by Tyler Norris
May 17, 2017
Ever since a team of Manhattan Project scientists created the first self-sustaining nuclear chain reaction in 1942, the United States has led a multi-decade project to develop new energy technologies.
Beginning with its invention of nuclear power and solar photovoltaics, peaking with its creation of the Department of Energy in 1977, and culminating in the Obama administration’s clean energy innovation programs, the United States spent tens of billions of dollars to transform the underlying technologies that power the modern world.
That historical period is now over.
On the surface, the proximate marker of its conclusion was the election of Donald Trump and his administration’s commitment to dismantle the Department of Energy and wipe out the Mission Innovation program that sought to double U.S. and international spending on energy research and development.
It is clear that the U.S. is not setting itself up to lead another wave of energy innovation — at least not one based on new technologies.
Beyond any single policy decision, the more fundamental reason for the end of the expansionary period in energy innovation has to do with basic technological realities. The world has largely exhausted the low-hanging fruits of the physical sciences, and the likelihood of a breakthrough changing the existing technological landscape is vanishingly small. Meanwhile, the developed world is increasingly unwilling to pursue complex, large-scale projects like nuclear power plants and carbon capture and storage systems, let alone the challenge of securing nuclear waste and subterranean carbon dioxide for millennia.
The other technological reality — one that has fully transpired in just a few short years to send shock waves through the energy industry — is that we have succeeded at making clean energy cheap. Solar is on track to become the world’s cheapest source of power on an unsubsidized basis, even in regions with modest insolation rates. By the 2030s or sooner, electrochemical battery costs will decline so far that stored solar electricity will be nearly as cheap as natural gas electricity, and electric vehicles as affordable as conventional cars — and even cheaper if the true costs of fossil fuels are accounted.
After 70 years of public and private investment, the U.S. and its allies have provided the world with the tools necessary to end the fossil fuel age.
To be sure, there will be further developments in energy technology, and Congress should attempt to salvage as much of the Department of Energy as possible. The Chinese government or a group of independent billionaires may decide to spend the billions of dollars necessary to scale an advanced nuclear reactor or carbon capture and storage system. These are possibilities that should be encouraged — and ones that are unlikely and unnecessary for decarbonization.
The energy endgame is in sight. The pre-eminent questions around the long-term habitability of the planet are whether the deployment of renewables and electric vehicles can be dramatically accelerated between now and mid-century, and whether existing sources of low-carbon baseload electricity can be preserved instead of replaced by fossil fuels. Even with cheap and scalable alternatives, we cannot wait for normal market structures to turn over the power system and vehicle fleet. What’s required instead is a relentless deployment push at every level, while preventing further development of fossil fuel infrastructure. Because we are already out of time.
Incremental revolutions
Between 2004 and 2014, venture capitalists plowed about $36 billion into cleantech startups. They had seen the disruptive change in information technology and wanted to apply a similar formula to the energy sector — and with it, reap the quick and outsized returns they expect. Their experience was quite different.
As dozens of companies started going bankrupt after 2010, venture capital firms gutted their clean energy investment portfolios by 75 percent — and while a few firms did well, up to half of that $36 billion may ultimately be lost.
One of the hard lessons we’ve learned since the turn of the millennium is that energy is among the least innovation-intensive sectors in the economy, especially in electric power and transportation.
In the power sector, business models are still largely based on investment in large-scale assets — hundreds of millions or billions of dollars per power plant or transmission system — that are designed to be built and left untouched for decades at a time. These assets sell a single, undifferentiated commodity — electrons — to consumers who until recently have cared little about their source, as long as the light switches work.
It’s no surprise then that electric utilities have traditionally spent around 0.1 percent of revenues on R&D, compared to more than 10 percent in other sectors. Distributed resources and changing utility business models are starting to change this reality, but small-scale solar still accounted for less than 0.5 percent of U.S. power generation in 2016.
The fracking revolution is often upheld as a shining example of energy innovation. More precisely, it should be understood as another example of technology change in upstream fossil fuel extraction, alongside mountaintop removal, tar sand production, and offshore drilling — all of which inflict serious environmental externalities.
