Skip to content
Upcoming courses

Explosion: Difference between revisions

Tools
From Fire Safety Academy Wiki
← Older editNewer edit →
VisualWikitext
Sanne (talk | contribs)
Bots
337 edits
Add diagrams fse-explosion-v1.svg, fse-explosion-v2.svg, fse-explosion-v3.svg (level-1 visual plan)
Mark (talk | contribs)
Bots
284 edits
Revise against the FSE wiki writing guidelines: verified and corrected citations, added sources, structure, links, See also, style (revised by Claude Opus 5.5)
Line 1: Line 1:
An '''Explosion''' is a sudden, violent expansion or production of gases that can be accompanied by heat, shock waves and disruption of nearby structural materials. Combustion explosions involve rapid reaction in a fuel-oxidant mixture, while other explosions can involve physical expansion and vessel failure. The resulting hazard depends on pressure development, flame, fragments and the surrounding geometry.<ref name="NUREG-F22">U.S. Nuclear Regulatory Commission (2005), p. F-22.</ref><ref name="Zalosh-3-402">Zalosh (2002), pp. 3-402–3-403.</ref>
An '''explosion''' is a rapid, potentially damaging rise in pressure, such as the one that follows when a combustible gas, vapour or dust mixed with air in a confined space is ignited. The flame travels through the mixture, and the confinement stops the hot combustion products from expanding freely.<ref name="Z402">Zalosh (2002), p. 3-402.</ref> NUREG-1805, a fire hazard analysis guide of the US Nuclear Regulatory Commission, gives a broader definition: a sudden and violent release of high-pressure gases into the surroundings.<ref name="IS-15-1">Iqbal and Salley (2004), p. 15-1.</ref> This definition does not depend on what produced the pressure.<ref name="IS-15-1" /> Explosions are often grouped into physical explosions, such as a heated tank that bursts, and chemical explosions, such as a burning cloud of gas or dust.<ref name="IS-15-2">Iqbal and Salley (2004), p. 15-2.</ref>


A deflagration propagates combustion at subsonic speed, whereas a detonation propagates it at supersonic speed through the unreacted medium. This distinction changes both the pressure distribution and suitable protection. It is therefore not enough to classify an event only as an explosion without identifying its mechanism.<ref name="Zalosh-3-411">Zalosh (2002), pp. 3-411–3-412.</ref><ref name="NUREG-F21">U.S. Nuclear Regulatory Commission (2005), p. F-21.</ref>
Explosions matter in fire safety because ordinary buildings are weak against pressure. Glass windows typically fail at 2 to 7 kPa, whereas a premixed gas-air mixture ignited inside a building can produce up to about 8 bar, or 800 kPa.<ref name="Harris">Harris (1983), as cited in Iqbal and Salley (2004), p. 15-13.</ref><ref name="IS-15-12">Iqbal and Salley (2004), p. 15-12.</ref> Engineers distinguish a [[deflagration]], in which the flame moves slower than sound, from a [[detonation]], in which it moves faster, because the two need different protection.<ref name="Z410">Zalosh (2002), p. 3-410.</ref><ref name="Z411">Zalosh (2002), p. 3-411.</ref> Protection works by preventing a flammable mixture, by removing ignition sources, or by limiting the pressure through venting, suppression or an enclosure strong enough to resist it.<ref name="Z402" /><ref name="IS-F20">Iqbal and Salley (2004), p. F-20.</ref>


Explosion assessment examines the formation of a hazardous condition, initiation, pressure or energy release and consequences. Protection can prevent flammable mixtures, control ignition, limit pressure or strengthen enclosures. Each measure has its own application limits and must be based on the relevant test and structural evidence.<ref name="Zalosh-3-402" /><ref name="Zalosh-3-414">Zalosh (2002), pp. 3-414–3-416.</ref>
== What counts as an explosion ==


== Combustion and pressure development ==
In its most widely accepted sense, an explosion is a bursting with a loud, sharp noise and an expanding pressure front, ranging from a supersonic shock wave to a relatively mild wind.<ref name="IS-15-1" /> The NUREG-1805 definition rests on two ideas. The gas must be at a higher pressure than its surroundings, and it must be released so fast that its energy escapes as a shock wave.<ref name="IS-15-1" /> The stored energy can be nuclear, chemical, electrical or pressure energy. A release counts as an explosion only if it produces a pressure wave that can be heard, even if it causes no damage.<ref name="IS-15-2" />


Combustible gases, vapours and dusts can form an explosion hazard when mixed with an oxidant between flammable limits and confined before ignition. Flame then propagates through the premixed fuel and oxidant. Restricted expansion of the hot combustion products produces a potentially damaging pressure rise.<ref name="Zalosh-3-402" />
The rate of release decides how violent an explosion is. The energy stored in a car tyre can cause an explosive burst, but the same energy can also be released gradually.<ref name="IS-15-16">Iqbal and Salley (2004), p. 15-16.</ref> In a combustion explosion the confinement keeps the hot gases from expanding freely.<ref name="Z402" /> The normal openings of process equipment are usually too small to prevent destructive pressures.<ref name="IS-15-3">Iqbal and Salley (2004), p. 15-3.</ref>


Flammability-limit tests determine whether flame propagates through a known mixture. Explosibility-limit tests instead focus on overpressure, and the two criteria need not produce identical values. The selected test must match whether the objective is flame prevention or pressure-hazard evaluation.<ref name="Zalosh-3-402" />
Explosions are often classified by how the high-pressure gas is produced.<ref name="IS-15-2" /> In a physical explosion the pressure comes from mechanical means without any chemical change. Examples are a tank that fails after external heating has raised its internal pressure, and a superheated liquid that is suddenly released and flashes into vapour.<ref name="IS-15-2" /> A pressure vessel that ruptures without its gas igniting also causes a physical explosion, driven by the expansion of the gas to atmospheric pressure.<ref name="Z419">Zalosh (2002), p. 3-419.</ref> In a chemical explosion the pressure comes from a chemical reaction. This group includes combustion explosions of dust, gas, mists and backdraft gases, thermal explosions in runaway reactions, and explosions of high explosives and propellants.<ref name="IS-15-2" />


Nonuniform mixtures can create local flammable regions even where the average concentration appears safe. Temperature, pressure and monitoring response can also change the relevant limits. A single room-average gas reading does not necessarily describe every region or time in the scenario.<ref name="Zalosh-3-403">Zalosh (2002), p. 3-403.</ref>
== How a gas or dust explosion develops ==


In a closed vessel deflagration, pressure depends on flame propagation, burned-gas temperature and composition, and heat losses. Complete reaction throughout the enclosure is a particular idealisation, not a description of every real event. Near-limit mixtures can react incompletely and lose heat to walls.<ref name="Zalosh-3-406">Zalosh (2002), pp. 3-406–3-407.</ref>
For a deflagration, NUREG-1805 names three things that must be present at the same time: a flammable mixture of fuel and oxygen, usually from air, a means of ignition, and an enclosure.<ref name="IS-15-3" /> The mixture can only explode between its lower and upper explosion limits, which are also called flammability limits.<ref name="IS-15-4">Iqbal and Salley (2004), p. 15-4.</ref> Outside this range a flame cannot spread through it.<ref name="IS-15-4" /> All glowing sparks are hot enough to ignite a mixture, but some carry too little energy, while a hot process surface may be cooler yet hold a large amount of energy.<ref name="IS-15-3" />


