Safety and health in the steel industry: Data report 2026

Foreword Potential serious injuries and fatalities (PSIF)Fatality analysis Process safety management (PSM) analysis All injury frequency rate (AIFR) and total recordable injury frequency rate (TRIFR)Lost time injury (LTI) analysis Sickness absence Download publication

Notes:

This report is based on data provided by our members and offers crucial insights into our industry’s ongoing commitment to the safety and wellbeing of our workforce.

All frequency rates are calculated per million hours worked.

For more details on terminology definitions and calculations, please refer to our glossary on worldsteel.org

The data in this document is provisional and may be updated or supplemented with additional information.

Foreword

Our industry has made remarkable progress over decades through a sustained focus on managing and reducing risk. As a result, it is safer today than it has ever been.

However, the reality is that last year, 74 people did not return home to their families at the end of their shifts.

Our work is not done.

Improvement is becoming harder, in part because the nature of fatal risk itself has changed over time. Fatalities are now concentrated in high-risk, non-routine activities involving a high degree of unpredictability, such as working at height, maintenance, vehicle movement, forklift operations and gas-related work. Importantly, these activities are also frequently carried out by contractors which explains the disproportionate number of contractor fatalities. To tackle this challenge, we need to focus our resources where a significant impact can be made.

The industry needs a shift from traditional activity-based safety approaches to critical risk control, barrier health assurance and potential serious injuries and fatalities (PSIF) thinking, using modern approaches like human and organizational performance (HOP).

This report reflects this shift within the industry with greater emphasis on PSIF, leading indicators and preventive action, alongside continued efforts to improve global alignment and transparency.

Shalini Kumari
Manager, Safety and Health

Across our members, there is now clear alignment that the most effective way to reduce fatalities and life-altering injuries is to prioritise the critical controls that prevent them.

worldsteel’s position on lagging indicators

We continue to emphasise that lagging indicators such as lost time injury frequency rate (LTIFR) or total recordable injury frequency rate (TRIFR) should not be part of individual, team or organisation objectives, bonuses, or incentives and they should not be used for comparisons between organisations or local sites. This is because of the following reasons:

These indicators correlate poorly with the severity of accidents and injuries and do not provide a valid or reliable measure of safety and health controls and initiatives.
They can foster a negative attitude towards reporting, limiting the organisational ability to learn and improve.
A smaller organisation’s LTIFR or TRIFR is more significantly affected by a single incident but this doesn’t necessarily indicate a more dangerous working environment.

Furthermore, although worldsteel has a definition for lost time injury (LTI) and total recordable injury (TRI), companies vary in their definition of what constitutes an LTI and TRI. For this reason, this report has an additional indicator, all injury frequency rate (AIFR), which includes every kind of injury irrespective of severity. However, this is also a lagging indicator, and our focus should be on indicators that monitor and prevent potential serious injuries and fatalities.

Potential serious injuries and fatalities (PSIF)

Serious injuries generally refer to long-term or permanent incapacity, as well as fatalities.

A potential serious injury or fatality is any incident, regardless of actual severity, that has the potential to lead to a life-threatening, life-altering, or fatal injury.

A PSIF precursor is any unmitigated high-risk situation that has potential to result in a serious injury or fatality if not controlled because management controls are absent, ineffective, or not complied with.

The number of individual sites reporting to worldsteel using the PSIF framework has increased in recent years. The figures below only represent the sites reporting PSIFs and combining contractors and employees.

In 2025, 92% of sites made use of a PSIF framework.

Figure 1: PSIF triangle 2025 for employees and contractors

* In 2025, there were 74 fatalities reported to worldsteel. In order to preserve the ratios within the PSIF section of this report, fatalities reported by organisations that were not able to provide PSIF information have been excluded from this analysis.

** All other injuries includes restricted work cases (RWC), medically treated injuries (MTI) and minor injuries (MI).

Stated injury statistics in this graphic are derived from companies that report PSIF information to worldsteel.

Fatality analysis

A fatality refers to death from a work-related injury, certified by a medical professional.

* Data received from China via Sinosteel Wuhan Safety & Environmental Protection Research Institute (SEPRI) does not distinguish between employees and contractors. Data received directly from Chinese worldsteel members does and is included in the analysis.

