Energy Efficiency Optimization of Air Conditioning Systems Towards Low-Carbon Cleanrooms: Review and Future Perspectives

Abstract

The advancement of high-tech industries, notably in semiconductor manufacturing, pharmaceuticals, and precision instrumentation, has imposed stringent requirements on cleanroom environments, where strict control of airborne particulates, microbial presence, temperature, and humidity is essential. However, these controlled environments incur significant energy consumption, with air conditioning systems accounting for 40–60% of total usage due to high air circulation rates, intensive treatment demands, and system resistance. In light of global carbon reduction goals and escalating energy costs, improving the energy efficiency of cleanroom heating, ventilation, and air conditioning (HVAC) systems has become a critical research priority. Recent efforts have focused on optimizing airflow distribution, integrating heat recovery technologies, and adopting low-resistance filtration to reduce energy demand while maintaining stringent environmental standards. Concurrently, artificial intelligence (AI) methods, such as machine learning, deep learning, and adaptive control, are being employed to enable intelligent, energy-efficient system operations. This review systematically examines current energy-saving technologies and strategies in cleanroom HVAC systems, assesses their real-world performance, and highlights emerging trends. The objective is to provide a scientific basis for the green design, operation, and retrofit of cleanrooms, thereby supporting the industry’s transition toward low-carbon, sustainable development.

1. Introduction

Diven by the accelerated advancement of industries such as semiconductor manufacturing, pharmaceuticals, and precision instruments, cleanrooms, as key infrastructure for ensuring product quality and cleanliness of the production environment, have been continuously expanding the application scope [1]. Cleanrooms are designed to tightly control indoor levels of particles, microorganisms, temperature, and humidity to protect product quality [2,3]. According to recent statistics, the cleanroom market size in the United States and Europe reached USD 6.54 billion in 2023 and is projected to grow to USD 7.93 billion by 2030 [4]. However, the stringent environmental control requirements of cleanrooms result in exceptionally high energy consumption [5]. Studies have shown that the unit energy consumption of cleanrooms is typically 30 to 50 times that of conventional commercial buildings, making them one of the major sources of energy use in the industrial field. Its global annual energy consumption is up to 110 billion kWh, which seriously restricts the green and sustainable development of related industries [6,7,8].

The air conditioning system constitutes one of the primary sources of energy consumption within cleanrooms, typically accounting for 40–60% of the total cleanroom energy use [9]. This high energy consumption phenomenon mainly stems from the need for high air change rates, complex air treatment processes, and continuous year-round operation to ensure environmental stability and cleanliness [10,11]. In addition, the multi-stage configuration of air filtration systems leads to increased airflow resistance, further elevating fan energy consumption. As illustrated in Figure 1, the energy consumption of cleanroom air conditioning systems even surpasses that of production processes [12]. This phenomenon underscores the critical role of air conditioning systems in the overall cleanroom energy management strategy, highlighting the urgent need for effective energy-saving operation strategies and control algorithms.

In recent years, driven by the intensifying challenges of global climate change and rising energy prices, the development of energy-saving technologies for cleanroom air conditioning systems have become the prominent focus. Current research on energy efficiency in cleanroom air conditioning systems is concentrated in several areas, including air delivery improvement [13,14,15,16], air handling process optimization [15,16,17,18], air filter innovation [19,20,21,22], and intelligent algorithm integration [6,23,24]. Specifically, in the field of air delivery optimization, effective adjustment of parameters such as air flow rate has been proven to significantly reduce energy waste caused by excessive ventilation. Research on the air handling process focuses on heat recovery and temperature–humidity decoupled strategies, aiming to minimize the energy loss caused by the cold–heat offset. In the area of filtration, emphasis has been placed on developing low-resistance filter materials and adopting electrostatic assisted technologies to further reduce system pressure losses. Also, the rise of AI has discovered new pathways for intelligent, dynamic control of cleanroom air conditioning systems. The application of machine learning, deep reinforcement learning, and other intelligent algorithms enables real-time optimization of system operating parameters, thereby further enhancing overall energy efficiency.

However, despite the extensive research on energy-saving technologies for cleanroom air conditioning systems, most existing studies focus on the optimization of specific components or single control strategies, lacking systematic and holistic considerations. In practical operation, cleanroom air conditioning systems involve the collaborative work of multiple modules, and the energy-saving potential relies heavily on the coupled optimization of the overall system. Meanwhile, the issues of long-term operational stability, reliability, and economic viability of these technologies also urgently need in-depth research. Furthermore, the variations in cleanroom applications and regional climatic conditions require energy-saving technologies to possess strong adaptability and flexibility.

Therefore, this article addresses the current technical bottlenecks and practical challenges in the energy-saving field of cleanroom air conditioning systems. It systematically reviews the existing research achievements and the applicability of various energy-saving technologies, thoroughly analyzes the respective advantages and limitations, and critically examines the practical application outcomes. Furthermore, the future research directions and potential technological breakthroughs for cleanroom air conditioning energy optimization are proposed. The goal of this review is to provide a scientific basis and practical reference for the energy efficient design and operational optimization of cleanroom air conditioning systems, thereby promoting the development of the cleanroom industry towards a greener, low-carbon, and more sustainable future.

2. Cleanroom Air Conditioning Systems

2.1. System Structures

Cleanroom air conditioning systems, as a specialized form of air conditioning equipment, are mainly applied in spaces that demand stringent environmental cleanliness, namely cleanrooms. These systems are designed to control the concentration of particles, harmful gases, and microbial contaminants, while simultaneously maintaining indoor temperature and humidity within specified ranges to meet the requirements of production processes and environmental standards. A typical cleanroom air conditioning system comprises several key components, including the make-up air unit (MAU), fan filter unit (FFU), dry cooling coil (DCC), and air handling unit (AHU).

In industrial cleanrooms, air conditioning systems are commonly composed of the MAU, FFU, and DCC (as shown in Figure 2). The MAU first pretreats the outdoor fresh air through filtration, cooling, heating, dehumidification (or humidification) processes to meet the cleanroom requirement. Then, the processed air is conveyed through air ducts to the cleanroom interlayer, where it mixes with return air that has been conditioned by the DCC in terms of temperature and humidity. Subsequently, the mixed air passes through the high efficiency particulate air filter (HEPA) of the FFU, thereby ensuring that the indoor air quality satisfies the cleanliness requirements [17].

In pharmaceutical cleanrooms, the MAU + AHU system is commonly adopted (Figure 3). For this configuration, the MAU performs necessary preliminary treatments for fresh outdoor air such as filtration and cooling. Then, the treated fresh air is mixed with the return air and directed into the AHU, where it undergoes further processing, including temperature and humidity control as well as additional filtration, to meet the required conditions. Subsequently, the conditioned air is pressurized by the fan, conveyed through the air duct, and passed through the HEPA filter before being supplied to the cleanroom.

2.2. Cleanroom Standards

Cleanroom standards are essential for air conditioning system selection and energy efficiency improvement. They establish strict limits on the maximum allowable number of suspended particles per unit volume of air to define different levels of cleanliness and provide essential guidelines for the design and operation of cleanroom air conditioning systems. The classification of air cleanliness levels in various national standards is shown in Table 1.

