Optimization Of Vapor Compression Systems Via Simulation

VapCyc is a vapor compression refrigeration simulation tool which features charge estimates, component inter-changeability and is specifically geared towards optimization of system level variables using gradient based and genetic optimization routines. System level variables currently accounted for are: system refrigerant charge, COP, weight, capacity, cost and selection of individual component models. The optimization is being implemented as a single objective optimization subject to various constraints. This optimization is carried out over a given set of components as well as a variable set that may include any independent variable of the selected component models. VapCyc is introduced here in its early stages to demonstrate its usefulness for design and optimization, and thus justify further work on its simulation capabilities, component models and optimization routines. To this end, this paper presents several simple component models, and the resulting combinations of these components, which represent optima of several different objective functions and constraint sets. NOMENCLATURE Charge Refrigerant Charge (kg) Q& Heat Rejected (W) W& Work Output (W) m& mass flow rate (kg/s) D Diameter (m) L Length (m) V Volume (m) U Overall Heat Transfer Coefficient (W/m K) ηv Volumetric Efficiency ηs Isentropic Efficiency ∆P Pressure Drop (Pa) f Fanning Friction Factor Cp Specific Heat (J/kg K) P Pressure (Pa) T Temperature (K) h Specific Enthalpy (J/kg) s Specific Etropy (J/kg K) ρ Density (kg/m) INTRODUCTION Vapor compression refrigeration represents a large part of the United States energy consumption. Many computer simulations have been created over the years, dating back to approximately 1976. These simulations are created for the purposes of design, analysis, but rarely optimization of the system [1-3]. The large majority of the simulations created are specific to a single system, although some simulations have a large number of independent variables that can be specified, making them more general in application. Even a simple vapor compression system can have many independent variables, making complete control of a system during experimentation difficult if not impossible. Simulation of vapor compression systems is therefore beneficial to establish system performance over a rigidly controlled set of independent system variables. Once simulation is established to reproduce the behavior of the system in question, optimization of the system’s simulated performance can be conducted, and applied to the real system. VapCyc is a steady-state vapor compression refrigeration simulation tool created expressly for this purpose. VapCyc combines four independent component models and simulates the performance of the set. Optimization is then carried out for a set of system level variables as a function of the system independent variables, as well as any relevant component independent variables. VapCyc currently allows for a set of system dependent variables consisting of: COP, capacity, weight and volume, and a set of independent variable, consisting of: system charge, and component models. This paper presents VapCyc, as a tool for vapor compression refrigeration system simulation and optimization. In addition, this paper presents simplified component models used to validate the VapCyc system model and offer examples of system level optimization. Two examples are presented to demonstrate the usefulness of such a tool, and justify further work on the system simulation tool, optimization routine and component models.