On October 17, 2026, four Canadian Armed Forces members lost their lives in a fire during a training exercise at Canadian Forces Base Wainwright in Alberta. Other service members were injured. National Defence has confirmed that the fire spread rapidly in high winds and overtook a unit administrative area. Investigations into the incident are continuing. For the families, friends and colleagues of those affected, the loss is immeasurable. An occupational health and safety discussion should begin with that recognition, not with speculation about why it happened. [1][2]

There is, however, a question the tragedy brings into focus for anyone responsible for hazardous civilian work. What happens when the conditions on which a safety plan was built change faster than the people carrying out the work can respond? A plan may have been carefully prepared, reviewed and communicated. Equipment may have been checked. Workers may know their roles. Yet weather can turn, access can deteriorate, an adjacent operation can create a new exposure, or the job can change shape midway through a shift. At that point, the original assessment is a record of earlier conditions. It isn't proof that the work is still safe.

For OHS leaders, the challenge is recognizing when that point has arrived and ensuring someone has the authority, information and time to respond.

A plan is a starting position, not a permanent clearance

Consider a fictional construction site where roofers begin their morning under acceptable conditions. By early afternoon, wind is stronger, unsecured material has started to move, and the crane operator is struggling with a suspended load. The crew has completed its pre-job review, and everyone is experienced. Even so, the risk profile has changed. A pre-task assessment from seven o'clock doesn't answer whether work should continue at two.

The Canadian Centre for Occupational Health and Safety (CCOHS) advises that job safety analyses be revised when equipment, processes, materials or the environment change. Its broader risk-management guidance calls for reassessment when conditions change, not merely at the next scheduled annual review. [3][4] That is a practical distinction. The paperwork can remain complete while the underlying assumptions are already out of date.

Dynamic risk assessment does not mean rewriting an entire safety manual each time a cloud appears. It means creating a reliable pause between noticing a meaningful change and continuing the work. What has changed? Which protections still work? Which may no longer be adequate? Who is deciding whether the task can proceed, and on what evidence? The answers may require a quick field conversation or a formal reassessment, depending on the hazard. Either way, the decision must happen before exposure continues.

The hazard often changes through a chain of small shifts

Few workplaces announce the moment a routine hazard becomes an urgent one. Conditions usually drift. A forestry crew sees wind pick up while a new fire is still distant. A utility team loses its preferred access road after heavy rain. A warehouse's loading pattern changes because an unexpected delivery arrives. An outdoor event crew notices that a temporary structure is taking more wind than forecast. In each case, the team may respond to one change at a time, without appreciating what the changes mean together.

That is why a good assessment considers interaction, not just individual hazards. Increased wind may affect flame spread, visibility, equipment stability and the reliability of a planned escape route. Rain can change ground stability, vehicle traction and the time needed for emergency access. Overtime can make every other risk harder to manage because the people interpreting conditions are tired. Risk may rise abruptly even when no single observation seems decisive.

The temptation is to treat each new problem as manageable because the last one was managed. Experienced workers are particularly vulnerable to this kind of normalization. Their confidence is earned, but familiarity can make gradual deterioration feel ordinary. Supervisors need a deliberate way to break that pattern, especially where the consequences could be catastrophic.

Triggers must be agreed upon before the pressure arrives

Telling people to 'use their judgment' is not an adequate control when the environment can deteriorate rapidly. Judgment matters, but it works best inside a decision framework. Employers should identify conditions that require a pause, a new assessment, additional controls or evacuation. Examples might include a prescribed wind limit for lifting or work at height, a weather alert affecting field operations, loss of a communication link, a blocked exit, failure of a critical safeguard, or an unexpected change to simultaneous operations. The appropriate trigger depends on the work; invented universal numbers are no substitute for engineering requirements, equipment limits and a site-specific assessment.

The critical step is assigning ownership. Who monitors the trigger? Who can stop the task? Can the person closest to the hazard make the call immediately, or must they wait for approval from a manager who is elsewhere? If conditions change outside normal hours, who receives the warning? A written trigger that no one watches is unlikely to protect anyone.

WorkSafeBC's August 2026 wildfire guidance urged employers to review site-specific evacuation risks, identify when workers may need to leave, and maintain primary and alternate routes. [5] The principles are broadly useful even in workplaces far from wildfire country. A threshold should prompt action early enough for the action to matter, and it should be linked to a feasible response.

