"Despite the skepticism of experts and criticism by naysayers, there is little doubt that we are heading into an era of unlimited and almost free clean energy," the Stanford technology maven Vivek Wadhwa declared in the Washington Post last month. The technology that most inspires his enthusiasm is solar energy—and while solar isn't close to "almost free" yet, it is indeed getting cheaper. The prices of solar photovoltaic (PV) modules have fallen steeply by more than 80 percent since 2008.
This trajectory seems to be following Swanson's Law, named for Richard Swanson, the founder of U.S. solar-cell manufacturer SunPower. Swanson suggested that the cost of the photovoltaic cells falls by 20 percent with each doubling of global manufacturing capacity. The pattern is a product of constantly improving manufacturing processes: more automation, better quality control, materials reduction, and so forth.
But how plausible is Wadhwa's prediction that solar power will be unlimited and nearly free? To get a handle on solar's future, let's look at a measure called the levelized cost of energy. This takes into account the capital costs, fuel costs, operations and maintenance costs, debt and equity costs, and plant utilization rates for each type of electric power generation. Many different groups have tried to calculate and compare the levelized costs for building, operating, and financing coal, natural gas, nuclear, hydro, solar, wind, geothermal, and biomass plants.
Let's start with the levelized cost analysis that is the most bullish with respect to solar photovoltaic. In September, the financial advisory firm Lazard reckoned that the levelized unsubsidized cost of utility-scale solar PV is as low as $72 per megawatt-hour. (A megawatt-hour is roughly equivalent to the amount of electricity used by 330 houses during one hour.) Lazard projects that these costs will drop to $60 per megawatt-hour by 2017. Meanwhile, the low-end of natural gas generation is now $61 per megawatt-hour; for coal generation, it's $66 per megawatt-hour; and for nuclear, it's $124 per megawatt-hour. With the current U.S. tax breaks, the low-end solar PV utility-scale costs is $56 per megawatt-hour. George Bilicic, a vice chairman of Lazard, concluded that utilities "still require conventional technologies to meet the energy needs of a developed economy, but they are using alternative technologies to create diversified portfolios of power generation resources."
Every couple of years the Electric Power Research Institute, a nonprofit think tank sponsored by the electric power generation industry, issues a report on the levelized cost of energy for various power generation technologies. Its Integrated Generation Technology Options 2012 report calculates the low-end levelized cost for solar PV next year at $107 per megawatt-hour. For natural gas, coal, and nuclear, the low-end costs are $33, $62, and $85 per megawatt-hour, respectively.
The institute calculates that by 2025, the low-end levelized costs of solar PV will fall to $81 per megawatt-hour. By that time, the institute expects that coal plants will be required to capture their carbon emissions, so the levelized cost of coal will be $102 per megawatt-hour. Natural gas plants without carbon capture will face levelized costs of $44 per megawatt-hour. The report cautions that its calculations with respect to renewable energy generation do not take into account additional costs, such as back-up generation or integration into the electric power grid.
One other authoritative analysis is the Annual Energy Outlook published by the U.S. Energy Information Administration (EIA). In its 2014 report, the agency reckons that in 2019, the low-end cost of solar PV will be $101 per megawatt-hour. Conventional coal, nuclear, and natural gas levelized costs stand correspondingly at $87, $92.60, and $61.10 per megawatt-hour.
To judge from these estimates, the era of unlimited, nearly free solar power has certainly not yet arrived. But things are moving quickly. As recently as 2011, the EIA did not even bother trying to calculate levelized solar PV costs. In that year's report, the agency projected that the country would have an installed solar PV capacity of 8.9 gigawatts by 2035. As of the second quarter of this year, the figure is already 15.9 gigawatts.
In 2008, global production capacity of solar cells/modules amounted to 7 gigawatts. It is now projected to be 64 gigawatts by 2015. This rate of increase suggests a manufacturing capacity doubling time of about two years. As capacity ramped up, Lazard reports that the levelized costs fell from $323 per megawatt-hour in 2009 to $72 now. If Swanson's Law proves true, the levelized cost solar PV could be expected to fall to around $24 per megawatt-hour in the next 10 years. That would not be too cheap to meter, but it would cost far less than any of the forecasts for fossil fuel electric power generation technologies.
Of course, this rough projection does not take into account the intermittency issue (the sun doesn't always shine) that makes solar power problematic as a baseload source of electricity. On the other hand, disruptive new innovations could both greatly improve the efficiency of solar power and battery storage capacity. Will Wadhwa's prophecy come true? Perhaps not, but wagering against human ingenuity has always been a bad bet.