Decision support methodology using rule-based reasoning coupled to non-parametric measurement for industrial wastewater network management

EU water framework directive [Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 Establishing a Framework for Community Action in the Field of Water Policy, Official Journal L 327, 22/12/2000 p. 0001-0073] encourages industrialists to improve the management of wastewater networks and thus reduce both pollution and treatment costs. Due to the complexity of wastewater treatment plant, the use of decision support system (DSS) becomes a complementary tool to assist decision-markers. A decision support methodology (DSM) was created which can optimise both the placement and number of physical and chemical sensors in industrial wastewater networks at petroleum refineries. Ongoing research concerns a promising method of analysing this dataflow to check whether the wastewater network is malfunctioning in any observable way. This method consists in five steps. First is to model the network to know each unit, node, etc. Second, provisional sampling points are selected in order to measure wastewater quality. Third, data is collected. Fourth, non-parametric measurement is studied: variability calculation on UV spectra. Fifth, the final selection of sampling points according to the variability study is made. With all these data, algorithm for the monitoring was developed to spot a dysfunction in the network (study of the monitoring section). These five steps of the DSM were developed and applied on a refinery wastewater treatment plant.