The first edition of ISO 14644-1 was published in 1999 and was rapidly adopted or referenced by many countries in formulating their own cleanroom standards [26]. The GB 50073-2013 (China) [27] and JIS B 9920: 2019 (Japan) [28] standards both drew heavily from ISO 14644-1 in defining cleanroom classifications. It has become the core international standard for cleanrooms, providing a crucial reference for the unification of related regulations across different countries. Among the ISO Class 1–9 levels, the core production areas of clean manufacturing facilities are predominantly classified between ISO Class 5 and ISO Class 7 [29].

Table 1. Classification of cleanroom cleanliness levels in different national standards.

Standard	Limitations	Country
	d = 0.5 μm (Unit: pcs/m3)	China		USA	Japan	Britain	German	Australian	ISO
		GBJ 73-1984 [30]	GB 50073-2013 [27]	FS 209E: 1992 [31]	JIS B 9920: 2019 [28]	BS 5295: 1989 [32]	VDI 2083: 2003 [33]	AS 1386: 1989 [34]	ISO 14644-1: 2015 [26]
Class	-		1		1				ISO 1
	4(0)		2		2		0		ISO 2
	10			M1
	35		3	M1.5	3	C	1	0.035	ISO 3
	100			M2
	352		4	M2.5	4	D	2	0.35	ISO 4
	1000			M3
	3520	100	5	M3.5	5	E, F	3	3.5	ISO 5
	10,000			M4
	35,200	1000	6	M4.5	6	G, H	4	35	ISO 6
	100,000			M5
	352,000	10,000	7	M5.5	7	R	5	350	ISO 7
	1,000,000			M6
	3,520,000	100,000	8	M6.5	8	K	6	3500	ISO 8
	10,000,000			M7
	35,200,000		9		9	L	7		ISO 9

Note: This table shows the cleanroom grades classified in various standards for particles with a diameter of 0.5 μm under different concentration limits.

Table 1. Classification of cleanroom cleanliness levels in different national standards.

Standard	Limitations	Country
	d = 0.5 μm (Unit: pcs/m3)	China		USA	Japan	Britain	German	Australian	ISO
		GBJ 73-1984 [30]	GB 50073-2013 [27]	FS 209E: 1992 [31]	JIS B 9920: 2019 [28]	BS 5295: 1989 [32]	VDI 2083: 2003 [33]	AS 1386: 1989 [34]	ISO 14644-1: 2015 [26]
Class	-		1		1				ISO 1
	4(0)		2		2		0		ISO 2
	10			M1
	35		3	M1.5	3	C	1	0.035	ISO 3
	100			M2
	352		4	M2.5	4	D	2	0.35	ISO 4
	1000			M3
	3520	100	5	M3.5	5	E, F	3	3.5	ISO 5
	10,000			M4
	35,200	1000	6	M4.5	6	G, H	4	35	ISO 6
	100,000			M5
	352,000	10,000	7	M5.5	7	R	5	350	ISO 7
	1,000,000			M6
	3,520,000	100,000	8	M6.5	8	K	6	3500	ISO 8
	10,000,000			M7
	35,200,000		9		9	L	7		ISO 9

Note: This table shows the cleanroom grades classified in various standards for particles with a diameter of 0.5 μm under different concentration limits.

For pharmaceutical cleanrooms, in addition to stringent limits on the maximum allowable concentration of suspended particles, specific requirements were also established for the concentration of airborne microorganisms. As shown in

Table 2. Classification of cleanliness levels for pharmaceutical cleanrooms [35].

Class	Maximum Allowable Particle Concentration (pcs/m3)				Maximum Allowable Microorganism Concentration
	At Rest		In Operation		Airborne Viable Particles cfu/m3	Sedimental Viable Particles (Ф90 mm) cfu/4 h	Surface Microorganisms
	≥0.5 µm	≥5 µm	≥0.5 µm	≥5 µm			Contact (Ф55 mm) cfu/Dish	5 Finger Gloves cfu/Glove
A	3520	20	3520	20	<1	<1	<1	<1
B	3520	29	352,000	2900	10	5	5	5
C	352,000	2900	3,520,000	29,000	100	50	25
D	3,520,000	29,000	-	-	200	100	50

, the Standard for design of pharmaceutical industry clean room (GB 50457-2019) classifies cleanroom environments into four air cleanliness grades (A, B, C, and D) and specifies clear requirements for both particle concentration and microbial concentration for each grade [35].

2.3. System Characteristics

Compared with conventional air conditioning systems, cleanroom air conditioning systems impose much stricter cleanliness requirements. The main characteristics include the following. (1) Large air supply volume: To satisfy the requirements for high air change rates and strict air cleanliness levels, while maintaining positive pressure of the room (to prevent the introduction of external contaminants due to pressure differentials), the system must process a substantial volume of supply air. [2] (2) High cooling and heating loads: Due to the significant difference between outdoor air conditions and the required clean air parameters, considerable energy is consumed for heating and cooling during fresh air treatment [36]. (3) Significant system resistance: The configuration of multi-stage filters (e.g., primary filter (PF), medium efficient particulate filter (MF), and HEPA filter) in cleanroom air conditioning systems generate significant airflow resistance, particularly at the HEPA stage [17]. Additionally, the resistance of each filter continues to increase over time due to particle accumulation on the filter surface during operation [37]. (4) High fan pressure requirement: To overcome the pressure drop caused by multi-stage filtration and meet the requirements of large air supply volume, fans must provide high static pressure and power output to maintain positive pressure state within the cleanroom [14]. (5) Continuous operation: To maintain stable indoor air quality, temperature, humidity, and pressure, cleanroom air conditioning systems are typically required to operate continuously (often 24 h/d) under high-performance conditions to meet the production or experimental needs [38].

In response to these energy-intensive characteristics of cleanroom air conditioning systems, researchers have proposed a series of optimization strategies aimed at enhancing operational efficiency. The characteristics of large air supply volumes and high fan pressures directly related to the need for optimization in the air delivery process. The high cooling and heating loads are mainly caused by strict temperature and humidity control. Therefore, the optimizations of air treatment processes are essential for reducing the energy consumption of systems. High system resistance is primarily caused by pressure losses across multi-stage filters. To address this cause, the improvement of the air filtration system occupies a core position. For the demand of continuous operation, optimizing control strategies and applying intelligent algorithms offer potential breakthroughs to improve overall system performance. Accordingly, the following sections categorize optimization strategies into four aspects: system air delivery, air handling processes, air filters, and intelligent algorithms. Each section reviews the research in relevant aspects and summarizes the achievements in design optimization and energy reduction.

3. Optimization Strategies Based on System Air Delivery

Optimization strategies of system air delivery play a crucial role in enhancing the energy efficiency of cleanroom air conditioning systems. The relevant improvements are mainly divided into two aspects: the optimization of air supply parameters and the optimization of fan design parameters. The former aims to reduce system energy consumption by adjusting parameters such as air velocity and air flow rate, while ensuring that cleanliness requirements are maintained. The latter focuses on reducing fan resistance and improving fan efficiency to maximize the operational efficiency of the system. Through the optimization of these two aspects, the overall performance of cleanroom air conditioning systems can be enhanced, achieving the goals of energy conservation and high-efficiency operation.

