Benchmark Simulation Model No. 1 Research Articles

Multiobjective Optimization of Papermaking Wastewater Treatment Processes under Economic, Energy, and Environmental Goals.

Due to the heterogeneity of recycled paper materials and the production conditions, pollutants in papermaking wastewater fluctuate sharply over time. Quality control of the papermaking wastewater treatment process (PWTP) is challenging and costly. As regulations are also growing about the environmental effects of the PWTP on greenhouse gas (GHG) emission, energy consumption, etc., the PWTP formulates a complex multiobjective optimization problem. This research established a multiagent deep reinforcement learning framework to simultaneously optimize process cost, energy consumption, and GHG emission in the PWTP, subjected to the effluent quality, to realize economic, energy, and environmental (3E) goals. The biological treatment process of wastewater in paper mills was simulated using benchmark simulation model no. 1 (BSM1). The data generated based on the BSM manual was utilized for model training, and real data acquired from a local papermaking factory was used to estimate the model performance. The results show that the proposed method outperforms conventional techniques in identifying the best control strategies for multiple targets.

Unified control of diverse actions in a wastewater treatment activated sludge system using reinforcement learning for multi-objective optimization

The operation of modern wastewater treatment facilities is a balancing act in which a multitude of variables are controlled to achieve a wide range of objectives, many of which are conflicting. This is especially true within secondary activated sludge systems, where significant research and industry effort has been devoted to advance control optimization strategies, both domain-driven and data-driven. Among data-driven control strategies, reinforcement learning (RL) stands out for its ability to achieve better than human performance in complex environments. While RL has been applied to activated sludge process optimization in existing literature, these applications are typically limited in scope, and never for the control of more than three actions. Expanding the scope of RL control has the potential to increase the optimization potential while concurrently reducing the number of control systems that must be tuned and maintained by operations staff. This study examined several facets of the implementation of multi-action, multi-objective RL agents, namely how many actions a single agent could successfully control and what extent of environment data was necessary to train such agents. This study observed improved control optimization with increasing action scope, though control of waste activated sludge remains a challenge. Furthermore, agents were able to maintain a high level of performance under decreased observation scope, up to a point. When compared to baseline control of the Benchmark Simulation Model No. 1 (BSM1), an RL agent controlling seven individual actions improved the average BSM1 performance metric by 8.3%, equivalent to an annual cost savings of $40,200 after accounting for the cost of additional sensors.

Interpretable GHG emission prediction for papermaking wastewater treatment process with deep learning

The content of pollutants in papermaking wastewater is highly varied due to heterogeneous raw materials, product types, and biochemical process conditions, resulting in significant uncertainty of GHG emissions mitigation in the treatment process. This study established a modified Benchmark Simulation Model no.1-based mechanism model to reveal the characteristics of GHG emissions and support GHG reduction in the papermaking wastewater treatment process. Due to its heavy dependence on the Benchmark Simulation Model No. 1 (BSM1) and plentiful parameters, it was instead further proposed Deep Neural Networks (DNN) models and Long Short-Term Memory (LSTM) models to estimate GHG emissions with high accuracy (R2 > 0.94) greatly reduced computational complexity (8920.6 times). The global sensitivity analysis upon DNN revealed key factors for GHG emission with strong explainability, supporting the development of GHG reduction control strategies for the papermaking wastewater treatment process.

Towards stable and efficient nitrogen removal in wastewater treatment processes via an adaptive neural network based sliding mode controller

Advanced controllers often offer an innovative solution to proper quality control in wastewater treatment processes (WWTPs). However, nonlinearity and uncertain disturbances usually make the conventional control strategies inadequate or impossible for the stable operations of WWTPs. To guarantee the stability of ammonia nitrogen concentration (SNH) control in WWTPs, a direct adaptive neural networks-based sliding mode control (ANNSMC) strategy has been proposed in this article. A sliding mode controller is designed and implemented with the help of an adaptive Neural Network (ANN), named Radial Basis Function Neural Network (RBFNN), which can approach the desired control law accurately. Also, the stability of a system installed with the ANNSMC is analyzed by using the Lyapunov theorem, which ensures system robustness and adaptability. Additionally, to deal with high energy consumption and low treatment efficiency problems in the wastewater denitrification processes, this paper proposes a dual-loop denitrification control strategy and validates it in the Benchmark Simulation Model No.2 (BSM2) platform. The strategy can strengthen the denitrification efficiency by collaborating the SNH with nitrate nitrogen (SNO) concentration in the WWTPs properly. The experimental results demonstrate that the proposed strategy can obtain remarkable stability and robustness, reducing energy consumption effectively compared with other standard and advanced control strategies.

