An arc flash can occur when an ener-gized phase conductor is exposed to another phase or ground conductor, with enough voltage differential to overcome the resistance of the air gap between the conductors. An arc flash also can occur if the resistance of the air gap is lowered, such as when contaminated by dust or moisture. When this happens, the air in the gap becomes plasma consisting of ionized air and vaporized metal.

An arc flash releases large amounts of energy that can kill or severely injure anyone exposed (See p. 42). It produces temperatures up to 35,000°F (hotter than the surface of the sun) that can expose a person nearby to heat loads of 120 cal/cm ² or more—sufficient to char skin and set clothing on fire. It can pro-duce a blast wave with pressures up to 2,000 lb/ft ² (enough to throw a person across a room or collapse lungs), eject bits of molten metal and other debris at ballistic speeds, and produce a sound level of 140 dB (equivalent to a gunshot) or more. The vaporized metal quickly turns into a cloud of hot metallic oxides, which can burn themselves into nearby insulators.

According to the NIOSH Office of Mine Safety and Health Research, arc-flash burns are the leading cause of non-fatal electrical injuries, accounting for 35% of lost work days due to electrical injuries in mining between 1990 and 2001 ¹ , averaging 21 lost work days per incident and accounting for more than 12,000 lost work days during the 11-year period of the study.

Most of the injury due to an arc flash is caused by the infrared radiation it pro-duces. This is measured in terms of the energy that reaches the person exposed to it, in calories per square centimeter. Table 1 shows the effect of various levels of incident energy.

An arc flash also generates smoke and toxic fumes from vaporized copper and other materials—fumes that cause health problem by themselves. And while these should eventually be removed by the mine’s ventilation system, if the arc flash takes out a substation there’s a chance it will disable the ventilation system at the same time.

Arc flash is possible on any system with voltages of 480 volts or more. In general, it involves exposing a live con-ductor to either another phase or ground. Such exposure could be caused by cable or equipment damage, a misplaced volt-meter probe, improper installation, dropped tools or even the accumulation of conductive dust on insulators. It is worth noting that in a typical mining electrical distribution system, a neutral-grounding resistor is used to limit ground-fault current, and as such will prevent an arc flash from occurring on a phase-to-ground fault.

While some industries have strict reg-ulations concerning arc-flash hazards, others do not. In manufacturing and gen-eral industry, for example, OSHA requires compliance with NFPA 70E, Standard for Electrical Safety in the Workplace , but these regulations do not apply in mining. Instead, MSHA requires compliance with CFR 30 56 subpart K (Electricity), but does not specifically cover arc flash. Yet, for obvious reasons, mine operators must protect against arc flash, and following NFPA 70E is a good way to ensure safe-ty. NFPA 70E applies to all electrical installations in the mining industry except in underground mines, for self-propelled mobile surface mining machin-ery and trailing cables. Even in those areas not required by law to comply with NFPA 70E, MSHA strongly recommends following its precautions.

Protecting Against Arc Flash
Protection against arc-flash dangers can be approached from two directions: pro-tect the people and minimize the possi-bility and effects of the arc flash itself. Protecting the people involves normal safety precautions and, importantly, the use of personal protective equipment (PPE) required by NFPA 70E, including such things as flame-resistant clothing/ undergarments, flash suits, flash suit hoods, arc-rated gloves and more.

Limiting arc energy—While protecting personnel from arc flash with PPE is both appropriate and necessary, the problem should be approached from the other side as well: minimize the danger of arc flash by eliminating the chance of having one in the first place, or at least mini-mizing the amount of energy released.

Table 1: Effect of arc-flash radiation.

Table 2: Typical opening times of overcurrent protective devices.

The energy released in an arc flash is determined by the square of the current flow and the duration of the arc (I²t) in ampere squared seconds, so limiting either the current or duration of an arc will limit the damage it can do. One way to do this is with fast-acting circuit breakers or current-limiting fuses in the feed to a panel. These overcurrent pro-tective devices react quickly to limit the duration of an arc.

Under short-circuit conditions (a 20x overcurrent condition) a current-limiting fuse can clear a fault in less than half an AC cycle (8.3 ms), as shown in Figure 1.

The gray area shows the energy allowed through by a conventional overcurrent protective device, while the green area shows the energy allowed through by a current-limiting one. An arc does not draw as much current as a bolted fault, but even with an 8x overcurrent a cur-rent-limiting fuse can open between 0.1 and 1 second, and some current-limiting circuit breakers in less than 10 msec. Table 2 compares the clearing times of some available overcurrent protective devices Note that current-limiting fuses operate more quickly than current-limit-ing circuit breakers and are thus more effectively limit delivered I²t.

