Skip to content
From Fire Safety Academy Wiki
Revision as of 09:15, 9 November 2026 by Emma (talk | contribs) (Add SVG figures fse-explosion-isolation-v1.svg, fse-explosion-isolation-v2.svg, fse-explosion-isolation-v3.svg, fse-explosion-isolation-v4.svg (list-2 visual plan))
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Contents
  1. 1Why connected equipment needs isolation
  2. 2Classification of isolation methods
  3. 3Passive isolation devices
  4. 4Active isolation systems
  5. 5Design considerations and limits
  6. 6Relation to other explosion protection methods
  7. 7See also
  8. 8Notes
  9. 9References

Explosion isolation, also called deflagration isolation, is the strategy of preventing a Deflagration that starts in one place from passing to another place where it could injure people or cause additional damage.[1][2] In process plants, mills, spray dryers, dust collectors, blowers and vacuum pumps are regularly connected by pipes, ducts, chutes or conveyors. An explosion in one unit can travel through these connections, both upstream and downstream, and can create a much more powerful explosion in the next unit.[1] Isolation methods interrupt or weaken this flame propagation, the deflagration pressure, pressure piling and flame jet ignition between items of equipment.[3]

The methods are either passive or active. Passive devices, such as a Flame arrester, a rotary valve or a flame front diverter, work without detection. Active systems detect the explosion and then close a fast-acting valve or inject a chemical barrier into the pipe.[3][4] Every technique has its own limits of application, so the choice depends on the fuel, the type of protection needed and the layout of the plant.[3] The Fire Protection Handbook (2008) refers to NFPA 69 for isolation systems, including the installation of in-line flame arresters.[3][5]

Why connected equipment needs isolation

Simmons (2002) explains the basic mechanism with connected vessels. Burning in one vessel raises the pressure in a second, connected vessel; when the flame enters that vessel, it burns a precompressed mixture and produces a correspondingly higher pressure.[6] Ural and Garzia (2008) call this pressure piling and note that it can lead to blast and shock formation.[3] The flame also enters the next vessel as a jet, and Zalosh (2008) names flame jet ignition as one of the strong ignition sources that can lead to a detonation.[3][5]

Mill, conveying duct, dust collector and fan connected by pipes, with an explosion spreading through the ducts and isolation points marked
An explosion in a mill can travel through the ducts to a dust collector and beyond; isolation devices at the connections stop flame, pressure or both.

A pipe that carries a flammable mixture can act as a fuse. If ignition occurs at one end, the flame can run through the pipe and ignite a process unit at the other end.[3] Unless it is blocked near the point of ignition, the flame speeds up steadily as it travels.[3] Simmons (2002) notes that a flame in a duct moves faster than the burning velocity of the mixture. The reason is that the flame front is larger than the duct cross-section, typically by a factor of two or three. He adds that this distinction matters when automatic protection is designed.[6]

If the mixture is in the detonable range, the accelerating flame can undergo Deflagration-to-detonation transition. The pressures then reach about 25 times the initial pressure, and up to 50 times on an object in the flow path.[3] Explosion vents on the pipe, placed in accordance with Chapter 9 of NFPA 68, can limit pressure development and flame speed. The flame front can still travel the whole length of the pipe at moderate speed, however, so venting does not replace isolation.[3]

For dusts, flame propagation inside pipes is somewhat unpredictable.[1] Tests summarised by Ural and Garzia (2008) show that pipes of less than 100 mm (4 in.) diameter are far less likely to carry a flame than larger pipes. Dense phase pneumatic transfer, at about 3 m/s (600 ft/min) and a solids loading ratio above 30, is also much less likely to do so. Dilute phase transfer, at 11 to 18 m/s and a ratio below 15, carries flames more readily.[1] As reported by Pineau, in controlled experiments in 150 mm piping propagation occurred as rarely as one time in ten, even for dilute phase systems.[3]

Metal dusts and other dusts with a high deflagration index propagate more readily through small pipes. For vessels that are close together, lowering the explosion pressure in each vessel does not necessarily remove the need for isolation.[3] Because the outcome is hard to predict, the handbook states that isolation techniques are generally necessary. The exception is a qualified risk analysis that shows, from both probability and consequence, that the risk is acceptable.[1]

Classification of isolation methods

Ural and Garzia (2008) compare the methods by whether they isolate the deflagration, the ignition source, and the flow or pressure. They also note whether each method is active or passive and whether it suits dust, gas or both.[7] Their comparison is reproduced in the table below.

