Skip to Content

Cheap Solar Power—at Night

New solar thermal technologies could address solar power’s intermittency problem.
April 8, 2014

When the world’s largest solar thermal power plant—in Ivanpah, California—opened earlier this year, it was greeted with skepticism. The power plant is undeniably impressive. A collection of 300,000 mirrors, each the size of a garage door, focus sunlight on three 140-meter towers, generating high temperatures. That heat produces steam that drives the same kind of turbines used in fossil-fuel power plants. That heat can be stored (such as by heating up molten salts) and used when the sun goes down far more cheaply than it costs to store electricity in batteries (see “World’s Largest Solar Thermal Power Delivers Power for the First Time”).

Searing sun: Thousands of mirrors focus sunlight on a tower to generate high temperatures at a power plant in Ivanpah, California.

But many experts—even some who invested in the plant—say it might be the last of its kind. David Crane, CEO of NRG Energy, one of three companies, including BrightSource Energy and Google, that funded the plant, says the economics looked good when the plant was first proposed six years ago. Since then, the price of conventional photovoltaic solar panels has plummeted. “Now we’re banking on solar photovoltaics,” he told a crowd of researchers and entrepreneurs at a conference earlier this year.

The allure of solar thermal technology is simple. Unlike conventional solar panels, it can generate power even when the sun isn’t shining. But in practice, it’s far more expensive than both fossil fuel power and electricity from solar panels. And that reality has sent researchers scrambling to find ways to make the technology more competitive.

One big challenge, says Philip Gleckman, chief technology officer of Areva Solar, is that the arrays of mirrors, as well as the motors and gearboxes used to aim them at the sun, are expensive. One potential fix, he says, comes from a San Francisco startup, Otherlab, which replaces the motors with pneumatics and actuators that can be made cheaply using the manufacturing equipment that’s currently used to make plastic water bottles.

The head of Otherlab’s solar efforts, Leila Madrone, says the technology could cut the cost of mirror fields for concentrating sunlight by 70 percent. But even this cost reduction, she says, won’t be enough to make the technology competitive with solar panels—even though the mirrors account for a third to a half of the overall cost of a solar thermal plant.

Getting overall costs down will require increasing the amount of power a solar thermal plant can generate, so it can sell more power for the same amount of investment. One approach to increasing power output is to increase the temperatures at which solar thermal power plants can operate, which would make them more efficient. They currently operate at 650 °C or less, but some researchers are developing ways to increase this to anywhere from 800 °C to 1,200 °C. That approach is being pursued by another startup, Halotechnics, which uses high-throughput screening processes to develop new materials—including new kinds of salt and glass—that can store heat at these high temperatures (see “Cheap Solar Power at Night”).

Another option, being funded by a new program at the U.S. Advanced Research Projects Agency for Energy, is to make power plants that add solar panels to solar thermal power plants. The basic idea is that solar panels can only efficiently convert certain wavelengths of light into electricity. Much of the energy in infrared and ultraviolet light, for example, doesn’t get converted, and is instead emitted as heat. The new projects look for ways to harness that heat.

Solar systems that combine heat and solar panels aren’t new. For many years, companies have offered solar systems that run water pipes behind solar panels—the waste heat from the panels makes the water hot enough for showers.

The new approach, however, is to look for ways to reach much higher temperatures—high enough to be used for generating electricity. Such methods typically involve concentrating sunlight to generate high temperatures, and then diverting some of that concentrated sunlight to solar panels.

In one case, nanoparticles suspended in a fluid absorb wavelengths of sunlight that solar panels don’t convert efficiently. Those nanoparticles heat up the fluid. Light that the solar panels can use pass through the fluid to a solar panel. Other researchers use mirrors that allow only certain wavelengths to pass through them.

Howard Branz, the program manager in charge of these projects at ARPA-E, says the hope is that the added cost of these hybrid systems will be made up for by two things. First, the systems will be more efficient, potentially converting more than half of the energy in sunlight into electricity, compared to 15 to 40 percent with existing conventional solar panels.

Second, the ability to store heat for use whenever it’s needed will become more valuable as more solar power is installed. Germany, which has far more solar power than any other country, sometimes has to pay its neighbors to take excess solar power generated on some sunny days. “This program is looking out to a future that might be tomorrow in Germany, three years away in California, five years away in Arizona,” Branz says. “But eventually this future will come to everywhere that people want to generate a lot of electricity with solar energy.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.