Recently, the NFPA standards that deal with combustible dusts have been going through significant revisions. The interest in these documents has been spurred by a number of large-loss incidents in the recent past and the reaction by federal regulatory agencies. The respective technical committees have worked to develop criteria in their standards that address these hazards.

The first revision was to move the recommended area limitations when using the prescriptive dust layer depth criteria in NFPA 6541 and NFPA 6642 from the annex of each standard into the body of the standard, making them mandatory. The second was the introduction of a calculated maximum permissible mass of dust for a compartment based upon the deflagration metrics of the dust and the compartment dimensions.

THE INCLUSION OF AN ENFORCEABLE AREA LIMITATION

The 2006 edition of NFPA 654 had an accumulated dust layer criterion that read:

6.2.3.1 When separation is used to limit the fire or dust explosion hazardous area, the hazardous area shall include areas where dust accumulations exceed 1/32 in. (0.8 mm) or areas where dust clouds of hazardous concentrations exist, unless otherwise permitted by 6.2.3.2.

6.3.2.3 The requirement of 6.2.3.1 shall not apply to dust accumulations with a bulk density less than 75 lb/ft3 (1,200 kg/m3) where the allowable thickness can be prorated by the following equation:

Allowable thickness (in) =

(1/32)(75) / bulk density (lb/ft3)

The requirements in this section were essentially the same as those initially adopted in the 1997 edition. This section was used as the basis for determining where there was a hazard of propagating deflagration ("flash fire”) through the building interior, triggering the requirement to manage that hazard. The annex text directed the user to Annex D, which provided a rough calculation that was used to suppor t the recommendation that the dust accumulation should not exceed 5% of the floor area or 1,000 ft2, whichever is smaller.

Similar language and calculations were used in the 2007 edition of NFPA 6642. The Technical Committee on Wood, Paper and Cellulosic Materials, which is responsible for NFPA 664, arrived at an 1/8th-inch (3 mm) dust layer criterion as its trigger for hazard management. If one corrects for the differences in the bulk density and the net heats of combustion for the "bounding value” combustible dusts addressed in each standard, the two standards permitted similar energy density per unit area of accumulation.

These requirements in the body of the standard and the recommendations in the annex have been in each standard for a number of revision cycles, initially appearing in the 1997 edition of NFPA 654 followed by the 1998 edition of NFPA 664. Consequently, there was a 14-year time period during which loss history suggested that the language in the standard was working. There were no dust deflagration events known to the technical committees where a facility complied with the requirements of the standard yet suffered a propagating deflagration or "flash fire” through the facility interior.

During the 2013 revision cycle for NFPA 654, the area limitation was moved into the body of the text via a "Tentative Interim Amendment.” The same area limitation was incorporated into the revised text for the 2012 edition of NFPA 664. The justification for these changes was based on the permissible dust mass calculation used in the "mass method” proposed for the 2013 edition of NFPA 654.

The area limitations increase the conservatism of each standard. In 1983, FM published "Dust Explosion Propagation In Simulated Grain Conveyor Galleries.”3 This research showed that very small dust accumulat ions, smaller than those permitted by the NFPA standards, could propagate a flame front in simulated grain elevator galleries. However, the authors of that research cautioned that the results should not be extended to other compartment geometries, and a number of other variables had not been studied.

A survey of some large loss incidents suggests a different problem. Table 1 shows the reported dust layers extant immediately prior to the explosion.

Loss Incident Reported Dust Layer, in (mm) Particulate
Malden Mills 6 – 12 (150 – 300) Flock fiber
Jahn Foundry 4 – 6 (100 – 150) Phenolformaldehyde resin
Hayes-Lemmerz 2 (50) Aluminum dust
West Pharmaceuticals 3 – 5 (70 – 130) Polyethylene powder
Deltic Lumber 4 (100) Wood dust
Imperial Sugar 4 – 12 (100 – 300) Sugar
Hoeganaes-Riverton, NJ 2 – 4 (50 – 100) Iron dust

Table 1. Dust Layer Data for Recent Large-Loss Dust Explosions (Data from the author’s reconstruction project files)

These losses occurred where the dust layers are orders of magnitude thicker than that permitted by either NFPA 654 or NFPA 664, even before the area limitations were introduced. But, in each case, the thickness of the accumulated fugitive dust layer was substantially greater than the layer thicknesses permitted by the relevant standard (with the exception of those facilities falling under NFPA 614, which has no dust layer criterion).


This loss history suggests that the dust layer thickness at which a propagating deflagration or "flashfire” becomes a likely possibility is some thickness between the former prescriptive layer thickness criteria in the relevant standards and the thicknesses shown in Table 1.


PERMISSIBLE DUST MASS CALCULATION

For the 2013 revision of NFPA 654, the technical committee responsible for the standard adopted a permissible alternative to the prescriptive layer depth criterion. The alternative was based upon relations for determining the maximum permissible dust mass for a building compartment before that compartment had to be designated either an "explosion” or a "flash fire” hazard. Different equations were derived for the two different hazards.

