Evaluating constructed wetlands with water hyacinth for greywater treatment: Media comparison and ANN-based predictive modelling

Q1 Environmental Science

Bioresource Technology Reports Pub Date : 2025-04-04 DOI:10.1016/j.biteb.2025.102112

Chandrashekhar Parab , Aakrut V. Patel , Kunwar D. Yadav , Vimalkumar Prajapati

{"title":"Evaluating constructed wetlands with water hyacinth for greywater treatment: Media comparison and ANN-based predictive modelling","authors":"Chandrashekhar Parab , Aakrut V. Patel , Kunwar D. Yadav , Vimalkumar Prajapati","doi":"10.1016/j.biteb.2025.102112","DOIUrl":null,"url":null,"abstract":"<div><div>Water demand is rising with population growth, making greywater reuse vital for sustainability. Constructed wetlands (CWs) utilize natural processes to treat greywater, but the role of water hyacinths and media in enhancing treatment efficiency remains unclear. This study assessed two CWs with water hyacinth: one with gravel media (WM) and one without (WOM). Over 90 days, the CW-WM showed significant removal efficiencies: 91.59 % for turbidity, 55.74 % for COD, 79.96 % for BOD, 75.97 % for phosphate, and 30.67 % for ammonia over CW-WOM. An artificial neural network (ANN) was employed to predict BOD and COD using input parameters like pH, EC, turbidity, TS, and TDS. The BOD model achieved an R-value of 0.8635 and MSE of 0.0182, while the COD model reached an R-value of 0.9041 and MSE of 0.0058. When tested on unknown data, the BOD model performed well (<em>R</em> = 0.9244), but the COD model's lower generalization (<em>R</em> = 0.7149) suggests room for improvement.</div></div>","PeriodicalId":8947,"journal":{"name":"Bioresource Technology Reports","volume":"30 ","pages":"Article 102112"},"PeriodicalIF":0.0000,"publicationDate":"2025-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Bioresource Technology Reports","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2589014X25000945","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"Environmental Science","Score":null,"Total":0}

引用次数: 0

Abstract

Water demand is rising with population growth, making greywater reuse vital for sustainability. Constructed wetlands (CWs) utilize natural processes to treat greywater, but the role of water hyacinths and media in enhancing treatment efficiency remains unclear. This study assessed two CWs with water hyacinth: one with gravel media (WM) and one without (WOM). Over 90 days, the CW-WM showed significant removal efficiencies: 91.59 % for turbidity, 55.74 % for COD, 79.96 % for BOD, 75.97 % for phosphate, and 30.67 % for ammonia over CW-WOM. An artificial neural network (ANN) was employed to predict BOD and COD using input parameters like pH, EC, turbidity, TS, and TDS. The BOD model achieved an R-value of 0.8635 and MSE of 0.0182, while the COD model reached an R-value of 0.9041 and MSE of 0.0058. When tested on unknown data, the BOD model performed well (R = 0.9244), but the COD model's lower generalization (R = 0.7149) suggests room for improvement.

Abstract Image

查看原文本刊更多论文

评估利用水葫芦建造湿地处理中水：介质比较和基于 ANN 的预测模型

随着人口的增长，对水的需求也在增加，因此中水的再利用对可持续发展至关重要。人工湿地（CWs）利用自然过程处理灰水，但水葫芦和介质在提高处理效率中的作用尚不清楚。本研究评估了两种水葫芦CWs：一种是砾石介质（WM），另一种是无砾石介质（WOM）。90 d后，CW-WM的去除率达到了91.59%，COD去除率为55.74%，BOD去除率为79.96%，磷酸盐去除率为75.97%，氨去除率为30.67%。通过输入pH、EC、浊度、TS和TDS等参数，采用人工神经网络（ANN）预测BOD和COD。BOD模型的r值为0.8635，MSE为0.0182，COD模型的r值为0.9041，MSE为0.0058。对未知数据进行检验时，COD模型表现良好（R = 0.9244），但COD模型泛化程度较低（R = 0.7149），表明有待改进。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