A new demo-software application simulates the efficiency with which two common treatment processes (reverse osmosis and activated carbon filtration) remove specific micropollutants from surface water under different operating conditions. The water companies Evides, Oasen, Vitens, Waternet and WML, carbon supplier CabotCorp and KWR have developed this software model over the past year and a half within TKI’s AquaPriori project. The new application is based on process models in which the relationships between substance properties, process conditions and the behaviour of a substance are inputted, and which then produces results for 2,400 substances contained in the associated databank.
Thousands of micropollutants in surface water
Surface water contains – despite legal and regulatory provisions – thousands of substances in very small quantities: so-called micropollutants. It is extremely costly and time-consuming to target experiments at each micropollutant to determine how effectively it is removed during water treatment. But water companies for example want to have a good idea of which substances they do, or can, effectively remove if they are present in the water. This is why the above-mentioned partners have developed this alternative to measurement experiments.
Database with 2,400 substances
The demo-software application permits experts at water companies and at suppliers of treatment processes and materials to simulate the removal of specific substances on a practical scale. The demo-software uses 60 known membrane-substance combinations for reverse osmosis and 100 known adsorption isotherms for activated carbon filtration. With these data, the software can predict the removal efficiency for 2,400 substances (contained in the databank supplied) under specific process conditions. This is important information, for instance, when a new substance is detected in the surface water.
Follow-up research under preparation
Consideration is currently being given to the project’s follow-up research. This research could for instance focus on the ways of making the application more accessible for users, on making it suitable for other treatment processes, or further improving its foundations. The latter would for example require more research into the applicability of the statistical approach to (groups of) substances, into the quantification of the model’s reliability and validation tests. Anyone interested in participating in follow-up research is invited to contact KWR.