Explosion Isolation in Energy & Power

Fuel handling is the one place where the connecting duct is the plant. A reclaim conveyor, a mill feed, a classifier return and a filter extraction add up to long runs between vessels, and a deflagration that starts anywhere in that chain has somewhere to go.

The fuel makes it worse. A plant commissioned on bituminous coal now takes wood pellets, sunflower husk, olive residue or torrefied material, often blended and rarely identical between deliveries. Each of those has its own Kst value, its own minimum ignition energy and its own dust fraction, and the isolation installed for the original fuel was certified against one set of numbers.

The ignition source is also different from most industries. Biomass self heats. A smouldering nest in a silo or a hot particle leaving the mill is not a rare abnormal event; it is the normal failure mode, and it feeds the duct directly.

A power or biomass plant changes its fuel more often than its equipment, and isolation certified once rarely covers what arrives later.

Every connection between protected vessels carries a tested isolation device within its certified range, assessed against the full fuel envelope the plant actually receives.

Detection that is appropriate to smouldering rather than only to flame, so the event is caught while it is still a hot nest and not after it has become a deflagration.

Devices that still work in February: outdoors, wet, cold, with the dust that fuel handling generates sitting on everything.

In a building materials plant the trap is the opposite assumption: the main product is inert, so the whole plant is treated as inert and the one combustible additive is missed.

The first piece of work is the fuel envelope. Not the design fuel, but the range of materials the plant is contractually and practically able to receive, with explosion characteristics measured on samples that include the fines.

Isolation is then placed where the chain can be broken into sections rather than everywhere. Long conveying runs are the awkward case: flap valves depend on flow direction and need a distance window, and a run that is too long for one device and too short for two is a layout problem rather than a device problem.

Detection is chosen for the ignition source that is actually present. Carbon monoxide monitoring in silos finds a smouldering nest hours before anything optical would, and that time is what lets the plant act with a fuel transfer rather than with a suppression system.

Then the device has to survive the environment. Fuel handling is outdoors or semi enclosed, the dust is sticky when wet, and a flap valve whose hinge has seized under a layer of damp fuel dust is a closed duct with a certificate on it.

• Fuel envelope defined and measured before devices are selected.
• Isolation placed to section the chain, with long runs treated as a layout question.
• Carbon monoxide monitoring used where smouldering is the credible ignition source.
• Devices specified for outdoor service, damp dust and winter temperatures.
• Inspection intervals set to the fouling rate, not to the certificate minimum.
• Fuel change treated as a change to the explosion protection case.

• Fuel envelope rather than design fuel: the worst credible Kst and the lowest credible ignition energy across everything the plant can receive.
• Duct length and run geometry, which in fuel handling decides between one device, two devices and a redesign.
• Self heating behaviour of the fuel, which sets the detection method as much as the explosion data does.
• Moisture and fouling, because a device that cannot move when dirty is not a device.
• Ambient conditions at the device position, including winter minimum and driving rain.
• Access for inspection on a conveying run that is often elevated and rarely convenient.