Solvay method of estimating the risk of electrostatic ignition

IEC TS 60079-32-1 “ATEX - Electrostatic hazards, guidance” gives us the best practice to avoid ignition from electrostatic sources. In Solvay we follow this guideline but, when we find that our existing practice, at a particular site for a particular chemical process in a particular piece of equipment, does not comply with that guidance, we are faced with a difficult question: what priority should we give to putting it right? To answer that question we have to analyse the risk associated with electrostatic ignition sources for the particular process and that particular piece of equipment. It depends on many factors, including the frequency of the presence of an explosive atmosphere, the Minimum Ignition Energy of the flammable gas or combustible dust and the number of people who would be exposed to the blast wave or shrapnel from an explosion or the thermal radiation from a fire. Such considerations allow us to prioritize the risk in three levels, ranging from Risk 1 (Unacceptable), through Risk 2 (Intermediate) to Risk 3 = Acceptable. This paper shows how to estimate the risk level of scenarios involving electrostatic and indeed other ignition sources using the Solvay method.


Objective
Solvay carries out Process Risk Analyses on all of its operations which involve the production or use of chemicals and of energy.Our objective is to protect our employees, our neighbors and the environment from harm.A Process Risk Analysis involves identifying potential scenarios with harmful effects, especially overpressure from explosions, thermal radiation from fires and the (eco)toxicity of any chemicals released.For each potential scenario, we try to quantify levels of the Severity of human and environmental consequences and the Probability of occurrence.We then place each scenario on our Risk Matrix, which is shown in a simplified form in Figure 1 below.From the Risk Matrix we can then deduce the level of Risk on three levels:  Risk 1 Unacceptable  Risk 2 Intermediate  Risk 3 Acceptable A Risk 1 situation for a project means the start-up is forbidden.For an existing installation a Risk Sheet is flagged up at corporate level and we have a one year period to introduce permanent means of prevention and protection to improve the situation.During this period temporary measures are applied.
We also try to improve Risk 2 situations whenever technically possible and the site and business managers will be asked to validate the start-up of a new unit or the operation of an existing unit when mitigation of a Risk 2 situation is not possible.
The overall result should be to have a Risk level of 3 (Acceptable) for all scenarios insofar as this is technically feasible and some scenarios of Risk level 2 (Intermediate) where it is not.on/off site Figure 1.Solvay Severity-Probability-Risk matrix in simplified form.

Working group
Scenarios are identified by a working group led by the plant manager for an existing installation or the project leader at the design stage of a new installation.The working group includes all of the expertise required especially:  Process Chemistry or Process Engineering  Operations  Instrumentation, Control and Automation. Process Safety

Identifying a scenario
The study carried out uses one or other of two main methods:  Safety Review on Diagrams (SRD)  Process Hazard Review (PHR) The Safety Review on Diagrams is very similar to the well-known and widely used HAZOP (HAZard and Operability) method.It involves identifying deviations such as "high", "low", "no", "reverse" to process parameters such as "level", "pressure", "temperature" or "flow rate".Operating instructions can also be examined, using guide words such as "do not", "too early" or "too late" to each one.
The Process Hazard Review identifies scenarios based on the classical dangers of the chemical industry:  Physical Explosion or Implosion with no combustion or chemical reaction  Gas phase Explosion involving the oxidation of a flammable gas such as methane mixed with air or another oxidant gas or the decomposition of an unstable gas, such as acetylene, in pure form  Dust Explosion involving the oxidation of a combustible dust suspended in air  Fire involving the oxidation of a flammable or combustible gas, liquid or solid in air  Thermal Explosion involving the generation of heat and gas by an uncontrolled chemical reaction or the decomposition of unstable material  Emission or Loss of Containment  Acute Toxicity of material released  Pollution of air, water or soil by material released  Biological hazards from, for example, legionella bacteria in water cooling towers.

Identifying a scenario
Whether the method is used is SRD or PHR, a potential scenario is built up from one or more Necessary, Sufficient and Independent (NSI) causes to an unwanted event such as the rupture of a vessel, a fire, an Unconfined Vapor Cloud Explosion (UVCE) or a loss of containment of toxic or harmful material.

