Tackling data challenges in forecasting effluent characteristics of wastewater treatment plants.

In wastewater treatment plants (WWTPs), the stochastic nature of influent wastewater and operational and weather conditions cause fluctuations in effluent quality. Data-driven models can forecast effluent quality a few hours ahead as a response to the influent characteristics, providing enough time to adjust system operations and avoid undesired consequences. However, existing data for training models are often incomplete and contain missing values. On the other hand, collecting additional data by installing new sensors is costly. The trade-off between using existing incomplete data and collecting costly new data results in three data challenges faced when developing data-driven WWTP effluent forecasters. These challenges are to determine important variables to be measured, the minimum number of required data instances, and the maximum percentage of tolerable missing values that do not impede the development of an accurate model. As these issues are not discussed in previous studies, in this research, for the first time, a comprehensive analysis is done to provide answers to these challenges. Another issue that arises in all data-driven modeling is how to select an appropriate forecasting model. This paper addresses these issues by first testing nine machine learning models on data collected from three wastewater treatment plants located in Iran, Australia, and Spain. The most accurate forecaster, Bayesian network, was then used to address the articulated challenges. Key variables in forecasting effluent characteristics were flow rate, total suspended solids, electrical conductivity, phosphorus compounds, wastewater temperature, and air temperature. A minimum of 250 samples was needed during the model training to achieve a great reduction in the forecasting error. Moreover, a steep increase in the error was observed should the portion of missing values exceed 10%. The results assist plant managers in estimating the necessary data collection effort to obtain an accurate forecaster, contributing to the quality of the effluent.

Development of data-driven models for the optimal design of multilayer sand filters for on-site treatment of greywater.

Greywater, with limited content of pathogens, makes up more than half of the produced wastewater in urban areas. Given the high cost of wastewater management and treatment, it causes sense to collect greywater separately at the source and employ an on-site treatment system to increase opportunities for on-site water reuse. For this purpose, this paper aims to propose a multilayer granular filter as an inexpensive and simple on-site treatment method for greywater reuse. Furthermore, as determining the optimal structure of multilayer filters is a serious challenge, a simulation-optimization model is developed for determining the best filter configuration. An Artificial Neural Network (ANN) is trained based on experimental results to simulate the filter performance with different combinations of layers and the Genetic Algorithm (GA) is used to find the optimal thickness of different layers based on ANN simulation results. The proposed filter in this paper for greywater treatment consists of silica sand (in three different gradings) and activated carbon (with fixed grading) and treatment measures for evaluation of filter performance are considered as Chemical Oxygen Demand (COD) and Electrical Conductivity (EC). Due to difficulties in collecting, transferring, and storing the real greywater, synthetic greywater was used in this study. 49 experiments with different combinations of filter media thicknesses were performed and the performance of the filter was analyzed. Generally, three-layer filters perform better in COD and EC reduction, however, the average COD and EC elimination equals 36.3% and 15.1%, respectively, which indicates more efficiency of filter in COD reduction in comparison with EC. Based on the optimization-simulation model and experimental results, a filter consisting of 33 cm of fine sand, 20 cm of activated carbon, and 7 cm of medium sand results in the maximum efficiency and can reduce the COD and EC of greywater by 72% and 30%, simultaneously. According to the optimization outputs, the ideal filter can treat greywater up to having EC of 1000 µS/cm and COD of 321 mg/L, which is generally suitable for irrigation purposes.

Developing a practical model for the optimal operation of wastewater treatment plant considering influent characteristics.

Wastewater treatment plants (WWTPs) play an important role in protecting the quality of water sources. The optimum operation of WWTPs in response to continuous changes in the characteristics of the influent of the WWTP is very important, and it can improve the quality of the effluent of the WWTP. In this study, an approach for optimal operation of the WWTP has been presented considering the quantitative and qualitative variables of influent. In the proposed method, first, the simulation model of WWTP is developed and calibrated using the recorded data of its influent and effluent characteristics as well as operation conditions. Then, the influent is classified into clusters quantitatively and qualitatively k-means clustering method. In the final step, after determining the effective operation parameters, the AMOEA-MAP optimization algorithm is used to determine the optimal values of operation parameters for each cluster of influents based on its quantitative and qualitative characteristics including flow rate, COD, ammonium, and temperature. The proposed approach was implemented on a WWTP in the South of Tehran, the capital of Iran. Dissolved oxygen (DO) in the aeration tank, waste-activated sludge flow rate (QWAS) and the ratio of the supernatant flow rate of the sludge dewatering unit to the effluent flow rate (Qd/Qe) were considered as operation parameters affecting the performance of the system in removing pollutants and their optimal values were obtained as DO, 0.25-1.7 mg/l, QWAS, 875-2000 m3/day, and Qd/Qe, 10-14%. Using this method, i.e., changing system operation conditions based on influent characteristics, has improved the performance of a system in reducing COD, ammonium, and nitrate in the effluent by 11-41, 17-20 and 15-34, respectively.

A hybrid constrained coral reefs optimization algorithm with machine learning for optimizing multi-reservoir systems operation.

