The main goal of disinfection, in plant and along the distribution network, is to ensure microbiologically safe water for human consumption, also considering the formation of disinfection by-products (DBPs). The use of forecasting models can support process optimization. The are many models present in the literature, but their applicability is currently unverified. The aim of this work is to generalize the decay models of sodium hypochlorite and chlorine dioxide and the formation of their DBPs in order to assess their effective applicability to case studies of drinking water distribution networks. Starting from the quality characteristics of water, through empirical linear regression models, the parameters of decay of the kinetics of the first and second order (the two models more common for sodium hypochlorite) have been estimated. The empirical models obtained were validated by real data collected in a monitoring campaign: both kinetics overestimate the concentration of disinfectant in the network with average absolute errors of 0.043 mg/L and 0.073 mg/L, respectively. The worst performance of second-order kinetic is probably due to the difficulty of estimating the kinetic constant from literature data. Concerning the formation of trihalomethanes (THM), 26 empirical models were compared, using the data monitored during the campaign carried out at a water distribution network. The models that best describe measured THM concentrations are those proposed by Tsitsifli & Kanakoudis (2020), with NRMSE of 0.19 and 0.20. On the other hand, in the literature there is a limited presence of kinetic and empirical models for chlorine dioxide decay and chlorite and chlorate formation: four models were chosen, validated with the parameters monitored in the water distribution network chosen as a case study, but with unsatisfactory results, overestimating the concentrations of disinfectant and DBPs.
L’obiettivo principale della disinfezione, in impianto e in rete, è garantire un’acqua microbiologicamente sicura per il consumo umano, anche considerando la formazione di sottoprodotti. Il ricorso a modelli previsionali potrebbe supportare l’ottimizzazione di processo. I modelli presenti in letteratura sono molteplici, ma la loro effettiva applicabilità pratica è ad oggi non verificata. Obiettivo del presente lavoro è la razionalizzazione dei modelli di decadimento di ipoclorito di sodio e biossido di cloro e di formazione dei relativi sottoprodotti, per valutarne l’effettiva applicazione a casi studio di reti di distribuzione di acqua potabile. A partire dalle caratteristiche di qualità dell’acqua, attraverso modelli empirici di regressione lineare, sono stati stimati i parametri di decadimento della cinetica di primo e di secondo ordine, attualmente i più diffusi per ipoclorito di sodio. I modelli empirici ottenuti sono stati validati tramite dati reali raccolti in una campagna di monitoraggio: entrambe le cinetiche sovrastimano la concentrazione di disinfettante in rete con errori assoluti medi pari rispettivamente a 0,043 mg/L e 0,073 mg/L. La peggiore performance della cinetica di secondo ordine è probabilmente dovuta alla difficoltà di stima della costante cinetica a partire da dati di letteratura. In relazione alla formazione di trialometani (THM), sono stati confrontati 26 modelli di letteratura, utilizzando i dati monitorati nella campagna effettuata presso una rete idrica scelta come caso studio. I modelli che meglio descrivono le concentrazioni di THM misurate in rete sono quelli proposti da Tsitsifli & Kanakoudis (2020), con NRMSE pari a 0,19 e 0,20. Vi è invece una presenza limitata in letteratura di modelli per la stima del decadimento di biossido di cloro e la formazione di clorito e clorato: sono stati scelti quattro modelli, validati con i parametri monitorati nella rete idrica scelta come caso studio, ma con esiti insoddisfacenti, sovrastimando le concentrazioni di disinfettante e di sottoprodotti.
Gestione ottimizzata della disinfezione in rete : monitoraggio e modelli
BONACINA, CHIARA
2020/2021
Abstract
The main goal of disinfection, in plant and along the distribution network, is to ensure microbiologically safe water for human consumption, also considering the formation of disinfection by-products (DBPs). The use of forecasting models can support process optimization. The are many models present in the literature, but their applicability is currently unverified. The aim of this work is to generalize the decay models of sodium hypochlorite and chlorine dioxide and the formation of their DBPs in order to assess their effective applicability to case studies of drinking water distribution networks. Starting from the quality characteristics of water, through empirical linear regression models, the parameters of decay of the kinetics of the first and second order (the two models more common for sodium hypochlorite) have been estimated. The empirical models obtained were validated by real data collected in a monitoring campaign: both kinetics overestimate the concentration of disinfectant in the network with average absolute errors of 0.043 mg/L and 0.073 mg/L, respectively. The worst performance of second-order kinetic is probably due to the difficulty of estimating the kinetic constant from literature data. Concerning the formation of trihalomethanes (THM), 26 empirical models were compared, using the data monitored during the campaign carried out at a water distribution network. The models that best describe measured THM concentrations are those proposed by Tsitsifli & Kanakoudis (2020), with NRMSE of 0.19 and 0.20. On the other hand, in the literature there is a limited presence of kinetic and empirical models for chlorine dioxide decay and chlorite and chlorate formation: four models were chosen, validated with the parameters monitored in the water distribution network chosen as a case study, but with unsatisfactory results, overestimating the concentrations of disinfectant and DBPs.File | Dimensione | Formato | |
---|---|---|---|
2021_07_Bonacina.pdf
Open Access dal 09/07/2024
Dimensione
2.64 MB
Formato
Adobe PDF
|
2.64 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/177816