Estimating the Severity level
The effects of this unwanted event are then examined in order to determine their human and environmental consequences.In the case of major hazard scenarios, the distance to irreversible and 1 % lethal effects is determined using the thresholds shown in table 1 below.The Severity level of a given scenario depends on the human and environmental consequences.The Severity of the human consequences and the Severity of the environmental consequences are estimated separately.The highest of those levels is retained for ranking the scenario on the Solvay Risk Matrix, as shown above in figure 1.

Estimating the Probability level
The level of Probability is determined from the number of Necessary, Sufficient and Independent causes and the expected Frequency of each one of these.If there is only one cause then it is easy to deduce the Probability of occurrence of the unwanted event over a period of one year.If there are two Necessary, Sufficient and Independent causes A and B then we assume in most cases that  either cause A occurs, is undetected for five days and cause B occurs during this period,  or cause B occurs, is undetected for five days and cause A occurs during this period.
In this way we can deduce the level of probability of an Unwanted Event, given the frequency of its Necessary, Sufficient and Independent causes, as shown below in figure 3.

Identifying the Necessary, Sufficient and Independent causes
The causes of a fire or explosion are usually easy to deduce from the fire triangle:  Presence of an oxidant gas (i.e.air for most scenarios)  Presence of a fuel (flammable gas or vapour or suspended dust inside explosivity limits)  Presence of an effective ignition source for the combination of fuel and oxidant gas.

Presence of an oxidant gas
In the case of a fire which occurs outside process vessels it is obvious that the presence of oxidant gas is "Given".In the case of a gas phase explosion inside a process vessel, it also given for processes which operate under air.On the other hand, certain processes operate under an inert gas such as nitrogen.In that case the presence of oxidant gas could be caused by a lack of nitrogen in the factory network or that the valves controlling the flow of nitrogen do not open when required.Typically that would be classified as "Frequent", i.e. between once every ten years and once every year.

Presence of a fuel
In the case of a leak of flammable material from process equipment, followed by a fire or UVCE the frequency of the presence of the fuel is equal to the frequency of the leak.For example a leak from a flexible hose is considered "Frequent" and a leak from flange gasket is considered "Possible".In the case of a gas phase explosion inside a process vessel containing a flammable liquid, it depends on the process temperature and the flash point.In the case of toluene (flash point 4 °C) stored at ambient temperature, it is "Given".In the case of nitrobenzene (flash point 88 °C) stored under the same conditions, the concentration of nitrobenzene in the headspace of the vessel will normally be well below the Lower Explosive Limit (LEL).But a failure of a temperature control loop might mean that the nitrobenzene is heated close to or above the flash point, such that the headspace concentration exceeds the LEL.Such a failure would typically be classified as "Frequent", i.e. between once every ten years and once every year.

Presence of an effective ignition source
Identifying effective ignition sources and assessing their Frequency is the most difficult aspect and is not treated satisfactorily in the open literature.The method we present below is that used in Solvay and is the result of 15 years of development.

Feedback from experience
Everyday experience shows that ignition frequency varies widely.For example inside the headspace of a closed drum of toluene there is a mixture of air and toluene vapor which is inside explosive limits all the time at ambient temperature.Thousands of closed drums of toluene are stored all over the world and they do not explode spontaneously.At the other extreme, at one time we operated a process involving silicon gum.This material is highly viscous and electrically insulating.It was mixed in horizontal vessels with a mechanical agitator as shown below in figure 3. Whenever the agitator of this vessel was running, spark discharges could be heard from outside the vessel.The vessel was made of steel as was the agitator.But the vessel was horizontal and operated with a level of gum above the shaft of the agitator.We think that gum tended to work its way into the bearing, so that the agitator shaft was no longer electrically bonded to the vessel.The movement of the agitator shaft on the silicone gum generated opposing electrostatic charges on the gum and the shaft.The charges on the agitator could not escape, so the electrical field built up until it approached the breakdown threshold for air (3 MV m -1 ) and a spark discharge occurred between the agitator shaft and the inside of the vessel.As the rotation continued, the electrical charges built up once more and the cycle was repeated.On one occasion a flammable liquid was fed in, leading to an explosion.
The difference in ignition frequency between the two situations is clearly at least six orders of magnitude, which is why we must attempt to assess it, even if the task is far from easy.