The long history of ever-more efficient extraction and the ultra-low prices it has enforced for oil, gas, and coal is a central reason why innovation in alternatives has been so difficult. That cheap gas has recently helped displace coal and balance intermittent renewables shouldn’t obscure the fact that cheaper fossil fuel extraction is an obstacle to the energy innovation we need.
The transportation sector is more favorable to innovation, given that consumers look for differentiation across vehicle models and retire or resell their cars and trucks more quickly than power assets. But even in transportation, the pace of change has been extremely slow. Hybrid vehicles were first introduced in the U.S. almost 20 years ago, but their share of total car and truck sales peaked at just 3 percent in 2013 — only to decline to 2 percent in 2016 amid low gasoline prices.
Tesla and other all-electric manufacturers can change this game, but even with major battery cost reductions, it will take decades to replace the conventional vehicle fleet. Even if electrics skyrocketed to 60 percent of all U.S. car and truck sales tomorrow and every year thereafter, it would take until the late 2030s to convert half the vehicle fleet.
The upshot is that most technological change in the energy sector occurs gradually based on evolutionary advancements in the performance and cost of technologies that accumulate over time, usually decades, rather than through scientific breakthroughs. The small number of exceptions — arguably in nuclear power, combined cycle gas turbines, and LEDs — proves the rule.
When a lab breakthrough does occur, it requires years to determine if it can be commercialized, subject to volatile market and policy fluctuations. And once a new energy technology has been successfully commercialized, it takes decades for it to diffuse throughout the capital stock. As energy historian Vaclav Smil has documented, it took 40 years for oil and 60 years for natural gas to grow from 5 percent to 25 percent of world primary energy supply.
Business as (un)usual
And yet, change happens. The gasoline cars we buy get slightly more efficient every year with the application of more advanced fuel injection systems. New natural gas power plants require slightly less gas for every kilowatt-hour of electricity they produce. Our upgraded appliances provide better services but our electric utility bill goes down. These changes appear small in any given year, but over decades they amount to large impacts on the energy balance.
The same holds true in renewable energy. Arguably the most important change occurring in the energy sector today is the steady cost decline in solar modules and lithium-ion batteries — about a 20 percent reduction in solar for every doubling in cumulative installations and 15 percent to 20 percent for lithium-ion batteries. Similar to Moore’s law for semiconductors, this “learning rate” accounts for cost reductionsfrom greater economies of scale, improved manufacturing processes and cell design, material costs, and learning-by-doing — together cutting hardware costs by 80 percent within a decade.
These cost reductions are now widely understood. What’s less recognized is that by 2030 or sooner, the installed cost of solar and battery systems will be so low — under $750 per kilowatt for solar power plants and $300 per kilowatt-hour for grid-scale battery systems — that stored solar electricity will be nearly as cheap as electricity from low-cost gas power plants, without subsidies. If the true costs of gas electricity are accounted for — that is, if gas plants aren’t allowed to dump greenhouse gases into our atmosphere for free — stored solar will be even cheaper.
All of that is before the tidal wave of dirt-cheap batteries arrives from retiring electric vehicles. In the U.S. alone, by 2035 a cumulative battery fleet capable of storing a terawatt-hour of electricity — nearly 10 percent of the entire country’s daily electricity use — will flood the market as old electric vehicles are retired, sending storage costs plummeting further, according to Pira Energy Group.
Meanwhile, batteries will be capable of operating for increasingly long durations, reaching as high as 6 hours for lithium-ion and up to 10 hours or more for flow batteries and other chemistries — solving the problem of daily renewable power intermittency.
None of these are particularly optimistic scenarios. They are the new business-as-usual forecast, and if anything, they are too conservative. After all, every major energy forecaster has underestimated the pace of cost reductions to date.
It is difficult to overstate the significance of this achievement. It is no exaggeration to say that it may someday be recognized as one of humanity’s most consequential technical accomplishments, since it currently represents our best hope for avoiding catastrophic climate change and providing energy access to billions of humans who have been left behind by existing energy systems. But to realize its potential, we must be clear-eyed about the remaining obstacles.
Baseload goes bust
A quiet revolution has also occurred in nuclear power and carbon capture and storage — in the opposite direction. Dramatic cost overruns, glacial technology development, fierce public opposition, and ongoing safety and regulatory issues have all conspired to dash the hopes of a nuclear renaissance or carbon capture fix. The recent Westinghouse bankruptcy may be only the tip of the iceberg, and even in developing economies, the risk is sharply to the downside.
