5

PILLAR 1: ELECTRICITY GENERATION

GROWING DEMAND

By 2050, billions of people will have new or increased access to energy. If clean energy becomes cheaper than expected—as it must in many cases to spread rapidly enough in developing countries—even more people will be able to afford larger amounts of energy. And the amount of sequestration required by 2050 will demand significant additional energy inputs.

Nearly all of this growing demand for energy services will have to be met with electricity—either directly, in the case of equipment that is already or will be electrified; or indirectly, through synthesized fuels (and some amount of biofuels) that can substitute for fossil fuels and which require large amounts of electricity to synthesize. Currently the world generates slightly under 25 petawatt-hours (PWh) of electricity per year. Most models show this doubling to around 45–50 PWh per year by 2050.¹

However, if we want to guarantee negative emissions by 2050, we have to aim for a dramatically higher number. Models that project around 45–50 PWh of electricity generation per year either show continued large amounts of emissions, with only modest increases in electrification, or they rely on the assumption that nearly every country will impose stringent efficiency policies and that energy demand will decrease dramatically even while economies continue to grow. Models that show near-zero emissions by 2050 also all rely on significant amounts of biomass-based fuel substituting for fossil fuels. As we’ll discuss in Chapters 7 and 9, this can certainly happen to some degree, but a strategy that relies so heavily on biomass would require strict regulation in every country (which we’d have to count on not only being implemented, but not ever being loosened once the energy system was dependent on so much biomass), which is highly unlikely in our thirty-year timeframe.

Under a framework that charts a path to negative emissions by 2050 without relying on strong regulations in every country, we must expect that electrified options (direct equipment and synthesized fuels) will have to be made significantly cheaper, and possibly subsidized by solution-focused countries, so they spread rapidly across the world. We must assume that fossil fuel use with carbon capture and storage (“CCS,” which most models also rely heavily on) will be limited to countries with strong mandates or incentives (see later in this chapter). We must assume that only a modest amount of biomass can contribute to the total energy demand, given the likelihood that growing biomass for fuel would contribute to deforestation emissions absent global regulation. We must assume that some amount of direct sequestration of atmospheric CO₂ is needed to close the gap to zero emissions and achieve negative emissions.

And therefore, we must assume that the world will need not 45–50 to fifty PWh of electricity by 2050, but more like 100–150 PWh. This is about the range we’ll need if 2050 end-use energy demand meets the current-trajectory projection but we electrify roughly 60% of energy use (the upper bound of most electrification models,²) with some portion of the other 40% served by synthesized fuels (the rest served by fossil fuels with CCS and biofuels), and with levels of electricity-powered sequestration in line with those discussed in Chapter 9. If countries don’t come through with efficiency measures to meet their existing energy commitments (e.g., those from the Paris Agreement, which indeed no major country is on track to meet), total demand and the need for non-regulatory solutions could push this number even higher, toward two hundred PWh per year of electricity generation. Thus, depending on how much policy actually gets implemented, electricity generation might need to grow to anywhere between two and eight times its current levels by 2050—most likely about four to six times.

The energy system in 2050 must be mostly powered by electricity—either directly through the grid, or indirectly through the electricity-powered synthesis of clean fuels. Electricity will also be needed to power some portion of the required sequestration each year. Biofuels and geothermal, and, in countries with strong policy, also fossil fuels with CCS, will make up the remainder of total energy demand.³

Most activists acknowledge that we are not on track to replace the current-size electricity generation system with clean equipment. Most models project that we will in fact need to build a clean electricity generation system double the current system’s size. But in reality, we have to plan for a solution that doesn’t rely on massive efficiency measures in countries like India, nor on strictly enforced biomass-growing regulations worldwide. A 100% solution can still be readily achieved without every country’s long-term cooperation, but only with two to three times more clean electricity generation than anyone is talking about.

If the goal is certainty under all policy scenarios for a negative-emissions 2050, we are two to three times more off track than most people are thinking.⁴

There’s one more challenge: most solar and wind facilities are only rated for thirty years or less, and though nuclear facilities last somewhat longer (forty to sixty years or sometimes more), most of those currently in use were built quite a while ago.⁵ So, most of the clean electricity generation capacity currently installed will be retired by 2050 (some hydro and bits of the others will remain). This means that the electricity generation system we need—one four to six times larger than what we currently have—will have to be built in thirty years almost from scratch.

