@article{fasaee_pesantez_pieper_ling_benham_edwards_berglund_2022, title={Developing early warning systems to predict water lead levels in tap water for private systems}, volume={221}, ISSN={["1879-2448"]}, url={https://doi.org/10.1016/j.watres.2022.118787}, DOI={10.1016/j.watres.2022.118787}, abstractNote={Lead is a chemical contaminant that threatens public health, and high levels of lead have been identified in drinking water at locations across the globe. Under-served populations that use private systems for drinking water supplies may be at an elevated level of risk because utilities and governing agencies are not responsible for ensuring that lead levels meet the Lead and Copper Rule at these systems. Predictive models that can be used by residents to assess water quality threats in their households can create awareness of water lead levels (WLLs). This research explores and compares the use of statistical models (i.e., Bayesian Belief classifiers) and machine learning models (i.e., ensemble of decision trees) for predicting WLLs. Models are developed using a dataset collected by the Virginia Household Water Quality Program (VAHWQP) at approximately 8000 households in Virginia during 2012-2017. The dataset reports laboratory-tested water quality parameters at households, location information, and household and plumbing characteristics, including observations of water odor, taste, discoloration. Some water quality parameters, such as pH, iron, and copper, can be measured at low resolution by residents using at-home water test kits and can be used to predict risk of WLLs. The use of at-home water quality test kits was simulated through the discretization of water quality parameter measurements to match the resolution of at-home water quality test kits and the introduction of error in water quality readings. Using this approach, this research demonstrates that low-resolution data collected by residents can be used as input for models to estimate WLLs. Model predictability was explored for a set of at-home water quality test kits that observe a variety of water quality parameters and report parameters at a range of resolutions. The effects of the timing of water sampling (e.g., first-draw vs. flushed samples) and error in kits on model error were tested through simulations. The prediction models developed through this research provide a set of tools for private well users to assess the risk of lead contamination. Models can be implemented as early warning systems in citizen science and online platforms to improve awareness of drinking water threats.}, journal={WATER RESEARCH}, author={Fasaee, Mohammad Ali Khaksar and Pesantez, Jorge and Pieper, Kelsey J. and Ling, Erin and Benham, Brian and Edwards, Marc and Berglund, Emily}, year={2022}, month={Aug} }
 @article{fasaee_monghasemi_nikoo_shafiee_berglund_bakhtiari_2021, title={A K-Sensor correlation-based evolutionary optimization algorithm to cluster contamination events and place sensors in water distribution systems}, volume={319}, ISSN={["1879-1786"]}, DOI={10.1016/j.jclepro.2021.128763}, abstractNote={Contaminants that are introduced to drinking water systems can threaten large populations, and the potential for catastrophic consequences accentuates the need for efficient post-disaster strategies, including optimal hydrant flushing. Efficient hydrant flushing can significantly reduce impacts on public health, but performance relies on information about the propagation of a contaminant and the affected regions in a water network. While observations from water quality sensors are useful in timely detections of contaminants, little information on its source, propagation, and affected regions can be inferred. In the absence of such information, opening or closing hydrants might not help discharge contaminants but could accelerate propagation of a plume through the water network due to drops in pressure. To address this limitation of sensor layout optimization models, this research has developed a new model to identify the optimal location of sensors to effectively support hydrant flushing mechanisms. The model has been developed in three steps: (1) contamination events were simulated in a water network; (2) spatially similar propagating contamination events were identified; and (3) the layout of water quality sensors was optimized. In the first step, a representative number of potential contamination events were simulated using a hydraulic model. The second step clustered contamination events based on spatial similarity in their propagation regimes. Finally, the last step identified locations for placing water quality sensors within clusters (identified in the previous step) while minimizing detection time and maximizing probability of detection. This model ensures that when a sensor alarm is activated, contaminated region where hydrants should be opened or closed are spatially restricted. The approach developed in this research was applied to design a sensor network for a benchmark case study, Mesopolis. The layout of 10 water quality sensors was optimized over a set of 9161 contamination events, leading to 76% probability of detection with an average detection time of 8.2 h. The solution was compared with sensor layouts based on existing approaches, and it was found that the new approach could improve the mass of contaminant that was removed from the pipe network through hydrant flushing strategies. The new approach model improves the effectiveness of hydrant flushing strategies by restricting the area where hydrants are flushed to predefined zones based on the activation of sensors.}, journal={JOURNAL OF CLEANER PRODUCTION}, author={Fasaee, Mohammad Ali Khaksar and Monghasemi, Shahryar and Nikoo, Mohammad Reza and Shafiee, M. Ehsan and Berglund, Emily Zechman and Bakhtiari, Parnian Hashempour}, year={2021}, month={Oct} }
