What Does an EMP Do? Cars, Electronics, the Grid and People

What does an EMP do? It induces sudden, damaging currents in electronics and long conductors such as power lines. A high-altitude nuclear burst produces the strongest and widest version. What it does to your body is nothing. What it might do to the machines around you is the real debate.

The popular version says one bomb in the sky erases all technology for a continent. The official record is more careful and more interesting. Tests show many devices survive. Experts disagree on how badly the grid would be hit and for how long. This article separates what was measured from what was modeled and what was guessed.

What an EMP is, and what causes one

CISA lists three ways to produce an EMP incident: a high-altitude nuclear detonation (HEMP), specialized conventional munitions, and non-nuclear directed-energy devices. The effects can be local, regional or continental depending on the source.

Nuclear HEMP is the one that worries planners. A warhead exploded tens to hundreds of miles up releases gamma rays. These knock electrons out of air molecules, and the electrons curve in Earth’s magnetic field, which radiates a pulse downward. Because the burst is high, its line of sight covers a huge circle on the ground. That is why one device could reach a whole region.

CISA says an adversary with basic nuclear and missile capability could cause regional disruption, and that a continental-scale grid collapse is considered unlikely. A highly capable adversary in an escalating conflict could do much more. CISA classes these as “hard problems”: low probability, high consequence.

Solar storms are not EMP. They are a related hazard with different physics, covered in the Carrington event article.

E1, E2 and E3: three pulses, three kinds of damage

A nuclear HEMP is not one pulse. It is three, arriving in sequence. The standard description comes from the Congressional EMP Commission and summaries such as the Congressional Research Service report.

The timing figures are approximate and come from secondary summaries. The key point: E1 is the electronics threat, E3 is the grid threat, and they are different problems with different fixes.

Starfish Prime: the one real test

The US exploded a 1.4-megaton bomb 250 miles (400 km) above the Pacific on July 9, 1962. About 900 miles (1,450 km) away in Hawaii, around 300 streetlights went dark, many burglar alarms went off, and a telephone microwave link was damaged, cutting calls from Kauai to the other islands.

That is the only large real-world data point, and it is weak evidence for modern systems. The 1962 Hawaii grid and electronics were very different from today. Also, the satellites that failed in the following months (Telstar 1 among them) were mainly hit by radiation trapped in Earth’s magnetic field, a separate effect from the EMP itself. Nobody has tested a modern city.

What an EMP does to a car

The best data is from the EMP Commission’s 2008 report, which put vehicles in a pulse simulator. These numbers are quoted from the report’s text by secondary sources, so check the original for exact wording:

• Cars (37, model years 1986 to 2002). Eight showed no anomaly. 25 had nuisance faults, such as blinking dashboard lights, that needed no driver action. Three stalled when the field passed about 30 kV/m.
• Trucks (18, 1991 to 2003). 13 responded while running; most responses were minor and temporary. Five showed no anomaly up to about 50 kV/m. One truck needed a tow.
• Vehicles switched off were not affected afterward.

The Commission did warn that if 10% or more of vehicles on a busy expressway stalled at once, crashes and injuries would follow. And these were older vehicles. Modern cars have far more electronics, and we found no comparable large public test of them. Critics such as Jeffrey Lewis also point out the report predicted stalled cars would coast to a stop, rather than showing what stalled cars did.

The fair summary: the “every car is dead” claim is not supported. A small share would likely stall, more would glitch, and the road system would still be in chaos.

How to read “kV/m”

Field strength is measured in volts per meter (here, kilovolts per meter). It describes how hard the pulse pushes on a conductor. The Commission tested up to about 50 kV/m, which is also the figure often quoted as the US military hardening standard. A weapon designed for maximum EMP might exceed that, as some advocates claim, but that figure is disputed and we found no unclassified measurement to settle it. So a car that survived 30 kV/m tells you something about a plain HEMP, and less about a purpose-built device.

Why the vehicle result matters for the bigger picture

If most vehicles keep running, then transport after an EMP is limited by fuel and roads rather than dead cars. Fuel pumps, refineries and pipelines depend on the grid and on control systems. So the car result is good news for the car and still bad news for the fuel supply chain. That is the same chain of dependencies described in our blackout timeline.

What an EMP does to small electronics

The usual claim is that phones, laptops and radios die. The record is thinner than the claim. E1 couples best into long wires and large loops, which act as antennas. A phone or a radio that is not plugged into anything has less metal to pick up energy. That is why the worst-hit items in a real event would likely be things connected to long cables: the grid, telephone lines, antennas and cable-fed equipment.

We did not find a public test of modern phones in a real HEMP field, and the Commission’s data cover specific hardware. So the honest range is: unplugged small electronics are less likely to be damaged than plugged-in ones, but the amount is unmeasured.

What an EMP does to the power grid

This is where the experts split. The EMP Commission and some trade summaries said E1 and E3 together could cause blackouts lasting months or years. A later industry study took a narrower view.

The EPRI study from 2017 analyzed E3 effects from one burst at each of 11 notional US locations. It found hundreds or thousands of transformers could see significant induced current, but only about 3 to 14 units, depending on the location, faced thermal damage risk. A multi-state regional blackout was possible; widespread immediate transformer failure was not expected. EPRI said its findings “do not support the belief that HEMPs can cause nationwide blackouts lasting for years.” It covered only the transmission system, not generation, distribution or end-use equipment.

