In everyday conversation, the word 'risk' is often used to mean the likelihood, or probability, of something (usually negative) happening. So we talk about the risks of having an accident while driving, of contracting heart disease or cancer, or of being struck by lightning in a thunderstorm. These risks are considered by, for example, insurance companies using statistical data as the basis for assessment.
In the context of engineering structures, 'risk' is often used in a more restrictive sense to express both the probability (likelihood) of an occurrence (such as the failure of the structure) and the degree of consequences arising from it (e.g. lives lost). The association of probability with consequences may be purely qualitative and judgmental (using terms like high, medium and low) and this may be useful in assessing relative risks. Where the probability of occurrence and the consequences can be quantified, then risk is formally defined as the product, i.e. -
risk = probability of occurrence x consequences arising from occurrence
Buildings, bridges, chemical plants and pipelines are usually built and maintained in accordance with standards and codes of practice that aim to achieve a certain target reliability (i.e. probability of not failing). But failure of, say, a pipeline will have very different consequences for persons if it occurs in a crowded city where people may be injured rather than in the unpopulated desert, no matter how unlikely the event. The risk to persons associated with the city-based pipeline is therefore higher, assuming that the two pipelines are designed and operated in the same way.
Risk reduction can involve reducing either the probability of failure of a structure or the consequences of failure, or both. Reduction of the probability of failure can be achieved by various engineering measures, such as:
- decreasing the operating pressure of a pipeline;
- controlling the number and weight of vehicles on a bridge;
- providing 'redundant' load paths in the structure so that failure of one component will not lead to failure of the entire structure.
Reduction of the consequences of failure, on the other hand, could involve a number of actions unrelated to the reliability of the structure, for example:
- excluding as many people as possible from the area of risk (e.g. building a fence around a chemical plant, carrying out pressure testing of equipment outside normal working hours);
- building a blast-wall between pressure vessels, so that failure of one does not lead to failure of an adjacent vessel;
- training of staff (and, where necessary, members of the local population) in emergency procedures.
'The tolerability of risk from nuclear power stations', HMSO, 1988 (revised 1992), ISBN 0 11 886368 1, http://www.onr.org.uk/documents/tolerability.pdf