Add SVG figures fse-arc-fault-v1.svg, fse-arc-fault-v2.svg, fse-arc-fault-v3.svg, fse-arc-fault-v4.svg (list-3 visual plan) |
m Remove source references from image captions |
||
| Line 5: | Line 5: | ||
The ''Fire Protection Handbook'' distinguishes three types by the path of the arc. A parallel arc fault results from direct contact between two conductors of opposite polarity, for example in a damaged extension cord. A ground arc fault is an arc between a single conductor and ground, for example in an improperly installed wall receptacle. A series arc fault is an arc across a break in one conductor, for example in a cable pierced by a nail.<ref name="Milatovich-8-148" /> The same distinction appears for cords, where a series fault is a partly or fully severed conductor and a parallel fault runs line to neutral or to ground through damaged insulation.<ref name="Milatovich-8-149">Milatovich (2008), p. 8-149.</ref> | The ''Fire Protection Handbook'' distinguishes three types by the path of the arc. A parallel arc fault results from direct contact between two conductors of opposite polarity, for example in a damaged extension cord. A ground arc fault is an arc between a single conductor and ground, for example in an improperly installed wall receptacle. A series arc fault is an arc across a break in one conductor, for example in a cable pierced by a nail.<ref name="Milatovich-8-148" /> The same distinction appears for cords, where a series fault is a partly or fully severed conductor and a parallel fault runs line to neutral or to ground through damaged insulation.<ref name="Milatovich-8-149">Milatovich (2008), p. 8-149.</ref> | ||
[[File:fse-arc-fault-v1.svg|thumb|upright=1.5|alt=Three panels showing a parallel arc between two conductors, a ground arc to a metal box and a series arc across a break in one conductor|The three types of arc fault: parallel (between conductors of opposite polarity), ground (from a conductor to ground) and series (across a break in one conductor) | [[File:fse-arc-fault-v1.svg|thumb|upright=1.5|alt=Three panels showing a parallel arc between two conductors, a ground arc to a metal box and a series arc across a break in one conductor|The three types of arc fault: parallel (between conductors of opposite polarity), ground (from a conductor to ground) and series (across a break in one conductor).]] | ||
An arc fault can also be a short circuit. An engineering glossary defines a short circuit as an abnormal connection of relatively low impedance between two points of different potential, explicitly including an arc.<ref name="IqbalSalley-F-50">IEEE (1988), as cited in Iqbal and Salley (2004), p. F-50.</ref> Conventional protective devices quickly detect the high current of the short circuits described in the handbook.<ref name="Milatovich-8-148" /> A high-impedance fault, by contrast, has current below the circuit breaker's trip point.<ref name="IqbalSalley-F-36">Generic Letter 86-10, as cited in Iqbal and Salley (2004), p. F-36.</ref> Low-level arcing in damaged cords can likewise continue without being detected by conventional protective devices.<ref name="Milatovich-8-148" /> | An arc fault can also be a short circuit. An engineering glossary defines a short circuit as an abnormal connection of relatively low impedance between two points of different potential, explicitly including an arc.<ref name="IqbalSalley-F-50">IEEE (1988), as cited in Iqbal and Salley (2004), p. F-50.</ref> Conventional protective devices quickly detect the high current of the short circuits described in the handbook.<ref name="Milatovich-8-148" /> A high-impedance fault, by contrast, has current below the circuit breaker's trip point.<ref name="IqbalSalley-F-36">Generic Letter 86-10, as cited in Iqbal and Salley (2004), p. F-36.</ref> Low-level arcing in damaged cords can likewise continue without being detected by conventional protective devices.<ref name="Milatovich-8-148" /> | ||
| Line 11: | Line 11: | ||
== Physics of the arc == | == Physics of the arc == | ||
