Wastewater treatment modelling in practice: a collaborative discussion of the state of the art

Coupling process-based modeling with machine learning for long-term simulation of wastewater treatment plant operations

Effective treatment of sewage by wastewater treatment plants (WWTPs) are essential to protecting water environment as well as people's health worldwide. However, operation of WWTPs is usually intricate due to precarious influent characteristics and nonlinear sewage treatment processes. Effective modeling of WWTPs can provide valuable decision-making support to facilitate their daily operations and management. In this study, we have built a novel hybrid model by combining a process-based WWTP model (GPS-X) with a data-driven machine learning model (Random Forest) to improve the simulation of long-term hourly effluent ammonium-nitrogen concentration of a WWTP. Our study results have shown that the hybrid GPS-X-RF model performs the best with a coefficient of determination (R²) of 0.95 and root mean squared error (RMSE) of 0.23 mg/L, followed by the GPS-X model with a R² of 0.93 and RMSE of 0.33 mg/L and last the Random Forest model with a R² of 0.84 and RMSE of 0.41 mg/L. Capable of incorporating wastewater treatment mechanisms and utilizing superior data mining capabilities of machine learning, the hybrid model adapts better to the large fluctuations in influent and operating conditions of the WWTP. The proposed hybrid modeling framework may be easily extended to WWTPs of various size and types to simulate their operations under increasingly variable environmental and operating conditions.

Automating process design by coupling genetic algorithms with commercial simulators: a case study for hybrid MABR processes

The development of commercial software and simulators has progressed to assist engineers to optimize design, operation, and control of wastewater treatment processes. Commonly, manual trial-and-error approaches combined with engineering experience or exhaustive searches are used to find candidate solutions. These approaches are becoming less favorable because of the increasingly elaborate process models, especially for new and innovative processes whose process knowledge is not fully established. This study coupled genetic algorithms (GAs), a subfield of artificial intelligence (AI), with a commercial simulator (SUMO) to automatically complete a design task. The design objective was the upgrade of a conventional Modified Ludzack-Ettinger (MLE) process to a hybrid membrane aerated biofilm reactor (hybrid MABR). Results demonstrated that GAs can (1) accurately estimate five influent wastewater fractions using eleven typical measurements - 3 out of 5 estimated fractions were nearly the same and the other two were within 7% relative errors and (2) propose reasonable designs for the hybrid MABR process that reduce footprint by 17%, aeration by 57%, and pumping by 57% with significantly improved effluent nitrogen quality (TN<3 mg-N/L). This study demonstrated that tools from AI promote efficiency in wastewater treatment process design, optimization and control by searching candidate solutions both smartly and automatically in replacement of manual trial-and-error methods. The methodology in this study contributes to accumulating process knowledge, understanding trade-offs between decisions, and finally accelerates the learning pace for new processes.

How much data is required for a robust and reliable wastewater characterization?

Water resource recovery facility (WRRF) modeling requires robust and reliable characterization of the wastewater to be treated. Poor characterization can lead to unreliable model predictions, which can have significant economic consequences when models are used to make important facility upgrade/expansion and operational decisions. Current wastewater characterization practice often involves a limited number of relatively short-duration intensive campaigns. On-going work at the Great Lakes Water Authority (GLWA) WRRF, serving 3.1 million residents in Southeast Michigan, provided an opportunity to conduct more detailed wastewater characterization over an annual cycle. The collection system includes a significant combined sewer component, and the WRRF provides primary and secondary treatment (high purity oxygen activated sludge) and phosphorus removal via ferric chloride addition. Detailed wastewater fractionation was conducted weekly over a one-year period. Daily conventional secondary influent and process operational data from that same period were used to evaluate the efficiency of various wastewater characterization strategies on the bioreactor mixed liquor volatile suspended solids (MLVSS) concentration calculated using an International Water Association (IWA) Activated Sludge Model Number 1 (ASM1) with minor modifications. An adaptive strategy consisting of a series of short-duration characterization campaigns, used to assess model fit for its intended purpose and continued until a robust and reliable model result, is recommended. Periods of unusual plant influent and/or operational conditions should be identified, and data from these periods potentially excluded from the analysis. Sufficient data should also be collected to identify periods when poor model structure, rather than wastewater characterization, leads to poor fit of the model to actual data.