3.1. Air Supply Parameter Optimization

3.1.1. Design Value Reduction

In the design of most cleanrooms, the air change rate (ACR) often significantly exceeds standard requirements. While such a design effectively enhances air cleanliness, it also results in excessive airflow redundancy, leading to a substantial increase in fan energy consumption [14]. Therefore, setting an appropriate ACR based on the actual requirements of cleanrooms can reduce airflow volume while maintaining cleanliness standards, thereby decreasing fan power consumption and achieving the goal of energy conservation.

Numerous studies have demonstrated that reduced design value can significantly lower energy use while maintaining required cleanliness levels. As summarized in

Table 3. Research on optimization of air supply parameters.

Type 1	Origin ACR/ Air Flow Rate	Strategy	Class 2	Power Consumption Reduction	Ref.
A	-	30% reduction in ACR	ISO Class 5	66%	[13]
A	-	39% reduction in fan speed	ISO Class 7	54% (from 151 kWh per week to 70 kWh)	[38]
A	15 h−1	Reduces to 12 h−1	-	20%	[39]
B	20 h−1	Reduces to 6 h−1	GMP C	97.3%	[40]
A	150,000 m3/h	5% reduction in flow velocity	-	3.9% (133 MWh)	[15]
A	150,000 m3/h	10% reduction in flow velocity	-	7.8% (266 MWh)	[15]
A	1,260,000 m3/h	Reduces to 1,200,000 m3/h	ISO Class 5 and 6	0.34% (729,568 kWh, annual energy saving)	[16]
A	32,000 m3/h	39.4% reduction in return air rate	ISO Class 5	22.6%	[14]
A	713,000 m3/h	58.3% reduction in return air rate	ISO Class 5	35.1%	[14]
B	21.3 h−1	31% reduction in air flow rate	ISO Class 8	31.2% (fan power)	[25]

¹ A: industrial cleanrooms; B: pharmaceutical cleanrooms. ² ISO: ISO 14644-1 [26]; GMP: Good Manufacturing Practice of Medical Products. It is the basic guideline for the production and quality management of medicines [41].

, various strategies, such as lowering ACR, reducing fan speed, and optimizing return air rate, have led to energy savings ranging from modest percentages to over 90%, depending on the cleanroom type and reduction extent. Tschudi et al. reduced the ACR by 30% without compromising ISO Class 5 cleanliness, resulting in a 66% reduction of energy consumption [13]. Hu et al. further explored the influence of the degree of flow velocity reduction (5% and 10%) on energy consumption [15]. However, reducing the air volume will also pose certain risks. For instance, in special circumstances where the dispersion of particulates intensifies, the reduced air supply volume may fail to effectively cover the critical areas.

3.1.2. Actual Demand Consideration

Cleanrooms are typically operated at constant ACR and work continuously for 24 h/d without interruption, leading to substantial energy waste. To address this issue, researchers have proposed methods for optimizing the ventilation mode based on the actual application requirements of cleanrooms. The relevant research is shown in

Table 4. Relevant research on control strategy of air supply parameters.

Type 1	Strategy	Class 2	Power Consumption Reduction	Ref.
A	Night and weekend setback	ISO Class 7	28% (from 151 kWh per week to 109 kWh)	[38]
A	DCF (particle counter)	ISO Class 7	40% (from 151 kWh per week to less than 91 kWh)	[38]
A	DCF (particle counter)	ISO Class 5	60–80%	[43]
A	Closes the supplementary air device for non-working hours	-	70–75%	[48]
A	DCF (occupancy sensor)	ISO Class 7	36% (from 151 kWh per week to less than 96 kWh)	[38]
A	DCF (occupancy sensor)	ISO Class 7	37–40%	[7]
B	DCF (occupancy sensor)	GMP C	68.6%	[40]
B	DCF (occupancy sensor)	GMP C	Over 70%	[46]
B	Multi-zone airflow network model	ISO Class 5 and 6	24.5%	[49]
B	Pressure gradient control	ISO Class 7 and 8	39.8% (for energy consumption in non-working mode)	[47]
A	Pressure gradient control	-	18.58%	[50]
-	Air balancing strategy	-	14.3% for fan frequency	[51]
-	Air balancing method with energy-saving constraint strategy	-	37.1%	[52]
-	Steady-state prediction model	-	20.9%	[53]

¹ A: industrial cleanrooms; B: pharmaceutical cleanrooms. ² ISO: ISO 14644-1 [26]; GMP: Good Manufacturing Practice of Medical Products. It is the basic guideline for the production and quality management of medicines [41].

. The demand control filtration (DCF) strategy dynamically adjusts the fan speed according to the changes in particulate contaminant concentration [38]. The studies conducted by Sun et al. [42] and Faulkner et al. [38] have all confirmed the energy-saving effectiveness of this strategy (achieving a reduction of 40–80%). However, such a DCF strategy typically relies on particle counters to monitor particulate concentrations, resulting in relatively long payback periods for cleanroom investments (generally ranging from one to four years) [42,43]. Since humans are the primary source of particulate matter in cleanrooms, occupancy can be used as an alternative indicator to particulate concentration for control purpose [44,45]. In comparison, the DCF strategy that adjusts cleanroom fan speed based on occupancy offers more considerable payback period (typically less than six months) and requires lower maintenance frequency [7]. The experimental results from Loomans et al. demonstrated that implementing the minimum fan speed during unoccupied conditions can reduce energy consumption by 68.6% [40]. On this basis, further investigation on DCF strategies that incorporates pressure level reduction was conducted. This approach achieved over 70% fan energy savings while maintaining the required cleanliness standards [46]. Beyond DCF strategies, energy-saving can also be achieved by optimizing cleanroom pressure control strategies. Liu et al. examined the differences in pressure demands between the operational and non-operational modes and proposed a pressure gradient control strategy that enables dynamic switching between these two conditions (Figure 4) [47]. This strategy can stabilize cleanroom pressure within a range of ±3 Pa and adjust the fan frequency during non-operational modes, resulting in 39.8% reduction in energy consumption [47]. However, the above control strategies have certain limitations. Some strategies rely on predefined operating modes and lack the flexibility to respond to actual operating condition changes, while others, although dynamically adjusting environmental parameters based on sensor feedback, still exhibit response delays under sudden load changes or complex disturbances. In the future, these strategies could be integrated with intelligent algorithms to enable real-time system condition recognition, thereby enhancing response speed and control accuracy.

