Development of a data driven FDD approach for building water networks: Water distribution system performance assessment rules

Abstract

While fault detection and diagnosis is a popular tool in the process industry, its application in building water distribution systems is however still largely absent. In this study, a new set of Water Distribution System Performance Assessment Rules (WDSPARs) were developed to identify common faults in a building water distribution system. The WDSPARs comprise a three-phase process which can be applied using flow and pressure sensor signals obtained in real-time and/or via analysis of historic data in conjunction with knowledge of water distribution system layouts. The performance assessment rules originated from analysis of behaviour in water consumption at two non-residential pilot sites over a 6 month trial. Implementation of WDSPAR at the two pilot studies revealed a number of faults and cases of non-optimal performance which were diagnosed and costed accordingly. The WDSPAR approach is intuitive and can be easily integrated into existing building management systems using sensor data. This study serves as the first practical guide for the implementation of the WDSPAR approach for adoption by large non-residential building end-users. Using the WDSPARs, the case studies outlined in this paper demonstrate 62% savings in water consumption which resulted in energy and carbon emission savings of the order of 50 kW h and 29.9 kg.CO2 per day respectively.