The most recent terminator, their data suggests, happened on December 13, 2021. In the days that followed, magnetic activity near the sun’s equator dissipated (signaling the death of one set of bands) while the number of sunspots at midlatitude rapidly doubled (signaling the solo reign of the remaining bands). Because this terminator arrived slightly sooner than expected, McIntosh predicts above-average activity for the current solar cycle, peaking at around 190 sunspots.

A clear victor in the modeling wars could emerge later this year. But McIntosh is already thinking ahead to the next thing—tools that can detect where a sunspot will emerge and how likely it is to burst. He yearns for a set of satellites orbiting the sun—a few at the poles and a few around the equator, like the ones used to forecast terrestrial weather. The price tag for such an early-­warning system would be modest, he argues: eight craft at roughly $30 million each. But will anyone fund it? “I think until Cycle 25 goes bananas,” he says, “nobody’s going to give a shit.”

When the next solar storm approaches Earth and the deep-space satellite provides its warning—maybe an hour in advance, or maybe 15 minutes, if the storm is fast-moving—alarms will sound on crewed spacecraft. Astronauts will proceed to cramped modules lined with hydrogen-rich materials like polyethylene, which will prevent their DNA from being shredded by protons in the plasma. They may float inside for hours or days, depending on how long the storm endures.

The plasma will begin to flood Earth’s ionosphere, and the electron bombardment will cause high-frequency radio to go dark. GPS signals, which are transmitted via radio waves, will fade with it. Cell phone reception zones will shrink; your location bubble on Google Maps will expand. As the atmosphere heats up, it will swell, and satellites will drag, veer off course, and risk collision with each other and space debris. Some will fall out of orbit entirely. Most new satellites are equipped to endure some solar radiation, but in a strong enough storm, even the fanciest circuit board can fry. When navigation and communication systems fail, the commercial airline fleet—about 10,000 planes in the sky at any given time—will attempt a simultaneous grounding. Pilots will eyeball themselves into a flight pattern while air traffic controllers use light signals to guide the planes in. Those living near military installations may see government aircraft scrambling overhead; when radar systems jam, nuclear defense protocols activate.

Through a weird and nonintuitive property of electromagnetism, the electricity coursing through the atmosphere will begin to induce currents at Earth’s surface. As those currents race through the crust, they will seek the path of least resistance. In regions with resistive rock (in the US, especially the Pacific Northwest, Great Lakes, and Eastern Seaboard), the most convenient route is upward, through the electrical grid.

The weakest points in the grid are its intermediaries—machines called transformers, which take low-voltage current from a power plant, convert it to a higher voltage for cheap and efficient transport, and convert it back down again so that it can be piped safely to your wall outlets. The largest transformers, numbering around 2,000 in the United States, are firmly anchored into the ground, using Earth’s crust as a sink for excess voltage. But during a geomagnetic storm, that sink becomes a source. Most transformers are only built to handle alternating current, so storm-induced direct current can cause them to overheat, melt, and even ignite. As one might expect, old transformers are at higher risk of failure. The average American transformer is 40 years old, pushed beyond its intended lifespan.

Modeling how the grid would fail during another Carrington-class storm is no easy task. The features of individual transformers—age, configuration, location—are typically considered trade secrets. Metatech, an engineering firm frequently contracted by the US government, offers one of the more dire estimates. It finds that a severe storm, on par with events in 1859 or 1921, could destroy 365 high-voltage transformers across the country—about one-fifth of those in operation. States along the East Coast could see transformer failure rates ranging from 24 percent (Maine) to 97 percent (New Hampshire). Grid failure on this scale would leave at least 130 million people in the dark. But the exact number of fried transformers may matter less than their location. In 2014, The Wall Street Journal reported findings from an unreleased Federal Energy Regulatory Commission report on grid security: If just nine transformers were to blow out in the wrong places, it found, the country could experience coast-to-coast outages for months.