 @article{fasaee_berglund_pieper_ling_benham_edwards_2021, title={Developing a framework for classifying water lead levels at private drinking water systems: A Bayesian Belief Network approach}, volume={189}, ISSN={["1879-2448"]}, DOI={10.1016/j.watres.2020.116641}, abstractNote={The presence of lead in drinking water creates a public health crisis, as lead causes neurological damage at low levels of exposure. The objective of this research is to explore modeling approaches to predict the risk of lead at private drinking water systems. This research uses Bayesian Network approaches to explore interactions among household characteristics, geological parameters, observations of tap water, and laboratory tests of water quality parameters. A knowledge discovery framework is developed by integrating methods for data discretization, feature selection, and Bayes classifiers. Forward selection and backward selection are explored for feature selection. Discretization approaches, including domain-knowledge, statistical, and information-based approaches, are tested to discretize continuous features. Bayes classifiers that are tested include General Bayesian Network, Naive Bayes, and Tree-Augmented Naive Bayes, which are applied to identify Directed Acyclic Graphs (DAGs). Bayesian inference is used to fit conditional probability tables for each DAG. The Bayesian framework is applied to fit models for a dataset collected by the Virginia Household Water Quality Program (VAHWQP), which collected water samples and conducted household surveys at 2,146 households that use private water systems, including wells and springs, in Virginia during 2012 and 2013. Relationships among laboratory-tested water quality parameters, observations of tap water, and household characteristics, including plumbing type, source water, household location, and on-site water treatment are explored to develop features for predicting water lead levels. Results demonstrate that Naive Bayes classifiers perform best based on recall and precision, when compared with other classifiers. Copper is the most significant predictor of lead, and other important predictors include county, pH, and on-site water treatment. Feature selection methods have a marginal effect on performance, and discretization methods can greatly affect model performance when paired with classifiers. Owners of private wells remain disadvantaged and may be at an elevated level of risk, because utilities and governing agencies are not responsible for ensuring that lead levels meet the Lead and Copper Rule for private wells. Insight gained from models can be used to identify water quality parameters, plumbing characteristics, and household variables that increase the likelihood of high water lead levels to inform decisions about lead testing and treatment.}, journal={WATER RESEARCH}, author={Fasaee, Mohammad Ali Khaksar and Berglund, Emily and Pieper, Kelsey J. and Ling, Erin and Benham, Brian and Edwards, Marc}, year={2021}, month={Feb} }
 @article{fasaee_nikoo_bakhtiari_monghasemi_sadegh_2020, title={A novel dynamic hydrant flushing framework facilitated by categorizing contamination events}, volume={17}, ISSN={["1744-9006"]}, DOI={10.1080/1573062X.2020.1758163}, abstractNote={ABSTRACT Security of water distribution systems is of highest importance to public health. In this study, a novel dynamic hydrant flushing strategy is developed to remove the injected contamination mass in the shortest possible time after being detected by the sensors. The coupled EPANET-IHS simulation-optimization model is developed to determine the hydrant flushing strategies (the hydrant sets and their opening/closing time). This allows water utility managers to utilize the hydrants instantly after the sensor alarm with no need for excessive investigations, which enhances the flushing effectiveness. Developed based on thousands of contaminant injection scenarios (contamination events), the proposed hydrant flushing strategies are reliable to remove any injected contamination mass. Moreover, the operation of hydrant sets is dynamically validated by closing/opening of hydrants throughout the total hydrant flushing duration, which enables the model to flush up to more 20%. The methodology is examined on the mid-sized WDS of Mesopolis city.}, number={3}, journal={URBAN WATER JOURNAL}, author={Fasaee, Mohammad Ali Khaksar and Nikoo, Mohammad Reza and Bakhtiari, Parnian Hashempour and Monghasemi, Shahryar and Sadegh, Mojtaba}, year={2020}, month={Mar}, pages={199–211} }