Trihalomethane prediction modelling in water distribution systems: calculation of the mean residence time
Contact of chlorine with organic matter can produce disinfection byproducts such as Trihalomethanes (THMs). These compounds are considered to be hazardous to health and their formation is influenced by pH, temperature, type of organic matter, dose of chlorine and reaction time. However, given the co...
| Autores Principales: | Araya-Obando, Andrés, Jones-Sánchez, Mark, Romero-Esquivel, Luis G |
|---|---|
| Formato: | Artículo |
| Idioma: | Español |
| Publicado: |
Editorial Tecnológica de Costa Rica (entidad editora)
2019
|
| Materias: | |
| Acceso en línea: |
https://revistas.tec.ac.cr/index.php/tec_marcha/article/view/4126 https://hdl.handle.net/2238/11858 |
| Sumario: |
Contact of chlorine with organic matter can produce disinfection byproducts such as Trihalomethanes (THMs). These compounds are considered to be hazardous to health and their formation is influenced by pH, temperature, type of organic matter, dose of chlorine and reaction time. However, given the costs associated with laboratory analyzes, mechanical and statistical models are commonly used for their prediction. Nevertheless, the determination of the mean residence time (TMR) in distribution networks requires computational tools that demand time and investment, so it is necessary to consider other methods for their estimation. In this sense, the present article emphasizes into the main aspects that must be considered for the construction of a prediction model, as well as the analysis of two calculation methods for the determination of TMR in distribution networks using tracers. For the latter, tests were carried out on a pilot scale distribution network made up of 12 mm diameter PVC pipes. Sodium chloride was used as a tracer by continuous addition. TMRs were determined at two sampling points and a difference of 2.40% and 3.31% respectively were obtained, demonstrating that both calculations methods are precise and easily understood. Finally, it is concluded that the models constructed from multiple regressions can be potentially used in Costa Rica, since they can be constructed simply from local conditions. |
|---|