Data scarcity in modelling and simulation of a large-scale WWTP: Stop sign or a challenge

Abstract Data scarcity can be considered as the main limitation for a more widespread utilization of mathematical models in the design, optimization and control of biological nutrient removal activated sludge systems (BNRAS). High cost and demanding workload related to experimental data and sufficient sampling campaigns make the data collection process an unpleasant necessity for managing stakeholders in modelling projects. Complicated use of online-sensors leading to frequent erroneous readings and dynamic nature of wastewater treatment processes can intensify the data scarcity problems. This paper investigates the influence of data scarcity on the development and calibration of wastewater treatment plant (WWTP) models. A straightforward methodology is proposed to address the challenges associated with data quality and quantity problems in modelling of a BNRAS in the largest Italian WWTP located in Castiglione, Italy. The plant operational modes, weather condition and sensor performance during the sampling campaigns were the main sources of the data scarcity. Influent, biokinetic, aeration, hydraulic and transport, clarifier, energy consumption and effluent sub-models were calibrated by use of the proposed extensive step-wise calibration process. The Monte Carlo analysis was performed to quantify the uncertainty of the modelling results. The proposed methodology could be implemented in engineering practice to develop and calibrate the WWTP models while it increases the awareness about modelling robustness and its characterized uncertainty to avoid bad modelling practice.