Assessment of Corrosion Potential Based on Water Quality Index in the Distribution Network of Urban Patna, Bihar, India

Abstract

Corrosion in the distribution network pipe can lead to pipe failure and water quality problems. This study assesses the corrosion or scaling potential based on the Water Quality Index (WQI) of drinking water in the distribution networks of Patna, Bihar, India. The water samples were collected from 18 points of the distribution network. In situ parameters like temperature, pH, electrical conductivity, and TDS were measured. Other parameters such as Alkalinity, Total hardness, Calcium, Magnesium, Chloride, Residual chlorine, Sulfate, Nitrate, and Dissolved oxygen were examined in the laboratory. Corrosiveness indicators, such as the Langelier saturation index (LSI), Ryznar stability index (RSI), Puckorius scaling index (PSI), Larson-Skold index (Ls), and Aggressive index (AI) are being used for water sample corrosion prediction. Experimental Corrosion rate (CR) is analyzed to show the actual prediction of corrosion. WQI was calculated to observe the effect of water quality on Corrosiveness indices and CR. A general conclusion was reached that LSI concludes 66.67% corrosive, 22.22% scaling, and 11.11% neutral, RSI concludes 88.88% corrosive, 5.56% scaling, and 5.56% neutral, PSI indicates 38.88% corrosive, 5.56% scaling, and 55.56% neutral, Ls indicates 94.44% scaling, and 5.56% corrosive, AI indicates 77.78% corrosive, and only 22.22% scaling. The average Experimental Corrosion rate is found at 1.91 mils per year. In this study, a weak correlation (r = 0.35) between Corrosion rate and WQI has been observed. A weak correlation is also observed between corrosion rate and corrosiveness indices (r < 0.5). It is concluded that the Corrosiveness Indices fail to represent the actual behavior of water.

Monitoring of traffic noise pollution in urban Patna, Bihar, India

“…The coefficient of correlation r less than 0.5 means weakly related and r greater than 0.5 means strongly related. 25,26 The correlation between noise indices is generated using IBM SPSS Statistics version 28 (given in Table 5). There is a strong correlation between NPL with TNI and NC (r > 0.5).…”

Research on deep-learning-based photovoltaic output estimation and multi-layer collaborative load prediction in distribution networks

“…Reference [8] proposes a deep convolutional neural network fault area localization method based on transfer learning and verifies that this method can accurately locate the fault area even in small sample situations. Reference [9] compares the LSTM prediction results with extreme learning machines, backpropagation neural networks, and K-nearest neighbor regression, indicating that the use of LSTM networks can significantly reduce prediction errors. Reference [10] proposes a distribution network day ahead optimization scheduling strategy based on improved deep reinforcement learning, and verifies that it is more flexible and targeted compared to traditional two-stage stochastic programming methods.…”