Modeling key intermediates during anaerobic digestion of lipid rich kitchen waste with an extended ADM1.

Quantitative dynamics of the key intermediates, gases and carbohydrates during anaerobic digestion of different lipid rich kitchen waste and lipid rich model kitchen waste were modeled. Six batch reactors loaded with 25g l ( 39 l ) kitchen waste and model kitchen waste during a batch experiment were considered in simulation. Observed dynamics of carbohydrates, volatile organic acids and gases were described by an extended benchmark simulation model no. 2 (BSM2). In this study the extended BSM2 included a more detailed -oxidation for prediction of caproic acid. Furthermore, the extensions included carbohydrate digestion with an additional intermediate before propionic acid was released. In addition, a novel simplification approach for initial pH estimation was successfully applied. For parameter estimation a Markov Chain Monte Carlo method was used to obtain parameter distributions. With the presented model it was possible even with no calibrated data to predict point of times of intermediates maxima and propionic acid with relative stable concentration over several days for kitchen waste.

Enhancement of energy and cost efficiency in wastewater treatment plants using hybrid bio-inspired machine learning control techniques

Global population growth drives demand for clean water, energy, and wastewater production in urban areas. Most existing wastewater treatment plants (WWTPs) are energy- and cost-intensive. To achieve the UN Sustainable Development Goals (SDGs) by 2030, many regions have implemented strict policies for energy consumption and contaminated water discharge, promoting environmental sustainability. This paper attempts to enhance the achievement of SDGs in the wastewater treatment industry by proposing a novel method for hierarchical control of wastewater treatment plants to ensure energy and cost efficiency with less knowledge of the process model, fewer measurements, and fewer control loops. This controller is the combination of a machine learning control technique with a bio-inspired computational algorithm. To evaluate and validate the performance of the proposed controller on the WWTP, the benchmark simulation model no. 1 (BSM1) has been used. The most significant contributions of this paper show that 488.96 KWh/d, 3620.08 Kg/d, and 3454.55 KWh/d of aeration energy cost index, sludge production cost index, and operational cost index are respectively saved when compared to the default control strategy (DCS). Also, 893.69 Kg pollution units per day of effluent fine-related costs have been successfully recovered when compared to the DCS. The experimental results illustrate that the proposed control yields significant amelioration of the energy and cost profile of the WWTP with no effluent violation in dry, rain, and storm weather conditions.

Modeling greenhouse gas emissions from biological wastewater treatment process with experimental verification: A case study of paper mill

Wastewater treatment plants (WWTPs) have been regarded as the main sources of greenhouse gas (GHG) emissions. This study compares the influent characteristics of industrial wastewater represented by the WWTP of paper mill and that of domestic sewage represented by the Benchmark Simulation Model No. 1 (BSM1) under stormy weather. The various sources of GHG emissions from the two processes are calculated, and the contribution of each source to the total GHG emissions is assessed. Firstly, based on the mass balance analysis and the recognized emission factors, a GHG emission calculation model was established for the on-site and off-site GHG emission sources from the WWTP of paper mill. Simultaneously, a GHG emission experimental model was established by determining the dissolved concentrations of carbon dioxide (CO2) and nitrous oxide (N2O) in the papermaking wastewater, to verify the accuracy of the developed GHG calculation model. Subsequently, an optimum aeration rate for the paper mill was investigated to comply with the discharging norms. Under the optimum aeration rate of 10 h−1, the obtained calculation accuracies of CO2 and N2O emissions were 94.6 % and 91.1 %, respectively. The mean total GHG emission in the WWTP of paper mill was 550 kg CO2-eq·h−1, of which 44.6 % came from the on-site emission sources and 55.4 % from the off-site emission sources. It was also uncovered that the electrical consumption for aeration was the largest contributor to the total GHG emissions with a proportion of 25.2 %, revealing that the control strategy of the aeration rate is highly significant in reducing GHG emissions in WWTP of paper mills.

