Applying Minimum Night Flow to Estimate Water Loss Using Statistical Modeling: A Case Study in Kinta Valley, Malaysia

Minimum night flow (MNF) is a common method used to evaluate water loss in a water network. In 2010, the average percentage of non-revenue water for the state of Perak in Malaysia was 29.4 %, a figure which resulted in major financial, supply, and pressure losses, as well as excessive energy consumption. In this study, a statistical analysis of the water distribution network and a modeling of MNF were carried out to estimate water loss in Kinta Valley, Perak. Flow and pressure for 361 zones were monitored for 24 h using PrimeWorks software (version: 1.5.57.0). Thirty study zones were randomly selected from 361 zones. MNF was screened within the time band of 1:00 am to 5:00 am. A total of 20 factors for physical, hydraulic, and operational variables were selected and correlated with MNF (L/s). Multiple linear regression was used as a statistical technique to determine factors that contributed to MNF (L/s). Consequently, pipe length (m) and pipe age (year) were the main contributors to MNF (L/s). The statistical model was finalized with R-Sq 0.706 and then improved to R-Sq 0.779. Results of the study revealed that 84.9 % of MNF frequencies for the 30 study areas were found at the time band 2:15 am to 4:15 am; therefore, the mean MNF for each zone in 2010 was determined to be between 1:00 am and 5:00 am. Statistical analyses showed that number of connections, total length of pipe, weighted mean of age of pipe, and type of pipe (100 mm asbestos cement) contributed to MNF. Moreover, approximately 97.5 % of registered repairs were conducted on pipes with small diameters of less than or equal to 50 mm. Pipes within this size range are usually used as service pipes and service connections. Copyright Springer Science+Business Media Dordrecht 2013