Practical Human Reliability Analysis an Effective Tool for Simultaneous Enhancement of Safety and Productivity, 1998 Engineering Conference Proceedings

In recent years, significant advances have been made in methods and approaches for analyzing human error and assessing the likelihood of people being successful in routine operations, maintenance, and responding to abnormal conditions. This success depends on a synergistic response by hardware systems, human operators and maintenance personnel. The area of study which focuses on people in the system is human reliability analysis @IRA). Many of the better-known models for HRA are hybrid approaches for modeling human errors using techniques such as probability trees, models of dependence, and performance shaping factors affecting an individual’s actions. These rigorous and methodical processes are effective tools to use when a comprehensive evaluation by HRA specialists is required to support a detailed quantitative risk assessment. However, for many safety assessments, a more practical, but effective, method can be applied for a fraction of the cost and time typically required to do rigorous quantitative analysis. Some trade-off of detail and precision for prompt and effective identification of potentiaily significant deficiencies is considered acceptable. One method that meets the “practical” criteria is Operational Human Performance Reliability Assessment (OHPRA). OHPRA provides added emphasis on human factors considerations in a systematic, yet practical way. Like classical human reliability models, the OHPRA mode1 considers all factors significantly affecting human performance, but focus is on identifying problems and potential improvements rather than generating human error probabilities. In the past five years, OHPRA has been used for 34 assessments covering refinery, petrochemical, chemical process and cogeneration facilities. Example case findings and results will be discussed.