View article

Case studies are presented where data-driven models are combined with engineering knowledge used to achieve improved project results for full-scale WRRF operations. The impact of irregular sampling rates, database/historian compression and other storage-related matters can be handled with appropriate algorithms, but the applied methods are only as good as the data that are organized, formatted and modelled appropriately. The effective application of the data-driven methods to WRRF projects requires process knowledge combined with advanced modeling methods applied in software that manages data and provides useful model outputs.