3.1.3. Air Distribution Improvement

The CFD simulation results of Villafruela et al. showed that the layout of air outlets and inlets has a certain impact on the ventilation efficiency of cleanrooms [54]. Hence, optimizing the airflow organization of cleanrooms also contributes to reducing ACR and decreasing system energy consumption [55]. Based on an evaluation of performance differences among various airflow organizations, Shi proposed that the “up-supply side-return” configuration offers advantages in terms of energy efficiency [55]. Also, according to Saidi et al., when the direction of airflow aligns with the movement path of contamination sources, ventilation efficiency can be improved, offering potential for energy savings [56]. Traditional cleanrooms typically adopt wall–return air recirculation systems. The extended circulation path and high airflow resistance increase the energy demand on the fan to maintain airflow, ultimately contributing to higher energy losses in the systems. In response, researchers have proposed a fan dry coil unit (FDCU) return system [57,58,59]. Compared to the conventional system, it consumes less electrical power, achieving a 4.3% reduction in energy consumption per unit flow rate (Figure 5) [59]. Nevertheless, most existing studies were conducted under static operating conditions, lacking in adequate consideration of air disturbances’ impact under dynamic production environments.

3.1.4. Comparison of the Advantages and Limitations

In terms of design value optimization, reducing airflow volume by minimizing redundant ACR represents a low-cost and easily deployable strategy, particularly suitable for cleanrooms with relatively stable contamination sources. However, its energy-saving potential is constrained by the initial design conditions. When the original design is already close to the standard threshold, the range for further optimization becomes limited. Strategies for actual demand consideration enable the dynamic adjustment of fan speeds, offering significant energy-saving potential. However, these strategies typically rely on sensors and automated control systems, resulting in higher initial investment costs. Moreover, some strategies depend on predefined operating modes, exhibiting limited adaptability to complex and variable operational conditions. The optimization of air distribution fundamentally adjusts the supply and return air paths, making it more appropriate for newly built or renovated cleanroom air conditioning systems. It requires high compatibility with building structures and existing air duct layouts, which limits the applicability. The summarized advantages and limitations are shown in

Table 5. The advantages and limitations of different optimization strategies for air supply parameters.

Method	Advantages	Limitations	Refs.
Design value reduction	Low cost Easy to deploy	Pressure difference reduction between adjacent cleanrooms Risks in special situations Limited energy-saving potential	[38,39]
Actual demand consideration	Great energy conservation potential Flexible Effectively reduces redundancy	High initial investment cost Long investment payback period High maintenance frequency Fixed pattern Response lag possibility	[42,43]
Air distribution improvement	Effectively reduces the ACR	Limited applicability	[54,55]

.

3.2. Fan Design Parameter Optimization

As a key component for directing airflow, the fan is mainly composed of the volute, impeller, inlet, outlet, etc. [60]. Studies have shown that the structural optimization of fans can significantly improve internal flow field distribution and enhance airflow delivery efficiency. Consequently, fan design optimization plays a critical role in improving the energy performance of equipment such as the FFU, MAU, and AHU [61]. Despite the limited research on the application of fan structural optimization in cleanroom HVAC systems, existing studies have indicated that improvements in fan efficiency and airflow parameters can effectively reduce the energy consumption of air supply systems [62]. Accordingly, fan optimization holds substantial potential for enhancing energy efficiency in cleanroom air conditioning systems. The relevant studies are summarized in

Table 6. Studies on optimization of fan design parameters.

Optimization Parameter	Method	Total Pressure Improvement	Efficiency Improvement	Ref.
Impeller (single-impeller structural parameter)
Blade shape	Kriging model and Genetic Algorithm-Particle Swarm Optimization (GA-PSO) algorithm	34 Pa	3.7%	[61]
Blade shape (for single-arc blades)	Class shape transformation function (CST) parameterization, numerical simulation using CFD	- (only shown as a figure)	3.1%	[66]
Blade shape (for multi-arc blades)	Hicks–Henne function parameterization, numerical simulation using CFD	- (only shown as a figure)	4.21%	[62]
Blade inlet angle	Numerical simulation using CFD	54 Pa	2.7%	[64]
Blade outlet angle	Numerical simulation using CFD	- (only shown as a figure)	Increased to 84.85%	[67]
Blade number	Numerical simulation using CFD	497.9 Pa	8.08%	[63]
Impeller (muti-impeller structural parameter)
Blade shape and position	Extreme Learning Machine (ELM) model and PSO algorithm	10.4 Pa	-	[68]
Blade number, and inlet and outlet blade angle	Response Surface Methodology (RSM), Numerical simulation using CFD	-	1.63%	[65]
Blade number, outlet blade angle, and impeller outlet width	Orthogonal design and BP neural network	Increased by 6.91%	0.5%	[69]
Inlet blade angle and impeller diameter	Numerical simulation using CFD	179.6 Pa	-	[70]
Volute
Volute tongue parameter	Partial coherence analysis, numerical simulation using CFD	27 Pa	3%	[71]
Volute geometric parameter	Self-adaption kriging surrogate model	-	1%	[72]
Volute geometric parameter	Numerical simulation using CFD	5.3%	4.6%	[73]
Efficient Global Optimization
Parameters for impellers, blades, and volutes	Numerical simulation using CFD	-	8.46%	[74]
Impeller and volute	Numerical simulation using CFD and physical testing	-	4.33%	[75]

. Varun Ch et al. studied the effect of blade number on the performance of centrifugal fans and achieved an optimum efficiency improvement of 8.08% when setting the blade number to 14 [63]. Blade inlet and outlet angles are also important parameters for the impeller. Zhao et al. set the blade inlet angle in the range from 24° to 36° and found that an inlet angle of 29° minimized turbulent kinetic energy, resulting in optimal aerodynamic performance and 2.7% improvement in efficiency [64]. Meng et al. adopted the response surface methodology to investigate the comprehensive effect of key parameters [65]. The result indicated that, under the optimal combination of blade numbers, inlet blade angle, and outlet blade angle, the efficiency can be increased to 93.7% [65]. In future cleanroom air conditioning system designs, the integration of fan parameter optimization could be further explored to enhance its energy-saving potential at the system level (taking into account the adaptability to the FFU, MAU, and AHU), thereby achieving greater overall energy efficiency.

4. Optimization Strategies Based on Air Handling Process

In cleanroom air conditioning systems, the treatment of air temperature and humidity (including cooling, heating, humidification, and dehumidification) is primarily conducted by the MAU. However, due to the need for multiple heat and moisture exchange, the overall energy consumption of the MAU is relatively high. Cold–heat offset is the key factor contributing to the energy waste. After cooling and dehumidification, the air must be reheated to meet the required temperature and humidity settings, resulting in the simultaneous consumption of cooling and heating energy, thereby causing unnecessary energy losses [76]. Research by Yin et al. indicated that cold–heat offset accounts for approximately 13.1–18.0% of the total energy consumption in the system [77]. Therefore, reducing cold–heat offset has become an important research focus for improving energy efficiency [78]. Various optimization strategies have been proposed, including demand reduction, heat recovery adoption, and temperature–humidity decoupled strategies, aiming to reduce system energy consumption and enhance energy utilization efficiency.

