Available via license: CC BY 4.0
Content may be subject to copyright.
energies
Article
Study on the Design and Operation of an Outdoor Air-Cooling
System for a Computer Room
JiHyun Hwang 1,2 and Taewon Lee 1, *
Citation: Hwang, J.; Lee, T. Study on
the Design and Operation of an
Outdoor Air-Cooling System for a
Computer Room. Energies 2021,14,
1670. https://doi.org/10.3390/
en14061670
Academic Editor: Fabrizio Ascione
Received: 4 January 2021
Accepted: 4 March 2021
Published: 17 March 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1Department of Fire Safety Research, Korea Institute of Civil Engineering and Building Technology,
Goyang 10223, Korea
2Department of Construction Environmental System Engineering, Sungkyunkwan University, Suwon 16419,
Korea; jhhwang0127@gmail.com
*Correspondence: twlee@kict.re.kr
Abstract:
Due to the recent growth of the Internet and the rapid development of information and
communications technology, the number of data centers and the amount of power consumed are
expected to increase rapidly. However, these data centers consume large amounts of electric power
throughout the year, regardless of working days or holidays; hence, measures to save energy at these
facilities are urgently required. A numerical analysis was conducted in this study by changing various
variables to examine the effects of various design and operating conditions on outdoor air-cooling
systems and to derive some operating plans to minimize electric power consumption through the
introduction of outside air into the computer room. When designing the system, it is desirable to
select airflow considering various factors, such as the heat generated by the computer equipment, the
efficiency of the fan, and the performance of the refrigerator, rather than the theoretical maximum
outdoor airflow. As a result, the optimal design air volume was calculated according to the equipment
load. Consequently, the optimal design air volume of the modeled computer system was obtained
as 26,200 m
3
/h. The application of this optimal design air volume is expected to yield an annual
energy-saving effect of 56.1% compared with the power consumption in the air-conditioning period
during which the application of outdoor air cooling is impossible.
Keywords:
computer room; outdoor air cooling; design and operation; airflow ratio; energy saving;
field application
1. Introduction
There are a total of 158 data centers in Korea, and the domestic data center market is
expected to grow rapidly over the next five years, from 2019. There will be 32 additional
data centers planned to be built by 2025 [
1
]. Apart from large commercial data centers,
several public- and private-sector small- and medium-sized buildings have their own
computer rooms. Considering the huge amount of electricity consumed by computer
rooms when operating air conditioning throughout the year to maintain a constant and
secure environment for information and communications technology (ICT) equipment,
measures to save energy are urgently required. In a space with a high cooling energy
demand, such as a computer room, a significant amount of energy may be saved by
installing an outdoor air-cooling system in conjunction with a thermo-hygrostat that can
cool the indoor space by introducing a relatively low outdoor air temperature.
As an instance of a data center that uses an outdoor air-cooling system, LG CNS’s
Busan Global Cloud Data Center obtained the highest rating in Green Data Center Cer-
tification with the lowest power usage effectiveness (PUE) (1.39) [
2
]. PUE is the ratio of
the total amount of energy used by a data center to the energy consumed by ICT equip-
ment. Similarly, after installing an outdoor air-cooling system, Chuncheon Naver Data
Center operated an air cooler for fewer than 30 days a year [
3
]. In countries located in
geographically cold regions, several data centers actively use outdoor air-cooling systems,
Energies 2021,14, 1670. https://doi.org/10.3390/en14061670 https://www.mdpi.com/journal/energies
Energies 2021,14, 1670 2 of 17
with their PUE maintained below 1.25 (Google, Hamina, Finland) [
4
] through a control
operation configured to maximize the use of outdoor air or even below 1.2 (Facebook,
Dublin, Ireland) [4] by relying completely on the outdoor air-cooling system.
We analyze various research cases that achieved energy savings by applying an
outdoor cooling system to the data center [
5
–
20
]. Among them, we provide an overview
of the major studies conducted thus far. Soe et al. [
21
] performed an energy simulation
and reported that an energy saving of 49% was achieved by employing an outdoor air
conditioning system. Jeong et al. [
22
] performed energy simulations to compare the energy-
saving performances of various methods and identified that the enthalpy control technique
had the highest energy saving rate of 54.5% compared with the reference value. In addition,
10 data centers in Iran with various weather conditions analyzed the effectiveness of the
application of a outdoor cooling system and found that the PUE index improved from
10% to 12% [
23
]. From the design perspective, the simulation of the energy saving effect
of outdoor cooling due to changes in supply air temperature and design volume of an
outdoor air-cooling system was evaluated. As a result, at the same supply-air temperature,
the energy consumption was reduced as the design volume [24].
These studies focused on the prediction or analysis of the energy-saving effect expected
or achieved by installing an outdoor cooling system in computer rooms. None of these
studies, even those investigating actual installation cases, verified the energy-saving effect
resulting from the use of an outdoor air-cooling system; consequently, they did not explore
effective designs and operation methods of outdoor air-cooling systems.