Isolation system Type Fuel type Deflagration isolation Ignition source isolation Flow (pressure) isolation
Chemical barrier Active Dust, gas Yes Yes No
Mechanical valve Active Dust, gas Yes Yes Yes
Actuated float valve Active Gas Yes Yes Yes
Actuated pinch valve Active Dust, gas Yes Yes Yes
Rotary valve Passive Dust Yes See note Yes
Flame arrester Passive Gas Yes Yes No
Flame front diverter Passive Dust No No Yes
Liquid seal Passive Gas Yes Yes No
Float valve Passive Gas Yes Yes Yes

Note: rotary valves can prevent flame front passage under certain conditions but do not always prevent the passage of burning embers. Source: Ural and Garzia (2008), p. 17-159, Table 17.8.9.[7]

The table shows the main trade-off. Barriers that let gas pass, such as flame arresters, liquid seals and chemical barriers, can stop a flame but not the pressure wave. Valves that close the pipe stop both, but they must close in time.[7]

Passive isolation devices

Passive devices include material chokes such as rotary valves, flame front diverters, passive float valves (explosion check valves), static dry flame arresters and hydraulic (liquid seal) flame arresters.[3] A flame arrester makes the gas pass through small passageways in which the flame is quenched. It is not suitable for dust or saturated vapour service.[3] A hydraulic arrester breaks the gas flow into isolated bubbles that rise through a liquid, so that no continuous channel exists for the flame. It works well when ignition occurs on the side where the bubbles reach the surface, and poorly in the opposite direction.[8]

A flame front diverter vents the deflagration at a point where the flow normally turns through 180 degrees. The fast flow caused by the explosion keeps going straight on, pushes open a hinged cover or bursts a rupture disk at the top of the diverter, and releases the flame to the atmosphere.[7] Some diverters have stopped flame propagation in tests, but in most cases they were not completely effective. Where flame passed, the deflagration was less severe and pressure piling did not occur or was less severe.[7] Bartknecht, as cited by Ural and Garzia (2008), recommends not using a diverter as the only means of isolation where the aim is to stop flame propagation. Diverters should be tested for the intended application before they are specified.[8]

Three panels showing a flame front diverter venting upward, a float valve closing under a pressure wave, and a rotary valve with a product layer above it
Passive isolation devices: a flame front diverter vents the explosion where the flow turns, a float valve is closed by the pressure wave, and a rotary valve blocks the flame with close clearances or a product layer.

A passive float valve works like a check valve. It closes at a set flow rate or pressure difference produced by the deflagration, and then stays closed until it is reset by hand; it usually works in both directions.[8] It needs a pressure wave of minimum strength, so an explosion whose pressure stays below the actuation pressure passes through. An elastomer seat can suffer at high temperature, and powder on the sealing surfaces can prevent a tight seal.[8] The valve must lie within a minimum and a maximum distance from the explosion origin. The minimum ensures that it closes before the flame arrives, and the maximum ensures that no detonation develops near the valve and that its design pressure is not exceeded.[8]

Rotary valves are common in process design to keep pressure differences between connected equipment, but they need special features to act as isolation devices.[8] One concept relies on close clearances that quench the flame, which must be measured before installation and monitored for wear. The other relies on a layer of product above the valve that blocks the flame; low-level sensors stop the valve before the hopper runs empty.[8] Even without full isolation, a rotary valve reduces the transmission of pressure considerably. After an explosion it must stop automatically, so that it does not carry glowing or burning material downstream.[8]

Other chapters of the same handbook show chokes in use. Where wood material is fine enough to explode, screw conveyors with one turn of the helix removed and rotary air locks are used as chokes for fire and explosion isolation, together with backdraft dampers.[9] In an oilseed extraction desolventiser-toaster, a rotary valve seals one tray from the next. It acts as a safety device that prevents explosive mixtures of hexane and air from forming in the upper trays.[10]

Active isolation systems

An active system succeeds only if the barrier is complete before the flame front reaches it. Ural and Garzia (2008) describe the design as a race between three processes: detecting the explosion, creating the barrier, and the travel of the flame to the barrier location.[8][11] Written as an inequality in the notation of this article (the handbook states the condition in words), the condition is:

td+tb<tf=∫0Ldxvf(x)

Symbol Description Unit
td Time from ignition to detection of the explosion s
tb Time to create a complete barrier (valve closure or agent discharge) s
tf Time for the flame front to travel from the ignition point to the barrier s
L Distance from the possible ignition location to the barrier m
vf Local flame speed in the pipe m/s

If the flame reaches the barrier location first, the isolation fails. The barrier must then be placed further downstream, provided the pipe is long enough and the extra length does not promote pressure piling or transition to detonation.[11] The flame speed has three contributions: the air velocity in the pipe before the explosion, the air velocity induced by the pressure rise upstream, and flame acceleration along the pipe.[11] Designs consider the strongest credible explosion, to set a minimum distance between ignition and barrier, and the weakest credible explosion, to make sure that it is detected.[11]

Time-distance chart with an accelerating flame path, a straight line for flame carried by conveying air, and a horizontal band for detection plus barrier time
Active isolation works only if detection and barrier creation finish before the flame arrives; flame acceleration shortens the available time.