The general form of the equation for explosion hazard is:

Where: Me = permissible dust mass, kg Pes = compartment enclosure strength, bar DLF = dynamic load factor (1.5) Cw = concentration at which Pmax is attained, kg/m3 Pmax = maximum deflagration pressure, bar A = compartment area, m2 H = compartment height, m ηD = entrainment efficiency factor (0.25)

This relation was derived from the partial volume deflagration relation in NFPA 68.5 The objective of this relation is to predict at what dust loading the building structural integrity becomes at risk. The relation yields the quantity of dust sufficient to exceed the strength of the weakest element of the building compartment enclosure, presuming a worst-case concentration throughout the compartment interior. It assumes that only a fraction (25%) of the accumulated dust is suspended from the accumulation layers.

The general form of the relation for the "flash fire” case is:

Where: Mf = permissible dust mass, kg p = probability of flame impingement, (5%) Pi = initial pressure during ASTM E 12266 test D = height of a person, 2.0 m

This relation is derived from the Annex D calculation in the 2006 edition of NFPA 654. The objective of this relation is to calculate the fraction of the building interior that will be occupied by both a deflagration flame front and a person. It also assumes that only a fraction (25%) of the accumulated dust is suspended from the accumulation layers.

Recent dust explosion losses can be reviewed to analyze these correlations. Deltic Lumber was a planer mill, running kiln-dried softwood lumber. The building was approximately 200 ft. (60 m) by 240 ft. (70 m) with a ceiling height of approximately 24 ft. (7.3 m).


From this author’s investigation of the event, the accumulated fugitive dust layer was approximately 4 in. (100 mm) deep near the planer and diminished to almost none at the opposite side of the building. Using an average of 2 in. (50 mm) for the calculations, the NFPA 654 mass method yields a permissible dust mass of 100 kg for the explosion scenario and the "flash-fire” scenario.

Converting these masses to an average accumulation depth over the building floor yields a thickness of approximately 0.08 mm (0.003 in.). These layer depths are two orders of magnitude smaller than the prescriptive layer depth criterion that has been in NFPA 654 from 1997 through 2012. The building suffered a loss of approximately 5% of its surface covering, even though it exceeded the accumulated dust load sufficient for building "explosion” by a factor of approximately 1,000. The deflagration impinged on two of the 12 employees in the building.

A second case is the explosion at Hayes-Lemmerz, Huntington, IN. The building area involved was approximately 100 ft. (30 m) by 75 ft. (23 m) with a ceiling height of approximately 24 ft. (7.3 m). From this author’s investigation of the event, the dust accumulations were reported by witnesses and survivors to be approximately 2 in. (50 mm) deep on the tops of the equipment, lights, roof support beams, ducts, and piping.

Assuming that the accumulation area was 25% of the floor area due to the nature of the facility, the NFPA 654 mass method yields a permissible dust mass of 20 kg for the explosion scenario and 13 kg for the "flash-fire” scenario. By converting these masses to an average accumulation depth for the building surfaces that were accumulating dust, one yields thicknesses of 0.1 mm (0.004 in.) for the explosion scenario and 0.07 mm (0.003 in.) for the "flash-fire” scenario. These layer depths are a full order of magnitude smaller than the prescriptive layer depth criterion that has been in NFPA 654 from 1997 through 2012. The building envelope was ruptured over approximately 10 % of its area.


Similar results are obtained when the explosions at Jahn Foundry and Malden Mills are considered. Indeed, applying the bulk density of commonly encountered particulates to the example calculation in the current Annex D of NFPA 654 and applying that volume of dust over the entire compartment area, one obtains layer depths that are on the order of 0.1 mm (0.004 in.) to 0.2 mm (0.008 in.). Again, these values are approximately an order of magnitude more stringent than the prescriptive layer depth criteria.


POSSIBLE SOURCES OF ERROR

One possible issue is the 25% entrainment fraction that is used. Possible variables are whether the entrainment fraction is uniform for all materials, whether it stays the same over time and space, and over variations in environmental conditions such as temperature and humidity. The calculations assume that the entire entrainable fraction will be entrained at the same time. The calculation assumes adiabatic conditions, but the deflagration flame will lose heat to the building structure and all the equipment housed within it.

From the above, there seems to be a case for questioning whether the validation inadvertently overlooked some factors in the course of developing these new criteria for the new dust standards.


John M. Cholin is with J.M. Cholin Consultants, Inc.

References:

  1. NFPA 654, Standard for the Prevention of Fire and Dust Explosions from the Manufacturing, Processing, and Handling of Combustible Particulate Solids, National Fire Protection Association, Quincy, MA, 2013.
  2. NFPA 664, Standard for the Prevention of Fires and Explosions in Wood Processing and Woodworking Facilities, National Fire Protection Association, Quincy, MA, 2012.
  3. Tamini, F. "Dust Explosion Propagation in Simulated Grain Conveyor Galleries,” FMRC Technical Report J.I. OF1R2.RK, Factory Mutual Research Corp., Norwood, MA,1983.
  4. NFPA 61, Standard for the Prevention of Fires and Dust Explosions in Agricultural and Food Processing Facilities, National Fire Protection Association, Quincy, MA, 2013.
  5. NFPA 68, Standard on Explosion Protection by Deflagration Venting, National Fire Protection Association, Quincy, MA, 2013.
  6. ASTM E 1226, Standard on Explosion Protection by Deflagration Venting, ASTM International, West Conshohocken, PA, 2012.