The continuous growing demand for water, prolonged periods of drought, and climatic uncertainties attributed mainly to climate change mean surface water reservoirs more than ever need to be managed efficiently. Several optimization algorithms have been developed to optimize multi-reservoir systems operation, mostly during severe dry/wet seasons, to mitigate extreme-events consequences. Yet, convergence speed, presence of local optimums, and calculation-cost efficiency are challenging while looking for the global optimum. In this paper, the problem of finding an efficient optimal operation policy in multi-reservoir systems is discussed. The complexity of the long-term operating rules and the reservoirs' upstream and downstream joint-demands projected in recursive constraints make this problem formidable. The original Coral Reefs Optimization (CRO) algorithm, which is a meta-heuristic evolutionary algorithm, and two modified versions have been used to solve this problem. Proposed modifications reduce the calculation cost by narrowing the search space called a constrained-CCRO and adjusting reproduction operators with a reinforcement learning approach, namely the Q-Learning method (i.e., the CCRO-QL algorithm). The modified versions search for the optimum solution in the feasible region instead of the entire problem domain. The models' performance has been evaluated by solving five mathematical benchmark problems and a well-known continuous four-reservoir system (CFr) problem. Obtained results have been compared with those in the literature and the global optimum, which Linear Programming (LP) achieves. The CCRO-QL is shown to be very calculation-cost-effective in locating the global optimum or near-optimal solutions and efficient in terms of convergence, accuracy, and robustness.

An investigation of the possible scenarios for the optimal locating of quality sensors in the water distribution networks with uncertain contamination.

One of the ways to reduce the risk of contaminated water consumption is to optimally locate the quality sensors. These sensors warn users in the case of contamination detection. Analyzing the actual conditions of the contamination which enters the network is faced with many uncertainties. These uncertainties include the dose of contamination, time and location of its entry which have received less attention. Also, the uncertainty in the nodes' water demand causes changes in the distribution and contamination diffusion within the network. The main impetus of the present study is to determine the optimal quality sensor locations in the water distribution network in order to reduce the damage caused by contaminated water consumption prior to the contamination event detection. For this purpose, a parameter is defined as the maximum possible damage for calculating which the vulnerability and importance of the nodes have been considered in addition to the uncertainties in the location and time of the contamination entry. The importance of each node differs from that of other ones. Ranking the importance of the nodes is influenced by both land use and covered population ratio. In this study, six scenarios are defined for the contamination event in the water distribution network. These scenarios consider the effects of varying pollutant dose and the contamination input from nodes which are prone to its entry. Also, the NSGA-II has been utilized in order to minimize the damage with minimum number of sensors. The proposed model is evaluated on a real network in Iran. The results indicate that adding only one or two contamination warning sensors to the proposed locations can lead to the decreasing damage caused by the contaminated water consumption from 54 to 82%. According to the proposed method, the best answer for scenarios 1-6 was obtained for 7, 6, 6, 2, 2 and 2 sensors, respectively. The results showed that the slope of the pollution rate diagram does not change much from 6 sensors upwards in the first three scenarios, and from 4 sensors upwards in the second three scenarios. In scenarios 1-3, with 7, 6 and 6 sensors, respectively, in different nodes, the best placement is for 203-224 equivalent attack population, and in scenarios 4-6, with sensors in nodes 4 and 43, the best placement is for 225-279 equivalent attack population.

Investigating the performance of agricultural wastes and their ashes in removing phenol from leachate in a fixed-bed column.

One of the major pollutants in leachate is phenol. Due to safety and environmental problems, removal of phenol from leachate is essential. Most of the adsorption studies have been conducted in batch systems. Practically, large-scale adsorption is carried out in continuous systems. In this research, the adsorption method has been used for phenol removal from leachate by using walnut shell activated carbon (WSA) and coconut shell activated carbon (CSA) as adsorbents in a fixed-bed column. The effect of adsorbent bed depth, influent phenol concentration and type of adsorbent on adsorption was explored. By increasing the depth of the adsorbent bed in the column, phenol removal efficiency and saturation time increase significantly. Also, by increasing the influent concentration, saturation time of the column decreases. To predict the column performance and describe the breakthrough curve, three kinetic models of Yon-Nelson, Adams-Bohart and Thomas were applied. The results of the experiments indicate that there is a good match between the results of the experiment and the predicted results of the models.

Improvement and prediction of OSA system performance in sludge reduction through integration with thermal and mechanical treatment.

The oxic-settling-anoxic (OSA) process is one of the sludge production reduction methods in the activated sludge process. In this method, sludge is stored in an anaerobic tank within the sludge return line before entrance into an aeration tank. Due to this method's flexibility in application to operating treatment plants and not being energy-consuming, its application is developing. In this research, the improvement of the OSA process is investigated via thermal and mechanical treatment in a sequencing batch reactor (SBR). A pilot-scale reactor and domestic wastewater are used. Sludge was subjected to high temperature in an anaerobic tank using a heat transformer and it was subjected to mechanical shear through mechanical mixing in the anaerobic tank. Different temperatures and voltages were tested. The OSA process reduced sludge production by 24% while the chemical oxygen demand (COD) removal rate decreased from 90% to 86%. Thermal treatment combined with the OSA process caused a maximum of 46% sludge production reduction. However temperatures above 90 °C are not recommended due to a high level of decrease in COD removal. Mechanical mixing in combination with the OSA process led to 34% sludge production reduction. The effluent quality is not affected by the OSA process itself but is slightly reduced by thermal treatment and mechanical mixing. Therefore, for reaching the maximum sludge reduction in OSA plus thermal and mechanical treatment it would be necessary to evaluate the effect of different sets of parameters on effluent quality beside the sludge reduction. For this purpose multi-layer perceptron artificial neural network models are developed to predict the effluent total suspended solids and COD removal efficiency as well as sludge production rate. The models perform well and would be useful tools in determining the optimal set of system operation parameters.