Accordingly, we examined and analyzed the effects of relevant parameters on the
energy consumption of cooling devices by installing an outdoor air-cooling system in a
computer room. We also performed numerical analysis by varying the design and operating
parameters to identify the operating strategies to minimize the power consumption of
computer rooms by introducing outdoor air into the room. Additionally, by extending the
results of a previous study that verified the energy-saving effect through field application
of an outdoor air-cooling system to a computer room operated by a research complex in
Korea [
25
], we verified through simulations, with the physical change in the facility system
hardly attainable, the effect of a outdoor air-cooling system in conjunction with the design
strategy of the cooling system.
2. Analysis Models and Methods
2.1. Analysis Model
2.1.1. Cooling Load Model
We selected a computer room building located in the northern part of Gyeonggi for
analysis and established the analysis model as illustrated in Figure 1to calculate the heat
load of the computer room based on the following assumptions:
1.
Heat exchange occurs only through the wall and window in contact with the outdoor
air, excluding all other parts.
2. Heat transmission occurs one-dimensionally.
3.
The indoor temperature and humidity are maintained at the values set by the thermo-
hygrostat throughout the year.
4.
The heat release rate of the ICT equipment is a constant given value throughout
the year.
5.
The indoor heat release is included in the heat release of the ICT equipment because
the former is negligible compared with the latter.
The heat load in this analysis model, considering the given climate conditions and the
heat release of the ICT equipment, can be calculated using Equation (1).
.
q=.
qequip ±.
qth ±.
qinf +.
qsol (1)
where
.
qequip
,
.
qth
,
.
qinf
, and
.
qsol
are the sensible heat load elements affecting the indoor
temperature, denoting the heat load released from the ICT equipment; the transmission
Energies 2021,14, 1670 3 of 17
load determined by the outdoor air and building structure; the outdoor air infiltration
load; and the solar load, respectively. The heat load released from the ICT equipment is
assumed to be given, whereas the transmission and infiltration loads are calculated using
Equations (2) and (3), respectively. Equation (2) is calculated as the difference between the
heat transmission coefficient, the area, and the indoor and outdoor temperature, which is
the resistance to heat transmission value calculated using the wall thickness and thermal
conductance of the material, and the indoor and outdoor surface heat transfer coefficient.
Equations (3) and (4) are calculated as the difference in volume, number of ventilating
systems, and indoor and outdoor temperatures or humidity of the space. Equation (5),
which can be used to calculate the solar load, consists of the insolation for each orientation
and SHGC (solar heat gain coefficient). The latent heat load of the infiltration load obtained
by converting the moisture content affecting indoor humidity into heat is expressed by
Equation (5). .
qth =K×A×(24 −To)(2)
.
qSH−inf=0.29 ×N×V×(24 −To)(3)
.
qLH−inf=N×V×597 ×(0.0093 −ϕo)(4)
.
qsol =.
qsol ×SHGC (5)
Figure 1. Analysis model for the computer room.
2.1.2. Fan Model
The heat load
.
q
mentioned above should be removed by the thermo-hygrostat or
outdoor air-cooling system. The capacity of the fan installed to perform outdoor air cooling,
i.e., the airflow ratio .
Q, can be calculated using Equation (6).
.
Q=
.
q
C×∆t(6)
where
C
denotes the specific heat of air and
∆T
denotes the temperature difference between
the outdoor air and indoor air.
Pf
and
PR
, which denote the power consumptions of the
fan and thermo-hygrostat, respectively, are used to eliminate the heat load and can be
calculated using Equations (7) and (8), respectively
Pf=
.
Q×Ps
102 ×3600 ×η(7)
PR=
.
q
COP ×864 (8)
2.2. Analysis Methods
We prepared an analysis tool to calculate the heat load and to perform numerical
experiments for the analysis model, and we performed the annual cooling and heating
load calculations at regular intervals. Table 1outlines the physical properties applied to the
Energies 2021,14, 1670 4 of 17
numerical experiments other than the parameter set for analysis. The total floor area of the
building was 156.48 m
2
, the ceiling height was 2.97 m, and the number of server racks was
32. The heat released by the operation of the ICT equipment was 22,560 W. The values of
the parameters employed in the analysis were as follows: the heat transmission rate of the
outer walls of the building = 0.64 W/m
2·
K, the heat transmission rate of the double-layered
glass = 4.1 W/m
2·
K, and the heat transmission rate of the floor = 0.61 W/m
2·
K. Each heat
transmission rate is referenced in the design drawings. The air infiltration rate (relative to
the reference infiltration air per unit area) = 276 m
3
/h. The solar load was calculated by
obtaining the insolation for each orientation and by applying the Sing class clear 6T solar
heat gain coefficient of 0.67.
Table 1. Analysis model condition.
