HazDat in Action

Using such a risk assessment model will aid in predicting injuries at your workplace.

FOR 30 years or more, industry and government managers have relied on injury statistics to rate the performance of safety programs and forecast the cost of worker's compensation insurance. While metrics such as Incident Rate and Lost Time Days are accurate measures of what has happened in the past, they are trailing indicators, do little to ferret out the cause of most accidents, and are of little use in understanding how well-trained, experienced machine operators manage to put their hands between immoveable objects and unstoppable blades.

Several years ago, in an effort to get ahead of the curve and eliminate the "points of opportunity" where an employee can be severely injured while engaged in a seemingly harmless task, industry professionals across the globe made a U-turn and began to use predictive models to assess the risk of injury at individual workstations and manufacturing cells. These models vary widely in the details, but most recent versions have a number of common elements: They are "task-based"; they factor the frequency of exposure and severity of the predictable injury into the risk formula; they establish a risk hierarchy for the purpose of setting priorities for taking remedial action; and they drive the organization toward reduction of the risk value and, presumably, lowering of the rate of serious injury.

The Key Skill: Imagination
When we began using the risk model as a hazard assessment tool, we had been retained by an international auto parts manufacturer. Working at a number of plants, we quickly found that experience in auditing and consulting for machine-intensive manufacturing companies would not be enough to enable us to develop a comprehensive risk-based assessment of the entire plant; there was simply too much going on. We needed help, and because we were convinced that hazard analysis would be best performed by a team, we set out to recruit supervisors and employees at the plants.

New members of the teams were selected for their understanding of the jobs being performed, but care was taken to avoid conflicts between hourly employees and their supervisors. We had no doubt, however, that hourly employees would help us avoid overlooking hazards associated with obscure, non-routine tasks and would facilitate the "buy in" to the recommended solutions.

Once the teams were formed, the training began, with heavy emphasis on identification of hazards--some obvious, many hidden or infrequently encountered. But it was clear the essential ingredient was imagination. It is easy to observe the activity around a power press; it is a challenge to conceive patterns of behavior and recognize those moments when critical decisions are made by an operator or die-setter. The team had to master the "what ifs" of the machine's operation.

Another important exercise was reliving the facility's history of accidents. In many cases, the information was available in the accident report; sometimes it was solely built on anecdote. We found past incidents were indicators of existing hazards or controls that were not adequate. They often furnished clues, enabling the team to recognize a hazardous condition that might not be obvious otherwise.

The Task-Hazard Pair
The system we developed was constructed around the concept of the task-hazard pair. This idea links each safety hazard to the task being performed by an employee when the hazard puts that employee at risk, thereby linking the mechanical equipment and the employee's routines and capabilities. While it is more time consuming than a pure "machine based" assessment, the "task-based" model considers the elusive quantity--human behavior--and factors it into the risk equation.

Hazards are not entered into the database alone; each is related to a task. This pairing is carried forward to the control strategy and ultimately to the report, which couples the multiple hazards associated each task, generating a Risk Value and Risk Category.

Task: Adjustment being made while the bore mill is running
Hazard: Pinch points.

Using the Risk Assessment Model
At the outset of the field work, the team is charged with identifying each task and defining the potential hazards or hazardous conditions associated with the task. The team members are taught to distinguish between a "hazard,"a condition or action that puts the employee at risk of an injury or illness, and the "risk," the probability the injury will occur.

When identifying task/hazard pairs, only the team's observations and a set of standardized risk criteria are considered. The program does the job of calculating the relative risk of injury; therefore, the team is instructed not to consider the overall risk. When the data being entered are complete, the calculated risk values will establish priorities for corrective action.

When assessing the task/hazard pairs at a workstation, the observation team must ask:

a) What action or condition would cause a hazard?
b) What type of injury or illness could occur?
c) How often does the hazard occur while the task is being performed?
d) How likely is it an injury or illness will occur?
e) If an injury or illness occurs, how severe could it be?
f) How quickly would the accident happen?
g) How much control over the situation would the employee have?
h) Does the PPE worn by the employee have any impact on the potential for injury or illness?
i) What ergonomic factors are contributing to the hazard?
j) Is the risk of injury related to an energy source that could be locked out when the activity takes place?

The answers to these questions are recorded on a field worksheet (see Figure 1). In order to normalize the responses of the team, numerical values are assigned to levels of risk for each factor. For example, the Injury Potential is scored using a number of considerations. The team must determine how likely it is the conditions leading to an injury or illness will occur and quantify that probability using the four choices found in the risk key.


