Before you spend money on shielded storage, there is a simple test with a phone that tells you whether a given bag or container is doing anything at all.
The phone test →A Faraday cage is a genuinely useful piece of physics wrapped in an unusual amount of nonsense. The physics is 190 years old, well understood, and easy to demonstrate on a kitchen table. The nonsense is everything sold on top of it. This page separates the two: what a Faraday cage actually does, what it measurably cannot do, and which of the things people put in one belong there.
The Physics, in One Paragraph
A conductor is full of electrons that are free to move. Bring an external electric field near a closed conductive shell and those electrons redistribute across the surface almost instantly, arranging themselves so their own field cancels the applied one everywhere inside. The interior ends up with essentially zero field. Michael Faraday demonstrated this in 1836 by lining a room with foil and sitting inside it while an electrostatic generator threw sparks at the outside. It works because of the geometry and the conductivity, not because of any special material.
For time-varying fields, meaning radio waves and pulses rather than static charge, two more effects join in. Some of the wave reflects off the conductive surface because of the impedance mismatch, and whatever gets in is absorbed as it penetrates the metal. Reflection does most of the work at the frequencies that matter here.
Skin Depth: Why Foil Is Thick Enough
The depth into a conductor at which a field falls to about 37% of its surface value is called the skin depth, and it shrinks as frequency rises. In aluminium at 100 MHz it works out to roughly 8 micrometres. Ordinary kitchen foil is around 16 micrometres thick, so a single layer is already two skin depths of absorption, on top of a substantial reflection loss at the surface.
That is the counterintuitive result worth holding onto: for the fast, high-frequency part of an electromagnetic pulse, thickness is almost never the limiting factor. A steel trash can is not meaningfully better than heavy foil because it is thicker. Anything that fails, fails somewhere else.
Apertures Are What Actually Leak
Every hole, seam, gap and unbonded joint is an antenna. How much energy gets through an opening depends on the size of that opening relative to the wavelength. A gap that is a small fraction of a wavelength barely radiates; one approaching half a wavelength passes energy almost freely.
At 1 GHz the wavelength is about 30 centimetres, so half a wavelength is 15 cm. The working rule engineers use is to keep the largest dimension of any opening under roughly a twentieth of the shortest wavelength you care about, which at 1 GHz means about 1.5 cm. This has consequences:
- Fine mesh is fine. Window screen with 1-2 mm openings is far below the threshold well past 1 GHz. A microwave oven door works on exactly this principle.
- A lid that merely rests on a can is not. A 3 mm gap running around a 30 cm rim is a slot antenna, and slots leak according to their long dimension, not their narrow one. This is the single most common failure in a homemade cage.
- Paint and anodising break the bond. Two metal surfaces need actual conductive contact around the whole seam. Powder coating is an insulator.
- Any wire passing through defeats the whole thing. A cable entering the enclosure carries energy straight past the shield, which is why real hardened facilities use filtered penetrations rather than holes.
The Three Components of an EMP
Most arguments about Faraday cages happen because people are talking about different threats. A high-altitude nuclear burst produces three distinct effects with wildly different timescales, and a cage only addresses one of them.
| Component | Timescale | What it threatens |
|---|---|---|
| E1 | Rise around 2.5 nanoseconds, over in under a microsecond | Small electronics directly. This is what a cage stops. |
| E2 | Microseconds to milliseconds, resembles lightning | Similar to a nearby strike. Ordinary surge protection applies. |
| E3 | Tens to hundreds of seconds, very low frequency | Long conductors: transmission lines, pipelines, transformers. |
E1 is fast and broadband, with meaningful energy up to around a gigahertz, which is precisely the regime where a decently sealed conductive box works well. E3 is a different animal. It is slow, it couples into conductors kilometres long, and it is the mechanism behind geomagnetic damage to grid transformers. A severe solar storm produces something like E3 and nothing like E1.
This matters practically: a Faraday cage does nothing whatsoever about E3, and therefore nothing about the scenario people actually fear most, which is the grid staying down for months. Your radio survives. The utility’s transformers are not in your trash can.
What a Faraday Cage Does Not Do
Worth stating plainly, because the marketing rarely does:
- It does not protect anything plugged in, connected to an antenna, or wired to anything outside the enclosure.
- It does not protect your house wiring, your well pump, your furnace board or your solar inverter, all of which are attached to long conductors by design.
- It does not keep the grid up, and it does not shorten an outage by a single hour.
- It does not need to be grounded to work. A grounded cage and an ungrounded one both shield their interior; grounding matters for safety and for the charge on the outside, not for the field inside.
- It does not stop the actual consequences of a long blackout, which are the ordinary ones covered in what breaks first when the grid goes down.
How Far Consumer Gear Is From the Standard
The military standard for HEMP-hardened facilities, MIL-STD-188-125-1, requires 80 dB of shielding effectiveness across a band from about 10 kHz to 1 GHz. Eighty decibels means the field inside is one ten-thousandth of the field outside, and achieving it takes welded seams, RF gaskets, filtered power penetrations and waveguide-below-cutoff vents.
Consumer shielded bags typically claim figures in the 40 to 80 dB range depending on frequency, and the honest ones publish a curve rather than a single number. That is not a scandal; 40 dB is a factor of a hundred and is probably ample for a spare radio. It is simply a different class of thing from a hardened shelter, and worth knowing before you pay a premium for a number on a package. Verifying which claims hold up is the whole point of faraday bags: which work and how to test.
What Is Genuinely Worth Putting In One
Think of it as insurance on the items that are hard to replace and useless to have only one of. A spare AM/FM/shortwave receiver, a handheld radio if you have a licence, a small solar charge controller, a spare inverter board, a cheap phone with offline maps loaded, a USB battery bank, and a printed copy of the things you would otherwise look up. Nothing here needs to be expensive.
What is not worth it is duplicating your entire household. The vehicle you drive, the furnace in the basement and the well pump 200 feet down are not going in a box, and the plan for those is the ordinary one: know what they draw, know what you would do without them, and start with a home energy audit.
My Honest Take
Build one, spend under thirty dollars on it, put a spare radio and a charge controller inside, and then stop thinking about it. The physics is real and the cost is trivial, so there is no reason not to. But if you have a shielded container and no stored water, no carbon monoxide detector and no plan for heat, you have optimised the wrong end of the problem by a very wide margin. The boring failures are the ones that actually happen.
Frequently Asked Questions
Does a Faraday cage need to be grounded?
No. The field inside a closed conductive shell is cancelled whether or not the shell is grounded. Grounding gives absorbed energy somewhere to go and is good practice on a large or permanently installed enclosure, but an ungrounded metal can with a sealed lid shields its contents perfectly well.
Does a microwave oven work as a Faraday cage?
Poorly. It is engineered to contain one frequency, around 2.45 GHz, and its door seal, power cord penetration and vents are not designed for broadband shielding. It will attenuate something. It is not a substitute for a properly sealed container.
Will a Faraday cage protect my car?
Not in any practical sense. A car body is a decent partial shield in its own right, thanks to a mostly metal shell, and testing on modern vehicles has generally found them far more robust to fast pulses than the folklore suggests. Building an enclosure for one is not a realistic project.
Do items inside need to be insulated from the walls?
Yes, keep a layer of cardboard, foam or bubble wrap between your gear and the metal. Direct contact lets current on the shell couple into whatever is touching it. It costs nothing and removes an easy failure mode.