One Atmospheric Nuclear Explosion Could Take Out the Power Grid

New study identifies vulnerabilities to EMP attack

3 min read
Satellite overhead view of the United States at night. The land mass is grey with bright yellow lights across the country.
Matthias Kulka/Getty Images

In July, Chinese researchers urged their government to increase the country's readiness for defending against a high-altitude electromagnetic pulse (EMP) attack. Just over a year ago, a group of American researchers released a report warning that China possessed the capability to conduct an EMP attack against the United States. Military and non-proliferation experts are worried about the growing temptation by nuclear-armed countries to engage in a first-strike EMP attack using nuclear weapons that, while avoiding direct casualties, could prove devastating to electric grids and electronic devices from smartwatches to supercomputers.

The enormous potential of an electromagnetic pulse released by the high-altitude detonation of a nuclear weapon has been recognized for some time. In 1962, the U.S. conducted an atmospheric test of a 1.45 megaton thermonuclear weapon, code-named Starfish Prime, 250 miles above Johnston Island in the Pacific Ocean. Over 1,000 miles away, the blast knocked out electricity supply in parts of Hawaii and disrupted telephone service for a period of time. In addition, radiation from the test damaged several satellites in low-Earth orbit, taking them out of service. Decades later, the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack determined as early as 2008 that the U.S. would face catastrophic consequences from an EMP attack given its growing dependence on electronics of all forms and complete reliance upon the electrical grid.

And yet, until now, government and industry risk assessments about EMP attacks and their effects on the power grid have been based on oversimplified models of the solid Earth that assume zero variation in depth or composition. But, as it turns out, the actual effects on the power grid of an electromagnetic pulse in outer space are strongly determined by the three-dimensional distribution of rocks beneath our feet.

Landmark research, the product of collaboration between the U.S. Geological Survey (USGS) and the University of Colorado, illuminates the role of the solid Earth in determining the magnitude of any EMP hazard. In an interview, USGS geophysicist Jeffrey J. Love, the lead author on the new report, explains that a high-altitude EMP produces three sequential waveforms with different impacts on electrical systems: E1, E2, and E3.

The high-frequency E1 pulse disrupts consumer electronics and tends to get the most attention; E2 behaves more-or-less like lightning, and, fortunately, our electrical systems are (largely) hardened against its effects. The E3 waveform is the lowest-amplitude part of the EMP signal, but because it is the longest-lasting part, covering periods from about 0.1 seconds to several hundred seconds, it has the potential to cause catastrophic damage to the electrical grid through its interactions with the solid Earth. (See the illustration, below.)

Diagram showing a geomagnetic field generated by space current caused by a nuclear explosion.A nuclear explosion in the upper atmosphere or in space (off screen) would push a current of ions and electrons through the atmosphere (B), producing a magnetic field (A) that would in turn induce currents and electric fields in the earth beneath it. Regions of high conductivity (D) would more readily carry that current, while regions of low conductivity (F), would not. Instead, the larger electric field in these regions (E) could help coax current out of the ground and through the wires of high-capacity power lines (C)—yielding a potentially grid-crippling power surge induced by the explosion's EMP. John MacNeill

Love notes that three factors conspire to form a geoelectric hazard for power grids: "The level of EMP magnetic disturbance, the conductivity of the surrounding Earth, and the specific parameters of the grid itself." The new study used existing survey data—originally collected for geological exploration—in a small region of the eastern-midcontinental U.S., covering portions of Arkansas, Missouri, Illinois, Mississippi, Kentucky, and Tennessee. Researchers from the USGS then secured permission from property owners to place sensors on the ground for measuring the natural variation in the Earth's magnetic field over several weeks. They also used voltmeters to measure the time-varying electric field at the same locations. These two measurement types provide estimates of surface impedance—an electromagnetic property that depends on rock conductivity.

Love and his co-authors used these survey data to assess the impact of an E3 EMP waveform generated by a high-altitude detonation of a nuclear weapon with a yield of several hundred kilotons—the benchmark description of an EMP event in the literature. They also included USGS research on natural, magnetic-storm disturbances across the continental U.S.—in regions such as the electrically resistive metamorphic and igneous rocks of northern Minnesota or east of the Appalachian Mountains as well as electrically conductive sedimentary rocks of Michigan or Illinois.

All in all, they found that EMP hazards had not been accurately mapped in complex geological settings. They call for USGS to analyze surface impedance across regions like the eastern mid-continent. They add that USGS should pay special attention to the eastern U.S, where magnetic-storm hazards are already known to be high. Of course this is also where many of the nation's largest cities are located.

"Better coordination across scientific disciplines is necessary," Love says. "By bringing together weapons engineers, space scientists, and geophysicists, we can achieve a holistic approach to EMP threat assessment and, with that, prioritize improvements."

This article appears in the October 2021 print issue as "One Atmospheric Nuclear Explosion Could Cripple the Entire Grid."

The Conversation (6)
Floyd Goldmann
Floyd Goldmann09 Sep, 2021

There are EMP issues far more insidious than those caused by a nuclear EMP. A nuclear bomb is not necessary to do significant damage to the grid infrastructure. None of the control electronics have any significant hardening to any type of EMP/IEMI. The controls are typically only tested to radiated susceptibility of 20 V/m. Processor core voltages are dropping to very low voltages (eg. 3.3V, 1.2V). and the control wires are exposed. One does not have to over current transformers or burn wires to take down the grid. If the control electronics are dead then the grid is dead. A 50kV/m high power pulse generator could be easily constructed. While the field strength decreases by the square of the radius E=(kQ)/(r^2) focusing using a high gain horn antenna could allow a bad actor to to be far away from the target system. It could be theoretically possible to cause a single backup turbine to synch out of phase and then come on line if the controls were disturbed by IEME. We would hope that the branch OCP devices would trip but if they were affected by the disturbance via conducted IEMI (induced by the radiated) they may not. This could cause down stream OCP devices to open. Sudden load losses could cause other turbines to over speed, draw excessive current and then trip more OCP devices causing a eventual cascade failure. A few simultaneous targeted attacks could disable all of ERCOT . Look what happened with the ice/snow storm which was a known hazard as it happened once before the 2021 incident. Michigan's grid infrastructure is so deteriorated that simple issues have caused cascaded failures. The American Society of Civil Engineers gave Michigan's infrastructure a D+ grade in its 2018 report, and the energy grid was given a C- grade.

J Peter Fitzsimmons
J Peter Fitzsimmons09 Sep, 2021

So, why would a nation risk exploding a nuclear weapon to knock out the grid when a retaliatory strike would turn them into a smoking poisonous ruin? hasn't our defense policy for the last 60 years been mutual assured destruction (MAD)? Or are you claiming our land based nuclear strike capability would be disabled by EMP. I thought we fixed that 50 years ago.

2 Replies
James Mercier
James Mercier09 Sep, 2021

Natasha, can I get permission to reprint your article in the Austin Gem & Mineral Society (AGMS) newsletter? AGMS is the rock collectors and lapidary hobby group, another 501(c)3 organization. This is just the type of information that would be considered interesting to the mostly non-technical members. James J. Mercier, P.E.

1 Reply