4.1. Demand Reduction

Demand reduction improves system energy consumption by lowering cooling and heating load requirements, thus minimizing the extent of air temperature and humidity regulation required by HVAC equipment. Two primary methods have been proposed to achieve demand reduction: (1) lowering the MAU outlet air temperature and (2) eliminating the reheating process. Lowering the MAU outlet air temperature reduces the loads on heating coils and dry cooling coils, thereby alleviating cold–heat offset to a certain extent [79]. Eliminating reheating, typically by deactivating heating coils during summer, reduces system resistance and decreases both cooling and heating energy consumption, thus contributing to overall energy savings [36]. A summary of related research findings is provided in

Table 7. Relevant research on demand reduction.

Method	Type 1	Specific Description	Power Consumption Reduction	Ref.
Reduce MAU outlet temperature	A	From 16.5 °C to 14.5 °C	0.22% of fab (annual)	[16]
	A	From 14 °C to 11 °C	8.63% of MAU (annual)	[79]
	A	From 16.7 °C to 12.2 °C	1,113,995 kWh (annual)	[80]
	A	From 19 °C to 14 °C	2.09% of fab (annual)	[15]
Cancel reheating	A	No heating required in summer	160 W/m2 heating source	[36]
	A	No heating required in summer	10.4–38.3%	[77]
	A	No heating required in summer	151–236 W/m2	[78]
	A	No heating required in summer and transitional seasons	10.7–17.2% of fan power	[17]
Other	A	Increases temperature by 1 °C (cleanroom)	0.1% of fab (annual)	[16]
	A	Increases humidity by 3% (cleanroom)	0.65% of fab (annual)	[16]

¹ A: industrial cleanrooms.

.

Extensive research has been conducted on reducing the outlet air temperature of the MAU. Chang et al. achieved annual electricity savings of 480,258 kWh by reducing the setpoint temperature by 2 °C (from 16.5 °C), with comparable outcomes observed in the study by Hu et al. [16,79]. On this basis, Hu et al. further investigated the relationship between energy consumption and outlet air temperature. By gradually reducing the temperature from 19 °C to 14 °C (in 1 °C increments), a linear correlation has been found. Each 1 °C reduction in outlet temperature corresponds to a 5.82% decrease in energy consumption of a high-temperature water chiller system [15]. Nevertheless, this study also emphasized that the temperature reduction range should be carefully controlled to prevent phenomenon such as filter frosting [15].

Cancelling reheating has also been studied in depth. Yin et al. proposed a novel air conditioning system in which the reheating process of the MAU is eliminated during summer operation. In this system, only a portion of the return air is cooled by the DCC, while the remaining potion directly enters the return air chute and mixes with fresh air that has only been subjected to cooling and dehumidification, thereby avoiding the need for reheating (Figure 6) [36]. It achieves a savings of 160 W/m² in MAU reheating energy consumption and reduces DCC cooling energy consumption by 40–52% [36,78]. In addition, an experiment conducted by Ma et al. indicated that simplifying the air handling process of the MAU can lower system pressure drops, hence saving fan energy [17].

Beyond the two approaches mentioned above, researchers have also proposed regulating cleanroom temperature and humidity setpoints to reduce the air handling loads of cleanroom air conditioning systems. By increasing the cleanroom temperature and humidity by 1 °C and 3%, respectively, Chang et al. achieved energy savings of 0.1% and 0.65% for a panel fabrication facility [16].

4.2. Heat Recovery Adoption

Heat recovery strategies recycle excess sensible heat or latent heat within the system (to precondition fresh air, etc.), thereby effectively reducing the cooling and heating requirements during the air handling process, improving overall energy efficiency [81]. Heat recovery strategies can be generally classified into four categories: DCC heat recovery, exhaust gas heat recovery, return air heat recovery, and MAU heat recovery. The relevant studies are summarized in

Table 8. Relevant research on heat recovery.

Heat Recovery Method	Type 1	Power Consumption Reduction	Ref.
From DCC to MAU	A	33.7% of MAU and DCC (Transition season) 64.3% of MAU and DCC (Winter)	[36]
From DCC to MAU	A	From 310.1 W/m2–1963.9 W/m2 to 305.9 W/m2–1180.3 W/m2 (includes all weather conditions)	[76]
From DCC to MAU	A	54.3% of MAU and DCC	[82]
From exhaust gas	B	7.1%, 13.5% 16.6%, 40.2% (four climate zones)	[83]
From exhaust gas	A	11.4%	[7]
From exhaust gas	A	12% of MAU pre-heating	[84]
From exhaust gas	-	21–51%	[85]
From return air (variable-frequency control)	A	3.58% of HVAC	[18]
From return air	A	40–52% of DCC for cooling	[78]
From MAU (Chillers)	A	17.26%, 20.77%, 17.93%, 20.50% (four cases)	[86]
From MAU (Run-around system)	A	7.62% of MAU	[87]
From MAU (Run-around system)	A	22.1%, 28.1% (Two cases)	[82]

¹ A: industrial cleanrooms; B: pharmaceutical cleanrooms.

. As the proportion of cold–heat offset between the heating of the MAU and the cooling of the DCC reaches 41.7%, implementing heat recovery from the DCC to the MAU offers significant energy-saving potential [36]. Tsao et al. explored the energy efficiency of various heat recovery systems and found that the DCC heat recovery system could simultaneously save both the precooling energy for the cooling coil and the reheating energy for the MAU, achieving a superior overall performance with an energy-saving rate of up to 28.1% [82]. The operating mechanism of this system is to utilize the outlet water from the DCC to preheat the fresh air within the MAU, while simultaneously cooling the DCC return water through heat exchange with the outdoor air. Yin et al. adopted a similar heat recovery design and achieved reductions of 33.7% and 64.3% in the MAU and DCC during the transitional season and winter, respectively (Figure 7a) [36]. The waste heat contained in exhaust gas (air discharged to the outdoor environment after circulating within the cleanroom) also presents potential for heat recovery. Kircher et al. reduced energy losses during the air preconditioning process by implementing heat exchange between the exhaust and intake air, resulting in an 11.4% reduction in annual energy consumption (Figure 7b) [7]. Liu et al. installed heat pipes within the air handling unit to recover waste heat from exhaust air and compared the resulting energy efficiency improvements across four different climate zones. The study revealed that the hot summer warm winter climate region exhibited the most significant energy savings, reaching up to 40.2% [83]. In addition, recovering heat from return air can also alleviate the cooling and heating demands associated with outdoor fresh air. Yu et al. connected a bypass air duct to a return air duct, allowing a portion of the untreated return air to directly mix with outdoor air and remaining conditioned return air before being supplied to the cleanroom. This design enhanced the utilization efficiency of return air heat, reduced the reheating power consumption of the heater, and achieved a 3.58% reduction in energy consumption (Figure 7c) [18]. Moreover, the run-around cooling coil system applied in the MAU can enable the reuse of thermal energy by absorbing the heat from the cooling coil side and transferring it to the reheating coil (Figure 7d) [82].

