J M Smallwood

Electrostatic Solutions Ltd

Abstract

Current practice assesses the electrostatic discharge ignition risk of a dust cloud by measurement of ignition sensitivity with a capacitive discharge circuit, expressed as the Minimum Ignition Energy. This implies that at lower discharge energies ignitions may not be obtained. A review of the literature demonstrates that this is misleading and ignition at lower energies may be obtainable series resistance or inductance to the circuit. Some workers have indicated a reduction in discharge energy required for ignition to 1% or less of the zero ohm energy value, when 10-100 kohm series resistance is inserted. Conversely, addition of series resistance reduces the incendivity of a capacitive discharge ignition (CDI) spark to gas mixtures.

Discharges from insulating surfaces are known to be less incendive to gas mixtures than CDI sparks due to their different space time characteristics. Nevertheless it is current practice to use gas mixtures to evaluate ignition risk from insulating surfaces for dust cloud hazard assessment.

There is a trend in modern practice to use static dissipative or conductive materials in the construction of items like FIBCs aimed at prevention of ESD ignition. These materials may have intermediate effective resistance values and could act like a capacitive discharge circuit with series resistance. The incendivity of a discharge from this type of circuit can be expected to be reduced for gas ignition but increased for dust cloud ignition.

This paper argues that conventional MIE values should be used with caution where a discharge may occur from an insulating or intermediate resistance source, as the incendivity of discharges from such materials to dust clouds has been inadequately studied. Where incendivity has been studied it has normally been using gas probe techniques, although gas mixtures have very different ignition characteristic compared to dust clouds. It is possible that the risk of dust cloud ignitions from insulators or materials of intermediate conductivity could be significantly underestimated using current MIE and incendivity evaluation methods.

The risk of ignition of a flammable mixture by electrostatic discharge (ESD) can be considered dependent on two main factors, the spark ignition sensitivity of the flammable mixture, and the incendivity of the discharge (Gibson & Harper 1988). The idea is that if the incendivity of the spark exceeds the ignition sensitivity of the mixture, ignition is likely to occur. Incendivity and ignition sensitivity are usually described in terms of energy. The sensitivity of the mixture is expressed as the Minimum Ignition Energy (MIE).

Minimum ignition energy and discharge incendivity

In practice MIE is usually measured using a capacitive discharge ignition (CDI) circuit (Lewis & von Elbe 1951, Glor 1988). A typical example is the Hartman tube apparatus for measurement of MIE of dust clouds (British Standards Institution 1991). A capacitor is charged to a certain voltage and on breakdown of the spark gap, the stored electrical energy is dissipated in the spark. The MIE is taken as the 0.5 C V² energy stored in the capacitor prior to discharge. In practice no spark circuit has pure capacitance, and series resistance and inductance are also present. These are normally intentionally low, and the discharge waveform produced has high peak current (tens or hundreds of amps) and a damped oscillatory form.

Some workers have studied the effects of CDI circuit resistance and inductance on ignition of certain flammable materials. The effect of adding series resistance to the circuit is to increase discharge duration (Rose & Priede 1959) and the reduce rate of energy dissipation. A significant amount of the stored energy is dissipated in the series resistance rather than the spark. For ignition of gas mixtures, adding series resistance increases the stored energy required for ignition, and hence the MIE value obtained.

In contrast, the effect of added series resistance on ignition of dust clouds can be to significantly reduce the stored energy required for ignition of dust clouds (Boyle & Llewellyn 1950). The MIE for granular aluminium and magnesium both decreased by a factor of about ten when series resistance of between 10⁴ and 10⁵ W were included in the circuit. Similar results were obtained by Line et al (1959) for lycopodium dust clouds. These results have been attributed to the ejection of dust particles from the ignition zone by the shock wave generated by the discharge, an effect which is reduced when the discharge is reduced in amplitude and lengthened in duration.

Inductance >1mH in the discharge circuit can also have the effect of reducing MIE (Line et al 1959, Glor 1988). The inductance also reduces the peak current and increases discharge duration, and can reduce the MIE measured for a dust cloud by more than a factor of ten.