Structural response depends on the pressure-time history and enclosure behaviour. A slowly developing pressure rise and a shock wave of short duration present different loads. Evaluation must therefore consider both the maximum pressure and how quickly it develops.<ref name="Zalosh-3-410">Zalosh (2002), pp. 3-409–3-412.</ref>
Once ignited, the flame spreads through the mixture and heats the gas it passes.<ref name="Z402" /> For a hydrocarbon fuel mixed with exactly enough air for complete burning, the temperature rises from 300 K to about 2,200 K.<ref name="Simmons">Simmons (2002), p. 1-147.</ref> If no heat is lost, the pressure in a closed vessel can then climb to a little over 7 bar. A vessel that is not designed for such pressures will rupture.<ref name="Simmons" /> Measured maximum pressures for ten of twelve flammable gases and vapours lie between 6.8 and 8.1 bar above atmospheric pressure.<ref name="Z407">Zalosh (2002), p. 3-407.</ref> Near the flammability limits the pressures are lower, because the flame burns only part of the enclosure and heat is lost to the walls.<ref name="Z406">Zalosh (2002), pp. 3-406–3-407.</ref>


== Deflagration and detonation ==
The pressure does not rise evenly. In a worked example by Zalosh for a spherical vessel 7.15 m in radius, the pressure had risen by only about 14 kPa when the flame was halfway to the wall. The final rise was about 640 kPa.<ref name="Z408">Zalosh (2002), pp. 3-408–3-409.</ref> In that example most of the pressure rise happened at the end, when flame filled almost the whole vessel.<ref name="Z408" /> Reaching the maximum pressure can take up to a few hundred milliseconds, depending on the size of the enclosure and the fuel.<ref name="IS-15-3" /> With connected vessels, burning in one vessel raises the pressure in the next, and a flame that then enters it produces a correspondingly higher pressure.<ref name="Simmons" />


[[Deflagration]] is thermal propagation of an explosion reaction at less than the speed of sound in the mixture. Pressure can develop throughout an enclosure as the flame expands. Although slower than detonation, it can still rupture vessels and buildings.<ref name="NUREG-F20">U.S. Nuclear Regulatory Commission (2005), p. F-20.</ref><ref name="Zalosh-3-406" />
Explosions in the open air, called unconfined explosions, differ fundamentally from confined explosions in vessels, pipework or buildings, and need different countermeasures.<ref name="IS-15-5">Iqbal and Salley (2004), p. 15-5.</ref> When a gas leaks into the open, mixes with air and is ignited, the flame front spreads through the cloud as a sphere wherever it can.<ref name="IS-15-6">Iqbal and Salley (2004), p. 15-6.</ref> Inside a building, gases rushing out through a vent, door or window can create turbulence that produces a second pressure peak.<ref name="IS-15-5" />


[[Detonation]] is propagation of a combustion zone at supersonic speed through the unreacted medium. Its pressure rise is highly nonuniform and occurs almost instantaneously as the shock front reaches a location. This is physically different from the distributed development of a closed-vessel deflagration.<ref name="NUREG-F21" /><ref name="Zalosh-3-411" />
== Deflagration and detonation ==
 
Deflagration pressures can be moderated by early venting or suppression during the developing event. Detonation leaves no comparable period for those systems to respond at a downstream point. Its impulsive load requires different structural analysis and protection.<ref name="Zalosh-3-411" /><ref name="Zalosh-3-414" />
 
A quasi-detonation describes a situation where flame speed is slightly below sound speed but pressure is nonuniform without a detonation shock wave. This illustrates why pressure distribution matters in addition to the speed label. The physical evidence should determine the design case.<ref name="Zalosh-3-411" />
[[File:fse-explosion-v1.svg|thumb|upright=1.5|alt=Deflagration fills a vessel over time; detonation produces a sharp local shock front.|Deflagration and detonation create different pressure conditions. Based on Zalosh (2002), pp. 3-411–3-412 and U.S. Nuclear Regulatory Commission (2005), pp. F-20–F-21.]]


Deflagration-to-detonation transition must be considered where geometry and flame acceleration make it credible. Zalosh distinguishes the need to assess this possibility before choosing a protection method. A system designed only for an initial deflagration may not remain adequate after transition.<ref name="Zalosh-3-412">Zalosh (2002), pp. 3-412–3-413.</ref>
A deflagration is an explosion in which the reaction spreads through the mixture by heat, at a speed below the speed of sound. It is slower than a detonation but can still cause damage.<ref name="IS-F20" /> In a detonation the combustion zone travels faster than sound through the unburned mixture.<ref name="IS-F21">Iqbal and Salley (2004), p. F-21.</ref> A shock wave runs ahead of the reaction and compresses the gas, which speeds up the reaction.<ref name="IS-15-3" /> A deflagration flame moves well below the speed of sound, about 340 m/s in room-temperature air, so its pressure rises almost evenly throughout an enclosure.<ref name="Z410" /> In a detonation the pressure rise is very uneven and almost instantaneous as the shock passes each point.<ref name="Z410" />
[[File:fse-explosion-v1.svg|thumb|upright=1.5|alt=Deflagration fills a vessel over time; detonation produces a sharp local shock front.|Deflagration and detonation create different pressure conditions. Based on Zalosh (2002), pp. 3-410–3-411 and Iqbal and Salley (2004), pp. F-20–F-21.]]


== Dust explosions and secondary events ==
Typical detonation speeds are about 1,500 to 3,000 m/s, and NUREG-1805 gives a value of the order of 2,000 m/s.<ref name="Simmons" /><ref name="IS-15-3" /> In a gas mixture that starts at atmospheric pressure, the pressure just behind a detonation can be up to 20 bar.<ref name="Simmons" /> Zalosh notes that this is roughly twice the pressure of the same mixture burning in a closed vessel.<ref name="Z411" /> Where a detonation is reflected from the closed end of a duct, the pressure can be up to 100 bar. Detonations are therefore much more destructive than an ordinary moving flame.<ref name="Simmons" />


[[Dust explosion]] occurs when combustible particles suspended in air ignite and burn rapidly, producing heat and pressure. Particle size, concentration, turbulence and ignition conditions determine whether propagation occurs. Confinement can make the consequences substantially more severe.<ref name="Zalosh-3-403" />
The difference matters for protection. The sudden, uneven pressure rise of a detonation rules out explosion venting and explosion suppression, and its high, short pressure peaks need a special structural assessment.<ref name="Z411" /> Most accidental explosions are deflagrations, and NUREG-1805 describes detonations as very rare.<ref name="Z413">Zalosh (2002), p. 3-413.</ref><ref name="IS-15-12" /> The minimum energy needed to start a detonation directly ranges from about 5 kJ for acetylene or hydrogen in air to an estimated 93,000 kJ for methane. These energies are far greater than those of accidental ignition sources.<ref name="Z412">Zalosh (2002), p. 3-412.</ref>


Dust requires dispersion before or during the event. Air injection, vibration and process disturbances can create clouds, while settled layers can be lifted by an earlier explosion. The assessment must include both the initial cloud and accumulated material that could feed a later one.<ref name="Zalosh-3-404" /><ref name="Zalosh-3-410" />
A deflagration can, however, accelerate and turn into a detonation, especially in long pipes, elongated enclosures or very turbulent conditions.<ref name="Z412" /> This deflagration-to-detonation transition comes sooner with a more reactive mixture, rough walls or obstacles, a larger diameter, a higher starting pressure or temperature, and more turbulence.<ref name="Z412" /><ref name="Z413" /> In tests with petrol vapour and air, a smooth pipe 15 cm wide and 16 m long gave no transition. A rough liner at the pipe entrance, however, produced a detonation after 74 pipe diameters.<ref name="Z413" /> Because the protection differs so much, Zalosh stresses that the possibility of a detonation has to be assessed, although there are no exact criteria for doing so.<ref name="Z412" />


