Ref 03 Stat Energy China

JID: ENBENV
ARTICLE IN PRESS [m5GeSdc;September 2, 2021;19:22]
Energy and Built Environment xxx (xxxx) xxx
Contents lists available at ScienceDirect
Energy and Built Environment

journal homepage: http://www.keaipublishing.com/en/journals/energy-and-built-environment/
Research status and development trends of evaporative cooling

air-conditioning technology in data centers
Junjie Chu a, Xiang Huang b,∗
a
Institute of Building Environment and Energy, China Academy of Building Research, Beijing 100013, China
b
School of Environmental and Chemical Engineering, Xi’an Polytechnic University, Xi’an 710048, China
a r t i c l e i n f o a b s t r a c t
Keywords: This paper systematically describes the technical principles, evaluation indicators, system forms and research
Data center progress of air-side evaporative cooling air conditioning systems, water-side evaporative cooling air conditioning
Evaporative cooling systems and freon-side evaporative cold coagulation heat air conditioning systems of Data center. In order to
Air-side
reduce the energy consumption of the refrigeration and air-conditioning system in the Data center, the applica-
Water-side
tion conditions and scenarios of the different forms of evaporative cooling air-conditioning systems should be
Freon-side
considered comprehensively. Therefore, it is very important that the renewable energy-dry air can be used to the
greatest extent. These efforts would contribute to China’s 2030 "Carbon Peak" and 2060 "Carbon Neutral."
1. Introduction a conference call to promote the integration of the national one net-
work and promote the development of 5G and data center business. In
A Data center is a construction site that provides operational envi- February, the Ministry of Industry and Information Technology issued
ronments for centralized electronic information devices, including host the "Notice on Adjusting the Internet Backbone Network Settlement Pol-
rooms, auxiliary districts, and support areas [1].and it can be a building icy", proposing to accelerate the integration of network resources and
or a part of a building. In recent years, the demand for mobile Internet, promote the development of the Internet industry and digital economy.
cloud computing, and big data applications has shown explosive growth. At the beginning of 2020, during the national coronavirus epidemic
At the same time, driven by innovations in information technology, the prevention and control, the government, enterprises and institutions
fields of smart terminals, VR, artificial intelligence, wearable devices, used a large number of scientific and technological epidemic prevention
Internet of Things, 5G, etc have developed rapidly, and "Internet +" has methods, including remote offices, distance education, and e-commerce,
accelerated its penetration into various industries. These have greatly in- which developed a large number of emerging customers, driving China’s
creased the scale of data storage, computing power, and network traffic IDC industry customer demand scale magnitude to increase, and pro-
and accelerated the construction of a new generation of Data center in- moting China’s IDC business market to maintain a stable growth situa-
frastructure into an acceleration period. As we all know, China’s mobile tion. As shown in , it is expected that in 2022, China’s IDC business mar-
Internet has the largest amount of information access and mobile phone ket will exceed 3200.5 billion yuan, a year-to-year increase of 28.8%,
ownership in the world. Likewise, among the top 30 listed Internet com- and enter a new round of outbreaks.
panies in the world, 10 of them are Chinese companies [2]. There is no At the same time, such large information visits and processing vol-
doubt that China is the most active country in the development of infor- umes bring about the rapid construction and development of the Data
mation industry technology. Based on the development and application center. With the development of the entire society, the demand for Data
of information technology, the government, enterprises and institutions centers continues to grow, and due to the high-density use of IT equip-
have increased the extent of data centers and network resources to prove ment and the uninterrupted operation of the Data center 24 hours a day,
rich infrastructure resource security for IDC (Internet Data center) busi- 365 days a year, the data center’s consumption of electric energy has in-
ness market development. Therefore, when the Standing Committee of creased sharply. Relevant data shows that the total power consumption
the Political Bureau of the CPC Central Committee held a meeting on of data centers in China in 2018 was 160.889 billion kWh, which is
March 4, 2020, it clearly emphasized the need to speed up the construc- more than the total power consumption of the whole society in Shang-
tion of new infrastructure such as 5G networks and data centers. On hai in 2018 (156.7 billion kWh), which is equivalent to the total power
March 2, the State Administration of Radio, Film and Television held consumption of the entire Three Gorges Dam in China. Annual power
∗
Corresponding author.
E-mail address: [email protected] (X. Huang).
https://doi.org/10.1016/j.enbenv.2021.08.004
Received 3 May 2021; Received in revised form 9 August 2021; Accepted 13 August 2021
Available online xxx
2666-1233/Copyright © 2021 Southwest Jiatong University. Publishing services by Elsevier B.V. on behalf of KeAi Communication Co. Ltd. This is an open access
article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Please cite this article as: J. Chu and X. Huang, Research status and development trends of evaporative cooling air-conditioning technology in data
centers, Energy and Built Environment, https://doi.org/10.1016/j.enbenv.2021.08.004
JID: ENBENV
J. Chu and X. Huang Energy and Built Environment xxx (xxxx) xxx
speaking, the refrigeration and air-conditioning system can introduce a

Nomenclature number of energy-saving design concepts and process technologies flex-
ibly to ensure the efficient operation of the Data center. Therefore, the
Symbols key to energy savings of the Data center is the energy saving design of
db dry bulb temperature [°C] the refrigeration air conditioning system.
wb wet bulb temperature [°C] Among them, the energy consumption of the traditional cooling sys-
dp dew point temperature [°C] tem of the data center includes components, such as compressors, water
DEC direct evaporative cooling pumps, and wind turbines, as shown in Fig. 4. Regardless of the air-
IEC indirect evaporative cooling cooled precision air conditioning system or the water-cooled precision
DIEC dew point indirect evaporative cooling air conditioning system, the energy consumption of refrigeration com-
𝜂 efficiency [%] pressors has reached more than 50% of the total energy consumption
𝜌 the density of product air of the cooler[𝑘𝑔∕𝑚3 ] of the Data center cooling system. Therefore, the key to reducing the
𝑄 cooling capacity of the cooler [kW] energy consumption of the Data center cooling system is to improve the
𝑉 the volume of air [𝑚3 ] efficiency of mechanical cooling or use natural cold sources instead of
𝑖 the air enthalpy value [𝑘𝐽 ∕𝑘𝑔] mechanical cooling. It is thus clear that free cooling in the Data center
𝜔 sub-wet bulb efficiency of indirect evaporative precool- cooling system can significantly reduce the energy consumption of the
ing unit [%] cooling system.
Γ𝑤 cooling efficiency of the water drenching filler on the
water-side [%] 2.2. Data center environment characteristics
Data center environmental parameter requirements have a signif-

generation [4]. As the density of IT equipment in data centers continues icant impact on the energy consumption of refrigeration and air-
to increase, the demand for cooling systems has increased dramatically conditioning systems and the reliability of IT equipment operation. The
in recent years. This change not only increases the demand for cooling US Heating Refrigeration and Air Conditioning Engineer Society estab-
system capacity, but also exposes the inefficiency of traditional cooling lished a technical committee 9.9 (ASHRAE TC 9.9) on Data centers in
methods in data centers. 2001. It is a professional research team composed of data communi-
cation equipment manufacturers, owners and managers of data equip-
2. Data center energy consumption and environmental ment end users, government agencies, research institutions, consulting
characteristics agencies and laboratory testing experts. To enable everyone in the in-
dustry to use universal languages and better understand the impact of
2.1. Data center energy consumption characteristics IT equipment cooling requirements on the Data center, ASHRAE TC 9.9
published the first edition of "Data Processing Environment Hot Guide"
The energy consumption content of the Data center is mainly di- in 2004 [5]. The temperature range recommended for level 1 and level
vided into three parts: (1) IT equipment and software; (2) power distri- 2 IT equipment is 20 to 25°C, and the relative humidity range is 40% to
bution systems; (3) refrigeration air conditioning systems. The energy 55%. The allowable air intake conditions for level 1 IT equipment are
consumption statistics of each part are shown in Fig. 3. In the propor- 15~32°C, the relative humidity is 20%~80%, and the maximum dew
tion of energy consumption, the energy consumption of IT equipment point temperature is 17°C; the allowable air intake conditions for level
accounts for about 40% of the total energy consumption of the Data cen- 2 IT equipment are dry-bulb temperature 10~35°C, relative humidity
ter, and the air conditioning system accounts for 40%. However, in the 20%~80%, maximum dew point temperature 21°C. To guide operators
development of energy-saving technology, IT equipment and software to maintain the high reliability of the environment in the Data center
energy consumption are mainly affected by the development of informa- and operate the Data center in the most energy-saving manner, ASHRAE
tion technology. Under the premise of ensuring Data center function and TC 9.9 continued to update this guide in 2008, 2011 and 2015 [6-8],
reliability, the space for improvement is relatively limited. Relatively and expand the recommended range of environmental air in the oper-
Fig. 1. 2014-2019 China IDC business market

size and growth [3].
1800 41.80% 39.30% 45.00%
37.80%
1600 40.00%
Hundred million yuan
32.40%
1400 29.80% 35.00%
27.20%
Growth rate
1200 30.00%
1000 25.00%
800 1662.5 20.00%
600 1228 15.00%
946.1
400 714.5 10.00%
518.6
200 372.2 5.00%
0 0.00%
2014 2015 2016 2017 2018 2019
Time
Market size Year-on-year growth
2
JID: ENBENV
Fig. 2. 2019 - 2022 China IDC business market

scale and forecast [3].
3500 30.00%
28.80%
3000 29.00%
Hundred million yuan
2500 27.20% 28.00%
Growth rate
26.90%
2000 27.00%
1500 25.30% 3200.5 26.00%

2485.7
1000 1958.2 25.00%
1562.5
500 24.00%
0 23.00%
2019 2020(E) 2021(E) 2022(E)
Time
Market size Year-on-year growth
Fig. 3. Proportion of China’s data center energy consumption in 2019.
Table 1 added value of environmental parameters in the 2015 updated version.

