Grounding for Electromagnetic Pulse

by Jerry Emanuelson


There are two cases where grounding for EMP is useful.  One is in making the ground connection for large surge suppressors (such as for whole-house or standby generator protection) if the surge suppressors are fast enough for EMP.  The other is for the outer shield on faraday cages that have conductors penetrating into the faraday cage.  Small faraday cages with no conductors penetrating the shield do NOT need to be grounded.  Large shields that do have conductors penetrating the shield need careful filtering and transient protection (surge suppression) to insure that electromagnetic fields are not re-radiated into the interior of the shielded volume.

On the other EMP pages on this site, I have warned against grounding for EMP protection unless you have a considerable amount of experience in grounding in high-intensity, high-frequency electromagnetic fields.  On this page, I will try to explain the dangers of the wrong type of EMP grounding, and explain what types of grounding might be useful.

First, consider the following thought experiment:   Imagine that you have a long copper rod vertically imbedded into the ground, but with 10 feet of the rod protruding vertically above the ground.  The copper rod is electrically well-grounded.  Now, suppose the copper rod is hit with a nuclear electromagnetic pulse that has an electric field strength of 20,000 volts per meter in the vertically polarized direction.  This means that the 10-foot (3.048 meter) section of the copper rod that is protruding from the ground will have more than 60,000 volts induced on the top of the rod.  It is true that the voltage will induce a current that will travel through the rod toward the ground at nearly the speed of light, but that will take about 10 nanoseconds.  The pulse is not an instantaneous, infinitely narrow spike, so as the current begins traveling down the rod, additional voltage is being induced at the top of the rod.  Therefore, depending upon the exact characteristics of the pulse, this well-grounded rod has a very high voltage pulse induced at its upper portions that will last for several nanoseconds.  The voltage pulse at the top of this grounded rod doesn't contain much total power, but it is far beyond the voltage that would be necessary to punch through a semiconductor junction, and completely destroy any transistor or integrated circuit.

If the vertical rod were replaced by a well-grounded ordinary copper wire, with a horizontal run, the situation could be worse since the horizontal component of the pulse is often of a much higher electric field strength.  If the 10-foot (a little more than 3 meters long) wire were exposed to a 50,000 volt per meter pulse, a voltage of 150,000 volts could be induced on the end of the wire for several nanoseconds.  If this wire were used to "ground" the shield around sensitive electronics equipment, it is easy to see that what you are actually doing is inducing a very high voltage spike on the entire shield, which then could be transmitted to the equipment inside (although there are ways to avoid this if you are experienced in shielding).  So this method of "grounding" an equipment shield can be a good way of assuring the destruction of the equipment that you are trying to protect.  (If the "ground" wire is curved or curled up in some way, the pulse induced on the equipment shield will be even worse!)

This is why grounding an EMP shield can be very difficult.  It is also why nested shields (or nested faraday cages) are so important (unless you are operating in a very well shielded and maintained room).  Even if you apply what you will learn later on this page about grounding, that does not reduce the importance of having layers of protection.  Alternating layers of a good conductor and a good insulator are important.  If you are trying to protect small equipment like radios, wrap the radio inside an insulator (such as a plastic Zip-loc® freezer bag), then wrap that insulated bag in aluminum foil, then wrap that in another layer of plastic, then use another layer of a good conductor (aluminum foil or aluminum or copper screen), and continue to add more layers if you need additional protection.

Layers of protection are also important for transient protection.  Layers (or chains) of transient protection, where ever possible, are always better than a single transient protector.  (The terms "transient protector" and "surge protector" can be used interchangeably.)

The same principle applies for electromagnetic shielding if you are building a shielded room.  For a room, you may be able to use an air space between the inner and outer shields as an electrical insulator between the inner and outer shields (depending upon your construction technique), but it is always much better to have at least two layers of shielding, separated from each other by an insulating material (or by air).  I realize that this is not always cost effective, and sometimes it may not even be possible.

