Risk Engineering and Engineers
By ROLAND O'BRIEN-BILLS, President, Accident Inspection Specialists Risk Engineering, Gloversville, N.Y. -- Consulting-Specifying Engineer, 11/1/2000
Why is risk engineering needed when the design is strong and the unit or system is maintained or needs upgrading?
When using engineering judgement, the engineer relies on knowledge. Unfortunately, today's knowledge may not predict tomorrow's needs. The demands that owners and users place on engineers may influence a marginal engineering decision, even in cases when the specification may meet the code minimum. Perhaps an engineer cannot foresee an accident that will occur 14 years later because of an engineering judgement that was good then, but marginal later.
A well-known risk-engineering method is failure mode and effects analysis, or FMEA, but there are other excellent methods. In some analyses, additional methods may be desired for best results. By combining FMEA with other risk-engineering techniques, discipline and system knowledge, the practice can be applied to mechanical and electrical systems. The following are simplified examples to illustrate the importance of risk engineering. The first step in each application is to ask how and what can fail in each circumstance.
In a case of poor power quality, the "what" is identified as the electrical distribution system, and the "how" is poor power quality, such as surges and spikes to harmonic disturbances. The next step is determining the probability of a failure. This includes many factors such as local grid stability, weather and harmonic distortions from variable-frequency drives.
Next, a severity is applied by ranking the impact of the failure. This can be measured in many ways, but the preferred method is dollars. At this point, risk modifiers are factored. These are safety systems, redundancies, periodic coordination studies and the like.
In this example, the outcome may show that the risk is worth taking with a few additional protection relays. The risk analysis may also show the other extreme, where the current system design is an unacceptable risk to take and may require in-house power generation or power-quality protection and correction components throughout the electrical distribution system.
Take another example of mechanical design: Suppose overpressurization is the failure mode on a boiler, pressure vessel or piping system. The "what" is the vessel and the "how" is overpressurization-lack of proper safety valves and relief valves to rupture disks. The next step is determining the probability or likelihood of a failure from an overpressure condition. Again, this encompasses many factors to consider, from location and sizing to discharge and valving. The severity is then applied by ranking the damages caused from the explosion. Severity is measured by the financial impact.
Finally, risk modifiers are factored, such as testing frequencies and overhaul schedules. Redundancies could be sentinel valves or audible alarms. In this example, the outcome may show that the risk requires that an alarm be installed when the pressure approaches the design pressure. On the other hand, the risk may be unacceptable and each vessel may require overpressurization protection instead of system-header protection.
These two examples clearly lack rigorous evaluation but intuitively show the insightful power of risk engineering. The challenge is accommodating human behavior that modifies the hazards through the life of the equipment.
In engineering, design adequacy assessments and engineering checklists are relied upon, but they do not quantify or qualify risk levels of the tradeoffs specified. Engineering judgements are made every day and risk engineering can help make those judgments by keeping potential failures in mind and accountable.