Recommended Values of the ASHRAE TC 9.9 Data center Environment Param- The additional value of the A1 to A4 environmental parameters is shown
eters. in the wet graph in Fig. 5.
Recommended Particulate matter or gas contamination in the Data center operation
environment will have a serious impact on the reliability of IT equip-
2004 version 2008/2011 version 2015 version
ment. Particulate pollution can cause intermittent failures of signal-
Dry bulb temperature limit 20°C 18°C 18°C sensitive circuits, and the accumulation of particulate matter on the
Dry bulb temperature limit 25°C 27°C 27°C
cooling surface of IT equipment can seriously affect its cooling perfor-
Dew point temperature limit — 5.5°C -9°C
Dew point temperature limit — 15°C 15°C mance. Gas contamination can lead to two common patterns of hard-
Relative humidity limit 40% — — ware failure: copper creep corrosion on printed circuit boards and silver
Relative humidity upper limit 55% 60% 60% terminal corrosion of surface mount resistors [9]. Therefore, ASHRAE
TC 9.9 issued the first edition of "Data communication environment
in the Data Communication environment" in 2009, which detailed the
ating environment of Data center IT equipment. The environmental air monitoring, prevention and control of particulate matter and gas pollu-
recommendation range of IT equipment has shown in Table 1, so that tion in Data centers. In 2013, ASHRAE TC 9.9 updated and published
an increasing number of locations around the world can take full advan- the second edition of the "Data communication environment in the data
tage of free cooling technology to dissipate heat in data centers over a communication environment". For particulate pollution, the Data center
longer period of time. Table 2 shows the recommendation value for the operating environment must be maintained to ISO 14644-1-8, as shown
Data center environment parameters of the four versions of ASHRAE. in Table 3, but also need to pay attention to the relative humidity of dust
The ambient temperature (inlet temperature) running the server is re- tidal solution, in which the particles absorb enough moisture to become
laxed from 20 to 25°C to 18 to 27°C. In the updated version of 2015, wet, so as to be conductive and corrosive to the metal on the printed
the dew point temperature of the recommended value continues to de- circuit board. [10,11]. According to the recommendations of the "rat-
crease, and the lower limit of the dew point temperature reaches -9°C. ing computer and data processing chamber unit air conditioner", indoor
Admittedly, this is even unprecedented. air can be continuously filtered using a MERV (Minimum Efficiency Re-
In addition to the recommended values of the environmental param- porting Value) 8 filter [12]. According to “particle and gas pollution in
eters, ASHRAE has continued to make more detailed amendments to the the data communication environment”, air entering the Data center can
3
JID: ENBENV
Fig. 4. Typical data center precision air conditioning system

(a) Air-cooled precision air conditioning system.
(b) Water-cooled precision air conditioning system.
Table 2
Additionally ASHRAE TC 9.9-2015 Data center Environmental Parameters [8].
Grade Dry bulb Humidity range (%) Maximum dew point

temperature (°C) temperature (°C)
A1 15~32 Dew point temperature-12°C Relative humidity 8%~ 17

Dew point temperature -27°C Relative humidity 80%
A2 10~35 Dew point temperature -12°C Relative humidity 8%~ 21
B 5~35 8%~80% 28
C 5~40 8%~80% 28
4
JID: ENBENV
Fig. 5. Additional Wet Map of the ASHRAE TC 9.9-2015 Data center Environmental Parameters [8].
Table 3
ISO Standard 14644-1: clean zone air particulate cleanliness level.
The maximum number of particles in the air (Particle particles per cubic meter is greater than or
equal to the designated size)
ISO
(Particle size)
> 0.1 𝜇m > 0.2 𝜇m > 0.3 𝜇m > 0.5 𝜇m > 1 𝜇m > 5 𝜇m
Level 1 10 2
Level 2 100 24 10 4
Level 3 1000 237 102 35 8
Level 4 10000 2370 1020 352 83
Level 5 100000 23700 10200 3520 832 29
Level 6 1000000 237000 102000 35200 8320 293
Level 7 352000 83200 2930
Level 8 3520000 832000 29300
Level 9 8320000 293000
be filtered through MERV 11 to MERV 13, and for data centers with midifiers, the use of these devices can offset the income of free cooling.
high levels of gas pollution, gas phase filtration is highly recommended The results also show that if the Data center cooling setpoint temper-
[13]. ature decreases, the energy savings will be reduced by 2.8 to 8.5% for
every 2°C reduction. The study of Siriwardana et al. [15]. suggested that
3. Data center free cooling technology introduction the energy-saving potential of direct air-side free cooling of Australia in
different climates, and their results showed that free cooling can reach
Under the conditions that meet the intake environment requirements more than 60% in the cold southern region. Depoorter et al. [16] consid-
of IT equipment, the Data center cooling system can be closely related ered that it is advantageous to use direct air-side free cooling anywhere,
to how many free cooling methods there are and that free cooling can which can reduce the energy consumption of the entire Data center by
operate closely to outdoor meteorological conditions in the geographic 5.4% to 7.9% (depending on the location). Oró et al. [17] studied the
location of the data center. Data centers in cold and cold areas typically potential of a direct air-side free cooling combined thermal energy stor-
use free cooling systems to dissipate heat for IT equipment. The classifi- age strategy. The study was conducted in five European cities affected
cation of free cooling technologies in the data center is shown in Fig. 6. by different climatic conditions and counted the number of hours oper-
The free cooling system in the data center is shown in Fig. 7. ated each year based on the ASHRAE recommendation guidelines. The
Lee and Chen [14] studied the energy-saving potential using direct results showed that Stockholm and Amsterdam have the greatest poten-
air-side free cooling in 17 climate zones. The results show that the use tial to cool in the use of external air, which can exceed 6500 h each
of energy-savings in mixing moist, warm, and warm marine climate dis- year.
tricts can significantly save energy, but in terms of dry and moist climate There is not only wind-side natural cooling technology, but also
zone efficiency, mainly due to high-energy fans, humidifiers and dehu- water-side natural cooling. Clidaras et al. [18] proposed a Data center
5
JID: ENBENV
Fig. 6. Data center Free cooling Technology Classification

(a) Direct free cooling on the air-side. (b) Indirect free cooling on the air-side
(c) Free cooling system on the water-side.
located next to or approaching the ocean, as shown in Fig. 8. The Data 4.1. Direct evaporative cooling
center is equipped with a closed-loop cooling system, which coolant
flows and heat is transferred to seawater through a heat exchanger. The 4.1.1. Technical principle
energy of seawater natural movement is captured and converted into The direct evaporative cooling system on the wind side of the data
the input power of the cooling pump. center is shown in Fig. 9, that is, the outdoor air sent to the data center
Evaporative cooling technology is an efficient free cooling technol- first comes into direct contact with water through the direct evaporative
ogy. There are rich with natural energy in areas of relatively high lat- cooling filler, and the temperature of the air drops due to the absorption
itudes such as solar, wind and dry air energy,so that there are unique of latent heat of vaporization by the evaporation of water, and the low-
advantages to building large-scale Data centers with green energy sav- temperature air is used for cooling the data center. Making the most of
ing. The cold and long duration of winter in these regions provides great outside air, the unit humidifies and cools the data center air with direct
space for data center cooling systems to adopt natural cooling sources; evaporative cooling (DEC) in the summer, while in the winter, the unit
the ambient air is hot and dry in summer, and the air itself has a large uses a portion of the server room return air to ensure control of the
dry bulb-wet bulb temperature difference, which is particularly suitable humidity of the supply air without the presence of condensation, which
for the application of evaporative cooling technology, and the combina- can affect the servers in the server room.
tion of evaporative cooling technology and data center natural cooling
technology. Such conditions can further extend the operation time of
natural cooling, thus achieving the maximum reduction of energy con- 4.1.2. Data center application case
sumption of data center cooling system. Domestic and foreign scholars have also conducted relevant research
and analysis on the application of evaporative cooling technology in the
air-side natural cooling system of data centers. Scofield [21] analyzed
the impact of applying free cooling on the air side of a direct evaporative
4. Data center air-side evaporative cooling cooler on reducing cooling energy consumption in the Data center. Cur-
rently, the most famous spray direct evaporative cooling method is the
Evaporative cooling technology is an energy-saving cooling technol- large data center built by Facebook in Plinville, USA [22–24], which
ogy that uses water as the cooling medium and reduces the temperature uses the air-side of the spray evaporative cooler to naturally achieve
of the cooled medium (air or circulating water and other media) directly enormous cooling energy savings. It is worth noting that they use their
or indirectly without the assistance of a compressor through the natu- own custom IT equipment and thermal environment range. The data
ral characteristic of heat absorption by evaporation when water is in center uses direct evaporative cooling with all fresh air. This is because
direct contact with air [19]. As can be seen from the above description, Prineville is located in the high desert with a very dry climate, and al-
energy-saving cooling technology in Data centers is mainly divided into though the maximum outdoor dry bulb temperature is 40.56°C through-
two basic forms: one is air-side free cooling, through the use of natu- out the year, the Facebook data center IT equipment can adapt to a high
ral cold air or cold water (lake water or sea water, etc.) from outside temperature of 43.3°C, so the energy saving effect is very significant. The
to directly or indirectly naturally cool the air that dissipates heat for Amazon Cloud Computing Data center at Zhongwei Western Cloud Base
IT equipment in the Data center; the other is water-side natural cool- in Ningxia, China [25] uses direct air-side free cooling in the data center
ing, through the use of natural cold air or cold water (lake water or sea cooling system, as shown in Fig. 10, based on local cold winters and cool
water, etc.) from outside to indirectly naturally cool the refrigerant wa- summers, with an annual average temperature between 7.3 and 9.5°C
ter that dissipates heat for IT equipment in the Data center. In order to throughout the year for outside air cooling sources. Mu Zhenghao and
be better integrated with the free cooling methods of Data centers, the Wang Yingzhen et al [26] explored seven operating conditions switch-
evaporative cooling technology can be divided into two main forms ac- ing strategies for the air-side free cooling system with direct fresh air
cording to the cooling medium: air-side evaporative cooling technology evaporative cooling as the main cooling source used in the Qihoo 360
and water-side evaporative cooling technology [20]. cloud computing data center in Zhongwei, Ningxia, as shown in Fig. 11.
6
JID: ENBENV
Fig. 7. A free cooling system schematic for a data center.
4.2. Indirect evaporative cooling bulb temperature efficiency equation:

