase agreements (PPAs), in which a buyer agrees to purchase a fixed amount of power at a fixed price over a fixed number of years.

A few things to note about this chart. First, where California (the green circles) used to dominate, there are now contracts being signed in all five regions — the market is spreading. Witness this map of utility-scale projects.

More and more dots over on the East Coast!

Second, PPA costs are falling: "As recently as 2011, solar PPA prices in excess of $100/MWh were quite common. Barely five years later, most PPAs in the 2015 sample are priced at or below $50/MWh levelized (in real, 2015 dollars), with a few priced as aggressively as ~$30/MWh." From $100/MWh to $50/MWh in four years. Not bad.

Third, all five regions now boast reasonably low average prices, with four of them under $50/MWh. (The Midwest is still catching up.)

Fourth, somewhat counterintuitively, there are no clear economies of scale in utility-scale solar. Bigger plants, represented by bigger circles in the chart, don’t seem to be producing cheaper power. The authors speculate that this is in part due to the increased regulatory and land-use hassles that come with plants over a certain size, which cancel out any savings. (Interestingly, there are economies of scale on the small-solar side. It seems solar power gets cheaper up to the 5 MW to 20 MW range, and then levels out. This has intriguing implications, which I’ll save for another post.)

And fifth, utility-scale solar is reaching prices, especially in California and the Southeast, that are competitive with local wind power (and cheaper at certain times of day). Natural gas is still so cheap that solar has trouble competing with existing plants, but when it comes to new gas plants, solar is getting within striking distance, especially if gas prices rise more than forecasted.

No one really has a clue what gas prices will do, of course, which helps highlight the fact that solar’s stable, fixed, long-term contracts can serve as a valuable hedge against volatile fuel prices.

Prices for small-scale solar are falling

Small-scale (under 5 MW), distributed solar falls into three categories: residential rooftops, nonresidential systems under 500 kW (think commercial rooftops), and nonresidential systems between 500 kW and 5 MW (think small, ground-mounted solar PV farms, like community solar projects).

Overall, the cost trend for distributed solar PV is clear.

Installed costs for distributed PV are headed down. Interestingly, what’s driving them down has changed over time. In its conclusion, the report provides a fascinating capsule history of recent PV prices.

Following a period of relatively steady and sizeable declines, installed price reductions began to stall around 2005, as the supply-chain and delivery infrastructure struggled to keep pace with rapidly expanding global demand. Beginning in 2008, however, global module prices began a steep downward trajectory, and those module price reductions were the driving force behind the decline in total system prices for PV from 2008 through 2012. Since 2012, however, module prices have remained relatively flat, yet installed prices have continued to fall as a result of a steady decline in non-module costs.

"Modules," if it’s not clear, are the panels themselves. "Non-module costs" are everything else. About 20 percent of the drop in non-module costs came through cheaper inverter and racking equipment. The rest came through "soft costs" — finding new customers, installation, maintenance, regulatory compliance, and various other administrative costs.

Soft costs are falling. And it’s a good thing, because module prices are holding fairly steady and are unlikely to plunge again like they did from 2008 to 2012. What’s more, state tax subsidies for PV installations are going bye-bye.

With module prices flat and subsidies declining, most future cost reductions will have to come through soft costs.

Luckily, there’s plenty of evidence that soft costs have more room to drop. First of all, there are wide variations in prices from project to project, especially in residential PV: "Among the residential systems installed in 2015, roughly 20% of systems were priced below $3.3/W (the 20 percentile value), while 20% were priced above $5.0/W (80th percentile)."

That’s a crazy wide spread. The report has several chapters devoted to explaining these variations — from installer to installer, state to state, customer-owned to third-party owned, retrofit to new build, different efficiencies and power electronics, and so on.

But the overarching point is that soft costs can be brought down much further just by everyone converging on current best practices.

Speaking of which, it’s worth noting that the US has some of the highest installed PV costs among developed countries.

Since module prices are pretty standardized the world over, almost all this international variation comes down to soft costs. The US can clearly do better.

The state of solar is strong

There’s lots more to draw from these reports — for instance, module efficiency is improving, and so is the average capacity factor of utility-scale PV plants.

The TL;DR version of all this is that the price of solar power continues to steadily decline in the US. There’s a good argument to be made that it needs to decline much more, and faster — there’s definitely no room to sit back and coast — but at least for now, the news is, provisionally, good.