Figure 5: Causes of fatalities 2016 – 2025

The following graphs show the distribution of causes split by employees and contractors.

*Note: Events registered as ‘other’ or ‘Unknown’ are not included.

Process safety management (PSM) analysis

Process safety management is a blend of engineering, operations and management skills focused on prevention of, preparedness for, mitigation of, response to, and restoration from catastrophic events, particularly explosions, fires, structural collapse and damaging releases of hazardous substances or energy from a process.

The manufacturing of steel involves processes with intrinsic hazards that need careful management.

The focus of process safety management is not limited to protecting the people within the company but also includes the environment, assets and the surrounding community.

Figure 8: Process safety management maturity assessment results 2025

The process safety management maturity analysis was derived from data provided by 38 organisations, representing 36.2% of worldsteel members.

Increasing maturity in process safety management can be observed. The industry has grown in experience and expertise in process safety management.

However, there are some important areas of improvement for elements 6 (process and equipment knowledge), 15 (measurement and metrics), and 16 (auditing).

Process Safety KPIs are different from Occupational Safety KPIs. A specific set of indicators and guidance are being developed for process safety.

Table 1: Significant process safety events 2025

	Fires	Explosions	Molten metal and water reactions	Gas and liquid releases
Quantity	757	84	68	317

In 2025, worldsteel members reported 106 tier 1 process safety events (PSEs) and 247 tier 2 PSEs.

All injury frequency rate (AIFR) and total recordable injury frequency rate (TRIFR)

All injuries comprise fatalities, lost time injuries (LTIs), restricted work cases (RWCs), medical treatment injuries (MTIs) and minor injuries (MIs).

Total recordable injuries comprise fatalities, lost time injuries (LTIs), restricted work cases (RWCs), and medical treatment injuries (MTIs). The frequency rate refers to the total number of injuries per million hours worked.

In the data survey, RWCs, MTIs and MIs are recorded separately. Working hours are counted only on sites that have reported at least one RWC or MTI or MI.

Safety pyramids can be used to compare the shape of a company’s incident triangle. In case the company triangle is narrower, it could mean that not all incidents are reported, or that fatality prevention requires more attention. It should be noted that the average incident triangle is not necessarily optimal; it should probably be wider. Reducing the number of more minor incidents does not directly reduce the probability of severe incidents, but having those reported allows action to be taken to mitigate risks that might otherwise cause severe incidents.

* In 2025, there were 74 fatalities reported to worldsteel. In order to preserve the ratios within the TRIFR section of this report, fatalities reported by organisations that were not able to provide TRIFR information have been excluded from this analysis.

Figure 14: Safety pyramids 2025 for employees and contractors

Figure 15: Regional safety pyramids

Table 2: Ratio LTIs vs fatalities 2012-2025

Cause	LTIs	Fatalities	Ratio LTIs per 1 fatality
Gassing and asphyxiation	272	155	1.75
Explosion	274	100	2.74
Rail	224	79	2.84
Fire	380	83	4.58
Electrical	508	89	5.71
Fall from height	2228	346	6.44
Structural failure	403	60	6.72
On site road vehicle	994	144	6.90
Overhead crane	1101	111	9.92
Forklift	368	36	10.22
Moving machinery	3344	321	10.42
Hot metal	809	76	10.64
Falling object	2500	185	13.51
Other mobile equipment	996	61	16.33
Hot substances	659	31	21.26
Off site road vehicle	345	14	24.64
Product loading	735	18	40.83
Product handling storage	2163	40	54.08
Exposure to chemicals	229	3	76.33
Manual tasks tools	3811	13	293.15
Slip, trip and fall	4497	6	749.50

The table shows the ratio between lost time injuries and fatalities by causes. Low ratios mean that per fatality, there are, on average, only a small amount of lost time injuries; a high ratio indicates there are many lost time injuries per fatality. If these were drawn as an incident triangle for each cause, the triangles for the low ratios would be very narrow.

These ratios show that to prevent fatalities, it is more important to undertake comprehensive and robust incident investigations and instigate preventive measures for those incidents with a low ratio. It is precisely these incidents that are most likely to lead to fatalities without effective prevention. These ratios also emphasise the usefulness of categorising potential serious injuries and fatalities among other incidents.

