For decades after they belched their last coal fumes, the cooling towers of the Richborough Power Station loomed over the low landscape of Kent, in the southeast of England, guiding fishing boats into the English Channel. That all ended on the morning of March 11, 2012, when a controlled demolition reduced the 30-story-high towers to rubble, clearing the site for a cleaner future.

The building that has risen in their place at the renamed Richborough Energy Park is one very tall story and painted in three layers of green and gray, as if determined to fade into the sky. It buzzes with 1,000 megawatts of electricity—enough to power a million homes—but emits no exhaust and has no need for towers or chimneys. Its energy is not produced here, only passing through, carried by a new piece of infrastructure known as Nemo Link. It is an “interconnector,” joining the power grids of the United Kingdom and Belgium, like a Chunnel for electrons. Its most salient feature is buried out of sight: a pair of copper cables, each as thick as a 2-liter soda bottle, stretching 87 miles across the English Channel from Richborough to the Belgian port of Zeebrugge, near the city of Bruges.

Nemo supplies both countries with renewable and reliable energy—two words that don’t often go together. We expect our electricity supply to be rock-steady, but wind and solar power are as intermittent as the breeze and the sun. With this high-voltage connection, when the U.K. is short of power, Belgium can send it; and vice-versa.

The goal is for Nemo’s operation to become part of the fabric of the two countries’ grids, allowing electricity to flow either way, on demand, in 60-minute time blocks. There are economic reasons for that: Both utilities, the U.K.‘s National Grid and Belgium’s Elia, can save (and make) money by selling their excess. There are technical reasons for it: The two nations can quickly share power in the case of a spike in demand—or a blackout. But above all there are environmental reasons: Renewable power must be consumed in equal measure to the amount generated because we don’t yet have grid-scale batteries. Time is fixed. But thanks to connectors like Nemo, space is now flexible.

Nemo is one piece of a broad strategy: the first completed segment in a boom of building interconnectors that is, ironically, drawing British infrastructure closer to Europe at just the moment when the island kingdom is pulling away politically. By 2023, the U.K. grid will join with that of France, the Netherlands, Belgium, Norway, and Denmark, across two existing and four newly created links. They all serve European decarbonization goals. In 2016, 17 percent of power consumed in the European Union came from renewable sources. By 2030, the E.U. wants that to be 32 percent—on the way to being climate neutral by 2050, meaning whatever emissions are left will be offset by removing an equal amount of carbon from the atmosphere. Recent studies have shown that achieving this will require interconnection capacity between E.U. nations to expand between 400 and 900 percent. The building boom has just begun.

The U.S. will someday have to solve the same equation. Our power plants still pump out lots of carbon, though wind and solar generation is growing quickly, rising to 10 percent this past year. But they serve localized, regional needs. That could change in the next decade. In particular, an upswing in offshore wind projects along the Eastern Seaboard will likely force us to rethink our networks and bring interconnectors like Nemo to American shores.

On the other side is where the magic happens,” Scott Williams, dressed in a neon-green safety jacket on a gray winter day, says as we peer through a thick pane of safety glass. We’re looking across a hall as big as an ice rink, filled with towers of equipment as tall as four-story buildings. Each mini high-rise is encircled by loops of tubing, making the whole thing look like a jungle gym for baby giants.

The assemblage is a converter station, where England’s alternating current (AC), which can run along wires in both directions, is turned into direct current (DC), which flows one way and is more efficient for long-distance transfer; at a similar station in Belgium, it’s switched back into AC. There’s an important bonus to Nemo’s setup: Using DC for the transfer means the two countries’ AC networks don’t need to be synchronized with each other to work. Over the past year, Williams has managed the construction and commissioning of the whole operation for Siemens, the German industrial conglomerate that designed both facilities and manufactured the equipment inside.

The magic, as Williams calls it, is a routine process carried out on a massive scale. A strawberry-size iPhone charger converts a wall socket’s AC power into 5 watts of DC power. Nemo converts 1,000 megawatts from AC to DC and back. The feat requires 2,304 of Siemens’ modules, each roughly the size of a golf bag. These are mounted as “six packs” and stacked up into the jungle-gym-like towers. The loops of tubing around them pump chilled demineralized water to dissipate the heat generated in the conversion between AC and DC. A pair of techs operates the whole thing from a nearby control room—but once the kinks are worked out, the system can also be run remotely.