An arc is a very intense discharge of electrons between two electrodes that carry an electric current. The electrons flow through ionised gas formed when the conductive material vaporises.<ref name="Nowlen-M-1" /> The arc has two main parts. The arc column is a conductive, high-temperature gas, or plasma, and the arc roots are the zones of high electric field and steep temperature gradient where the arc meets the electrodes.<ref name="Nowlen-M-2">Nowlen et al. (2005b), p. M-2.</ref> Temperatures of 6000 K (about 5700 °C) have been observed in open arcs and 25,000 K in high-power circuit breakers at maximum current.<ref name="Nowlen-M-2" /> For comparison, copper, the usual electrode material in such equipment, melts at 1356 K (about 1080 °C).<ref name="Nowlen-M-2" /> | An arc is a very intense discharge of electrons between two electrodes that carry an electric current. The electrons flow through ionised gas formed when the conductive material vaporises.<ref name="Nowlen-M-1" /> The arc has two main parts. The arc column is a conductive, high-temperature gas, or plasma, and the arc roots are the zones of high electric field and steep temperature gradient where the arc meets the electrodes.<ref name="Nowlen-M-2">Nowlen et al. (2005b), p. M-2.</ref> Temperatures of 6000 K (about 5700 °C) have been observed in open arcs and 25,000 K in high-power circuit breakers at maximum current.<ref name="Nowlen-M-2" /> For comparison, copper, the usual electrode material in such equipment, melts at 1356 K (about 1080 °C).<ref name="Nowlen-M-2" /> | ||
[[File:fse-arc-fault-v2.svg|thumb|upright=1.5|alt=Schematic of an electric arc between two copper electrodes with arc roots and a plasma column, with a temperature scale|Structure of an electric arc between two electrodes, with observed arc temperatures compared with the melting point of copper | [[File:fse-arc-fault-v2.svg|thumb|upright=1.5|alt=Schematic of an electric arc between two copper electrodes with arc roots and a plasma column, with a temperature scale|Structure of an electric arc between two electrodes, with observed arc temperatures compared with the melting point of copper.]] | ||
The intensity of an arc and the energy it releases depend largely on the current and on the resistance of the contact, and also on the voltage and the inductance of the circuit load.<ref name="Milatovich-8-127" /> In alternating current the voltage sine wave repeatedly falls to zero. Above 250 V to ground, however, the voltage can restrike the arc, so that burning continues at the location of a ground fault.<ref name="Milatovich-8-140" /> The damage done by an arcing ground fault is directly proportional to the time the arc is allowed to burn.<ref name="Milatovich-8-140" /> | The intensity of an arc and the energy it releases depend largely on the current and on the resistance of the contact, and also on the voltage and the inductance of the circuit load.<ref name="Milatovich-8-127" /> In alternating current the voltage sine wave repeatedly falls to zero. Above 250 V to ground, however, the voltage can restrike the arc, so that burning continues at the location of a ground fault.<ref name="Milatovich-8-140" /> The damage done by an arcing ground fault is directly proportional to the time the arc is allowed to burn.<ref name="Milatovich-8-140" /> | ||
| Line 34: | Line 34: | ||
Cable trays above the cabinet get a separate rule. Unprotected cables in the first overhead cable tray are ignited at the moment of the fault if the tray is within 1.5 m (5 ft) vertically of the top of the cabinet. The time to ignition of that tray is then zero instead of the normal 5 minutes.<ref name="Nowlen-M-13">Nowlen et al. (2005b), p. M-13.</ref> The rule also covers trays above the aisle in front of or behind the cabinet if part of the tray lies within 0.3 m (1 ft) horizontally of its face.<ref name="Nowlen-M-13" /> For a bus duct, the entire length of the duct and any cable immediately adjacent to it are assumed damaged or ignited.<ref name="Nowlen-M-14" /> | Cable trays above the cabinet get a separate rule. Unprotected cables in the first overhead cable tray are ignited at the moment of the fault if the tray is within 1.5 m (5 ft) vertically of the top of the cabinet. The time to ignition of that tray is then zero instead of the normal 5 minutes.<ref name="Nowlen-M-13">Nowlen et al. (2005b), p. M-13.</ref> The rule also covers trays above the aisle in front of or behind the cabinet if part of the tray lies within 0.3 m (1 ft) horizontally of its face.<ref name="Nowlen-M-13" /> For a bus duct, the entire length of the duct and any cable immediately adjacent to it are assumed damaged or ignited.<ref name="Nowlen-M-14" /> | ||