Lost time injury (LTI) analysis

Any work-related injury resulting in the employee or contractor being unable to return to work for their next scheduled work period. Returning to work with work restrictions does not constitute an LTI, no matter how minimal or severe the restrictions, provided it is at the employee’s next scheduled shift. However, if an injury deteriorates and time is later lost, a LTI should be recorded. The LTIFR is calculated by combining fatalities and LTIs.

Figure 19: Causes of LTIs 2016-2025 for employees and contractors combined

The following graphs show the distribution of causes, separated by employees and contractors.

Figure 20: Causes of LTIs in 2025 for employees and contractors

Sickness absence

Sickness absence refers to absence from work on the grounds of incapacity to work due to any sickness, work-related or not, and which could qualify for ‘disability income’. All other cases of absence, such as pregnancy, childbirth, leave, training and seminars, are not included in the definition of sickness absence.

Sickness absence rate is calculated as the total number of hours of sickness absence per scheduled hours. This indicator is only calculated for employees and is expressed as a percentage.

Table 3: Sickness absence per region 2016 – 2025

% Sickness Absence	2016	2017	2018	2019	2020	2021	2022	2023	2024	2025
Africa and Middle East (AME)	0.25	0.23	0.29	1.06	0.18	1.01	1.53	2.8	1.19	0.62
Asia/Pacific	0.08	0.08	0.19	0.15	0.05	0.11	0.69	0.74	0.2	0.26
Russia & other CIS + Ukraine	1.26	1.41	0.6	1.38	1.88	1.73	2.33	2.21	1.66	0.75
Europe	2.02	2.26	1.92	3.32	3.89	4.36	6.38	4.95	4.34	4.96
North America	0.14	0.14	0.23	0.22	0.42	0.14	1.31	1.27	3.5	NA
South America	0.53	0.48	0.43	0.75	0.46	0.42	3.03	1.59	1.28	1.08

Reported occupational sickness levels vary between regions due to differences in industry composition, regulatory standards, and access to healthcare. Cultural attitudes towards reporting workplace-related illnesses and socioeconomic factors also play a role in the disparities. Efforts to improve healthcare access can help promote healthier working environments worldwide.

.
.
.

	Employees	Contractors	Combined
2016	0.017	0.040	0.025
2017	0.018	0.034	0.024
2018	0.030	0.029	0.029
2019	0.014	0.032	0.021
2020	0.015	0.038	0.024
2021	0.023	0.037	0.029
2022	0.015	0.041	0.027
2023	0.014	0.025	0.019
2024	0.011	0.022	0.016
2025	0.011	0.025	0.018

	Africa and Middle East (AME)	Asia / Pacific	Europe	North America	Russia & other CIS + Ukraine	South America
2016	0.051	0.022	0.023	0.026	0.037	0.02
2017	0.006	0.012	0.031	0.022	0.05	0.033
2018	0.014	0.033	0.015	0.008	0.056	0.011
2019	0.014	0.019	0.018	0.018	0.039	0.004
2020	0.024	0.02	0.025	0.018	0.049	0.005
2021	0.091	0.028	0.022	0.022	0.039	0.012
2022	0.018	0.03	0.032	0.004	0.078	0.011
2023	0.019	0.018	0.02	0.013	0.066	0.019
2024	0.003	0.015	0.026	0.003	0.032	0.02
2025	0.019	0.019	0.009	0.024	0.033	0.014

	Total AIFR	Contractors AIFR	Employees AIFR
2017	10.5	7.5	12.4
2018	11.2	6.8	14.4
2019	7.78	4.80	10.11
2020	8.78	4.21	11.90
2021	9.35	5.59	12.10
2022	11.18	5.52	16.62
2023	11.68	5.18	18.47
2024	6.93	3.79	10.53
2025	5.67	3.45	8.11

	Africa and Middle East (AME)	Asia / Pacific	Europe	North America	Russia & other CIS + Ukraine	South America
2017	5.53	5.54	19.9	18.79	5.21	10.71
2018	39.22	5.46	22.77	17.93	5.02	10.3
2019	16.7	3.42	20.86	13.36	3.16	7.84
2020	5.77	3.36	24.4	15.12	10.17	7.54
2021	7.67	4.82	27.1	14.59	6.67	8.3
2022	20.37	4.4	23.97	12.25	50.07	7.47
2023	32.1	4.24	29.86	12.84	39.31	8.34
2024	20.4	2.48	28.09	12.29	3.29	7.99
2025	4.57	1.87	20.11	10.93	2.78	8.2