Hybrid supervised hierarchical control of a biological wastewater treatment plant.

In wastewater treatment intensification, hierarchical control structures are developed to improve the plant's performance. This paper proposes two novel hybrid supervised hierarchical control structures for specifying the dissolved oxygen concentration in the last aerobic reactor of the wastewater treatment plant (WWTP) based on the nitrification rate and the ammonia level in this reactor. These structures combine the optimum disturbance rejection PI control (OPI), adaptive neuro-fuzzy inference system (ANFIS), and genetic algorithms (GA) to reduce energy consumption and operational costs, improve effluent quality, and reduce the number and percentage of times the established maximum concentration of pollutants in the effluent of the WWTP is violated. The proposed control strategy is implemented and evaluated using benchmark simulation model no. 1 (BSM1). The OPI-ANFIS-GA configuration significantly enhances effluent quality in dry, rainy, and stormy weather conditions, reducing total nitrogen violations by 50.17%, 63.35%, and 47.35%, respectively. Then, 6.79% and 7.12% of aeration energy and 1.44% and 1.46% of operational costs are reduced in dry and rain weather conditions. The OPI-ANFIS configuration enhanced significant energy savings and a cost reduction in storm weather conditions. Both configurations led to a 49.89% decrease in total suspended sludge (TSS) during stormy weather conditions. The proposed controller significantly improves the performance of the WWTP in all weather scenarios compared to the default controller and similar controllers found in the literature.

A multi-subsystem collaborative Bi-LSTM-based adaptive soft sensor for global prediction of ammonia-nitrogen concentration in wastewater treatment processes

Ammonia-nitrogen concentration is a key water quality indicator, which reflects changes in pollutant components during wastewater treatment processes. The timely and accurate detection results contribute to optimizing control and operational management of wastewater treatment plants (WWTPs), but current detection methods only focus on the effluent location. This paper proposes a multi-subsystem collaborative Bi-LSTM-based adaptive soft sensor to achieve the global prediction of ammonia-nitrogen concentration. Firstly, the wastewater treatment process is divided into several independent subsystems depending on the reaction mechanism, and the variable selection is performed using mutual information. Subsequently, the bidirectional long short-term memory network (Bi-LSTM) is employed to construct a model for predicting ammonia-nitrogen concentration within each subsystem, and the outputs between neighboring subsystems are incorporated as a set of new variables added into the training dataset to strengthen their connection. Finally, to address performance degradation caused by environmental factors, a probability density function (PDF)-based dynamic moving window method is proposed to enhance the robustness. The effectiveness and superiority of the proposed soft sensor are validated in the Benchmark Simulation Model no. 1 (BSM1). The experimental results demonstrate that the proposed soft sensor can accurately predict the global ammonia-nitrogen concentration in the face of different weather conditions including sunny, rainy, and stormy days. This study contributes to the stable operation of WWTPs with higher treatment efficiency and lower economic costs.

Data-Driven Robust Adaptive Control With Deep Learning for Wastewater Treatment Process

Due to high complexity and time-variant operation, as well as increasingly requirements for water quality, stability and reliability, wastewater treatment process (WWTP) is regarded as an adaptive control problem. In this study, a data-driven adaptive control with deep learning (DRAC-DL) is developed to improve the operational performance of WWTP. First, a feedback controller is designed to construct the closed-loop control scheme. Second, an adaptive deep belief network (ADBN), based on the data-driven self-incremental learning strategy, is proposed to approximate the ideal control law. Third, the stability of DRAC-DL scheme is analyzed in detail. The main advantage of DRAC-DL lies in its improved robustness and efficiency, which benefit from Lyapunov-based closed-loop strategy and efficient ADBN controller. Finally, the feasibility and applicability of DRAC-DL are verified by two parts: 1) Simulation on nonlinear system; and 2) Application to WWTP on the benchmark simulation model No.1 (BSM1). The experimental results show the applicability and effectiveness, among which DRAC-DL reduces the output fluctuation (Var) by no less than 82% and realizes the better stability and robustness.