Site modeling
Site: Koyang, Gyeonggi
Floor area: 156.48 m2
Height: 2.97 m
Heat transmission rate of the wall: 0.64 W/m2K
Heat transmission rate of the window: 4.1 W/m2K
Heat transmission rate of the floor: 0.61W /m2K
Infiltration: 276 m3/h
SHGC (solar heat gain): 0.67
Servers Number of racks: 32 (600 ×1000 ×2000)
Heat release rate from the servers: 22,560 W (Estimated)
Figure 2shows a schematic diagram of the modeled computer room. It was cooled
though the air supply/exhaust duct installed in the high-temperature area and bottom
discharge. To improve the total-air recirculation cooling method using a general thermo-
hygrostat, the air supply and exhaust, and outdoor air damper were modeled, which was
integrated into the modeling of an outdoor air-cooling system capable of controlling the air
inlet rate. As cooling through an air inlet is impossible when the outdoor air temperature
exceeds a certain threshold, the cooling operation was modeled according to the total-air
recirculation method, in which the outdoor air was blocked and cooling was performed
only with a thermo-hygrostat. Similarly, when the outdoor temperature fell below a certain
degree relative to the indoor set temperature, the cooling system was stopped and cooling
continued only with outdoor air. A mixed operation method was also modeled, in which
the cold outdoor air in winter was mixed with the air recovered from indoor spaces to
achieve the desired air temperature for supply.
Figure 2. Schematic diagram of the outdoor air-cooling system.
Table 2lists the values obtained from the modeling of the outdoor air-cooling system
and the capacity of the thermo-hygrostat. The cost and space for air supply could be
Energies 2021,14, 1670 5 of 17
reduced by using both the thermo-hygrostat and the air supply fan. The total air volumes
of the air supply fan and the exhaust fan were 20,000 m
3
/h each, and a constant air volume
was supplied to the air supply fan and a variable air volume was supplied to the exhaust
fan.
Table 2. Capacity model of the outdoor air-cooling system and thermo-hygrostat.
Equipment Air Volume (m3/h) Electricity Consumption (kW)
SA Fan 10,000 ×1 EA 3.2 ×2 EA
Thermo-hygrostat Compressor - 7.2 ×4 EA
Outdoor air cooling EA Fan, Inverter 20,000 ×1 EA 4.7 ×1 EA
Based on the indoor temperature and relative humidity of a computer room as rec-
ommended by the American Society of Heating, Refrigeration, and Air Conditioning
(ASHRAE) in the flexible range of 18–27
◦
C and 20–80%, we set the indoor temperature
and relative humidity at 24
◦
C and 60%, respectively. Figure 3illustrates the fan model
used for the calculation.
Figure 3. Fan model used for the calculation (normalized).
The decision to use the outdoor air-cooling system considered in this study depended
on the indoor setting conditions, outdoor air condition, and the heat release level of the ICT
equipment as well as on the capacity and performance of the fan and thermo-hygrostat.
Among them, the outdoor air condition was the most decisive factor. That is, the application
of the outdoor air-cooling system was greatly influenced by the outdoor air condition. In
fact, while it can be used when the outdoor air temperature is significantly lower than the
indoor temperature, its application is impossible when the outdoor air temperature exceeds
a certain threshold and a thermo-hygrostat should be operated. Even when outdoor air
cooling is applied under the same indoor set temperature and humidity conditions, the air
volume and power consumption of the fan vary depending on the outdoor air condition.
We selected the climate model for Incheon, located close to the study site, to consider
the outdoor air condition, a determinant factor for the application of outdoor air cooling.
Figure 4plots the daily averaged outdoor air temperature using the 2019 data for Incheon
provided by the Korea Meteorological Administration. We selected three daily outdoor
temperature patterns considered representative of the season, i.e., when the daily averaged
outdoor air temperature is the highest (summer, S), lowest (winter, W), and middle (mid-
season, M), as listed in Table 3. Table 4lists the design and operating parameters for the
outdoor air-cooling system of the computer room used for the numerical analysis.
Energies 2021,14, 1670 6 of 17
Figure 4. Annual daily averaged outdoor air temperature.
Table 3. Averaged outdoor air temperature for representative days.
Types of Weather Conditions Averaged Temperature (◦C)
S (summer) 32.6
M (middle season) 11.9
W (winter) −11.7
Table 4. Design and operating parameters.
Design Parameters Operating Parameters
Outdoor Air-Cooling
Air Volume (m3/h) Equipment Heat Load (kW) Indoor Temperature (◦C)
20,000 20 15
30,000 30 17
40,000 40 19
50,000 50 21
60,000 60 23
70,000 70 26
80,000 80 28
90,000 90
100
110
Using the analysis model and tools developed in the previous section, we performed
numerical analysis to analyze the outdoor air-cooling system application plan and its effect
according to the design and operating parameters of the computer room. For numerical
analysis, the heat load was first calculated by considering the climatic conditions given at
regular intervals, the specifications of the building, and the heat load of the ICT equipment.
Then, after choosing the mode of cooling between outdoor air cooling using the cool inlet
air and apparatus cooling using the thermo-hygrostat, and by considering the indoor and
outdoor conditions in the calculation time slot, design parameters (fan air volume and the
heat load released from the equipment), and operating parameters (indoor temperature)
with regard to the specified fan and thermo-hygrostat capacities, we calculated the fan
air volume, the heat removed by outdoor air cooling, the power consumption of the fan,
and the power consumption of the thermo-hygrostat for both cooling modes. In the case
of outdoor air cooling using a fan, the air volume of the fan was variably controlled
Energies 2021,14, 1670 7 of 17
according to the heat load and outdoor air temperature, and the calculation was repeated,
incrementing the time interval.
