Reliability and Safety Assessment

Reliability and safety assessment methods have developed over a number of years from a variety of different initiatives. The pioneer work in Germany on missile systems during World War II, the development of reliability methods for defence equipment by the US Department of Defense, and the contributions on hazard analysis by Trevor Kletz of ICI Ltd and on reliability methods by Green and Bourne of the UKAEA are all worthy of note. Milestones also exist such as the reports of the public enquiries following the accidents at Windscale, Flixborough, Piper Alpha, the well-know WASH 1400 report on nuclear safety, the Canvey Island risk assessment report and many others. All of these initiatives have contributed to the increased awareness of the general public to potential hazards and highlighted the need for better methods for ensuring the reliability of complex systems and improving the safety of hazardous plant.

After the first world war reliability became a major interest of the aircraft industry. Comparisons were made between the reliability of single and multi-engined aircraft although at that time little attempt was made to express reliability in quantitative terms. Most attempts at reliability improvement were based on trial and error. When something failed its replacement would be designed and manufactured using any improvements in of technology together with experience gained from investigating the failure. As time went by information was gradually collected on system failures which led naturally to the concept of expressing reliability in terms of the failure rate for a particular type of aircraft or system. In the 1940s the reliability requirement was given for maximum permissible failure rates of 1 per 100,000 flying hours. It was also estimated that there was a fatal accident risk of 1 in 106 landings prior to automatic landing systems. When automatic landing systems were introduced in the 1960s the reliability requirement of the system was specified in terms of the fatal accident risk being not greater than 1 in 107 landings. So quantification gradually became part of the design specification.

The missile and space industry is a product of the second world war and is the area where quantitative reliability became formalised. After the initial poor reliability of the V1 missile Lusser, a German mathematician, queried the original assumption that the reliability of a chain of components was determined by the strength of its weakest link. He showed that a large number of fairly ‘strong’ links can be inherently more unreliable than a single ‘weak’ link because of the variability of component strengths and operational loading. Quality assurance was therefore applied to increase the reliability of all components resulting in a dramatic improvement of V1 reliability which ultimately achieved a success rate of over 60%. Because of the increase in complexity of modern systems the reliability of such devices is still in the in the region of 60-70%.

The American armed forces took an increasing interest in reliability and its measurement in the second world war because the unreliability of vital equipment and systems was causing significant problems. In the Korean War the unreliability of electronic equipment was reported as costing $2 per year to maintain each dollar’s worth of equipment. In the UK at this time it was also noted that the lack of reliability and maintainability of equipment was forcing the armed services to spend over half its resources on maintenance rather than operations.

Arising from this appreciation of the importance of reliability and maintainability a series of US Department of Defense Standards (MIL-STD’s) were introduced and implemented. Subsequently the UK Ministry of Defence also introduced similar standards. By and large their introduction has significantly improved the reliability and maintainability of military equipment and systems.

Reliability and safety assessment methods have also been introduced into the process industries. Here the objectives are to improve the safety and availability of new and existing plant. The technology of reliability has benefited significantly from this interest and the concepts of unrevealed failures and techniques such as fault tree analysis, network analysis, Markov analysis and simulation, have found many applications in process plant assessment.

For more information on safety assessment software visit Isograph’s web site at www.isograph.com.

Visit the Isograph Blog

The Isograph Blog contains a wealth of additional information on Isograph products and services.

More Information on Isograph products

Reliability and safety assessment software