Prediction of Membrane Failure in a Water Purification Plant Using Nonhomogeneous Poisson Process Models

The prediction of membrane failure in full-scale water purification plants is an important but difficult task. Although previous studies employed accelerated laboratory-scale tests of membrane failure, it is not possible to reproduce the complex operational conditions of full-scale plants. Therefore...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Takashi Hashimoto, Satoshi Takizawa
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	membrane filtration membrane failure nonhomogeneous Poisson process bootstrap module replacement Chemical technology TP1-1185 Chemical engineering TP155-156
Acceso en línea:	https://doaj.org/article/72fc8f97886b415b966485ea1c4b6c89
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:72fc8f97886b415b966485ea1c4b6c89
record_format	dspace
spelling	oai:doaj.org-article:72fc8f97886b415b966485ea1c4b6c892021-11-25T18:19:26ZPrediction of Membrane Failure in a Water Purification Plant Using Nonhomogeneous Poisson Process Models10.3390/membranes111108002077-0375https://doaj.org/article/72fc8f97886b415b966485ea1c4b6c892021-10-01T00:00:00Zhttps://www.mdpi.com/2077-0375/11/11/800https://doaj.org/toc/2077-0375The prediction of membrane failure in full-scale water purification plants is an important but difficult task. Although previous studies employed accelerated laboratory-scale tests of membrane failure, it is not possible to reproduce the complex operational conditions of full-scale plants. Therefore, we aimed to develop prediction models of membrane failure using actual membrane failure data. Because membrane filtration systems are repairable systems, nonhomogeneous Poisson process (NHPP) models, i.e., power law and log-linear models, were employed; the model parameters were estimated using the membrane failure data from a full-scale plant operated for 13 years. Both models were able to predict cumulative failures for forthcoming years; nonetheless, the power law model showed higher stability and narrower confidence intervals than the log-linear model. By integrating two membrane replacement criteria, namely deterioration of filtrate water quality and reduction of membrane permeability, it was possible to predict the time to replace all the membranes on a water purification plant. Finally, the NHPP models coupled with a nonparametric bootstrap method provided a method to select membrane modules for earlier replacement than others. Although the criteria for membrane replacement may vary among membrane filtration plants, the NHPP models presented in this study could be applied to any other plant with membrane failure data.Takashi HashimotoSatoshi TakizawaMDPI AGarticlemembrane filtrationmembrane failurenonhomogeneous Poisson processbootstrapmodule replacementChemical technologyTP1-1185Chemical engineeringTP155-156ENMembranes, Vol 11, Iss 800, p 800 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	membrane filtration membrane failure nonhomogeneous Poisson process bootstrap module replacement Chemical technology TP1-1185 Chemical engineering TP155-156
spellingShingle	membrane filtration membrane failure nonhomogeneous Poisson process bootstrap module replacement Chemical technology TP1-1185 Chemical engineering TP155-156 Takashi Hashimoto Satoshi Takizawa Prediction of Membrane Failure in a Water Purification Plant Using Nonhomogeneous Poisson Process Models
description	The prediction of membrane failure in full-scale water purification plants is an important but difficult task. Although previous studies employed accelerated laboratory-scale tests of membrane failure, it is not possible to reproduce the complex operational conditions of full-scale plants. Therefore, we aimed to develop prediction models of membrane failure using actual membrane failure data. Because membrane filtration systems are repairable systems, nonhomogeneous Poisson process (NHPP) models, i.e., power law and log-linear models, were employed; the model parameters were estimated using the membrane failure data from a full-scale plant operated for 13 years. Both models were able to predict cumulative failures for forthcoming years; nonetheless, the power law model showed higher stability and narrower confidence intervals than the log-linear model. By integrating two membrane replacement criteria, namely deterioration of filtrate water quality and reduction of membrane permeability, it was possible to predict the time to replace all the membranes on a water purification plant. Finally, the NHPP models coupled with a nonparametric bootstrap method provided a method to select membrane modules for earlier replacement than others. Although the criteria for membrane replacement may vary among membrane filtration plants, the NHPP models presented in this study could be applied to any other plant with membrane failure data.
format	article
author	Takashi Hashimoto Satoshi Takizawa
author_facet	Takashi Hashimoto Satoshi Takizawa
author_sort	Takashi Hashimoto
title	Prediction of Membrane Failure in a Water Purification Plant Using Nonhomogeneous Poisson Process Models
title_short	Prediction of Membrane Failure in a Water Purification Plant Using Nonhomogeneous Poisson Process Models
title_full	Prediction of Membrane Failure in a Water Purification Plant Using Nonhomogeneous Poisson Process Models
title_fullStr	Prediction of Membrane Failure in a Water Purification Plant Using Nonhomogeneous Poisson Process Models
title_full_unstemmed	Prediction of Membrane Failure in a Water Purification Plant Using Nonhomogeneous Poisson Process Models
title_sort	prediction of membrane failure in a water purification plant using nonhomogeneous poisson process models
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/72fc8f97886b415b966485ea1c4b6c89
work_keys_str_mv	AT takashihashimoto predictionofmembranefailureinawaterpurificationplantusingnonhomogeneouspoissonprocessmodels AT satoshitakizawa predictionofmembranefailureinawaterpurificationplantusingnonhomogeneouspoissonprocessmodels
_version_	1718411326491983872

Prediction of Membrane Failure in a Water Purification Plant Using Nonhomogeneous Poisson Process Models

Ejemplares similares