Introduction

There is some confusion about the hazards of dust deposits, especially surrounding installations, within buildings. It is generally known that such deposits create a hazard of secondary explosions: if there is an explosion inside equipment that is not well protected, the flame jet (in combination with the air flow and pressure wave) from this primary explosion may blow up such deposits and cause secondary dust explosions.

However, if equipment is well protected (for example explosion venting into open air, with adequate explosion isolation and no leaks), it could be argued that there is no risk for such secondary explosions and no measures regarding dust explosion prevention would be required in a dusty room. Such measures include ATEX zoning and the use of ATEX certified equipment within the zoned areas.

At first, this publication will present the current legislation and guidance with respect to zoning when there are dust deposits. Next, the hazards of dust deposits will be evaluated. Finally, the characteristics that are relevant for dust layers will be discussed.

Zoning

The European directive 1999/92/EC (ATEX 137) provides the definitions of zones in Annex I. Concerning dust layers following is stated: Layers, deposits and heaps of combustible dust must be considered as any other source which can form an explosive atmosphere.

According to this directive, therefore, it should be evaluated if it is possible that dust layers could create dust clouds. In general, for dust layers on floors, this is most unlikely. If air currents are not to be expected, it might even be excluded and no zoning would be required. For dust layers on high surfaces (on ducting, cable trays, beams), the situation is different. Especially very thick layers may easily fall down and create a dust cloud, for example due to vibrations or impact.

It is important to note that zoning, according to ATEX 137, is based on normal operation. This means that secondary explosions, which are created by primary explosions blowing up and igniting dust layers, are not taken into account. This might be confusing, because it is generally accepted that such secondary explosions are an important risk of dust layers. However, the purpose of defining zones is to allow the adequate choice of equipment that is installed in such zones, see annex IIB of ATEX 137. Since the choice of equipment in dusty areas has no effect on the probability of such secondary explosions, it does not influence the zoning.

A very important requirement in ATEX 137 is the risk assessment. This risk assessment should take into account the scale of the anticipated effects (see article 4) which of course includes the probability of secondary explosions.

The IEC standard on dust zoning (IEC 60079-10-2:2009) has exactly the same approach: dust layers need to be taken into account as possible sources of dust clouds. In annex B the hazard that dust layers are ignited is discussed.

Since such an approach (determine for each dust source, including dust layers, the likelihood that an explosive atmosphere is created) is rather complicated and very time consuming, several guidelines were developed providing a practical approach. The Dutch NPR7910-2:2010 is widely used (also outside the Netherlands). This guideline is very strict on dust layers:

• Dust layers are non-negligible if the thickness is > 0,1 mm.
• If layers (> 0,1 mm) are present for > 8 hours (uninterrupted) a zone 21 is defined. Below 8 hours there is a zone 22.

This approach is often criticized as being too conservative: obviously NPR assumes that any dust layer can easily result into a dust cloud.  However, it needs to be noted that, if one wants to prevent a complicated analysis and use simplified guidance, this usually means that such a simplified approach is conservative, to prevent it is unsafe. Since it is merely a guideline, it is of course allowed to deviate from this approach, if supported by a well-founded analysis.

The layer thickness of 0,1 mm is also sometimes considered as being very conservative. Therefore, following indicative estimation is given. Assume there is a dusty room, with a height of 3 m. The dust involved has a specific weight of 1000 kg/m³ and a LEL of 30 g/m³. If the whole floor is covered with a layer of dust of uniform thickness, a thickness of 0,09 mm is already sufficient to enable formation of a uniform dust cloud in the whole room with a concentration at the LEL. If there is frequent cleaning and there is no dust on the floor, but only on cable trays, piping and beams which have (as an example) an overall surface area of 1 % of the floor area, the required thickness to enable a dust cloud at the LEL would be 9 mm. Deposits with lower thicknesses, or local dust deposits only, might not be able to fill the whole room at LEL, but might still result in local explosive atmospheres. Since the UEL of an average dust is about 1000 times the LEL, it is most unlikely that, in case of dust layer disturbances, a dust cloud is created with a concentration beyond the UEL.

A practical problem, is that, although this NPR is only a guideline, it is sometimes imposed by inspectors. Especially in the Netherlands, SZW (labour inspectorate) requires strictly application of the NPR. Meaning a more sophisticated approach, instead of this very simple approach, is not accepted.

Hazards of dust layers

There are a number of hazards related to dusty buildings, including:

• Secondary dust explosions

A primary explosion might blow up dust deposits, causing a dust cloud which is ignited by the primary explosion.

• Maintenance involving hot work

In case of hot work activities there are usually requirements that the area should be cleaned up to X m around the work place (to prevent dust layers from being ignited). Dust layers that are located far above the work place (e.g. on cable trays), however, are often overlooked. If during the same maintenance, such deposits are very likely to fall down, a hazardous dust cloud at the work place may still be created. This could easily occur if an electrician is removing cables from a cable tray above. Therefore, for such hot work activities, it is especially important that the presence of dust deposits above the work place is taken into account in the risk analysis. The best solution would be a thorough cleaning of the overall building, but this is not always feasible.