A 100% solution can still be readily achieved without every country’s long-term cooperation, but only with two to three times more clean electricity generation than anyone is talking about.

RAPID BUILDOUT

That’s already a difficult undertaking, but the world should aim for the even higher goal of deploying as much of that new generation as possible in the near future rather than waiting until the 2040s for most of the electricity generation transition. Partly that’s because the sooner any given sources of emissions are eliminated, the less cumulative CO₂ will be in the atmosphere by 2050 and therefore the less sequestration will be needed. Because of its centralized nature, fossil fuel–based electricity generation, especially from coal power plants, holds the single biggest chunk of emissions that can be eliminated in the near future.

Furthermore, the sooner clean electricity generation is definitively cheaper than fossil electricity generation, the easier it will be to deploy, because it can compete with new fossil equipment that would otherwise be built. Convincing a company or country to build a solar or nuclear plant instead of a new coal plant they would otherwise build is fairly easy if the clean option is cheaper. But convincing them to replace a just-built, operating coal plant with a clean plant is very hard. The payback would have to be extremely fast—maybe clean electricity generation can get there eventually, but to maximize the chance of fully eliminating fossil fuel electricity generation by 2050 or sooner, it would be wise to frontload as much of the deployment of clean generation as possible so that it can mostly compete with potential new fossil plants, not existing ones.

That’s a large obstacle given that a key limitation of all clean electricity generation sources is how expensive they are up front despite costing little or nothing to operate (in the case of wind and solar, most important is the up-front cost of storage and load-balancing technologies to enable significant use of these intermittent sources). At least one clean generation option must come down in capital costs, so that when short-term decisions are made based on the cheapest way to add an additional bit of electricity generation output, a clean option is always chosen over fossil fuels.

SOLAR

If our window for solving climate change were more like a hundred years, solar might be the perfect technology. Millennial author Varun Sivaram laid out all the potentials and limitations of solar in his insightful book Taming the Sun.⁶ Solar comes with great benefits for the long-term: it’s easy to scale from tiny to huge installations, so you can test out new solar cell technologies without investing huge amounts of money. Eventual technologies for organic or “perovskite” solar cells will be printed on roll-to-roll manufacturing machines, a form of rapid, continuous production that will be much faster and cheaper than current batch-based and energy-intensive silicon processing.

Roll-to-roll printing of solar cells.

While such roll-to-roll solar cells may eventually become extremely cheap, there are two problems. First, they aren’t there yet, and we need something ready to scale up for the initial short-term electricity buildout the world requires (they might still contribute meaningfully toward the total system by 2050). Second, to power a high percentage of the electricity generation system, they’d need batteries (or other forms of storage or grid flexibility), which are subject to the same development and manufacturing ramp-up constraints with less certainty for success in low cost. That means that although solar might contribute heavily to the 2050 system, it can’t form the majority of electricity generation by then. Eventually, by the end of the century we expect battery technologies will have caught up and solar will be viable to use for a larger portion of generation.

And so, solar should be expanded massively, but given the enormous amounts of material and land required for solar manufacturing and deployment, and the lack of cheap batteries, it will probably amount to less than a third of the total 100–150 PWh demand for electricity generation by 2050. Indeed, Sivaram, who is optimistic about solar’s long-term potential and who works on commercial-scale solar deployment, puts meeting one-third of global demand as the upper goal for 2050 under the normal-assumption scenario requiring only 45–50 PWh of total generation.⁷

To get there, research and testing will have to pursue organic and perovskite roll-to-roll printed solar cells, so that the cost of solar keeps declining in the next few decades. The more solar is deployed on the grid, the more costly it becomes to deploy because of the need to keep flexible fossil fuel generation, or storage, available to run at a moment’s notice (but not to use most of the time, which makes that backup uneconomical in itself). After a point, additional solar will generate electricity almost entirely during times when excess electricity is already being generated by existing solar, making the value of new solar drop as more is deployed. To continue being economical, its costs will also have to drop, and energy storage innovations, as well as policy designs to transform grids, will have to be deployed. These could include adding more long-distance lines to bring excess solar from midafternoon in Western Europe to serve evening demand in Eastern Europe, or adding more responsive end-use equipment (for example, heaters that choose the time of day to heat based on electricity availability).