4.3. Temperature–Humidity Decoupled Strategies

Traditional cleanroom air conditioning systems normally perform cooling and dehumidification simultaneously. The fresh air is first cooled to dew point to achieve dehumidification and then reheated to ensure the cleanroom requirements for supply air temperature are met. This alternating cooling and heating processes lead to significant cold–heat offsets, resulting in substantial energy waste. Therefore, temperature–humidity decoupled strategies offer considerable potential for improving the energy efficiency of air conditioning systems. These strategies separate the control processes of temperature and humidity, independently treating sensible and latent loads to avoid the cold–heat offset. Currently, decoupled strategies can be classified into three types: the partially decoupled strategy (PD), the fully decoupled strategy (FD), and the adaptive full-range decoupled ventilation strategy (ADV). Relevant studies are summarized in Table 9.

PD utilizes dry outdoor air (or outdoor air that is first cooled and dried by a pre-cooling air handling unit (PAU) if the humidity is high) for effective dehumidification of the cleanroom air (Figure 8a). This method enables dehumidification and temperature control without the need for additional cooling or reheating [88]. Unlike PD, FD incorporates two separate cooling coils within the air handling unit, which are responsible for temperature adjustment and humidity regulation, respectively (Figure 8b) [8]. However, both PD and FD operate based on a single control strategy, their energy-saving potentials are limited by environmental conditions. ADV, which proposed based on the PD, FD, and interactive control (IC), leverages the strengths of each approach. Equipped with an adaptive economizer, it can switch among three modes: following sensible load, following latent load, and lower-limit humidity control. This ensures the precise selection of the most economic mode under varying climatic conditions and internal load scenarios, thereby minimizing energy consumption [89].

Recent studies have demonstrated the effectiveness of various temperature–humidity decoupled strategies in improving cleanroom energy performance. Shan and Wang applied the PD approach to a pharmaceutical cleanroom, achieving substantial reductions in cooling and heating energy consumption (69.6% and 87.8%) [88]. Zhao et al. implemented the FD strategy in a high-tech cleanroom. Under different climatic conditions, the retrofitted cleanroom HVAC system achieved energy savings ranging from 39.2% to 79.6% [8]. Further advancement was made by Zhuang et al., who proposed the ADV strategy and employed it in an ISO Class 8 cleanroom. Compared to PD and FD, ADV delivered additional annual energy savings of 21.64% and 7.77%, respectively [89]. Beyond temperature–humidity decoupled strategies, replacing traditional steam humidification with adiabatic humidification in conventional air conditioning systems can also reduce energy waste. Zhao et al. applied a pressurized water atomizer in the MAU, reducing the annual energy consumption of the air conditioning system by 8% [90].

Table 9. Relevant research on temperature and humidity control strategies.

Method	Type 1	Class 2	Power Consumption Reduction	Ref.
PD	B	ISO Class 7/8	69.6% of cooling (48.4 kW) and 87.8% of heating (50.9 kW)	[88]
PD	A	ISO Class 4/5	39.2–79.6% (Different climatic conditions)	[8]
FD	B	-	15.04% (Annual energy consumptions compared to PD)	[89]
ADV	B	-	21.64% (Annual energy consumptions compared to PD)	[89]
Adiabatic humidification	A	ISO Class 5/6	Saved 6311 kW	[91]
Adiabatic humidification	A	-	8% and 23% (Two cases)	[90]

¹ A: industrial cleanrooms; B: pharmaceutical cleanrooms. ² ISO: ISO 14644-1. [26].

Table 9. Relevant research on temperature and humidity control strategies.

Method	Type 1	Class 2	Power Consumption Reduction	Ref.
PD	B	ISO Class 7/8	69.6% of cooling (48.4 kW) and 87.8% of heating (50.9 kW)	[88]
PD	A	ISO Class 4/5	39.2–79.6% (Different climatic conditions)	[8]
FD	B	-	15.04% (Annual energy consumptions compared to PD)	[89]
ADV	B	-	21.64% (Annual energy consumptions compared to PD)	[89]
Adiabatic humidification	A	ISO Class 5/6	Saved 6311 kW	[91]
Adiabatic humidification	A	-	8% and 23% (Two cases)	[90]

¹ A: industrial cleanrooms; B: pharmaceutical cleanrooms. ² ISO: ISO 14644-1. [26].

4.4. Comparison of the Advantages and Limitations

The three approaches for reducing cold–heat offsets all have advantages and limitations (

Table 10. The advantages and limitations of different optimization strategies for air handling process.

Category	Method	Advantages	Limitations	Note	Refs.
Demand reduction	Reduce MAU outlet temperature	Low cost Easy to deploy	Cold–heat offsets still exist Risks in special situations Limited energy-saving potential	-	[16,79]
	Cancel reheating	Effectively avoids cold–heat offsets Significantly improves energy efficiency	Available for limited seasons	Applicable in summer	[77,78]
Adopt heat recovery		Great energy-saving potential Available for all seasons Various forms	High initial investment cost Increases complexity of system Increases maintenance cost	Heat recovery from returning air is applicable in summer	[36,78,87]
Decoupled strategy	PD and FD	Simple operational modes Lower investment cost Shorter investment payback period	Available for limited situations Large space for supply and return air ducts	Better in hot/mild regions than cold regions Better in buildings with low internal latent loads	[8,88,89]
	ADV	Available for various situations Great energy-saving potential	Complex control logic Higher investment cost	-	[89]

). For demand reduction, reducing the MAU outlet temperature is a relatively simple method to apply. However, it cannot fully mitigate the cold–heat offset, and the setting of the optimal supply air temperature setpoint remains uncertain and environmentally dependent. In contrast, canceling reheating can effectively avoid the additional energy consumption associated with cold–heat offsets and significantly improve system efficiency under suitable operating conditions. Nevertheless, this approach is mainly applicable during summer in most regions. Heat recovery adoption, which enables the recycling of excess sensible or latent heat within the system, can reduce the load demand from outdoor air. Integrating heat recovery with reheating cancellation offers an effective solution for optimizing cooling and heating loads and enhancing overall system energy efficiency [76]. But, implementing heat recovery systems will increase the complexity of the system design and equipment layout. For temperature–humidity decoupled strategies, PD and FD are more suitable for buildings with low internal latent loads, whereas the ADV strategy is capable of optimizing energy utilization under various internal loads and outdoor climatic conditions. Nonetheless, ADV requires more sophisticated control logic, thereby increasing the complexity of system design and operation [89].

5. Optimization Strategies Based on Air Filters

As essential components of cleanroom air conditioning systems, air filters are crucial for maintaining required cleanliness levels [92,93]. However, a high filtration efficiency often results in increased airflow resistance, leading to significant energy consumption. The initial pressure drops of the PF, MF, and HEPA in the MAU are approximately 150 Pa, 200 Pa, and 300 Pa, respectively, and will gradually increase over time during operation [2]. Sun et al. reported that when the particulate load accumulated on the filter media reaches 6 g/m², the pressure drop increases by approximately 400 Pa, leading to a significant rise in energy consumption [94]. Therefore, reducing filtration resistance and enhancing dust-holding capacity have emerged as key strategies for energy-saving optimization. Although various optimization methods for filters have been proposed in conventional air conditioning systems, research specific to cleanroom air conditioning systems remains limited. Lin et al. replaced traditional fiberglass filter media with polytetrafluoroethylene (PTFE), effectively reducing airflow resistance and enabling the FFU to operate at lower air flow rates, thereby achieving an energy-saving rate up to 40.66% [19]. This demonstrates the energy-saving potential of low-resistance materials in cleanroom air filtration. Hence, this section is not limited to the technologies already applied in cleanrooms, but expands to a broader field, reviewing the optimization of filter media in terms of structure, materials, and electrostatics, aiming to guide the energy-saving optimization of cleanroom filters in the future.