𝑡𝑝,𝑑𝑏,𝑖𝑛 − 𝑡𝑝,𝑑𝑏,𝑜𝑢𝑡
4.2.1. Ordinary indirect evaporative cooling 𝜂𝑤𝑏 = × 100% (1)
𝑡𝑝,𝑑𝑏,𝑖𝑛 − 𝑡𝑝,𝑤𝑏,𝑖𝑛
4.2.1.1. Technical Principle. Air quality in most regions is often incon-
sistent. Therefore, in the construction of data centers in these places, the In the equation: 𝜂𝑤𝑏 —Wet bulb efficiency, %;
solution of introducing fresh outdoor air directly into the data center is 𝑡𝑝,𝑑𝑏,𝑖𝑛 —The dry bulb temperature of the inlet air of the cooler, °C;
very risky and not accepted. in order to reduce the energy consump- 𝑡𝑝,𝑑𝑏,𝑜𝑢𝑡 —The dry bulb temperature of the product air of the cooler,
tion of data centers and to make greater use of natural cooling sources, °C;
indirect evaporative natural cooling technology becomes the primary 𝑡𝑝,𝑤𝑏,𝑖𝑛 —The wet bulb temperature of the inlet air of the cooler, °C.
choice. A schematic diagram of a typical data center air-side indirect (1) Refrigeration capacity
evaporative free cooling system is shown in Fig. 12. This indirect evap-
orative free cooling system has three working modes during the whole The cooling capacity refers to the change in enthalpy of the output air
year, which are dry mode (cooler air/air heat exchange), wet mode as it passes through the dry passages of the indirect evaporative cooling
(wind-side indirect evaporative cooling), and coupled mode (wind-side heat exchanger [27] and is expressed as follows:
indirect evaporative cooling + mechanical cooling).According to the dy- 𝑄 = 𝜌𝑓 𝑉𝑝,𝑜𝑢𝑡 (𝑖𝑝,𝑑𝑏,𝑖𝑛 − 𝑖𝑝,𝑑𝑏,𝑜𝑢𝑡 ) (2)
namic changes of the outdoor environment parameters throughout the
In the equation: 𝑄—Cooling capacity of the cooler, kW;
year, the three working modes are dynamically switched in turn to pro-
𝜌𝑓 —The density of product air of the cooler,𝑘𝑔∕𝑚3 ;
vide a guarantee for the thermal environment of the data center.
𝑉𝑝,𝑜𝑢𝑡 —The volume of product air of the cooler,𝑚3 ;
𝑖𝑝,𝑑𝑏,𝑖𝑛 —The air enthalpy value at the inlet of the cooler,𝑘𝐽 ∕𝑘𝑔;
𝑖𝑝,𝑑𝑏,𝑜𝑢𝑡 —The air enthalpy value at the outlet of the cooler, 𝑘𝐽 ∕𝑘𝑔.
4.2.1.2. Indicative parameters. Since the air is cooled with equal humidity in the dry channel of the
indirect evaporative cooling heat exchanger, the enthalpy change of the
(1) Wet-bulb efficiency air can be expressed in terms of the temperature drop of the air as it
flows through the dry channel. For this reason, the above equation can
The cooling efficiency is an important standard for the performance be written as:
evaluation of evaporative cooling units. The wet bulb efficiency is a pa-
𝑄 = 𝑐𝑓 𝜌𝑓 𝑉𝑝,𝑜𝑢𝑡 (𝑡𝑝,𝑑𝑏,𝑖𝑛 − 𝑡𝑝,𝑑𝑏,𝑜𝑢𝑡 ) (3)
rameter relationship describing the temperature of the cold air and the
imported working air wet bulb temperature in the indirect evaporative In the equation: 𝑐𝑓 —The specific heat capacity of air, 𝑘𝐽 ∕(𝑘𝑔 ⋅ 𝐾)
cooling unit, where the evaporative cooling unit outputs the cold air wet The Eq. (3) is only valid for constant heat capacity.
7
JID: ENBENV
Fig. 8. Schematic diagram of a seawater direct

cooling system in a marine data center.
4.2.1.3. Research progress. Table 4

Comparison of the heat transfer efficiency of an indirect evapo-
(1) Research on the structure of the cooler rative cooler [27,28].
Cooler type Dry conditional heat Wet manual heat

Data centers use different structure types for coolers, shown in transfer efficiency transfer efficiency
Fig. 13.
Crossflow Plate 40%~50% 60%~85%
By summarizing the reference, it is found that the heat transfer ef-
Heat wheel 70%~80% 60%~70%
ficiency of different indirect evaporative coolers has a large difference. Heat pipe 50%~60% 65%~75%
The cooler of the data center air conditioner considers not only the heat Tube NA 40%~80%
exchange efficiency during wet working mode in summer, but also the
heat transfer efficiency between air/air during the dry operating mode
in winter and transition season. As shown in Table 4, the heat transfer
efficiency of the runner type indirect cooler is higher for dry conditions, ing time of the year. Therefore, plate fin and tube heat exchanger forms
but the heat transfer efficiency of the runner type heat exchanger is are more commonly used in data center air conditioning. At the same
lower for wet operation mode. Considering that the indirect evaporative time, the tube indirect evaporative cooler is recognized in data center
cooler of the data center needs to spray water during most of the operat- applications because it can effectively solve the problem of blockage.
8
JID: ENBENV
Fig. 9. Data center air-side direct evaporative cooling system

(a) Direct evaporative cooling filler. (b) Fan wall.
Fig. 10. Ningxia Zhongwei Amazon cloud

computing data center direct air-side free cool-
ing.
Fig. 11. The direct air-side free cooling of

Ningxia Zhongwei 360 cloud computing data
center.
(1) Research on the cooler material and established a test bench to test and analyze its performance, and the
hydrophilicity of porous ceramic effectively improves the heat transfer
The hydrophilic, anti-corrosive and evaporative characteristics of
performance of the indirect evaporative cooler. Huang Xiang [31] pro-
cooler materials are also prominent research areas. Manfredo Guilizzoni
posed to cover the surface of a tubular indirect evaporative cooler with
et al. [29] experimentally analyzed two different indirect evaporative
a layer of hygroscopic material and conducted theoretical and experi-
cooling (IEC) systems, one of which used a plate heat exchanger with a
mental studies on this layer of hygroscopic material to verify its use. Fan
standard epoxy coating (STD) and another that used a new hydrophilic
Lijuan [32] et al. pointed out through comparative analysis that the use
coating (HPHI). The results showed that the contact angle of the HPHI
of hydrophilic aluminum foil tubes, wrapping stainless steel mesh struc-
coating was always lower than the contact angle of the STD coating.
ture material, metal fiber fabric or polypropylene fiber fabric outside the
Sun T [30] et al. proposed a porous ceramic indirect evaporative cooler
9
JID: ENBENV
Fig. 12. Data center air-side indirect evaporative cooling schematic

(a) Dry mode (b) Wet mode (c) Mixed mode.
tube, and coating organic inorganic composite hydrophilic film outside carry 1,000 server cabinets, more than 10,000 servers, the project scale
the heat exchanger tube are several proven measures to enhance heat of the first phase of 300 cabinets, including 4kW/cabinet in the low-
and mass transfer in the heat exchanger tube in tubular indirect evapora- density area, 8kW/cabinet in the high-density area, using a total of 10
tive cooling systems. Xu P [33] From experimental tests on various fiber sets of air-side indirect evaporative cooling air conditioning, 3 months
fabrics, it can be concluded that various fabrics (textiles) are compared to complete All components were installed in 3 months, shortening the
with kraft paper that has been commonly used as a surface medium for construction cycle by 50% and saving 11.3% of the annual electricity
wet channels in indirect evaporative coolers. It can be found that most cost of the data center. It is the first data center in the world to adopt
of the textile fabrics have superior properties under the capillary action indirect evaporative cooling technology with a PUE value below 1.25.
of moisture, which are greater diffusion coefficient and better evapo- From 2017 to the present, Tencent data centers have been mas-
ration capacity. Duan Z [34] et al. compared the capillary core suction sively laid out in first-tier cities around China, such as Huailai in Hebei,
level of porous fiber materials such as kraft paper and non-woven fabrics Yizheng in Jiangsu and Qingyuan in Guangdong, carrying out large-
with time to provide a basis for indirect evaporative coolers. Chu Junjie scale planning and construction of T-block, the fourth-generation data
[35] et al., by testing the core suction height of 6 types of hydrophilic center technology, shown in Fig. 15. As an important technical achieve-
material samples, compared the drip diffusion, moisture absorption and ment of Tencent’s new infrastructure practice, where the PUE is about
moisture and surface structure, as well as the hydrophilic performance 1.2 and can be as low as 1.1.
of the material. It is considerd that the efficiency of indirect evaporative
cooling systems can be greatly excavated by the excellent hydrophilic 4.2.2. Dew point indirect evaporative cooling
materials. Bruno [36] et al. constructed a fork-flow heat exchanger using 4.2.1.1. Technical principle. In 2003, Maisotenko [38] proposed a new
a special medium to improve the hydrophilic properties of the wet chan- form of indirect evaporative cooling heat circulation, which allows any
nel while moisture cannot pass through the dry channel. Tests showed type of gas or liquid to cool below the wet bulb temperature until the
that the dew point efficiency of this heat exchanger was 75%, which is it approaches the dew point temperature. Internationally, this thermal
still relatively low for a given operating condition. Velasco [37] et al. ex- cycle is called M-cycle, and since it is itself a form of indirect evapo-
perimentally studied indirect evaporative cooling heat exchanger based rative cooling technology with more compact structure and higher heat
on polycarbonate. Benefit from low cost and low weight, as well as the transfer efficiency, it is more often called dew point indirect evaporative
absence of corrosion problems, limiting the reduction of plate thickness cooling technology by Chinese scholars. It almost overturned the percep-
caused by low thermal conductivity. The results show that as the in- tion of the traditional evaporative cooling technology of large volume
let air temperature gradually increases from 25°C to 40°C, the cooling and low efficiency, dew point indirect evaporative cooling has the char-
capacity of the cooler is significantly improved. acteristics of being able to provide output air after wet cooling such as
approaching the dew point, small footprint, flexible installation form,
4.2.1.4. Data center application case. The Huawei Big Data center of etc. Therefore, its importance is increasingly attached to many scholars
Linyi is located in Linyi city, Shandong Province, shown in Fig. 14, and is at home and abroad, and is the main direction of evaporative cooling
a big data center for cloud computing that radiates to the economic cir- technology.
cle in the southern Shandong Province and northern Jiangsu Province. Unlike the conventional indirect evaporative cooling process, in dew
The Huawei Linyi Data center project has a total investment of 1.5 bil- point indirect evaporative cooling, small holes are punched in the sheet
lion yuan, covers an area of 95 mu, with a total construction area of on the wet and dry channel side, as shown in Fig. 16(a), or holes are
13,315 square meters, has 6 standardized modular server rooms, can punched at the end of the dry channel to allow some of the primary
10
JID: ENBENV
Fig. 13. Indirect evaporative cooling heat exchanger’s physical map