In the near term, policy mandates can keep accelerating solar deployment and can help bring new solar technologies to the demonstration and early commercialization phases. Policymaking and activist messaging should keep in mind that the physical realities of engineering a grid with solar are seriously different in different parts of the world. In northern regions, solar produces so little compared to its theoretical peak output that it can only contribute to a form of “efficiency,” that is, reducing total demand for other generation but not shifting the source of that generation away from fossil fuels. In regions with strong winters, the difference in solar production in summer vs winter is enough to require a backup generation system (or months-long storage system, which currently no technology can come close to providing at any reasonable cost) to replace most of the solar generation during the winter. In regions closer to the equator with more consistent sunshine, solar faces few of those problems and merely requires hydro or methane backup (or nuclear backup, or storage, if either is developed to be affordable soon) for nighttime.

WIND

Wind provides the cheapest renewable electricity generation in many regions that don’t have abundant hydropower.⁸ And unlike hydro, which requires specific conditions that are not abundant worldwide, wind can work in a larger number of places. The total potential for wind generation is not nearly as much as the theoretical potential for solar, but it is far more than hydro and possibly close to the economical potential for solar. Wind could probably be scaled up to power about a third of total demand by 2050, if it was pushed significantly.⁹

Unlike solar, though, wind has already been improved technologically to achieve about the lowest costs we can expect.¹⁰ Offshore wind dramatically expands the areas viable for wind deployment, but is significantly more expensive than onshore wind. Like solar, onshore wind can compete on cost with methane power plants when the wind is blowing, but wind also suffers from the problem of intermittency, and from the large amount of land and materials required to harvest a diffuse energy source. All the same grid and demand policies and storage technology improvements that must be applied to solar could also make wind a viable provider of a larger share of global electricity generation demand.

In this analysis from energy consultancy NorthBridge, which modeled a hypothetical 100% wind and solar system for New England (with the proportion of wind and solar that minimized surpluses and deficits), electricity would have to be stored for many weeks to several months at a time.¹¹

Together—depending on how dramatically solar and storage technologies improve in cost, and on the volume of flexible grid demand that is built—solar and wind might be able to power one- to two-thirds of the total 2050 electricity generation demand. Much of the buildout, however, will likely happen in the second half of the transition, partly because it will take time and innovation effort for solar and storage technologies to drop further in cost, and particularly because the flexible demand loads that will enable much larger percentages of solar and wind to be economical on the grid will be much more common when fuel synthesis and sequestration scale up. To the extent that solar and wind buildout can accelerate, policy incentives and mandates will have to drive most of the increased pace for the next decade or two.

HYDRO

Hydro is currently the largest clean electricity generation source in the world, but the limited availability of good sites for hydroelectric dams restrict where it can be deployed, and most of the best ones have been taken. As discussed in Chapter 2, hydro is cheaper than fossil fuels where it is plentiful. Because it is a fairly centralized and concentrated form of energy, one new hydro project adds a lot of generation, meaning hydro can grow quickly as an electricity source. But new sites will encroach on natural ecosystems and native lands, and will therefore face political hurdles that limit the pace of buildout. Small-scale hydro faces fewer hurdles, but also has a much smaller potential for adding up to large amounts of generation.

In total, hydro may be able to double globally, possibly triple if we pushed it to the extreme.¹² Even a tripling would still cover only about 10% of the 2050 electricity generation demand.

The advantage of hydro for the larger electricity system is that it is “dispatchable,” meaning it can generate more or less in a given instant depending on grid demand. Water gets stored up behind dams and released at peak demand times. The rate of water flow can be varied to alter the power moment to moment. This means hydro can serve as an excellent backup to solar and wind, filling in for their generation when the sun goes behind a cloud or sets for the night and when the wind calms. The more hydro that is added, the more solar and wind will be viable to add in the same regions.

NUCLEAR

Unlike solar and wind, nuclear deployment is not currently accelerating. In fact, for several decades the nuclear industry has totally floundered, and in Europe and the United States the costs for the few new nuclear plants that have been built have been extraordinarily high. However, in China and South Korea nuclear has been built in the past decade for about the same costs as some fossil fuel plants (in Europe and the United States in the early days of nuclear plant deployment, it was even cheaper).¹³ There remains some hope that if scaled up with better supply chain and manufacturing practices, nuclear costs could drop below those of the cheapest fossil fuel plants.