The importance of mixture composition
The mixture composition, i.e. the combination of oxidant gas and fuel is critical, as shown by the picture in figure 4 below and the Minimum Ignition Energies listed in table 3.This reactor contained a mixture of isopropanol and other materials at atmospheric pressure.It was initially under nitrogen.But an unexpected chemical side reaction occurred, generating oxygen and forming a mixture of isopropanol vapour and oxygen inside explosive limits.A gas phase explosion occurred inside the vessel and blew out the gasket of the manhole cover.It is estimated that the pressure reached inside the vessel was between 12 and 13 bar gauge.The reactor was designed for 6 bar gauge but it suffered no significant damage [8].As shown by the data in table 3, passing from air with 21 % volume of oxygen to pure oxygen reduces the Minimum Ignition Energy of a most fuels by a factor of around 100.That means that a gaseous mixture of pure oxygen with typical fuels such as acetone, hexane or methane is easier to ignite than a gaseous mixture of air containing 21 % oxygen with very sensitive fuels such as acetylene or hydrogen.For this reason, in our method, we consider that the frequency of ignition of a gaseous mixture of air containing over 21 % volume of oxygen with any flammable gas or vapor is "Given", whatever precautions are taken.We apply the same rule to mixtures involving chlorine, nitrous oxide etc. as there is insufficient feedback from experience to define an ignition probability.

The different types of ignition sources
In any given situation there is more than one type of ignition source to be considered.Indeed there are no less than thirteen different types to be evaluated, according to the ATEX standard EN 1127-1 "Explosive atmospheres -Explosion prevention and protection -Part 1: Basic concepts and methodology.The list given in the standard is as follows.
 To make matters worse, one of those thirteen, the electrostatic ignition sources, really needs to be assessed in terms of five quite different phenomena: That makes seventeen different ignition sources to think about.In Solvay we have risen to this challenge by defining our own method, which we believe to be unique.One of its first principles is that we cannot realistically put a precise figure on the ignition probability of each of the seventeen ignition sources in a given situation, but we do not need to: we only need to evaluate the frequency of the most likely ignition source for the combination of oxidant gas and fuel under the process conditions.

The effect of the release rate of flammable gas
The larger the area covered by a gas cloud inside explosive limits, the more likely it is to find an effective ignition source.For example, in the case of the accident of 4th January 1966 at Feyzin, an operator opened a manual valve on the bottom of spherical storage vessel containing liquefied propane gas under pressure.He wanted to run off water, which had collected inside the sphere as a lower layer.At first, what came out was indeed water, but it was followed by liquid propane and the operator could not reclose the manual valve.The propane mostly vaporized forming a dense cloud of propane vapor, which spread offsite.After about 30 minutes the cloud of vapor was ignited by a car on a minor road.The storage sphere was engulfed by fire and exploded, killing 18 people and injuring 81 others.
It is clear that in the case of a major release, leading to a cloud of flammable gas inside explosive limits, which goes outside the boundaries of our site, we have no control and we must conclude that the frequency of ignition, for example by a road vehicle, is "Given".
On the other hand, many studies reported in the literature indicate that for a release of flammable gas which stays on site the probability of ignition is clearly less than 1.The studies all confirm that the higher the release rate and the higher the area covered by the flammable gas cloud, the higher is the probability of ignition.For example the UK Energy institute [10], [11] gives the following equation for ignition probability, P, (value from 0 to 1 with no units) as a function of release rate F (kg/s).
A release rate of 58 kg/s corresponds to a probability of ignition of 0.1.The same study estimates that the ignition probability is proportional to the area covered by a flammable gas cloud.For 10,000 m 2 , the ignition probability is 0.21.On that basis, the ignition probability would be 0.1 for a release over an area of 4,800 m 2 , which corresponds to a circle with a radius of 60 m.This is applicable to many scenarios of release of flammable materials that stay on a production site.
In most if not all of our sites, employees park their cars outside the site fence and traffic onsite is restricted as far as possible.In our method, for a release to area inside site boundaries, but where non ATEX equipment is to be expected, we propose an ignition Frequency of "Very Frequent", which corresponds to an ignition probability of 0.1.

