When it comes to physical assets, the in operation and maintenance stages demand substantial effort and budget allocation, compared to the installation stage. This is partly due to their long life in use. More importantly, any deviation from normal operating condition could have a negative impact on safety, economics and the environment.
To ensure that the required equipment performance level is maintained, industry is shifting focus from a corrective maintenance approach to more predictive and proactive strategies. In addition, a continual maintenance improvement strategy can help retain the designed functionality of equipment.
A key tool for the visualisation of maintenance strategies in the context of an asset lifecycle is the P-F curve, which defines the path between potential failure, P, and functional failure, F. Interpreting the curve helps to predict the probability of equipment failure increases, and performance degradation, over time after an impending failure is detected. The continuation of maintenance strategies that involve operational or time-based intervals may not be an appropriate choice after point P because of the rise in failure rate.
Instead, there is the need for a continuous measurement of equipment mechanical conditions. These measurements use non-intrusive testing techniques, visual inspections and performance readings to assess asset conditions, as shown in the P-F curve. Therefore, performance-based maintenance planning allows for the scheduling of tasks only when notified by asset condition.
Most assets’ lifecycle can be modelled by the P-F curve. It is worth noting that the curve is a conceptual representation only and does not have any units or scale. However, maintenance cost could increase as the asset approaches its failed state. Measurement techniques, such as vibration analysis and ultrasonic testing, therefore indicate the relationship between the technology requirements and the stage that a potential failure is detected.
LOW, MID, HIGH
ICT enables the operational status of critical equipment to be visualised, analysed and used. This makes ICT-based monitoring strategies producers of key information within broader decision-making processes for equipment maintenance. Systems operate in a multi-variate continuous spatial field and have measurable responses. Methods of condition monitoring are capable of monitoring and analysing these responses to identify any deviation from the desired outcome, which could indicate the development of a fault or failure.
The objective of equipment condition monitoring varies across industries, as well as in application, depending on the importance of the equipment to the process:
● Condition monitoring process in equipment of low criticality usually involves offline and/or manual data collection
● Online monitoring and scanning are examples of condition monitoring application to equipment of medium criticality
● Continuous online monitoring is ideally applied to highly-critical equipment.
Key objectives of condition monitoring applications can be stated as data collection to monitor equipment performance, diagnostic identification of early warnings of potential failure, and prognostic measures to identify the nature and root cause of failure.
While failure detection, diagnostic and prognostic measures can present visual information and spreadsheets of data about equipment condition, the most important task in equipment maintenance is the interpretation of data to inform decisions on maintenance tactics development for reliability improvement and optimisation.
In the maintenance strategy process flow, the ‘maintenance strategy improvement’ entity would include elements such as equipment and maintenance performance review and defect elimination data review.
Condition monitoring can provide data for these elements, including historical failure time and the data of dependent/independent variables with significant influence on performance measures, such as vibration, temperature, and flow.
Through the application of statistical and reliability modelling, these variables and corresponding equipment response can be characterised using the best-fit probability density functions, such as the Weibull distribution – a continuous probability distribution. Therefore, these modelling outcomes can be used to improve or update existing equipment failure characteristics, such as the failure rate and mean time between failures.
Consequently, condition monitoring data can also improve the scheduling of maintenance tasks through updated frequency or time intervals and updated mechanical response parameters.
Good condition monitoring practice can contribute significantly to maintenance strategy development and lifecycle analysis through equipment reliability improvement and optimisation.
A continuous knowledge of how equipment operates and how it will fail can help maintenance professionals select a suitable strategy to prolong the functional life of the equipment.
Abayomi Obisesan is currently tasked with full implementation of reliability-centred maintenance. This includes the building of asset register, technical basis of maintenance, maintenance strategies, tasks and frequencies, bill of materials, criticality analysis, maintenance plans and cost management, as well as lifecycle optimisation within the mining, oil and gas industries, coupled with the development of engineering and management standard reporting. His employer, Optimal Asset Maintenance Solutions, is a physical asset management and engineering company, specialising in enterprise asset management solutions. Its HQ is in Aberdeen, Scotland, with subsidiaries in South Africa and Tanzania.