by Bill Gates
I wasn’t always aware of how much we rely on electricity, but over the years I’ve gradually come to see how essential it is. And I really appreciate what it takes to deliver this miracle. In fact, it’s fair to say that I’m in awe of all the physical infrastructure that makes electricity so cheap, available, and reliable. It’s downright magical that you can simply turn a switch almost anywhere in a well-off country and expect the lights to come on for a fraction of a penny. Literally: In the United States, leaving a 40-watt lightbulb turned on for an hour costs you about half of one cent.
I’m not the only one in the family who feels this way about electricity: My son, Rory, and I used to visit power plants for fun, just to learn how they worked.
After a family visit to the Þríhnúkagígur volcano in Iceland in 2015, Rory and I checked out the geothermal power plant next door.
I’m glad I’ve invested all that time learning about electricity. For one thing, it was a great father-son activity. (Seriously.) Besides, figuring out how to get all the benefits of cheap, reliable electricity without emitting greenhouse gases is the single most important thing we must do to avoid a climate disaster. That’s partly because producing electricity is a major contributor to climate change, and also because, if we get zero-carbon electricity, we can use it to help decarbonize lots of other activities, like how we get around and how we make things. The energy we give up by not using coal, natural gas, and oil has to come from somewhere, and mostly it will come from clean electricity. This is why I’m covering electricity first, even though manufacturing is responsible for more emissions.
Plus, even more people should be getting and using electricity. In sub-Saharan Africa, less than half of the population has reliable power at home.*1 And if you don’t have access to any electricity at all, even a seemingly simple task like recharging your mobile phone is difficult and expensive. You have to walk to a store and pay 25 cents or more to plug your phone into an outlet, hundreds of times more than people pay in developed countries.
860 million people don’t have reliable access to electricity. Fewer than half the people in sub-Saharan Africa are on the grid. (IEA)
I don’t expect most people to get as excited about grids and transformers as I do. (Even I can recognize that you have to be a pretty big nerd to write a sentence like “I’m in awe of physical infrastructure.”) But I think if everyone stopped to consider what it takes to deliver the service we now take for granted, they would appreciate it more. And they’d realize that none of us want to give it up. Whatever methods we use to get to zero-carbon electricity in the future will have to be as dependable and nearly as affordable as the ones we use today.
In this chapter I want to explain what it will take to keep getting all the things we like from electricity—a cheap source of energy that’s always available—and deliver it to even more people, but without the carbon emissions. The story starts with how we got here and where we’re headed.
* * *
—
Considering how ubiquitous electricity is today, it’s easy to forget that it only became an important factor in most Americans’ lives a few decades into the 20th century. And one of our early major sources of electricity wasn’t any of the ones that we think of today, like coal, oil, or natural gas. It was water, in the form of hydropower.
Hydropower has a lot going for it—it’s relatively cheap—but it also has some big downsides. Making a reservoir displaces local communities and wildlife. When you cover land with water, if there’s a lot of carbon in the soil, the carbon eventually turns into methane and escapes into the atmosphere—which is why studies show that depending on where it’s built, a dam can actually be a worse emitter than coal for 50 to 100 years before it makes up for all the methane it’s responsible for.*2 In addition, the amount of electricity you can generate from a dam depends on the season, because you’re relying on rain-fed streams and rivers. And, of course, hydropower is immobile. You have to build the dams where the rivers are.
Fossil fuels don’t have that limitation. You can take coal, oil, or natural gas out of the ground and move it to a power plant, where you burn it, use the heat to boil water, and let the steam from the boiling water turn a turbine to make electricity.
Because of all these advantages, when demand for electricity in the United States took off after World War II, we met it with fossil fuels. They provided most of the new capacity we built in the second half of the 20th century—some 700 gigawatts, nearly 60 times more than we had installed before the war.
Getting all the world’s electricity from clean sources won’t be easy. Today, fossil fuels account for two-thirds of all electricity generated worldwide. (bp Statistical Review of World Energy 2020)
Over time, electricity has become extraordinarily cheap. One study found that it was at least 200 times more affordable in the year 2000 than in 1900. Today, the United States spends only 2 percent of its GDP on electricity, an amazingly low number when you consider how much we rely on it.
The main reason it’s so cheap is that fossil fuels are cheap. They’re widely available, and we’ve developed better and more efficient ways to extract them and turn them into electricity. Governments also go to considerable effort to keep the prices of fossil fuels low and encourage their production.
In the United States, we’ve been doing this since the earliest days of the Republic: Congress enacted America’s first protective tariff on imported coal in 1789. In the early 1800s, recognizing how important coal was for the railroad industry, states began to exempt it from some taxes and created other incentives for its production. After the corporate income tax was established in 1913, oil and gas producers got the right to deduct certain expenses, including drilling costs. In all, these tax expenditures represented roughly $42 billion (in today’s dollars) in support for coal and natural gas producers from 1950 through 1978, and they’re still in the tax code today. In addition, coal and gas producers benefit from favorable leasing terms on federal lands.
