Efficiency and efficacy of ozonation for disinfection at the Weesperkarspel drinking water treatment plant

Abstract

Steering the ozonation process for drinking water production at Weesperkarspel (Waternet, Amsterdam) is a challenge. The aim is to keep disinfection at the desired level: a constant disinfection is optimal, as using more ozone than necessary brings along costs and bromate formation. But verifying disinfection also remains a challenge: grab samples take time to analyze, may be inaccurate and do not enable direct response to changes in the influent water.

To tackle these problems, two steps were taken: quantitative disinfection research was conducted through literature study to gain better insight into disinfection kinetics of ozone in water. Using this knowledge, different steering mechanisms for the ozonation process were modeled to see which gives the best results. Disinfection research has shown that there is a significant difference (p > 0,005) in inactivation between lab-grown bacteria and environmental bacteria, where inactivation constants can be 7 times larger for lab bacteria compared to environmental bacteria, which must be taken into account during disinfection experiments and research. The relation between CT (ozone concentration [mg/L] * contact time in the reactor [min]) and disinfection is usually assumed to be linear, but environmental data from this research shows this relation is more likely to be logarithmic.

Steering the ozonation process was carried out in four ways: steering for a constant ozone dosage, steering for a constant CT, steering for a constant disinfection and steering for a constant percentage decrease of UV254 absorbance of the water. Variation in bromate formation, ozone dosage and disinfection differs per steering mechanism: steering for a constant disinfection is the best way to meet disinfection goals while keeping ozone costs low. Five percent ozone could be saved on a yearly basis compared to steering for a constant ozone dosage.