[[File:fse-arc-fault-v3.svg|thumb|upright=1.5|alt=Side view of a switchgear cabinet with a 0.9 m damage zone at front and rear and a cable tray within 1.5 m above it|Zone of influence of the energetic phase of a high-energy arcing fault in a switchgear cabinet, as used in United States nuclear fire risk analysis | [[File:fse-arc-fault-v3.svg|thumb|upright=1.5|alt=Side view of a switchgear cabinet with a 0.9 m damage zone at front and rear and a cable tray within 1.5 m above it|Zone of influence of the energetic phase of a high-energy arcing fault in a switchgear cabinet, as used in United States nuclear fire risk analysis.]] | ||
Risk is quantified per scenario. The core damage frequency (CDF) of HEAF scenario <math>i</math> is the product of five terms. These are the generic HEAF frequency, two weighting factors that share it out over locations and individual cabinets, the probability of no suppression and the conditional core damage probability (CCDP):<ref name="Nowlen-M-16">Nowlen et al. (2005b), p. M-16.</ref><ref name="Nowlen-xlix">Nowlen et al. (2005b), p. xlix.</ref> | Risk is quantified per scenario. The core damage frequency (CDF) of HEAF scenario <math>i</math> is the product of five terms. These are the generic HEAF frequency, two weighting factors that share it out over locations and individual cabinets, the probability of no suppression and the conditional core damage probability (CCDP):<ref name="Nowlen-M-16">Nowlen et al. (2005b), p. M-16.</ref><ref name="Nowlen-xlix">Nowlen et al. (2005b), p. xlix.</ref> | ||
| Line 57: | Line 57: | ||
The plant-wide generic HEAF frequency in the 2005 U.S. nuclear-plant method is 1.5 × 10<sup>−3</sup> per reactor year; the reciprocal of this model mean is about 667 reactor years.<ref name="Nowlen-6-4">Nowlen et al. (2005b), p. 6-4.</ref> It rests on three events in medium-voltage switchgear and one uncertain event, counted as half, in a 480 V load centre.<ref name="Hyslop-4-1">Hyslop et al. (2010), p. 4-1.</ref> The 2010 supplement splits the bin in two for U.S. nuclear plants. Low-voltage panels of 480–1000 V have a plant-wide mean frequency of 4.8 × 10<sup>−4</sup> per year. Medium-voltage panels above 1000 V have a mean of 1.4 × 10<sup>−3</sup> per year.<ref name="Hyslop-4-3">Hyslop et al. (2010), p. 4-3.</ref> | The plant-wide generic HEAF frequency in the 2005 U.S. nuclear-plant method is 1.5 × 10<sup>−3</sup> per reactor year; the reciprocal of this model mean is about 667 reactor years.<ref name="Nowlen-6-4">Nowlen et al. (2005b), p. 6-4.</ref> It rests on three events in medium-voltage switchgear and one uncertain event, counted as half, in a 480 V load centre.<ref name="Hyslop-4-1">Hyslop et al. (2010), p. 4-1.</ref> The 2010 supplement splits the bin in two for U.S. nuclear plants. Low-voltage panels of 480–1000 V have a plant-wide mean frequency of 4.8 × 10<sup>−4</sup> per year. Medium-voltage panels above 1000 V have a mean of 1.4 × 10<sup>−3</sup> per year.<ref name="Hyslop-4-3">Hyslop et al. (2010), p. 4-3.</ref> | ||
[[File:fse-arc-fault-v4.svg|thumb|upright=1.5|alt=Logarithmic chart of high-energy arcing fault frequencies with mean and 5 to 95 percent ranges for low- and medium-voltage panels|Generic frequencies of high-energy arcing faults in United States nuclear power plants: the single 2005 bin and the 2010 split into low- and medium-voltage panels (mean, median and 5–95% range) | [[File:fse-arc-fault-v4.svg|thumb|upright=1.5|alt=Logarithmic chart of high-energy arcing fault frequencies with mean and 5 to 95 percent ranges for low- and medium-voltage panels|Generic frequencies of high-energy arcing faults in United States nuclear power plants: the single 2005 bin and the 2010 split into low- and medium-voltage panels (mean, median and 5–95% range).]] | ||
Voltage level matters for both frequency and consequence. Experts cited in the 2010 supplement reported that recorded HEAF events were dominated by 480 V equipment, partly because there was more of it. They also stated that a majority of arc flash events were initiated by human error.<ref name="Hyslop-4-2">Hyslop et al. (2010), p. 4-2.</ref> An arc fault in low-voltage gear is more likely to stay confined to its panel than one in 4.16 kV switchgear.<ref name="Hyslop-4-3" /> Sustained arcing faults at 208 V are rare.<ref name="Hyslop-4-2" /> | Voltage level matters for both frequency and consequence. Experts cited in the 2010 supplement reported that recorded HEAF events were dominated by 480 V equipment, partly because there was more of it. They also stated that a majority of arc flash events were initiated by human error.<ref name="Hyslop-4-2">Hyslop et al. (2010), p. 4-2.</ref> An arc fault in low-voltage gear is more likely to stay confined to its panel than one in 4.16 kV switchgear.<ref name="Hyslop-4-3" /> Sustained arcing faults at 208 V are rare.<ref name="Hyslop-4-2" /> | ||
