10.4.2.2 Where Protection is Applied.
AS noted above, HEMP protection must be designed to accommodate hardness verification procedures. The
most easily verified protection requires the least number of tests and the least number of assumptions to
establish the integrity of the protection.
For example, suppose the facility is a node with a 100-pair cable
linking it to other parts of a network. Because of unbalanced and nonlinear terminations, there may be 2002
two-wire stresses and susceptibilities to evaluate at the cable penetration of the facility barrier. Inside the
facility, the 200-wire cable may branch into 1000 or more wires and equipment terminals. Thus if the HEMP
stress is allowed to be dominant (larger than known peacetime stresses) inside the barrier, 10002 transient
stresses and susceptibilities must be evaluated (or assumed unimportant). In addition, in the latter case, all the
internal interactions between the 1000 wires and other internal circuits must be assessed (or assumed
unimportant). Typically, both the number of features to be evaluated and the number of assumptions necessary
increase with the depth into the system at which hardness verification is attempted.
Therefore, for facilities whose protection against HEMP has high value (i.e., where confidence in the protection
is important), the protection is placed at the system-level barrier, and the protection at this level is sufficient
that the HEMP-induced stress is not dominant inside this barrier.
10.4.2.3 Terminal Protection Devices.
Problems from HEMP are expected to arise from the antennas and
connecting cables, long interconnecting leads and cables between equipments, and the ac power lines.
Antennas, connecting cables, and the front-end of the associated communications equipments in particular will
be subjected to very large voltages and currents.
The protective technique or device must protect the
equipment without adversely affecting its performance, and must be capable of withstanding the effects of both
EMP-induced transients and other transients in the system. The latter two considerations may severely limit
applications of many of the protective devices at rf unless they are modified or used in conjunction with other
Spark Gaps and Gas Tubes.
Spark gaps are one of the oldest forms of surge arrester. A spark gap is
a pair of electrodes, insulated by air or other gas, spaced so that the gap will break down when the voltage
exceeds a specified level.
The insulating gas pressure varies from a fraction of an atmosphere to several
atmospheres, and the electrode spacing varies from a few millimeters in carbon blocks to several inches in large
lightning arresters used for power equipment. Firing voltages range from about 1 kV for some carbon blocks to
hundreds of kV for large lightning arresters.
Large spark gaps can handle large charge transfers (many
coulombs). In the nonconducting state, spark gaps behave as open circuits or small capacitances. The spark-gap
Thus, for the large rates-of-rise
firing voltage increases with the rate-of-rise of the applied voltage.
encountered in EMP-induced voltages, the firing voltage may be several times as large as the rated static firing
voltage. When spark gaps are used on energized lines, some provision must be made to assure that the discharge
will be extinguished. Frequently, a metal-oxide varistor (MOV) is used in series with the spark gap to ensure arc
extinction after the surge.
Gas tubes are spark gaps with a low-pressure gas so that lower firing voltages can be achieved. Firing voltages
below 100 V are available for commercial gas tubes. The gas tubes are generally more limited in their peak
current and charge transfer capability than the spark gaps.
Gas tubes are used primarily for secondary
protection of wire pairs entering a facility from a long external shielded cable, or for exposed intrafacility