EPANET MODEL CLEAN WATER NETWORK CALLIBRATION USING MODIFIED ROUGHNESS C NUMBER HAZEN WILLIAM METHOD

Abstract

The Epanet model in clean water pipelines is intended for time efficiency in calculating hydraulic behaviour. Calibration of the water pipe network is to compare the simulation results of the model to actual field observations and to see that the model that has been made is truly valid and reliable as a tool to determine the hydraulic behaviour of the network system when a change input into the system is given (e.g. the addition of pipelines and number of customer tapping debits). The pipe roughness coefficient (C) is a number that indicates the amount of energy loss due to friction between the flowing fluid and the pipe wall. The increasing age of the pipe will cause the pipe wall roughness to increase, and the energy loss will be even more significant. This research aims to calibrate the pipe network model by adjusting the C value of the actual pipe roughness coefficient according to the installed age. An initial simulation model was carried out by entering the standard pipe C value in the Hazen Williams energy loss equation in the Epanet link/pipe properties. Then a simulation is carried out by modifying the value of C according to the pipe's service life. For 600 mm steel pipe, change the C value from 150 to 87 (pipe age >30 years), HDPE 315 mm pipe changes in C value from 140 to 100 (pipe age >10 years) and 25-50 mm PVC pipe from 140 to 100 (pipe age >10 years). From the simulation results of the Epanet model-field observations, the average compressive height was obtained for five observation nodes.of 2 tails was performed paired sample test the pressure height value of the model-observation. Before modifying the value of C, the paired mean difference test results showed a significant difference between the simulation results of the model on field observations with a model significance value of 0.004, which means <0.005. There is a considerable difference between the simulations and observations, and the model is considered invalid. However, after modifying the value of C according to the actual age of the pipe, a paired sample test T average value of the compression model-observation was 0.098 > 0.005), which means that there is no significant difference between the simulation and observation. This means that the model made is quite valid. Product moment correlation between the model simulation height - field observations, the value of R = 0.967, the relationship is robust between the two results.