Latest revision as of 08:22, 13 November 2026
An arc fault is an unintentional electrical discharge that can ignite combustible materials.[1] On household branch circuits, its current can be low and erratic, allowing an arcing fault in a damaged cord or cable to continue undetected by ordinary fuses and circuit breakers.[1] In United States home structure fires in 1999–2002, arcing was the heat source in 12% of the fires, second only to radiated or conducted heat from operating equipment.[2] Arcing is one of the two basic forms of electrical ignition, the other being overheating.[3] In switchgear and other high-energy equipment, a high-energy arcing fault releases heat, vaporised copper and mechanical force so quickly that neither automatic nor manual suppression can prevent the initial damage.[4][5] The arc-fault circuit interrupter (AFCI) detects low-level arcing on branch circuits; ground-fault protection of equipment is a separate measure for large electrical services under the National Electrical Code (NEC).[1][6]
Definition and types
Some arcing is a normal part of switching. A nuclear-industry glossary defines arcing as the flashing that occurs at electrical terminals when a circuit is opened or closed.[7] Normal arcs occur when a switch is opened or a plug is pulled from a receptacle, and an arc-detecting device should not trip on them.[1] An arc fault is the unwanted case: an unintentional discharge of the kind defined above.[1]
The Fire Protection Handbook distinguishes three types by the path of the arc. A parallel arc fault results from direct contact between two conductors of opposite polarity, for example in a damaged extension cord. A ground arc fault is an arc between a single conductor and ground, for example in an improperly installed wall receptacle. A series arc fault is an arc across a break in one conductor, for example in a cable pierced by a nail.[1] The same distinction appears for cords, where a series fault is a partly or fully severed conductor and a parallel fault runs line to neutral or to ground through damaged insulation.[8]
An arc fault can also be a short circuit. An engineering glossary defines a short circuit as an abnormal connection of relatively low impedance between two points of different potential, explicitly including an arc.[9] Conventional protective devices quickly detect the high current of the short circuits described in the handbook.[1] A high-impedance fault, by contrast, has current below the circuit breaker's trip point.[10] Low-level arcing in damaged cords can likewise continue without being detected by conventional protective devices.[1]
Physics of the arc
An arc is a very intense discharge of electrons between two electrodes that carry an electric current. The electrons flow through ionised gas formed when the conductive material vaporises.[4] The arc has two main parts. The arc column is a conductive, high-temperature gas, or plasma, and the arc roots are the zones of high electric field and steep temperature gradient where the arc meets the electrodes.[11] Temperatures of 6000 K (about 5700 °C) have been observed in open arcs and 25,000 K in high-power circuit breakers at maximum current.[11] For comparison, copper, the usual electrode material in such equipment, melts at 1356 K (about 1080 °C).[11]
The intensity of an arc and the energy it releases depend largely on the current and on the resistance of the contact, and also on the voltage and the inductance of the circuit load.[3] In alternating current the voltage sine wave repeatedly falls to zero. Above 250 V to ground, however, the voltage can restrike the arc, so that burning continues at the location of a ground fault.[6] The damage done by an arcing ground fault is directly proportional to the time the arc is allowed to burn.[6]
Temperature alone does not guarantee ignition. Fire investigators speak of a competent ignition source when its energy and temperature can raise a particular fuel to its ignition temperature within the exposure time. An electric arc may be far above the ignition temperature of a fuel yet last too briefly to deliver enough energy.[12] Duration also matters for flammable atmospheres. A longer spark ignites a dust cloud with less energy, so a spark in electrical equipment, such as from a switch, needs less energy than an electrostatic discharge to ignite the same dust.[13] Near-stoichiometric gas mixtures in air can be ignited by only a few tenths of a millijoule.[14]
