Studies on energy performance of high ambient temperature green data centres.
Chua, Wee Jian.
Date of Issue2012
School of Mechanical and Aerospace Engineering
This report presents a simulation of cooling in data center within the context of Singapore’s climate. The modelling and simulation of the data center is executed with TRNSYS. The programme will carry out year-round simulations based on recorded local weather data. A data center space of 4000m³ was set up and studies were carried out to understand and select the potential energy saving methodologies available. All simulation parameters are within specifications of ASHRAE’s whitepaper titled ‘2011 Thermal Guidelines for Data Processing Environment – Expanded Data Center Classes and Usage Guidance’. The project first explores the suitability of an air-cooled system versus a water-cooled system, which is followed by the viability of adding an economizer into the conventional cooling system. Studies were also conducted to explore the suitability of a water-side economizer versus an air-side economizer. The results show that water-cooled systems are more efficient due to its higher efficiency rating. It also discovered that water-side economizers are not suitable for Singapore’s weather context as it operates best in areas with wet bulb temperatures of 12.8oC for more than 3000 hours per year. On the contrary, air-side economizers are a viable option as they managed to clock 537 hours to produce a 6.21% energy savings against the energy consumption of a whole year (8640 hours). The report also compares the PUE (Power Usage Effectiveness) of the conventional cooling system with the economizer-added system. It is found out that the PUE manages to achieve a rating of 1.83. Finally, the report will lead to the derivation of the optimized HVAC set-up with potential energy savings of up to 37% when compared to a conventional HVAC set-up.
DRNTU::Engineering::Mechanical engineering::Alternative, renewable energy sources
Final Year Project (FYP)
Nanyang Technological University