Modern buildings provide comfortable, clean and safe shelters for human beings to live and work in at the cost of energy consumption. Building energy consumption, which will further increase with improved indoor thermal comfort levels, accounts for nearly 30% of total social energy consumption in China and more than 40% in Europe and OECD countries ^[1][2][3][1,2,3]. Such huge energy consumption caused by buildings makes carbon emissions the primary cause of global warming, climate change and air pollution [4]. Thus, improving the energy efficiency of the building energy consumption system has become the focus of global attention. As the key building energy consumption system, heating, ventilation and air-conditioning (HVAC) systems, sometimes called air-conditioning systems (ACS), are responsible for 40–60% of the total building energy consumption ^[5][6][5,6]. How to reduce this part of energy consumption has become a key issue in promoting energy conservation and emission reduction in the building sector.

ACS is usually classified into two categories, i.e., split air conditioner (SAC) and centralized ACS (CACS) [7]. The SAC, or so-called split ACS [8], plays a major role in air-conditioning requirements of domestic and small office buildings in summer [9]. However, the performance of the SAC drops significantly under the operation condition of high outdoor ambient temperature, and the indoor air quality cannot be guaranteed [10]. Moreover, the condensate water from the SAC is usually discharged on the spot, although the daily production of the condensate water reaches 52.99 kg during the main air-conditioning season for an SAC, resulting in the waste of the cold energy stored in the condensate water [11]. For the CACS, it is commonly serviced for large-scale public buildings [12]. The indoor air quality of the CACS is improved significantly as fresh air is induced into the room, compared with the SAC. However, processes of reheating in the air handling unit (AHU) and the indoor air exhaust bring corresponding additional energy consumption and energy waste for the CACS, such as the air–air system ^[13][14][13,14]. In fact, there is an abundance of waste heat and low-grade energies with the potential to be utilized in ACS in nature. Numerous researchers have optimized the performance of the ACS in terms of the selection of natural cold sources, the application of heat recovery systems and the utilization of high efficiency terminals.

Natural cold source cooling technologies, such as direct ground-coupled cooling ^[5][15][5,15], natural ventilation ^[16][17][16,17] and novel radiative sky cooling ^[18][19][18,19], can cut down the running time of electric-driven chillers, thereby achieving the goal of reducing energy consumption of the ACS. Yuan et al. [5] performed thermodynamic and economic analysis for a ground-source heat pump system coupled with borehole free cooling. The results showed that free cooling with borehole water matched the indoor air parameter with the design indoor conditions and also improved the average annual cooling efficiency of the ground source heat pump system to 16.22. The average annual energy consumption and operation costs could be decreased to 55.9% and 56.0%, respectively. Gratia et al. [20] chose a narrow plan building suited for natural ventilation to study the use of night ventilation to cool office buildings in the moderate climate of Belgium. They found that single-sided ventilation was as efficient as cross night ventilation, which reduced the cooling load by approximately 40%. As a natural cooling technology, nocturnal sky cooling follows the principle of objects on the Earth’s surface at night giving off heat to the colder sky and becoming colder ^[21][22][21,22]. Cooling air or water produced by nocturnal sky radiators can be used to counteract the building heat gain and maintain building thermal comfort in summer. Aili et al. [18] developed a kW-scale, 24-h continuously operational radiative sky cooling system and experimentally investigated the corresponding heat rejection performance at different flow rates. The results showed that after a 14-h nighttime operation, the storage tank temperature decreased from initial 27 to final 7 °C, with the temperature reduction of 20 °C. This demonstrated that the radiative sky cooling radiator had a cooling capacity of 120 ± 10 W/m² during the nighttime in a typical summer month in Phoenix. Previous studies have confirmed that natural cooling sources can reduce energy consumption in conventional refrigeration systems. However, natural cold sources still need to be transported to the indoor terminals by fans or water pumps corresponding with the energy consumption.

A dew-point air supply system has the advantages of a simple air handling process and energy saving. However, the humidity regulation of the room is poor, as the air stream leaving the cooling coil (CC) is usually too high in relative humidity and too low in temperature, easily causing uncomfortable feelings of the people living in it [23]. Moreover, the risk of illness will increase if people stay in a room with high relative humidity for a long time [24]. In order to create a better, simple air supply control condition and a thermal comfort environment, the cold supply air is always reheated before entering the room. However, this process will consume additional energy.

To maintain indoor air quality at a high level, the ACS always possesses ventilation functions. However, if the outdoor air is taken into the room and indoor air is exhausted to the outside directly, it will increase the demands of the cooling and heating loads for buildings, resulting in the increase in the energy consumption of the ACS. In addition, there is a large amount of low-grade energy in the air exhausted from indoor spaces. This part of energy has great potential to reduce the energy consumption of ACS through heat recovery devices. To reduce the energy consumption of ACS, energy recovery facilities, such as rotary wheel and fixed plate HE, are commonly used in the heat recovery system [25].

As terminals in the ACS transfer heat to the indoor space and affect the indoor comfort and energy efficiency of the ACS directly, high-efficiency terminals are necessary for the ACS. The conventional terminal, such as the fan coil, has a fast heat exchange performance. However, it cannot avoid mechanical energy consumption. Therefore, more and more scholars are aiming at the research of space radiant heating and cooling terminals, which can improve the performance of the ACS without lowering the desired level of comfort conditions of buildings ^[26][27][28][26,27,28]. However, the thermal response speed of radiant terminals is relatively slow due to the high thermal inertia and low heat transfer coefficient compared to that of a convective terminal [29].