Secondary explosions arise when a primary event breaches equipment and disturbs surrounding dust. The blast and airflow suspend the dust, which can then ignite in the larger building. This sequence can produce casualties and structural damage beyond the original process enclosure.<ref name="Zalosh-3-410" />
== Explosion scenarios in fire safety ==


Dust explosion tests report maximum pressure and size-normalised rate-of-pressure-rise parameters under specified conditions. These results support comparison and vent design, but they depend on ignition source, dispersion delay and vessel geometry. They should not be used as context-free material constants.<ref name="Zalosh-3-410" /><ref name="Zalosh-3-415" />
A [[dust explosion]] occurs in a cloud of combustible powder suspended in air. Dusts pose this hazard when their typical particle size is below about 100 to 400 micrometres, depending on the material.<ref name="Z403">Zalosh (2002), p. 3-403.</ref> The concentration must also lie between the lower and upper explosive limits, and for dusts finer than about 100 micrometres the lower limit is typically 30 to 60 g/m³.<ref name="Z403" /> A particular danger is the secondary explosion, in which a first explosion bursts a piece of equipment and its blast lifts dust that has settled on walls and floors.<ref name="Z409">Zalosh (2002), pp. 3-409–3-410.</ref> Flame escaping from the equipment often ignites this new cloud, which in an occupied building can cause many casualties and structural damage.<ref name="Z409" />


== Vapour cloud explosions ==
A [[vapour cloud explosion]] occurs when a very large cloud of flammable gas or vapour is ignited in a congested area, such as a chemical plant crowded with process equipment and storage vessels.<ref name="Z420">Zalosh (2002), p. 3-420.</ref> In risk studies of process plants, it usually involves more than about a tonne of flammable vapour, which forms a large cloud and explodes after delayed ignition.<ref name="Barry-184">Barry (2002), p. 5-184.</ref> The strength of its blast depends on the flame speed that develops and on how strongly the cloud is confined.<ref name="Z420" /> In a flash fire the flame is slower: it spreads quickly through the flammable part of the cloud but causes no blast damage.<ref name="Barry-184" />


A [[Vapour cloud explosion]] involves ignition of a large flammable cloud with blast overpressure. Zalosh highlights highly obstructed chemical and petrochemical environments, where flame acceleration and confinement strongly affect blast strength. The term should not be applied to every cloud fire.<ref name="Zalosh-3-420">Zalosh (2002), pp. 3-420–3-421.</ref><ref name="Barry-5-184">Barry (2002), pp. 5-184–5-185.</ref>
A [[BLEVE]], short for boiling liquid expanding vapour explosion, occurs when a pressurised vessel of liquefied gas ruptures and the liquid suddenly expands into gas, producing blast waves.<ref name="IS-5-15">Iqbal and Salley (2004), p. 5-15.</ref> It typically happens when the metal of a container has been overheated above about 538 °C. Usually only the wall around the vapour space is at risk, because the liquid absorbs heat from the wetted wall.<ref name="IS-5-15" /> Barry describes it as a delayed event that generally follows 8 to 15 minutes of exposure to a fire.<ref name="Barry-184" /> The bursting vessel throws fragments and sends out a blast wave, and most BLEVEs also produce a rising fireball.<ref name="Z419" /> Fireballs in accidents with liquefied petroleum gas have usually been of the order of 100 m across and have lasted a few seconds.<ref name="Beyler">Beyler (2002a), p. 3-303.</ref>
[[File:fse-explosion-v2.svg|thumb|upright=1.5|alt=Heated vessel fails, releasing expanding vapour, fragments, blast and a fireball.|A BLEVE combines physical expansion with several possible hazards. Based on Iqbal and Salley (2004), p. 5-15 and Zalosh (2002), p. 3-419.]]


A flash fire can burn rapidly through the flammable zone without producing the same blast damage. Barry distinguishes that outcome from an unconfined vapour cloud explosion, which creates explosion overpressure around the area. Flame presence alone does not establish which consequence model applies.<ref name="Barry-5-184" />
Fires inside buildings can produce explosions of their own. When a fire is short of air, unburned pyrolysis gases can collect and form, once mixed with air, a flammable mixture that needs only ignition energy.<ref name="VR-4.8">Van Rossum (2027), §4.8.</ref> A [[smoke explosion]] is the very rapid burning of such gases from decomposed or incompletely burned fuel.<ref name="VR-4.10">Van Rossum (2027), §4.10.</ref> It typically requires the gases and air to be reasonably premixed within the explosion limits, and it can be ignited by a spark, by glowing particles or by autoignition.<ref name="VR-4.10" /> A backdraft is different: hot fire gases that lack only oxygen ignite explosively when fresh air suddenly enters, for example when a door is opened.<ref name="VR-4.9">Van Rossum (2027), §4.9.</ref> NUREG-1805 adds that a backdraft behaves like a premixed fuel-air deflagration, and that smoke explosions have been observed in smouldering polyurethane foam fires.<ref name="IS-15-4" />


Blast strength depends on flame speed and confinement around the cloud. Obstacles and equipment can accelerate flame and change the pressure field. A model based only on total cloud energy can miss those near-field effects.<ref name="Zalosh-3-420" />
Electric vehicles add a newer example. Battery cells in thermal runaway release flammable gases. If these do not ignite at once, they can build up into an explosive mixture in a confined or poorly ventilated space.<ref name="VdG-27">Van der Graaf et al. (2021/2023), p. 27.</ref> When the mixture reaches its lower explosion limit and is ignited, a fire gas explosion can occur.<ref name="VdG-27" /> Fires in electric vehicles can also involve repeated minor explosions, flying parts and flash fires.<ref name="VdG-29">Van der Graaf et al. (2021/2023), p. 29.</ref>


Zalosh explains that early ideal-blast estimates were inaccurate for vapour cloud explosions, particularly because real blast strength and impulse did not decay like an ideal point-source blast. Newer approaches use flame-speed estimates, nonideal propagation calculations or computational fluid dynamics. Their suitability depends on scenario knowledge and user competence.<ref name="Zalosh-3-420" />
== Effects on people and buildings ==


Release size, duration and isolation influence the cloud that can develop. Barry's discussion calls for realistic evaluation of detection and shutdown timing. Consequence assessment therefore begins before ignition, with the release and dispersion scenario.<ref name="Barry-5-184" />
An explosion sends out a blast wave, a thin shell of strongly compressed air and hot gases that spreads in all directions. It can travel faster than sound, which is why a large explosion is heard as a loud bang.<ref name="IS-15-7">Iqbal and Salley (2004), p. 15-7.</ref> Its strength is expressed as the peak overpressure, the highest pressure in the wave above normal atmospheric pressure.<ref name="IS-15-7" /> This overpressure is highest near the source and falls quickly with distance, and nearby reflecting surfaces strongly influence the damage.<ref name="IS-15-7" /> An explosion can also cause heat damage, damage from flying objects, ground shock, craters and injuries, although not every explosion does all of these.<ref name="IS-15-8">Iqbal and Salley (2004), p. 15-8.</ref>


== BLEVE and vessel failure ==
Buildings fail at low overpressures. Typical failure pressures are 2 to 3 kPa for room doors, 2 to 5 kPa for light partition walls and 7 to 15 kPa for unrestrained brick walls.<ref name="Harris" /> A gas explosion inside a building, at up to about 800 kPa, can therefore exceed the strength of such elements many times over.<ref name="IS-15-12" /> For process plants, Barry lists about 3.4 to 6.9 kPa (0.5 to 1 psi) as causing glass damage and injuries from flying glass. He gives 17 to 34 kPa (2.5 to 5 psi) for heavy damage to buildings and process equipment.<ref name="Barry-186">Barry (2002), p. 5-186.</ref>