If this can happen, the advantages are tremendous. Nuclear is the most concentrated source of energy we know of for generating electricity. That means that the amounts of land, raw materials, manufacturing labor, and waste are all much smaller than those for solar or wind. It also means nuclear can be built out faster in terms of generation capacity.¹⁴ And unlike solar and wind, nuclear power is not intermittent, and tends to run near its peak capacity almost all the time. That means that for every unit of peak capacity added, nuclear is generating far more actual electricity than solar, wind, or hydro.

So nuclear holds the potential to add clean electricity generation at the incredibly rapid pace needed. There are two problems: politics and economics.

The political problem is that people are scared of nuclear power, especially nuclear waste. This may be starting to change among the most intensely focused Millennial climate activists—for example, the Sunrise Movement has usually been careful with its rhetoric to refer to “clean” rather than “renewable” electricity so as to leave the door open to nuclear (“renewable” means that the supply is theoretically inexhaustible as long as the sun is shining on the earth; as nuclear permanently consumes small amounts of fuel it is not renewable, but is nearly inexhaustible for practical purposes)—but generally these fears are still pervasive. The idea of radiation triggers disgust and taints nuclear power as something “unnatural.” Scientific measurements point in a different direction: even during the famous Fukushima nuclear incident, radiation exposure outside the plant never exceeded medically established limits, which are themselves well below natural levels of background radiation that exist in some locations in the world.¹⁵ We’re all walking around in a soup of radiation, but we don’t usually think about it.

Radiation from nuclear power plants has only ever killed people—or even made anyone measurably sick—once, at the Chernobyl power plant. At that plant, which did not have the containment vessel that all reactors are built with today (and always have been, outside the Soviet Union), operators decided to carry out a foolhardy experiment with few safety precautions and a perfect storm of conflicting personalities. As the experiment started to go wrong, a series of bad decisions led to a set of explosions and the destruction of the reactor. Two people were killed by the explosions, and a few dozen more from acute radiation exposure. From all the radiation released, estimates put the upper limit of potential deaths over the next few decades at about four thousand (more recently, even that number has been called into question and numbers lower than one hundred have been suggested).¹⁶ For perspective, four thousand is about the number of people killed by coal every couple days—mostly from respiratory illness caused by the particulate matter released when coal is burned.¹⁷

Nuclear is in fact the safest form of energy we have—even slightly safer than solar (which sometimes kills installers who fall off roofs) and much safer than hydro (which occasionally kills whole communities when a large dam breaks and towns are flooded).¹⁸

The revulsion people have about nuclear waste is similarly connected to a philosophical uneasiness about radiation. But this is perhaps the most misplaced concern of all those held against nuclear: the waste involved is easy to contain, so radiation from nuclear waste has never been a problem. And because nuclear is such a concentrated form of energy, its waste comes in far smaller volumes than waste from fossil fuels. In fact, nuclear produces even less waste than solar because of the chemicals used in manufacturing solar panels and the shorter lifetime of solar cells themselves.¹⁹ All the wastes in question are toxic, but in this case radiation is actually an advantage: over time, radioactive waste decomposes and reduces its own volume. The other wastes stay toxic forever.

For true perspective, we must consider the question of wastes in the context of climate change: we need massive, immediate additions of clean electricity generation. We don’t have long-term sustainable disposal systems for nuclear, solar, or coal wastes. We’re currently generating massive amounts of coal waste, and to replace it we’re going to have to generate more (but far less massive than coal) solar and nuclear wastes. Those, especially in the tiny volumes that nuclear waste comes in, can be safely stored in “intermediate” storage for many decades, giving us time to focus on the dramatically more urgent problem of solving climate change. Eventually, we can re-focus on developing sustainable disposal systems for all of these wastes—maybe by that time, current proposals for nuclear reactors that can consume (and therefore dispose of) nuclear wastes as fuel will have even been developed to commercial viability.