Towards low-carbon papermaking wastewater treatment process based on Kriging surrogate predictive model

Papermaking wastewater contains a diverse range of organic carbon, nitrogen and other pollutants, which could lead to the emission of greenhouse gases (GHG) during biochemical treatment process. Despite the importance of decarbonizing the papermaking wastewater treatment process, the mechanisms underlying the generation of greenhouse gases remain elusive and no models currently exist to describe the process. To address this issue, this study employed benchmark simulation model No. 1 (BSM1) to simulate, the papermaking wastewater treatment process, and applied the steady-state output value of simulation to calculate the GHG emissions. Kriging surrogate model was applied to establish the emission model for GHG in the papermaking wastewater treatment process. The model provides a reference for the control and optimization of GHG emission reduction and a high accuracy was achieved with determination coefficient (R2) of above 96%. Further analysis of the parameter impact of process control on GHG emissions and the optimization effects of controllable variables contributes to the goal of reducing GHG emissions in the papermaking wastewater treatment process.

Static calibration of wastewater treatment plant models: Investigating calibration processes and objective functions

Mathematical modeling provides a reliable tool for optimal management and operation of wastewater treatment plants (WWTPs). Accurate calibration is crucial for producing reliable results, and static calibration is a critical initial step for this purpose, ensuring enough accuracy and minimizing later difficulties. The two main areas to investigate in static calibration are the appropriate calibration process and objective function. This study proposes a three-stage methodology for static calibration of model parameters. Nine different calibration objective functions considering various target variables are investigated to evaluate how different objective functions can affect the calibration process accuracy. The proposed approach is applied in a real WWTP with a carousel oxidation ditch for secondary treatment. Benchmark Simulation Model No.2 (BSM2) is modified to be used for oxidation ditch modeling. The calibration results show less than 3 % prediction error for COD, TN, and TSS, and 6.7 % mean overall error. The validation resulted in 22.5 %, 10.6 %, and 31.5 % MARE (Mean Absolute Relative Error) for effluent COD, TN, and, TSS, respectively, indicating successful calibration. Comparing different objective functions indicates that in static calibration, the objective functions that consider more targets usually yield more accurate results in a relatively acceptable time.

Systematic Performance Evaluation of Reinforcement Learning Algorithms Applied to Wastewater Treatment Control Optimization.

Treatment of wastewater using activated sludge relies on several complex, nonlinear processes. While activated sludge systems can provide high levels of treatment, including nutrient removal, operating these systems is often challenging and energy intensive. Significant research investment has been made in recent years into improving control optimization of such systems, through both domain knowledge and, more recently, machine learning. This study leverages a novel interface between a common process modeling software and a Python reinforcement learning environment to evaluate four common reinforcement learning algorithms for their ability to minimize treatment energy use while maintaining effluent compliance within the Benchmark Simulation Model No. 1 (BSM1) simulation. Three of the algorithms tested, deep Q-learning, proximal policy optimization, and synchronous advantage actor critic, generally performed poorly over the scenarios tested in this study. In contrast, the twin delayed deep deterministic policy gradient (TD3) algorithm consistently produced a high level of control optimization while maintaining the treatment requirements. Under the best selection of state observation features, TD3 control optimization reduced aeration and pumping energy requirements by 14.3% compared to the BSM1 benchmark control, outperforming the advanced domain-based control strategy of ammonia-based aeration control, although future work is necessary to improve robustness of RL implementation.

Machine learning approaches for data-driven process monitoring of biological wastewater treatment plant: A review of research works on benchmark simulation model No.1(BSM1).

In the past decade, machine learning techniques have seen wide industrial applications for design of data-based process monitoring systems with an aim to improve industrial productivity. An efficient process monitoring system for wastewater treatment process (WWTP) ensures increased efficiency and effluents meeting stringent emission norms. Benchmark simulation model No. 1 (BSM1) provides a simulation platform to researchers for developing efficient data-based process monitoring, quality monitoring, and process control systems for WWTPs. The present article presents a review of all research works reporting applications of various machine learning techniques for sensor and process fault detection of BSM1. The review focuses on process monitoring of biological wastewater treatment process, which uses a series of aerobic and anaerobic reactions followed by secondary settling process. Detailed information on various parameters monitored, different machine learning techniques explored, and results obtained by different researchers are presented in tabular and graphical format. In the review, it was observed that principal component analysis(PCA) and its variants account for the maximum number of research works for process monitoring in WWTPs and there are very few applications of recently developed deep learning techniques. Following the review and analysis, various future scopes of research (such as techniques yet to be explored or improvement of results for a particular fault) are also presented. These information will assist prospective researchers working on BSM1 to take forward the research.