Equipment #:

Mill #12345



HazDat #:



John Doe, Mel Gibson


Preset Machining--Upper Surface


Tooling change, mid run


Right hand caught between


Mill clamp and chamfer plate


Hazard Type:


Ergonomic Factors

Injury Type









Hazard Frequency



Extend up


Extend down




Injury Potential









Severity Potential









PPE Impact
















Injury Potential Criteria:
1) Remote: Generally understood to mean the hazard is unlikely to result in an injury to the employee(s) working in the area because the cause of the injury approaches the employee slowly (< 1 inch per second), and the employee can usually prevent injury by triggering a stopping mechanism or rapidly changing her position.

2) Moderate: Generally understood to mean injury to the employee(s) working in the area may be prevented if a minimum of caution is exercised; the cause of injury approaches the employee at moderate speed (1-3 inches per second); and/or the employee can sometimes prevent injury by triggering a stopping mechanism or rapidly changing her position.

5) High: Generally understood to mean the hazard is likely to result in an injury to the employee(s) working in the area unless diligence is exercised to prevent such an incident; the cause of injury approaches quickly but can be seen (> 3 inches per second); and/or the employee has very limited access to a stopping mechanism and can rarely avoid an impending injury.

10) Very Likely: Generally understood to mean the hazard is very likely to result in an injury to the employee(s) working in the area despite diligence exercised to prevent such an incident; the cause of injury approaches very rapidly, cannot be seen (high-speed ram action or disintegration of metal parts, explosion, arc flash); and/or the employee has no access to a stopping mechanism and cannot avoid the impending injury.

These risk criteria values are established based on the facility's general condition and can be modified to achieve specific goals. This is also true of the ergonomic criteria. These are checked off when a risk of illness or injury related to the employee's posture, the weight of objects being lifted or moved, or the repetition of a specific movement is present.

Calculating and Sorting Risk Values
When a workstation has been evaluated and discussed, the data can be entered into the electronic model at any convenient time. Our database has been set up in MS Access and is generally user friendly. When the task/hazard page is selected, the HazDat Number, Equipment or Zone Number, Date, and the risk criteria information can be entered. The model uses an equation that combines the criteria values to generate a risk value and risk category.

The risk value is useful in setting priorities for modifying equipment, training, or even redesigning systems. The risk category, which is considered very important in EU risk assessment systems, generally is used to establish the level of control that will be used to eliminate or reduce the risk of injury at a workstation.

Finally, the database generates a report that summarizes the risk values for each workstation, sorted by task/hazard pair, greatest to least risk. It also indicates to the facility the preponderance of hazard types (i.e., ergonomic, frequency), which can be used for developing broad-based risk reduction strategies.

The Value of Risk Assessment
There has been considerable debate among industry representatives, consultants, and trade associations about the usefulness of risk assessment as a tool for reducing injuries. While its success is difficult to gauge and is, to a large extent, dependent on the resources that are focused on the program by individual manufacturers, it is our experience that use of the system invariably draws attention to workplace hazards that might otherwise be overlooked and forms a solid base of data from which to track improvements.

The other notable positive impact has been the active participation by hourly employees, with apparent pride in their work. Only incident history over time will enable us to fully judge the technique.

This article appeared in the July 2006 issue of Occupational Health & Safety.

This article originally appeared in the July 2006 issue of Occupational Health & Safety.

Download Center

  • Safety Metrics Guide

    Is your company leveraging its safety data and analytics to maintain a safe workplace? With so much data available, where do you start? This downloadable guide will give you insight on helpful key performance indicators (KPIs) you should track for your safety program.

  • Job Hazard Analysis Guide

    This guide includes details on how to conduct a thorough Job Hazard Analysis, and it's based directly on an OSHA publication for conducting JHAs. Learn how to identify potential hazards associated with each task of a job and set controls to mitigate hazard risks.

  • A Guide to Practicing “New Safety”

    Learn from safety professionals from around the world as they share their perspectives on various “new views” of safety, including Safety Differently, Safety-II, No Safety, Human and Organizational Performance (HOP), Resilience Engineering, and more in this helpful guide.

  • Lone Worker Safety Guide

    As organizations digitalize and remote operations become more commonplace, the number of lone workers is on the rise. These employees are at increased risk for unaddressed workplace accidents or emergencies. This guide was created to help employers better understand common lone worker risks and solutions for lone worker risk mitigation and incident prevention.

  • EHS Software Buyer's Guide

    Learn the keys to staying organized, staying sharp, and staying one step ahead on all things safety. This buyer’s guide is designed for you to use in your search for the safety management solution that best suits your company’s needs.

  • Vector Solutions

Featured Whitepaper

OH&S Digital Edition

  • OHS Magazine Digital Edition - June 2022

    June 2022


      Corporate Safety Culture Is Workplace Culture
      Keeping Workers Safe from Heat-Related Illnesses & Injuries
      Should Employers Consider Oral Fluid Drug Testing?
      Addressing Physical Differences
    View This Issue