Combustible dust explosions remain one of the most persistent and underestimated threats in chemical processing. That was the starting point for "Dust to Disaster: Managing Combustible Dust Hazards," a roundtable hosted by Chemical Processing in July 2026.

The panel brought together two practitioners who approach the problem from different ends. Ashok Dastidar, vice president, dust and flammability testing and consulting services at Fauske & Associates, has spent roughly 32 years in dust and vapor flammability research and consulting. He also serves on the technical committee for NFPA 660 and chairs the ASTM E27.05 subcommittee on dust explosion test methods. Alex Noonan, applications engineering manager, industrial air filtration at Donaldson Company Inc., has nearly 13 years of experience helping customers work through dust hazard analyses (DHAs) and explosion and fire mitigation for dust collectors.

The discussion was moderated by Jonathan McPherson and Todd Smith of Kansas State University's Advanced Manufacturing and Bulk Solids Technology Center. Driven largely by audience questions, the conversation covered the issues that come up again and again on plant floors. Below are the panel's answers to those questions.

How do you get new operators to recognize a dust hazard?

McPherson opened with a question many plant managers face: how do you take someone new to the industry, who may not know dust can burn at all, and train them to spot a problem and raise it with management?

Dastidar recommended treating combustible dust like any other core safety topic. "The best way is the same onboarding process that you use for all your employees, where you're training them on forklift operation or confined space entry," he said. A short awareness course, even 30 minutes to an hour, can explain how the fire triangle and explosion pentagon come together with dust and introduce common mitigation steps such as housekeeping and electrical area classification. More advanced training can follow as employees take on more responsibility.

Noonan described one of the more effective programs he has seen: walking new hires through a condensed version of the facility's DHA. "Just trying to get them to understand and get familiar with what good looks like in the process, what acceptable looks like, so that they can then identify some of those hazards," he said. Even when a new employee can't name the specific hazard, they may recognize that something doesn't look normal.

What does an effective housekeeping program look like?

The most effective programs, Noonan said, have a defined cadence and specific checkboxes to complete each shift or each week, so day and night crews work to the same standard. Success comes when housekeeping "is built into that culture and it's really part of the daily work."

The most common failure he sees is dust accumulating where it's hard to reach: rafters, I-beams, cranes and anything hanging from the ceiling. "More facilities than not would probably have what I would assume an unacceptable level of dust buildup in those rafters," he said. "Just because it's out of sight, it's not out of mind."

"The best housekeeping program is one that actually gets executed," said Dastidar. Many facilities have a program on paper that isn't followed on the plant floor, especially when crews are busy. Checklists and auditing help close that gap. He noted that floors tend to get cleaned, "but as soon as you get above chest height, things start falling through the cracks."

He also flagged a less obvious location: the inside of electrical boxes and light fixtures. "Dust, unlike vapors and gases, has a tendency to migrate," he said. "Because it's gritty, it can work itself through seals." His advice was to open those enclosures and check.

Which NFPA and OSHA standards apply?

An audience member asked which standards guard against dust explosions. Dastidar explained that the familiar occupancy standards, including NFPA 652, NFPA 654, NFPA 61 for agricultural dust and NFPA 484 for metals, have been consolidated into a single document: NFPA 660.

"All the old standards aren't changed. Their redundancies are eliminated," he said. The former standards now appear as chapters; agricultural dust is chapter 22 and metals is chapter 23, for example. The goal is to make the requirements easier for code enforcers to apply. Some regional fire codes may still reference the older standards, but Dastidar expects new fire codes to adopt NFPA 660.

NFPA 660 is the primary document for combustible dust, but it isn't the only one. Dastidar pointed to NFPA 77 on static electricity and NFPA 70, the National Electrical Code. Noonan added NFPA 68 and NFPA 69, which govern explosion venting and explosion prevention systems, and which he works with regularly when specifying protection for dust collectors.

On the regulatory side, OSHA's Combustible Dust National Emphasis Program and the agency's general duty clause also come into play, as the panel discussed when the conversation turned to compliance.

How do you assess a material you've never handled before?

Greenfield lines and product changeovers raise a chicken-and-egg problem: you need test data to design protection, but you may not have the material yet. Dastidar noted that toll manufacturers face this constantly, switching from one chemical to another every few months.

