In January 2012, an explosion at the Babine Forest Products sawmill in Burns Lake, B.C. killed two workers and injured 19. Three months later a second explosion, this one at the Lakeland Mills sawmill in Prince George, B.C., killed two people and injured 22. In both cases, combustible dust from mountain pine beetle-killed wood was implicated as a contributing factor. Paul Amyotte, a chemical engineer at Dalhousie University, just published a book An Introduction to Dust Explosions. ACCN spoke with Amyotte to find out what it will take to make such tragic accidents a thing of the past.
How did you become an expert in dust explosions?
I started in 1984, about mid-way through my PhD at the Technical University of Nova Scotia, which is now part of Dalhousie. At that time the federal government was operating coal mines through the Cape Breton Development Corporation, or DEVCO. After graduating in 1986, I had some coal dust explosion research contracts with the corporation. In 1992 the Westray Mine in Pictou County, N.S. experienced just such an explosion, triggered when a leak of methane gas was ignited. All 26 workers underground were killed. I was involved in some aspects of the Westray investigation, and continue to do that kind of work today. I acted as a consultant to WorkSafeBC during its initial investigation last year.
Although triggered by the ignition of methane, the bulk of this conflagration at the Bruceton Experimental Mine in Pittsburgh, Pa. is due to exploding coal dust. Dalhousie University’s Paul Amyotte, pictured above, has written a book that explains why such dust explosions happen and how they can be prevented. Photo credit: K.L. Cashdollar
What exactly is a dust explosion?
In fire prevention, people talk about the ‘fire triangle,’ which consists of a fuel, an oxidant (usually oxygen from the air) and an ignition source. To go from a fire to an explosion, you need two additional components — hence we talk about an ‘explosion pentagon.’ The first component is mixing. Left to its own devices, solid dust will settle on surfaces. It can still burn, and I do have one chapter on dust fires, which are actually more common than dust explosions. But to have an explosion, the dust needs to burn all at once, which requires it to be well mixed with oxygen, and that only happens in a dust cloud.
The second additional component is confinement. If you have a dust cloud in the open air, it will burn very rapidly, but the pressure is typically dissipated. Confinement means you fix the volume and the ideal gas law tells us that if the volume is fixed, a rise in temperature must lead to a rise in pressure, which can be even more devastating than the fire itself. So when you have all these components of the pentagon: fuel, oxidant, ignition source, mixing and confinement, you can have a dust explosion.
Dust explosions have a long history. The first recorded account comes from 1795 and occurred in a flour warehouse in Turin, Italy. In the mid-19th century, Michael Faraday, famous for his work in electromagnetism, wrote a detailed report on an explosion in an English coal mine. He was the first to show that such explosions were not just due to ‘fire damp’ (as methane or natural gas was then called) but that the coal dust itself played a role.
In your book, you say that some people believe the myth that dust does not explode. Really?
I’ve had it expressed to me, yes. It’s true that not all dust is combustible: things like sand (silicon dioxide, SiO2) or limestone (calcium carbonate, CaCO3) are already fully oxidized, so they can’t explode. But I would guess that from 70 to 80 per cent of the powdered materials that are handled in industry are capable of exploding: cotton, wood, food and feed, coal and coal products, oil shale, petroleum coke, plastics, rubber, pharmaceutical intermediates, metals, sulphur and many more. Any time you have a fine particle size of an un-oxidized material, you should be concerned about it.
Dust explosions are an ever-present problem, but interest in them tends to be cyclical. For example, when I got started, a lot of the interest was driven by four explosions that happened at grain elevators in the United States in the late 1970s. In the past decade, a lot of work was driven by four major explosions that were investigated by the U.S. Chemical Safety Board. These involved polyethylene, phenolic resin, aluminum dust and sugar dust. In Canada, the major incidents have been the Westray explosion in 1992 and the two recent explosions in B.C.
What are the lessons learned from those accidents in B.C.?
To my knowledge the investigations are still ongoing, so I can’t say for sure. But if beetle-killed wood played a role in those explosions, it speaks to the importance of managing any changes in the material you’re handling. For example, if beetle-killed wood dust is dryer, or if its particle size is smaller, or if there is more resin involved, those are the kinds of things you need to manage in order to keep your hazard, and therefore your risk, under control.
How much dust does it take to make an explosion?
Very little. If you were in a dust cloud that was just at the lower limit of being able to explode, you would not be able to see through that cloud. That may sound like a lot, but remember that these dust clouds originate from layers on the floor and other surfaces that get lifted up into suspension. When you do the calculation to see how thick a dust layer you would need to get an opaque cloud in a room, you find that you’re talking on the order of a millimetre-thick layer. Of course, it depends on the type of dust, the minimum explosible concentration, the height of the room, etc. Regulations recommend cleaning up dust layers only 0.8 millimetres thick. A general rule is that if you can write your initials in the dust, there’s too much there.