When the calculation results showed variations in the cooling load and rated air
volume depending on the outdoor air temperature, we checked the air volume necessary
for outdoor air cooling and analyzed the air volume of the fan required for cooling and
the heat load removed by the fan according to the seasonal change of the representative
days. In addition, the annual power consumption and removed heat load were compared
and checked while varying the parameters. Based on the calculation results, we derived
the design air volume according to the heat load released from the ICT equipment in the
computer room studied and analyzed the effect of the indoor set temperature, which is an
operating parameter. Finally, we predicted the annual energy-saving rates for the cases
using the conventional cooling system and the thermo-hygrostat and applying the design
air volumes and design air volume.
3. Results and Discussion
3.1. Verification of the Calculation Results
We compared the cooling load obtained from the analysis with the corresponding
result of the experiment performed in the actual computer room modeled to verify the
validity of the results obtained through the theoretical analysis. To obtain the experimental
results, we performed outdoor air cooling using a fan in winter when air conditioning was
not used. Thus, the cooling load was calculated using the fan air volume per hour and the
temperatures of the inlet and outlet air.
Figure 5and Table 5present the results of the comparison between the theoretical and
experimental results in time series. Among the Measurement and Verification calculation
methods, we used the root mean square error (RMSE) method to calculate the error between
the simulation and measurement values shown in Equation (9) to verify the error rate. The
error between the theoretical and experimental results was 0.19%, which indicated that the
results of the theoretical analysis may be used in this study.
Mean error of the predicted value =r(P1−P2)2)
n−p−1
=q(3.77)2
12−1−1=0.19
(9)
Figure 5.
Predicted result of the daily cooling load compared with the measured one for a typical
day.
Energies 2021,14, 1670 8 of 17
Table 5. Comparison between the predicted and measured cooling loads.
Time
(h)
Outdoor Air Temperature
(◦C)
Cooling Load (kWh) (Y1−Y2)2
Y1-Predicted Y2-Measured
0:00 2.5 16.0 15.2 0.64
2:00 1.8 15.7 15.4 0.09
4:00 1.5 15.5 14.9 0.36
6:00 0.3 15.0 14.7 0.09
8:00 0.7 16.0 15.8 0.04
10:00 5.1 19.6 19.2 0.16
12:00 7.5 20.7 20.2 0.25
14:00 9 21.1 20.2 0.81
16:00 8.6 19.6 19.4 0.04
18:00 6.1 17.5 16.7 0.64
20:00 2.1 15.8 15.2 0.36
22:00 0.4 15.1 14.9 0.04
24:00 −0.1 14.9 14.4 0.25
Total 3.77
3.2. Changing Characteristics of Cooling Load
First, Figure 6shows the cooling load of the ICT equipment and outside air for one
year and the combined total load of the two according to the daily average outside air
temperature in the area where the computer room is located. As the cooling load is directly
affected by the state of the outside air, it has a large positive value when the outdoor
air temperature is high, whereas it may have a negative value when the outdoor air
temperature is low. This generally indicates that heating, not cooling, is needed at this time.
On the other hand, the cooling load of the ICT equipment is constant throughout the year
regardless of the condition of the outside air. Depending on the operating conditions of the
ICT equipment, the power consumption or the amount of heat generated may vary and the
cooling load may vary accordingly. Finally, the sum of these two cooling loads is the total
cooling load, and as the load of the ICT equipment has a constant value, the total load is
shifted upward by the amount of the external air load plus the load of the ICT equipment.
Figure 6. Variations of several related loads with the daily averaged outdoor air temperature.
Figure 7plots the fan air volume for outdoor air cooling applied to maintain the space
temperature at the set value of 24
◦
C. The graph shows that only a small amount of air is
necessary to remove the heat load released from the ICT equipment in the winter months.
This is because the load caused by the outside air is negative or has little effect owing to
the low outdoor air temperature. In contrast, as the air temperature gradually rises toward
Energies 2021,14, 1670 9 of 17
the middle period, the air volume increases as the cooling load increases. Except for some
transient periods during which the outdoor air temperature is relatively low, outdoor air
cooling is no longer possible and cooling must be performed by the thermo-hygrostat
installed additionally for hot seasons.
Figure 7. Annual variation of outdoor air volume with fan used for cooling.
Based on the above results, the air volume required to remove the heat generated
indoors including the heat released from the ICT equipment can be plotted according to the
variations in outdoor air temperature, as shown in Figure 8. As the outdoor air temperature
increases, a general tendency of increasing fan air volume is observed, tending toward
an exponential increase with a further increase in the outdoor air temperature. That is,
while the heat released from the ICT equipment can be removed even with a relatively
small air volume in areas with low outdoor air temperature, the fan air volume required
for offsetting the heat load increases drastically. Outdoor air cooling is possible only up to
the rated capacity of the fan (20,000 m
3
/h) and an outdoor air temperature of 20.4
◦
C, as
applied to the analysis model.
