Performance and reliability analysis of a horizontal roughing filtration (hrf) system in the treatment of urban pond water: a case study of kolkata city

Abstract

Almost all real life datasets consist of missing values. These missing values require special attention on analysing the data. One approach to solve incomplete data problems is adoption of imputation techniques. This paper discusses the usage of linear interpolation technique to estimate the missing values. The study involved conducting experiments with a horizontal roughing filter which was present at Fluvial Hydraulics Laboratory, School of Water Resources Engineering, Jadavpur University to assess its capacity to lower turbidity levels. The raw water is collected from Jheel of university and measured turbidity values at inlet and outlet of HRF system during filtration to gauge the filter's efficiency. Then calculate the turbidity values by linear interpolation taking each fourth value of actual turbidity as reference. To assess the effectiveness of linear interpolation, the estimated turbidity values were compared with actual measurements. The thesis investigates the accuracy of the interpolation by calculating the percentage error between interpolated and actual turbidity values. This analysis provides insights into how well linear interpolation performs in filling gaps within the dataset and its impact on the overall evaluation of the filter’s performance. The results indicate that the horizontal roughing filter significantly reduces turbidity 82.92%, demonstrating its effectiveness as a water treatment solution. Linear interpolation was found to be a useful technique for estimating missing turbidity values, with error 9.825%. Generally within an acceptable range, suggesting that it is a reliable method for handling incomplete data in turbidity studies. This study contributes to the understanding of horizontal roughing filter systems and their application in turbidity removal. It also highlights the importance of accurate data interpolation in water quality assessment and provides insights into the reliability of different methodologies for handling incomplete data.