How arc faults start fires
An arc can ignite the insulation and covering of the conductor and melt the conductor metal. Burning embers and splatters of hot metal may fall onto combustible material and set it on fire.[3] Unwanted arcing produces high temperatures that can ignite nearby combustibles such as wood, paper and carpets.[1] In cords the process develops in stages: a series or parallel fault leads to carbon tracking within the cord, then to leakage current, arcing and finally combustion.[8]
The handbook lists the common situations in which arcing leads to fire in homes. They include insulation that has deteriorated with age, extension cords caught in a doorjamb or crimped between furniture and a wall, and cords damaged by sunlight or a hot air duct.[1] Frayed appliance cords, impaired internal wiring of appliances, and wire stapled so tightly against a stud that its insulation is cut are further examples. Overheated cords or wires, exposed stranded wiring on old plug caps and receptacles with loose connections complete the list.[1] Lightning can create a delayed hazard of the same kind. A strike conducted into a building can melt conductor insulation inside walls, reducing its dielectric strength, and arcing and a fire within the walls may follow at a later date.[15]
Arcs also ignite furnishings. In United States homes in 1999–2002, arcing was the heat source in 800 of 9,900 upholstered furniture fires a year (8%) and in 1,400 of 17,500 mattress and bedding fires (8%).[16] Concealed spaces are a particular concern. In aircraft, electrical failure and arcing within concealed spaces, often behind the decorative cabin liner, can remain undetected for a long time and grow beyond the reach of hand-held extinguishers.[17]
Industrial settings add flammable atmospheres. In wastewater plants, sustained arcing faults can extensively damage switchgear and motor control centres and ignite flammable gases or vapours, including those released by pyrolysis of the insulation.[18] Starters, generators, distributors and magnetos of stationary engines can arc and ignite flammable vapours. Where an engine pumps or compresses flammable liquids or gases, the handbook calls for shielding, flame arresters, purging or ventilation around arcing components.[19] In United States underground coal mines, by contrast, only about 2.5% of ignitions and explosions result from electric arcs, a share described as surprisingly low given the extensive use of electrical equipment.[20]
High-energy arcing faults
In switchgear, load centres and bus bars or bus ducts of 440 V and above, an arc fault can take an especially violent form, the high-energy arcing fault (HEAF). It releases electrical energy rapidly as heat, vaporised copper and mechanical force, and it is also called an energetic or explosive electrical equipment fault.[4] The fire risk method for United States nuclear power plants divides the event into two phases. The first is the short, rapid release of energy; the second is the ensuing fire in the device and in exposed combustibles nearby, such as overhead cable trays. Any closed cabinet hit by a HEAF is treated as opened to a fully ventilated fire.[4]
Experience from United States nuclear plants between 1979 and 2001 forms the basis of the method. Heavy smoke could delay identification of the fire origin, and in nearly all events the HEAF started in the feed breaker cubicle, where most of the electrical energy resides.[11] In severe events, portable extinguishers could be ineffective whatever the agent (carbon dioxide, halon or dry chemical), and the fires were eventually suppressed with water by the fire brigade. Fire durations ranged from minutes to over an hour, and in about half of the events damage stayed within the electrical device.[21] One 1984 event in 480 V switchgear was counted as half an event in the frequency estimate. In it, high resistance in the main disconnecting contacts of one phase of an air circuit breaker caused an arc that spread to the other phases.[22]
The method defines a zone of influence (ZOI) for the energetic phase. Vulnerable components and movable elements within 0.9 m (3 ft) horizontally of the front or rear panels, and at or below the top of the faulting cabinet, are assumed to suffer physical damage. The method excludes fixed structural elements and large mechanical components from this rule. Exposed cables and other combustibles in the region are assumed ignited.[23]