Prolonged national grid failure is new territory for humankind. Documents from an assortment of government agencies and private organizations paint a dismal picture of what that would look like in the United States. Homes and offices will lose heating and cooling; water pressure in showers and faucets will drop. Subway trains will stop mid-voyage; city traffic will creep along unassisted by stoplights. Oil production will grind to a halt, and so will shipping and transportation. The blessing of modern logistics, which allows grocery stores to stock only a few days’ worth of goods, will become a curse. Pantries will thin out within a few days. The biggest killer, though, will be water. Fifteen percent of treatment facilities in the country serve 75 percent of the population—and they rely on energy-intensive pumping systems. These pumps not only distribute clean water but also remove the disease- and chemical-tainted sludge constantly oozing into sewage facilities. Without power, these waste systems could overflow, contaminating remaining surface water.

As the outage goes on, health care facilities will grow overwhelmed. Sterile supplies will run low, and caseloads will soar. When backup batteries and generators fail or run out of power, perishable medications like insulin will spoil. Heavy medical hardware—dialysis machines, imaging devices, ventilators—will cease to function, and hospital wards will resemble field clinics. With death tolls mounting and morgues losing refrigeration, municipalities will face grave decisions about how to safely handle bodies.

This is roughly the point in the worst-case scenario when the meltdowns at nuclear power plants begin. These facilities require many megawatts of electricity to cool their reactor cores and spent fuel rods. Today, most American plants run their backup systems on diesel. Koroush Shirvan, a nuclear safety expert at MIT, warns that many reactors could run into trouble if outages last longer than a few weeks.

ILLUSTRATION: MARK PERNICE
–

If you thumb through enough government reports on geomagnetic storms, you’ll find that one name comes up almost every time: John G. Kappenman. He has published 50 scientific papers, spoken before Congress and NATO, and advised half a dozen federal agencies and commissions. The soft-spoken utility veteran is the man behind the cataclysmic Meta­tech projections, and he is either a visionary or an alarmist, depending on whom you ask. Kappenman spent the first two decades of his career climbing the ladder at Minnesota Power, learning the ins and outs of the utility industry. In 1998, he joined Metatech, where he advised governments and energy companies on space weather and grid resilience.

His end-of-days predictions first gained national traction in 2010, setting off such alarm that the Department of Homeland Security enlisted JASON, an elite scientific advisory group, to pull together a counter-study. “We are not convinced that Kappenman’s worst-case scenario is possible,” the authors concluded in their 2011 report. Notably, however, JASON did not challenge Kappenman’s work on its merits, nor did the group offer a competing model. Rather, its objections were rooted in the fact that Metatech’s models are proprietary, and utility industry secrecy makes it hard to run national grid simulations. Still, the authors echoed Kappenman’s essential conclusion: The US grid is dramatically underprepared for a major storm, and operators should take immediate action to harden their transformers.

The good news is that a technical fix already exists. Mitigating this threat could be as simple as outfitting vulnerable transformers with capacitors, relatively inexpensive devices that block the flow of direct current. During the 1989 storm in Quebec, the grid fell offline and stopped conducting electricity before the current could inflict widespread damage. One close call was enough, though. In the years after, Canada spent more than $1 billion on reliability upgrades, including capacitors for its most vulnerable transformers. “To cover the entirety of the US, you’re probably in the ballpark of a few billion dollars,” Kappenman says. “If you spread that cost out, it would equal a postage stamp per year per customer.” A 2020 study by the Foundation for Resilient Societies arrived at a similar figure for comprehensive grid hardening: about $500 million a year for 10 years.

To date, however, American utility companies haven’t widely deployed current-blocking devices to the live grid. “They’ve only done things, like moving to higher and higher operating voltages”—for cheaper transmission—“that greatly magnify their vulnerability to these storms,” Kappenman tells me.