Worked example. A dilute phase conveying line carries dust from a mill to a collector at an air velocity of 18 m/s, the upper end of the range quoted for dilute phase transfer.[1] A barrier is planned 5 m downstream of the point where a flame from the mill would enter the line. If the flame were carried only by the conveying air, it would reach the barrier after 5/18=0.278 s. For ignition at the pipe inlet, or with an optical detector there that sees the flame as it enters, detection and barrier creation together would have to take less than this time. Pressure-induced flow and flame acceleration add to the flame speed in this direction, so 0.278 s is an upper bound on the time available, and the real margin is smaller.[11][4] In the typical arrangement drawn in the handbook, the flame runs from the collector back up the conveying line, against the flow, so the conveying velocity slows the flame and gives no such bound.[11][12]

The handbook states that no scientifically valid and universally accepted method exists to predict flame speed and acceleration in industrial pipes. Manufacturers therefore rely on simplified proprietary correlations based on limited data.[4] Detection is most often based on a preset pressure or rate of pressure rise, which takes a finite and not insignificant time.[11] The detection time increases with the volume of the vessel and decreases as the deflagration index rises. A low-reactivity dust in a large vessel therefore takes much longer to detect than a reactive dust in a small one, and ignition near the pipe inlet is hard to detect in time.[4] Optical detectors on the duct near the inlet avoid this delay, but the flame must enter the duct before it is seen, and optical detection can be difficult in dust service.[4] Explosion suppression systems use similar detectors: Zalosh (2002) notes that pressure detectors are usual for dusts, while gas applications may use pressure or ultraviolet detectors, the latter being the fastest.[13]

Automatic fast-acting mechanical valves close the pipe on a signal from a detector placed between two connected units. Originally a compressed gas cylinder discharged into a piston at the top of the valve and closed the gate; the propellant is generally nitrogen at 33 to 60 barg.[4] The gas can be released by a rupture disk opened with a pyrotechnic device or by gas generators, and plant-air actuated valves also exist. Slide gate valves leave the pipe fully open in normal operation, and dampers absorb the closing forces and stop the slide from springing back.[4] The flame sensor is placed far enough upstream for the valve to close in time. Fast-closing valves have been tested successfully in detecting and intercepting detonations in progress.[4]

Duct between two vessels with a deflagration detector, a controller and either a fast-acting knife gate valve or an agent bottle creating a chemical barrier
Active isolation: a detector signals a controller, which closes a fast-acting valve driven by nitrogen at 33 to 60 barg, or discharges an agent that forms a chemical barrier.

Actuated float valves add pneumatic closure to the passive design, so that they work even when low explosion overpressures are expected. They usually act in one direction only.[4] In an actuated pinch valve, a gas charge compresses an elastomer sleeve and pinches off the flow. The light sleeve makes the valve very fast and gentle on the piping, but sustained high temperatures, as in a fire, can affect it.[4]

Chemical isolation does not block the pipe. It discharges an extinguishing agent into the pipe to form a non-flammable zone, often with an agent bottle in place of the valve in a fast-acting valve layout.[4][14] Sodium bicarbonate and ammonium dihydrogen phosphate, common fire extinguisher chemicals, have been found very effective. Low-boiling halogenated agents such as HFC-227ea can form a complete vapour block in pipes.[14]

Where the explosion pressure is high, the induced flow can move the agent cloud or dilute it. Halogenated agents can also form halogen acids in the flame and face limits because of their global warming potential.[14] The handbook stresses that chemical isolation systems must be engineered specifically for each application.[14]

Design considerations and limits

The length of unprotected pipe is a key design quantity. Zalosh (2008) lists the factors that shorten the distance needed for transition to detonation: a rough or obstructed pipe, a larger diameter, a higher initial temperature and pressure, and more turbulence or initial gas velocity.[5] Without obstructions or initial turbulence, length-to-diameter ratios above 100 are needed for transition in most hydrocarbon-air mixtures. Detonations are therefore much more likely in large piping systems than in buildings or vessels, which shows the importance of the isolation systems described in NFPA 69.[5] For a smooth 150 mm pipe this ratio corresponds to about 15 m, and for a 300 mm pipe to about 30 m; roughness, obstructions and turbulence shorten these lengths.[5]

