Suspension of Disbelief
No. 02 | February 19, 2024
Seven decades ago, Bell Labs announced its “solar battery”, the first commercial application of photovoltaic cells. The solar battery was six percent efficient; as Nature described it, its cells “are linked together electrically to deliver power from the sun at the rate of 50 W[att] per square yard of surface.” To describe the market as nascent is unfair to nascency. In 1954, “the world had less than a watt of solar cells capable of running electrical equipment.”
C BY 2.0
From that year until the year 2000, the world installed a total of just over one gigawatt of PV power generation capacity. By 2005, annual installations reached a one gigawatt run rate; by 2010, the annual run rate exceeded 10 gigawatts and by 2017, it surpassed 100 gigawatts.
At the same time, year-on-year annual installation growth slowed from more than 100% in 2010, to 24% in 2020. A simple, and defensible, theory of growth and change would say that solar installation rates should be slowing as the market grew by orders of magnitude.
But then in 2022, annual installations jumped 38%, to more than 250 gigawatts, the most peak power generation capacity installed of any technology ever. And then in 2023, annual installations jumped more than 70%. Last year, the world installed 440 gigawatts of solar, and in one year added more than 30% to the cumulative base of solar installed in the previous seven decades. Solar’s share of global power generation is still small, but the line it describes is math-function smooth, arcing up from infinitesimal 15 years ago, to measurable and notable today.
At the same time the technology is also improving in efficiency and decreasing in cost. Today’s nameplate solar module efficiency is 5.9% higher (in absolute terms) and 38% higher (in relative terms) than in 2012 – from a product that also costs 89% less than it did in 2012.
It goes without saying too that the industry has aggregate scale, but it is also useful to think of its unit scale too. Today’s residential PV modules range from 350 to 450 Watts of output, while the biggest utility-scale modules are significantly larger. Simple math tells us that a market of entirely large residential modules would have installed a billion individual units last year, while a market of entirely large-scale utility modules would have installed merely 700 million or so. The reality is a figure somewhere in between, and a unit market probably four times the size of the global television market.
Keeping track of a market growing at this rate – which requires intra-year revisions to annual projections and frequent exhortations to trust the data, not just the model – is challenging. It is perhaps particularly challenging for those who do it for a living. It requires an almost literary approach: one must suspend professional disbelief in a large market growing at a growing rate, and put faith in high-frequency, high-validity information instead. Rather than suspending disbelief in something that is impossible in reality, though, it requires suspending disbelief in something that is impossible in the model.
What, then, gives us an advance look on the future in an era of large and growing markets? The photovoltaic industry has an upstream, fortunately, that helps: the installed and planned manufacturing capacity for polysilicon, wafers, cells, and modules. Each of those is on track to meet the IEA’s net zero scenario for solar manufacturing output, and to do so handily, in fact, at a 40% capacity utilization rate. Other sectors, save batteries, are not as far along.
This is one vision of the future that is visible already, even if disbelief sometimes dims its clarity. Today, the theories for why solar will not continue to grow rapidly considering its inchoate capacity are elaborate and highly contingent. The theory for why it will continue to grow rapidly, in contrast, is parsimonious: it always does.
Media
Thank you to all who have read and shared my annual presentation.
I had the pleasure of recording four smart, and different, podcasts discussing it.
Plain English with Derek Thompson. Fast-paced and focused on how we think about energy superabundance. Link
Catalyst with Shayle Kann. So in-depth it actually drops in two parts. Listen along with the slides on hand as we call out specific pages, and stay until the end of Part Two for a deuteragonist-worthy riff on artificial intelligence and crystals. Part One Part Two
Smarter Markets with Dave Greely. A nuanced and markets-driven look at the data, with one of my favorite interviewers. Link
Thank you for reading.