[[BLEVE]] stands for boiling liquid expanding vapour explosion. It involves sudden expansion of a liquid into gas when a pressurised vessel fails, with energy that can create a blast wave. The NFPA definition quoted by NUREG-1805 concerns failure of a major container while its liquid is above its normal-atmospheric boiling point.<ref name="NUREG-5-15">U.S. Nuclear Regulatory Commission (2005), pp. 5-15–5-16.</ref>
People can withstand fairly high blast pressures, and when blast waves kill people it is usually through falling objects rather than through the pressure itself.<ref name="IS-15-12" /> Data quoted in NUREG-1805 give about 35 kPa as the threshold for eardrum rupture and about 70 kPa for lung damage.<ref name="Fischer">Fischer et al. (1995), as cited in Iqbal and Salley (2004), p. 15-12.</ref> The blast can also hurl fragments and debris, and it can throw people against the ground, walls or other fixed objects.<ref name="IS-15-9">Iqbal and Salley (2004), p. 15-9.</ref> NUREG-1805 notes that the consequences for people depend strongly on whether anything nearby can strike them.<ref name="IS-15-12" />


A vessel can fail from impact, corrosion, internal overheating, manufacturing defects or other weakening mechanisms. NUREG-1805 explains that the unwetted internal surface is particularly exposed to fire heating because it lacks the liquid's heat absorption. The physical condition of the vessel must therefore be part of the scenario.<ref name="NUREG-5-15" />
A common way to estimate blast effects is to express an explosion as an equivalent mass of TNT (trinitrotoluene), whose blast effects have been studied in detail.<ref name="IS-15-10">Iqbal and Salley (2004), p. 15-10.</ref> Such methods treat the explosion as an ideal blast released at a single point.<ref name="Z419" /><ref name="IS-15-15">Iqbal and Salley (2004), p. 15-15.</ref> They cannot be valid inside or close to a flammable vapour cloud.<ref name="IS-15-15" /> For vapour cloud explosions the real blast also weakens far less quickly with distance than an ideal one.<ref name="Z420" /> Current methods instead estimate the flame speed for the type of application, or use computer models that calculate how the flame spreads and accelerates.<ref name="Z420" />
[[File:fse-explosion-v2.svg|thumb|upright=1.5|alt=Heated vessel fails, releasing expanding vapour, fragments, blast and a fireball.|A BLEVE combines physical expansion with several possible hazards. Based on U.S. Nuclear Regulatory Commission (2005), pp. 5-15–5-16 and Zalosh (2002), p. 3-420.]]


A BLEVE can produce a large fireball and intense radiant heat when the released contents ignite. The fireball is a separate hazard from blast and vessel fragments. Assessment should consider fuel mass, fireball size, duration, radiation and fragment trajectories.<ref name="NUREG-5-15" /><ref name="Zalosh-3-420" />
== Prevention and protection ==


Not every BLEVE is simply a combustion deflagration inside a vessel. The initial phase change and vessel rupture are central to its mechanism. Subsequent fire and thermal radiation may then expose nearby people and property.<ref name="NUREG-5-15" /><ref name="Barry-5-184" />
Explosion protection interrupts the chain of events that leads to an explosion. It can prevent a flammable fuel-air mixture from forming, for example by ventilation or [[inerting]]. It can remove ignition sources, for example by earthing equipment so that static electricity cannot discharge as a spark. It can also limit the pressure rise by explosion venting or explosion suppression.<ref name="Z402" /> Deflagration pressure containment instead designs a vessel and its fittings to withstand the maximum pressure of an internal deflagration.<ref name="IS-F20" /> In the United States, Factory Mutual Data Sheet 7-32 (1993 edition) recommends at least one of four protections for process equipment that could explode: venting, enough strength to resist the overpressure, inerting or suppression.<ref name="IS-B27">Iqbal and Salley (2004), pp. B-27–B-28.</ref>


The source notes that many BLEVEs involve containers less than half to completely full, but this is a general observation from the cited NUREG section, not a design rule. It should not be used to infer that another fill condition is safe.<ref name="NUREG-5-15" />
Inerting adds a gas such as nitrogen or carbon dioxide to reduce the oxygen concentration to a safe level. The highest safe level is called the limiting oxidant concentration.<ref name="Z402" /><ref name="Z403" /> For the flammable gases and vapours listed by Zalosh, apart from hydrogen, this limit is 9.5 to 12 per cent oxygen by volume with nitrogen. With carbon dioxide it is 11.5 to 14.5 per cent, while hydrogen needs about 5 per cent with either gas.<ref name="Z403" /> In the United States, NFPA 69 (1997 edition) states that oxygen should be kept at least 2 percentage points below this limit under continuous monitoring, and further below it otherwise.<ref name="Z403" /> Uneven mixtures and the accuracy and response time of oxygen monitors also have to be considered.<ref name="Z403" />


== Smoke explosion ==
[[Explosion venting]] releases the combustion gases through openings during a deflagration, so that the pressure stays below the level that would damage the enclosure.<ref name="Z414">Zalosh (2002), p. 3-414.</ref> The openings are usually covered by blow-off panels, rupture discs or other vent devices, which usually open only after ignition. Venting works best when the vents open early, are as large as possible and let the gases escape freely. In the outside walls and roofs of buildings, the lowest practical opening pressure is usually just above the highest expected wind pressure, typically 0.96 to 1.44 kPa.<ref name="Z414" /> Explosion suppression instead detects an explosion in its first moments and discharges an extinguishing agent that quenches the flame.<ref name="Z417">Zalosh (2002), p. 3-417.</ref> It releases no flame or fuel, so it suits indoor equipment and toxic materials, but it is expensive to install and to reset.<ref name="Z417" />


A [[Smoke explosion]] involves very rapid combustion of gases formed by thermal decomposition or incomplete combustion. Van Rossum describes a typical requirement for air and pyrolysis gases to be reasonably premixed and within the lower and upper explosion limits. An ignition source or autoignition can initiate the event.<ref name="VR-4-10">Van Rossum (2027), §4.10.</ref>
Neither venting nor suppression can protect against a detonation. Vents open only after the explosion has started, while a detonation reaches its maximum pressure the instant its shock front arrives.<ref name="Z414" /><ref name="Z411" /> Venting guidelines are also limited because they are based on tests and do not account for the turbulence and equipment obstructions of a particular application.<ref name="Z416">Zalosh (2002), p. 3-416.</ref> Most of them also ignore how much of the enclosure is filled with the flammable mixture.<ref name="Z416" />
 
Oxygen-limited combustion can allow unburned pyrolysis gases to accumulate. If air later mixes with those gases, a flammable cloud can form. The mixture then requires suitable temperature or another ignition source before rapid combustion occurs.<ref name="VR-4-8">Van Rossum (2027), §4.8.</ref><ref name="VR-4-10" />
 
This differs from a backdraft, which Van Rossum defines as explosive ignition of pyrolysis gases when oxygen suddenly enters a room. The two phenomena are related to ventilation and unburned gases, but they should not be treated as identical labels for every rapid smoke ignition.<ref name="VR-4-9">Van Rossum (2027), §4.9.</ref><ref name="VR-4-10" />
 
Smoke explosion assessment must therefore include mixture history. It is not sufficient to ask whether a flame is currently visible. Fuel decomposition, oxygen depletion, gas accumulation and later mixing can establish the conditions for a later explosion.<ref name="VR-4-8" /><ref name="VR-4-10" />
 
== Explosion protection ==
 
Protection can interrupt the sequence before ignition. Ventilation or inerting can prevent a flammable mixture, while elimination of ignition sources can stop initiation. These measures should address actual process and environmental conditions rather than generic fuel names alone.<ref name="Zalosh-3-402" />
 
