Prediction models for manganese, iron and ammonium in raw water for a drinking water treatment plant Butoniga (Croatia)

Abstract

Drinking water treatment plant Butoniga is one of the main water supply facilities for potable water in Istria (Croatia). Water for treatment process is captured from the Butoniga reservoir which is a small and relatively shallow reservoir. As such, reservoir is very sensitive to eutrophication and degradation processes caused by climate change and human activities in surrounding watershed. In summer months during tourist season, when water production is the highest and the reservoir water level is the lowest, the water temperature is the most critical parameter during treatment process. Because of this, to capture colder water, raw water for treatment is captured from the lowest water intake which captures water from the lowest water layer in the Butoniga reservoir. This layer have another problem, which are increased concentrations of manganese, iron and ammonium under lower pH values.  Regarding this, prediction models seven days in advance were made for manganese, iron and ammonium, which are some of the most critical parameters during summer months and have significant influence on treatment process of raw water. For modelling purposes, machine learning software Weka was used to build models in form of model trees. Obtained prediction models for manganese, iron and ammonium have high accuracy compared to the measured data with a good prediction of the peak values. Therefore, obtained models can help in optimization of treatment processes on treatment plant which are depending on the quality of raw water in Butoniga reservoir.