Scholarlink

<p>Recycling wastewater is one of the viable solutions to overcome freshwater scarcity; however, its suitability for reuse needs a systematic quality assessment. This study proposes an integrated hydrochemical and AI-based framework to evaluate the suitability of untreated domestic wastewater for a residential population of approximately 10,000 people at NIT Durgapur, West Bengal, India. Thirty-two samples from four different sources (boys, girls, residential, and junction point) were assessed to develop the Irrigation Water Quality Index (IWQI). Piper diagrams, Pearson correlation analysis, and principal component analysis (PCA) were used to interpret hydrochemical variations. A strong correlation was observed among EC, TDS, and major ions, while BOD and COD showed positive correlation and a negative relationship with dissolved oxygen. The wastewater was predominantly classified as mixed Ca<sup>2</sup>⁺–Mg<sup>2</sup>⁺–Cl⁻ type, reflecting anthropogenic influence, while PCA showed that the first two principal components explained 45.5% of the total variance, with salinity and organic contamination as significant controlling variables. IWQI values varied from 49.4 to 79.1 revealed that 84.4% of samples fall into combined poor (78.2%) and very poor (6.2%) categories, indicating the unsuitability for reuse. The sodium absorption ratio (SAR) values are between 0.41 and 1.03, signifying low sodium hazard. An artificial neural network (ANN) model (6–6–1–1 architecture) demonstrated strong predictive performance (<i>R</i><sup>2</sup> = 0.968, RMSE = 1.725), representing its reliability for IWQI estimation. Despite low sodium risk, elevated organic and ionic loads necessitate appropriate treatment before reuse. The research work provides an effective tool for decentralized wastewater management and reuse planning.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hydrochemical characterization and AI-based prediction for irrigation suitability of domestic wastewater in Durgapur, West Bengal, India

  • 期刊论文

  • 发表日期：

    2026年6月16日

查看全文
  • Shivam Saw,
  • Suchhanda Mondal,
  • Md Sarfraz Ahmad,
  • Rajib Ghosh Chaudhuri,
  • Supriya Pal,
  • Hirok Chaudhuri,
  • Bera Tushar Kanti,
  • Mrinal Kanti Mandal

摘要

Recycling wastewater is one of the viable solutions to overcome freshwater scarcity; however, its suitability for reuse needs a systematic quality assessment. This study proposes an integrated hydrochemical and AI-based framework to evaluate the suitability of untreated domestic wastewater for a residential population of approximately 10,000 people at NIT Durgapur, West Bengal, India. Thirty-two samples from four different sources (boys, girls, residential, and junction point) were assessed to develop the Irrigation Water Quality Index (IWQI). Piper diagrams, Pearson correlation analysis, and principal component analysis (PCA) were used to interpret hydrochemical variations. A strong correlation was observed among EC, TDS, and major ions, while BOD and COD showed positive correlation and a negative relationship with dissolved oxygen. The wastewater was predominantly classified as mixed Ca2⁺–Mg2⁺–Cl⁻ type, reflecting anthropogenic influence, while PCA showed that the first two principal components explained 45.5% of the total variance, with salinity and organic contamination as significant controlling variables. IWQI values varied from 49.4 to 79.1 revealed that 84.4% of samples fall into combined poor (78.2%) and very poor (6.2%) categories, indicating the unsuitability for reuse. The sodium absorption ratio (SAR) values are between 0.41 and 1.03, signifying low sodium hazard. An artificial neural network (ANN) model (6–6–1–1 architecture) demonstrated strong predictive performance (R2 = 0.968, RMSE = 1.725), representing its reliability for IWQI estimation. Despite low sodium risk, elevated organic and ionic loads necessitate appropriate treatment before reuse. The research work provides an effective tool for decentralized wastewater management and reuse planning.