The Failure Progression Model
Following many years of research, development, testing, trials, and observations, we created a failure progression model. The model has its roots in a structured process developed by Boeing for investigating failure events called the Maintenance Error Decision Aid© or MEDA©. The MEDA process was expanded and re-designed to apply the concepts of human factors, errors, violations, underlying conditions, and contributing factors to equipment failures.
The model begins with the realization that all equipment failures produce risk. The risk can be analyzed and ranked using specific risk acceptance criteria in various value areas such as safety, environment, production, or costs. Risk is mitigated by lowering the consequence or lowering the probability (or both) of the equipment failure.
The most effective way to lower the risk of failure is to lower the probability. The full failure progression model shows the many opportunities to lower the probability of failure with the ultimate objective of preventing the failure altogether. As the model shows, the probability of equipment failure increases as stress is applied to the equipment. The stress applied to equipment is rarely the result of aging or normal wear but the result of an event or series of events. Pumps cavitate, compressors surge, lubrication degrades, connections loosen, lines plug……all of these events apply stress to equipment that can lead to an equipment failure.
There is much evidence to support that 80% – 95% of all events have two root causes. One, violations of the process design limits often referred to as the Integrity Operating Window (IOW) and two, errors occurring during some human interaction with the process or equipment. The errors and the violations raise the probability that events will occur which raises the probability that equipment failures will occur.
Finally, even the probability of violations and errors are influenced by underlying conditions and contributing factors. The contributing factors heavily influence human performance. There are hundreds of contributing factors that influence human performance, however, there are twelve that tend to dominate most working environments known as The Dirty Dozen.