ABSTRACT

This study presents a data-driven system analysis approach intended to improve greywater reuse at the household/community level using a case study of ¨Granular Activated Carbon-Membrane Integrated-Multigrade Effluent¨(GAC-MI-ME) treatment system. It consists of a matrix of advanced filtration units and provides multi-grade water reuse options. Using experimental data with varying greywater strengths, artificial neural networks were applied to develop effluent prediction models for water quality parameters including turbidity, total dissolved solids, five-day biochemical oxidation demand, oxidation redox potential and pH. The results show that the GAC-MI-ME can improve water quality with terminal removal efficiency of 99% and above for turbidity and BOD. At the same time, the model simulations can effectively predict effluent quality, which can be applied as decision-making tools in selecting and the treatment trains of GAC-MI-ME. The concept may help broaden the greywater reuse by ensuring it meets the requirements of diverse water reuse regulations.

摘要

这项研究提出了一种数据驱动的系统分析方法，旨在以“颗粒活性炭膜-多级污水综合处理” （GAC-MI-ME）处理系统为案例，在家庭/社区一级改善中水回用。它由高级过滤单元组成，并提供多级水回用选项。利用具有不同灰水强度的实验数据，人工神经网络被用于为水质参数开发污水预测模型，包括浊度，总溶解固体，五天生化氧化需求，氧化还原电位和pH。结果表明，GAC-MI-ME可以改善水质，浊度和生化需氧量的最终去除率达到99％以上。同时，模型仿真可以有效地预测出水水质，可以作为GAC-MI-ME选型和处理过程的决策工具。该概念可以通过确保满足各种中水回用法规的要求来帮助扩大中水回用。