Episode 3
HEAT REJECTION
Heat from the servers in data centers is usually rejected – via a heat sink – into the atmosphere. In some cases, it may be rejected into a lake or the ground. There are various methods for transferring this heat, including adiabatic coolers, dry coolers, cooling towers and air cooled chillers with remote condensers.
The choice depends on the location, the availability of water and power requirements. For data centers seeking to use less energy, free cooling solutions that use naturally cool air or water, instead of mechanical refrigeration, are increasingly attractive options. Factors such as the maximum temperature required for the servers and the availability of water and ambient temperatures will influence which cooling system to select.
FREE COOLING POTENTIALS FOR
DIFFERENT LOCATIONS
We have run sample calculations for 5 different locations to show the free cooling potential for each heat rejection method on each location.
Calculations are based on 1 MW with average local temperature data over the past 40 years.
Mean Temperature
Humidity
FOCAL-POINT: REDUCE PUE
AND WUE SELECTING THE RIGHT
HEAT REJECTION METHOD
Heat exchangers, used for heat dissipation, represent a modern technology for reducing energy consumption.
In this focal paper we consider factors that determine the choice of the right solution, such as location and ambient conditions.
We will discuss the different systems and the effects on your free cooling potential.
ā
READ MORE and DOWNLOAD OUR TECH-PAPER
BEIJING
Dry Cooler
Adiabatic Cooler
Cooling Tower
Mean Temperature
Relative
Humidity
Annual Temperature Distribution
Temperature [°C]
Hours per year
FREE COOLING
73 MWh
0 m³
74%
FREE COOLING
58 MWh
3767 m³
89%
FREE COOLING
16 MWh
18290 m³
94%
Power Consumption Water Usage
Calculations are based on 1 MW with average local temperature data over the past 40 years.
DALLAS
Dry Cooler
Adiabatic Cooler
Cooling Tower
Mean Temperature
Relative
Humidity
Annual Temperature Distribution
Temperature [°C]
Hours per year
FREE COOLING
101 MWh
0 m³
61%
FREE COOLING
86 MWh
6161 m³
80%
FREE COOLING
21 MWh
18656 m³
93%
Power Consumption Water Usage
Calculations are based on 1 MW with average local temperature data over the past 40 years.
DUBAI
Dry Cooler
Adiabatic Cooler
Cooling Tower
Mean Temperature
Relative
Humidity
Annual Temperature Distribution
Temperature [°C]
Hours per year
FREE COOLING
161 MWh
0 m³
28%
FREE COOLING
136 MWh
13836 m³
54%
FREE COOLING
34 MWh
19883 m³
69%
Power Consumption Water Usage
Calculations are based on 1 MW with average local temperature data over the past 40 years.
FRANKFURT
Dry Cooler
Adiabatic Cooler
Cooling Tower
Mean Temperature
Relative
Humidity
Annual Temperature Distribution
Temperature [°C]
Hours per year
FREE COOLING
37 MWh
0 m³
93%
FREE COOLING
29 MWh
514 m³
99%
FREE COOLING
9 MWh
17793 m³
100%
Power Consumption Water Usage
Calculations are based on 1 MW with average local temperature data over the past 40 years.
TRONDHEIM
Dry Cooler
Adiabatic Cooler
Cooling Tower
Mean Temperature
Relative
Humidity
Annual Temperature Distribution
Temperature [°C]
Hours per year
FREE COOLING
16 MWh
0 m³
99%
FREE COOLING
15 MWh
44 m³
100%
FREE COOLING
7 MWh
17643 m³
100%
Power Consumption Water Usage
Calculations are based on 1 MW with average local temperature data over the past 40 years.
HEAT RECOVERY
Data centers produce tons of heat, which is usually released into the atmosphere via a heat exchanger. There are several methods for optimizing the free cooling potential to see positive effects on your PUE and WUE.
Another way to improve the energy footprint is to use the waste heat. This means that the waste heat can be reused to heat a building or fed into a district heating network.
Heat recovery is another step towards your sustainability balance and we will talk more about this in episode 4.
PUE
Power Usage Efficiency
WUE
Water Usage Efficiency