Keywords
Drinking water, natural organic matter (NOM), enhanced coagulation, response surface methodology (RSM), environmental decision support system (EDSS)
Start Date
15-9-2020 6:40 PM
End Date
15-9-2020 7:00 PM
Abstract
Drinking water treatment plants (DWTPs) are used to deal with surface water quality and quantity fluctuations. Basically, changes affect natural organic matter (NOM) content. NOM is the major precursor of disinfection by-products (DBPs), pollutants controlled by restrictive drinking water quality requirements due to its consumption-related carcinogenic properties. To ensure treatment effectiveness, DWTPs install a series of treatments to reduce NOM content as much as possible. One of these process is coagulation-flocculation (CF) step, typically located at first stages of DWTPs. For NOM monitoring, there are several parameters analysed during water treatment. The aim of this study is to analyse NOM removal throughout treatment using parameters at two case-study DWTPs. Also the application of artificial intelligence (AI) techniques to develop CF models and implement them in a real NOM enhanced CF environmental decision support system (EDSS). CF models has been designed based on response surface methodology (RSM). First, model design and analysis are discussed. Then, the results of this study reveal that NOM selected parameters to track NOM at two case study DWTPs are not equal removed through unit operations. In both cases, CF emerges as the most powerful process in terms of NOM parameters percentages of removal. Designed RSM models are implemented in a three-level based EDSS architecture: Data acquisition, control and supervision levels. Model optimization has been analysed with 2D, 3D and overlay plots. Optimization criteria was fixed for each response in EDSSs control level, to ensure, at least: 62, 21 and 25 percentage of removal for turbidity, TOC and UV254 at Montfullà DWTP and 26, 67 and 33 for the same responses at Abrera DWTP, based on raw waters content.
Enhanced coagulation with RSM and HPSEC analysis for DWTPs optimization
Drinking water treatment plants (DWTPs) are used to deal with surface water quality and quantity fluctuations. Basically, changes affect natural organic matter (NOM) content. NOM is the major precursor of disinfection by-products (DBPs), pollutants controlled by restrictive drinking water quality requirements due to its consumption-related carcinogenic properties. To ensure treatment effectiveness, DWTPs install a series of treatments to reduce NOM content as much as possible. One of these process is coagulation-flocculation (CF) step, typically located at first stages of DWTPs. For NOM monitoring, there are several parameters analysed during water treatment. The aim of this study is to analyse NOM removal throughout treatment using parameters at two case-study DWTPs. Also the application of artificial intelligence (AI) techniques to develop CF models and implement them in a real NOM enhanced CF environmental decision support system (EDSS). CF models has been designed based on response surface methodology (RSM). First, model design and analysis are discussed. Then, the results of this study reveal that NOM selected parameters to track NOM at two case study DWTPs are not equal removed through unit operations. In both cases, CF emerges as the most powerful process in terms of NOM parameters percentages of removal. Designed RSM models are implemented in a three-level based EDSS architecture: Data acquisition, control and supervision levels. Model optimization has been analysed with 2D, 3D and overlay plots. Optimization criteria was fixed for each response in EDSSs control level, to ensure, at least: 62, 21 and 25 percentage of removal for turbidity, TOC and UV254 at Montfullà DWTP and 26, 67 and 33 for the same responses at Abrera DWTP, based on raw waters content.
Stream and Session
false