Cable trays above the cabinet get a separate rule. Unprotected cables in the first overhead cable tray are ignited at the moment of the fault if the tray is within 1.5 m (5 ft) vertically of the top of the cabinet. The time to ignition of that tray is then zero instead of the normal 5 minutes.[24] The rule also covers trays above the aisle in front of or behind the cabinet if part of the tray lies within 0.3 m (1 ft) horizontally of its face.[24] For a bus duct, the entire length of the duct and any cable immediately adjacent to it are assumed damaged or ignited.[23]
Risk is quantified per scenario. The core damage frequency (CDF) of HEAF scenario is the product of five terms. These are the generic HEAF frequency, two weighting factors that share it out over locations and individual cabinets, the probability of no suppression and the conditional core damage probability (CCDP):[25][26]
| Symbol | Description | Unit |
|---|---|---|
| Core damage frequency of scenario | per reactor year | |
| Generic frequency of high-energy arcing faults | per reactor year | |
| Location weighting factor | – | |
| Ignition source weighting factor | – | |
| Probability of no suppression; 1.0 if all targets lie inside the ZOI | – | |
| Conditional core damage probability for the targets damaged in scenario | – |
The plant-wide generic HEAF frequency in the 2005 U.S. nuclear-plant method is 1.5 × 10−3 per reactor year; the reciprocal of this model mean is about 667 reactor years.[27] It rests on three events in medium-voltage switchgear and one uncertain event, counted as half, in a 480 V load centre.[28] The 2010 supplement splits the bin in two for U.S. nuclear plants. Low-voltage panels of 480–1000 V have a plant-wide mean frequency of 4.8 × 10−4 per year. Medium-voltage panels above 1000 V have a mean of 1.4 × 10−3 per year.[29]
Voltage level matters for both frequency and consequence. Experts cited in the 2010 supplement reported that recorded HEAF events were dominated by 480 V equipment, partly because there was more of it. They also stated that a majority of arc flash events were initiated by human error.[30] An arc fault in low-voltage gear is more likely to stay confined to its panel than one in 4.16 kV switchgear.[29] Sustained arcing faults at 208 V are rare.[30]
Worked example. The method itself gives a switchgear cabinet with a stack of three cable trays above it, the lowest 0.9 m above the cabinet.[25] One target cable lies in the first tray and one in the third. The first tray is within 1.5 m of the cabinet top, so its target is taken as ignited at the moment of the HEAF, at time zero, and = 1.0 for that target.[24][25] Fire then spreads upwards: 4 minutes from the first to the second tray and 3 minutes from the second to the third, so the second target is damaged at 0 + 4 + 3 = 7 minutes. For the scenario with both targets, is the probability that manual suppression fails before those 7 minutes have passed, taken from the manual suppression curve for HEAFs in the method.[25]
Detection and protection
Overcurrent protection limits the effects of a fault inside equipment by reducing arcing and the ejection of hot metal.[31] Standard circuit breakers provide protection against overload and short circuit, and ground-fault protection for people where specified. They do not protect against the fire hazard of low-level arcing faults.[1] Where more than 10,000 A is available under short-circuit conditions, current-limiting devices can keep the fault current substantially below what the circuit could otherwise deliver.[32]
In the United States, the NEC first required ground-fault protection of equipment in its 1971 edition, for services of 1000 A or more at 480Y/277 V, because of an unusually high number of burndowns.[6] As summarised in the 20th edition of the Fire Protection Handbook (2008), the maximum setting of the ground-fault sensors is 1200 A. The NEC also limits the duration of faults above 3000 A to no more than 1 second, which reduces the damage done by an arcing ground fault.[6] Up to six disconnects of 800 A each may be used instead, in which case this protection is not required.[6]