He suggested several routes. Produce small pilot batches, perhaps running them three or four times, to generate enough powder for testing. Test a surrogate with similar chemistry or particle size. Or buy the product from a competitor that already makes it. For raw materials, he advised putting the burden on suppliers: "Make that as a component to you accepting the material from them as part of the buying order, that they have to provide the explosibility and combustibility data."

Noonan described a recent project for a new aluminum boat plant. Donaldson obtained dust from the customer's sister site for testing. If that dust came back at a Kst of 130 bar·m/s, he explained, Donaldson might design the system for a Kst of 200 or 250 to leave margin, then require the customer to verify the actual dust once production starts. "You can make assumptions, but you need to be able to verify and back those assumptions up later with good data," he said.

Dastidar explained why that margin matters. If a system is designed for a Kst of 130 and the real dust turns out to be 180, "then you've undersized all your explosion protection." Overdesigning costs more up front, but Noonan said that cost "pales in comparison to the retrofit cost." An undersized explosion vent may mean cutting a new opening in the collector and redoing the calculations.

Can you trust Kst and Pmax values on a safety data sheet?

McPherson observed that explosibility data such as Kst and Pmax now show up on safety data sheets (SDSs) far more often than they did 15 years ago. Is that good enough for design?

"I would not. I would not at all," Dastidar said. Gas and vapor explosions happen at the molecular level, but dust behavior depends on particle size distribution, particle morphology, moisture content and mixture composition. The number on an SDS might describe the material as shipped, but your process may change those properties entirely. Worse, he said, "sometimes those values are totally fictitious," pulled from an internet search for a similar material.

His recommendation: ask the supplier for the laboratory report behind the value. If the report's particle size distribution and moisture content match what you'll have in your facility, the data can be trusted. If no report exists, don't rely on the SDS.

Noonan agreed and added a point specific to dust collectors. Collectors capture the finest particles in a process, which can concentrate inside the unit. "That particle size distribution inside the dust collector might not represent the big bulk bag that you're unloading on the front end of your process," he said.

Who decides whether a facility is in compliance?

"My answer to that question would always be the AHJ, or the authority having jurisdiction," Noonan said. But the AHJ isn't always the same entity. Depending on location, it could be a local fire marshal, a building inspector or an insurer. He recently helped a customer in a town of about 6,000 people work out that their AHJ was the town fire marshal. His rule of thumb: start with whoever would fine you for being out of compliance, and bring them into the conversation early.

Noonan stressed that equipment suppliers and consultants contribute to that process but don't own it. "At the end of the day, it is the process owner that truly owns that process and the risk associated with it," he said.

Dastidar added OSHA to the list, noting it can cite facilities under the general duty clause or for housekeeping and hazard communication violations. He also argued that owners are, in a sense, an AHJ themselves, and can defend their approach with data. NFPA standards include an equivalency clause, he explained. Inspectors often treat a standard as a yes-or-no checklist, but an alternative that provides an equivalent level of risk reduction is permitted. Owners can also pursue a performance-based design that meets the standard's process safety and life safety goals without following its prescriptive solution. The catch, he said, is that "you need to have the documentation to support it."

Is grounding motors and equipment enough to control static?

Noonan said more customers are specifying fully grounded dust collectors "out of an abundance of caution," in part because it's usually inexpensive. Interest in bonding and grounding has grown noticeably over the past three or four years, he said.

To confirm a collector is grounded, Donaldson tests continuity from the filter to the grounding point, such as a collector leg or an anchor bolt. Depending on the collector, anti-static filter media may be needed to complete the ground path.

Dastidar cautioned that motor grounding has a common weak point. Vibration dampeners under motor mounts can electrically isolate the motor, so an independent ground can help drain charge. Smith noted that in those cases, a grounding strap may be needed to bridge the isolation pad.

Which processes and industries pose the biggest hazards?

Noonan distinguished between how often events happen and how severe they can be. He sees the most fires and flash fires in welding and woodworking. But some of the most energetic dusts he encounters come from emerging sectors, including thermal spray, 3D printing and battery manufacturing, where finely divided metal powders are difficult to collect and protect.

