Model-based predictive control optimization of chiller plants with water-side economizer system

Abstract

A chiller plant with a water-side economizer (WSE) system is an environmental technology that utilizes the low-grade natural cooling source to optimize the cooling supply and reduce the energy consumption. Many studies focused on the real-time optimization of conventional chiller plants, but few researchers aimed at optimizing chiller plants with a WSE system. There is also a lack of optimization platforms that can provide a systematic evaluation of the energy saving potential of WSE under different optimization scenarios. Therefore, this study proposes a model-based predictive control (MPC) method to optimize multiple parameters of the WSE system. First, WSE system models and an optimization platform were developed using the equation-based Modelica modeling method. An actual real-world chiller plant system was considered to validate the chiller plant models. Then, the energy saving potential was demonstrated by considering 15 simulation cases, implementing three optimization strategies and five optimization frequencies. The obtained results indicated that the hourly optimization strategy can achieve the maximum energy saving potential of around 14.3% compared to the baseline model, followed by the daily optimization, with a reported 12% energy savings. The difference in energy consumption between the other three optimization frequencies was found to be small. The largest chiller plants’ energy efficiency ratio was 7.04 when performing hourly optimization for multiple parameters, which is 16.7% higher than that obtained by running annually optimization for a single parameter. Although multi-parameter optimization can save more energy than single-parameter optimization, it is time-consuming and requires extensive computational resources. The proposed MPC optimization method can provide a reference for engineers to select appropriate optimization parameters and frequency to be used for applications in practice.