For much of the past decade, safety improvement in the steel industry followed a familiar pattern. Injury frequency declined gradually, reporting frameworks became more structured, and safety management systems matured. Yet despite this progress, fatalities continued to surface year after year — often linked to the same high-energy hazards. Improvement was visible, but it rarely translated into confidence that the most severe risks were truly under control.
The data from 2024 breaks that pattern.
Not because steel has become a low-risk industry — it has not — but because for the first time, improvements in injury frequency, severity, and fatality outcomes aligned within a single year. That alignment is rare in heavy industry. More importantly, it signals a structural shift in how risk is being understood and managed across steel operations.
A Year Where Safety Indicators Finally Converged
At headline level, 2024 delivered the strongest safety performance the global steel industry has recorded.
Individually, each metric is positive. Collectively, they are significant. In prior years, the industry often saw progress in one area offset by stagnation or deterioration in another — for example, falling LTIFR without a corresponding reduction in fatalities. In 2024, that divergence narrowed materially.
This convergence matters because it suggests system-level improvement, not isolated performance spikes or reporting effects. It indicates that risk is being addressed earlier in the operational cycle — before exposure escalates into high-consequence outcomes.
More Important Insight: Hazards Did Not Change — Exposure Did
A deeper look at the data reveals a critical truth that is easy to overlook when reading annual reports. Despite record improvements in outcomes, the hazard profile of the steel industry has remained largely unchanged for more than a decade.
Falls from height, moving machinery, on-site vehicles, gassing and asphyxiation, and falling objects continue to dominate fatality statistics. These hazards were responsible for fatalities in 2015, remained prominent in 2020, and continued to account for most fatal outcomes in 2024.
The implication is clear: fatality reduction in 2024 did not occur because hazards disappeared. It occurred because uncontrolled exposure to those hazards reduced.
This distinction is fundamental for HSE leaders. It reframes steel safety away from the unrealistic goal of eliminating risk entirely and toward a more practical objective: controlling when, where, and how people are exposed to high-energy work. The breakthrough of 2024 lies in this shift from hazard identification to exposure control.
Why PSIF Thinking Changed the Trajectory
One of the most consequential developments underpinning 2024’s performance is the broader adoption of Potential Serious Injury or Fatality (PSIF) frameworks across steel operations.
Traditional safety systems implicitly treated incidents as broadly equivalent, prioritising volume-based metrics such as total injuries or near misses. PSIF analysis challenged this assumption by demonstrating that fatal risk is highly concentrated within a small subset of incidents.
In 2024, steel sites reported millions of precursors and near misses globally. Yet only a fraction of these events carried the potential to escalate into serious injury or fatality.
This concentration explains why earlier improvements in injury frequency did not always translate into fewer deaths. In 2024, organisations that focused on identifying and intervening in PSIF precursors began to see fatality outcomes improve as well. Leadership attention moved upstream — closer to where fatal outcomes are actually created.
What Steel Industry Behaviour Reveals About Safety Maturity
Another signal that 2024 represents a structural shift lies in how safety excellence is now being interpreted across the industry.
Increasingly, performance is associated not with low injury frequency alone, but with engineering-led interventions and process-level controls that remove high-severity risk from operations. This shift is reflected in the types of practices being recognised and replicated across leading steel organisations. Where earlier recognition often centred on lagging indicators, recent emphasis has moved toward demonstrable control of high-energy hazards through design and process integrity.
The fact that organisations such as Tata Steel were recognised for safety excellence in 2025 reinforces this shift away from frequency-led narratives toward genuine risk elimination. Tata Steel’s recognition was anchored in a process safety intervention addressing one of the most catastrophic risk scenarios in integrated steel plants: the derailment of rail-based torpedo ladle cars transporting molten metal. Rather than relying on procedural controls or post-incident indicators, the company implemented a safety instrumented system across its rail network to eliminate manual point-setting errors — a known precursor to derailments, loss of containment, hot metal–water explosions, and multi-fatality events. This approach directly removed the possibility of exposure at source.
Its recognition reflects what the 2024 industry data already indicates: sustainable safety improvement is achieved by engineering out high-severity risk before exposure occurs, not by managing outcomes after the fact.
Process Safety: The Boundary 2024 Made Explicits
While 2024 delivered historic gains in injury and fatality metrics, it also exposed a clear boundary around what traditional occupational safety programs can achieve.
Steel sites continued to report significant numbers of fires, explosions, molten metal–water reactions, and gas or liquid releases during the year. These events are not behavioural deviations. They are loss-of-containment failures with the potential for multi-fatality outcomes, regardless of how low LTIFR appears.
The breakthrough here lies not in elimination, but in visibility. These events are increasingly measured, disclosed, and discussed as leadership-level risks rather than being absorbed into general incident statistics. The implication is unavoidable: low LTIFR does not equate to low catastrophic risk.
Role of Technology in the Next Phase of Steel Safety
As safety systems mature, a practical limitation becomes evident. Many fatal precursors are visual, transient, and repetitive — unsafe proximity to moving equipment, vehicle–pedestrian interactions, deviations during work at height, and PPE breaches in hazardous zones. These conditions are difficult to capture consistently through manual supervision alone.
This is where technologies such as computer vision and video analytics for safety are gaining relevance within HSE programs. By enabling continuous monitoring and objective classification of unsafe conditions, vision-based systems support earlier detection of PSIF precursors and more consistent intervention in high-risk environments.
The lesson from 2024 is not that technology solved steel safety. It is that manual, retrospective systems have reached their ceiling in complex, high-energy operations. The next phase of improvement will depend on augmenting human oversight with technology-enabled visibility.
What 2024 Ultimately Proved for Steel Industry
2024 was a breakthrough year because it demonstrated that meaningful fatality reduction is possible without changing the fundamental hazard profile of the steel industry.
It showed that:
- Frequency and severity can improve together
- Fatal risk can be isolated rather than averaged
- Process safety must sit alongside occupational safety
- Safety maturity is increasingly defined by risk elimination, not metric optimisation
Steel safety is improving. What happens next will depend on whether organisations can sustain this transition — from counting incidents to controlling the moments that carry irreversible consequences.
