Risk Assessment and Accident Investigation in Context

Risk Assessment as defined in the South African mining context is the pursuit of managing the future safety of workers. It involves a detailed and systematic examination of any activity, location or operational system to identify risks, understand the likelihood and potential consequences of the risks, and review the current or planned approaches to controlling the risks, resulting in the instituting of additional controls where required.

Successful risk control can include outcomes such as improved safety, health, production, environmental protection or community acceptance.

From this explanation it is clear that essentially there are no differences between the accident prevention risk assessment and accident investigation processes as both aim to prevent unwanted incidents and both have the same components: hazard identification, consequence or impact, controls and failure modes. The consequence or impact stems from hazards being realised. Controls are active management interventions to eliminate or reduce the risks. Failure modes are the primary reasons for controls failing.

Despite the similarities in the two processes some procedural differences do exist. The main distinction is the timing within the management cycle of activation. Risk assessment occurs at the beginning of a management system cycle (pro-active) and accident investigation occurs after accidents (re-active) in the management cycle. However, the information contained in the outcome of accident investigations is fed into the risk assessment process and the cycle is repeated. A graphic representation of the main components of the two processes is given in Figure 1.

Accident investigation has a defined re-active outcome (specific information after a once-off activity) as opposed to risk assessment that pro-actively attempts to prevent incidents.

During the accident investigation process (left triangle in Figure. 1) the starting point is after the impact when the consequences are known. The first step is to confirm the hazard responsible for the accident. The next step is to determine the failure modes in order to reactively develop additional controls to prevent a re-occurrence. Since the consequences are known it is reasonably easy to identify the hazards. The challenge is to correctly identify the failure modes and then develop control measures to eliminate, minimize or control the risk at source, as personal protective equipment should only be issued in response to risks remaining after instituting controls.

A risk assessment process (right triangle in Figure 1) starts before the impact and potential consequences are known. The first step is therefore to identify the risks in each step of the process. The next step is to identify the potential failure modes for all the identified hazards. The aim of this process is to identify and ascertain the consequences pro-actively to develop controls that will prevent the various potential failure modes. It goes without saying that this process is much more complex since all the potential consequences and all the potential failure modes and existing controls should be evaluated for effectiveness. Should it be found that existing controls do not reduce residual risk to tolerable levels, additional controls should be developed and implemented.

In essence, there is only a very small difference between an accident prevention risk assessment and an accident investigation. It is only the sequence of events and the complexity of assessment that differ. It is therefore possible to utilise the same model for both accident investigations and risk assessment.

© 2009 Carl Marx