Research - Determining water and energy use of cooling towers of data centers and commercial buildings

The increasing use of AI by almost everyone from their daily lives to using it as a vital tool in their jobs has spurred a construction frenzy of data centers.

Such data centers can range in energy consumption from kilowatts to megawatts (with some already in construction in the gigawatt scale), and these massive installations require 24/7 cooling so they can keep operating efficiently.

Evaporating cooling is one technology often used to satisfy this cooling demand, typically via cooling towers (CT). These systems utilize water as the working fluid: heat from the server rooms is transferred to a chiller, which in turn transfers the heat to the water of the cooling tower so that it may in turn be cooled down and recirculated 

Understanding how much water is utilized with this type of cooling system is vital to help design sustainable buildings and minimize the use of our most vital resource. This was the aim of my time as a research assistant at Arizona State University, my work focused on developing thermodynamic models in python to estimate the water use of such cooling system. These models were then compared to real-world data to determine accuracy and potential areas of improvement

The final model was designed for the user to be able to optimize only airflow or only water flow or both and return relevant results. The latest version of the model tries to effectively implement a Model Predictive Control (MPC) system to optimize the air and water flow given the initial conditions. 

The input data is based on local average weather conditions, amount of cooling towers to be analyzed, and the specifications of the cooling tower model. The code then utilizes the data to evaluate the thermodynamic model and produce an approximate of water use and cooling achieved per cooling tower and for all the cooling towers combined.