Development of gradient boosting-assisted machine learning data-driven model for free chlorine residual prediction

Highlights

● A machine learning approach was applied to predict free chlorine residuals.

● Annual data were obtained from chlorination unit at a 98 MGD water treatment plant.

● The last model iteration returned a high prediction value ( R ² = 0.937).

● Non-intuitive parameters were found to be highly significant to predictions.

Abstract

Chlorine-based disinfection is ubiquitous in conventional drinking water treatment (DWT) and serves to mitigate threats of acute microbial disease caused by pathogens that may be present in source water. An important index of disinfection efficiency is the free chlorine residual (FCR), a regulated disinfection parameter in the US that indirectly measures disinfectant power for prevention of microbial recontamination during DWT and distribution. This work demonstrates how machine learning (ML) can be implemented to improve FCR forecasting when supplied with water quality data from a real, full-scale chlorine disinfection system in Georgia, USA. More precisely, a gradient-boosting ML method (CatBoost) was developed from a full year of DWT plant-generated chlorine disinfection data, including water quality parameters (e.g., temperature, turbidity, pH) and operational process data (e.g., flowrates), to predict FCR. Four gradient-boosting models were implemented, with the highest performance achieving a coefficient of determination, R², of 0.937. Values that provide explanations using Shapley’s additive method were used to interpret the model’s results, uncovering that standard DWT operating parameters, although non-intuitive and theoretically non-causal, vastly improved prediction performance. These results provide a base case for data-driven DWT disinfection supervision and suggest process monitoring methods to provide better information to plant operators for implementation of safe chlorine dosing to maintain optimum FCR.

Graphical abstract

Keywords

Machine learning / Data-driven modeling / Drinking water treatment / Disinfection / Chlorination

References

[1] Abdullah M P , Yee L F , Ata S , Abdullah A , Ishak B , Abidin K N Z . (2009). The study of interrelationship between raw water quality parameters, chlorine demand and the formation of disinfection by-products. Physics and Chemistry of the Earth Parts A/B/C, 34(13–16): 806–811

[2] André Felipe L , Fábio Cosme Rodrigues Dos S , Cleber Gustavo D . (2018). Artificial neural networks to control chlorine dosing in a water treatment plant. Acta Scientiarum. Technology, 40(1): 1–9

[3] Boulos P F . (2017). Optimal scheduling of pipe replacement. Journal−American Water Works Association, 109(1): 42–46

[4] BuysschaertBVermijsLNakaABoonNDe GussemeB (2018). Online flow cytometric monitoring of microbial water quality in a full-scale water treatment plant. npj Clean Water, 1(1): 16

[5] Clark R M , Sivaganesan M . (2002). Predicting chlorine residuals in drinking water: second order model. Journal of Water Resources Planning and Management, 128(2): 152–161

[6] CriderYSultanaSUnicombLDavisJLubyS PPickeringA J (2018). Can you taste it? Taste detection and acceptability thresholds for chlorine residual in drinking water in Dhaka, Bangladesh. Science of the Total Environment, 613–614: 840–846

[7] Delpla I , Jung A V , Baures E , Clement M , Thomas O . (2009). Impacts of climate change on surface water quality in relation to drinking water production. Environment International, 35(8): 1225–1233

[8] Di Nardo A , Di Natale M , Greco R , Santonastaso G F . (2014). Ant algorithm for smart water network partitioning. Procedia Engineering, 70: 525–534

[9] FishKOsbornA MBoxallJ B (2017). Biofilm structures (EPS and bacterial communities) in drinking water distribution systems are conditioned by hydraulics and influence discolouration. Science of the Total Environment, 593–594: 571–580

[10] Frateur I , Deslouis C , Kiene L , Levi Y , Tribollet B . (1999). Free chlorine consumption induced by cast iron corrosion in drinking water distribution systems. Water Research, 33(8): 1781–1790