Figure 8.
Variation in design air volume used for outdoor air cooling according to outdoor air
temperature.
In other words, when an outdoor air-cooling system is installed in a computer room,
the period of outdoor air cooling can be extended by increasing the fan air volume; however,
excessive use of the fan sharply increases the power consumption, offsetting the cost-saving
effect compared with air conditioning. Therefore, instead of choosing the maximum air
Energies 2021,14, 1670 10 of 17
volume to perform outdoor air cooling as long as possible, an optimal air volume should
be selected by considering the heat released from the ICT equipment, the rated capacity of
the fan, and the performance of the thermo-hygrostat.
Figure 9plots the fan air volume required for cooling and the heat removed by the fan
when the indoor set temperature is maintained at 24
◦
C for the three models indicated in
Table 3, viz. for the three days representative of winter (W), summer (S), and midseason (M).
In Model S, outdoor air cooling was impossible because the outdoor air temperature was
higher than the indoor temperature, whereas in Model M, outdoor air cooling was observed
to be possible except for two hours (15:00–17:00) when the outdoor air temperature was
higher than the indoor temperature. This suggests that the air volume required for cooling
tends to change in proportion to the outdoor air temperature and that outdoor air cooling
is impossible when the rated capacity of the fan is exceeded or when the outdoor air
temperature is higher than the indoor set temperature. In the case of Model W, outdoor air
cooling could be performed throughout the day owing to the low outdoor air temperature.
This is because the lower the outdoor air temperature, the higher the cost-saving effect
because indoor heat can be removed with smaller air volume.
Figure 9.
Daily variations in outdoor air volume and removed cooling load for representative days.
The daily variations in heat removed by the fan show a general tendency similar to
the variations in the air volume. Model S, which did not perform outdoor air cooling, did
not incur any power consumption. In Models W and M, 100% and 90% of indoor heat,
respectively, were removed by outdoor air cooling. Thus, the performance characteristic
of the fan is reflected in the correlation between the air volume used for cooling and the
power consumption.
3.3. Effects of Design Parameters
The design parameters to be considered when installing an optimal outdoor air-
cooling system in a computer room are the equipment capacity and load, the capacity of
the fan for outdoor air inlet, the capacity and performance of the thermo-hygrostat, and
the insulation level of the building. We performed a range of numerical experiments using
the analysis model to identify the capacity of the fan capable of performing the required
cooling while consuming the least amount of energy by performing outdoor air cooling
according to the heat release level of the ICT equipment.
Figure 10 plots the variations in the power consumption of the fan and the cooling
load removed by the outdoor air depending on the design air volume of the fan with
respect to the equipment capacity, i.e., the equipment load, to investigate the effect of the
design air volume on the operating characteristics of the outdoor air-cooling system. As
expected, the equipment capacity was positively related to the annual power consumption
and the cooling load removed by the outdoor air. In addition, for any ICT equipment
Energies 2021,14, 1670 11 of 17
capacity, an increase in the design air volume was associated with a decrease in the annual
power consumption, followed by an inverse trend of increasing again. This suggests that,
under the given circumstances, including the ICT equipment capacity or heat release rate,
there is an optimal fan capacity for the most efficient execution of outdoor air cooling. It
was also observed that, as the ICT equipment capacity or heat release rate increases, the
optimal design air volume tends to increase as well.
Figure 10.
Variations in the annual electricity consumption with the design air volume by equipment
load.
Based on these results, Figure 11 plots the design air volumes when the annual power
consumption reaches the minimum value with respect to the capacity and load of the ICT
equipment, i.e., when the highest cost-saving effect is achieved. Table 6lists the annual
power consumption corresponding to each optimal design air volume. As the load of the
ICT equipment increases, the power consumption and design air volume tend to increase.
Notably, however, the design air volume increases more rapidly when the capacity of the
ICT equipment is smaller but decreases after the capacity reaches a certain level.
Figure 11.
Design outdoor air volume and electricity consumption according to the equipment load.
Energies 2021,14, 1670 12 of 17
Table 6. Electricity consumption according to the design outdoor air volume by equipment load.
Equipment Load
(kW)
Design Outdoor Air Volume
(m3/h)
Electricity Consumption
(MWh)
20 25,000 115
30 30,000 122
40 35,000 129
50 42,000 135
60 50,000 141
70 60,000 145
80 69,000 149
90 75,000 153
100 79,000 157
110 80,000 158
3.4. Effects of Operating Parameters
Once the outdoor air-cooling system is installed, the indoor temperature and the
design outdoor air temperature switching to outdoor air cooling are considered crucial
operating parameters. It is necessary to select an outdoor air temperature capable of
outdoor air cooling at the indoor set temperature. Particular care should be given to the
selection of outdoor air, especially because it is closely associated with the capacity of the
fan that executes outdoor air cooling.
