Water treatment plants are in general robust and designs are based on the performance of individual processes with pre-set boundary conditions. It is assumed that an integral approach of the entire treatment plant can lead to more efficient operation. Taking into account the developments in sensoring, automation and computation, it is a challenge to improve quality and reliability of the treatment plants and to make maximal use of the installed infrastructure, postponing new investments.
At Amsterdam Water Supply (AWS), the first steps have been taken to come to an integral dynamic model of the total water treatment plant and the use of this model as an instrument for integral control. The parameters influencing the performance of the water treatment process will be incorporated in an overall model evaluating the goal factors quality (good, constant and reliable), quantity, costs, environmental impact (low residuals level), redundancy and flexibility. For several individual processes at AWS models have already been developed during the last few years, like models for the ozone process, biological activated carbon filtration and pellet softening. For the final calibration and validation pilot reactors are automated and on-line data are collected. Criteria for evaluation are developed to realise an optimal control of the individual processes in interaction with the goal factors of the total treatment process.