Ignition frequency to be associated with ATEX equipment
The various ATEX standards specify Equipment Protection Level different types of equipment as a function of the type of zone where they may be installed, as shown in table 4 below.The Equipment Protection Levels are defined so that the higher the frequency of an explosive atmosphere the lower must be the frequency of occurrence of an effective ignition source.We can use this idea in the case of explosive atmospheres in and around process equipment.The table shows that, when using the Solvay method, ignition with equipment specified for zone 2 or 22 (Gc or Dc respectively) is "Frequent", compared with "Very Frequent" for ordinary equipment items found outside ATEX zones.Ignition with equipment designed for zone 1 or zone 21 (Gb or Db respectively) is "Possible".Finally, Ignition with equipment designed for zone 0 or zone 20 (Ga or Da respectively) is "Improbable".
Note that the ignition frequency is not a function of the zone classification but of the specification of the equipment which is installed inside it and which is reached by the explosive atmosphere.

Summary of the method for non-electrostatic ignition sources
The first question we ask is whether we are talking about air with up to 21% volume oxygen or other oxidant gases (air enriched in oxygen, chlorine, nitrous oxide etc.).For all oxidant gases other than air with up to 21% volume oxygen, the ignition frequency is "Given".
The second question is whether the fuel is a gas from IEC group IIC (hydrogen, acetylene, etc.) or a gas from IEC groups IIA or IIB (methane and other common hydrocarbons, ethanol etc.) or indeed a combustible dust in suspension.Hydrogen, acetylene and other gases in group IIC are very easy to ignite.If all equipment present is specified for these gases and all other relevant precautions have been taken to avoid ignition, then we consider that the ignition frequency is "Very Frequent".Otherwise, the ignition frequency for these gases is "Given".Now if we are talking about gases from IEC gas groups IIA or IIB or suspensions of combustible dust in air with up to 21% volume oxygen, we consider four levels of ignition frequency:  "Given" if there is at least one "Given" ignition source identified  "Very Frequent" if there is at least one "Very Frequent" ignition source identified  "Frequent" if there is at least one "Frequent" ignition source identified  "Possible" if all good practice for handling flammables is followed and all equipment installed complies with zone 1 or zone 21.
This is shown as a logic diagram in figure 5 below.

Rating the frequency of electrostatic ignition sources
The last piece in the jigsaw is how we rate the frequency of each of the five electrostatic ignition sources.Appendix A gives tables 6 to 10 showing electrostatic ignition sources found inside process vessels in order of decreasing frequency using the Solvay method.Appendix B gives tables 11 to 13 for electrostatic ignition sources found outside process vessels.The tables only apply to IEC group IIA or IIB gasses and combustible dusts in air with up to 21 % volume of oxygen.They do not include corona discharges, which are not incendive to such mixtures.The logic used to determine ignition frequency (Solvay terminology) or probability is shown below in table 5. Brush discharge inside a vessel where an insulating liquid is filled at speed > limit Frequent P = 0.01 All good practice has been followed especially IEC 60079-32-1 Brush discharge inside a vessel where an insulating liquid is filled at speed < limit Possible P = 0.001