Figure 12 plots the variations in annual power consumption for outdoor air cooling
according to the indoor set temperature. A linear inverse correlation is observed between
the indoor set temperature and the power consumption. Our analysis revealed that each
increase of 1
◦
C in the indoor set temperature results in a power-saving effect amounting
to 441 kWh per year, which is tantamount to 0.4% less power consumed every year.
Figure 12. Variation in the annual electricity consumption with the indoor set temperature.
Figure 13 plots the variations in the limit in outdoor air temperature allowed for
outdoor air cooling according to the design fan capacity with respect to the capacity of the
ICT equipment, i.e., equipment load. The graphs show that the outdoor air temperature
allowing outdoor air cooling increases with the increase in design fan air volume but
that the increase rate decreases asymptotically. Moreover, in areas with lower design
air volume, the difference caused by the difference in equipment load is larger, but the
difference decreases with an increase in the design air volume. This result can be used
Energies 2021,14, 1670 13 of 17
when setting the outdoor air-cooling switching time (outdoor air temperature value) for
efficient operation of the selected system.
Figure 13.
Outdoor air-cooling temperature according to the design outdoor air volume by equipment
load.
3.5. Outdoor Air-Cooling Operation Time
As examined above, the most determinant factor for the outdoor air-cooling system is
the outdoor air condition, which also determines the time periods of the day during which
the operation of the outdoor air-cooling system is allowed. For example, the system can be
operated throughout the day in winter and cannot be operated during the summer months.
Figure 14 shows the results of the analysis of the time periods of the day during
which outdoor air cooling is allowed depending on the annual daily average outdoor air
temperature under the given circumstances, as described in the previous section. As the
daily average outdoor air temperature becomes lower, the period of outdoor air cooling
becomes longer. In one of the experimental models of this study, it was estimated that
outdoor air cooling was possible in an area with the daily average outdoor air temperature
at 11
◦
C or lower; however, in an area with the temperature at 20.4
◦
C or higher, outdoor air
cooling was not possible. In the temperature range of 11–19.9
◦
C, outdoor air temperature
and outdoor air-cooling time displayed an inversely proportional relationship.
Figure 14. Outdoor air-cooling operating time according to the average outdoor air temperature.
3.6. Analysis of the Effects of Outdoor Air Cooling
The equipment load of the modeled computer room was 22,560 W, and the design
air volume of the outdoor air-cooling system was 20,000 m
3
/h. The optimal design air
Energies 2021,14, 1670 14 of 17
volume with respect to the equipment load, which was estimated in the previous section,
was observed to be 26,200 m
3
/h. Figure 15, Table 7presents the results of the monthly
comparison of power consumption and cost-saving ratio among three outdoor air-cooling
application models: (i) pre-application (conventional), (ii) post-application (design air
volume), and (iii) optimized application (optimal design air volume). The applications
of the design air volume (20,000 m
3
/h) and optimal design air volume (26,200 m
3
/h)
yielded the annual cost-saving ratios of 55.1% and 56.1%, respectively, relative to the power
consumption in the air-conditioning period during which the application of outdoor air
cooling is impossible.
Figure 15.
Comparison of the monthly electricity consumption and saving ratio by cooling method.
Table 7. Comparison of the monthly electricity consumption and saving ratio by cooling method.
Month Conventional
(kWh)
Outdoor Air-Cooling System
(kWh)
Saving Ratio
(%)
Optimal Design Air Volume Applied
(kWh)
Saving Ratio
(%)
1 21,193 5240 75.3 5639 73.4
2 19,254 4718 75.5 5067 73.7
3 21,828 5180 76.3 5535 74.6
4 22,030 5264 76.1 5447 75.3
5 23,281 7144 69.3 6772 70.9
6 22,847 14,699 35.7 13,339 41.6
7 23,783 21,509 9.6 20,446 14.0
8 23,857 23,749 0.5 23,030 3.5
9 22,859 15,995 30.0 14,871 34.9
10 23,136 6321 72.7 6164 73.4
11 21,591 5012 76.8 5326 75.3
12 21,515 5213 75.8 5592 74.0
Total 267,174 120,044 55.1 117,228 56.1
Table 8outlines the results of the analysis of the time periods during which the
application of outdoor air cooling was allowed under the conditions of design air volume
and optimal design air volume. The total durations of application of outdoor air cooling
were estimated to be 5545 h and 6739 h, respectively, out of 8760 h for the design air volume
(20,000 m3/h) and optimal design air volume (26,200 m3/h).
Energies 2021,14, 1670 15 of 17
Table 8. Expected monthly operation time of the outdoor air-cooling system.
Month 1 2 3 4 5 6 7 8 9 10 11 12 Total
Outdoor air-cooling system
operating time (h) 743 672 744 676 492 82 0 0 95 576 720 745 5545
Optimal design air volume applied
operating time (h) 743 672 744 716 704 422
155 42
351 725 720 745 6739
4. Conclusions
We examined the effects of various factors on the energy consumption of cooling
equipment in the context of installing an outdoor air-cooling system in a computer room.