[11] Fujioka T , Hoang A T , Aizawa H , Ashiba H , Fujimaki M , Leddy M . (2018). Real-time online monitoring for assessing removal of bacteria by reverse osmosis. Environmental Science & Technology Letters, 5(6): 389–393

[12] Gagnon G A , Rand J L , O’leary K C , Rygel A C , Chauret C , Andrews R C . (2005). Disinfectant efficacy of chlorite and chlorine dioxide in drinking water biofilms. Water Research, 39(9): 1809–1817

[13] Gang D C , Clevenger T E , Banerji K S . (2003). Modeling chlorine decay in surface water. Journal of Environmental Informatics, 1(1): 21–27

[14] Gao H , Zhong S , Zhang W , Igou T , Berger E , Reid E , Zhao Y , Lambeth D , Gan L , Afolabi M A , Tong Z , Lan G , Chen Y . (2022). Revolutionizing membrane design using machine learning-Bayesian optimization. Environmental Science & Technology, 56(4): 2572–2581

[15] Gray M J , Wholey W Y , Jakob U . (2013). Bacterial responses to reactive chlorine species. Annual Review of Microbiology, 67(1): 141–160

[16] HolzingerAGoebelRFongRMoonTMüllerK RSamekW (2022). xxAI-beyond explainable artificial intelligence. In: Proceedings of International Workshop on Extending Explainable AI Beyond Deep Models and Classifiers, Vienna, Austria, 18 July 2020. Cham: Springer, 3–10

[17] Hsu L H H , Hoque E , Kruse P , Ravi Selvaganapathy P . (2015). A carbon nanotube based resettable sensor for measuring free chlorine in drinking water. Applied Physics Letters, 106(6): 063102

[18] Li L , Rong S , Wang R , Yu S . (2021). Recent advances in artificial intelligence and machine learning for nonlinear relationship analysis and process control in drinking water treatment: a review. Chemical Engineering Journal, 405: 126673

[19] LiuX (2016). Methods and Applications of Longitudinal Data Analysis. Oxford: Academic Press, 441–473

[20] Lowe M , Qin R , Mao X . (2022). A review on machine learning, artificial intelligence, and smart technology in water treatment and monitoring. Water (Basel), 14(9): 1384–1411

[21] Mac Kenzie W R , Hoxie N J , Proctor M E , Gradus M S , Blair K A , Peterson D E , Kazmierczak J J , Addiss D G , Fox K R , Rose J B . . (1994). A massive outbreak in Milwaukee of Cryptosporidium infection transmitted through the public water supply. New England Journal of Medicine, 331(3): 161–167

[22] MerrickLTalyA (2020). The Explanation Game: Explaining Machine Learning Models Using Shapley Values. Cham: Springer International Publishing, 17–38

[23] Onyutha C , Kwio-Tamale J C . (2022). Modelling chlorine residuals in drinking water: a review. International Journal of Environmental Science and Technology, 19(11): 11613–11630

[24] Powell J C , Hallam N B , West J R , Forster C F , Simms J . (2000). Factors which control bulk chlorine decay rates. Water Research, 34(1): 117–126

[25] Reid E , Igou T , Zhao Y , Crittenden J , Huang C H , Westerhoff P , Rittmann B , Drewes J E , Chen Y . (2023). The minus approach can redefine the standard of practice of drinking water treatment. Environmental Science & Technology, 57(18): 7150–7161

[26] Richardson S D , Kimura S Y . (2020). Water analysis: emerging contaminants and current issues. Analytical Chemistry, 92(1): 473–505

[27] Rittmann B E , Snoeyink V L . (1984). Achieving biologically stable drinking water. Journal–American Water Works Association, 76(10): 106–114

[28] Romano M , Kapelan Z , Savić D A . (2014). Automated detection of pipe bursts and other events in water distribution systems. Journal of Water Resources Planning and Management, 140(4): 457–467

