For solar power to become truly mainstream, how much should it cost? And is the industry on track to meet that target?
We tackle each of those questions in an article released today in the journal Nature Energy. For solar power to supply nearly one-third of the world’s electricity by 2050, it will ultimately need to cost around 25 cents per watt (in today’s dollars), fully installed. And that target may be out of reach without a major technological shift.
Why would solar need to meet such an aggressively low cost target in the future if it is already competing with fossil fuels in some markets today? We argue:
Cost-competitiveness for solar is a moving target. As solar’s share of the electricity mix increases, the cost of each new solar project must fall to compete. This ‘value deflation’ effect of solar at higher penetrations is a well-known theoretical concept but is rarely discussed as a matter of practice in the solar industry. […] Thus, the installed cost of solar must fall dramatically to enable 30% penetration by 2050. Existing literature suggests a value deflation effect of roughly 70% by that time. Therefore, if unsubsidized solar at US$1.00 per W would be competitive at low penetrations, a cost target of US$0.25 per W would enable solar to outrun value deflation in the long term.
What about the role of energy storage and load-shifting to mitigate this value deflation effect? Though they are important, we argue, these applications may provide only a partial solution:
The quantities of storage required to substantially offset value deflation are significant and diverse -- storage would need to buffer variability between different parts of the day (diurnal storage) as well as between seasons as solar’s output fluctuates in short and long cycles. One study of the California grid finds that, if the cost of storage in 2030 turns out to be 80% lower than existing benchmark projections, then value deflation for renewable energy at 30% penetration will be roughly one-third less severe…The same study of the California grid under 30% renewable penetration found that highly elastic demand -- responding to rates varying in real time -- would only alleviate 15% of the value deflation effect.
Existing solar technology, based on silicon, has consistently (and sometimes dramatically) fallen in cost. So why wouldn't the solar industry just incrementally improve its products to meet the cost target that the market might demand in the future? We reply:
At the very ambitious, lower end of that range [of possible future costs], silicon solar PV would be close to meeting the US$0.25 per W target to outrun the value deflation effect. But it would be a mistake for the solar industry to put on blinders in a sprint toward silicon solar cost reduction -- in decades, the industry may find it backed the wrong horse.
So which technology should the industry back? Rather than pick a new favorite, firms in the solar industry should invest widely to develop alternatives to existing technology. And those alternatives should include both new materials and processes to make solar panels as well as new applications, which together could upend solar economics. For example, highly efficient solar coatings integrated into a cityscape could help supply urban energy needs while adding little to the cost of new construction.
Our article expands on the justification behind our long-term cost target. And in this post, we’ll explain why long-term targets are a tried-and-true mechanism for industries to invest today in pursuit of breakthrough products tomorrow.
The most famous example of a long-term roadmap for technology development is Moore’s law. In 1965, Gordon Moore predicted that the number of transistors on an integrated circuit -- or microchip -- would double every two years. Over the subsequent 50 years, the industry met Moore’s target every two years with the regularity of a metronome.
Although the pace might finally be slowing, Moore’s legacy is indelible: he proved that an industry that prioritizes innovation can consistently meet targets that were once considered unrealistic.
Other industries got the message. Some of them were closely related to the semiconductor industry and benefited directly from its roadmapping efforts. For example, firms producing micro-electro-mechanical systems (MEMS) -- which include sensors like the accelerometer in your smartphone -- adapted the integrated circuit roadmap to design their own roadmaps for disruptive MEMS technology.
But even outside of high tech, numerous industries created technology roadmaps that set long-term targets and galvanized firms to invest in R&D. For example, the U.S. steel industry, in partnership with the Department of Energy, executed a research roadmap from 1997 to 2008. As a result, the energy intensity of steel production in the United States fell 30 percent, and the industry continues to fund long-term breakthrough technology development.
A more ambitious solar industry
The solar photovoltaic (PV) industry actually has a technology roadmap, developed by a consortium of silicon solar manufacturers. But the roadmap is less a set of ambitious targets than a compilation of forecasts for how the industry is most likely to evolve. That is, the solar roadmap reacts to industry trends instead of shaping them.
This stands in stark contrast to the roadmaps introduced above, which set ambitious targets that firms could only achieve through innovation. The Department of Energy’s SunShot Initiative is more ambitious and has been instrumental in U.S. solar cost reductions over the past five years. But its target so far extends only through 2020.
Now is an ideal time to set longer-term targets. Producers of solar cells and panels enjoyed higher margins in 2015 than in the prior five years. And with the five-year extension of the U.S. federal Investment Tax Credit for solar, as well as supportive policies in China, India and other major economies, the market will continue to grow. During this period of relative stability, the industry should step back and re-evaluate its long-term technology trajectory.
Companies should pool resources to fund collective research and standards-setting efforts, a model successfully demonstrated by the semiconductor industry. And they should not assume that the only improvements worth investing in are incremental to existing products.
Ultimately, even with rapid technological advancement, 25 cents per watt may be out of reach by mid-century. But striving to achieve such an ambitious goal will only benefit the industry in the interim. By investing in long-term innovation, the solar industry can lay the foundation for its prolonged global success.
Listen to a discussion between Varun Sivaram, Shayle Kann and Stephen Lacey on why the solar industry needs to target these ambitious cost reductions:
Varun Sivaram is the Douglas Dillon fellow at the Council on Foreign Relations. Shayle Kann is the Senior Vice President of GTM Research. You can access the article here.