Demo case #3: Intelligent control for wastewater treatment (The Netherlands)


Context: WATNL’s Wastewater Treatment Plant West (WWTP West) has a capacity of 1 Million population equivalent and serves the city of Amsterdam, itself a FIWARE supporter city as part of the Open and Agile Smart Cities Initiative. The process automation (PA) of WWTP West will be renewed and the IT architecture of WATNL is in fast development. This to enable WATNL to optimally benefit from implementation of upcoming technologies such as Artificial Intelligence (AI), Internet of Things (IoT), Business Intelligence (BI) and streaming analytics to improve operational efficiency and asset management, and to reduce climate impact and costs. Currently the control loops of WWTP West are for a large part dedicated to a single process. With the use of real-time plant data, process models and external data sources a more optimal plant-wide control can be achieved. However, this is not possible with the current single process table or PID controllers.

Therefore, WATNL will make one lane in WWTP West available for introducing and testing of additional sensors, data driven control strategies and decision support based on newly to be developed AI models and data fusion. The objective is to minimise N2O emission, energy use and sludge production at minimum costs while meeting effluent water quality targets. Minimising N2O emission has a large potential for further reduction of the CO2 emission and thus for the ambition of WATNL to become climate neutral.

Challenges: The first challenge for WWTP West is the reduction of N2O emission and energy demand to achieve climate neutral, which can be largely overcome via a more stable operation. The second challenge is the growing number of inhabitants of Amsterdam will increase the influent flow. By increasing the capacity of the current treatment plant with a more stable operation, investments in eight aspects can possibly be postponed leading to significant costs saving.

Role of Fiware4Water: FIWARE applications on WWTPs are not yet developed. The project will increase knowledge through building historical data based on real-time sensing and use of data analytics to make predictions of the dynamic behaviour of the treatment processes. The sensors will be installed in one of the seven lanes of the full-scale WWTP West and sensor readings will be verified with lab measurements. FIWARE tools will be developed and used to determine oxygen set points, nitrification capacity or in-situ respiration. Robust and transferable control algorithms to reduce N2O emissions, to reduce energy demand and to improve overall treatment efficiency are relevant for many treatment plants around the world. FIWARE compatibility ensures the introduction of the tools easier at different sites.

Advantage include: (i) increased knowledge of the state of the process due to previously unknown values of essential process variables (ii) optimal control settings determined in a multivariate process for more substantiated decisions for control, taking in to consideration the future state of the process for optimal wastewater treatment (interdependency between effluent water quality, energy use, greenhouse gas emissions and costs) (iii) lower treatment costs and GHG emissions.



26/11/2021 - Demo Case Update from KWR-Waternet-EURECAT (2/2)
22/11/2021 - Demo Case Update from KWR-Waternet-EURECAT (1/2)
28/05/2021 - Demo Case Update from Waternet - Wastewater Treatment Plant Amsterdam West
17.12.2020 - Demo Case Update from Waternet - Wastewater Treatment Plant Amsterdam West
02.10/2020 - Demo Case Update from KWR - Amsterdam West Wastewater Treatment Plant (WWTP)
10.06.2020 - Demo Case Update From Waternet – Wastewater Treatment Plant Amsterdam West 
03.26.2020 - Demo Case Update From Waternet - Wastewater Treatment Plant Amsterdam West