Dastidar explained the 3D printing risk. In powder bed fusion, vaporized metal condenses into extremely fine particles that collect on filters. "As those particles get smaller and smaller, specifically metals, they tend to become pyrophoric," he said. When the filter is removed and exposed to oxygen, those submicron particles can react rapidly and ignite the filter media. Insurers are increasingly raising the issue, he added.

Among traditional operations, Noonan pointed to milling and metal fabrication because of the sheer number of ignition sources. Arc gouging is especially difficult because it throws sparks across large areas, and fumes from pickled or oiled steel can still be combustible.

Dastidar said metal fabrication suffers from a false sense of security. "There's this idea in the industry that metals don't burn," he said. A glowing piece of steel doesn't burst into flame, but the same metal in finely divided form can be combustible or explosible.

Milling carries its own overlooked hazard, he added. Facilities often protect the mill itself but forget that material leaving the mill is hot. Once it lands in a hopper or bin, a spark-heated particle or the material's own self-heating can create a smoldering nest.

Should explosion vents discharge inside or outside?

"You're not allowed to vent them inside," Dastidar said of conventional explosion vents. The deflagration fireball must be directed outdoors. The alternative for indoor equipment is flameless venting, which surrounds the vent with a mesh that quenches the flame. Hot gases still escape, so the discharge area must keep people out of the jet's path.

Chemistry matters too, he warned. Materials containing sulfur or halogens such as fluorine, chlorine or bromine can generate acid gases, including HF, HCl, HBr, H2S and sulfuric acid vapor, when they burn. Those shouldn't be released into an occupied space.

So where should a dust collector go? "I will always say outside if possible," Noonan said, because outdoor installations have more mitigation options. He has seen conventional vents used indoors only once, in a facility so large that workers drove pickup trucks from one end to the other.

When an indoor collector is ducted to the outside, duct length becomes critical. Noonan said Donaldson aims to keep the vent duct within one hydraulic diameter of the vent. Beyond that, the Kst and Pmax the collector can handle drop quickly. In practice, the collector must sit right next to an exterior wall or the roof, with a vent duct of perhaps three or four feet. Otherwise, he said, flameless venting or chemical suppression are the alternatives.

Dastidar noted there can be valid reasons for indoor placement, such as protecting the quality of material that will be reclaimed. In that case, owners must choose among ducted venting, flameless venting or suppression.

How should sifters be designed, and how do you keep sensors from clogging?

One attendee described seeing open sifters handling combustible dust, as well as enclosed, nitrogen-inerted sifters whose oxygen sensors kept clogging.

For open sifters, Noonan said a hood with a capture velocity of roughly 100 to 150 feet per minute will catch nuisance dust, though some will still settle. An enclosed sifter needs far less airflow, since air is drawn only from the unit. If operators open a door to load batches, the system may need to be sized around that opening. "It can be sometimes tough to speak in general terms when we're dealing with combustible dust," he said, because process details can change both the solution and its effectiveness.

The panel offered several practical fixes for clogged sensors:

• Use a T-connection on pressure taps. Noonan said Donaldson uses a T-connection with two open ends on differential pressure sensors, which clogs less than a single port.
• Relocate the sensor. Moving an oxygen sensor to a cleaner spot can help.
• Add a filter, carefully. A sintered filter on the sensor inlet can keep dust out, but Noonan cautioned that filters can alter differential pressure readings. Follow the sensor manufacturer's recommendations.
• Pulse it with nitrogen. Dastidar suggested a brief nitrogen puff every three or four minutes to clear the sensor face, which can extend the interval between filter changes. Readings will dip momentarily after each pulse.
• Install a stagnation tube. McPherson described using a larger connection, such as a 2-inch tube, at the vessel wall so dust can settle out before reaching the smaller sensor line.

The takeaway: data, verification and ownership

Across every topic, the panel returned to a few themes. Test the dust you actually have, not the number on a data sheet. Design conservatively when data is uncertain, then verify once the process is running. Make housekeeping and training part of daily work rather than a binder on a shelf. And remember that the facility owner, not the consultant or equipment supplier, owns the risk.

As Noonan put it, doing your homework early "will save you a lot of heartache on the back end for sure."