On 125 V branch circuits of 15 or 20 A, the AFCI uses electronics to recognise the current and voltage characteristics of an arcing fault and interrupts the circuit when one occurs.[1] It should not trip on normal arcs, such as those of a switch or a plug. A ground-fault circuit interrupter can detect some arcing faults to ground, but not the hazardous across-the-line arcing faults that can cause fires.[1] Leakage current detection and interruption devices in power supply cords sense damage to the cord and disconnect power in less than 25 milliseconds.[8] Bonding of metallic masses in a lightning protection system keeps conductive parts at the same potential and so minimises arcing between them.[15]
Fire investigation and engineering use
In fire investigation, arcs are evidence. The path of the electrical circuits that were energised at the time of the fire is examined for arcing, because the locations of the arcs help to determine how the fire progressed. This technique is called arc fault mapping.[33] Investigators then ask whether a given arc was a competent ignition source for the fuel found nearby, since a short arc may not deliver enough energy.[12]
Arc faults also serve as design fire scenarios. In the handbook example of a gaseous agent system for a telecommunications facility, the chosen design fire is an electrical fault in a power circuit within a data processing equipment enclosure. The sustained arcing is not interrupted by the circuit protection and ignites cable insulation, which spreads to other cables in the cabinet.[34] The design objectives include extinguishing the fire at or below 2 kW and preventing reignition in energised equipment for the hold time, or at least 10 minutes.[35]
Such fires are hard to extinguish while the circuit stays live. With halocarbon clean agents, flames can rekindle because electrical energy continues to reach the fuel, and the agent concentration needed depends on whether the heat comes from an arc, ohmic heating or conduction.[36] Dry chemical leaves no lasting inert atmosphere, so it does not give permanent extinguishment if persistent electrical arcing is present.[37]
Case histories from United States nuclear plants show how arcing starts and spreads fires. At the San Onofre nuclear station in March 1968, operators saw blue arcing above the door window of a 480 V switchgear room, followed by fire in three cable trays above the door. The fire was declared out 39 minutes after the first alarms.[38] At Shearon Harris in October 1989, ground faults in a bus duct, apparently caused by aluminium debris, produced arcing over a length of about 15 m (50 ft). The arcing led to faults in a transformer bushing box, a cracked bushing and an oil fire. Arcing from a second short to ground burned holes in generator-related piping, which contributed to a further fire.[39]
The main limits of current knowledge concern prediction. The HEAF frequencies rest on very few events, three and a half in the 2005 estimate.[28] The 2005 method noted that treating HEAFs could shift the perceived risk importance of switchgear as an ignition source, and that suppression may not prevent the initial damage.[40] For low-energy arc faults, whether a given arc ignites a material depends on its energy and duration as well as its temperature. Investigators therefore test each candidate arc against the competent ignition source criterion.[12]
See also
Notes
- ↑ 1.00 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 1.10 1.11 1.12 1.13 Milatovich (2008), p. 8-148.
- ↑ Ahrens et al. (2008), p. 3-72.
- ↑ 3.0 3.1 3.2 Milatovich (2008), p. 8-127.
- ↑ 4.0 4.1 4.2 4.3 Nowlen et al. (2005b), p. M-1.
- ↑ Nowlen et al. (2005b), p. M-15.
- ↑ 6.0 6.1 6.2 6.3 6.4 6.5 Milatovich (2008), p. 8-140.
- ↑ Iqbal and Salley (2004), p. F-5.
- ↑ 8.0 8.1 8.2 Milatovich (2008), p. 8-149.
- ↑ IEEE (1988), as cited in Iqbal and Salley (2004), p. F-50.
- ↑ Generic Letter 86-10, as cited in Iqbal and Salley (2004), p. F-36.
- ↑ 11.0 11.1 11.2 11.3 Nowlen et al. (2005b), p. M-2.
- ↑ 12.0 12.1 12.2 Custer and Sheppard (2008), p. 3-37.
- ↑ Zalosh (2002), p. 3-405.
- ↑ Beyler (2002b), p. 2-181.
- ↑ 15.0 15.1 Caloggero (2008), p. 11-73.
- ↑ Babrauskas (2008b), pp. 6-106–6-107.
- ↑ Boucher (2008), p. 21-140.
- ↑ Wheeler (2008), p. 9-149.
- ↑ Biggins (2008), pp. 8-106–8-107.
- ↑ L. J. Moore (2008), p. 9-214.
- ↑ Nowlen et al. (2005b), p. M-3.
- ↑ Nowlen et al. (2005b), p. M-4.
- ↑ 23.0 23.1 Nowlen et al. (2005b), p. M-14.
- ↑ 24.0 24.1 24.2 Nowlen et al. (2005b), p. M-13.
- ↑ 25.0 25.1 25.2 25.3 Nowlen et al. (2005b), p. M-16.