5.1. Material Optimization

Material optimization serves as a fundamental approach to enhancing filtration efficiency while reducing air resistance. High-performance filter materials not only effectively capture particulate matter but also maintain low energy consumption over long operation periods. Current research on filter materials is primarily classified into three categories: synthetic polymer fibers, natural polymer fibers, and porous materials (Table 11).

Synthetic polymer fibers include PTFE, polyvinylidene fluoride (PVDF), PAN, etc. [95,96]. PTFE filter media can maintain comparable filtration efficiency to fiberglass while exhibiting an initial pressure drop of less than 60% of the latter [97,98,99]. However, the resistance growth rate of PTFE tends to be higher than that of fiberglass HEPA media [95]. Therefore, scholars have further developed the properties of synthetic polymer fibers by adopting advanced fabrication techniques, e.g., the fabrication of composite nanofiber membranes via electrospinning [100,101,102], the incorporation of functional components [103,104,105], etc. The filters utilizing composite nanofiber materials can achieve filtration efficiencies exceeding 99.5% for ~0.3 μm particles, while maintaining low pressure drops between 60 and 160 Pa. Similarly, materials embedded with functional components can also deliver high filtration performance. In addition, due to the presence of metal oxides and other substances, these materials exhibit antimicrobial or microorganism-inactivating capabilities. Natural polymers and porous materials have also been regarded as promising filtration materials. Through functional modification and composite processing, both materials have shown notable advancements in achieving high filtration efficiency with low pressure drop. Nevertheless, their applicability in cleanroom air conditioning systems remains unclear and requires further investigation

Table 11. Studies on filter materials.

Type	Specific Materials	Air Flow Rate	Resistance	Filtration Efficiency	Particle Size (μm)	Ref.
Synthetic polymer fibers	Nylon 6 nanofibers	5 cm/s	150 Pa	99.993%	0.3	[106]
	PTFE + Poly (vinyl pyrrolidone) (PVP)	-	89.9 Pa	99.72%	-	[100]
	Polyvinyl Chloride (PVC) + Polyurethane (PU)	5.3 cm/s	144 Pa	99.5%	0.3–0.5	[107]
	Polyacrylonitrile (PAN) + Polycaprolactone (PCL)	0.3 m/s	121 Pa	99.95%	PM2.5	[108]
	PAN + Poly (acrylic acid) (PAA)	5.3 cm/s	160 Pa	99.994%	0.3–0.5	[101]
	PAN + Poly (ethylene oxide) (PEO) + Polysulfone (PSU)	32 L/min	95 Pa	99.992%	0.3–0.5	[109]
	PAN + PA-6	5.3 cm/s	117.5 Pa	99.9998%	0.3	[102]
	PSU + PAN + PA-6	-	118 Pa	99.992%	0.3	[110]
	PVDF + FPU	-	67 Pa	99.98%	0.3	[111]
	PVDF + HFP	5.3 cm/s	71.4 Pa	99.50%	PM0.3	[112]
	PSU + TiO2 nanoparticles	30 L/min	45.3 Pa	99.997%	0.3–0.5	[113]
	PVDF + Negative Ions Powder (NIP)	-	40.5 Pa	99.99%	PM2.5	[114]
	PVDF + SiO2 + Ag	0.049 m/s	101.2 Pa	99.94%	0.3	[115]
	PAN + Ag	0.05 m/s	68.13 Pa	>98%	0.3	[103]
	PAN + TiO2	0.05 m/s	183.47 Pa	~100%	0.3	[103]
	PAN + TiO2	4 cm/s	-	99%	0.3	[104]
	Polylactic Acid (PLA) + TiO2	-	128.7 Pa	99.996%	-	[105]
Natural polymer fibers	Zein	8 L/min	15.18 Pa	97.36%	PM2.5	[116]
	Gelatin	0.06 m/s	148 Pa	97–99%	0.3–5	[117]
	Sericin	145 m3/h	-	~100%	PM2.5	[118]
Porous materials	Metal foams	0.1 m/s	~58 Pa	96.6%	0.1–0.4	[119]
	Metal foams	0.5 m/s	10.8 Pa	78.9%	0.3	[120]
	Carbon foam	-	112 Pa	>99%	PM2.5	[121]

Table 11. Studies on filter materials.

Type	Specific Materials	Air Flow Rate	Resistance	Filtration Efficiency	Particle Size (μm)	Ref.
Synthetic polymer fibers	Nylon 6 nanofibers	5 cm/s	150 Pa	99.993%	0.3	[106]
	PTFE + Poly (vinyl pyrrolidone) (PVP)	-	89.9 Pa	99.72%	-	[100]
	Polyvinyl Chloride (PVC) + Polyurethane (PU)	5.3 cm/s	144 Pa	99.5%	0.3–0.5	[107]
	Polyacrylonitrile (PAN) + Polycaprolactone (PCL)	0.3 m/s	121 Pa	99.95%	PM2.5	[108]
	PAN + Poly (acrylic acid) (PAA)	5.3 cm/s	160 Pa	99.994%	0.3–0.5	[101]
	PAN + Poly (ethylene oxide) (PEO) + Polysulfone (PSU)	32 L/min	95 Pa	99.992%	0.3–0.5	[109]
	PAN + PA-6	5.3 cm/s	117.5 Pa	99.9998%	0.3	[102]
	PSU + PAN + PA-6	-	118 Pa	99.992%	0.3	[110]
	PVDF + FPU	-	67 Pa	99.98%	0.3	[111]
	PVDF + HFP	5.3 cm/s	71.4 Pa	99.50%	PM0.3	[112]
	PSU + TiO2 nanoparticles	30 L/min	45.3 Pa	99.997%	0.3–0.5	[113]
	PVDF + Negative Ions Powder (NIP)	-	40.5 Pa	99.99%	PM2.5	[114]
	PVDF + SiO2 + Ag	0.049 m/s	101.2 Pa	99.94%	0.3	[115]
	PAN + Ag	0.05 m/s	68.13 Pa	>98%	0.3	[103]
	PAN + TiO2	0.05 m/s	183.47 Pa	~100%	0.3	[103]
	PAN + TiO2	4 cm/s	-	99%	0.3	[104]
	Polylactic Acid (PLA) + TiO2	-	128.7 Pa	99.996%	-	[105]
Natural polymer fibers	Zein	8 L/min	15.18 Pa	97.36%	PM2.5	[116]
	Gelatin	0.06 m/s	148 Pa	97–99%	0.3–5	[117]
	Sericin	145 m3/h	-	~100%	PM2.5	[118]
Porous materials	Metal foams	0.1 m/s	~58 Pa	96.6%	0.1–0.4	[119]
	Metal foams	0.5 m/s	10.8 Pa	78.9%	0.3	[120]
	Carbon foam	-	112 Pa	>99%	PM2.5	[121]

