Explosion protection in process plants

Grinding a solid into a powder multiplies its surface area enormously. Reaction rate depends on available surface, so a material that will barely burn as a block can propagate a flame front through a suspended cloud in milliseconds. Sugar, flour, starch, milk powder, many pigments, most organic chemicals and a good number of metals all behave this way. Materials practically inert in consolidated form become quite hazardous once converted to powders and granules.

For a dust explosion you need five things at once: combustible dust, dispersion into a cloud within its explosible concentration range, an oxidant, an ignition source, and confinement. Remove any one and you do not get an explosion. The whole of dust explosion engineering is an argument about which of those five is cheapest and most reliable to remove in a given piece of equipment.

Confinement is what turns a flash fire into a destructive event. Inside a silo, a mixer, a mill, a bucket elevator or a dust collector, pressure has nowhere to go, so it rises until the vessel fails. This is why the risk concentrates in exactly the equipment that a powder plant is made of.

Precautions must be taken in connection with processes such as grinding, atomising, conveying, collecting, drying, screening, grading, blending, weighing and packing. Each of those operations either creates fine particles, suspends them in air, or both. A dust collector is the clearest example: its entire purpose is to concentrate dust in a moving air stream, which makes it the most likely ignition location in many plants.

Mills and grinders combine a dust cloud with high energy input and the possibility of tramp metal striking a rotor. Dryers add heat and, sometimes, self-heating material. Bucket elevators combine a vertical duct, a dust cloud and a belt that can slip and generate friction heat. Pneumatic lines transport a cloud by design and can accumulate static charge along their length.

Then there is secondary explosion, which is what usually causes the casualties. A primary event inside equipment blows settled dust off beams, ledges and floors into suspension, and that much larger cloud ignites from the first. It is why housekeeping and the elimination of horizontal ledges are genuine engineering controls, not tidiness.

Safe powder handling equipment has to be designed for two things: to remove all of what are identified as thirteen possible sources of ignition, primarily by making sure the equipment is properly grounded, and to be as easy to clean as possible. Both objectives are structural. Neither can be retrofitted with a procedure.

The practical list runs through electrostatic discharge, hot surfaces, mechanical sparks and friction, self-heating, open flame and hot work, electrical faults, lightning, stray currents and exothermic reaction. In powder plants, electrostatic discharge is the one that catches people out most often. Powder moving through a pipeline or falling into a vessel separates charge, and an ungrounded metal component in an otherwise bonded system becomes an isolated conductor capable of a spark energetic enough to ignite a cloud.

So continuity is checked and recorded, not assumed. Every conductive part — pipe sections, flexible connectors, filter cages, valve bodies, bag unloader frames — is bonded and earthed, non-conductive hoses are avoided or replaced with antistatic equivalents, and bonding straps are inspected as part of routine maintenance because that is the control most likely to have quietly gone missing.

Prevention is the first line, but for equipment where an explosible cloud is unavoidable, the design assumes ignition happens and limits the consequence. Explosion venting gives the vessel a deliberate weak point that opens at a low pressure and directs the flame front to a safe place. Where a vent cannot discharge safely — indoors, or with toxic material — suppression injects an extinguishing agent within milliseconds of detection, or the vessel is built to contain the full pressure.

Isolation is the part most often left out. Vessels in a powder plant are connected by ducts and pipes, and those connections are exactly how a local event becomes a plant-wide one. A deflagration in a dust collector can travel back along the duct into the silo it serves, where the pressure rise is far more destructive. Isolation devices in the connecting duct interrupt that path.

None of this can be specified from a general principle. The severity of an event depends on the specific dust's explosibility characteristics, which are measured in a laboratory, not estimated. Design that starts from measured dust properties for the actual material is the difference between protection that works and protection that is decorative.

Powder handling equipment operators should be involved in the process and correctly trained. That extends beyond operators to maintenance technicians, cleaning staff, and anyone who comes into contact with the equipment. The person vacuuming a spill with the wrong vacuum, or replacing an antistatic hose with an ordinary one, can undo a well-engineered system in an afternoon.

Maintenance is where most controls degrade. Bonding straps get removed and not replaced. Vent panels get painted or blocked by later pipework. Filter media gets substituted for a cheaper grade with different antistatic properties. Access doors get left with a fastener missing. A short, specific inspection routine covering exactly those items is worth more than a thick manual nobody reads.

Hot work during a shutdown deserves separate mention, because it introduces a strong ignition source into equipment that has spent years accumulating dust in places nobody can see. Cleaning before cutting, and treating internal dust deposits as fuel, belong in the permit rather than in judgement on the day.