PROCEEDINGS, 43rd Workshop on Geothermal Reservoir Engineering Stanford University, Stanford, California, February 12-14, 2018 SGP-TR-213 1 Hybrid Geothermal Heat Pumps for Cooling Telecommunications Data Centers David P. Zurmuhl 1 , Maciej Z. Lukawski 1 , Gloria A. Aguirre 1 , George P. Schnaars 1 , Koenraad F. Beckers 1,2 , C. Lindsay Anderson 1 , and Jefferson W. Tester 1 * 1 Cornell University: Cornell Energy Institute, Sibley School of Mechanical and Aerospace Engineering, Smith School of Chemical and Biomolecular Engineering, Earth & Atmospheric Sciences, Biological and Environmental Engineering, Ithaca NY 14853, USA 2 National Renewable Energy Laboratory, 15013 Denver West Parkway, Golden, CO 80401, USA *corresponding author: jwt54@cornell.edu Keywords: geothermal heat pump (GHP), ground-source heat pump (GSHP), data center cooling, computer room air conditioning (CRAC), air-source heat pump (ASHP), techno-economic modeling, TRNSYS ABSTRACT The technical and economic performance of geothermal heat pump (GHP) systems supplying year-round cooling to representative small data centers with cooling loads less than 500 kWth were analyzed and compared to air-source heat pumps (ASHPs). A numerical model was developed in TRNSYS software to simulate the operation of air-source and geothermal heat pumps with and without supplementary air cooled heat exchangers – dry coolers (DCs). The model was validated using data measured at an experimental geothermal system installed in Ithaca, NY, USA. The coefficient of performance (COP) and cooling capacity of the GHPs were calculated over a 20-year lifetime and compared to the performance of ASHPs. The total cost of ownership (TCO) of each of the cooling systems was calculated to assess its economic performance. Both the length of the geothermal borehole heat exchangers (BHEs) and the dry cooler temperature set point were optimized to minimize the TCO of the geothermal systems. Lastly, a preliminary analysis of the performance of geothermal heat pumps for cooling dominated systems was performed for other locations including Dallas, TX, Sacramento, CA, and Minneapolis, MN. 1. INTRODUCTION Geothermal or ground-source heat pumps (GHPs) utilize the relatively shallow ground as a heat source or sink to provide space or water heating and/or cooling. Depending on factors such as climate, particularly ambient air temperature and humidity, and price of electricity, GHPs are often the most energy efficient and cost-effective systems for space cooling. The IT equipment in data centers produces large amounts of heat and typically requires year-round cooling. About 40% of the energy consumed by a data center is typically used for cooling the IT equipment, which corresponds to 0.5% of the world’s electricity demand (Song, et al., 2015). The most commonly used data center cooling technologies rely on air-source heat pumps (ASHPs), which use the atmosphere as a heat sink. An alternative solution is to use GHPs utilizing a set of vertical boreholes which are typically more efficient because the ground remains at a moderate temperature year-round, whereas the ambient air temperature fluctuates throughout the year. Although the initial cost of GHP systems can be significantly higher than the cost of ASHPs, the reduced electricity consumption of geothermal systems over the course of their lifetimes can allow the initial cost to be recovered within a reasonable timeframe, while reducing the carbon footprint of data centers. Unlike GHP systems used in residential buildings for both space heating and cooling, a system used for data center cooling needs to transfer heat to the subsurface year-round. One of the main concerns of such systems is the potential increase in temperature of the geothermal well field over the lifetime of the system, resulting in diminished efficiency and cooling capacity of the heat pumps. In order to mitigate the expected temperature increase, an air-cooled heat exchanger – a dry cooler (DC) – can be added to the system to transfer heat generated by the IT equipment and stored in the subsurface to the atmosphere when ambient temperatures are low. An earlier study comparing the performance and economics of different cooling systems for cellular tower shelters with cooling loads of approximately 8 kWth (including ASHPs and GHPs equipped with DCs and/or air economizers) was conducted in our group and is documented in Beckers et al. (2014), Beckers (2016), and Aguirre et al. (2017). The study utilized computer models validated against data from a cellular tower demonstration site in Varna, NY. The main outcome of the study was that in most cases, an ASHP combined with an air economizer provided the lowest total cost of ownership while a GHP combined with an air economizer provided the lowest lifetime electricity consumption. A nationwide analysis of the cooling systems was then conducted using climate and hydrogeological data to produce maps of the total cost of ownership of the GHP and ASHP systems. An experimental study comparing the performance of GHPs and ASHPs over a one year period was conducted by Urchueguía et al. (2008). Although the studied systems were used for both heating and cooling, the experimental site was located in the warm climate of Valencia, Spain, making for a cooling-dominated application. The study concluded that GHPs compared to ASHPs save 43±17% of energy in heating mode and 37±18% of energy in cooling mode.