In-Sensor-Experiments are proposed as a means of providing highly informative data concerning the bioprocesses occurring in N-removal systems. It is highlighted that the changing nature of both wastewater and activated sludge enforces the continuous adjustment of the proposed In-Sensor-Experiments to maintain the quality of the sensor's outputs. Central to the automated design of the experiments performed in these adaptive sensors is a mathematical representation of the processes occurring in the device. This model is continuously updated on the basis of previously acquired data. It is illustrated how different design criteria (objective functions), such as most reliable model selection capability or minimal parameter variance, influence the experimental designs. The concepts are illustrated with real-life data from two types of In-Sensor-Experiments. First, short-term (fed-)batch respirometric experiments are used to estimate the biokinetics of the nitrifying population. Second, a new device is presented in which a sensing element is placed at the end of a plug-flow reactor, hence the term “Plug-Flow-Sensor”. The optimally designed and continuously updated variation of the flow rate through the plug flow reactor results in a retention time distribution. This allows us to monitor a variable as function of the reaction time and this within a narrow (user-specified) window of reaction times, increasing the measuring frequency and accuracy compared to equivalent batch experiments. As a first example, ORP is applied as measured variable. By using ORP “nitrate knees” can be detected after a certain reaction time. This information is an indicator of the denitrifying capacity of the sludge.
Research Article|January 01 1995
Optimal design of in-sensor-experiments for on-line modelling of nitrogen removal processes
Water Sci Technol (1995) 31 (2): 149-160.
P. Vanrolleghem, F. Coen; Optimal design of in-sensor-experiments for on-line modelling of nitrogen removal processes. Water Sci Technol 1 January 1995; 31 (2): 149–160. doi: https://doi.org/10.2166/wst.1995.0091
Download citation file: