To predict the energy performance of a chilled water system more accurately, the hydraulic resistances of its water pipe network should be calibrated before simulation. However, it is a challenge to calibrate the hydraulic resistance of such a complex pipe network that are compose of chillers, terminal units, variable-speed pumps, valves and many pipes installed in different floors of a high-rise building. In this study, a new calibration method is proposed elaborately to adapt the vertical structure of the water pipe network in a high-rise building. The proposed calibration method utilized an optimization model and a general pipe network hydraulic solver. To overcome the severe nonlinear characteristic of the pipe network, Genetic Algorithm (GA) is used to solve the optimization model. Then, the proposed calibration method is validated in a real-life chilled water system in a high-rise building. With the hourly measured data from the chilled water system in operation in a typical summer day, the hydraulic resistances of 200 terminal units, 46 valves and 912 pipes are calibrated in detail. The calibrated hydraulic resistances are used to predict pressures and flow rates of the chilled water system in the next day. Compared with the uncalibrated simulation results, the average pressure error between the calibrated simulation results and measured data from the 42 onsite pressure meters is reduced from 2.2% to 0.6%. The average flow rate error between the calibrated simulation results and measured data from the 3 onsite flow rate meters is reduced from 5.3% to 0.9%.
Keywords chilled water system, high-rise building, hydraulic resistance calibration, optimization model, GA