Roughly-proportional amounts of waste from coal, solar, and nuclear visualized.²⁰

The immediate economic problem of nuclear power is much more pressing. Even the cheapest nuclear plants built in the last decade are too expensive to outcompete fossil fuels globally, let alone do so in the immediate future. To bring the costs of nuclear down enough to rapidly outcompete coal and methane, the key is standardization. So far, nuclear reactor designs have changed every few reactors built. Any single design has been executed a couple dozen times at the most. This means the technology is constantly starting at the beginning of the cost-lowering scale up process that solar and lithium-ion batteries have undergone dramatically in the past two decades. “First of a kind” plants are always much more expensive than “nth of a kind” plants.

Bringing down the cost of nuclear through standardization and scale-up of manufacturing is also the most likely route to the short-term addition of clean electricity generation we need.

Government efforts can be directed to convene industry players and spur adoption of best practices—for example, construction planning practices to minimize financing costs (loans taken out to pay for each new plant). Public funding, through direct procurement, power purchase agreements, or simply loan guarantees, can kick-start the construction of the first batch of standardized plants.

South Korea and China both have existing designs that they have built several times at reasonable costs.²¹ International collaborations to begin mass-manufacturing those designs could be the fastest way to add clean electricity generation immediately. Such collaborations, though, might be hindered by the recent shift in South Korean political inclinations against nuclear, akin to stigmas in the United States and Europe. And Chinese politics are always hard to predict because of China’s centralized government system. Nonetheless, leaders who could collaborate with them should explore a scale-up in exporting one or both countries’ reactors.

At the same time, there are many designs for advanced, “fourth generation” nuclear reactors, and a few dozen startups trying to commercialize their version.²² None of these are on track to add large amounts of clean electricity generation as rapidly as we need, but public convening, funding, and support (including testing and demonstration facilities, and guaranteed government purchases if new designs meet certain targets) could accelerate these fourth-gen reactors to commercialization in time to contribute to the short-term buildout of clean generation we need. Many fourth-gen designs have other benefits for the eventual electricity system, such as flexibility to ramp up and down at a moment’s notice (a feature known as “dispatchable generation”), fewer operators needed, passive safety features to ease public concerns and make the politics of deployment easier, and modular sizes to allow deployment in a wider range of locations.

Often, proposals for cheaper, standardized nuclear plants revolve around mass-manufacturing them in a factory or shipyard. Shipyards are known for on-time and at-cost delivery of durable large industrial products. There are only a few with enough size and quality to contribute to a rapid buildout of nuclear, so more would have to be built soon, but they could each process dozens of plants per year. These plants could be floated to their eventual deployment sites (some proposals even suggest skipping the step of towing them onto land and preparing a construction site and instead operating the plants in the ocean or a river right next to shore).

With smart business management, mass-manufactured nuclear plants could probably drop below the cost of the cheapest fossil fuel plants in the near future.

With smart business management, mass-manufactured nuclear plants could probably drop below the cost of the cheapest fossil fuel plants in the near future. Scaling up the supply chain (uranium mining and enrichment, shipyards and their workforce, turbine manufacturing, and the manufacturing of the nuclear plant) fast enough to transition significantly from fossil fuels to nuclear in a decade or so is a tall order, but within the realm of possibility.

FOSSIL FUELS WITH CCS

The final option for decarbonized electricity generation is to continue using coal and methane, but to filter out the CO₂ from the plant exhaust (some processes actually separate the CO₂ before burning the fossil fuel, leaving pure hydrogen to be burned and power the turbine), transport it to appropriate geologic sites, and bury it deep underground where it will not escape for a long time, if ever.

It might be a stretch to call this carbon capture and storage (CCS) clean, in that coal causes some degree of pollution, disease, and death (especially among coal miners) even before it gets burned, and methane causes large amounts of greenhouse gas emissions when it leaks during extraction and distribution (see Chapter 8). But fossil fuel power plants with CCS are certainly much cleaner from a climate change perspective. If political leaders can accept the continued health damage from coal (and hey, they’ve accepted it for a century already), then coal with CCS would be essentially emissions-free. Methane with CCS could be considered emissions-free if and when methane drilling and pipeline technologies improve to prevent leaks. Because of these uncertainties, CCS isn’t explicitly a part of this five pillar framework—it could of course decarbonize factories without electrifying them or using synthesized fuels, but we don’t know at the moment whether the methane ones would be truly zero emissions, or whether CCS in general has a significant chance of scaling.

