Detailed and reduced order modeling of steady state counterflow mechanical draft cooling towers for analysis of Data Center energy efficiency

Air-cooled Data Centers absorb heat in central room air conditioners or air handlers and most commonly reject the heat through cooling towers to ambient air. The energy efficiency of the overall cooling system is determined by the energy efficiency of its components, its thermodynamic design and layout, and the protocol for its control, operation and its operational set points. The current work is part of an effort to develop a comprehensive analysis tool for thermodynamic modeling and analysis of such cooling systems. The tool, called Villanova Thermodynamic Analysis of Systems (VTAS), allows integration of sophisticated thermodynamic models of data center components into system layouts and then allows simulation of both steady state and transient system behavior from chip to cooling tower. This paper reports on the development of a steady-state model for a constant flow rate, counterflow mechanical draft cooling tower using a finite difference method and the subsequent synthesis of data from this model into a fast reduced order model developed using an Artificial Neural Network (ANN) approach. The outlet water temperature, moist air thermodynamic state, heat transfer, and exergy destruction are computed. By comparing the heat transfer and the exergy destruction from this model, an optimum range for the air flow rate was found for a set of inlet conditions based on the minimization of exergy destruction. An ANN model was developed using data generated from the detailed model. The ANN model is fast and accurate and is easily integrated into a system simulation code such as VTAS.