(a) Plate fin (b) Heat wheel (c) Circular tube (d) Plate tube.
air to be introduced into the wet channel, as shown in Fig. 16(b). Such temperature of the inlet air and reaches a sub-wet bulb temperature.
secondary air is often part of the primary air, secondary air in the wet Part of the cooled primary air as secondary air is introduced into the
channel before entering the heat exchanger core has been pre-cooled, wet channel, along the way with the water film heat mass exchange,
so that the wet bulb temperature of the secondary air after cooling will the state from point 2 → 2wb → 3 Indicative parameters
be further reduced, enlarging the heat transfer temperature difference
(1) Dew point efficiency
between the dry and wet channels, making the primary air cooling in the
dry channel will increase, which can be its own wet bulb temperature Dew point efficiency is defined as the ratio between the tempera-
to sub-wet bulb temperature [39]. ture difference of the cold air inlet and outlet of the indirect evapora-
Dew point indirect evaporative cooling is represented on an tive cooling unit and the temperature difference between the dry bulb
enthalpy-wet diagram, as in Fig. 17, where primary air 1 enters the dry temperature of the unit’s inlet air and its dew point temperature. This
channel of the heat exchanger and is cooled with equal humidity, with reflects the degree of influence of the unit output cold water tempera-
a state from point 1→2 (close to 1dp); it breaks through the wet bulb ture on the temperature of the inlet working air, where the evaporative
11
JID: ENBENV
Fig. 14. Huawei Linyi Data center.
Fig. 15. Tencent T-Block Data center.
Fig. 16. Dew point indirect evaporative cooling schematic

(a) Cross flow (b) Countercurrent flow.
12
JID: ENBENV
Fig. 17. Dew point indirect evaporative cooling process wet.
cooling unit outputs a cold wind dew point efficiency; see Eq. (4) as 5. Data center water-side evaporative cooling
follows:
𝑡𝑝,𝑑𝑏,𝑖𝑛 − 𝑡𝑝,𝑑𝑏,𝑜𝑢𝑡 5.1. Cooling tower cooling system
𝜂𝑑𝑝 = × 100% (4)
𝑡𝑝,𝑑𝑏,𝑖𝑛 − 𝑡𝑝,𝑑𝑝,𝑖𝑛
In an air-side evaporative free cooling system, the produced cold air
In the equation: 𝜂𝑑𝑝 —Dew point efficiency, %; needs to be sent into the cold aisle of the data center through the air
𝑡𝑝,𝑑𝑝,𝑖𝑛 —The dew point temperature of the inlet air of the cooler, °C. supply duct. Therefore, the air supply duct of the system should not be
too long, nor should there be too many elbows and inflection points.
(1) Approximation Thus, air-side evaporative free cooling systems are often suitable for
The unit output cold air or cold water approximation is defined as warehouse-type and flat-layer data centers. The building forms of data
the difference between the unit output cold air dry bulb temperature or centers in the actual process are diverse, and at the same time, since the
output cold water temperature and the dew point temperature of the cooling capacity of wind is much smaller than that of water, the water-
incoming air, where the indirect evaporative cooling unit output cold side evaporative natural cooling system has become a hot research topic
air approximation, as shown in Eq. (5): for data center thermal environment control. The water-side free cooling
system adds a plate heat exchanger between the cooling water and the
𝐴 = 𝑡𝑝,𝑑𝑏,𝑜𝑢𝑡 − 𝑡𝑝,𝑑𝑝,𝑖𝑛 (5) chilled water. When the outdoor temperature is low, the chilled water is
directly cooled by the low-temperature cooling water by the controlling
In the equation: A—Degree of approximation,°C.
part or all of the cooling water to flow to the new heat exchanger. This
reduces the compressor operation and achieves energy savings, shown
4.2.2.2. Dew point heat exchanger. Scholars at home and abroad have in Fig. 20. Traditionally, cooling towers are used to directly supply cool-
conducted a lot of research on the flow channel form, heat transfer core ing capacity to the data center to increase the free cooling time of the
size, operation conditions, and cooling effect of dew point evaporative system.
coolers. Since this paper only focuses on the research and application of Taylor S T et al. [55] noted when the air temperature in the cold
the dew point cooler in the data center, only the relevant experimental aisle of the data center exceeds 27°C and the temperature in the hot
results of the dew point cooler are shown in Table 5. aisle exceeds 38°C, conversely, these air temperatures allow the cooling
water supply temperature to exceed 18°C, which is easily reached by the
4.2.2.3. Data center application case. There have also been attempts by cooling tower during most of the year under most climatic conditions.
scholars on how to apply dew point coolers to data centers. The Na- J Thomas et al. [56] listed two different free cooling methods
tional Renewable Energy Laboratory [52,53] in the Colorado National through water-side cooling technology. One is that a single cooling
Ice Snow data center adopted a direct air-side free cooling system for tower can be used for both a water-side economizer and heat extrac-
indirect evaporation of cooling heat exchangers in conjunction with a tion from the chiller, and the other one has two cooling towers, one of
dew point cooler, as shown in Fig. 18. The results show that system en- which is dedicated as a water-side economizer. These two configura-
ergy consumption is reduced by 75% in the summer, while in winter it tions have their own advantages, depending on the climatic conditions
is possible to reduce it by 90%. that need to be considered.
Yin Bi [54] proposed indirectly evaporating the new refurbishment Y Udagawa et al. [57] noted that a free cooling system is one of
dew point to the data center cooling system, as shown in Fig. 19, and the effective way to reduce the energy consumption of air condition-
found that the energy utilization efficiency (PUE) of the data center is ing systems. In recent years, free cooling systems have attracted signif-
found by using a new dew point cooling system. Between 1.10∼ 1.22, it icant attention in Japan, so their introduction has been investigated,
is much higher than the use of traditional cooling technology. Using a and simulation-based evaluation calculations have been conducted. The
new dew point cooling system and the best supply management system, estimation results show that the water-side cooler needs to double the
an 87.7∼91.6% electricity consumption saving for space cooling in DCs capacity of the cooling tower and consider the installation space and
could be achieved in 10 typical cities at 10 selected climatic conditions. cost.
13
JID: ENBENV
Table 5
Summary of the dew point indirect evaporative cooling experimental results.
Researcher Flow path form Experiment entrance Accuracy Research content Experimental results
conditions Cooling efficiency
Riangvilaikul et al. [40] Countercurrent flow Imported Dry bulb Air supply temperature Effects of various Wet bulb efficiency:
(2010 Thailand) temperature = 25-45°C error 5% imported air parameters 106% -109%
Air humidity = 6.9-26.4 g/kg Wet bulging efficiency and structural Dew point efficiency:
Air flow rate = 1.5-6 m/s error 10% parameters and other 65% -86%
Secondary/primary = 0.33 factors on the thermal
performance
Zhan et al. [41-42] Cross flow Imported Dry bulb Wet bulb efficiency error Comparative analysis of Wet bulb efficiency:
(2011 UK) temperature = 25-40°C 0.2-0.4% the fork flow heat 116%
Entering air relative transfer and
humidity = 50% countercurrent heat
Air flow = 130 m3/h exchanger under the
Countercurrent flow Imported Dry bulb The air supply same geometric Dew point efficiency:
temperature = 25-45°C temperature error is less dimensions and 80% -90%
Entrance air moisture than 3.4% operating conditions,
weight = 6.9-26.4 g/kg and perform
Entrance air flow rate = 2.4 corresponding parameter
m/s methods
Secondary/primary = 0.33
Lee and Lee [43] Countercurrent flow Imported Dry bulb Combine better Analysis of the plus rib Wet bulb efficiency:
(2013 South Korea) temperature = 27.5-32°C coolers placed in an 118% -122%
Imported air moisture artificial climate under Dew point efficiency:
weight = 9.19-18.11 g/kg different operating 75% -90%
Air flow rate = 1.0 m/s conditions
Secondary air flow rate = 0.6
m/s
Secondary/primary = 0.3
Sergey, Demis et al [44] Cross flow Air flow = 95-400 m3/h The deviation of the The convergence cooler Air moisture is 25 g/kg
(2015 Poland) Secondary air flow = 95-400 primary air flow outlet for the fork agent was as an example
m3/h temperature is up to 5%, evaluated, compared Wet bulb efficiency: 90%
Imported air Dry bulb the deviation of the with the results of the -110%
temperature = 20-44°C cooling capacity is as test and numerical Dew point efficiency:
Entrance air moisture high as 7%, and the simulation, further 63% -68%
weight = 9-26 g/kg maximum difference analyzing the method of
between the wet bulb improving the cooling
and dew point efficiency capacity and efficiency
is 6%.
Rogdakis Ed et al [45] Cross flow Imported Dry bulb The air supply Discusses the effects of Wet bulb
(2013 Greece) temperature = 33.1-36.5°C temperature error is less different meteorological efficiency:97%-115%
Import air relative than 1°C, and the conditions and the
humidity = 30.3-33.5% moisture error in the secondary air and
Secondary/primary = 1 secondary air exit is less primary air volume ratio
than 10%. on the cooler’s cooling
performance and water
consumption in Greece
X. Cui et al [46] Countercurrent flow Imported Dry bulb The air supply The air import Wet bulb
(2015 Singapore) temperature = 25-40°C temperature error is less parameters are efficiency:122%-132%
Imported air moisture = 8-20 than 7.3%, the air discussed, indoor air is Dew point
g/kg temperature difference discounted as a back efficiency:81%-93%
Air flow rate = 1.0-2.5 m/s error is less than 7.4% ventilation, geometric
Secondary air flow size, and rib on the
rate = 1.0-2.5 m/s performance of the
cooler performance.
Liu Jiali et al [47] Cross flow Imported Dry bulb results of testing In high humidity regions, Wet bulb
(2014 China) temperature = 36.2-36.7°C the performance efficiency:103%
Imported wet ball parameters, such as the
temperature = 26.5-28.8°C wind pressure, efficiency,
Secondary/primary = 0.67- secondary/one air
0.95 volume, length, and
water consumption are
actually tested.
Song Wei, etc. [48] Cross flow Imported Dry bulb results of testing Experimental test Wet bulb efficiency: 91%
(2014 China) temperature = 36-38.7°C analysis using dew point
imported wet ball evaporative cooling air
temperature = 28.4-29.3°C conditioners in the
imported dew point simulated data machine
temperature = 25.3-27.3°C room
Relative
humidity = 47-61.4%
Jiang Xiaoqiang, etc. Cross flow Imported Dry bulb results of testing A new type of dew point Wet bulb efficiency:
[49] temperature = 38°C, evaporative cooling air 126%
(2015 China) Imported wet ball conditioning unit is
temperature = 23°C, designed, and the
Imported dew point experimental platform is
temperature = 16.1°C designed for the analysis.
(continued on next page)
14
JID: ENBENV
Table 5 (continued)
Researcher Flow path form Experiment entrance Accuracy Research content Experimental results
conditions Cooling efficiency
Wang Yugang et al [50] Cross flow Select 8 outdoor temperature The experimental effect A new type of dew point 150 m3/h
(2015 China) and humidity parameters is better than the indirect evaporative Wet bulb efficiency: 80%
that mainly represent the city simulation effect. cooling air conditioning -99%
system was established, Dew point efficiency:
and the study was 50% -75%
conducted via
experimental simulations
under 8 different
climatic conditions.
Peng Xu et al. [51] Countercurrent flow Select 5 outdoor temperature results of testing Simulations under 5 In the Dry bulb
(2017 UK) and humidity parameters different climatic temperature of 37.8°C,
from the world around the conditions, inside the the wet ball temperature
world laboratory is 21.1°C under
conditions
Wet bulb efficiency:
114%
Dew point efficiency:
75%
COP up to 52.5
Fig. 18. Schematic of the new cooling system

for the NSIDC (winter operation).
Fig. 19. Schematic drawing of the novel dew point cooling system.
15
JID: ENBENV
Fig. 20. Schematic diagram of the data center cooling tower cooling system.
Christy Sujatha et al. [58] compared the energy consumption of a and capital costs can be achieved. Yury Y. Lui et al. [64] proposed that
data center using a traditional cooling system with a data center us- the application of water-side economizers to data center projects re-
ing a water-side economizer in view of the fact that the outdoor air in quires additional consideration to meet unique design standards that are
the three different areas of Chicago, Atlanta, and Phoenix is relatively not common in other commercial building types. Design considerations
cool most of the time. The results show that the performance of the wa- for water-side economizers include appropriate equipment size, cooling
ter economizer is always better than that of the air economizer, which tower antifreeze, pump design in parallel and series configurations, and
significantly improves the efficiency and cost of the cooling system of control strategies for system stability.
the data center. The performance ratios of conventional air condition- Quan Zhang et al. [65] proposed a cooling system in which the data
ers, air-side economizers, and water-side economizers are 50%, 76%, center directly uses lake water and a cooling tower and conducted sim-
and 79%, respectively, indicating that economizers have greater energy- ulation and actual measurement analysis, shown in Fig. 21. The sim-
saving effects than conventional systems. ulation and test results show that when the air supply temperature is
Bruce A. Hellmer et al. [59] analyzed the power and water con- set to 24°C and the water supply temperature is set to 15°C, the energy
sumption of cooling systems, involving the use of dry coolers, water-side consumption of the cooling system reaches a minimum.
economizers, air-side economizers, steam, and ultrasonic and evapora- Zuiliang Ma et al. [66] introduced the principle, system form and
tive humidifiers in telecommunication/data centers across the United foreign application examples of cooling tower cooling, and analyzed
States to perform refrigeration. The results show that the air-side econ- the cooling temperature and system equipment selection, local climatic
omizers is the most energy-efficient mode, although the energy saved in conditions, building load characteristics and other factors that affect the
the tropics is quite small. cooling tower cooling economy, while looking forward to the applica-
Dong Wei et al. [60] proposed a mode conversion strategy based on tion prospects of this technology in China.
a prediction method. When the cold water return temperature is greater Jianli Zhe et al. [67] determined that the free cooling system of the
than the predicted heat exchanger outlet temperature (PHXLWT) plus cooling tower can operate stably within a specific temperature range
a threshold, the economizer can save energy. This transition strategy through tests. The location and form of the data center conditioning
was implemented on a TRNSYS-MATLAB test bench. The results show terminal has great significance to extending the free cooling time of
that, compared with the traditional wet bulb temperature constant tem- the cooling tower. The closer the indoor terminal is to the cabinet, the
perature method, this method has a better energy-saving effect, the free smaller the cooling loss, the smaller the requirement for outdoor meteo-
cooling operation time can be extended to approximately 120%, and the rological parameters, and the longer the free cooling time of the cooling
total energy consumption is reduced by 7%. tower.
Saurabh K. Shrivastava et al. [61] proposed that there are two main Cooling tower and natural cooling source cooling technology in the
types of air containment systems: hot aisle closure and cold aisle closure. data center related engineering cases, but the cooling tower water tem-
They also estimated the energy consumption of the three types of chan- perature is relatively high, the application range is narrow and the use of
nel sealing structures used in three different types of data center cooling mode is simple, only as a supplementary means of mechanical cooling.
systems (dry coolers, cooling towers and water-side economizers). They The use of natural cooling source direct cooling, the actual project will
determined that dry coolers with hot aisle seals were the most energy be subject to the greater constraints and impact of natural conditions.
efficient.
Jinkyun Cho et al. [62]considered that the outdoor environment af- 5.2. Indirect evaporative cooling chiller cooling system
fects the energy consumption of facilities. The terms and definitions of
two different energy-saving modes were also given, and their energy 5.2.1. Technical Principle
performance was compared with key data center attributes in temper- Indirect evaporative cooling chillers use the dry air energy contained
ate or subtropical regions. The terminal power failure and power us- in the unsaturated air as the driving force. Before the air and water are
age efficiency (PUE) values were estimated in typical years. The water- directly contacted for the evaporative cooling process, the air is first
side economizer can achieve a maximum energy saving performance ap- cooled via isohumidity to produce cold water. The limit temperature
proximately 16.6% higher than the reference traditional cooling system, of the cold water is the dew point temperature of the air. The indirect
while the air-side economizer can save approximately 42.4% energy. evaporative cooling chiller is generally composed of two sections: an
Robert Tozer et al [63] have the potential to achieve zero energy indirect evaporative cooling precooling section and a water spray filling
cooling for every type of free cooling system across the United States section, as shown in Fig. 22. The enthalpy diagram of the cooling process
when air is supplied to the data halls within the ASHRAE recommen- is shown in Fig. 23.
dations and allowances. Zero-energy cooling solutions offer significant
capital cost savings due to the reduction of mechanical equipment and 5.2.2. Indicative parameters
associated supporting electrical infrastructure (e.g., switchboards, gen- (1) Sub-wet bulb efficiency
erators, distribution panels, etc.). In many U.S. climates, 100% natural The sub-wet bulb efficiency of the indirect evaporative cooling
cooling (zero cooling) is possible, and significant savings in operating chiller is defined as the difference between the wet bulb temperature
16
JID: ENBENV
Fig. 21. Schematic diagram of the lake water direct cooling system of the lake water data center.
Fig. 22. Schematic diagram of the indirect evapora-

tive cooling chiller
of the unit’s inlet air and the wet bulb temperature after the indirect 𝑡𝑤𝑏,𝑜 : wet bulb temperature of ambient air, °C;
section, as well as the ratio of the difference between the wet bulb tem- 𝑡𝑑𝑝,𝑜 : dew point temperature of ambient air, °C;
perature of the unit’s inlet air and the dew point temperature of the 𝑡𝑤𝑏,𝑐 : wet bulb temperature of precooled ambient air, °C.
unit’s inlet air. Reflecting the effect of the indirect section and the de- (2) Water-side cooling efficiency of the packing section of the chiller
gree to which the unit inlet air wet bulb temperature approaches the The precooled ambient air and the return water of the heat-carrying
unit inlet air dew point temperature; see equation (6): unit are in direct contact with the water spray filler for evaporative
cooling, heat, and humidity exchange. During the change process, the
𝑡𝑤𝑏,𝑜 − 𝑡𝑤𝑏,𝑐
𝜔= (6) limit temperature that the return water of the unit can reach is the wet
𝑡𝑤𝑏,𝑜 − 𝑡𝑑𝑝,𝑜 bulb temperature after the ambient air is precooled. Therefore, after the
water-sprinkling filling section is processed, the reduction degree of the
Among them,
return water temperature of the unit can be described via expression (7),
𝜔: sub-wet bulb efficiency of indirect evaporative precooling unit,
which is called the "water-side cooling efficiency of the water-sprinkling
%;
17
JID: ENBENV
Fig. 23. The indirect evaporative cooling chiller is

shown on the enthalpy diagram.
filling". Generally, the cooling efficiency of the water-side of the spray The cold water produced by the mechanical refrigeration chiller is di-
filler is between 60% and 80%. rectly provided to the air conditioning terminal of the data center, and
𝑡𝐻 − 𝑡𝐺 the indirect evaporative cooling chiller is used to discharge the heat
Γ𝑤 = (7) of the mechanical refrigeration chiller. The temperature of cold water
𝑡𝐻 − 𝑡𝑤𝑏,𝑐
produced via indirect evaporative cooling chillers is lower than that of
Among them, ordinary cooling towers, thereby improving the energy efficiency of the
Γ𝑤 : cooling efficiency of the water drenching filler on the water-side, entire air conditioning system.
%;
𝑡𝐻 : return water temperature of the unit, °C; 5.2.4. Research progress
𝑡𝐺 : supply water temperature of the unit, °C; Data center water-side indirect evaporative free cooling technology
𝑡𝑤𝑏,𝑐 : wet bulb temperature of precooled ambient air, °C. is the latest attempt to apply indirect evaporative cooling chillers in a
data center. Water-side indirect evaporative free cooling chillers have
5.2.3. System type multiple functional sections and diverse usage modes. The cooling rate
Fig. 23 shows a schematic diagram of the indirect evaporative cool- is greater compared with cooling towers. In the reality of increasing heat
ing system on the data center’s water side. The core equipment of the generation in data centers, research on indirect evaporative cooling on
water-side indirect evaporative cooling system is the indirect evapora- the water-side has attracted increasing attention.
tive cooling chiller [68-70]. When the outdoor environment dry bulb Huang Xiang et al. [71] proposed an indirect evaporative cooling
temperature is lower than 3°C, it is the dry operation mode of the air- chiller that uses an indirect evaporative cooler to precool the inlet air.
side indirect evaporative free cooling air conditioning system, as shown It was applied to the data center cooling project in Xinjiang Uygur Au-
in Fig. 24 (a). At this time, the air-conditioning system uses the an- tonomous Region of China, as shown in Fig. 25, and it achieved 100%
tifreeze cooling medium (such as glycol) in the cooling coil to exchange free cooling of the data center air conditioning system throughout the
heat with the cooler outdoor air to remove heat for the data center. year. The system is safe and reliable in operation, energy savings and
When the outdoor environment temperature rises and the dry bulb tem- water savings and has considerable economic benefits.
perature is higher than 3°C, the effective heat removal of the data cen- Feng Xiaoxiao et al. [72] proposed a data room cooling system based
ter cannot be guaranteed by relying only on an air/antifreeze cooling on an indirect evaporative cooling tower. In summer, the indirect evap-
medium heat exchange. At this time, the indirect evaporative cooling orative cooling tower is used as the cooling tower of the mechanical
chiller begins working, and the air conditioning system switches to wet refrigerator, which can prepare cooling water lower than the wet bulb
mode. The operation mode is indirect evaporative cooling to produce temperature, thereby improving the COP of the refrigerator; in winter,
the cold water operation mode. The cold water produced by the indirect the indirect evaporative cooling tower is used as the cold source of the
evaporative cooling chiller removes the heat of the data center through cooling system to independently supply the cooling capacity to the data
the plate heat exchanger, as shown in Fig. 24 (b). When the outdoor room. Because of the unique structure of the indirect evaporative cool-
environment temperature continues to rise and the ambient wet bulb ing tower, it can realize the function of antifreezing in winter.
temperature is higher than 18°C, the indirect evaporative cooling mode Geng Zhichao et al. [73] analyzed and noted that in Xinjiang Urumqi,
is still unable to effectively remove heat, so the mechanical refrigeration Qinghai Xining, Gansu Lanzhou, Ningxia Yinchuan and Shaanxi Xi’an,
chiller is turned on and the indirect evaporative cooling and mechani- the indirect water-side free cooling system of air precooling direct evap-
cal refrigeration coupled operation mode is run, as shown in Fig. 24 (c). orative cooling chillers operates in the full free cooling mode throughout
18
JID: ENBENV
Fig. 24. Schematic diagram of the indirect