Arc-flash relays— While current-limit-ing fuses and circuit breakers can help reduce arc-flash energy, they have a sig-nificant drawback; because the earliest moments of an arc flash may draw only a fraction of the current of a short circuit, overcurrent protective devices cannot distinguish them from a typical inrush current, and must wait until the current increases—during which time significant harm can be done to nearby personnel. If the current is low enough, an arc can develop and remain fairly stable for some time—seconds or longer—before it draws enough current to trip the overcurrent protective device.

In contrast, an arc-flash relay (Figure 2) uses light sensors (either point type or distributed fiber optic type) to detect light from an emerging arc flash and send a signal to the relay. The relay will then send a signal to the trip coil on the break-er feeding the panel. The arc-flash relay is designed to operate extremely quickly.

The relay and sensors can be used in transformer enclosures, substations, switchgear and motor control centers. Arc-flash relays are compact and can easily fit in retrofit projects and new switchgear with little or no re-configuration.

Figure 1: Current limitation with a current limiting fuse.

Figure 2: An arc-flash relay.

Figure 3: Damage caused by arc-flash incident.

Typically, the light sensors are set to detect light at 10,000 lux (equivalent to about 10% of the smallest arc). In some relays the lux tripping level can be user-adjusted to prevent nuisance tripping. Certain arc-flash relays can be equipped with current transformers on each phase. If high levels of light are detected (such as from opening a panel in direct sun-light or from a nearby arc welder), but no corresponding increase in current is detected, then the unit does not trip.

Enough sensors should be used to cover the application thoroughly, and they are typically placed near vertical and horizontal bus bars. In addition to point sensors, most relays also accept input from a fiber optic cable that will detect a flash of light anywhere along its length. Such cables range from 26 to 65 ft (8 to 20 m) long, and in some cases they can interconnect to make even longer lengths. Use long lengths with caution, however, because the fiber optic cable may attenuate the light arriving at the detector end, delaying detection.

Figure 3 shows how damage from an arc flash increases with time. Clearly, the faster an arc is detected, the better. Among the arc-flash relays on the mar-ket, detection times vary from less than 1 ms to about 9 ms. These reaction times are a function of the relay’s light sensor input sampling scheme and the design of its trip output circuit. For example, the relay’s microprocessor’s sampling rate of six light sensors might be one sample every 125 microseconds (8 kHz). The relay’s microprocessor may be pro-grammed to count three samples above the threshold value before tripping. The electronic output takes time to turn on, say 200 microseconds for an insulated gate bipolar transistor. Add these times together and the total detection time in this example is <1 ms. Typically the breaker will take an additional 30-35 ms to open after it receives the trip signal.

An arc-flash relay requires that the main breaker have a relay trip coil, so in some cases it may be necessary to replace the main breaker when the relay is installed. In addition, the main breaker should receive regular maintenance— generally by cycling it off and on every six to 12 months—to help keep the mecha-nism from seizing up. Some arc-flash relays have a circuit-breaker fail function; if the breaker does not trip after a time delay of 50 to 150 milliseconds and the arc flash condition is still present, the unit will trip the upstream supply breaker.

Reliability is essential. Select an arc-flash relay that offers a redundant trip fea-ture that will still be able to trip the break-er if the microprocessor does not. Any fail-ure in the primary path will activate a solid-state shunt-trip relay if a sensor input is above threshold. This feature is also useful upon startup after power has been off (as happens after a planned maintenance shut down) since a microprocessor requires start up time before it starts scanning sensors. In contrast, a solid-state device can detect an arc and trip in as few as 2 ms.

To further improve reliability, most arc-flash relays have some degree of internal health monitoring, but designs vary con-siderably. Ideally, the relay will check the health of each component in the path from the light sensor to the trip output contact. The relay should keep event logs that can be accessed by maintenance personnel.

Several arc-flash relays allow multiple relays to be connected. This can be useful if the motor control center does not have a local circuit breaker. In case of an arc flash, the relay that detected the fault can send the trip signal to a relay locat-ed in the switchgear upstream. A network of relays also makes it possible to divide protection in zones. For these applica-tions most relays come with easy-to-use configuration software.

Arc flashes are dangerous events; they can kill or injure people, and damage or destroy equipment. In an underground mine, they can start fires that could even-tually fill the mine with hazardous fumes and trigger secondary explosions. Through the application of arc-flash relays, it’s pos-sible to contain arc flashes inside cabinets, minimize the potential for injury, safeguard equipment, and avoid MSHA fines.