This flyer featuring a coal facility in Connellsville, Pennsylvania, dates from around 1900.
The United States isn’t alone. Most countries take various steps to keep fossil fuels cheap—the International Energy Agency (IEA) estimates that government subsidies for the consumption of fossil fuels amounted to $400 billion in 2018—which helps explain why they’re such a steady part of our electricity supply. The share of global power that comes from burning coal (roughly 40 percent) hasn’t changed in 30 years. Oil and natural gas together have been hovering around 26 percent for three decades. All told, fossil fuels provide two-thirds of the world’s electricity. Solar and wind, meanwhile, account for 7 percent.
As of mid-2019, some 236 gigawatts’ worth of coal plants were being built around the world; coal and natural gas are now the fuels of choice in developing countries, where demand has skyrocketed in the past few decades. Between 2000 and 2018, China tripled the amount of coal power it uses. That’s more capacity than in the United States, Mexico, and Canada combined!
Can we turn this around and get all the electricity we’ll need without any greenhouse gas emissions?
It depends on what you mean by “we.” The United States can get pretty close, with the right policies to expand wind and solar along with a big push for specific innovations. But can the whole world get zero-carbon electricity? That will be much harder.
* * *
—
Let’s start with the Green Premiums for electricity in the United States. It’s actually good news: We can eliminate our emissions with only a modest Green Premium.
In the case of electricity, the premium is the additional cost of getting all our power from non-emitting sources, including wind, solar, nuclear power, and coal- and natural-gas-fired plants equipped with devices that capture the carbon they produce. (Remember that the goal isn’t to use only renewable sources like wind and solar; the goal is to get to zero emissions. That’s why I’m including these other zero-carbon options.)
How much is the premium? Changing America’s entire electricity system to zero-carbon sources would raise average retail rates by between 1.3 and 1.7 cents per kilowatt-hour, roughly 15 percent more than what most people pay now. That adds up to a Green Premium of $18 a month for the average home—pretty affordable for most people, though possibly not for low-income Americans, who already spend a tenth of their income on energy.
(You’re probably familiar with kilowatt-hours if you pay a utility bill, because they’re how we’re charged for electricity in our homes. But in case you’re wondering, a kilowatt-hour is a unit of energy that’s used to measure how much electricity you use in a given time period. If you consume one kilowatt for an hour, you’ve used one kilowatt-hour. The typical U.S. household uses 29 kilowatt-hours a day. On average, across all types of customers and states in the United States, a kilowatt-hour of electricity costs around 10 cents, though in some places it can be more than three times that much.)
It’s great that America’s Green Premium could be so low. Europe is similarly well situated; one study by a European trade association suggested that decarbonizing its power grid by 90 to 95 percent would cause average rates to go up about 20 percent. (This study used a different methodology from the way I figured America’s Green Premium.)
Unfortunately, few other countries are so lucky. The United States has a large supply of renewables, including hydropower in the Pacific Northwest, strong winds in the Midwest, and year-round solar power in the Southwest and California. Other countries might have some sun but no wind, or some wind but little year-round sun, or not much of either. And they might have low credit ratings that make it hard to finance big investments in new power plants.
Africa and Asia are in the toughest position. Over the past few decades, China has accomplished one of the greatest feats in history—lifting hundreds of millions of people out of poverty—and did it in part by building coal-fired electric plants very cheaply. Chinese firms drove down the cost of a coal plant by a remarkable 75 percent. And now they understandably want more customers, so they’re making a big play to attract the next wave of developing countries: India, Indonesia, Vietnam, Pakistan, and nations throughout Africa.
What will those potential new customers do? Will they build coal plants or go clean? Consider their goals and their options. Small-scale solar can be an option for people in poor, rural areas who need to charge their cell phones and run lights at night. But that kind of solution is never going to deliver the massive amounts of cheap, always-available electricity these countries need to jump-start their economies. They’re looking to do what China did: grow their economies by attracting industries like manufacturing and call centers—the types of businesses that demand far more (and far more reliable) power than small-scale renewables can provide today.
If these countries opt for coal plants, as China and every rich country did, it’ll be a disaster for the climate. But right now, that’s their most economical option.
* * *
—
It’s not immediately obvious why there’s such a thing as a Green Premium in the first place. Natural gas plants have to keep buying fuel as long as they’re running; solar farms, wind farms, and dams get their fuel for free. Also, there’s the truism that as you take a technology to broad scale, it gets cheaper. So why does it cost extra to go green?
One problem is that fossil fuels are so cheap. Because their prices don’t factor in the true cost of climate change—the economic damage they inflict by making the planet warmer—it’s harder for clean energy sources to compete with them. And we’ve spent many decades building up a system to extract fossil fuels from the ground, get energy from them, and deliver that energy, all very cheaply.
Another reason is that, as I mentioned earlier, some regions of the world simply don’t have decent renewable resources. To get close to 100 percent, we’d have to move lots of clean energy from where it’s made (sunny places, ideally near the equator, and windy regions) to where it’s needed (cloudy, windless ones). That would require building new transmission lines, a costly and time-consuming task—especially if it involves crossing national borders—and the more power lines we install, the more the price of power goes up. In fact, transmission and distribution are responsible for more than a third of the final cost of electricity.*3 And many countries don’t want to rely on other countries for their electricity supply.