Tom Berger, former director of the US government’s Space Weather Prediction Center, also expressed doubts about grid operators. “When I talk to them, they tell me they understand space weather, and they’re ready,” he says. But Berger’s confidence waned after the February 2021 collapse of the Texas power grid, which killed hundreds of people, left millions of homes and businesses without heat, and caused about $200 billion in damage. That crisis was brought on by nothing more exotic than a big cold snap. “We heard the same thing,” Berger says. “‘We understand winter; it’s no problem.’”

I reached out to 12 of the country’s largest utility companies, requesting information on specific steps taken to mitigate damage from a major geomagnetic event. American Electric Power, the country’s largest transmission network, was the only company to share concrete measures, which it says include regularly upgrading hardware, redirecting current during a storm, and quickly replacing equipment after an event. Two other companies, Consolidated Edison and Exelon, claim to have outfitted their systems with geomagnetic monitoring sensors and be instructing their operators in unspecified “procedures.” Florida Power & Light declined to meaningfully comment, citing security risks. The other eight did not respond to multiple requests for comment.

At this point, curious minds may wonder whether utility companies are even required to plan for geomagnetic storms. The answer is complicated, in a uniquely American way. In 2005, when George W. Bush, a former oil executive, occupied the Oval Office, Congress passed the Energy Policy Act, which included a grab bag of giveaways to the oil and gas industry. It rescinded much of the Federal Energy Regulatory Commission’s authority to regulate the utility industry. Reliability standards are now developed and enforced by the North American Electric Reliability Corporation—a trade association that represents the interests of those same companies.

Some find the NERC reliability standards laughable. (Two interviewees audibly laughed when asked about them.) Kappenman objected to the first set of standards, proposed in 2015, on the grounds that they were too lenient—they didn’t require utilities to prepare for a storm on par with 1859 or 1921. Berger took issue too, but for a different reason: The standards made no mention of storm duration. The ground-based effects of the Carrington Event lasted four or five consecutive days; a transformer built to withstand 10 seconds of current is very different from one ready for 120 hours.

Under pressure from the federal government, NERC enacted stricter standards in 2019. In a lengthy written statement, Rachel Sherrard, a spokeswoman for the group, emphasized that American utilities are now expected to deal with an event twice as strong as the 1989 Quebec storm. (Comparison with an old storm like Carrington, she noted, “is challenging because high-fidelity historical measurement data is not available.”) Though the new standards require utilities to fix vulnerabilities in their systems, the companies themselves determine the right approach—and the timeline.

If the utilities remain unmotivated, humanity’s ability to withstand a major geomagnetic storm will depend largely on our ability to replace damaged transformers. A 2020 investigation by the US Department of Commerce found that the nation imported more than 80 percent of its large transformers and their components. Under normal supply and demand conditions, lead times for these structures can reach two years. “People outside the industry don’t understand how difficult these things are to manufacture,” Kappenman says. Insiders know not to buy a transformer unless the factory that made it is at least 10 years old. “It takes that long to work out the kinks,” he says. In a time of solar crisis, foreign governments—even geopolitical allies—may throttle exports of vital electrical equipment, Kappenman notes. Some spare-part programs have cropped up over the past decade that allow participants to pool resources in various disaster scenarios. The size and location of these spares, however, are unknown to federal authorities—because the industry won’t tell them.

One day regulators may manage to map the electrical grid, even stormproof it (provided a big one doesn’t wipe it out first). Engineers may launch a satellite array that gives us days to batten down the hatches. Governments may figure out a way to stand up emergency transformers in a pinch. And there the sun will be—the inconceivable, inextinguishable furnace at the center of our solar system that destroys as indiscriminately as it creates. Life on this little mote depends entirely on the mercy of a cosmic nuclear power with an itchy trigger finger. No human triumph will ever change that. (But we should still buy the capacitors. Soon, please.)

–

Cover: Styling by Jeanne Yang and Chloe Takayanagi. Styling assistance by Ella Harrington. Grooming by April Bautista using Oribe at Dew Beauty Agency. Prop styling by Chloe Kirk.

Let us know what you think about this article. Submit a letter to the editor at mail@wired.com.