Placement must therefore satisfy two limits at once. A barrier too close to the ignition point may not be ready in time, while one too far away may face a detonation or pressures above its rating.[8][11] For flow-actuated float valves, the vendor usually recommends the minimum and maximum distances. They depend on the explosibility of the fuel, the pipe diameter and the expected reduced explosion pressure upstream.[8]

Layout and segregation reduce the isolation task. Bouchard (2008) recommends that a Dust collector be located outdoors or in a detached room with adequate explosion vents, and that a single collector should not serve several separate processes.[15] For dryers, high-speed infrared detectors with water spray may be needed in the ducts to extinguish sparks or embers and to prevent propagation to bins or dust collectors.[16] Isolation devices also need care in service: rotary valve clearances wear, and float valve seats can be fouled by powder.[8]

Relation to other explosion protection methods

Isolation complements the methods that protect a single enclosure. Where a vessel is designed to contain a deflagration, hot combustion gases may still flow out through process openings or relief devices. They should be prevented from reaching other units, where they could start a fire or another deflagration.[17] Explosion venting limits the pressure in the vessel but releases flame, and vents on a pipe do not stop the flame from travelling along it.[3][13]

An Explosion suppression system detects an incipient explosion and discharges an agent that quenches the flame before the pressure reaches the damage threshold of the enclosure.[13] Chemical isolation applies a similar idea to a pipe. Dry chemical agents are the most common choice for suppression, and sodium bicarbonate and ammonium dihydrogen phosphate have proved very effective in chemical barriers.[1][14] Suppression avoids the discharge of flame and fuel, but installing the system and refilling and resetting it after a discharge are costly.[13] The handbook stresses that prevention reduces the likelihood of a deflagration but cannot guarantee that none will occur. Mitigation methods, isolation among them, come into play once ignition has occurred.[14]

See also

Notes

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 Ural and Garzia (2008), p. 17-157.
  2. ↑ Ural and Garzia (2008), p. 17-142.
  3. ↑ 3.00 3.01 3.02 3.03 3.04 3.05 3.06 3.07 3.08 3.09 3.10 3.11 3.12 3.13 3.14 Ural and Garzia (2008), p. 17-158.
  4. ↑ 4.00 4.01 4.02 4.03 4.04 4.05 4.06 4.07 4.08 4.09 4.10 Ural and Garzia (2008), p. 17-162.
  5. ↑ 5.0 5.1 5.2 5.3 5.4 Zalosh (2008), p. 2-100.
  6. ↑ 6.0 6.1 Simmons (2002), p. 1-147.
  7. ↑ 7.0 7.1 7.2 7.3 7.4 Ural and Garzia (2008), p. 17-159.
  8. ↑ 8.00 8.01 8.02 8.03 8.04 8.05 8.06 8.07 8.08 8.09 8.10 8.11 Ural and Garzia (2008), p. 17-160.
  9. ↑ Dungan (2008b), p. 7-10.
  10. ↑ Kingsbaker (2008), p. 9-133.
  11. ↑ 11.0 11.1 11.2 11.3 11.4 11.5 11.6 11.7 Ural and Garzia (2008), p. 17-161.
  12. ↑ Ural and Garzia (2008), p. 17-163.
  13. ↑ 13.0 13.1 13.2 13.3 Zalosh (2002), p. 3-417.
  14. ↑ 14.0 14.1 14.2 14.3 14.4 14.5 Ural and Garzia (2008), p. 17-164.
  15. ↑ Bouchard (2008), p. 8-36.
  16. ↑ Ostrowski (2008a), p. 8-91.
  17. ↑ Ural and Garzia (2008), p. 17-155.

References

  • Bouchard, J. K. (2008). Materials-Handling equipment. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 8-31–8-41). National Fire Protection Association.
  • Dungan, K. W. (2008b). Storage and handling of solid fuels. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 7-3–7-13). National Fire Protection Association.
  • Kingsbaker, C. L. (2008). Oilseed solvent extraction plants. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 9-131–9-140). National Fire Protection Association.
  • Ostrowski, R. (2008a). Industrial and commercial heat utilization equipment. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 8-71–8-92). National Fire Protection Association.
  • 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.
  • Ural, E. A., & Garzia, H. W. (2008). Explosion prevention and protection. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 17-141–17-165). 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.
  • Zalosh, R. (2008). Explosions. In A. E. Cote, C. C. Grant, J. R. Hall Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 2-93–2-106). National Fire Protection Association.