But cheap oil and expensive transmission lines aren’t the biggest drivers of the electricity Green Premium. The main culprits are our demand for reliability, and the curse of intermittency.
The sun and the wind are intermittent sources, meaning that they don’t generate electricity 24 hours a day, 365 days a year. But our need for power is not intermittent; we want it all the time. So if solar and wind represent a big part of our electricity mix and we want to avoid major outages, we’re going to need other options for when the sun isn’t shining and the wind isn’t blowing. Either we need to store excess electricity in batteries (which, I’ll argue in a moment, is prohibitively expensive), or we need to add other energy sources that use fossil fuels, such as natural gas plants that run only when you need them. Either way, the economics won’t work in our favor. As we approach 100 percent clean electricity, intermittency becomes a bigger and more expensive problem.
The clearest example of intermittency is when the sun goes down, cutting off our supply of solar-generated electricity. Suppose we want to solve that problem by taking one kilowatt-hour of excess electricity that’s generated during the day, storing it, and using it that night. (You’d need much more than that, but I’m picking one kilowatt-hour to make the math easy.) How much would that add to our electric bill?
That depends on two factors: how much the battery costs, and how long it’ll last before we have to replace it. For the cost, let’s say we can buy a one-kilowatt-hour battery for $100. (This is a conservative estimate, and I’ll ignore for the moment what happens if we have to take out a loan for this battery.) As for how long our battery will last, let’s assume it can go through 1,000 charge-and-discharge cycles.
So the capital cost of this one-kilowatt-hour battery will be $100 spread out over 1,000 cycles, which works out to 10 cents per kilowatt-hour. That’s on top of the cost of generating the power in the first place, which in the case of solar power is something like 5 cents per kilowatt-hour. In other words, the electricity we store for nighttime use will cost us triple what we’re paying during the day—5 cents to generate and 10 cents to store, for a total of 15 cents.
I know researchers who think they can make a battery that lasts five times longer than the one I’ve described here. They haven’t done it yet, but if they’re right, that would drive the premium down from 10 cents to 2 cents, a much more modest increase. In any case, the nighttime problem is solvable today, if you’re willing to pay a big premium, and with innovation I’m confident we can reduce that premium.
Unfortunately, nighttime intermittency isn’t the hardest problem to deal with. The seasonal variation between summer and winter is an even bigger hurdle. There are various ways to try to deal with it—like adding in power from a nuclear plant or a gas-fired electric plant fitted with a device that captures its emissions—and any realistic scenario will include these options. I’ll get to them later in the chapter, but for the sake of simplicity for now I’ll just use batteries to illustrate the problem of seasonal variation.
Say we want to store a single kilowatt-hour not for a day but for a whole season. We’ll generate it during the summer and use it in the winter to run a space heater. This time, the battery’s life cycle isn’t an issue, because we’re charging it only once a year.
But suppose we have to finance the purchase of the battery. Now we’ve tied up $100 in capital. (Obviously you wouldn’t finance a $100 battery, but you might need financing if you were buying enough to store several gigawatts. And the math is the same.) If we pay 5 percent interest on the capital, and the battery costs $100, that’s an
additional $5 cost to store our single kilowatt-hour. And remember how much we’re paying for solar power during the day: just 5 cents. Who would pay $5 to store a nickel’s worth of electricity?
Seasonal intermittency and the high cost of storage cause yet another problem, especially for big users of solar power—the problem of overgeneration in the summer and undergeneration in the winter.
Because the earth is tilted on its axis, the amount of sunlight that hits any given part of the planet varies across the four seasons, as does the intensity of the sunlight. Just how big the variation is depends on how far you are from the equator. In Ecuador, there’s essentially no change. In the Seattle area, where I live, we get about twice as much sunlight on the longest day of the year as on the shortest day. Parts of Canada and Russia get about 12 times more.*4
To see why this variation matters, let’s do another thought experiment. Imagine there’s a town near Seattle—we’ll call it Suntown—that wants to generate a gigawatt of solar power year-round. How big should Suntown’s solar array be?
One option would be to install enough panels to produce a gigawatt during the summer, when sunlight is plentiful. But the town would be out of luck in the winter, when it’ll get only half as much sunlight. That’s undergeneration. (And the town council is well aware that storage is excessively expensive, so they’ve ruled out batteries.) On the other hand, Suntown could put up all the solar panels it needs for the short, dark days of winter, but then by the time summer arrives, it would be generating way more than necessary. Electricity would be so cheap that the town would be hard-pressed to recoup the expense of installing all those panels.
Suntown could deal with this overgeneration problem by turning off some of its panels during the summer, but then it’d be sinking money into equipment that gets used only for part of the year. That would raise the cost of electricity even more for every home and business in town; in other words, it would add to the town’s Green Premium.