Common practice in high-frequency, high-power electromagnetic environments is to use a wide copper strap for grounding.  The width of the copper strap should be the widest that is both feasible and affordable for the particular situation.  The copper strap is used because high frequencies, including those in a fast rise-time pulse, tend to travel mainly on the outer surface of a conductor.  This makes the DC resistance of the grounding material less relevant.  Especially now that the price of copper has become so high, it is important not to waste such a valuable commodity by having most of the copper in the interior of a grounding conductor.

Another alternative for the grounding conductor to maximize the surface area to total volume is to use copper tubing.  Flexible copper tubing (such as that used for plumbing) may be used, but the diameter of the tubing should be the largest that is feasible and affordable.  For many individuals, copper tubing is much easier to obtain than a wide copper strap.

For copper tubing, you can take the grounding effectiveness to an even higher level by having a copper shield around the copper tubing that is used for the ground (with the outer shield electrically insulated from the main inner grounding tube except at the actual ground point).  This is rarely done in practice because of the cost of the additional material for the shield.  In this procedure, the tubing that you are using as a ground is shielded in order to prevent it from becoming an "accidental antenna" by having a large current pulse induced in the tubing that is used as the ground wire.   If you are using this method, the central ground tube and the outer shield must both be connected together underground, but the outer shield must not be connected to anything at the top end.  The central tubing (using this method) protrudes a short distance from the outer shield, as necessary, to make the connection to the grounding point on the equipment or on the outer shield of shield of the faraday cage.  As I mentioned, this is an extraordinary method and is seldom used.

Remember that a faraday cage without any wire penetrations does not need to be grounded, and any attempt to ground such an uncomplicated faraday cage is simply asking for problems.

It is imperative that the above-ground portion of all ground conductors be as short and direct as possible.

I see many reports of people trying to "ground" something by running a long ordinary-sized wire halfway across a house or other building.  This may work as a static drain or to protect against electric shock; but for an EMP ground it is worse than worthless, it is an antenna.

If you are grounding an equipment shield, it is best to ground only the outer shield.  Professionally designed and maintained EMP shielded rooms often have a ground connection to all of the shields, but doing this properly is completely out of the reach of individuals unless you happen to be both wealthy (with access to some very expensive and specialized test equipment) and a knowledgeable engineer.

Great care should be used when making outdoor connections, and especially connections that will go underground.  Although soldering can be used for such connections, it is definitely not the best option.  A brazed or Cadwelded® connection for the underground connection or for connections to the ground rod are better options.

If you are having a whole-house generator installed, the electrician may insist upon a vertical rod pounded into the ground, even if the ground is very dry, rocky, non-conductive soil.   If you are faced with this problem, I would suggest using posthole diggers to dig down as far as possible, then fill up the hole with bentonite or a proprietary grounding compound.   (Bentonite is easier for most people to obtain.)   By pounding the rod through the center of the bentonite and on into the soil below, a much better ground will be obtained due to the bentonite surrounding the upper part of the copper grounding rod.

All grounding rods should normally be kept as close as possible to the building electrical ground.  There should be one grounding point that is regarded as the ultimate ground.  Although multiple ground rods can improve the grounding system, if multiple ground rods are used they should be connected together with copper tubing that is buried in a shallow trench running back to the ultimate grounding point.

Another method for improving the effectiveness of a ground is to dig one or more trenches running outward from the ultimate grounding point.  The trenches should be as long and as deep as practical.  Run copper tubing through the trenches and connect all of the copper tubing to a copper grounding rod at (or as close as possible to) your ultimate grounding point.  Fill the bottom part of each trench (covering the copper tubing) with bentonite or a proprietary grounding material unless your soil is always very moist and electrically conductive.

Of course, whenever you are digging holes, pounding rods or digging trenches, be sure to use standard safety methods and practices to avoid hitting other underground lines.  Until you know where those underground lines are, you shouldn't be digging at all.