Inerting reduces oxygen below the limiting concentration required for combustion. The necessary concentration depends on the fuel, inert gas and process conditions. Monitoring accuracy, nonuniformity and response time must be included in its reliability assessment.<ref name="Zalosh-3-403" />
 
[[Explosion venting]] discharges gases during a deflagration to keep pressure below the enclosure damage threshold. It uses engineered openings that deploy after initiation. Early opening and unrestricted discharge are central to the cited performance, but required area depends on the enclosure and mixture.<ref name="Zalosh-3-414" />
 
Suppression detects and arrests combustion while it remains incipient. Containment instead specifies an enclosure able to withstand the expected deflagration pressure. These approaches act differently and should not be presented as substitutes for every explosion scenario.<ref name="NUREG-F20" /><ref name="Zalosh-3-402" />
[[File:fse-explosion-v3.svg|thumb|upright=1.5|alt=Prevention, venting and containment are mapped to deflagration; venting is excluded for detonation.|Protection must match the explosion mechanism. Based on Zalosh (2002), pp. 3-411–3-416.]]
[[File:fse-explosion-v3.svg|thumb|upright=1.5|alt=Prevention, venting and containment are mapped to deflagration; venting is excluded for detonation.|Protection must match the explosion mechanism. Based on Zalosh (2002), pp. 3-411–3-416.]]


Explosion venting and many suppression methods are unsuitable for detonation. The local shock reaches a point too quickly for a developing pressure rise to be relieved in time. Protection must therefore begin with the correct deflagration-versus-detonation assessment.<ref name="Zalosh-3-411" /><ref name="Zalosh-3-414" />
== Assessment and uncertainty ==
 
== Testing, modelling and uncertainty ==
 
Explosion tests establish flammability, pressure and rate-of-pressure-rise parameters under defined conditions. Vessel size, ignition strength, turbulence and geometry influence results. Those conditions must accompany any value used in a design or comparison.<ref name="Zalosh-3-404" /><ref name="Zalosh-3-415" />
 
Vented-explosion models add vent area, deployment pressure and vent-device behaviour. They also need to represent enclosure scale, flame acceleration and equipment obstruction. A correlation developed from one test programme should not be applied to an unrelated geometry without justification.<ref name="Zalosh-3-414" /><ref name="Zalosh-3-416">Zalosh (2002), pp. 3-415–3-417.</ref>
 
Computational fluid dynamics can represent spatial variation in gas velocity, composition and thermal properties. Zalosh notes that these models are more complicated than two-fluid approaches and are used for special applications. Their added detail does not remove the need for suitable input data and validation.<ref name="Zalosh-3-414" />
 
Blast and structural assessment must consider reflected waves, impulse and fragment hazards where relevant. A single peak-pressure number does not capture all those effects. The selected model should correspond to the physical event and the structural question.<ref name="Zalosh-3-411" /><ref name="Zalosh-3-420" />
 
Risk assessment combines frequency and consequence, but both are scenario dependent. Barry describes its use for prevention, protection and emergency planning. An explosion model's output becomes useful for decisions only when its scenario, limitations and exposed targets are explicit.<ref name="Barry-5-176">Barry (2002), pp. 5-176–5-177.</ref>
 
Explosion scenarios should include the period before the hazardous mixture forms. Fuel storage, release rate, dispersion and isolation determine how much material can participate and where it travels. Limiting a release can change the scenario before ignition-control or pressure-relief measures are needed.<ref name="Barry-5-184" /><ref name="Zalosh-3-420" />
 
Mixture evidence should be interpreted by location and time. A cloud that is below a flammability limit at one point may be within it elsewhere, and continued release can change the average. The assessment should not treat a single sensor value as a complete description of a developing cloud.<ref name="Zalosh-3-403" />
 
Fuel form changes the relevant test and model. Gas, vapour, dust and pressurised liquefied gas have different dispersion and energy-release processes. Their scenarios should not be combined merely because all can produce an explosion.<ref name="Zalosh-3-402" /><ref name="NUREG-5-15" />


Structural assessment should retain the difference between pressure resistance and dynamic response. An enclosure may be designed for a specified deflagration pressure, but a detonation creates a different impulse and spatial load. Fragments and reflected waves can also create hazards beyond the enclosure.<ref name="Zalosh-3-411" /><ref name="Zalosh-3-420" />
An explosion hazard evaluation usually has three steps. It recognises where a flammable fuel-air mixture could form, identifies ignition sources that could then be present, and works out the resulting pressures and possible structural damage.<ref name="Z402" /> In the chemical process industries, fire and explosion risk can also be assessed quantitatively, as a function of how often an event occurs and how severe its consequences are.<ref name="Barry-176">Barry (2002), p. 5-176.</ref> Such studies help to compare design options and to map explosion overpressure zones for emergency planning.<ref name="Barry-176" /> The amount of fuel in an accident depends on the release rate and duration, and so on how quickly a leak is detected and isolated.<ref name="Barry-182">Barry (2002), p. 5-182.</ref><ref name="Barry-184" />


Operational controls should preserve the protection assumptions. Vent openings must remain available, inerting systems need suitable monitoring, and process changes can alter dust or vapour formation. A protection measure that exists on a drawing may not be effective after an alteration or impairment.<ref name="Zalosh-3-403" /><ref name="Zalosh-3-414" />
Evaluations have to allow for uneven or layered mixtures of gas and air.<ref name="Z403" /> Zalosh gives the example of about 200 g of butane leaking from an aerosol can into a filling room of 4 by 4 by 3 m.<ref name="Z410" /> Spread evenly through the 48 m³ room, this gas would be too little to form a flammable mixture. If it mixes in the right proportion with only part of the air and is ignited, the pressure in the unvented room still rises by about 0.46 bar.<ref name="Z410" /> That is more than rooms in industrial buildings can withstand, so some form of explosion venting and suppression would be needed.<ref name="Z410" />


The final scenario record should identify the initiating event, fuel and oxidant state, ignition mechanism, enclosure, pressure response and exposed targets. It should also state which uncertainties could change the mechanism, including possible transition to detonation. This keeps the selected protection tied to the physical event rather than to a generic explosion label. The record should also distinguish demonstrated performance from assumptions that still require test evidence or further analysis. Each assumption should remain traceable to its supporting document, test or explicitly justified judgement.<ref name="Zalosh-3-402" /><ref name="Zalosh-3-412" />
Several limits make explosion predictions uncertain, and NUREG-1805 notes that the pressures produced are usually difficult to predict.<ref name="IS-15-12" /> Dust explosions are harder to model than gas explosions.<ref name="Z414" /> Detailed computer models of gas explosions are too complex for routine vent design, but some are used for special cases such as congested offshore platform modules.<ref name="Z414" /> Laboratory results depend on the ignition source and on the turbulence in the test vessel at the moment of ignition.<ref name="Z404">Zalosh (2002), p. 3-404.</ref> A standard method for dust tests was therefore developed so that different laboratories obtain reproducible results.<ref name="Z409" /> Barry also reports that human factors, such as management controls and operator errors, account for about 70 to 85 per cent of the primary causes of fires and explosions in the process industries.<ref name="Barry-188">Barry (2002), p. 5-188.</ref>


== See also ==
== See also ==
Line 116: Line 78:
* [[Fireball]]
* [[Fireball]]
* [[Flammability limits]]
* [[Flammability limits]]
* [[Hazardous area classification]]
* [[Premixed flame]]
* [[Premixed flame]]
* [[Thermal runaway]]
* [[TNT equivalence]]
* [[TNT equivalence]]