Explosion inside a toluene tank which is normally under nitrogen
Toluene (flash point 4 °C) is stored in a 100 m 3 atmospheric tank under nitrogen.During normal operation toluene is pumped to other vessels as required.Nitrogen gas is supplied to the tank as required to make sure that the pressure does not go below atmospheric.The tank also has a PSV which opens in case of under pressure or overpressure.Periodically, the tank is filled with toluene from a road tanker.An obvious scenario is that if the nitrogen supply is lost or the nitrogen valves fail closed then as toluene is pumped out air is drawn in via the PSV and the headspace inside the vessel will contain a gaseous mixture inside explosive limits.This is not an issue so long as there is no ignition source.The flow rate of toluene into the tank complies with that specified in IEC 60079-32-1, such that a brush discharge inside the tank is not to be expected.Now suppose that, at the same time as the problem of nitrogen supply, a short length of the toluene inlet pipe, inside the headspace of the tank, is not grounded and bonded.When toluene is filled from a road tanker, toluene flows through this short length of pipe.The flow of toluene generates equal and opposite electrostatic charges on the toluene and the short length of pipe.The charges on the pipe are not evacuated so they build up, generating an electrical field.When the strength of the electrical field reaches 3 MV m -1 a spark discharge occurs between the short length of pipe and the shell of the tank etc.The MIE of toluene is 0.24 mJ [9].The energy of the spark discharge may easily reach a few mJ, so it is incendive to the mixture of toluene vapor and air inside the headspace of the tank.There will be a gas phase explosion inside the tank which may reach a pressure of 8 to 10 bar gauge, resulting in the rupture of the tank and a spillage of the toluene, which may catch fire.The distance corresponding to 1 % lethal effects would be around 20 m.For a typical tank farm, the number of people inside that distance would be less than 10, corresponding to a Severity level of H. Three causes are Necessary, Sufficient and Independent: 1. Problem of nitrogen supply from the factory network or with the valves on the nitrogen line, 2. Presence of toluene vapor inside explosive limits, 3. Loss of bonding of a fixed metal part.
The first is the failure of a dynamic system and is considered "Frequent", meaning between once every ten years and once every year.
The second is "Given".
The third is the failure of a static system and is considered "Possible", meaning between once every thousand years and once every ten years.
The combination of one "Frequent" and one "Possible" cause leads to a Probability, over a one year period, of 10 -4 , corresponding to level 3 on the Solvay Risk matrix.
The combination of Severity H with Probability 3 corresponds to Risk level 3 ("acceptable") on the Solvay Risk Matrix and indeed the storage of a flammable liquid under nitrogen is considered best practice in the chemical industry.

Explosion inside a toluene tank which is no longer under nitrogen
Now suppose that major alterations are to be carried out on the nitrogen network.They are expected to last a few weeks and during that time, nitrogen cannot be supplied to the tank.The risk of a gas phase explosion is clearly increased, but will it be level 2 (intermediate) or level 1 (unacceptable)?
With the nitrogen supply disabled, two of the causes are "Given".Now only one cause is required: the loss of effective bonding of a short piece of piping inside the tank, which has a frequency of "Possible"., meaning between once every 1000 years and once every 10 years This corresponds to a Probability, over a one year period of 10 -2 , corresponding to level 2 on the Solvay Risk matrix.
The combination of Severity H with Probability 2 corresponds to Risk level 2 ("intermediate") on the Solvay Risk Matrix and although the storage of a flammable liquid under air is not considered best practice in the chemical industry, it can be found on many installations.

