A stochastic chiller optimization operation strategy based on uncertainty analysis

Abstract

Deterministic chiller optimization control strategies, such as COP optimization strategy, are intended to save energy based on deterministic sensor measuring data and equipment characteristics. However, the sensor data and the equipment characteristics are typically uncertain due to poor calibration of sensors, poor maintenance of chillers, etc., which could harm the energy-saving performance of deterministic chiller optimization operation strategies. In order to tackle this problem, a stochastic chiller optimization operation strategy based on uncertainty analysis is proposed in this paper. The strategy consists of three steps: (1) Analyze the uncertainty of the HVAC system and specify the probability distribution of each uncertain parameter. (2) Calculate the mathematical expectation value of energy consumption and return chilled water temperature in each operation plan under uncertainty. (3) Select the operation plan with the least energy consumption expectation and limited return chilled water temperature. The performance of the proposed strategy is validated on TRNSYS with measured hourly cooling load data of an office building located in Shanghai. Compared with the deterministic optimized operation strategy, the proposed stochastic strategy performs better on robustness (i.e., keeping return chilled water temperature within safe criteria) because of the consideration of measurement uncertainty. Also, compared with traditional operation strategy without optimization, the proposed strategy performs better on saving energy