Recently, the milestones have been coming weekly: the initial electric charge of the converter, the first power flowing through it, and then a 1,000-megawatt test, which Williams marked with a snapshot of the controller’s screen, showing it off like a baby picture. Turning on this machine—in technical terms, a high-voltage direct-current link—was a big moment. But “once you’ve done it three days in a row, they weren’t even paying attention anymore,” Williams says of the control-room staff. The next week, he would officially pass the project from Siemens to a joint venture co-owned by National Grid and Elia.

It is only when Williams leads us into an adjacent hall that the full force of the energy soon to flow through the building becomes apparent. A streetlamp-size arm, called a disconnector, lowers and raises as a physical break between the converter modules and the undersea cable. It is a safety feature, allowing techs to be 100 percent sure that the giant power cord is disconnected from the other end before they touch it. (A control box with a physical lock and key makes that even clearer.)

“You wouldn’t want someone in Belgium to flip on the switch when you’re working next to it,”

Williams says. That’s no swipe at the Belgians—only an inherent risk of joining two electric systems.

In a conference room overlooking London’s Trafalgar Square, National Grid’s Nigel Williams explains that his old job with the British utility involved carefully tuning the available domestic supply of electricity to match demand—directing the production of different types of energy like the conductor of an orchestra. But in his new role managing the construction of North Sea Link, an interconnector to Norway that will be three times the size of Nemo, he is thinking about how to prepare the energy infrastructure of the country over decades rather than minutes.

“I think everyone has begun to realize that running your own country like a fiefdom without interconnecting is a crazy thing to do,”

Williams says. Belying the slick corporate meeting room, he wears a checked shirt, jeans, and sturdy boots. His phone vibrates on the table with calls from Norway, from which he’s just returned the evening before, on yet another trip to shepherd his utility’s next big interconnector into existence.

“It makes sense that if you’re good at producing something, it needs to be exported,”

he says. In particular, Britain is getting good at producing electricity from wind. In just the past decade, its turbines in the North Sea have gone from turning out almost nothing to a capacity of 8 gigawatts. On one particularly blustery day this past November, the U.K. generated enough wind energy to power one-third of the country—an event that garnered triumphant headlines. With the opening of Nemo this year, any surplus can go to Belgium. But what about the calm days?

Substation electrical equipment at a converter station, where AC power becomes high-voltage direct current.Courtesy Siemens AG

To keep the lights on, generation plants feeding the grid need to be nimble and responsive.

“Supply and demand have to meet second by second,”

Williams says. Coal plants are flexible, with electricity-generating turbines that are easy to speed up or slow down depending on demand, but they are dirty. Nuclear power is clean—or at least low-carbon—but “you can’t swing it around,” Williams says, meaning turn its generation up or down as demand warrants.

Interconnectors offer a new menu of options—such as importing extra nuclear power from Belgium, hydropower from Norway, or wind power from Denmark. They move electricity from where it’s generated to where it’s needed. They allow a utility to order up extra power, or offload excess when renewable generation is high—whether seasonally, such as when the Norwegian snow is melting, or daily, when a North Sea storm blows through. Or, as Williams puts it,

“Interconnectors allow electricity to be traded and treated in exactly the same way as your can of Coke.”

But these links alone will not solve the challenge of moving the grid to renewable energy. They are a stopgap, buying time for each country involved to build more of its own turbines and develop new technologies. Iben Fürst Frimann-Dahl, an analyst at Rystad Energy in Norway, tracks the global renewables market. “Currently we are dependent on having backups, especially now that countries are phasing out coal and nuclear,” she says. Today that means power plants designed to quickly increase output to satisfy peak demand—but by virtue of often relying on coal or gas, these “peaking plants” work against low-carbon goals. In the future it will mean storage. “When we develop these batteries, most of the issues with renewables will disappear,” says Frimann-Dahl. Until then, interconnectors are the bridge not only between grids, but also to a completely renewable future.