Techno-Economic Assessment of PEM Electrolysis for O2 Supply in Activated Sludge Systems—A Simulation Study Based on the BSM2 Wastewater Treatment Plant

The conversion of renewable energy into hydrogen (H2) by power-to-gas technologies involving electrolysis is seen today as a key element in the transition to a sustainable energy sector. Wastewater treatment plants (WWTP) could be integrated into future green H2 networks as users of oxygen (O2) produced alongside H2 in water electrolysis. In WWTPs, O2 is required for biological treatment steps, e.g., in activated sludge (AS) systems. However, the production costs of electrolysis O2 should be competitive with those of conventional O2 production processes. In this study, mathematical models of a polymer electrolyte membrane electrolyser (PEME) plant and the WWTP of the Benchmark Simulation Model No. 2 (BSM2) were used to simulate electrolysis O2 supply to an AS system and estimate net costs of production (NCP) for produced O2 via a techno-economic assessment (TEA). Assuming that produced H2 is sold to a nearby industry, NCPs for O2 were calculated for two different PEME plant dimensions, four alternatives regarding electricity supply and costs, and three sets of assumptions regarding system performance and market conditions. The analyses were performed for 2020 as a reference year and 2030 based on forecasts of relevant data. Results of the dimensioning of the PEME show the O2 demand of a municipal WWTP with an installed capacity of 80,000 population equivalents (PE), such as the one of the BSM2, can be covered for more than 99% of the simulated period by either a 6.4 MW PEME operated for 4073 full load hours or a 4.8 MW PEME operated for 6259 full load hours. Investment costs for the PEME stacks and the operational costs for electricity make up most of the NCP of electrolysis O2. The projected decrease in PEME stack costs and renewable energy prices in favourable market conditions can result in a competitive NCP for electrolysis O2 in 2030. The approach described in this study can be applied to analyse O2 supply to biological wastewater treatment in WWTPs with different characteristics, in processes different from AS, and under different assumptions regarding economic conditions.

Event-Driven Model Predictive Control With Deep Learning for Wastewater Treatment Process

Wastewater treatment processes (WWTPs) have been considered as complex control problems, because effluent water standard, stability and multi-operational conditions need to be taken into account. In this work, an event-driven model predictive control with deep learning (EMPC-DL) is proposed for the control problems to improve the running performance of WWTPs. First, several events are defined based on different operational conditions reflected by operational data. Then, an event-driven deep belief network (EDBN) is developed based on deep learning to approximate the nonlinear characteristics of the WWTPs. Second, a quadratic optimization is designed to solve the control law of MPC based on the predictive output of the EDBN. The major advantage of quadratic optimization is its efficiency, which is achieved by an efficient strategy that only needs one-step prediction of EDBN during one-time rolling optimization. Third, this paper gives convergence and stability analysis of EMPC-DL. Finally, the feasibility and applicability of EMPC-DL are demonstrated on the benchmark simulation model No. 1 (BSM1). The experimental results show that EMPC-DL achieves the more satisfactory performance in modeling, controlling, and tracking water quality parameters than its peers.

Design of Broad Learning-Based Self-Healing Predictive Control for Sludge Bulking in Wastewater Treatment Process