Until now, the majority of the world’s energy analysts assumed that new nuclear reactors and fossil fuel power plants with carbon capture would be a linchpin in meeting climate targets. This is no longer a reasonable assumption.
The smart money will no longer support new nuclear, and where it does, required interest rates and public insurance will be so extreme that only the most committed governments will underwrite their development. Some carbon capture projects will go forward, but without a high carbon price and major public investment in technology scale-up, it will not achieve more than marginal penetration.
The burden of evidence is increasingly on the side of nuclear and carbon capture developers to demonstrate that new technologies can make a meaningful contribution to decarbonization. In the meantime, the world cannot wait or craft its plans with rose-colored glasses. For this reason, it is understandable why there is such deep skepticism about nuclear and CCS, especially in comparison to solar and wind — and why so many aspire to 100 percent renewable energy, even if the pathway there isn’t entirely mapped out yet.
Deep decarbonization
In 2016, solar and wind generated about 7 percent of total U.S. electricity, with another 7 percent from hydropower and 2 percent from biomass and geothermal — compared to 64 percent from fossil fuels and 20 percent from nuclear. If a large expansion of nuclear, carbon capture, and hydropower is off the table, and if biomass and geothermal double or triple (at most), there is effectively one option: Natural gas and coal will have to drop to nearly zero and solar and wind together will have to grow to supply 60 percent to 80 percent of electricity.
The good news is that after years of real-world observation and modeling, we now know this is technically and economically feasible. The bad news is that anything beyond this level quickly increases system costs due to weather fluctuations that make available sunlight and wind speeds uncertain. Even with very large amounts of low-cost storage for daily and inter-seasonal purposes — batteries, thermal systems, pumped hydro, and other technologies — creating a grid with over 80 percent solar and wind would require an expensive overbuild of renewable capacity and a huge fleet of backup fossil fuel generators.
The best analysis today tells us that making solar and wind the dominant sources of energy will be substantially cheaper if we maintain a minority share of low-carbon baseload. That’s why it’s essential to preserve and extend as many existing nuclear plants as possible and continue making long-term public investments in advanced nuclear and carbon capture technology, even if their scale-up is less than likely — and even if the United States government doesn’t lead.
The energy endgame
The long-term challenge of fully decarbonizing the energy system requires planning and investment. But the much bigger challenge facing us today, in the here and now — the one that demands an immediate mobilization of resources at a scale and pace we have not seen in peacetime — is dramatically accelerating the deployment of renewable power and electric vehicles.
Whether or not this can be accomplished will determine the bulk of the world’s carbon trajectory over the next half-century and the allocation of tens of trillions of dollars.
For years, the prevailing assumption among policy analysts was that subsidies and regulations should be phased out as low-carbon technologies reach cost parity with fossil fuels to create a “level playing field.” This has it backwards. Policy ambition needs to continually rise, not fall, as the technologies get cheaper if there’s to be any chance of meeting emission targets. And no policy will be a silver bullet or a final victory — not even a high carbon price.
The essential realization for policymakers is that existing market structures and regulations cannot get us there. Even with cheap renewables and storage, the turnover rate in the capital stock — how quickly the existing fleet of power plants and vehicles is retired and replaced — is simply too long to wait for these technologies to diffuse on their own. Electricity markets need to be reinvented. Low-carbon power standards, fuel economy requirements, and pollution regulations have to be repeatedly ratcheted up. Operating coal and gas plants will need to shut down, and in many cases, new fuel pipelines will have to be blocked.
Every year is a race against time. Every solar installation, and every battery shipment, is a step down the cost curve. Every state and city policy, every corporate procurement, and every investor commitment matters — especially in the absence of U.S. federal leadership.
As his parting reflection on his experience directing the Manhattan Project, Lieutenant General Leslie Groves wrote: “Those of us who saw the dawn of the Atomic Age that early morning…know now that when humanity is willing to make the effort, it is capable of accomplishing virtually anything.”
He may have been wrong about the technology. But if we are to secure civilization once more, we must mobilize today as if he was right about our capabilities. We have the tools. Time is short. Much is at stake.