	Total AIFR	Contractors AIFR	Employees AIFR
2017	4.5	4.0	4.8
2018	4.9	2.9	6.3
2019	4.84	2.87	6.16
2020	3.48	2.31	4.19
2021	4.02	2.69	4.93
2022	4.74	2.71	6.43
2023	6.29	2.65	9.46
2024	3.55	1.83	5.55
2025	3.09	1.79	4.32

	Employees	Contractors	China (indirect data)	Totals
2016	45	57	27	129
2017	40	49	15	104
2018	70	48	22	140
2019	32	49	18	99
2020	38	56	12	106
2021	45	55	11	111
2022	28	71	7	106
2023	25	48	9	82
2024	19	43	6	68
2025	21	49	4	74

Attribute	Value
Fall from height	14
Gassing and asphyxiation	11
Hot metal	6
Falling object	6
Other mobile equipment	5
On site road vehicle	5
Moving machinery	5
Overhead crane	3
Manual tasks tools	3
Hot substances	3
Product handling storage	2
Explosion	2
Unknown	1
Structural failure	1
Rail	1
Forklift	1

Attribute	Values
1. Fatalities*	70
2. Lost time injuries	2384
3. Restricted work cases	1213
4. Medical treatment injuries	3724
5. Minor injuries	12006
6. Near misses (incidents)	147157
7. Unsafe acts and situations	2611434

Attribute	Value
1. Fatalities	21
2. Lost time injuries	1626
3. Restricted work cases	856
4. Medical treatment injuries	2727
5. Minor injuries	8202
6. Near misses	110817
7. Unsafe acts and situations	1467972

	Employees	Contractors	Total
2016	3129	925	4054
2017	2835	743	3578
2018	2429	923	3352
2019	2211	1002	3213
2020	2167	881	3048
2021	2075	870	2945
2022	2004	1083	3087
2023	1897	979	2876
2024	1855	857	2712
2025	1647	807	2454

	Africa and Middle East (AME)	Asia / Pacific	Europe	North America	Russia & other CIS + Ukraine	South America
2017	2.99	2.79	7.82	4.58	3.18	4.4
2018	20.79	1.63	11.55	4.91	1.51	4.05
2019	14.36	1.92	11.89	5.03	2.15	3.88
2020	3.72	1.86	11.08	4.69	1.93	3.56
2021	6.54	1.86	12.46	4.08	1.74	3.82
2022	19.24	1.71	11.32	3.43	1	3.49
2023	36.57	1.76	14.97	4.33	0.84	3.67
2024	17	1.12	15.3	4.08	0.84	3.24
2025	3.91	1.36	10.57	3.78	1.04	2.92

	Employees	Contractors	Combined
2016	1.21	0.64	1.01
2017	1.25	0.52	0.97
2018	1.03	0.56	0.84
2019	0.94	0.65	0.83
2020	0.87	0.60	0.77
2021	1.05	0.59	0.85
2022	1.05	0.63	0.85
2023	1.03	0.51	0.76
2024	1.03	0.43	0.72
2025	0.88	0.41	0.64

	Africa and Middle East (AME)	Asia / Pacific	Europe	North America	Russia & other CIS + Ukraine	South America
2016	0.7	0.49	4.09	1.19	0.85	1.2
2017	0.7	0.2	4.42	1	0.78	1.14
2018	1.25	0.24	3.63	0.88	0.79	0.79
2019	1.05	0.24	3.77	0.83	0.71	0.86
2020	0.6	0.27	3.79	0.84	0.76	0.77
2021	1.59	0.39	3.84	0.89	0.69	0.74
2022	1.21	0.38	4.12	0.67	0.64	0.62
2023	0.94	0.31	4.26	0.65	0.34	0.64
2024	0.91	0.31	3.93	0.62	0.35	0.57
2025	0.94	0.29	3.32	0.67	0.49	0.76