Probabilistic reliability assessment

Everything has the potential to fail. The weather can have disastrous effects on the natural cycle of life. The conditions which lead to a flood or drought can, in turn, affect essential commodities:­ their availability and pricing. Despite our best efforts at long-term weather forecasting our success and mitigations continue to be poor.

But in the case of the non-natural world, our success rates are good; science engineers can now forecast with great accuracy everything from the integrity of a steel beam to the reliability of a jet engine. For machines or components which are expected to work once, first time, every time, the engineering has to be perfect, enabled by computer modelling of reliability.

KGAL’s engineers understand the techniques of reliability analysis and estimating component tolerances, sensitivities and stability. Some of the applications to which our assessments are modelled include failure rates in hydraulic steel structures but, more generally, are used in helping to forecast energy reliability from renewables. Using templates and bespoke mapping, we merge data from multiple factors: the ‘knowns’ (for example location, capacity) with the more abstract ‘unknowns’ (disruptive weather, seismic activity, even future changes to legislation) to provide clients with long-term planning intelligence both for power systems and specific assets.

Our consultancy service advice from a probabilistic reliability assessment (PRA) can become pivotal to an energy client’s investment plans. Consider, for example, a scenario where transmission lines are approaching their capacity: there is the potential for overload tripping even though the energy source itself is operating normally. The revenue forgone by any outage can be seen against the grid improvement cost to determine justification. Our role would be to forecast the likelihood and duration of any such outage.

PRAs form a fundamental part of mission critical projects where failure can be truly catastrophic: air traffic control systems, certain medical equipment, data centres, public transport signalling. In these situations 99.999% reliability still means there are over five minutes every year when the anaesthetist’s ventilator won’t work.

And if to achieve that remaining 0.001% costs four times the investment to get to 99.999% then the reliability of the reliability assessment itself is critical.

A component malfunction, system design fault and installation error are amongst the most common causes of critical failures. In this regard a risk assessment forms an essential prologue to PRA and an overall risk analysis should be seen as a mandatory requirement.

For both types of risk control investigation, KGAL has the experience to undertake dependable research upon which planning and investment decisions can be made.