In 1961, President John F. Kennedy commissioned a grand infrastructure project known as the Pacific Intertie, the jewel in the crown of an even more massive public works. Over the previous three decades, the federal government had built a series of hydroelectric plants in the Columbia River Basin of the Pacific Northwest. All together, they produce 29 gigawatts of electricity, or 44 percent of the hydropower currently generated in the U.S. To get that energy from its rural source to its urban customers, the Pacific Intertie provides a high-voltage DC link from the edge of the Columbia River on the border between Oregon and Washington state to the fringes of the sprawling L.A. metropolitan region, more than 800 miles away.

The Intertie was completed in 1970, but nothing of its scale has been constructed in the States since. “We are not, as a nation, in a nation-building mode,” says Jesse Jenkins, a postdoctoral fellow at the Harvard Kennedy School, who researches the engineering and economics of the grid. “We don’t do long-distance, large-scale, national-scale infrastructure projects.”

In part to answer the question of what we should build if we did, Jenkins and his colleagues recently published their analysis of 40 studies from around the world of “deep de-carbonization,” defined as an 80 to 100 percent reduction in emissions for electricity generation. What they found is that the strategies require some combination of energy generation and storage, reductions in consumption, and increases in transmission. Strikingly, there is no coherent plan in the U.S. for any of those three things.

But there has been progress. The technology for wind- and solar-electricity generation is mature, with costs competitive with fossil-fuel generation, but storage is still the biggest sticking point. Relieving the grid of having to match supply and demand instantaneously would go a long way toward surviving the inevitable troughs when the wind doesn’t blow or the sun doesn’t shine, but the technology does not yet exist. Despite the hype around Elon Musk’s house-size batteries, everyone in the industry recognizes how limited the current options are for storage.

Europe’s interim solution to the storage problem is interconnectors, which effectively let one country harness another’s winds. The need for an improved grid in the U.S. is no less important. “The larger and more robust our transmission system, the more we can spread the variability of the weather patterns across the country,” Jenkins says. According to one study by the National Renewable Energy Laboratory, reaching 80 percent renewable electricity in the United States (with 50 percent from wind and solar) would require a 56 to 105 percent increase in long-distance transmission capacity. We need a new grid as big as the old one. The unanswered political question is whether such a multibillion-dollar project will ultimately save more money than the catastrophically costly consequences of a carbon-filled atmosphere.

Still, even without a revamped grid, renewable energy in the United States has been employed at a heady pace, with wind and solar increasing from less than 5 percent to more than 10 percent of production over the past decade. By 2020, they are expected to make up 13 percent of total U.S. electricity generation, surpassing hydropower. Eventually we’ll have no choice but to come up with a way to accommodate its intermittency.

In 2016, five turbines rose out of the blue Atlantic waters off Block Island, their 600-foot masts cradled by four-legged steel bases, painted bright yellow and blinking with navigation lights. They were the first of what will eventually be many. This past year, the state of Massachusetts awarded a major contract for offshore wind generation, allowing for the construction of 84 turbines producing 800 megawatts of electricity.

The project, known as Vineyard Wind, is the starting gun for a massive build-out up and down the Eastern Seaboard, with some of the projects managed (and in some cases funded) by experienced European wind operators. According to one analysis by Citi, the next decade could see the construction of 2,000 turbines off the East Coast, with 22 gigawatts of offshore capacity.

This is a startling number. Denmark installed the world’s first offshore wind farm in 1991, but then it took almost 30 years for total global offshore capacity to reach 18.8 gigawatts, in 2017.

Injecting 22 gigawatts of power—intermittent power—into the broader U.S. grid will be a challenge. Edward Krapels, CEO of Anbaric Development Partners, which designs and builds large-scale electrical transmission systems like Nemo, is angling to construct an offshore grid for the East Coast’s burgeoning wind fields. At the scale of development being discussed, it will have to be robust—with far greater capacity than National Grid’s new interconnectors. But the principle is the same: The power has to move to the people. Soon, the U.S. will need infrastructure like Nemo, feeding all that wind-driven energy into high-voltage transmission lines, efficiently and flexibly sending it to shore. The technology is ready. Now all we need is the will.

(First published in Popular Science, issue ‘Make It Last’ in Summer 2019)