Self-healing control plays a crucial role in taking remedial action to minimize the adverse impacts of sludge bulking in wastewater treatment process (WWTP). However, since sludge bulking is a strong nonlinear and complex process with multiple fault conditions, the conventional self-healing control is difficult to obtain reliable performance. Thus, the purpose of this paper is to design a broad learning-based self-healing predictive controller (BL-SHPC) for sludge bulking in WWTP. The main innovations of the proposed controller are threefold. First, a dynamic fuzzy broad learning system (DFBLS) with an adaptive expansion strategy (AES) is used to identify the fault conditions of sludge bulking. Then, the fault features of sludge bulking can be comprehensively extracted with desirable performance. Second, a prioritized multi-objective optimization algorithm (PMOOA)-based predictive control, which considers the objective correlation and preference of fault conditions, is presented to obtain the optimal solutions to achieve self-healing. Then, the proposed controller can feasibly and precisely readjust manipulated variables to eliminate the sludge bulking. Third, the stability of the developed controller is proved by the Lyapunov stability theorem. Then, the stability analysis can ensure the successful application of BL-SHPC. Finally, the proposed BL-SHPC is tested on the Benchmark Simulation Model No.2 (BSM2) to validate its merits. The simulation results indicate that the proposed controller can obtain superior self-healing ability for sludge bulking in WWTP.

Heterogeneous multi-population competitive algorithm and its application in ASM1 parameter estimation.

Activated sludge model 1 (ASM1) is a biokinetic model of the activated sludge wastewater treatment process, which has several uncertain parameters and complex chemical reaction processes. This model is often used in simulations to assess whether the effluent quality of wastewater meets discharge standards. Accurate and rapid estimation of the parameters of ASM1 is a necessary prerequisite for its successful application in industrial practice. However, ASM1-based parameter estimation is difficult to implement in practical operations due to the slow convergence rate and low convergence accuracy of most parameter estimation algorithms as well as the non-uniqueness of parameter estimation. In view of this, a novel metaheuristic optimization algorithm combining Legendre function network and dynamic partitioning strategy, named heterogeneous multi-group competitive algorithm (HMCA), is proposed to overcome the problem of low algorithmic efficiency and poor practicality in parameter estimation. The performance of the algorithm in terms of convergence speed, convergence accuracy and applicability is verified by two sets of test functions and eight state-of-the-art comparison algorithms. The superior performance of the algorithm in parameter estimation is then validated in conjunction with Benchmark Simulation Model no. 1 (BSM1) and four sets of operational data.

Economic Oriented Dynamic Matrix Control of Wastewater Treatment Plants

Wastewater Treatment Plants (WWTPs) are industrial facilities, which are important for the protection of the environment, because they remove pollutants from wastewater, before it reaches natural bodies of water. WWTPs consist of complex physical, chemical, and biological energy-intensive processes, which are subject to significant disturbances and uncertainties, due to large variations in the load and quality of the influent. Rising energy prices and increasingly stringent effluent requirements have amplified the need of developing more efficient control schemes for WWTPs. In this paper a novel Economic Dynamic Matrix Control (EDMC) configuration is proposed for WWTPs, where the objective is to minimize the plant’s operating costs in terms of energy savings, while maintaining the effluent quality within acceptable regulatory limits. The novelty of the proposed scheme lies in the combination of the standard Dynamic Matrix Control (DMC) methodology, with economic oriented control strategies. The EDMC predictive models are derived from the application of step tests on the COST/IWA Benchmark Simulation Model No. 1 (BSM1). Based on the BSM1 model, the proposed method is compared to standard Multiple Input–Multiple Output (MIMO) DMC controllers, to the default BSM1 control strategy and to other economic control methods, which have been proposed in the literature. The results illustrate that the proposed EDMC scheme is superior to alternative control strategies in terms of minimizing the energy consumption while, the effluent quality of the plant is maintained at acceptable levels.

Data-based robust model predictive control for wastewater treatment process

Data-driven methods are widely used in wastewater treatment processes (WWTPs). However, disturbances are not fully considered during the control process, which may affect the stable operation of WWTPs. To address this issue, a data-driven robust model predictive control (DRMPC) method is proposed for WWTPs with disturbances. First, an interval type-2 fuzzy neural network (IT2FNN), which has high robust performance to disturbances, is used to identify dynamic behavior of WWTPs. Second, based on IT2FNN, a robust stabilizing controller is developed to weaken the influence of disturbances on control performance and stability. Third, the proposed controller ensures constraint satisfaction and input-to-state stability (ISS). Finally, the feasibility of DRMPC strategy is verified on the benchmark simulation model No. 1 (BSM1). The experimental studies indicate that the proposed DRMPC is capable of running stably and tracking accurately.