Key
22571
Conference Year
2016
Keywords
dynamic simulation, control, heat pump, Modelica, optimization
Abstract
Because vapor compression air-conditioners and heat pumps consume significant amounts of electrical power in today's residential and commercial buildings, energy optimization of these systems is becoming increasingly important from the perspectives of both environmental conservation and economic value. Corresponding efforts to improve the energy efficiency of these machines require attention to all stages of system design, installation, and operation, due to the myriad factors influencing power consumption. Among the many variables that must be optimized, one particularly salient variable is the mass of refrigerant contained within the cycle, or the refrigerant charge; this variable is strongly coupled to many other variables in the system, including the electrical power consumption, the system pressures, and the degree of subcooling and superheat in the heat exchangers. As such, the mass of refrigerant in the system must be carefully tuned for a given set of operational conditions to maximize the system's energy efficiency. In practice, field-installed vapor compression systems are often not charged with the mass of refrigerant that optimizes energy efficiency for the conditions in which systems actually operate. In accordance with the conventional view of the refrigerant charge as a static system parameter, the mass of refrigerant is often specified to maximize the average energy efficiency over a set of multiple conditions. This approach results in suboptimal energy efficiency at any one of the conditions within the rating set, and furthermore often results in lower energy efficiency at non-rated conditions. Such an impact is especially evident in reversible heat pump cycles because the optimal refrigerant mass for a cycle over a range of conditions in cooling mode is often very different than the optimal refrigerant mass in heating mode. As today's system manufacturers sell equipment across large geographic ranges with a wide range of ambient conditions and operational requirements, the cumulative impact of operating these systems with suboptimal refrigerant charge is generally a much higher rate of energy consumption than would be observed with cycles that incorporate an optimally specified refrigerant charge. In this paper, we describe a system architecture for a vapor compression system that enables the circulating refrigerant charge to be modulated as a function of time, effectively allowing the refrigerant charge to be optimized for a predicted or observed set of operational conditions. This is accomplished by dynamically controlling the amount of refrigerant sequestered in a storage vessel (referred to as a dynamic receiver) that is continuously coupled to the other components of the system. We first explore alternate system architectures that have been previously proposed for similar purposes, and elaborate on the opportunities that are afforded by this particular candidate architecture. A set of first-principles physics-based dynamic models are then developed using the Modelica language, and a candidate controller architecture is discussed that directly optimizes the electrical power consumption by using this new dynamic receiver. Finally, we will compare energy performance of this proposed system with that of conventional system architectures to evaluate its benefits over a range of operational conditions.