• Deposits on hot surfaces

If surface temperatures are sufficiently high, deposits on such surfaces may start smouldering. With instable products decomposition reactions may start, which might also result into very hot, or smouldering, deposits.

• Sparks settling down in dust layers

Deposits of some dusts are rather sensitive to ignition by mechanical sparks. Also electrical or electrostatical sparks might result into smouldering or burning deposits.

• Electrical equipment

Common electrical equipment is not always dust tight. Meaning, in a dusty environment, it has to be taken into account that dust might enter the equipment and create dust deposits. In case of conductive dusts, this might cause a short circuit, which could result into a fire.  But also for non-conductive dusts there are fire hazards: due to the presence of dust, electrical equipment might become overheated (insulation of such equipment by the dust) or dust deposits might be ignited by electrical sparks.

Because of these hazards, it certainly makes sense to apply the conservative method according to NPR7910-2 and require the use of ATEX certified equipment  even if the dust layers are unlikely to cause dust clouds.

Relevant characteristics for dust layers

Most explosion characteristics (explosion limits, Pmax, Kst, MIE, MIT) refer to the hazards of dust clouds. If the considered hazard of the dust layers is the probability that dust clouds are created (under normal conditions or due to a primary explosion) these characteristics are also relevant for dust layers, as potential dust clouds. Otherwise these are not relevant. There also a number of specific characteristics that are relevant for dust layers only and which are very useful in the evaluation of the various hazards mentioned in the previous paragraph:

• Layer Ignition Temperature (LIT)

This is the temperature of a hot surface, covered with a 5 mm dust layer, that is just capable to cause an ignition (smouldering or fire) of this dust layer. It needs to be taken into account that the ignition temperature of dust layers actually depends on the thickness of the dust layers. When dust layers thicker than 5 mm are involved, a hot layer with a temperature below the LIT might still be dangerous. On the other hand, in a very clean installation where layer thicknesses are always far below 5 mm, using the LIT as a limit for maximum surface temperatures is a conservative approach.

• The combustion class (Brennzahl or BZ value)

This value defines the probability that a dust layer might be ignited by sparks. In the test it is tried to ignite a dust heap with a glowing platinum wire at 1000°C, which simulates a spark or glowing particle. The results range from BZ1 (nothing happens) up to BZ6 (a very fast inflammation of the whole heap).

• Electrical conductivity

A value that is not an explosion characteristic but certainly relevant for the hazards of dust layers is the electrical conductivity of the dust. With conductive dusts, there is an increased risk of short circuits in electrical equipment. With non-conductive dust, parts may become isolated and charged resulting into electrostatic discharges.

• Dustiness

There are many other parameters, such as particle size, specific weight, shape of the particles, stickiness of the dust, etc. These determine the probability that dust layers may cause a dust cloud. There is a new (German) characteristic that takes this into account: the Dustiness. In the test concerned, a sample of the dust is dropped in a well-controlled way and the arising dust concentration is measured.

Conclusion

Dust zoning is only intended to define the likelihood of an explosive dust-air mixture being present during normal operation, in order to enable the adequate choice of equipment in such zoned areas. In this context, dust layers should only be seen as a potential source of an explosive atmosphere. If the dust layer is not expected to be whirled up in normal operation, strictly formal, dust zoning is not required.

Apart from the hazard of dust cloud formation, there are two additional hazards related to deposits of combustible dusts:

1. If dust clouds (due to deposits) are only to be expected in case of a primary dust explosion, no zoning is required, but this needs to be taken into account in the risk analysis.
2. Dust deposits might also promote ignition sources. There are several characteristics of dust layers that help to define this ignition hazard. This also needs to be taken into account in the risk analysis.

The Dutch NPR7910-2 seems rather conservative: any dust layer thicker than 0,1 mm requires zoning. However, even with frequent and extensive cleaning, it is most unlikely that there will be no dust layers on surfaces which are hard to reach. Since such deposits may easily fall down, zoning would in fact be required.

Moreover, since common electrical equipment is generally not dust tight, such dust layers cause an increased fire hazard. It clearly makes sense to install dust tight equipment in such areas. In practice, most dust certified ATEX equipment is dust tight. Therefore, although the zoning of NPR7910-2 might be conservative, the installation of ATEX certified equipment is certainly useful, however, dust tight equipment might also be sufficient.

NPR7910-2 can be considered as a simplified tool which makes the zoning and corresponding choice of equipment a rather easy task. In some situations, the simplified approach might lead to unnecessary investments. Since NPR7910-2 is a guideline only, deviation from the guidance should be allowed, on condition that such a deviation is on well-founded arguments.