== Notes ==
== Notes ==


<ref name="Zalosh-3-404">Zalosh (2002), pp. 3-404–3-405.</ref>
<ref name="Zalosh-3-415">Zalosh (2002), pp. 3-415–3-416.</ref>
<references />
<references />


== References ==
== References ==


* Barry, T. F. (2002). Quantitative risk assessment in chemical process industries. In P. J. DiNenno (Ed.), ''SFPE handbook of fire protection engineering'' (3rd ed.). National Fire Protection Association.
* Barry, T. F. (2002). Quantitative risk assessment in chemical process industries. 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. 5-176–5-213). National Fire Protection Association; Society of Fire Protection Engineers.
* U.S. Nuclear Regulatory Commission. (2005). ''Fire dynamics tools: Quantitative fire hazard analysis methods for the U.S. nuclear regulatory commission fire protection inspection program'' (NUREG-1805). Author.
* Beyler, C. L. (2002a). Fire hazard calculations for large, open hydrocarbon fires. 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. 3-268–3-314). National Fire Protection Association; Society of Fire Protection Engineers.
* 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.
* Simmons, R. F. (2002). Premixed burning. 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. 1-144–1-154). National Fire Protection Association; Society of Fire Protection Engineers.
* Van der Graaf, J., Hessels, T., & Rosmuller, N. (2023). ''Fire safety of indoor car parks accommodating electrically powered vehicles''. Nederlands Instituut Publieke Veiligheid. (Original work published 2021)
* Van Rossum, E. (2027). ''Introduction to Fire Safety Engineering''. Fire Safety Academy.
* Van Rossum, E. (2027). ''Introduction to Fire Safety Engineering''. Fire Safety Academy.
* Zalosh, R. (2002). Explosion protection. In P. J. DiNenno (Ed.), ''SFPE handbook of fire protection engineering'' (3rd ed.). 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.


[[Category:Explosions & Industrial Fires]]
[[Category:Explosions & Industrial Fires]]

Revision as of 18:52, 2 November 2026

An explosion is a rapid, potentially damaging rise in pressure, such as the one that follows when a combustible gas, vapour or dust mixed with air in a confined space is ignited. The flame travels through the mixture, and the confinement stops the hot combustion products from expanding freely.[1] NUREG-1805, a fire hazard analysis guide of the US Nuclear Regulatory Commission, gives a broader definition: a sudden and violent release of high-pressure gases into the surroundings.[2] This definition does not depend on what produced the pressure.[2] Explosions are often grouped into physical explosions, such as a heated tank that bursts, and chemical explosions, such as a burning cloud of gas or dust.[3]

Explosions matter in fire safety because ordinary buildings are weak against pressure. Glass windows typically fail at 2 to 7 kPa, whereas a premixed gas-air mixture ignited inside a building can produce up to about 8 bar, or 800 kPa.[4][5] Engineers distinguish a deflagration, in which the flame moves slower than sound, from a detonation, in which it moves faster, because the two need different protection.[6][7] Protection works by preventing a flammable mixture, by removing ignition sources, or by limiting the pressure through venting, suppression or an enclosure strong enough to resist it.[1][8]

What counts as an explosion

In its most widely accepted sense, an explosion is a bursting with a loud, sharp noise and an expanding pressure front, ranging from a supersonic shock wave to a relatively mild wind.[2] The NUREG-1805 definition rests on two ideas. The gas must be at a higher pressure than its surroundings, and it must be released so fast that its energy escapes as a shock wave.[2] The stored energy can be nuclear, chemical, electrical or pressure energy. A release counts as an explosion only if it produces a pressure wave that can be heard, even if it causes no damage.[3]

The rate of release decides how violent an explosion is. The energy stored in a car tyre can cause an explosive burst, but the same energy can also be released gradually.[9] In a combustion explosion the confinement keeps the hot gases from expanding freely.[1] The normal openings of process equipment are usually too small to prevent destructive pressures.[10]

Explosions are often classified by how the high-pressure gas is produced.[3] In a physical explosion the pressure comes from mechanical means without any chemical change. Examples are a tank that fails after external heating has raised its internal pressure, and a superheated liquid that is suddenly released and flashes into vapour.[3] A pressure vessel that ruptures without its gas igniting also causes a physical explosion, driven by the expansion of the gas to atmospheric pressure.[11] In a chemical explosion the pressure comes from a chemical reaction. This group includes combustion explosions of dust, gas, mists and backdraft gases, thermal explosions in runaway reactions, and explosions of high explosives and propellants.[3]

How a gas or dust explosion develops

For a deflagration, NUREG-1805 names three things that must be present at the same time: a flammable mixture of fuel and oxygen, usually from air, a means of ignition, and an enclosure.[10] The mixture can only explode between its lower and upper explosion limits, which are also called flammability limits.[12] Outside this range a flame cannot spread through it.[12] All glowing sparks are hot enough to ignite a mixture, but some carry too little energy, while a hot process surface may be cooler yet hold a large amount of energy.[10]

Once ignited, the flame spreads through the mixture and heats the gas it passes.[1] For a hydrocarbon fuel mixed with exactly enough air for complete burning, the temperature rises from 300 K to about 2,200 K.[13] If no heat is lost, the pressure in a closed vessel can then climb to a little over 7 bar. A vessel that is not designed for such pressures will rupture.[13] Measured maximum pressures for ten of twelve flammable gases and vapours lie between 6.8 and 8.1 bar above atmospheric pressure.[14] Near the flammability limits the pressures are lower, because the flame burns only part of the enclosure and heat is lost to the walls.[15]

The pressure does not rise evenly. In a worked example by Zalosh for a spherical vessel 7.15 m in radius, the pressure had risen by only about 14 kPa when the flame was halfway to the wall. The final rise was about 640 kPa.[16] In that example most of the pressure rise happened at the end, when flame filled almost the whole vessel.[16] Reaching the maximum pressure can take up to a few hundred milliseconds, depending on the size of the enclosure and the fuel.[10] With connected vessels, burning in one vessel raises the pressure in the next, and a flame that then enters it produces a correspondingly higher pressure.[13]

Explosions in the open air, called unconfined explosions, differ fundamentally from confined explosions in vessels, pipework or buildings, and need different countermeasures.[17] When a gas leaks into the open, mixes with air and is ignited, the flame front spreads through the cloud as a sphere wherever it can.[18] Inside a building, gases rushing out through a vent, door or window can create turbulence that produces a second pressure peak.[17]

Deflagration and detonation

A deflagration is an explosion in which the reaction spreads through the mixture by heat, at a speed below the speed of sound. It is slower than a detonation but can still cause damage.[8] In a detonation the combustion zone travels faster than sound through the unburned mixture.[19] A shock wave runs ahead of the reaction and compresses the gas, which speeds up the reaction.[10] A deflagration flame moves well below the speed of sound, about 340 m/s in room-temperature air, so its pressure rises almost evenly throughout an enclosure.[6] In a detonation the pressure rise is very uneven and almost instantaneous as the shock passes each point.[6]

Deflagration fills a vessel over time; detonation produces a sharp local shock front.
Deflagration and detonation create different pressure conditions. Based on Zalosh (2002), pp. 3-410–3-411 and Iqbal and Salley (2004), pp. F-20–F-21.

