Abstract:
World energy demands in buildings is increasing and expected to rise at greater
rates due to population growth, industrialization and rapid advancement in information
technology (IT). Big data computation and communication, internet of things (IoT) and
cloud computing demand expansion in existing and establishment of new data centers.
At present, data centers constitute 1% of total world energy consumption and within
data centers almost half of energy is consumed by the heating, ventilation and air conditioning (HVAC) system. Overall HVAC systems consume 10% of total world energy
and expected to increase three times by 2050. Improving the efficiency of HVAC system
is need of the hour. Energy consumption by air conditioners (AC) can be reduced by
using multi speed compressors or increasing the set point temperatures. In this research,
computational fluid dynamics (CFD) simulation of a localized data center has been done
for velocity field and temperature distribution in the server room. User defined functions (UDFs) have been used for switching air conditioners based on American Society of
Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) recommended maintaining temperatures for air cooled data centers. Energy consumption has been further
reduced by shifting temperature sensing location from air conditioner’s return to server’s
inlet. During initial transient of 250 seconds 12.8% of power consumption can be saved
by adhering to standard temperature range whereas, 33.24% energy can be saved by
shifting temperature sensing location from AC’s return to server’s inlet. In case of of
steady state approximately 32.4% energy can be saved by shifting sensor location from
AC’s return to server’s inlet.
Similar methodology is also applied to study and analyse the effect of flow augmentation in a non conventional, exhaust dependent air flow through engine radiator in
available specialized equipment. 11.89% increase at 3000 and 36.52% increase at 1000
engine’s revolution per minutes (RPMs) in heat transfer coefficient is achieved through
v
fixed flow augmentation. Numerical study shows that the flow augmentation is more
effective at low RPMs