A study that looked only at the size of induced currents had concluded that widespread transformer damage was likely. Both sides have worked on the same problem and reached different answers. So we give a range: from a regional outage measured in days or weeks to a long recovery if a large number of large transformers fail. Large transformers are slow to replace: a GAO review found utilities ranked manufacturing lead times as their top supply problem, and Wood Mackenzie data put waits for large units at roughly 80 to 210 weeks. That is the link to what happens when the grid goes down.

What an EMP does to a person

Nothing, directly. The pulse induces currents in conductors and circuits. We found no government source that reports direct injury to people from an EMP, and CISA’s framing is entirely about infrastructure. The harm comes from what depends on that infrastructure: hospitals, water, fuel, traffic signals, communications.

The blast, heat and fallout of a nuclear weapon are a different problem. A high-altitude burst is far from the ground, but a ground-level or low-altitude one is not. Those effects are in our nuclear attack guide.

Can you protect electronics from an EMP?

Yes, partly. Real protection means enclosing equipment in continuous conductive shielding with every opening dealt with. Professional standards for this (IEEE Std 299 for shielding measurement and MIL-STD-188-125 for facility hardening) exist because it is hard to do well. For the home version, see our step-by-step guide on how to protect electronics from EMP.

At home, the practical points are these:

• Seams and gaps are the enemy. In one home test, a metal trash can gave only about 10 dB of shielding, and much of the energy was thought to leak around the lid seam. That is a single hobbyist report, not a standard result.
• Corroded or painted contact surfaces create gaps.
• A foil-wrapped box is not verified protection unless it is tested.
• Testing with a phone or FM radio only checks the frequencies those devices use. An EMP is broadband.
• Surge protectors do not reliably stop E1 because it rises faster than most protectors respond.

If you want a cheap backup, keep a spare of the few items you cannot replace, stored unplugged and away from long cables. A battery radio, a small flashlight and a spare phone charger cover most needs. Our survival gear guide lists what is worth owning.

Note: For most people, preparing for a multi-day blackout protects against far more likely events than an EMP. The same list works for both.

How to tell hype from evidence

Most EMP claims online fall into a few patterns. Here is a quick filter.

1. A number with no source. Claims that most of the population would die within a year circulate widely. We did not find such a figure in the sources we reviewed, and we do not use it.
2. Testing claims that skip the method. “I put my phone in a box and it lost signal” shows the box blocks cell frequencies, not that it stops a broadband pulse.
3. Treating one study as settled. The EMP Commission and EPRI disagree on grid damage. Each has its limits: the Commission’s work has been criticized by some analysts, and EPRI covered only the transmission system.
4. Mixing sources. Solar storms, nuclear EMP and local directed-energy devices are three different things.
5. Selling a fix for a scenario. If a product page cites only its own test, treat it as marketing.

The useful question is not “will an EMP happen?” but “what do I need if the grid is down for a week or more?”

Non-nuclear EMP and “EMP weapons”

CISA lists conventional munitions and directed-energy devices as separate EMP sources. Their reach is short: they affect a building or a facility, not a region. These are real, but they are local security issues, not a grid-wide scenario.

Lab verdict

An EMP is real physics with a real, uncertain effect on the grid and a smaller, less certain effect on most gadgets. The measured data say most cars would keep running, people would not be harmed by the pulse, and the grid is the main exposure. Estimates of how long a blackout would last range from regional and short to long and severe, depending on which study you trust.

Plan for the consequence, not the cause. A week without power is the same problem whatever triggered it.

Frequently asked questions

Does an EMP destroy all electronics?

No. Damage depends on the equipment, the field strength and whether it is connected to long wires. The EMP Commission’s vehicle tests showed most vehicles with no effect or minor glitches. Long cables, grid equipment and control systems are more vulnerable than small unplugged devices.

Would my car still run after an EMP?

Probably yes in most cases. In the Commission’s tests, 3 of 37 cars stalled, and 25 had minor glitches that did not need driver action. Switched-off vehicles were unaffected. Newer cars were not part of that test, so the result is not certain for modern models.

Can an EMP kill you?

Not directly. The pulse does not injure the body. The danger is indirect: loss of electricity, hospital function, water and transport. Officials such as CISA frame the threat as damage to critical infrastructure.

Does a Faraday cage really work?

A proper, sealed, tested enclosure works. A random metal box often does not, because gaps and seams let the pulse in. A single home test found about 10 dB of shielding in a taped-up trash can. Testing with a phone is a weak check.

How likely is an EMP attack?

No agency publishes a probability. CISA calls high-altitude EMP a low-probability, high-consequence risk. Executive Order 13865 (March 2019) directs DHS to coordinate preparedness for EMP and geomagnetic disturbances. Treat any single number you see online with suspicion.

Sources

• CISA: Electromagnetic Pulse
• EMP Commission (Commission to Assess the Threat to the US from EMP Attack)
• Congressional Research Service: HEMP and High Power Microwave Devices, RL32544
• Wikipedia (as a pointer): Starfish Prime
• SurvivalBlog: excerpts of EMP Commission vehicle test results (secondary, check the report)
• ArmsControlWonk: critique of EMP vehicle claims
• GAO-23-106180: Electricity grid transformer supply chain
• T&D World: EPRI research on E3 and transformers