Explosion inside a dust filter
A dust filter is used to filter air from the ventilation network of a powder handling and packaging station, as shown in figure 6 below.Dirty air from the packaging station arrives in the filter and clean air passes through the filter membrane then through a ventilator to the atmosphere.Dust falls to the bottom of the filter and passes through a rotary valve to a super-sack.As the powder can clog the filter membrane, there is an air-blow system to remove it.The filter membrane is supported on steel cylinders which move when the air blow system operates to shake off the dust.The characteristics of the powder are as follows.
 Resistivity 5 X 10 10 Ω.m,  MIE = 3 to 10 mJ after grinding  LEL = 30 g/m 3   The concentration of the air from the bagging station is only a small fraction of LEL.Nevertheless the danger of a dust explosion inside the filter must be considered.
A classic scenario is that there is a fault in the grounding of one of the steel cylinders supporting the filter membrane.A grounding strap is missing.So the steel is not bonded to the rest of the filter.The movement of the powder leads the formation of equal and opposite electrostatic charges on the powder and the metal cylinders.The charges on the steel cylinder are not evacuated so they build up, generating an electrical field.When the strength of the electrical field reaches 3 MV m -1 a spark discharge occurs between the steel cylinder and the body of the filter.The MIE of the powder is 3 to 10 mJ.The energy of the spark discharge may easily reached a few mJ and it is incendive to the suspension of dust inside the filter.There will be a dust explosion inside the filter which may reach a pressure of 8 to 10 bar gauge, resulting in the rupture of the filter.The distance corresponding to 1 % lethal effects would be around 10 m.The first cause is "Given".
The second cause is present for around 10 % of the time, corresponding to the operation of the airblow system.It is considered "Very Frequent".
The third cause is the failure or the absence of the grounding strap of one of the steel cylinders.These items are mobile, increasing the likelihood of failure of their grounding straps.This failure is considered "Frequent", meaning between once every ten years and once every year.
The combination of one "Very Frequent" and one "Frequent" cause leads to a Probability, over a one year period, of 10 -2 , corresponding to level 2 on the Solvay Risk matrix.The combination of Severity H with Probability 2 corresponds to Risk level 2 ("Intermediate") on the Solvay Risk Matrix.

Grounding strap missing
Although many such filters operate without problems for a number of years, it is clearly desirable to use safety measures such as a correctly sized rupture panel on the filter and a valve designed to prevent the transmission of the flame to the process area.Such safeguards, if they are properly designed, inspected and maintained, are considered to reduce the probability by a factor of 100.In this case that would give a Probability, over a one year period, of 10 -4 , corresponding to level 3 on the Solvay Risk matrix.The combination of Severity H with Probability 3 corresponds to Risk level 3 ("acceptable") on the Solvay Risk Matrix.

Flash fire whilst charging a combustible dust to a hopper
The following incident occurred in 1994 in a Rhône-Poulenc facility in France.The formulation of a pharmaceutical product involved mixing the active ingredient (an antibiotic) with starch.The starch and the antibiotic were supplied in 25 kg cardboard drums with plastic liners.The starch and the antibiotic were charged to a hopper via a charging chute, before passing the mixture to a special mixer known as a "turbosphere", as shown below in figure 7.Both the hopper and the charging chute were made of stainless steel.The charging chute had a stainless steel grid to stop plastic liners falling into the hopper.The hopper was under air.This process had been operated without incident for some years.After the incident it was noticed that the metal grid in the charging chute had no grounding line and was not bonded to the charging chute.So on starting a campaign with clean equipment, the contact of the metal grid with the metal charging chute would be sufficient to ensure effective grounding, but this natural grounding could not be expected to last indefinitely, because insulating solids would build up around the grid inside the charging chute.
As solid passed over the grid, equal and opposite electrostatic charges are expected to be generated.That is to say, charge of one sign would stay on the solid and opposing charges would be go to the metal grid.When the installation was perfectly clean, the charges on the solid were evacuated.But one metal to metal contact had been lost, they would build up on the grid.This would generate a localized electrical field, whose intensity would increase until it reached the breakdown threshold of air (3 MV m -1 ), giving rise to a spark discharge between the grid and the charging chute.
Three causes are Necessary, Sufficient and Independent: 1. Presence of air.
2. Presence of a suspension of dust inside explosive limits.
3. Loss of bonding of a metal part.
The first cause is "Given".
The second cause is present for around 10 % of the time, corresponding to the solids charging operation.It is considered "Very Frequent".
The third cause is the failure of a grounding of the steel grid, which is not fixed and has no grounding or bonding.It is considered "Very Frequent".
The combination of two "Very Frequent" causes leads to a Probability, over a one year period, of 10 -1 , corresponding to level 1-2 on the Solvay Risk matrix.
The combination of Severity H with Probability 1-2 corresponds to Risk level 1 ("Unacceptable") on the Solvay Risk Matrix.