The latter uncertainty is because of the economics of CCS: by definition, CCS involves extra equipment and extra steps on top of any industrial process. It will always be cheaper to run the exact same process without CCS. Therefore, virtually no company will ever adopt CCS technology for economic reasons alone. It will happen only in countries that either mandate it or put a significant enough price on carbon pollution that it becomes cheaper for a factory to use CCS technology than to pay the carbon price for its emissions. Part of the reason CCS doesn’t feature in this framework is this fact that policy is absolutely essential for its widespread use, which makes it unlikely to decarbonize as large a portion of electricity generation and industrial emissions as technologies that can come down in cost to the point that they actually save money.²³

However, all that said, CCS has one major advantage in countries where it is politically viable to mandate its use: CCS equipment could be added to most existing power plants for some significant but not prohibitive cost, and that would mean an extremely rapid first step in reducing power plant and factory emissions. In some countries, CCS might play a significant role in the 2050 energy mix, either because of political choices to allow fossil industries to transition less abruptly and therefore win support for larger policy proposals, or as a last-mile effort to eliminate remaining emissions when we approach 2050.

In other countries, CCS might not play a major role by 2050, but it does present a major opportunity for immediate emissions reductions in countries willing to enact sweeping policy. Remember that the sooner large sources of emissions are eliminated, the smaller cumulative amount of emissions will need to be sequestered from the atmosphere in 2050 (and atmospheric capture and sequestration is much more expensive than factory or power plant based capture and sequestration). CCS is highlighted here as a possible intermediary measure in countries with the political clout to use it, while we start mass-manufacturing standardized nuclear plants, and while solar and storage research is pushed forward to take a larger role later in the transition.

CCS equipment could be added to most existing power plants for some significant but not prohibitive cost, and that would mean an extremely rapid first step in reducing power plant and factory emissions.

OTHER EXISTING TECHNOLOGIES

Aside from solar, wind, hydro, nuclear, and fossil fuels with CCS, a couple of existing technologies could supply modest portions of the 2050 electricity generation mix. Geothermal heat can power some industrial processes, heat buildings and water, and sometimes create steam to generate electricity with.²⁴ Existing geothermal technology is limited to locations with the right geologic features and doesn’t have massive total potential for the energy that could be harvested globally, so it will likely play a small role.

Biofuels can substitute for fossil fuels, though as we will discuss in Chapters 7 and 9, they can only be scaled up modestly before they start causing more emissions, through deforestation, than they save by replacing fossil fuels. There are various kinds of biofuels, and some (such as from anaerobic digesters on farms) may be more sustainable than others (such as ethanol from corn that takes away land from food crops). Some models of a decarbonized 2050 rely on vast amounts of biomass energy,²⁵ which could be viable if most national governments impose ambitious regulations. But we cannot count on them to be carbon neutral in the more likely scenario in which national governments don’t implement strict enough rules or monitoring programs to guarantee biomass is grown without deforestation impacts. Therefore, biomass figures minimally in this framework, which is meant to show the viability of solving climate change without the ambitious and sustained international cooperation that currently seems unlikely.

EMERGING TECHNOLOGIES

Moonshot efforts for electricity generation should focus most immediately on a rapid scale-up of nuclear (or CCS where viable to mandate) to steeply reduce electricity generation emissions. These projects should also aim to bolster the acceleration of solar and wind deployment, particularly with a mind toward the extra level of competitiveness solar (and storage) might reach for the second half of the needed energy transition when more varied uses for electricity will become prominent.

At the same time, moonshot efforts should immediately start supporting basic research or “valley of death” phase work on new electricity generation technologies that may have no certainty of contributing, but that could turn out to play a major role if they were developed successfully.

Perhaps the most promising proposal is for deep geothermal heating and electricity generation, which can be deployed in far more locations than traditional geothermal, and which could supply higher temperatures to run power plants and some factories with no intermittency.²⁶

Another idea is for generating thorium-fueled (rather than uranium-fueled) nuclear power.²⁷ One advantage is that key large countries, including India, have more abundant domestic supplies of thorium than uranium.