evaporative cooling system on the water-side
of the data center
(a) Dry mode
(b) Wet mode
(c) Tandem hybrid mode
(d) Parallel hybrid mode.
19
JID: ENBENV
Fig. 25. Schematic diagram of the indirect evaporative cooling system on the water-side of the new data center
(a) Cold season (b) Normal season (c) High temperature and humidity season.
the year. The hours are 40.8% (2508 h), 20.5% (1467 h), 38.0% (2195 method will be used in the data center cooling system, namely, state
h), 25.3% (1473 h), and 38.1% (1576 h) more hours, respectively than point liquid cooling (SPLC) [75], as shown in Fig. 26 (a). SPLC uses
the indirect water-side free cooling system of the direct-entry evapora- a membrane separation layer to produce cold water instead of cold air
tive cooling chiller. through indirect evaporative cooling technology, as shown in Fig. 26 (c).
Guo Zhicheng et al. [74], aimed at the free cooling air-conditioning In the water flow channel, a small part of the processed return water will
system of a data center in Xinjiang, used the abundant dry air energy and pass through the membrane separation layer and directly contact the
low-temperature air in Xinjiang as free cooling sources, and achieved working air in the air flow channel to evaporate heat absorption in or-
the first application of a cooling air-water air-conditioning system in der to achieve a cooling effect on the processed return water. The mem-
the field of a data center computer room air-conditioning system based brane separation layer prevents cross-contamination between the water
on direct evaporative cooling technology, indirect evaporative cooling and air streams. The SPLC air handling unit, as shown in Fig. 26 (b),
technology and glycol free cooling technology, and achieved 100% free is mainly composed of precooling coils, air-water film heat exchangers,
cooling throughout the year. The actual measurement shows that the direct evaporative cooling units, dry cooling coils, and electrical control
water consumption of the chiller unit and the fresh air unit are 0.34 units. The application form of the cooling system in the data center is
m3 /h and 0.06 m3 /h, respectively. The SCOP of the system is 16.64, and shown in the figure. The server cabinets in the computer room are ar-
the COP is 6.65. The theoretical PUE of this data center is 1.285, which is ranged in a closed hot aisle, and the cold water taken out by the unit
compared with the two commonly used air-conditioning systems in the system is exchanged through the heat exchange coils on both sides of
computer room. This method has a maximum annual energy saving rate the computer room. The return air in the closed hot aisle of the ma-
of 73%, a saving operating costs of 2.663 million yuan, and an annual chine room is cooled by the heat exchange coils on both sides of the
cabinet revenue of 5.15 million yuan. The system is safe and reliable machine room and then sent to the cold aisle using a fan. The return
in operation, energy savings and water savings and has considerable water of the heat exchange coil is returned to the SPLC unit for cooling.
economic benefits. When the SPLC air handling unit is applied in the data center cooling
Facebook is preparing to build Asia’s first data center in Singapore, system there are three operating modes: (1) when the outside ambient
and it is expected to begin operation in 2022. A new type of cooling air temperature is low, the return water of the heat exchange coil in the
20
JID: ENBENV
Fig. 26. Facebook Schematic diagram of the SPLC system in the Singapore data center
(a) SPLC System diagram (b) SPLC Schematic diagram of the unit
(c) Schematic diagram of the internal membrane separation heat exchanger of the unit.
computer room is directly returned to the dry cooling coil in the SPLC contact with the water film on the surface of the coil. For the water film
unit, and the outside cold air directly exchanges heat through the dry and the air exchange heat and mass under the combined action of the
cooling coil to produce cold water; (2) when the temperature of the out- temperature difference and the water vapor partial pressure difference,
side ambient air is high, the return water from the heat exchange coil a direct evaporative cooling process occurs. The surface temperature of
in the computer room returns to the dry cooling coil in the SPLC unit, the water film is reduced and then passes through the heat exchange
the external ambient air needs to be precooled by the direct evaporative plate. The heat conduction effect of the tube cools the medium in the
cooling unit, and then the dry cooling coil is used to exchange heat to tube, condensing the high-temperature and high-pressure refrigerant va-
produce cold water; (3) when the temperature of the external ambient por into a low-temperature refrigerant liquid. The heat of condensation
air is high, the return water of the heat exchange coil in the computer is then discharged to the outside by the air outside the tube through the
room returns to the air-film heat exchanger, and part of the cold water fan.
taken out is transported to the heat exchange coil in the machine room,
and the other part is transported to the precooling coil in the SPLC unit.
6.2. Research progress
The outside air first passes through the precooling coil. After the cold
coil is precooled, it enters the membrane heat exchanger to perform in-
For the application of evaporative condensers to data center systems,
direct evaporative cooling on the back water of the machine room to be
some related scholars have also researched and engineering applications
processed.
in recent years. Zongwei Han et al. [76] proposed an evaporative cooling
composite air-conditioning system with the goal of improving the energy
6. Data center freon-side evaporative cooling efficiency of data centers. On this basis, a mathematical model based on
the operating characteristics of the system was established. Through ex-
6.1. Technical Principle perimental verification, the energy-saving performance of the system
under different climatic conditions was studied. The simulation results
Freon-side evaporative cooling technology is an air conditioning sys- show that when the system PUE value is lower than 1.40, the system
tem that uses evaporative cooling technology to dissipate heat from the has good energy-saving advantages in different regions. Dongqing Wu
condenser of the chiller, and the cold water that is produced is passed [77] used an evaporative condensing air-conditioning system to supply
through the air-conditioning terminal to bear the heat load and wet load cooling capacity to the data room in Xinjiang, China, shown in Fig. 28.
of the air-conditioning area. The evaporative condenser "combines the A 45% ethylene glycol (freezing point temperature is -32°C) solution
cooling tower and the condenser into one" using air and water as the was added to the air conditioning cold water system to prevent the cold
cooling medium, shown in Fig. 27. When working, the water pump and water from freezing in winter, achieving the free cooling mode of the
fan are started. At this time, the water in the water tank is sprayed system in winter. After adopting this technology the PUE was 1.362,
through the nozzle by the suction action of the water pump and sprayed reaching the required target, and the energy saving effect was remark-
on the surface of the heat exchange coil to form a uniform water film. able. Tingting Li [78] introduced an engineering example of the appli-
The air then enters from the air inlet and is forced out after sweeping cation of evaporative condensers in a data center in Yinchuan, Ningxia.
across the surface of the coil due to the action of the fan. The air is in Through actual tests on the air conditioning system, thermal environ-
21
JID: ENBENV
Fig. 27. Schematic diagram of the evaporative con-