Typical detonation speeds are about 1,500 to 3,000 m/s, and NUREG-1805 gives a value of the order of 2,000 m/s.[13][10] In a gas mixture that starts at atmospheric pressure, the pressure just behind a detonation can be up to 20 bar.[13] Zalosh notes that this is roughly twice the pressure of the same mixture burning in a closed vessel.[7] Where a detonation is reflected from the closed end of a duct, the pressure can be up to 100 bar. Detonations are therefore much more destructive than an ordinary moving flame.[13]

The difference matters for protection. The sudden, uneven pressure rise of a detonation rules out explosion venting and explosion suppression, and its high, short pressure peaks need a special structural assessment.[7] Most accidental explosions are deflagrations, and NUREG-1805 describes detonations as very rare.[20][5] The minimum energy needed to start a detonation directly ranges from about 5 kJ for acetylene or hydrogen in air to an estimated 93,000 kJ for methane. These energies are far greater than those of accidental ignition sources.[21]

A deflagration can, however, accelerate and turn into a detonation, especially in long pipes, elongated enclosures or very turbulent conditions.[21] This deflagration-to-detonation transition comes sooner with a more reactive mixture, rough walls or obstacles, a larger diameter, a higher starting pressure or temperature, and more turbulence.[21][20] In tests with petrol vapour and air, a smooth pipe 15 cm wide and 16 m long gave no transition. A rough liner at the pipe entrance, however, produced a detonation after 74 pipe diameters.[20] Because the protection differs so much, Zalosh stresses that the possibility of a detonation has to be assessed, although there are no exact criteria for doing so.[21]

Explosion scenarios in fire safety

A dust explosion occurs in a cloud of combustible powder suspended in air. Dusts pose this hazard when their typical particle size is below about 100 to 400 micrometres, depending on the material.[22] The concentration must also lie between the lower and upper explosive limits, and for dusts finer than about 100 micrometres the lower limit is typically 30 to 60 g/m³.[22] A particular danger is the secondary explosion, in which a first explosion bursts a piece of equipment and its blast lifts dust that has settled on walls and floors.[23] Flame escaping from the equipment often ignites this new cloud, which in an occupied building can cause many casualties and structural damage.[23]

A vapour cloud explosion occurs when a very large cloud of flammable gas or vapour is ignited in a congested area, such as a chemical plant crowded with process equipment and storage vessels.[24] In risk studies of process plants, it usually involves more than about a tonne of flammable vapour, which forms a large cloud and explodes after delayed ignition.[25] The strength of its blast depends on the flame speed that develops and on how strongly the cloud is confined.[24] In a flash fire the flame is slower: it spreads quickly through the flammable part of the cloud but causes no blast damage.[25]

A BLEVE, short for boiling liquid expanding vapour explosion, occurs when a pressurised vessel of liquefied gas ruptures and the liquid suddenly expands into gas, producing blast waves.[26] It typically happens when the metal of a container has been overheated above about 538 °C. Usually only the wall around the vapour space is at risk, because the liquid absorbs heat from the wetted wall.[26] Barry describes it as a delayed event that generally follows 8 to 15 minutes of exposure to a fire.[25] The bursting vessel throws fragments and sends out a blast wave, and most BLEVEs also produce a rising fireball.[11] Fireballs in accidents with liquefied petroleum gas have usually been of the order of 100 m across and have lasted a few seconds.[27]

Heated vessel fails, releasing expanding vapour, fragments, blast and a fireball.
A BLEVE combines physical expansion with several possible hazards. Based on Iqbal and Salley (2004), p. 5-15 and Zalosh (2002), p. 3-419.

Fires inside buildings can produce explosions of their own. When a fire is short of air, unburned pyrolysis gases can collect and form, once mixed with air, a flammable mixture that needs only ignition energy.[28] A smoke explosion is the very rapid burning of such gases from decomposed or incompletely burned fuel.[29] It typically requires the gases and air to be reasonably premixed within the explosion limits, and it can be ignited by a spark, by glowing particles or by autoignition.[29] A backdraft is different: hot fire gases that lack only oxygen ignite explosively when fresh air suddenly enters, for example when a door is opened.[30] NUREG-1805 adds that a backdraft behaves like a premixed fuel-air deflagration, and that smoke explosions have been observed in smouldering polyurethane foam fires.[12]

Electric vehicles add a newer example. Battery cells in thermal runaway release flammable gases. If these do not ignite at once, they can build up into an explosive mixture in a confined or poorly ventilated space.[31] When the mixture reaches its lower explosion limit and is ignited, a fire gas explosion can occur.[31] Fires in electric vehicles can also involve repeated minor explosions, flying parts and flash fires.[32]

Effects on people and buildings

An explosion sends out a blast wave, a thin shell of strongly compressed air and hot gases that spreads in all directions. It can travel faster than sound, which is why a large explosion is heard as a loud bang.[33] Its strength is expressed as the peak overpressure, the highest pressure in the wave above normal atmospheric pressure.[33] This overpressure is highest near the source and falls quickly with distance, and nearby reflecting surfaces strongly influence the damage.[33] An explosion can also cause heat damage, damage from flying objects, ground shock, craters and injuries, although not every explosion does all of these.[34]

Buildings fail at low overpressures. Typical failure pressures are 2 to 3 kPa for room doors, 2 to 5 kPa for light partition walls and 7 to 15 kPa for unrestrained brick walls.[4] A gas explosion inside a building, at up to about 800 kPa, can therefore exceed the strength of such elements many times over.[5] For process plants, Barry lists about 3.4 to 6.9 kPa (0.5 to 1 psi) as causing glass damage and injuries from flying glass. He gives 17 to 34 kPa (2.5 to 5 psi) for heavy damage to buildings and process equipment.[35]

People can withstand fairly high blast pressures, and when blast waves kill people it is usually through falling objects rather than through the pressure itself.[5] Data quoted in NUREG-1805 give about 35 kPa as the threshold for eardrum rupture and about 70 kPa for lung damage.[36] The blast can also hurl fragments and debris, and it can throw people against the ground, walls or other fixed objects.[37] NUREG-1805 notes that the consequences for people depend strongly on whether anything nearby can strike them.[5]

A common way to estimate blast effects is to express an explosion as an equivalent mass of TNT (trinitrotoluene), whose blast effects have been studied in detail.[38] Such methods treat the explosion as an ideal blast released at a single point.[11][39] They cannot be valid inside or close to a flammable vapour cloud.[39] For vapour cloud explosions the real blast also weakens far less quickly with distance than an ideal one.[24] Current methods instead estimate the flame speed for the type of application, or use computer models that calculate how the flame spreads and accelerates.[24]

Prevention and protection

Explosion protection interrupts the chain of events that leads to an explosion. It can prevent a flammable fuel-air mixture from forming, for example by ventilation or inerting. It can remove ignition sources, for example by earthing equipment so that static electricity cannot discharge as a spark. It can also limit the pressure rise by explosion venting or explosion suppression.[1] Deflagration pressure containment instead designs a vessel and its fittings to withstand the maximum pressure of an internal deflagration.[8] In the United States, Factory Mutual Data Sheet 7-32 (1993 edition) recommends at least one of four protections for process equipment that could explode: venting, enough strength to resist the overpressure, inerting or suppression.[40]

Inerting adds a gas such as nitrogen or carbon dioxide to reduce the oxygen concentration to a safe level. The highest safe level is called the limiting oxidant concentration.[1][22] For the flammable gases and vapours listed by Zalosh, apart from hydrogen, this limit is 9.5 to 12 per cent oxygen by volume with nitrogen. With carbon dioxide it is 11.5 to 14.5 per cent, while hydrogen needs about 5 per cent with either gas.[22] In the United States, NFPA 69 (1997 edition) states that oxygen should be kept at least 2 percentage points below this limit under continuous monitoring, and further below it otherwise.[22] Uneven mixtures and the accuracy and response time of oxygen monitors also have to be considered.[22]

Explosion venting releases the combustion gases through openings during a deflagration, so that the pressure stays below the level that would damage the enclosure.[41] The openings are usually covered by blow-off panels, rupture discs or other vent devices, which usually open only after ignition. Venting works best when the vents open early, are as large as possible and let the gases escape freely. In the outside walls and roofs of buildings, the lowest practical opening pressure is usually just above the highest expected wind pressure, typically 0.96 to 1.44 kPa.[41] Explosion suppression instead detects an explosion in its first moments and discharges an extinguishing agent that quenches the flame.[42] It releases no flame or fuel, so it suits indoor equipment and toxic materials, but it is expensive to install and to reset.[42]

Neither venting nor suppression can protect against a detonation. Vents open only after the explosion has started, while a detonation reaches its maximum pressure the instant its shock front arrives.[41][7] Venting guidelines are also limited because they are based on tests and do not account for the turbulence and equipment obstructions of a particular application.[43] Most of them also ignore how much of the enclosure is filled with the flammable mixture.[43]

Prevention, venting and containment are mapped to deflagration; venting is excluded for detonation.
Protection must match the explosion mechanism. Based on Zalosh (2002), pp. 3-411–3-416.