We also performed numerical analysis by considering various design and operating pa-
rameters to identify the optimal operational strategy to minimize power consumption
in computer rooms by integrating an outdoor air-cooling system into an existing HVAC
system. The analysis results can be summarized as follows:
1.
A computer room has high cooling load throughout the year because of high equip-
ment load. When an air volume of 20,000 m
3
/h was applied, which is the rated
capacity of the fan used for the analysis model of outdoor air cooling, outdoor air
cooling could be allowed only up to the outdoor air temperature limit of 20.4
◦
C. An
analysis of the fan air volume required for cooling and the heat removed by the fan on
the representative days of three seasons (summer, winter, and midseason) indicated
that, the lower the outdoor air temperature, the greater is the efficiency of outdoor air
cooling because indoor heat could be removed with smaller air volume.
2.
While an increase in air volume can prolong the operation period of outdoor air
cooling, it also results in an increase in the power consumption of the fan. Therefore,
at the system design stage, it is recommended to select the optimal air volume by
considering the theoretically possible maximum air volume along with the equipment
load, the efficiency of the fan, and the performance of the thermo-hygrostat. We
performed numerical analysis and identified the design air volume when the annual
power consumption reached the minimum value with respect to the capacity and
load of the ICT equipment, i.e., when the highest cost-saving effect was achieved.
3.
Once the outdoor air-cooling system is installed, the indoor temperature and the
design outdoor air temperature switching to outdoor air cooling are considered
crucial operating parameters. Our analysis revealed that each increase of 1 ◦C in the
indoor set temperature results in a power-saving effect amounting to 441 kWh per
year, which is tantamount to 0.4% less power consumed every year.
4.
The results of the analysis of the time periods during which the application of outdoor
air cooling was allowed under the conditions of design air volume and optimal design
air volume are as follows. The total durations for application of outdoor air cooling
were estimated to be 5545 h and 6739 h, respectively, out of 8760 h for the design air
volume (20,000 m3/h) and optimal design air volume (26,200 m3/h).
5.
The optimal design air volume was calculated according to the equipment load.
Consequently, the optimal design air volume of the modeled computer system was
obtained as 26,200 m
3
/h. Application of this optimal design air volume is expected to
yield an annual energy-saving effect of 56.1% compared with power consumption
in the air-conditioning period during which application of outdoor air cooling is
impossible.
Energies 2021,14, 1670 16 of 17
Author Contributions:
Conceptualization, T.L.; Data curation, J.H.; Formal analysis, J.H.; Funding
acquisition, T.L.; Investigation, J.H.; Methodology, J.H., T.L.; Project administration, T.L.; Resources,
T.L., J.H.; Software, T.L., J.H.; Supervision, T.L.; Validation, J.H.; Visualization, J.H.; Writing—original
draft, J.H.; Writing—review & editing, T.L., J.H. Both authors have read and agreed to the published
version of the manuscript.
Funding:
This study was supported by a grant for the 2020 Urban Architecture Research Project
funded by the Ministry of Land, Infrastructure, and Transport (grant No. 20AUDP-B099686-06).
Informed Consent Statement:
Informed consent was obtained from all subjects involved in the
study.
Data Availability Statement: Not applicable.
Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design
of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or
in the decision to publish the results.
Abbreviations
A Area of the space (m2)
C Specific heat of air (kcal/kg·◦C)
ICT Information and Communications Technology
K Heat transmission rate (W/m2·K)
N Ventilation rate (m3/h)
P_f Fan power consumption (kW)
P_s Fan static pressure (mmAq)
P_R Power consumption of thermo-hygrostat (kW)
.
Q Airflow ratio (m3/h)
.
q HVAC load (kal/h)
.
qequip ICT equipment heat load (kal/h)
.
qth Transmission load (kal/h)
.
qSH−i n f (Sensible heat) infiltration load (kal/h)
.
qLH−i n f (Latent heat) infiltration load (kal/h)
.
qsol Solar radiation load (kal/h)
SHGC Solar heat gain coefficient
T_o Outdoor air temperature (◦C)
∆T Difference between the indoor and outdoor temperatures (◦C)
V Volume of the space (m3)
ηFan efficiency (%)
ϕAbsolute humidity (kg/kg)
Subscript
equip ICT equipment
f Fan
inf Infiltration
LH Latent heat load
o Outdoor air
R Thermo-hygrostat
s Static pressure
SH Sensible heat load
sol Solar radiation load
th Heat transmission rate
References
1.
Nam, D.Y. Data Center Is Growing at ‘Hyperscale’ Scale in 5G Era. 2019. Available online: https://www.techm.kr/news/
articleView.html?idxno=72449 (accessed on 10 June 2020).
2.
LG CNS Infraservice Group. Data Centers Joins the Energy-saving Movement with Sustainable Technology. LG CNS Infraservice
Group. Available online: https://blog.lgcns.com/881 (accessed on 24 August 2015).
3. Lee, K.B. Cool Wind on Mount Gubong; Naver Dater Center: Chuncheon, Korea, 2019.
Energies 2021,14, 1670 17 of 17
4.