Conclusion
Roughly half of the scenarios identified in a Process Risk Analysis of a chemical plant are fires or gas phase explosions or dust explosions.All of these involve ignition sources.Over the last 15 years Solvay has developed a method of estimating ignition frequency including electrostatic ignition sources.This allows us to estimate the level of risk in any given situation and so prioritize the action to be taken to reduce the risk.The method takes into account both gases and dusts and allows us to adjust ignition frequency to take account of the characteristics of the materials encountered.If we apply the method to typical installations we find the following:  Real incidents correspond to Risk 1 (unacceptable) situations  Substandard practice corresponds to Risk 2 (intermediate) situations  Best practice corresponds to Risk 3 (acceptable) situations.Crystallisation of a solid from an insulating solvent (hexane) in a glass lined steel vessel [7].

APPENDIX A Electrostatic ignition sources and their frequency inside vessels
Cone discharge E = 10 to 100 mJ as a function of vessel diameter and median particle size    To qualify for a frequency of P, the volume of the sample pot must not exceed 1 litre.Also, care must be taken not to rub it with a dry cloth.Otherwise it can become electrically charged and give a brush discharge when it gets near to the surface of a conductive liquid or to the metal surrounding a manhole.It should be cleaned using a damp cloth and left to dry.Filling of a vessel with a conductive liquid Filling from top or bottom.No velocity limit so long as a continuous stream of liquid is not formed.In practice we have run processes with injection of conducting liquids at up to 40 m/s to the headspace of vessels with an atmosphere within explosive limits.No issue with protruding metal.Filling of a vessel or a container with a liquid Filling from top or bottom Velocity < limit IEC 60079-32-1.

Filling of a vessel with a dissipative or conductive powder
Resistivity between 10 8 and 10 To justify the frequency P, the operator should walk on a clean conductive floor and wear dissipative shoes and gloves to ensure that charges are dissipated from the antistatic liners to ground.

Filling or emptying of insulating powder from type C super sack
The super sack should be grounded to ensure that charges are dissipated.

Propagating brush discharge E ≤ 1000 mJ Filling a type A supersack
For the following materials, following tests, the frequency of ignition is considered to Possible (P):  Hydroquinone  Pyrocatechol  Paracetamol.For other insulating powders the frequency is higher (F).Emptying a type A supersack For the following materials, following tests, the frequency of ignition is considered to Possible (P) so long as the relative humidity is controlled at > 50%.
 Hydroquinone  Pyrocatechol  Paracetamol.N.B.For the above named materials, the ignition frequency is higher (F) if the humidity is not controlled.For other insulating powders the frequency is even higher (VF).

Figure 3 .
Figure 3. Mixer for silicone gum (shown in yellow)

Figure 4 .
Figure 4. Blown flange gasket of a reactor vessel Hot surfaces  Flames, hot gases and incandescent particles  Mechanical sparks  Adiabatic compression and shock waves  Exothermic reactions, including autoignition of powders  Electrical equipment  Stray electric currents  Lightning  Radiofrequency waves : 10 4 Hz to 3 × 10 11 Hz  Infrared, visible and ultraviolet light 3 × 10 11 Hz to 3 × 10 15 Hz  Ionizing radiation  Ultrasound  Electrostatic ignition sources

Figure 5 .
Figure 5. Logic diagram to determine the Frequency of Ignition

Figure 6 .
Figure 6.Dust filter Three causes are Necessary, Sufficient and Independent: 1. Presence of air.2. Presence of a suspension of dust inside explosive limits.3. Loss of bonding of a mobile metal part.

Figure 7 .
Figure 7. Charging of solids to a hopper On the day of the incident, during the charging of the antibiotic, a flash fire occurred and the operator suffered serious burns.The properties of the antibiotic were as follows:  MIE < 1 mJ  Resistivity 10 13 Ω.m.  Median particle size 25 µm

Filling
of a vessel with an insulating powder Powder having resistivity > 10 12 Ω.m.Transfer rate 1-9 m/s.Powder having resistivity between 10 9 and 10 12 Ω.m and transfer rate > 10 m/s.