Assessment and uncertainty

An explosion hazard evaluation usually has three steps. It recognises where a flammable fuel-air mixture could form, identifies ignition sources that could then be present, and works out the resulting pressures and possible structural damage.[1] In the chemical process industries, fire and explosion risk can also be assessed quantitatively, as a function of how often an event occurs and how severe its consequences are.[44] Such studies help to compare design options and to map explosion overpressure zones for emergency planning.[44] The amount of fuel in an accident depends on the release rate and duration, and so on how quickly a leak is detected and isolated.[45][25]

Evaluations have to allow for uneven or layered mixtures of gas and air.[22] Zalosh gives the example of about 200 g of butane leaking from an aerosol can into a filling room of 4 by 4 by 3 m.[6] Spread evenly through the 48 m³ room, this gas would be too little to form a flammable mixture. If it mixes in the right proportion with only part of the air and is ignited, the pressure in the unvented room still rises by about 0.46 bar.[6] That is more than rooms in industrial buildings can withstand, so some form of explosion venting and suppression would be needed.[6]

Several limits make explosion predictions uncertain, and NUREG-1805 notes that the pressures produced are usually difficult to predict.[5] Dust explosions are harder to model than gas explosions.[41] Detailed computer models of gas explosions are too complex for routine vent design, but some are used for special cases such as congested offshore platform modules.[41] Laboratory results depend on the ignition source and on the turbulence in the test vessel at the moment of ignition.[46] A standard method for dust tests was therefore developed so that different laboratories obtain reproducible results.[23] Barry also reports that human factors, such as management controls and operator errors, account for about 70 to 85 per cent of the primary causes of fires and explosions in the process industries.[47]

See also

Notes

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 Zalosh (2002), p. 3-402.
  2. ↑ 2.0 2.1 2.2 2.3 Iqbal and Salley (2004), p. 15-1.
  3. ↑ 3.0 3.1 3.2 3.3 3.4 Iqbal and Salley (2004), p. 15-2.
  4. ↑ 4.0 4.1 Harris (1983), as cited in Iqbal and Salley (2004), p. 15-13.
  5. ↑ 5.0 5.1 5.2 5.3 5.4 5.5 Iqbal and Salley (2004), p. 15-12.
  6. ↑ 6.0 6.1 6.2 6.3 6.4 6.5 Zalosh (2002), p. 3-410.
  7. ↑ 7.0 7.1 7.2 7.3 Zalosh (2002), p. 3-411.
  8. ↑ 8.0 8.1 8.2 Iqbal and Salley (2004), p. F-20.
  9. ↑ Iqbal and Salley (2004), p. 15-16.
  10. ↑ 10.0 10.1 10.2 10.3 10.4 10.5 Iqbal and Salley (2004), p. 15-3.
  11. ↑ 11.0 11.1 11.2 Zalosh (2002), p. 3-419.
  12. ↑ 12.0 12.1 12.2 Iqbal and Salley (2004), p. 15-4.
  13. ↑ 13.0 13.1 13.2 13.3 13.4 13.5 Simmons (2002), p. 1-147.
  14. ↑ Zalosh (2002), p. 3-407.
  15. ↑ Zalosh (2002), pp. 3-406–3-407.
  16. ↑ 16.0 16.1 Zalosh (2002), pp. 3-408–3-409.
  17. ↑ 17.0 17.1 Iqbal and Salley (2004), p. 15-5.
  18. ↑ Iqbal and Salley (2004), p. 15-6.
  19. ↑ Iqbal and Salley (2004), p. F-21.
  20. ↑ 20.0 20.1 20.2 Zalosh (2002), p. 3-413.
  21. ↑ 21.0 21.1 21.2 21.3 Zalosh (2002), p. 3-412.
  22. ↑ 22.0 22.1 22.2 22.3 22.4 22.5 22.6 Zalosh (2002), p. 3-403.
  23. ↑ 23.0 23.1 23.2 Zalosh (2002), pp. 3-409–3-410.
  24. ↑ 24.0 24.1 24.2 24.3 Zalosh (2002), p. 3-420.
  25. ↑ 25.0 25.1 25.2 25.3 Barry (2002), p. 5-184.
  26. ↑ 26.0 26.1 Iqbal and Salley (2004), p. 5-15.
  27. ↑ Beyler (2002a), p. 3-303.
  28. ↑ Van Rossum (2027), §4.8.
  29. ↑ 29.0 29.1 Van Rossum (2027), §4.10.
  30. ↑ Van Rossum (2027), §4.9.
  31. ↑ 31.0 31.1 Van der Graaf et al. (2021/2023), p. 27.
  32. ↑ Van der Graaf et al. (2021/2023), p. 29.
  33. ↑ 33.0 33.1 33.2 Iqbal and Salley (2004), p. 15-7.
  34. ↑ Iqbal and Salley (2004), p. 15-8.
  35. ↑ Barry (2002), p. 5-186.
  36. ↑ Fischer et al. (1995), as cited in Iqbal and Salley (2004), p. 15-12.
  37. ↑ Iqbal and Salley (2004), p. 15-9.
  38. ↑ Iqbal and Salley (2004), p. 15-10.
  39. ↑ 39.0 39.1 Iqbal and Salley (2004), p. 15-15.
  40. ↑ Iqbal and Salley (2004), pp. B-27–B-28.
  41. ↑ 41.0 41.1 41.2 41.3 41.4 Zalosh (2002), p. 3-414.
  42. ↑ 42.0 42.1 Zalosh (2002), p. 3-417.
  43. ↑ 43.0 43.1 Zalosh (2002), p. 3-416.
  44. ↑ 44.0 44.1 Barry (2002), p. 5-176.
  45. ↑ Barry (2002), p. 5-182.
  46. ↑ Zalosh (2002), p. 3-404.
  47. ↑ Barry (2002), p. 5-188.

References

  • Barry, T. F. (2002). Quantitative risk assessment in chemical process industries. 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. 5-176–5-213). National Fire Protection Association; Society of Fire Protection Engineers.
  • Beyler, C. L. (2002a). Fire hazard calculations for large, open hydrocarbon fires. 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. 3-268–3-314). National Fire Protection Association; Society of Fire Protection Engineers.
  • 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.
  • Simmons, R. F. (2002). Premixed burning. 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. 1-144–1-154). National Fire Protection Association; Society of Fire Protection Engineers.
  • Van der Graaf, J., Hessels, T., & Rosmuller, N. (2023). Fire safety of indoor car parks accommodating electrically powered vehicles. Nederlands Instituut Publieke Veiligheid. (Original work published 2021)
  • Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.
  • 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.
Retrieved from "https://fire-safety.academy/w/index.php?title=Explosion&oldid=2683"