Water quality management
Water quality monitoring largely takes place by instantaneous sample taking and analysis in laboratory. If deviations are discovered a new sample is taken 1 – 2 days later and action is taken. Water quality issues are established by customer complaints and doctor reports on illnesses. In the last 10 years online and real time sensors for specific parameters and screening sensors for generic parameters have been developed. If sensors for specific parameters are used, many different sensors are needed. Therefore, for water quality management specific sensors are not an option on their own. Generic sensors and a combination of generic and specific sensors are a promising solution (less data, less sensors, less investments and costs; maintenance). A number of specific and generic screening sensors have been developed and tested successfully in lab and pilot scale situations showing detection of relevant parameters. Relevant developments include:
- Optiqua EventLab generic screening sensor: the sensor measures differences in water quality online based on refractive index;
- Specific S::Can spectrometry sensor: The sensor measures turbidity, Nitrate, Colour and organic parameters (Total Organic Carbon and Dissolved Organic Carbon) online based on absorption;
- Specific Intellisonde sensor from Intellitect: the sensor measures flow, temperature, free chlorine, monochloramines, dissolved oxygen, pH, ORP, conductivity, color, turbidity, pressure, ISE online based on differences in potential;
- Generic Mycometer biologic sensor: the sensor measures biologic activity based on enzyme activity;
- Aqua Explorer generic sensor for biologic activity: the sensor measures biologic activity based on antibodies.
Currently, sensors are applied ad hoc and interpretation insights are linked to the individual sensor.
For determining the source locations of contaminations and effects of contamination water, water demand prediction models and hydraulic network simulations can be used. However, in practice a number of limitations exist. Building a detailed water quantity model for a network will take months of time. Once developed, the model is tweaked and tuned and not adapted later on. As demand patterns will change and the network itself will be adapted, hydraulic models are almost immediately out of date when ready. Therefore, continuous calibration is needed for determining the sources location and effects of contaminations, but does hardly take place.
For localisation and effect prediction of water quality events KWR’s SIMDEUM model has been developed for modelling of water usage at household level. This model is needed for generating usage patterns for water modelling at street level. Thus the framework for water quality models exists for water distribution networks, specific settings and effects on water quality issues, however, require further study.