The dust hazard analysis

A dust hazard analysis (DHA) is a systematic review of a facility or process to identify where combustible dust fire, flash fire and explosion hazards exist, what could go wrong, what safeguards are already in place and what more is needed. It is the backbone of a combustible dust program. Lesson 5 explained that NFPA 652 made the DHA a requirement, with a deadline of 7 September 2020 for existing facilities, and that NFPA 660 carries the requirement forward. This lesson explains what a DHA covers, who should do it, how to carry it out step by step and how to keep it current.

HSE-311 covered process hazard analysis methods such as what-if analysis, checklists, HAZOP and layer of protection analysis. A DHA can use any of these methods. What makes it a DHA is its focus: it looks specifically at combustible particulate solids, it considers both the inside of process equipment and the building spaces around it, and it uses explosibility data to decide whether a hazard is credible.

What a DHA must cover

NFPA 660 describes the minimum content of a DHA. In plain terms, the analysis must identify and evaluate the fire, flash fire and explosion hazards of the materials, and it must look at each part of the process and each building compartment where combustible dust is present. For each hazard it must identify the safeguards in place and recommend additional safeguards where needed. The results must be documented. The standard does not require a particular method, which leaves room to scale the analysis to the size and complexity of the facility, but the analysis must be done or led by a qualified person.

Step 1: Gather material data

Start by listing every combustible solid material handled, generated or stored, including raw materials, intermediates, products, wastes and dust generated by cutting, sanding or grinding. For each one, collect explosibility data from testing of representative samples, as described in Lesson 3. Where test data are not available and the material is not well characterized, NFPA 660 requires the material to be either tested or assumed to be combustible. Record the particle size, moisture content and the source of the data. Also gather drawings, equipment lists, process flow diagrams, previous incident reports, inspection findings, insurance surveys and any earlier hazard studies.

Step 2: Decide whether the dust is combustible or explosible

If testing shows that a material is not combustible or not explosible in the form handled, the analysis can document that and focus elsewhere. Remember that a material can change. A material that is coarse and moist when it arrives may become fine and dry after milling or drying. The analysis must consider the material in the condition in which it actually exists at each point in the process.

Step 3: Divide the facility into systems and compartments

Break the facility into manageable parts. Process systems are groups of connected equipment, such as a receiving and storage system, a milling system, a drying system, a conveying system, a packaging line and a dust collection system. Building compartments are rooms or enclosed spaces separated from each other, such as a mill room, a packing room, a warehouse and a basement conveyor gallery. Each process system is analyzed for hazards inside the equipment, and each building compartment is analyzed for hazards from dust that escapes into it.

Step 4: Identify hazard scenarios

For each system and compartment, ask where a dust cloud could form, what could ignite it, and what would happen if it did. Inside equipment, clouds usually exist during normal operation, so the questions focus on ignition sources and on what happens when an explosion occurs: could it rupture the equipment, injure people nearby, or travel through ducts and conveyors to other equipment or into the building? In building compartments, the questions focus on how much dust escapes, how much accumulates, and whether a primary event could lift enough dust to cause a flash fire or secondary explosion. Fire hazards such as smoldering in collectors, dryers and silos must also be considered.

Step 5: Evaluate existing safeguards

For each scenario, list the safeguards already in place and judge whether they are adequate. Safeguards may prevent the event, such as spark detection and extinguishing in a duct, magnetic separators to remove tramp metal, bonding and grounding, and housekeeping. Others mitigate the consequences, such as explosion vents, suppression systems, isolation valves, chemical isolation barriers, and building separation. Compare what is in place with the requirements of NFPA 660 and related standards such as NFPA 68 and NFPA 69. Where a safeguard exists but is not inspected, tested or maintained, it should not be given full credit.

Step 6: Recommend, document and track

Where safeguards are inadequate, recommend specific actions. Good recommendations are clear, practical and assigned to an owner with a target date. Document the analysis, including the data used, the method, the team members and their qualifications, the systems and compartments reviewed, the scenarios, the safeguards and the recommendations. Track recommendations through to completion. A DHA with a long list of open recommendations provides little protection, and it may become evidence against the employer if an incident occurs.

Who should do the DHA

NFPA 660 requires the DHA to be performed or led by a qualified person. Such a person has the knowledge, training and experience to understand combustible dust hazards, the process and the relevant standards. In practice, a good DHA team includes the qualified leader, people who operate and maintain the equipment, an engineer who understands the process, and someone who knows the facility's history. Operators and maintenance staff often know where dust escapes, which machines run hot, and which safeguards are bypassed, which is information no drawing can provide.

Ranking risks

Many DHAs use a simple risk matrix to set priorities, combining an estimate of likelihood with an estimate of severity. HSE-223 covers risk matrices in detail. For combustible dust, severity is often high because explosions can kill, so the main differences between scenarios lie in the likelihood and in how many people are exposed. A quick ranking helps a facility decide what to fix first, but it is not a substitute for meeting the requirements of the applicable standards.

# Simple DHA risk ranking (illustrative 1-5 scales)
scenarios = [
    ("Bin explosion, no vent, packing room occupied", 3, 5),
    ("Collector fire spreading to building via duct", 3, 4),
    ("Secondary explosion from beam accumulations", 2, 5),
    ("Flash fire during manual bag dumping", 3, 3),
    ("Hot residue in dryer after shutdown", 2, 3),
]
ranked = sorted(scenarios, key=lambda s: s[1] * s[2], reverse=True)
for name, likelihood, severity in ranked:
    score = likelihood * severity
    level = "high" if score >= 12 else "medium" if score >= 6 else "low"
    print(f"{score:2d} {level:6s} {name}")

15 high   Bin explosion, no vent, packing room occupied
12 high   Collector fire spreading to building via duct
10 medium Secondary explosion from beam accumulations
 9 medium Flash fire during manual bag dumping
 6 medium Hot residue in dryer after shutdown

Keeping the DHA current

A DHA is a living document. NFPA 660 expects it to be reviewed and updated at least every five years. It must also be updated whenever a change could affect the dust hazard, such as a new raw material, a new piece of equipment, a change in production rate, a change in particle size, a modification to the dust collection system, or a change in building layout. This is where management of change, covered in HSE-311, links to dust safety. The 2012 BC sawmill explosions showed what can happen when the raw material changes but the hazard assessment does not.

Common weaknesses found in DHAs

Regulators and insurers often find the same weaknesses. Some DHAs analyze the equipment but ignore the building compartments. Others rely on generic data rather than testing the facility's own dust. Some are written by a consultant who never spoke to operators, so the analysis does not reflect how the plant actually runs. Recommendations may be vague, such as improve housekeeping, without saying what, where or how often. Many DHAs are completed once and never updated. The best protection against these weaknesses is to treat the DHA as a working tool that the facility owns, reviews and acts on.