Lee, J.Y.; Kim, J.H.; Yoen, C.K.; Jung, C.S. Action plan for the outdoor air cooling system in a data center. In Proceedings of the
2014 Winter Symposium of the Korea Institute of Civil Engineering and Building Technology, Seoul, Korea, 22 November 2014;
pp. 214–218.
5.
Cho, J.K.; Shin, S.H. Power and heat load of IT equipment projections for new data center’s hvac system design. Korean J. Air
Cond. Refrig. Eng. 2012,24, 212–217.
6. Cho, J.K. Energy Efficiency Technology in Data Centers. Korean J. Air Cond. Refrig. Eng. 2012,41, 60–67.
7.
Cho, J.K.; Shin, S.H.; Lee, J.Y. Case study and energy impact analysis of cooling technologies as applied to green data centers. J.
Archit. Inst. Korea Plan. Des. 2013,29, 327–334.
8.
Park, S.; Seo, J.; Jung, Y.H.; Chang, H.J.; Hwang, S.H. Energy Consumption analysis based on filter differential pressure when
adopting an air-side economizer system for a data center. Korean J. Air Cond. Refrig. Eng. 2013,25, 371–376.
9.
Jung, Y.H.; Jang, H.J.; Joe, J.H.; Park, B.Y. Cooling energy saving effects by hybrid economizer cycle in data center. Korean Inst.
Archit. Sustain. Environ. Build. Syst. 2013,11, 311–314.
10. Park, S.H.; Seo, J. Economic Evaluation of Air-side Economizer System for Data Center. Korean J. Air Cond. Refrig. Eng. 2014,26,
145–150.
11.
Kim, J.Y.; Chang, H.J.; Jung, Y.H. Study on the development of energy conservation system for data center by utilizing multi-staged
outdoor air cooling. Archit. Inst. Korea 2014,34, 267–268.
12.
Kim, S.H.; Yeon, J.Y.; Choi, Y.K.; Lee, E.K. A study on free cooling system for the energy saving in the data center. Korean J. Air
Cond. Refrig. Eng. 2011,7, 810–814.
13. Lee, J.Y.; Kim, J.H.; Yoen, C.K.; Jung, C.S. Green data design examples. Mag. Soc. Air Cond. Refrig. Eng. Korea 2013,42, 40–46.
14.
Lee, J.Y.; Kim, J.H.; Yoen, C.K.; Jung, C.S. Air-side economizer strategies for data centers. Mag. Soc. Air Cond. Refrig. Eng. Korea
2014,11, 214–218.
15.
Ham, S.W.; Kim, M.H.; Jeong, J.W. Energy saving potentials of various air-side economizers in a modular center for supply air
conditions and geological locations. J. Archit. Inst. Korea Plan. Des. 2014,30, 263–270. [CrossRef]
16. Bae, J.K. A case study on the building of green internet data center for green IT. Glob. Bus. Assoc. 2011,12, 329–348.
17.
Jo, H.; Kim, S.H.; Lee, S.K. A strategic design of green data center: The case of data center in the domestic public sector. J. Korean
Inst. Inf. Technol. 2012,10, 143–152.
18.
Jon, F. Dell’s Next Generation Servers: Pushing the Limits of Data Center Cooling Cost Savings. 2012. Available online:
http://dell.postclickmarketing.com/Global/FileLib/12g-vs-hp/12g-data-center-fresh-air.pdf (accessed on 28 February 2021).
19.
Masanet, E.; Shehabi, A.; Lei, N.; Smith, S.; Koomey, J. Recalibrating global data center energy-use estimates. Science
2020
,367,
984–986. Available online: https://science.sciencemag.org/content/367/6481/984 (accessed on 28 February 2020). [CrossRef]
[PubMed]
20.
Jaume, S. Cost-Effective Renewable Cooling on Green Data Centres; Catalonia Institute for Energy Research: Barcelona, Spain, 2014;
Available online: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.708.6831&rep=rep1&type=pdf (accessed on 10
December 2014).
21.
Seo, J.H.; Park, S. A study on the application of air-side economizer system for the data center. Korean Sol. Energy Soc.
2013
,33,
150–153.
22.
Ham, S.W.; Kim, M.H.; Choi, B.N.; Jeong, J.W. A study of direct air-side economizer control strategies for a modular data center.
Archit. Inst. Korea 2014,34, 243–244.
23.
Mahdi, D.D.; Namanlo, S.V. Potentiometric and economic analysis of using air and water-side economizers for data center cooling
based on various weather conditions. Int. J. Refrig. 2019,99, 213–225.
24.
Kim, J.H.; Aum, T.Y.; Jeong, C.S. Energy performance of air-side economizer system for data center considering supply tempera-
ture and design airflow rate of CRAH. Archit. Inst. Korea 2019,35, 181–188.
25.
Hwang, J.H.; Lee, T.W. Outdoor air-cooling system for a computer room and its corresponding energy-saving effect. Energies
2020,13, 5719. [CrossRef]
