Plenary 5: Monitoring and Associated Technologies 1 - Measurement Methods
1:30pm - 1:50pm
Towards non-contact pollution monitoring in sewers with hyperspectral imaging
1ETH Zürich, Institute of Environmental Engineering, 8093 Zürich, Switzerland; 2Eawag, Swiss Federal Institute of Aquatic Science and Technology, 8600 Dübendorf, Switzerland; 3Headwall Photonics, Bolton, USA
Monitoring continuously pollution in the urban drainage system is challenging. Traditional approaches such as sampling campaigns or spectrometric probes have limitations. Using a hyperspectral imaging system to measure light reflection spectra of the wastewater surface is a promising approach for non-contact online measurement of pollution. We acquired hyperspectral data-cube from 18 synthetic wastewater samples. After pre-processing and pixel selection, a cross-validation partial-least square regression was used to predict turbidity (R2=0.923). Further research is planned before the SPN10 to improve the measurement, generate and analyse more data, and explore other indicators such as organic pollution.
1:50pm - 2:10pm
On-site measurement of organic micropollutants with transportable HRMS platform at combined sewer overflows
1Eawag, Switzerland; 2ETH Zürich
2:10pm - 2:30pm
Deployment of a non-invasive optical monitoring system in wastewater pumping stations
1Deltares, Netherlands, The; 2Municipality of Rotterdam, The Netherlands; 3Un poids une mesure, Lyon, France; 4Norwegian University of Science and Technology, Trondheim, Norway
Accumulation of floating solids (such as plastics, fat, oil and grease) in wastewater pump sumps is a relevant cause of malfunctioning, loss of efficiency and frequent maintenance activities. Recently, we developed a deep-learning based solution to automatically monitor the surface dynamics of floating layers in wastewater pumping stations. We present here the application of this technique to 7 pumping stations (6 in the Netherlands and 1 in France) representative of different urban drainage systems.
2:30pm - 2:50pm
On dirty water and cheap video: Real-time flow measurements derived from camera footage using an Open-Source ecosystem
1Norwegian University of Science and Technology; 2Kompetenzzentrum Wasser Berlin; 3National Technical University of Athens
Sensors used for wastewater flow measurements are robust and expensive pieces of hardware that must be maintained regularly to function in the hazardous environment of sewers. Remote sensing can remedy these issues. We utilize off‐the‐shelf cameras and convolutional neural networks to extract the water level and surface velocity from camera images directly, without the need for artificial markers in the sewage stream. In a laboratory setting, our method estimates the water level with an accuracy of ±2.48% and the surface velocity with an accuracy of ±2.08% —a performance comparable to other state‐of‐the‐art solutions.