Strategies for achieving energy neutrality in biological nutrient removal systems - a case study of the Slupsk WWTP (northern Poland)

The paper presents a model-based evaluation of technological upgrades on the energy and cost balance in a large biological nutrient removal (BNR) wastewater treatment plant (WWTP) in the city of Slupsk (northern Poland). The proposed upgrades include chemically enhanced primary sludge removal and reduction of the nitrogen load in the deammonification process employed for reject water treatment. Simulations enabled to estimate the increased biogas generation and decreased energy consumption for aeration. The proposed upgrades may lead the studied WWTP from the energy deficit to energy neutrality and positive cost balance, while still maintaining the required effluent standards for nitrogen. The operating cost balance depends on the type of applied coagulants/flocculants and specific costs of electric energy. The choice of the coagulant/flocculent was found as the main factor determining a positive cost balance.

Modelling and characterization of primary settlers in view of whole plant and resource recovery modelling

Characterization and modelling of primary settlers have been neglected pretty much to date. However, whole plant and resource recovery modelling requires primary settler model development, as current models lack detail in describing the dynamics and the diversity of the removal process for different particulate fractions. This paper focuses on the improved modelling and experimental characterization of primary settlers. First, a new modelling concept based on particle settling velocity distribution is proposed which is then applied for the development of an improved primary settler model as well as for its characterization under addition of chemicals (chemically enhanced primary treatment, CEPT). This model is compared to two existing simple primary settler models (Otterpohl and Freund; Lessard and Beck), showing to be better than the first one and statistically comparable to the second one, but with easier calibration thanks to the ease with which wastewater characteristics can be translated into model parameters. Second, the changes in the activated sludge model (ASM)-based chemical oxygen demand fractionation between inlet and outlet induced by primary settling is investigated, showing that typical wastewater fractions are modified by primary treatment. As they clearly impact the downstream processes, both model improvements demonstrate the need for more detailed primary settler models in view of whole plant modelling.

Calibration and validation of a phenomenological influent pollutant disturbance scenario generator using full-scale data

The objective of this paper is to demonstrate the full-scale feasibility of the phenomenological dynamic influent pollutant disturbance scenario generator (DIPDSG) that was originally used to create the influent data of the International Water Association (IWA) Benchmark Simulation Model No. 2 (BSM2). In this study, the influent characteristics of two large Scandinavian treatment facilities are studied for a period of two years. A step-wise procedure based on adjusting the most sensitive parameters at different time scales is followed to calibrate/validate the DIPDSG model blocks for: 1) flow rate; 2) pollutants (carbon, nitrogen); 3) temperature; and, 4) transport. Simulation results show that the model successfully describes daily/weekly and seasonal variations and the effect of rainfall and snow melting on the influent flow rate, pollutant concentrations and temperature profiles. Furthermore, additional phenomena such as size and accumulation/flush of particulates of/in the upstream catchment and sewer system are incorporated in the simulated time series. Finally, this study is complemented with: 1) the generation of additional future scenarios showing the effects of different rainfall patterns (climate change) or influent biodegradability (process uncertainty) on the generated time series; 2) a demonstration of how to reduce the cost/workload of measuring campaigns by filling the gaps due to missing data in the influent profiles; and, 3) a critical discussion of the presented results balancing model structure/calibration procedure complexity and prediction capabilities.

Assessment of multiple sustainability demands for wastewater treatment alternatives: a refined evaluation scheme and case study

Current estimation schemes as decision support tools for the selection of wastewater treatment alternatives focus primarily on the treatment efficiency, effluent quality, and environmental consequences for receiving water bodies. However, these schemes generally do not quantify the potential to convert pollutants in wastewater to recoverable resources. This study proposes a refined evaluation scheme for choices of wastewater treatment processes that quantifies not only adverse environmental effects but also bioenergy and nutrient recovery indices. An original means of data processing was established and clear estimate indicators were consequently obtained to allow a smooth overall estimation. An array of wastewater treatment alternatives that meet three effluent limits were used as case studies to demonstrate how the present scheme works, simultaneously, to identify optimum choices. It is concluded in the overall estimation that the lower sustainability of wastewater treatment contributed by increasingly stringent discharge demands was offset and mitigated by the resource-recovery scenarios involved, and the scenario of recovering nutrients via excess-sludge composting was of more benefit. Thus, before tightening wastewater discharge requirements, one should bear in mind the situation of multiple sustainability by setting a goal to achieve not only the greatest reduction in environmental burden but also the maximum resource-recovery benefits.