denser.
Fig. 28. Schematic diagram of the evaporative cooling system on the freon-side of the data center.
ment, and power consumption of various electrical equipment in the fects, which provides a basis for the future energy-saving transformation
computer room, the test results showed that the computer room’s av- of the data center. Xiaoqing Su [79] studied the applicability of evapora-
erage temperature of the environment is 23.3°C, which can essentially tive condensing chillers in computer room air conditioning applications.
reach the design standard of 23±3°C. The system can also meet the cool- Through a comparison with water-cooled chillers, it was concluded that
ing demand of the computer room and has obvious energy-saving ef- evaporative condensing chillers save approximately 29.5%. It is a ben-
22
JID: ENBENV
Table 6 Table 8
PUE reference values. RD reference values of evaporative cool-
ing air conditioning.
Standard Good Better
2.0 1.4 1.1
6 5 4
Table 7
WUE reference values of evaporative cool-
ing air conditioners in data centers. cold source of the data center air conditioning system. The water con-
sumption indicators of different forms of evaporative cooling equipment
vary greatly. Evaporative cooling equipment should be reasonably con-
2.0 1.0 0.5 figured in order to reduce data, specifically the WUE value of the central
evaporative cooling equipment.
eficial attempt to apply evaporative condensation technology to data 7.3. More land savings
centers. However, it is very risky for the system to be used throughout
the year, especially when the host is running with spray water during Evaporative cooling air conditioners are weather air conditioners,
the transition season (outdoor temperature is lower than 0°C). Winter and the equipment volume is usually relatively large. When it is applied
operation strategies and risks are fully evaluated and prepared in ad- in a data center, the volume of the cooling equipment also becomes
vance. a factor that must be considered because of the large amount of heat
generated in the data center, the large cooling capacity required, and
7. Development of evaporative cooling technology in data center the huge number of evaporative cooling air conditioning equipment.
We use the refrigeration equipment cooling density RD as an evaluation
7.1. More power savings index to measure the size of the cooling capacity per unit volume of the
equipment.
Energy saving is always the eternal theme of the development of data
center air conditioning. PUE (power usage effectiveness) is the earliest 𝐶 𝑜𝑜𝑙𝑖𝑛𝑔 𝐶 𝑎𝑝𝑎𝑐𝑖𝑡𝑦
𝑅𝐷 = (10)
energy efficiency indicator proposed by the Green Grid [80]. PUE is 𝐸𝑞𝑢𝑖𝑝𝑚𝑒𝑛𝑡 𝐴𝑟𝑒𝑎
defined as the ratio of the total power to run the data center facility to RD—refrigeration capacity per unit area of refrigeration equipment,
the total power drawn by all IT equipment: kW/m2 ;
𝑇 𝑜𝑡𝑎𝑙 𝐹 𝑎𝑐𝑖𝑙𝑖𝑡𝑦 𝑃 𝑜𝑤𝑒𝑟 This index reflects the ratio of the refrigeration capacity of the evap-
𝑃 𝑈𝐸 = (8) orative cooling equipment to the area of the equipment and character-
𝐼𝑇 𝐸𝑞𝑢𝑖𝑝𝑚𝑒𝑛𝑡 𝑃 𝑜𝑤𝑒𝑟
izes the refrigeration capacity of the evaporative cooling equipment per
An average data center has a PUE of 2.0; however, several recent
unit area. Under the same conditions of cooling capacity, the greater
superefficient data centers have been known to achieve a PUE as low as
the cooling density is, the smaller the equipment area is. The reference
1.1.The reference values of PUE in the data center are shown in Table 6.
values of RD in the data center are shown in Table 8.
In an air-conditioning system combined with evaporative cooling
Currently, indirect evaporative cooling air-conditioning units are
and mechanical refrigeration, to meet the safe and reliable operation
used in data centers to improve the heat exchange efficiency and re-
of the data center throughout the year the mechanical refrigeration ca-
duce the footprint of the unit, mainly using efficient and compact indi-
pacity is generally designed to be 100% in the preliminary design, and
rect evaporative coolers. From the relatively large structure of the tube
the subsequent unreasonable operation will increase the energy con-
type to the current plate-fin type, it will develop in the direction of mi-
sumption of the system, failing to achieve the purpose of saving elec-
crochannel heat exchangers in the future. However, the more compact
tricity. The operation mode of the data center evaporative cooling air-
the structure of the evaporative cooler is, the higher the air quality re-
conditioning system is then freely switched according to the changes of
quirements are. Therefore, to prevent the microchannel compact heat
outdoor meteorological parameters, so its energy-saving effect depends
exchanger from clogging, the air must be filtered to avoid the risk of
on the efficient control strategy of the system and the later standard-
clogging the heat exchange core, thereby improving the operating effi-
ized operation and maintenance management. This is also a relatively
ciency of the unit.
lacking link at present.
7.2. More water savings 8. Conclusion
The application of evaporative cooling air conditioning in data cen- (1) The application of evaporative cooling air-conditioning sys-
ters leads to water consumption. How to save water is also a research tems in data centers mainly includes air-side evaporative cool-
focus of evaporative cooling air conditioning in data centers. WUE (wa- ing air-conditioning systems, water-side evaporative cooling air-
ter usage effectiveness) is a data center water efficiency evaluation index conditioning systems, and freon-side evaporative cooling air-
proposed by the Green Grid [81]. conditioning systems. Evaporative cooling air-conditioning tech-
nology can reduce the annual PUE of the data center to below
𝐴𝑛𝑛𝑢𝑎𝑙 𝑆𝑖𝑡𝑒 𝑊 𝑎𝑡𝑒𝑟 𝑈 𝑠𝑎𝑔𝑒 1.25, which has great promotion value in data center applica-
𝑊 𝑈𝐸 = (9)
𝐼𝑇 𝐸𝑞𝑢𝑖𝑝𝑚𝑒𝑛𝑡 𝑃 𝑜𝑤𝑒𝑟 tions. The applications of evaporative cooling technology on dif-
WUE is defined as the ratio of the total water consumption of the ferent aspects has matured. For the different forms of evaporative
data center evaporative cooling equipment to the power consumption cooling air-conditioning systems in data centers, their application
of IT equipment throughout the year. The closer the WUE is to 0, the conditions and scenarios should be comprehensively considered
higher the water saving degree of the evaporative cooling equipment in to achieve the maximum use of renewable energy-dry air energy
a data center.The reference values of WUE in the data center are shown in data center air-conditioning systems, thereby reducing the en-
in Table 7. ergy consumption of data center air-conditioning systems.
Facing the enormous water consumption and shortage of water re- (2) The following should be given attention: optimizing the structure
sources in data centers, evaporative cooling equipment is used as the and size of evaporative cooling refrigeration and air-conditioning
23
JID: ENBENV
equipment in data centers and adopting a modular and highly in- [14] K.P. Lee, H.L. Chen, Analysis of energy saving potential of air-side free cooling for
tegrated design concept to realize the prefabrication and produc- data centers in worldwide climate zones, Energy Build 64 (2013) 103–112.
[15] J. Siriwardana, S. Jayasekara, S.K. Halgamuge, Potential of air-side economizers for
tization of data center refrigeration and air-conditioning technol- data center cooling: a case study for key Australian cities, Appl. Energy 104 (2013)
ogy. At the same time, combining free cooling technology with 207–219.
other refrigeration and renewable energy technologies, a better [16] V Depoorter, E Oro, J. Salom, The location as an energy efficiency and renewable en-
ergy supply measure for data centres in Europe, Appl. Energy 140 (2015) 338–349.
combination of energy technology can better cater to the needs [17] E. Oró, V. Depoorter, N. Pflugradt, J. Salom, Overview of direct air free cooling and
of rapid development and construction of data centers. thermal energy storage potential energy savings in data centres, Appl. Therm. Eng.
(3) The equationtion of relevant advanced and applicable evapora- 85 (2015) 100–110.
[18] Clidaras J, Stiver DW. Water-based data center: U.S. Patent 8,853,872[P]. 2014-10-
tive cooling and air-conditioning technical standards for data
7.
centers should be strengthened. It is necessary not only to put [19] 15k515, Evaporative Cooling Ventilation Air Conditioning System Design and In-
forward high-quality requirements for different forms of air- stallation, China Planning Press, Beijing, 2015.
[20] Huang Xiang, Evaporative Cooling Ventilation Air Conditioning System Design
conditioning products, but also to pay more attention to system
Guide, China Construction Industry Press, 2016.
design and operation. The efficient operation of the refrigeration [21] CM Scofield, TS. Weaver, Using wet-bulb economizers: data center cooling, ASHRAE
and air-conditioning system in the data center has been real- J. 50 (8) (2008) 52–58.
ized, and the energy consumption of the refrigeration and air- [22] Heiliger, J., Building Efficient Data centers with the Open Compute Project. 2011.
[23] Park J. Open Compute Project Data center v1.0. The Open Compute Project 2011.
conditioning system in the data center has been effectively re- [24] Prineville, Partners S. Facebook Prineville Data center /Sheehan Partners. Archdaily.
duced. Contributions should be made to China’s 2030 "carbon [25] ChinaDCC. Data center design concept award-Ningxia Zhongwei western cloud base
peak" and 2060 "carbon neutral" long-term plans. data center.2015. http: mp.weixin.qq.com/s/lDKKTChIHhr7u3gzmgMXxg.
[26] Mu Zhenghao, Wang Ying, Air conditioning design of Ningxia Zhongwan calculation
data center, Haicone Air Condition. 46 (10) (2016) 23–26.
Author Statement [27] Z Duan, C Zhan, X Zhang, et al., Indirect evaporative cooling: Past, present and
future potentials, Renew. Sustain. Energy Rev. 16 (9) (2012) 6823–6850.
[28] American Society of HeatingRefrigerating and Air-Conditioning Engineers. ASHRAE
Junjie Chu: Methodology, Software, Writing-Original Draft, Review HANDBOOK HVAC APPLICATIONS, ASHRAE, 2015.
and Editing [29] A Mg, A Sm, B Pl, et al., Effect of plates coating on performance of an indirect
evaporative cooling system, Int. J. Refrigerat. 104 (2019) 367–375.
Xiang Huang: Conceptualization, Funding acquisition, Supervision,
[30] T Sun, X Huang, Y Chen, et al., Experimental investigation of water spraying in an
Review and Editing. indirect evaporative cooler from nozzle type and spray strategy perspectives, Energy
Buildings 214 (2020) 109871.
[31] Huang Xiang, Ji Yao, Di Yonghui, et al., Theory and experimental study on elliptical
Conflict of interest tube cost elliptical tube of water-absorbent materials, J. Refrigerat. J. (02) (2006)
48–54.
The authors declare that they have no known competing financial [32] Fan Lijuan, Experimental Study on Hydrophilic Performance of Tube Indirect Evap-
orative Cooler, Xi’an University of Engineering, 2009.
interests or personal relationships that could have appeared to influence [33] P Xu, X Ma, X Zhao, et al., Experimental investigation on performance of fabrics for
the work reported in this paper. indirect evaporative cooling applications, Build. Environ. 110 (2016) 104–114.
[34] Duan Zhiyin, Zhao Xudong, Zhan Changhong, et al., Energy saving potential of a
counter-flow regenerative evaporative cooler for various climates of China: Experi-
Acknowledgments ment-based evaluation, Energy Build. 148 (aug.) (2017) 199–210.
[35] Chu Junjie, Huang Xiang, Sun Tiezhu, Wu Yi, Huang Kaikin, Study on the hy-
drophilicity of the indirect evaporative cooler moisture channel in the dew point,
This study was supported by the Shenzhen Sustainable Development Chin textile Technol. 46 (01) (2018) 40–44.
Science and Technology Project (2021N033), the National Key Research [36] F. Bruno, On-site experimental testing of a novel dew point evaporative cooler, En-
and Development Program (2016YFC0700404), the National Natural ergy Build. 43 (12) (2011) 3475–3483.
[37] E V Gómez, A T González, F J R Martínez, Experimental characterisation of an in-
Science Foundation of China (51676145).
direct evaporative cooling prototype in two operating modes, Appl. Energy 97 (3)
(2012) 340–346.
References [38] Maisotsenko V, Gillan L E, Heaton T L, et al. Method and plate apparatus for dew
point evaporative cooler: U.S. Patent 6,581,402[P]. 2003-6-24.
[1] GB 50174-2017, Data center Design Specification, China Planning Press, Beijing, [39] Xia Qing, Huang Xiang, Zhou Haidong, et al., Summary of evaporative cooling air
2017. treatment of sub-wet ball temperature, Refrigerat. Air Condition. 13 (5) (2013)
[2] China Information Communications Research InstituteChina Internet Industry De- 25–30.
velopment Situation and Sports Index Report (2019), 2019. [40] B Riangvilaikul, S. Kumar, An experimental study of a novel dew point evaporative
[3] China IDC Circle Industry Research CenterChina Idc Industrial Development Re- cooling system, Energy Build. 42 (5) (2010) 637–644.
search in 2019-2020, 2020 http://www.idcquan.com/special/2020baogao/. [41] C Zhan, X Zhao, S Smith, et al., Numerical study of a M-cycle cross-flow heat ex-
[4] Greenpeace, North China Electric Power UniversityLightning Green Cloud: Research changer for indirect evaporative cooling, Build. Environ. 46 (3) (2011) 657–668.
on China Data center Energy Consumption and Renewable Energy Use Potential, [42] C Zhan, Z Duan, X Zhao, et al., Comparative study of the performance of the M-cycle
2019. counter-flow and cross-flow heat exchangers for indirect evaporative cooling–paving
[5] American Society of Heating, Refrigerating and Air-Conditioning EngineersThermal the path toward sustainable cooling of buildings, Energy, 36 (12) (2011) 6790–6805.
Guidelines for Data Processing Environments, ASHRAE, 2004. [43] J Lee, D Y Lee, Experimental study of a counter flow regenerative evaporative cooler
[6] American Society of Heating, Refrigerating and Air-Conditioning EngineersThermal with finned channels, Int. J. Heat Mass Transfer 65 (2013) 173–179.
Guidelines for Data Processing Environments, ASHRAE, 2008. [44] D Pandelidis, S Anisimov, W M Worek, Performance study of counter-flow indirect
[7] American Society of Heating, Refrigerating and Air-Conditioning EngineersThermal evaporative air coolers, Energy Build. 109 (2015) 53–64.
Guidelines for Data Processing Environments, ASHRAE, 2011. [45] J Lin, K Thu, T D Bui, et al., Study on dew point evaporative cooling system with
[8] American Society of Heating, Refrigerating and Air-Conditioning EngineersThermal counter-flow configuration, Energy Convers. Manage. 109 (2016) 153–165.
Guidelines for Data Processing Environments, ASHRAE, 2015. [46] X Cui, K J Chua, M R Islam, et al., Performance evaluation of an indirect pre-cooling
[9] M Iyengar, R Schmidt, V Kamath, et al., Energy efficient economizer based data cen- evaporative heat exchanger operating in hot and humid climate, Energy Convers.
ters with air cooled servers, in: Thermal and Thermomechanical Phenomena in Elec- Manage. 102 (2015) 140–150.
tronic Systems (ITherm), 2012 13th IEEE Intersociety Conference on, IEEE, 2012, [47] Liu Jiali, Huang Xiang, Sun Zhe, Huang Huoliang. Experimental study on indirect
pp. 367–376. evaporative of new composite dew points, Chiza Machinery (05) (2014) 61–66.
[10] American Society of HeatingRefrigerating and Air-Conditioning Engineers, Inc. Par- [48] Song Wei, Application of Intersection Of Interstitial And Evaporative Cooling Air
ticulate and Gaseous Contamination in Datacom Environments. United States of Conditioning Unit In Communication Machine Room /Base Station, Xi’an University
America, 2009. of Engineering, 2015.
[11] American Society of HeatingRefrigerating and Air-Conditioning Engineers, Inc. Par- [49] Jiang Xiaoqiang, He Huaming, Xia Wei, Chiang Yiyi, Dai Guilong, Performance test
ticulate and Gaseous Contamination in Datacom Environments. United States of and adaptability analysis of a new dew point indirect evaporative air conditioning,
America, 2013. Build. Sci. 35 (06) (2019) 54–57 + 134.
[12] ASHRAEANSI/ASHRAE Standard 127-2007,“Method of Testing for Rating Computer [50] Wang Yugang, Wang Yi, Huang Xiang, et al., Experimental study on the applicability
and Data Processing Room Unitary Air Conditioners”, American Society of Heating, of indirect evaporative cooler in different climate districts in my country, Chiza
Atlanta, 2007 Refrigeration and Air- Conditioning Engineers. Machinery 43 (8) (2015) 68–72.
[13] GPE Hwaiyu, T. Han, in: 16. Particulate and Gaseous Contamination in Data centers [51] P Xu, X Ma, X Zhao, et al., Experimental investigation of a super performance dew
Data center Handbook, John Wiley & Sons, Inc, 2014, pp. 307–312. point air cooler, Appl. Energy 203 (oct.1) (2017) 761–777.
24
JID: ENBENV
[52] BA Weerts, D Gallaher, R Weaver, O. Vangeet, Green data center cooling: achieving [66] The horse is the best, Sun Yuxi. Principle and analysis of cooling tower cooling
90% reduction: airside economization and unique indirect evaporative cooling, IEEE technology, HVAC 28 (6) (1998) 27–30.
Green Technol. Conf. (2012) 1–6. [67] Discovering. Research on Cooling Tower Cooling System in Data center, Xi’an Uni-
[53] DOE. NSIDC data center: energy reduction strategies airside economization and versity of Engineering, 2017.
unique indirect evaporative cooling. U.S. Department of Energy. Federal Energy [68] Y Jiang, X. Xie, Theoretical and testing performance of an innovative indirect evap-
Management Program; 2012. orative chiller, Solar Energy 84 (12) (2010) 2041–2055.
[54] Y Bi, Y Wang, X Ma, et al., Investigation on the energy saving potential of using a [69] X Xie, Y Jiang, An indirect evaporative chiller, Front. Energy 4 (1) (2010) 66–76.
novel dew point cooling system in data centres, Energies 10 (11) (2017) 1732. [70] X Xie, Y Jiang, Comparison of two kinds of indirect evaporative cooling system: to
[55] ST Taylor, How to design and control waterside economizers, ASHRAE J. 56 (6) produce cold water and to produce cooling air, Proc. Eng. 121 (2015) 881–890.
(2014) 30–36. [71] Huang X, Guo Z, Tian Z, et al. Application of new evaporative cooling air-
[56] J Thomas, J Bruschi, R Anliker, et al., Data center Best Practices Guide Energy Effi- conditioning system in a data center in Xinjiang. 2020.
ciency Solutions for High-Performance Data centers, Pacific Gas and Electric Com- [72] Research and Application of Indirect Evaporative Cooling Technology in Data center,
pany, Sacramento, CA, USA, 2012. Tsinghua University, 2017.
[57] Y Udagawa, S Waragai, M Yanagi, et al., Study on free cooling systems for data [73] Qi Zhichao, Research ON Indirect Evaporative of Free cooling Air Conditioning Sys-
centers in Japan, in: Intelec 2010, IEEE, 2010, pp. 1–5. tems in Dry Areas Data center, Xi’an university of engineering, 2018.
[58] S. Abimannan, Energy efficient free cooling system for data centers, in: 2011 IEEE [74] Guo Zhicheng, Research on Free cooling Air Conditioning System for a Data center
Third International Conference on Cloud Computing Technology and Science, IEEE, in Xinjiang, Xi’an University of Engineering, 2019.
2011, pp. 646–651. [75] StatePoint Liquid Cooling system: A new, more efficient way to cool a data
[59] B A Hellmer, Consumption analysis of telco and data center cooling and humidifi- center. https: engineering.fb.com/2018/06/05/data-center-engineering/statepoint-
cation options, ASHRAE Trans. 116 (2010) 118. liquid-cooling.
[60] Dong W E I, Baiyu L I, Yancai S H I, et al. Energy-efficient mode transition strategies [76] Z Han, X Sun, H Wei, et al., Energy saving analysis of evaporative cooling composite
of data center cooling systems. 2016. air conditioning system for data centers, Appl. Therm. Eng. 186 (5) (2020) 116506.
[61] S K Shrivastava, A R Calder, M. Ibrahim, Quantitative comparison of air contain- [77] Wu Dongqing, Wu Xueyuan, Application of indirect evaporative condensed technol-
ment systems, in: 13th InterSociety Conference on Thermal and Thermomechanical ogy in Northern Data center, Haicone Air Condition. 49 (08) (2019) 72–76.
Phenomena in Electronic Systems, IEEE, 2012, pp. 68–77. [78] Li Tingting, Huang Xiang, Luo Ye, Folding, Liu Kai Lei, Yan Zhi Chao, Yang Liran,
[62] J Cho, T Lim, B S Kim, Viability of datacenter cooling systems for energy efficiency The evaporative condenser test analysis of a data center in Yinchuan, Refrig. Air
in temperate or subtropical regions: case study, Energy Build. 55 (2012) 189–197. Condition. (Sichuan) 32 (04) (2018) 380–386.
[63] R Tozer, S. Flucker, Zero Refrigeration for Data Centres in the USA, ASHRAE Trans. [79] Su Xiaoqing, Huang Xiang, Wang Jun, Application analysis of evaporative conden-
118 (2) (2012). sate concentration concentrates in data center, J. Xi’an Univ. Eng. 30 (01) (2016)
[64] Y Y Lui, Waterside and airside economizers design considerations for data center 37–42.
facilities, ASHRAE Trans. 116 (1) (2010). [80] Grid TG. Green grid metrics: describing datacenter power efficiency. Online at http:
[65] L Ling, Q Zhang, Y Yu, et al., Energy saving analysis of the cooling plant using lake www.Thegreengrid.org/gg_content/Green_Grid_Metrics_WP.pdf, 2007.
water source base on the optimized control strategy with set points change, Appl. [81] D Azevedo, S C Belady, J. Pouchet, Water usage effectiveness (WUETM ): a green grid
Therm. Eng. 130 (2017) 1440–1449. datacenter sustainability metric, Green Grid. (2011).
25

