Human Reliability Assessments (HRA) start by analyzing operations proactively, looking at the types of potential human error and the probability of these happening. An organization must understand what effect such risk will have on safety. It needs to understand how to control or reduce human error so it’s As Low As Reasonably Practicable (ALARP).
To assess the potential human errors within the organization and the possible effects, we can use a Hazard Analysis method called Human Reliability Assessment (HRA).
There are three (3) main steps of a basic HRA:
1) Identifying human error/hazards
2) Screening hazard analysis
3) Quantifying human error
Step 1: Identifying human error
This is a critical part of human error analysis. We need to know about all the potential risks, so we don’t underestimate the threat to system safety and performance. This stage is used to assess human errors qualitatively, which forms the basis for quantifying error in stage 3.
UP can use several proven techniques, particularly a human HAZOP (hazard and operability study). This variant of an engineering HAZOP focuses on human error(s). The human HAZOP can form an integral part of your safety engineering approach by adjusting it accordingly (including guide words for identifying human errors and their underlying causes).
Key steps of the HAZOP
1) We break down your operator’s tasks into component actions with a formal analysis
2) Actions are assigned to the relevant system components to be analyzed in the HAZOP
3) Each system component is assessed in terms of operator actions
4) For each action involving the component, we identify potential human errors
We identify human errors with an extensive list of External Error Modes (EEMs). Each EEM is a simple keyword or term (for example, “failure to detect indicator signal”, “incorrect interpretation,” or “selects wrong procedure”). These are categorized by the type of activity.
The HAZOP analysis uses a similar categorized list of the underlying Psychological Error Mechanisms (PEMs). These are the cognitive explanations of the errors (for example, “lapse of attention”, “inadequate time perception,” and “cognitive overload”).
For each PEM, we use a list of possible error reduction mechanisms.
Step 2: Screening analysis
Here, we identify errors that can effectively be excluded from further analysis and those which should be assessed further. We assign a probability of 1.0 to each human error and then examine the consequences. The probability of 1.0 reflects a worst-case scenario so that any negative effects will be exaggerated. Errors found to have a negligible effect are not analyzed further. Significant errors are then quantified.
Step 3: Quantifying human error
If we’ve carried out screening analysis, we only need to look at errors that have a significant consequence. We assess system safety numerically, calculating human error likelihood in various ways, including the Human Error Assessment and Reduction Technique (HEART)
The output allows tasks to be compared relative to each other. We then assess each individual task in turn.
First, we specify the generic nature of the task.
We then identify error-producing conditions (EPCs). For example, “unfamiliarity.”
Next, we specify the extent of each EPC effect, assigning a value between 0 and 1.
We calculate the final error probability.