Flow measurement uncertainty quantification for building central cooling systems with multiple water-cooled chillers using a Bayesian approach

Abstract

Measurement uncertainty has significant negative impacts on the operation and control of heating, ventilation and air conditioning systems. It is a big challenge and should be solved urgently. Existing studies focus on reducing the impacts of measurement uncertainty by developing uncertainty tolerant methods without quantifying the measurement uncertainties themselves. They therefore fail to fundamentally solve them. This study aims to directly quantify the measurement uncertainties of water flow meters in multiple water-cooled chiller systems using a Bayesian approach. A measurement uncertainty quantification strategy is proposed based on Bayesian inference and energy balance models, and the Markov chain Monto Carlo method is used to achieve the strategy. The site data collected from a chiller system are used to test the strategy. Four simulation tests with different levels of measurement uncertainty are conducted to further test and systematically validate the strategy. Test results show that the measurement uncertainties (both systematic and random uncertainties) of the water flow meters in the chiller systems can be quantified effectively and with acceptable accuracy. The strategy performs very well in quantifying random uncertainties of flow meters, and the relative errors range from 0% to 12.8%. The performance of the strategy in quantifying systematic uncertainties is also satisfactory, and the relative errors range from 0.1% to 36.57%. The proposed strategy is able to quantify measurement uncertainties and can be used to optimize the control of chiller systems and improve the reliability of chiller systems.