[29] Saboe D , Hristovski K D , Burge S R , Burge R G , Taylor E , Hoffman D A . (2021). Measurement of free chlorine levels in water using potentiometric responses of biofilms and applications for monitoring and managing the quality of potable water. Science of the Total Environment, 766: 144424

[30] Sedlak D L , Von Gunten U . (2011). The chlorine dilemma. Science, 331(6013): 42–43

[31] Smeets P W M H , Medema G J , Van Dijk J C . (2009). The Dutch secret: How to provide safe drinking water without chlorine in the Netherlands?. Drinking Water Engineering and Science, 2(1): 1–14

[32] Suffet I H , Corado A , Chou D , Mcguire M J , Butterworth S . (1996). AWWA taste and odor survey. Journal–American Water Works Association, 88(4): 168–180

[33] SundararajanMNajmiA (2020). The many Shapley values for model explanation. In: Hal D III, Aarti S, editors. Proceedings of Machine Learning Research. Brookline, MA, USA: 119, 9269–9278

[34] Tinelli S , Juran I . (2019). Artificial intelligence-based monitoring system of water quality parameters for early detection of non-specific bio-contamination in water distribution systems. Water Science and Technology: Water Supply, 19(6): 1785–1792

[35] Valdivia-Garcia M , Weir P , Graham D W , Werner D . (2019). Predicted impact of climate change on trihalomethanes formation in drinking water treatment. Scientific Reports, 9(1): 9967

[36] Warton B , Heitz A , Joll C , Kagi R . (2006). A new method for calculation of the chlorine demand of natural and treated waters. Water Research, 40(15): 2877–2884

[37] Wilson R E , Stoianov I , O’hare D . (2019). Continuous chlorine detection in drinking water and a review of new detection methods. Johnson Matthey Technology Review, 63(2): 103–118

[38] WorldHealth Organization (2017). Principles and Practices of Drinking-Water Chlorination: a Guide to Strengthening Chlorination Practices in Small to Medium Sized Water Supplies. New Delhi: World Health Organization Regional Office for South-East Asia

[39] Zhang B , Kotsalis G , Khan J , Xiong Z , Igou T , Lan G , Chen Y . (2020a). Backwash sequence optimization of a pilot-scale ultrafiltration membrane system using data-driven modeling for parameter forecasting. Journal of Membrane Science, 612: 118464

[40] Zhang K , Zhong S , Zhang H . (2020b). Predicting aqueous adsorption of organic compounds onto biochars, carbon nanotubes, granular activated carbons, and resins with machine learning. Environmental Science & Technology, 54(11): 7008–7018

[41] Zhong S , Lambeth D R , Igou T K , Chen Y . (2022). Enlarging applicability domain of quantitative structure–activity relationship models through uncertainty-based active learning. ACS ES&T Engineering, 2(7): 1211–1220

[42] Zhong S , Zhang K , Bagheri M , Burken J G , Gu A , Li B , Ma X , Marrone B L , Ren Z J , Schrier J . . (2021). Machine learning: new ideas and tools in environmental science and engineering. Environmental Science & Technology, 55(19): 12741–12754

Acknowledgements

This research was partially supported by: US Department of Agriculture’s National Institute of Food and Agriculture, Agriculture and Food Research Initiative, Water for Food Production Systems (No. 2018-68011-28371); National Science Foundation (USA) (Nos. 1936928, 2112533); US Department of Agriculture’ National Institute of Food and Agriculture (No. 2020-67021-31526); and US Environmental Protection Agency (No. 840080010).

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Data Accessibility Statement

The data and code that support the findings of this study are available from the corresponding author, Prof. Yongsheng Chen, upon reasonable request.

Electronic Supplementary Material

Supplementary material is available in the online version of this article at https://doi.org/10.1007/s11783-024-1777-6 and is accessible for authorized users.

RIGHTS & PERMISSIONS

2024 Higher Education Press