As a result of the above problems, most of the waste heat and natural low-grade energies are not utilized efficiently or cannot be utilized at all. Therefore, more and more researchers are focusing on the advantages of a heat pipe (HP) for waste heat recovery and natural low-grade energy applications. The HP, which usually consists of the evaporator, adiabatic and condenser sections, is a passive thermal transfer device based on the internal gas–liquid phase change principle without consuming external mechanical power [30]. The heat transfer coefficient of the HP is hundreds of times that of the copper pipe and the thermal response is very fast [31]. This means that a large quantity of heat can be transported through a small cross-sectional area over a considerable distance. The working principle of an HP is that when the evaporator section is heated, the liquid working medium inside absorbs the heat and evaporates. Then, the vapor travels along the HP to the condenser section and condenses back into liquid, accompanied by the latent heat release. Finally, the working fluid returns to the evaporator section by gravity or capillary force. With the continuous evaporation and condensation processes of the working medium in the HP, heat is continuously transferred from the evaporator to the condenser section. There are four types of HPs commonly used in HVAC systems, such as tubular HP (THP) [32], loop HP (LHP) [33], pulsating HP (PHP) [31] and flat HP (FHP) [34].

In past years, HPs with different forms have been widely applied in engineering applications, such as heat recovery systems ^{[35][36][37][38]}[35,36,37,38], electronics cooling equipment ^[39][40][41][39,40,41], solar water heater systems ^{[42][43][44][45]}[42,43,44,45], photovoltaic cooling systems ^[46][47][48][46,47,48], thermoelectric cooling ^[49][50][51][49,50,51], data center cooling systems ^[52][53][54][52,53,54] and ACSs ^[55][56][57][55,56,57]. The ACS integrated with HP mainly concerns the hot spots of the natural cooling system and performance improvements with respect to the natural cooling system, SAC, CACS and cooling terminal devices.

Two types of THPs are usually used to improve the performance of ACS. One is the traditional HP containing a wick structure, named wicked HP. The driving power of this HP is the capillary force generated by the wick. The other is gravity (gravity-assisted) HP (GHP) [32], or so-called wickless HP and two-phase closed thermosyphon (TPCT) [58]. This type of HP has a smooth inner wall and is driven by gravity force [59]. It should be noted that the installation position of the evaporator section of the GHP is lower than that of the condenser section.

LHP [33], or so-called separate HP (SHP) ^[60][61], and two-phase thermosyphon loop HP (TPTLHP) ^[61][62] is a high-performance passive two-phase heat-transfer device, of which the evaporator and condenser sections are connected by the riser (or the vapor line) and downcomer (or the liquid line) end-to-end to form a loop. There are two types of LHP. One of the LHP has a compensation chamber in the evaporator section. However, the other type has no compensation chamber. The working fluid in the LHP is driven by capillary, gravity or join force coming from the capillary and gravity ^[33][62][63][33,63,64]. The LHP performs better in long-distance heat transfer, design and manufacturing flexibility compared with the traditional THP ^[63][64].

PHP, or so-called oscillating HP (OHP), invented by Akachi ^[64][65] in the 1990s, is made by bending a long capillary tube without the wicking structure into multiple turns. When PHP is running, the working fluid in the capillary tube has two flow states. One of the flow states is random oscillation. This will affect the heat transfer of the PHP, especially when the oscillation pause occurs ^[65][66]. Thus, some scholars have improved the heat transfer by installing the check valve in the PHP to control the liquid slug flow in one direction.

Flat micro-HP (FMHP) array is a relatively new HP technology relying on gravity and capillary force to drive the circulating flow of the working fluid, which was invented by Zhao et al. [34] in recent years. Sometimes, it is abbreviated as the FHP ^[66][68]. An FMHP array is made of several micro-HPs [45]. Multiple parallel micro HPs solve the problem of low heat transfer capacity caused by micro scale in traditional HPs. Moreover, there are many small ribs (or micro-grooves) in the micro channel to enhance the heat transfer performance of the FHP.

The classification of the four types of HPs mentioned above is mainly based on their structures and the driving modes of the working fluid. Each HP has advantages and disadvantages. THP has the advantages of compact structure and low cost, but its efficiency is greatly affected by the installation angle, which limits its application in practical engineering. The structure of the LHP is relatively flexible, and the LHP solves the problem of the short heat transfer distance of the conventional THP. However, the working fluid in the LHP is subjected to multiple forces, and the heat transfer mechanism is relatively complex, resulting in challenges for designers. PHP has the advantage of transmitting both latent and sensible heat simultaneously. It is less affected by gravity force and has a flexible heating mode. However, PHP has the problem of oscillation stagnation, which will seriously affect its heat transfer performance. FHP has the advantage of a flat surface and easy adhesion to the surface of cooled objects. Its heat transfer is relatively large compared with that of conventional THP. However, the disadvantage of the FHP is the same as that of the GHP. Moreover, FHP is more difficult to manufacture.