Power generators are critical assets in wastewater treatment plants (WWTPs) in Australia and many countries. Better managing the lifetime, minimising failures, improving reliability and availability, and reducing operating and maintenance costs of the power generation assets are still challenging topics for water utilities. This case study aims to develop power generation system reliability and availability modelling considering redundancy to minimise operation and maintenance costs. The two‐parameter Weibull model was used to assess system reliability and availability to power generation engines in WWTPs. The Kaplan‐Meier method (a time‐driven estimation technique) and the log beta‐Weibull model (which is suitable for modelling censored and uncensored data) were used to analyse and validate the modelling results. Shape and scale parameters of the Weibull models were estimated by maximising the log‐likelihood function using non‐linear optimisation. Hazard and reliability functions were calculated using the Weibull model. Results using two‐parameter Weibull, Kaplan‐Meier, and log beta‐Weibull models display low reliability and high hazard rate over time, which was associated with spark plug failure due to a suboptimal start and stop operation strategy.

No abstract

A water utility requires myriads of data for effective decision-making. As the sources and ranges of data are becoming increasingly complex, the use of a metadata framework can play a significant role in effective data management. Using case study method, this research analyzed data needs of a water supply system in a small town in South Australia and designed a demo portal of a metadata framework. As part of the case study, the project team undertook a broad investigative approach using focus group (interviews), observation, exploration of potential data sources, identification of knowledge leaders and information technology systems. The metadata framework comprised two separate but interconnected metadata groups, (1) metadata elements to describe the metadata source and (2) metadata elements to describe the datasets held in each data source. The metadata framework was populated to describe data sources and data held in each of the sources. The data catalogue created by this process showed that it was accomplishable and appropriate to describe data sources and datasets via a metadata framework.

Power generation engines in wastewater treatment plants (WWTPs) are critical and strategic assets as major elements of an effective energy management system. Therefore, water utilities seek a smart tool to optimise the maintenance program on overall cost and reliability of these assets. In the previous study, a reliability model with consideration of redundancy was developed to estimate failure modes, reliability and availability of power generation engines in a WWTP. This study examines a joint model of reliability index using the Weibull model, operating maintenance cost using the steepest descent method and quasi‐Newton algorithm. This research is the first application of this joint model to power generation engines in WWTPs. A genetic algorithm has been employed to analyse and validate the modelling results as a global optimisation method. Results of the optimal solution are compared with conventional maintenance regimes of these engines based on recommended original equipment manufacturer (OEM). Results show an average of 22.3% reduction in maintenance costs of three engines after the implementation of the proposed cost‐optimization model. In addition, the manufacturer's recommended maintenance regime is not an appropriate maintenance strategy due to differences between the assumed and actual operating conditions. Also, hazard and reliability estimate using historical failure, operating and maintenance cost data has been ignored in the recommended OEM maintenance strategy.