Stop-work authority has to survive contact with production pressure

Most employers say workers can stop unsafe work. The harder question is whether that authority holds when a shutdown would delay a major project, inconvenience a client or put a supervisor under scrutiny. Workers quickly learn the real rule from what happens after someone stops a job. If management treats the pause as a failure of commitment, the policy may exist only on paper.

In Alberta, workers have a legal right to refuse dangerous work where they believe there is an undue hazard, subject to the process established by provincial OHS law. [6] Other jurisdictions have their own statutory procedures. But a strong safety system should not wait until the threshold for a formal work refusal becomes the only way to raise concern. Supervisors should encourage timely reassessment whenever controls appear compromised, and workers should know exactly how to escalate a condition that is changing faster than the job plan anticipated.

Some decisions will be uncomfortable. A contractor may need to halt a lift with the crane already on site. A municipal operations crew may need to abandon a planned route. A field team may have to withdraw before a client agrees that the situation is serious. Leadership is tested in those moments because the short-term cost is visible while the harm being prevented is hypothetical.

Emergency response begins before anyone hears an alarm

Emergency plans tend to concentrate on what happens once an incident has been declared: who calls for help, which alarm sounds and where workers assemble. Those arrangements matter, but they are the second half of the problem. The first half is knowing when conditions require a move from ordinary operations into heightened readiness, controlled shutdown or evacuation.

CCOHS describes emergency planning as a process that starts with vulnerability assessment and includes credible scenarios, resources, procedures and responsibilities. It also identifies the value of phased alerts, alternate communication methods and more than one escape route where needed. [7] A robust plan therefore has to connect the early-warning picture to the emergency response. It should not leave a gap between 'we have noticed a developing hazard' and 'the emergency is now underway.'

For a remote crew, that might mean moving vehicles to maintain an outward-facing exit before deteriorating weather closes a road. For a manufacturing site, it may mean accounting for workers in a contractor-controlled area before an incident forces an evacuation. For utility workers, it could mean identifying a safe withdrawal point and a communication backup before a storm compromises radio coverage. These are examples of planning choices, not claims about what occurred at Wainwright.

Test the moment when the original plan stops working

Many emergency drills are tidy. The alarm works, the main exit is clear, the muster point is usable and the designated supervisor is present. Real emergencies are rarely so accommodating. A credible exercise should sometimes remove one of the assumptions on which the plan depends. What if the usual road is blocked? What if one crew is out of radio range? What if a subcontractor has moved its people? What if the weather warning arrives while the supervisor is dealing with another incident?

The point is not to frighten workers or manufacture a spectacular disaster scenario. It is to discover the decision points and practical obstacles while the consequences are still confined to an exercise. WorkSafeBC advises employers to train workers in emergency procedures, conduct drills and revise arrangements when circumstances change. [8] That makes review an ongoing management duty rather than an annual administrative event.

After the exercise, ask not only whether everyone reached a safe location, but how long it took to recognize the changed conditions, issue a clear instruction and confirm that every worker received it. If the scenario reveals confusion about authority or incompatible communication systems, the plan needs repair before the next drill.

What senior leaders should demand from the system

Executives should ask for more than completed hazard assessments. They need to understand whether the organization detects meaningful changes during work and whether its decisions keep pace. A useful monthly review might examine jobs paused because conditions changed, recurring failures in communication, near misses involving weather or simultaneous operations, delayed escalations and instances where evacuation or withdrawal plans proved impractical. Those aren't necessarily signs of a weak culture. An increase in reported pauses may indicate that workers are using the system as intended.

Leaders should also examine the decisions they make upstream. Tight staffing, unrealistic schedules and poorly coordinated contractors can erode the time needed to reassess risk. Buying better monitoring equipment won't solve that problem if supervisors are rewarded for continuing regardless of what the data shows.

The Wainwright investigation will determine what can properly be learned from that specific event. Civilian employers should not borrow conclusions before investigators reach them, nor assume that military operations operate under the same legal framework as civilian workplaces. But the broader OHS question is already worth asking: can our safety system recognize when its original assumptions are no longer valid, and can it change course quickly enough to protect people? The answer lies less in the quality of the form completed at the start of the day than in the decisions made after conditions begin to change.