Ref 03 Stat Energy China

Uploaded by

Copyright:

Available Formats

Ref 03 Stat Energy China

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ref 03 Stat Energy China

Uploaded by

Copyright:

Available Formats

JID: ENBENV

ARTICLE IN PRESS [m5GeSdc;September 2, 2021;19:22]

Energy and Built Environment xxx (xxxx) xxx

Contents lists available at ScienceDirect

Energy and Built Environment

Research status and development trends of evaporative cooling

speaking, the refrigeration and air-conditioning system can introduce a

Data center environmental parameter requirements have a signif-

Fig. 1. 2014-2019 China IDC business market

Fig. 2. 2019 - 2022 China IDC business market

2500 27.20% 28.00%

1500 25.30% 3200.5 26.00%

Fig. 3. Proportion of China’s data center energy consumption in 2019.

Table 1 added value of environmental parameters in the 2015 updated version.

Fig. 4. Typical data center precision air conditioning system

Grade Dry bulb Humidity range (%) Maximum dew point

A1 15~32 Dew point temperature-12°C Relative humidity 8%~ 17

Fig. 6. Data center Free cooling Technology Classiﬁcation

Fig. 7. A free cooling system schematic for a data center.

4.2. Indirect evaporative cooling bulb temperature eﬃciency equation:

Fig. 8. Schematic diagram of a seawater direct

4.2.1.3. Research progress. Table 4

Cooler type Dry conditional heat Wet manual heat

Fig. 9. Data center air-side direct evaporative cooling system

Fig. 10. Ningxia Zhongwei Amazon cloud

Fig. 11. The direct air-side free cooling of

Fig. 12. Data center air-side indirect evaporative cooling schematic

Fig. 13. Indirect evaporative cooling heat exchanger’s physical map

Fig. 14. Huawei Linyi Data center.

Fig. 15. Tencent T-Block Data center.

Fig. 16. Dew point indirect evaporative cooling schematic

Fig. 17. Dew point indirect evaporative cooling process wet.

Fig. 18. Schematic of the new cooling system

Fig. 22. Schematic diagram of the indirect evapora-

Fig. 23. The indirect evaporative cooling chiller is

Fig. 24. Schematic diagram of the indirect

Fig. 27. Schematic diagram of the evaporative con-

7.2. More water savings 8. Conclusion

You might also like