6.7 System failures and system reliability Flashcards Preview

C06. Work equipment (workplace machinery) > 6.7 System failures and system reliability > Flashcards

Flashcards in 6.7 System failures and system reliability Deck (6)
Loading flashcards...
1

The overall reliability of equipment depends on both the reliability of all components and the way in which they are arranged, which may be as: 3

 Series systems  Parallel systems  Combinations of both series and parallel systems - i.e. mixed (complex) systems.

2

The overall reliability of a series system is calculated by

by multiplying the reliability of each individual component

3

A designer is designing a safety-critical, electrical control system. The system consists of a number of components arranged in series. One of the components is a detector which has a reliability of 0.95. The designer is considering installing two identical detectors in parallel to improve the reliability of the system. (a) Calculate the improvement in reliability that using two identical detectors in parallel would give compared with a single detector. 4 marks (b) Outline the issues that would need to be considered when assessing whether the proposed extra detector in parallel should be adopted. 3 marks (c) Assuming that the decision is taken to use two detectors in parallel, outline other ways in which the reliability of the control system could be improved. 9 marks (d) The designer has been warned that his assumptions of improved reliability from two detectors might be undermined by common mode failure. Outline the meaning of ‘common mode failure’ AND outline why it may affect the reliability as calculated in (a) above. 4 marks

(a) The following calculation should have been produced to demonstrate a full understanding of the problem: The reliability of the parallel components: 1 – [(1 - 0.95)(1 - 0.95)] = 1 – (0.05 x 0.05) = 1 – 0.0025 = 0.9975 …or when rounded to 3 decimal places … 0.998 or 99.8% The improvement in reliability would therefore be: 0.998 – 0.95 = 0.048 or 4.8% (b) Issues that would need to be considered include:  the probability of system failure and its consequences  legal requirements and advice contained in industry and HSE codes of practice and guidance  the initial cost of the additional detector  the subsequent expense connected with the detector's on-going maintenance and inspection  risk tolerability criteria such as those contained in ‘Reducing Risks, Protecting People’. (c) Additional ways in which the reliability of the control system could be improved include:  the use of design stage failure tracing techniques, such as HAZOP  introducing purchasing quality control arrangements to ensure the most reliable detectors are used  using two different types of equipment to minimise the risk of common mode failure  ensuring the system components are tested before installation and that they are correctly installed by competent personnel  arranging for the introduction of procedures for the periodic inspection, testing and maintenance of the system, including the replacement of components within their useful life  providing training to employees in operating the system and in fault detection using indicators or warnings to indicate component failure. (d) Common mode failure is as a type or cause of failure that could affect more than one component at a time, even when the components are supposed to be arranged to operate independently of each other. The reliability calculations for components in parallel assume independent failure modes and the existence of common mode failures would mean that the actual reliability was less than that calculated.

4

HRA steps 4

 Identify all points in a sequence of operations at which incorrect human action, or inaction (‘sins of omission'), may lead to adverse consequences for plant and/or for people.  Assign a degree of probability on a numerical scale to each event in the chain.  Aggregate the probabilities to arrive at an overall figure for the probability of human failure for the whole chain of events.  Identify steps that need to be taken to reduce the likelihood of failure at certain points by introducing organisational, procedural, ergonomic or other changes.

5

The major limitation of HRA techniques is the potential inaccuracy of the assigned values. Different techniques use different sources of data, but essentially the figures are derived from: 3

 expert opinion  historical data  experimental data. Each approach is prone to various selection and measurement biases.

6

Practical methods for improving system reliability include: 7

 Use of reliable components  Quality assurance  Parallel redundancy  Standby systems  Minimising failures to danger  Planned preventive maintenance  Minimising human error.