The discharge from an insulating material or highly resistive material is rather different from a discharge from a metal object or CDI circuit (Gibson & Harper 1988, Gibson & Lloyd 1965). A brush discharge occurs, which is known to be less incendive to gas mixtures than the capacitive discharge. A conventional CDI circuit does not simulate this type of discharge well. Gibson & Lloyd 1965 proposed that the incendivity brush discharges should be characterised by an "equivalent energy", measured as the MIE of a gas mixture which can just be ignited by the discharge. They recognised that the difference in incendivity was due to the differences in spatial and temporal characteristics of brush and capacitive discharges.

Safety issues in current ignition test practice

Although the complex nature of ignition by electrostatic discharges has been recognised by many authors, some well established practices have arisen in ESD ignition hazard testing, which now should be re-examined as they could seriously erode the safety margins in operational circumstances. To summarise the factors giving rise to this situation;

• MIE of dust clouds and gas mixtures are generally measured using capacitive discharges with low circuit resistance and inductance.
• This type of CDI spark gives optimum ignition for gas mixtures. However, for dust clouds a factor of ten or more reduction of MIE can be obtained by altering the space-time characteristics of the discharge using added series resistance or inductance.
• Discharges from insulating surfaces have space-time characteristics which are not optimal and give reduced incendivity for gas ignition. Nevertheless it is current practice to evaluate their incendivity using ignition of gas mixtures.
• Incendivities of discharges from insulators evaluated using gas mixtures are then applied in assessment of ignition risk to dust clouds.
• There is little or no information available as to whether discharges from insulators are more, or less, incendive to dust clouds than CDI discharges. Although gas mixtures are less easily ignited, there is no reason to expect this is true of dust clouds.

We must conclude that incendivity data obtained using ignition of gas mixtures, combined with dust cloud MIE data obtained using CD discharges, could underestimate the ignition risk of discharges from insulators.

A typical example is in the current practice of investigating ignition risk in FIBCs using a gas probe. The discharge from the FIBC surface is made to pass through a flammable gas of known MIE. A judgement is made of the ignition risk arising from the FIBC design based on whether or not ignition can be made to occur.

A second cause for concern is the use of resistive materials in static dissipative equipment designed for ignition hazard avoidance. An example is a carbon fibre matrix woven into a cloth. In electrical circuit terms, a carbon fibre matrix of intermediate conductivity could be expected to resemble a CDI circuit with series resistance. Again, current practice would be to evaluate ignition risk using a flammable gas mixture, although a reduced incendivity to gas mixtures could be expected. In contrast, an order of magnitude increase in the incendivity of the discharge to a dust cloud could be expected.

Again we must conclude that incendivity data obtained using ignition of gas mixtures, combined with dust cloud MIE data obtained using CD discharges, could underestimate by an order of magnitude the ignition risk due to discharges from materials of intermediate resistance.

1. Conclusions

The MIE of flammable mixtures is usually measured as the stored energy in a capacitive discharge circuit which leads to ignition of the mixture. The circuit usually has minimal and undefined series resistance or inductance. This type of discharge has been shown to have high incendivity to gas mixtures. However, inclusion of 10⁴-10⁵ ohms series resistance, or >1mH inductance can increase the incendivity to dust clouds by an order of magnitude but reduces the incendivity to gas mixtures due to changes in the discharge space time characteristics.

Discharges from insulating surfaces, or conductors of intermediate resistivity, have different space time characteristics than CD discharges used in MIE test. Their incendivity to dust clouds has been little studied, and is usually evaluated using gas ignition. While the incendivity of these discharges to gas mixtures is reduced, their incendivity to dust clouds compared with CDI spark source is uncertain.

Thus CDI circuits give ESD in a form which ignites gas mixtures more easily than discharges from insulators or objects of intermediate resistance. However we have no reason to expect the same behaviour for dust cloud ignition.

It is therefore unsafe to rely on ignition of gas mixtures as an indicator of ignition risk to dust clouds from discharges from insulators or conductors of intermediate resistance. This concern is exacerbated by the measurement of MIE of dust clouds using discharges from CDI circuits with little series resistance or inductance, which are not optimised for incendivity to dust clouds.