- ↑ Nowlen et al. (2005b), p. xlix.
- ↑ Nowlen et al. (2005b), p. 6-4.
- ↑ 28.0 28.1 Hyslop et al. (2010), p. 4-1.
- ↑ 29.0 29.1 Hyslop et al. (2010), p. 4-3.
- ↑ 30.0 30.1 Hyslop et al. (2010), p. 4-2.
- ↑ Milatovich (2008), p. 8-145.
- ↑ Milatovich (2008), p. 8-146.
- ↑ Custer and Sheppard (2008), p. 3-35.
- ↑ Harrington (2008), p. 17-138.
- ↑ Harrington (2008), p. 17-136.
- ↑ Harrington (2008), p. 17-129.
- ↑ Lake (2008), pp. 17-19–17-20.
- ↑ Iqbal and Salley (2004), p. H-6.
- ↑ Iqbal and Salley (2004), p. H-18.
- ↑ Nowlen et al. (2005a), p. 3-12.
References
- Ahrens, M., Frazier, P., & Kelch, G. (2008). Use of fire incident data and statistics. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-61–3-92). National Fire Protection Association.
- Babrauskas, V. (2008b). Upholstered furniture and mattresses. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 6-103–6-128). National Fire Protection Association.
- Beyler, C. L. (2002b). Flammability limits of premixed and diffusion flames. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 2-172–2-187). National Fire Protection Association; Society of Fire Protection Engineers.
- Biggins, J. B. (2008). Stationary combustion engines and fuel cells. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 8-105–8-111). National Fire Protection Association.
- Boucher, B. (2008). Aviation. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 21-135–21-149). National Fire Protection Association.
- Caloggero, J. M. (2008). Lightning protection systems. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 11-69–11-81). National Fire Protection Association.
- Custer, R. L. P., & Sheppard, D. T. (2008). Fire loss investigation. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-31–3-41). National Fire Protection Association.
- Harrington, J. L. (2008). Application of gaseous agents to special hazards fire protection. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 17-123–17-139). National Fire Protection Association.
- Hyslop, J. S., Kratchman, J., Siu, N., Afshar, C., Najafi, B., Joglar, F., Li, Y., Henneke, D., & Warner, M. (2010). Fire probabilistic risk assessment methods enhancements: Supplement 1 to NUREG/CR-6850 and EPRI 1011989 (NUREG/CR-6850, Supplement 1; EPRI 1019259). Electric Power Research Institute; U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research.
- Iqbal, N., & Salley, M. H. (2004). Fire Dynamics Tools (FDTs): Quantitative fire hazard analysis methods for the U.S. Nuclear Regulatory Commission Fire Protection Inspection Program (NUREG-1805). U.S. Nuclear Regulatory Commission, Office of Nuclear Reactor Regulation.
- Lake, J. D. (2008). Chemical extinguishing agents and application systems. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 17-17–17-29). National Fire Protection Association.
- Milatovich, R. M. (2008). Electrical systems and appliances. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 8-127–8-173). National Fire Protection Association.
- Moore, L. J. (2008). Mining and mineral processing. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 9-203–9-223). National Fire Protection Association.
- Nowlen, S. P., Najafi, B., Wyant, F. J., Forester, J., Kazarians, M., Kolaczkowski, A., Joglar, F., Funk, D., Hannaman, G., & Anoba, R. (2005a). EPRI/NRC-RES fire PRA methodology for nuclear power facilities: Vol. 1. Summary and overview (NUREG/CR-6850; EPRI 1011989). Electric Power Research Institute; U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research.
- Nowlen, S. P., Najafi, B., Wyant, F. J., Forester, J., Kazarians, M., Kolaczkowski, A., Joglar, F., Funk, D., Hannaman, G., & Anoba, R. (2005b). EPRI/NRC-RES fire PRA methodology for nuclear power facilities: Vol. 2. Detailed methodology (NUREG/CR-6850; EPRI 1011989). Electric Power Research Institute; U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research.
- Wheeler, J. F. (2008). Protection of wastewater treatment plants. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 9-141–9-155). National Fire Protection Association.
- Zalosh, R. (2002). Explosion protection. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 3-402–3-421). National Fire Protection Association; Society of Fire Protection Engineers.