Protection from lightning damage is a major consideration for most installations. A typical lightning strike event involves currents of around 10,000 to 50,000 amperes with a 10% to 90% risetime of between 1 and 10 microseconds. Thus, the peak voltages involved are usually determined more by the inductive reactance than by the D.C. resistance of the lightning current's path to ground. To protect equipment and living creatures, we need to keep them from becoming a conductive path for lightning currents by keeping everything in their immediate environment at nearly the same electrical potential.
A holistic engineering approach for achieving this in a specific installation might consist of the following steps:
- Identify the zones to protect.
- Specify the elements to be protected within these zones and quantify the desired level of protection for each one.
- Identify all paths through which dangerous amounts of lightning energy could enter these zones. These will usually include:
- Direct strikes: Capture them and divert them to ground.
- Surges on A.C. power lines: Install surge protection devices (SPD's) to divert them into the grounding system, using multiple stages as needed to achieve the specified level of protection;
- Surges on data, communications and control cables: Same thing.
- Currents flowing in the earth: Install an equipotential conductive ring below grade around the building, bonded to the grounding system.
- Inductive coupling of lightning currents: This is mostly a problem for cables and equipment mounted on the tower when the tower takes a direct strike. Bond all conductors to the tower at both ends and every 20m along the tower, either directly (coax shields and other grounds) or using appropriate SPD's.
- Capacitive coupling: This is rarely a separate problem for us.
- Design an adequate grounding system to absorb the strike and surge current;
- Bond all grounds together;
- Install the surge protection devices;
- Monitor and maintain the protection system.
The grounding system forms the core of any plan for dealing with lightning, similar to the role of the city's drainage system in a plan for dealing with tsunamis. We need to do all we can to minimize both the magnitude and duration of the voltage difference due to the sudden flood of electrons by providing a low-inductance path to earth for the strike current. We then devise a way to keep whatever remains from pushing damaging currents through anything we want to protect. Low inductance is the key: a couple of ground rods connected with a few feet of heavy wire might be an adequate ground for a 60 Hz electrical distribution system but not be for a 20kA lightning strike with much of its energy concentrated at tens or hundreds of kiloherz. A lower inductance system of multiple buried radials with multiple ground rods distributed along them may be needed.
Towers are grounded structures somewhat higher than the surrounding terrain, and thus may attract lightning strikes. Towers more than about 100m high can have a triggering effect, but we mostly don't need to worry about this, since our masts and towers are rarely more than about 10m-20m high. According to the formulas given in  (Vol. 1, 3.5.1), the attractive radius for a conductive structure 20m high is about 124m, corresponding to an attractive area of about .048 km^2. At our latitude of about 17 degrees north here in Oaxaca, we can expect about 13% of observed lightning flash events to be cloud-to-ground strikes. Assuming a flash density of 20 flashes per km^2 per year, we can expect direct strikes to our 20m tower to happen on the average about once every 8 years. The tower itself should always be bonded to a low-inductance grounding system. To prevent arcing, any cabling should be kept near tower potential by bonding it to the tower at both ends and every 20m in between, either directly for shields and grounds or through SPD's for other conductors.
Tower cables that need to enter a building should do so via a metal entry panel located as close to grade level as possible. This entry panel should be bonded to the building's equipotential ring at exactly one place with the lowest-inductance connection possible, usually via a short, wide copper strap. This panel will ideally serve as the single-point ground reference for the building. Surge-suppression devices for all cables entering the building through this panel should be installed there.
In most installations, surges on the incoming A.C. power lines are far more common than direct lightning strikes to a tower or building, and therefore a more urgent problem. SPD's should be installed so as to provide defense-in-depth to anything we want to protect. It's common practice to install an SPD at the service entry which absorbs most of the energy from the largest surges but only clamps the line to a fairly high voltage level, between 600-1200 volts, for example. Then on sub-branch circuits and at the point-of-use we can put other SPD's that will clamp the line to a lower voltage level, once the surge has already been decapitated by the upstream SPD's and smoothed out a bit by the upstream wiring inductance.
Point-of-use SPD's marketed mainly to consumers, such as multicontact surge-suppressing outlet strips, rarely come with useful specifications. The number of Joules absorbed doesn't tell us very much (this depends a lot on the applied waveform), nor does the turn-on time of the devices used to clamp the voltage (delays due to wiring inductance will usually be much greater anyway). It's far more meaningful to us to know what the maximum voltage clamping level is under what current waveform conditions. There are various standard test waveforms in common use. The most common of these, at least in the U.S., is probably the 1.2/50-8/20us combination waveform described in ANSI/IEEE Std C62 (see below). There are others, and there is still much debate as to how well various standard test waveforms model actual lightning events. In the 1990's, NEMA (U.S. National Electrical Manufacturers' Association) issued a standard, LS-1, defining a common format for specifying SPD's. It was withdrawn (by the invisible hand?) in 2009 because of controversies associated with its use in marketing. Various performance and safety standards have been developed by the likes of the IEEE (Institute of Electrical and Electronics Engineers), IEC (International Electrotechnical Commission) and U.L. (Underwriters' Laboratories, U.S.). Unfortunately, these performance and safety-conscious folks keep their standards locked up behind a paywall. Of the IEEE publications relevant to SPD's, the most important are:
- IEEE Std C62.41.1, "Guide on the Surge Environment in Low-Voltage (1000 V and less) AC Power Circuits"
- IEEE Std C62.41.2, "Recommended Practice on characterization of Surges in Low-Voltage (1000 V and less) AC Power Circuits"
- IEEE Std C62.45, "Recommended Practice on Surge Testing for Equipment Connected to Low-Voltage (1000 V and Less) AC Power Circuits"
Other publications and standards:
- U.L. 1449 is a U.S. safety and performance standard for SPD's, for both manufacturers and users.
- IEC SC37A is a European standard.
 U.S. Dept. of 'Defense', "MILITARY HANDBOOK: GROUNDING, BONDING AND SHIELDING FOR ELECTRONIC EQUIPMENTS AND FACILITIES", MIL-HDBK-419A, Dec. 29, 1987
Very comprehensive. Volume 1 is theory, volume 2 is applications. "If only they had used their powers for Good, instead of Evil..."
 Bogdan “Bogey” Klobassa, "A Lightning Protection System for Wireless Telecom Sites", Above Ground Level, Dec. 2010
How to protect cell towers and equipment buildings.
 Le & Nouanesengsy, "Lightning Protection for Cellular Tower Mounted Electronics", Andrew Corporation
A good summary. Andrew is a major manufacturer of coaxial cable used for cellular towers.
 Block, Ron KB2UYT, "Lightning Protection for the Amateur Radio Station", ARRL.org
Lightning Protection for the Amateur Radio Station -- Part 1 Lightning Protection for the Amateur Radio Station -- Part 2 Lightning Protection for the Amateur Radio Station -- Part 3
A systems-oriented approach to protecting equipment in a ham radio station, written by the brother of Polyphaser's founder.
 NEMA (National Electrical Manufacturers' Association, U.S.) Surge Protection Institute webpages:
Nice listing of relevant standards.
 Erico's telecom protection products:
A great resource. Many well-written guides/appnotes in the "Documents" section for using SPD's, with an emphasis on protecting telecom and industrial sites. Erico developed the exothermic welding process used nowadays by nearly everybody for bonding conductors in grounding systems for lightning protection. They manufacture a wide variety of stuff for lightning protection systems.
 Polyphaser / Transtector / Smiths Power
Lots of White Papers mainly oriented towards RF surge protection. Polyphaser is a leading manufacturer of RF surge suppression devices.