Partani Sadegh, Arzhangi Amin, Azari Hamidreza, Moheghi Hamidreza
Faculty of Engineering, Civil Engineering Department, University of Bojnord, Northern Khorasan, Iran.
Environmental Engineering Department, Centrale Méditerranée, Marseille, Provence- Alpes-Côte d'Azur, France.
Sci Rep. 2025 Aug 17;15(1):30108. doi: 10.1038/s41598-025-15866-0.
The interaction between chemical oxygen demand (COD) and biological oxygen demand (BOD5) in wastewater from Tehran's Paytakht and Nasirabad Industrial Parks is investigated in this work. Monitoring platforms of industrial parks were the base frame of monthly collection data for laboratory measurements (for BOD5 and COD) and in-situ measurements (for DO, EC and Temperature-T°C) with a frequency of 4-hour samples/day. Backward elimination regression analysis was employed as an integrated procedure to find out effective model removing ineffective independent variables. Multivariate Regression analysis showed a relatively strong linear relationship between COD and BOD, with independent variables with R²=0.64 and R²=0.59, respectively. A prediction model for BOD based on COD was found by analyzing important effluent quality variables using simple linear regression and a strong linear association (BOD = 0.433COD + 222) with R² = 0.94, MSE = 38,829, RMSE = 197.05 was obtained. In all of these regression analyses, model accuracy was assessed by conducting statistical tests on the residuals. To verify and improve the reliability and practicability of model, it is applied of industrial parks' wastewater records of countries around the world such as Egypt, France, India, Pakistan and Malaysia. The extracted model applied on some of the mentioned countries' records and the results of BOD prediction was matched by observations in 95% of reliability domain. Variation of BOD-COD ratio was least affected by pH and temperature; the results underline the requirement of localized validation resulting from industry-specific differences and promote cost-effective, quick wastewater evaluation, hence lowering reliance on laboratory-based BOD testing. It defiantly provides the opportunity of analytical and applied researches in south countries toward sustainable industrial wastewater management.
本研究调查了德黑兰帕塔克特和纳西拉巴德工业园区废水中化学需氧量（COD）与生化需氧量（BOD5）之间的相互作用。工业园区的监测平台是每月收集数据的基础框架，用于实验室测量（BOD5和COD）和现场测量（溶解氧、电导率和温度-T°C），采样频率为每天4小时一次。采用向后消除回归分析作为一种综合程序，以找出有效的模型并去除无效的自变量。多元回归分析表明，COD和BOD之间存在相对较强的线性关系，自变量的R²分别为0.64和0.59。通过简单线性回归分析重要的出水水质变量，发现了基于COD的BOD预测模型，得到了较强的线性关联（BOD = 0.433COD + 222），R² = 0.94，均方误差（MSE）= 38829，均方根误差（RMSE）= 197.05。在所有这些回归分析中，通过对残差进行统计检验来评估模型的准确性。为了验证和提高模型的可靠性和实用性，将其应用于埃及、法国、印度、巴基斯坦和马来西亚等世界各国工业园区的废水记录。将提取的模型应用于上述一些国家的记录，BOD预测结果在95%的可靠性范围内与观测值匹配。BOD-COD比值的变化受pH值和温度的影响最小；结果强调了因行业特定差异而进行本地化验证的必要性，并促进了具有成本效益的快速废水评估，从而降低了对基于实验室的BOD测试的依赖。它无疑为南方国家在可持续工业废水管理方面提供了分析和应用研究的机会。