Other ideas include undersea water current power, which is akin to offshore wind turbines, but has the turbines spun by water near the bottom of shallow ocean areas. Related proposals include tidal power and wave power systems.²⁸

What underwater current turbines might look like.

Nuclear fusion (rather than fission, which all current nuclear power is) has been said to be “thirty years away” for the last fifty years, but it would indeed be revolutionary if we could commercialize it successfully.²⁹ It deserves some level of research and testing support.

And then there are more out-there ideas that may be in the earliest lab stages. Who knows what someone may discover and whether it could have a chance of scaling to commercial usefulness? A portion of funding should be used to help along these innovations with low probabilities of success but with serious game-changing potential if they succeed.

STORAGE TECHNOLOGIES

Finally, solar and wind can only reach a high percentage of electricity generation if electricity storage technologies become much cheaper. Nuclear would also benefit significantly from the deployment of storage: Unlike solar and wind’s uncontrollable fluctuation with weather, current nuclear plants are hard to turn on and off, so they run at a steady power almost all the time, even when grid demand fluctuates. As nuclear becomes a large share of electricity generation, there may be times when the total output from nuclear plants is considerably more than total demand at that moment. Affordable storage technologies would improve the economics of this grid system tremendously.

The storage in question could be lithium-ion batteries, as dominate now. They will probably come down somewhat more in cost. But current projections don’t show them coming down enough that weeks or months of storage could be viable.³⁰ Other batteries may be invented or scaled up—already, “liquid flow” batteries are being demonstrated in early commercial deployment and show promise of eventually beating lithium-ion costs for grid-scale storage.³¹

The cheapest current form of electricity storage is pumped hydro—the practice of pumping water up a hill when there’s excess electricity on the grid, and letting it down through a hydro turbine when electricity is needed. This requires specific sites (mountains next to rivers or lakes), so its potential is limited, but it could be scaled up a bit. Similar gravity-based storage systems have been proposed, such as ones using concrete blocks and cranes, though none have shown serious commercial promise.³² In the physical storage category, pressurized air systems can store electricity, but they require either specific geological formations to pump air into, or undersea bags of air that get inflated.³³

A different and more likely category of storage is using excess electricity to synthesize fuels—some of which can simply be the synthesized fuels sold to other sectors, and some of which could power traditional power plants (or fuel cells) and therefore act like giant batteries.

Fuel synthesis being used as a form of electricity storage.

Relatedly, any process—such as atmospheric sequestration—which uses a lot of electricity but can run at any time (or vary its power demand throughout the day) can balance out the grid demand with the intermittent solar and wind generation and constant nuclear generation.

These “flexible loads” will become more common on the grid as fuel synthesis and sequestration technologies scale up. Flexible loads could also include water desalination plants and other equipment not related to the energy system. The key factor in how flexibly these loads can operate is how cheap the capital cost is for the equipment in question. If it costs a lot to build a fuel synthesis plant, a company may only make its money back within the lifetime of the plant by operating it close to full time. If a plant can be built extremely cheaply, it might be economical to run it only a fifth of the time when there is a lot of excess electricity. Therefore, public funding for research and startups working on low-capital-cost flexible loads should also be a major part of a strategy to help solar, wind, and nuclear scale up as much and as fast as possible.

BASIS OF A CLEAN ENERGY TRANSITION

Because three of the other four pillars rely heavily on a plentiful supply of cheap, clean electricity, Pillar 1 needs to be carried out most immediately. Doing so will also bring the benefit of early large-chunk emissions reductions that limit later sequestration needs.

Government convening could jump start scale-up projects in nuclear using standardized designs based on roughly current technology. Policies such as state/provincial or national clean electricity standards, carbon pricing, subsidies, and simplified permitting regulations for the siting and construction of solar, wind, and nuclear plants can speed up the deployment of ready-to-go technologies.

With mass-manufacturing and policy efforts combined, and possible mandates for CCS, governments can guarantee the first wave of additional clean electricity generation is deployed. Efforts should aim to deploy enough clean equipment to at least replace current electricity generation levels by around 2030, on the way to generating the 100–150 PWh of electricity per year needed by 2050. Government-supported research can ensure that technology for all generation options continues making significant strides so that the remaining deployment can happen cheaply in the following twenty years.