Planning plant maintenance – how to assess its long-term impact on safety and economics

The goal of Asset Life Management is to achieve the best return on investment in major physical assets. An underlying concern for a plant operator in the private sector is to deploy its plants in a way which maximises the value to shareholders. In practice, profitability can be enhanced through a combination of improved plant performance (output and availability) and reduced operational costs and capital expenditure. Nuclear utilities around the world have adopted different approaches to the key issue of enhancing profitability whilst maintaining an acceptable level of safety. An optimum Asset Life Management strategy aims to secure an appropriate balance between plant performance enhancement and cost reduction. Any measures adopted to improve profitability should be demonstrated to have a net benefit over the full life of the plant: for example, there is a need to avoid the short term cost savings achieved through reductions in plant maintenance being offset by failure of a component before the end of its intended service life. Ageing management programmes, such as the US Nuclear Regulatory Commission maintenance rule and the Periodic Safety Reviews carried out in the UK, require effective ageing management of safety critical or safety related areas. The aim of these management programmes is to take account of the potential for increasing uncertainty in system/component condition and performance as ageing processes develop, through a demonstration of continuing plant safety. The ageing management plan generally includes some combination of a performance review, surveillance testing and preventive maintenance. Whilst safety provides a natural focus for developing the plant management plan, the impact on wider business issues of any decisions taken should also be examined to ensure an overall economic benefit. This raises issues such as the following: • The incidence and cost of past failures, repairs and refurbishments, and downtime due to component ageing. • The likelihood and extent of future failures, their remedial costs and the confidence which can be placed in these estimates. • The likely loss of cash flow from future failures, bearing in mind the future electricity trading market in an era of industry deregulation. The overall objective is to counter developing economic risks as key systems/components degrade in ways which are more (or less) understood and whose condition is more (or less) uncertain. This is not a simple process and it is becoming increasingly clear that practical software tools can provide valuable support to management decisions. MAINTENANCE PLANNING A risk based methodology provides a rational basis for evaluating the impact of various Asset Life Management strategies on overall business objectives. Taking the important area of plant maintenance management, a risk based approach can be adopted to examine allowed outage times and to optimise test intervals. This approach has the potential to deliver significant economic benefits if it can be demonstrated that a more relaxed testing schedule can be adopted with no adverse impact on system availability. Adoption of such an approach to maintenance planning requires the following information: • A PSA (probabilistic safety assessment) for either the whole plant or a single system. • The capability to model time-dependent and configuration changes in the PSA. This should include a maintenance model, information on the maintenance period and effectiveness, component ages, failure rates with age etc. • Software to derive the risk based test information. PSA Tools for Managers (PTM) is a software tool developed by AEA Technology which can be used to optimise test intervals and to determine the risk implications of different testing regimes. The tool is best applied as part of a consultancy package which takes into account the practical and statutory constraints on the testing regimes, since it is recognised that the predicted benefits cannot always be achieved as a result of other constraints. The system selected for the demonstration of these principles was the back up cooling system for the Hunterston B AGR. This system provides a segregated and diverse source of cooling water in the event of failure of the main boiler feedwater supply. A PSA-based test optimisation approach was adopted by applying the APSA program (which is part of PTM) to the existing system fault tree supplied by British Energy. APSA was run until the system unavailability reached a preset upper limit. A risk reduction worth calculation was then performed to determine which components, if tested, would reduce the system unavailability the most, and these components were tested until the system unavailability fell to a preset lower limit. Repetition of the process for the required period of time results in a net output in the form of an optimised test schedule. The results showed that 89% of the mean system unavailability was attributable to only 5 of the 179 components included in the PSA. The results indicated strong evidence for over testing: an alternative testing schedule could be devised, with no adverse impact on risk, in which there is more frequent testing of only 5 components (at 4 month intervals instead of the 2 years currently specified) but in which many components which are now tested weekly or monthly are tested at greater than 2 year intervals. Overall, the number of tests between major outages could be reduced from about 6000 to about 200, with a concurrent decrease in mean system risk of 30. The benefits of adopting such an approach include both a significant cost reduction and the potential to operate the plant more flexibly through a justification of changes to the Allowed Outage Times. The demonstration was only for one system and clearly there would be even greater scope for changes in maintenance practice if the plant as a whole was examined. STRATEGIC DECISION SUPPORT The use of the PSA as part of the PTM analysis moves the PSA from being simply a tool to demonstrate regulatory compliance to being a valuable tool to support operational decisions: this significantly increases the value obtained from the substantial investment already made in constructing the PSA for the plant. However, there is scope for making further use of the PSA, through including it in a model which can be used to support strategic decisions for the plant as a whole. As already noted, the impact of any decisions taken with regard to one item of plant needs to be assessed as part of it’s contribution to plant performance and profitability over the full plant life. This is not a simple assessment and clearly an overall framework for decision taking is desirable. This will structure all the conflicting decisions which must be addressed and enable an overall balance between cost and benefits (economic and safety) to be achieved. Such tools are becoming increasingly important because of the rapidly changing circumstances in which utilities operate and the complex nature of the decisions which must be taken. The first step in making such an assessment is to identify clearly what are the key strategic areas for the utility as a whole and to identify the purpose of each asset in that context: that is, the critical question which must be answered is “what are the assets for and how do they contribute to business performance?” A decision can then be made on what must be modelled to enable the performance of each asset to be linked directly to critical company values, thus allowing an assessment to be made of the contribution to business performance which is made by that asset. This assessment requires an analysis of financial impact, process impact (for example, maintenance practices) and data collection and analysis. Using scenario planning and the decision support software, the impact of a range of possible decisions can be identified over the life of the plant. In addressing these types of issues, a risk based approach can continue to help to logically structure decision making and to support engineering conclusions in terms that are readily understandable by corporate decision makers. This requires a broadening of the measures of risk to include, for example, economic and reputation factors. The diagram illustrates a framework which has been developed by AEA Technology to link management options to business impact. Encapsulating this approach within customised software has supported the decision making process by enabling the rapid handling of credible scenarios and data uncertainties (with the software ranging from spreadsheet models through to major information systems). Whilst such systems are still in their infancy, there is considerable opportunity for innovation building on existing technology and risk based methods to enhance the business performance of nuclear utilities through more robust decisions. Successful implementation requires: • Risk based methodologies to assess the impact of failure, in terms of business impact (safety, lifetime, availability etc.). This builds on the PTM model. • Technologies which provide a means of measuring current condition and also models for predicting the future operational performance of components. This combination of methodology and technology enables improved asset replacement programmes to be devised, maintenance strategies to be optimised and spares holdings to be minimised, all within the wider context of improving business performance for customers in capital intensive industries. The framework also has important implications for assessing data collection requirements and associated technology development. Through the explicit linking of management options and business impact, it is possible to identify where focused technical development will improve overall profitability (and, conversely, those areas where development will have lower returns). In summary, significant benefits could be gained by the adoption of advanced methods of evaluating the contribution of assets to overall business performance. Such methods build on existing PSAs for the plant and have the potential to provide increased safety assurance combined with economic benefits.