A Data Visualisation Tool for Treatment Process Monitoring in Web Browsers

Abstract

Traditional water quality–based processes of assessment using grab samples are simple to implement but far too retrospective and intermediate to adequately deal with water quality change situation. This case study demonstrated a customised software package using a web-based prototype portal with data integration, visualisation, prediction, and anomaly detection functions for complex real-time field monitoring data sets as a visualisation and decision support tool for treatment plant operators. The prototype was developed to analyse two-dimensional data obtained from a spectrophotometer (absorbance against wavelength) and data integrated from other sources (such as rainfall and temperature). It supports normal functions of time series data visualisation, such as, raw data plotting, zooming, smoothing, Boolean band etc. More importantly, it is specialised in supporting real-time monitoring of data quality, time series data comparison, anomaly detection, and future value prediction. The application of the prototype revealed that if this was placed in an operational environment, it has the potential to inform better operational decision by early detection of changes in water quality and impacts on the treatment process. This work demonstrated the potential usefulness of such real-time display and visualisation of online water quality data and the feasibility for prediction tools to provide an early warning system for process upsets.