The 2003 explosion at West Pharmaceutical Services in Kinston, N.C., was fuelled by polyethylene dust created from the polymers used to coat rubber vial stoppers. The dust accumulated above a suspended ceiling where it couldn’t be seen by workers. Thus, despite rigorous housekeeping practices, the hazard remained. Photo credit: U.S. Chemical Safety Board
One of chapters in your book discusses primary and secondary explosions; can you explain these?
Explosible dust clouds are very optically thick; as a result they don’t routinely exist in areas where people work because it makes it hard to breathe. But they do exist inside process equipment, like grinders or mixers. The sequence that has been documented in numerous incidents is that there is an explosion inside a piece of equipment sufficient enough to rupture the equipment itself. If, around that equipment, you have layers of dust, the pressure wave from that primary explosion will lift that layered dust up into the air. The flame from the primary explosion ignites a secondary explosion and that’s the one that causes the real damage. This is what happened at Westray, where a localized methane explosion created a pressure wave that lifted coal dust up in the air.
So if you can’t see any dust, you’re safe?
Not necessarily. Another example is from the early 2000s, at the West Pharmaceutical plant in North Carolina. They were making rubber strips for vial stoppers, but these were coated in polyethylene to keep the rubber from sticking together, like an anti-tacking agent. The rubber strips were dipped in a polyethylene-water slurry, then dried with fans to drive off the water. Workers knew that the polyethylene dust was a problem and actually had a fairly meticulous housekeeping program. Unfortunately, the fans carried small amounts of the polyethylene up into a sort of false ceiling, depositing a layer of polyethylene dust that workers couldn’t see. When this was triggered by a disturbance somewhere else in the plant, they had a really serious polyethylene dust explosion.
Does the size of the dust particles matter?
In general, smaller particles have greater surface area and therefore greater reactivity, so as the particles get smaller, the dust is easier to ignite, and the explosion overpressure that you’ll generate increases. A lot of our latest work focuses on getting down into the nano size, for example titanium nanoparticles. These are 1,000 times smaller than the kind of dusts we normally deal with, but fortunately, when they’re dispersed in a cloud, they very quickly agglomerate. It’s too early to say for sure, but so far it looks like explosion consequences for nano-sized powders will not be that much more severe than for micron-sized particles. That said, the explosion likelihood increases because the ignition energy is lowered, so there may be some differences there.
What can be done to prevent dust explosions?
Venting is the most commonly applied remedy. Unfortunately, it’s not a prevention measure; it really only comes into play once the explosion has happened and you’re trying to reduce the consequences. Having relief vents or explosion panels sounds simple enough, but in practice you really want to rely on people who know what they’re doing to design and size them.
Another way to have safer designs is by mixing in inert materials. For example, inert gases like nitrogen or argon can lower the oxygen concentration inside a process vessel. You can also add inert dusts to the fuel itself. That’s the most common approach in coal mines: spreading limestone or rock dust throughout the mine so that if dust started to lift from some kind of disturbance, it wouldn’t explode. One of the main conclusions from the Westray report was that this had not been done to a sufficient degree.
In a sugar or flour mill, the dust is your product, so you can’t add inert dust without contaminating it. In those cases, it’s important to have good housekeeping, cleaning up dust and keeping the concentration out of the explosible range. That’s hard to do because, as I said, the minimum concentration you need is quite low, and you don’t want to use simple air hoses, because that will just raise dust clouds, which is exactly what you’re trying to prevent. There are explosion-proof vacuums you can use. You can also make your equipment strong enough that it will withstand an explosion; that’s commonly done with hammer mills, which are used for particle size reduction, grinding everything from corn and wheat to rocks and minerals. Finally, you can use automatic suppression as well as pressure detectors and fast-acting valves to essentially isolate the equipment downstream of the explosion.
The last myth you mention is, “it won’t happen to me.” Why is this attitude so pervasive?
At one time, I would have thought that all we needed were technical solutions; after all, I’m an engineer. But all that technology I mentioned has been available for a long time, which makes it hard to explain why dust explosions keep happening. I’ve become convinced that the solutions are also in the management and the social sciences as well. In the book I write, “without the will to implement dust explosion risk reduction measures in a systematic and organized manner, all our technical solutions are for naught.”
Dust explosions are not isolated phenomena; they need to be seen in the context of an entire safety management system. Too often, people don’t report near misses, because they’re afraid that they’re going to be disciplined or fired. What we need is a safety culture, where we can learn from our mistakes and let the people best equipped to make the decisions actually make them.
Are you optimistic?
This is my life’s work, so I have to be! Anyone who works in the field of process safety knows that you’re always dealing with things that can go wrong, but you try to look for success stories wherever you can, and find effective ways to communicate them to the people who really need to hear the message. That’s the reason I wrote this book.