Filling of a vessel
with an insulating powder Powder having resistivity > 10 12 Ω.m.Transfer rate ≤ 1 m/s.Powder having resistivity between 10 9 and 10 12 Ω.m and transfer rate 1-9 m/s.

Table 1 .
Threshold values used for determining the effects distances of potential scenarios For example, in the case of a reactor explosion, giving rise to a blast wave, the Severity of the human consequences depends on the number of people who are inside the zone of 1 % lethal effects (140 mbar): 2 S etc.), exposure time and jurisdiction With this information, the number of people who are potentially affected can be determined.

Table 2 .
Frequency of causes -5   10 -3 Complete rupture of 25 mm diameter pipe or nozzle

Table 3 .
[9]imum Ignition Energy of some fuels in air and in pure oxygen[9].

Table 4 .
ATEX zones and Equipment Protection Levels.

Table 5 .
Logic used to determine ignition frequency or probability

Table 6 .
Electrostatic discharges occurring inside of vessels whose frequency of ignition is Given for gases of NFPA groups C & D or IEC/ATEX groups IIA & IIB or for combustible dust in air (≤21% O2)

Table 7 .
Electrostatic discharges occurring inside of vessels whose frequency of ignition is Very Frequent (VF) for gases of NFPA groups C & D or IEC/ATEX groups IIA & IIB or for combustible dust in air (≤21% O2)

Table 8 .
Electrostatic discharges occurring inside of vessels whose frequency of ignition is Frequent (F) for gases of NFPA groups C & D or IEC/ATEX groups IIA & IIB or for combustible dust in air (≤21% O2)

Table 9 .
Electrostatic discharges occurring inside of vessels whose frequency of ignition is Possible (P) for gases of NFPA groups C & D or IEC/ATEX groups IIA & IIB or for combustible dust in air (≤21% O2)

Table 10 .
Electrostatic discharges occurring inside of vessels whose frequency of ignition is Possible (P) for gases of NFPA groups C & D or IEC/ATEX groups IIA & IIB or for combustible dust in air (≤21% O2) 21APPENDIX B

Electrostatic ignition sources and their frequency outside vesselsTable 11 .
Electrostatic discharges occurring outside of vessels whose frequency of ignition is "Very Frequent (VF)" for gases of NFPA groups C & D or IEC/ATEX groups IIA & IIB or for combustible dust in air (≤21% O2)

Table 12 .
Electrostatic discharges occurring outside of vessels whose frequency of ignition is "Frequent (F)" for gases of NFPA groups C & D or IEC/ATEX groups IIA & IIB or for combustible dust in air (≤21% O2) rail tank car is grounded via the rails.Frequency F in the event of no other grounding, for example in a case where the operator has not connected the grounding line.If the grounding is not effective, a discharge will occur between the drum or the super sack and a nearby metal object or an operator.The grounding error frequency may be taken to be between 1/100 and 1/1000 times the frequency of the operation, which corresponds to F in most cases.Non-conductive pipe diameter > 3 cm The frequency is "Possible (P)" if the plastic items are unlikely to be rubbed by operators.

Table 13 .
Electrostatic discharges occurring outside of vessels whose frequency of ignition is Possible (P) for gases of NFPA groups C & D or IEC/ATEX groups IIA & IIB or for combustible dust in air (≤21% O2) Operator To justify the frequency P, the operator should wear dissipative shoes and walk on a clean conductive floor.The shoes and flooring should ensure that the electrical resistance between a metal object held in the hand of an operator and ground is between 10 5 Ω and 10 8 Ω.The resistance should be tested at least once per year and the shoes replaced as necessary.Unprotected concrete is sufficiently conductive.Anti-acid coatings should be examined on a case by case basis.frequency P, the operator should wear clothing with an outer layer made of ≥ 65% cotton, or which complies with a recognized standard for antistatic clothing, such as EN 1149.Clothing should not be removed in explosive zones.