Sensitivity and Uncertainty Analysis for Chiller Sequencing Control of the Variable Primary Flow System

Chiller sequencing control is a key function for the automatic control of chiller plants in air-conditioning system. It switches chillers on/off according to the building cooling load for cooling supply and energy saving. However, when facing uncertainties, sequencing controller may switch-on/off chillers improperly or frequently, leading to unstable operation or energy waste. In order to enhance the robustness of chiller sequencing control under uncertainties, an uncertainty analysis was performed for the primary variable flow system, taking account of possibly confronted uncertain factors in real operations. A chiller plant was constructed using the simulation platform TRNSYS. A typical sequencing control strategy was coded using MATLAB and inserted into TRNSYS by the open linking type. Associated uncertainties such as measurement uncertainty, control uncertainty or threshold uncertainty were extracted from site operational data of a real chiller plant and modeled using statistic methods. The impacts of each uncertainty were evaluated with Monte Carlo simulations. Their sensitivities were also ranked. Further, this study provides an guidance for the robustness enhancement of chiller sequencing control.