5.2. Structural Optimization

Based on the chosen filter materials, structural optimization can effectively improve airflow resistance by altering the airflow path and particle interception trajectory, thereby reducing energy consumption and enhancing filtration performance. This approach can be implemented at both the macroscopic and microscopic levels. At the macroscopic level, optimizing the geometry of the filter media (pleat ratio and pleat shape) can increase the effective filtration area, improve the dust-holding capacity and decrease the pressure drop, thereby reducing the fan energy consumption and operating load throughout the filter’s service life (Figure 9). At the microscopic level, modifications to the fiber structure offer new pathways for further enhancing filtration performance. Relevant studies are summarized in Table 12. Examples include a protrusion structure [122,123], a branched structure [105,124,125], and a bead-on-string structure (Figure 10) [126,127,128]. Fibers with porous or hollow structures exhibit high porosity, which significantly enhances the capture efficiency of fine particles [129]. These materials can achieve >99.99% filtration efficiency for particles smaller than 0.3 μm under pressure drops below 100 Pa. Similarly, protrusion and bead-on-string structure materials improve filtration performance by enhancing the fiber surface roughness and increasing the specific surface area, thus offering enhanced particle capture capabilities under low-resistance conditions [130].

Table 12. Studies on filter material structures.

Specific Materials	Type	Air Flow Rate	Resistance	Filtration Efficiency	Particle Size (μm)	Ref.
PLA nanofibers	Porous structure	32 L/min	90.35 Pa	99.9989%	PM0.3	[131]
PLA fiber composite	Porous structure	5.3 cm/s	93.3 Pa	99.999%	0.26	[132]
BaTiO3/PES	Protrusion structure	-	67 Pa	99.99%	-	[122]
SiO2 nanofilament-based Poly (m-phenylene isophthalamide) (PMIA)	Protrusion structure	-	170 Pa	97.33%	PM2.5	[123]
Chitosan/ PEO@MOF-	Protrusion structure	3.4 m3/h	44 Pa	99.95%	PM2.5	[133]
PLA	Bead-on-string structure	5.8 cm/s	165.3 Pa	99.997%	0.26	[126]
Graphene oxide-in-PAN composite	Bead-on-string structure	5.31 cm/s	8 Pa	99.97%	PM2.5	[127]
PVDF/SiO2 hollow fibers	Hollow structure	-	100 Pa	99.9999%	0.01–0.42	[134]
PVDF-PEG hollow fibers	Hollow structure	9.2 cm/s	-	99.999%	0.03	[135]
PVDF/TLNMs	Branched structure	32 L/min	124.2 Pa	99.999%	0.26	[105]
PS/BTA	Branched structure	-	22 Pa	99.8%	0.3–10	[124]
Polyamide-66 (PA-66)/BaCl2/ Polypropylene (PP)	Net structure	60 ± 2 L/min	~245 Pa	99.9%	-	[136]
PMIA	Net structure	32 L/min	92 Pa	99.999%	0.3–0.5	[137]
PVDF	Net structure	32 L/min	66.7 Pa	99.985%	-	[138]
PSU/PAN/PA-6	Net structure	32 L/min	118 Pa	99.992%	PM0.3	[110]
N-6 PAN NNF	Net structure	90 L/min	~75 Pa	99.99%	PM0.3	[139]

Table 12. Studies on filter material structures.

Specific Materials	Type	Air Flow Rate	Resistance	Filtration Efficiency	Particle Size (μm)	Ref.
PLA nanofibers	Porous structure	32 L/min	90.35 Pa	99.9989%	PM0.3	[131]
PLA fiber composite	Porous structure	5.3 cm/s	93.3 Pa	99.999%	0.26	[132]
BaTiO3/PES	Protrusion structure	-	67 Pa	99.99%	-	[122]
SiO2 nanofilament-based Poly (m-phenylene isophthalamide) (PMIA)	Protrusion structure	-	170 Pa	97.33%	PM2.5	[123]
Chitosan/ PEO@MOF-	Protrusion structure	3.4 m3/h	44 Pa	99.95%	PM2.5	[133]
PLA	Bead-on-string structure	5.8 cm/s	165.3 Pa	99.997%	0.26	[126]
Graphene oxide-in-PAN composite	Bead-on-string structure	5.31 cm/s	8 Pa	99.97%	PM2.5	[127]
PVDF/SiO2 hollow fibers	Hollow structure	-	100 Pa	99.9999%	0.01–0.42	[134]
PVDF-PEG hollow fibers	Hollow structure	9.2 cm/s	-	99.999%	0.03	[135]
PVDF/TLNMs	Branched structure	32 L/min	124.2 Pa	99.999%	0.26	[105]
PS/BTA	Branched structure	-	22 Pa	99.8%	0.3–10	[124]
Polyamide-66 (PA-66)/BaCl2/ Polypropylene (PP)	Net structure	60 ± 2 L/min	~245 Pa	99.9%	-	[136]
PMIA	Net structure	32 L/min	92 Pa	99.999%	0.3–0.5	[137]
PVDF	Net structure	32 L/min	66.7 Pa	99.985%	-	[138]
PSU/PAN/PA-6	Net structure	32 L/min	118 Pa	99.992%	PM0.3	[110]
N-6 PAN NNF	Net structure	90 L/min	~75 Pa	99.99%	PM0.3	[139]

Figure 9. Schematic illustration of rectangular-type pleat, U-shaped pleat, semi-triangular-type pleat, and triangular pleat [140]. Copyright 2023 Springer Nature.

Figure 9. Schematic illustration of rectangular-type pleat, U-shaped pleat, semi-triangular-type pleat, and triangular pleat [140]. Copyright 2023 Springer Nature.

Figure 10. SEM images of different fiber structures. (a) Porous fiber structure [141]. Copyright 2017 American Chemical Society. (b) Protrusion fiber structure [122]. Copyright 2018 Elsevier. (c) Bead-on-string fiber structure [128]. Copyright 2019 Elsevier. (d) Hollow fiber structure [134]. Copyright 2018 Elsevier. (e) Branched fiber structure [125]. Copyright 2018 Elsevier. (f) Net fiber structure [142]. Copyright 2020 John Wiley & Sons.

Figure 10. SEM images of different fiber structures. (a) Porous fiber structure [141]. Copyright 2017 American Chemical Society. (b) Protrusion fiber structure [122]. Copyright 2018 Elsevier. (c) Bead-on-string fiber structure [128]. Copyright 2019 Elsevier. (d) Hollow fiber structure [134]. Copyright 2018 Elsevier. (e) Branched fiber structure [125]. Copyright 2018 Elsevier. (f) Net fiber structure [142]. Copyright 2020 John Wiley & Sons.

5.3. Electrostatic Force Combination

Building on material and structural optimizations, the combination of electrostatic force can further enhance filtration performance and reduce energy consumption. The application of electrostatic forces improves the particle capture efficiency of filter media, enabling high filtration performance to be maintained while minimizing airflow resistance. According to Zaatari et al., replacing high-resistance filters with low-resistance filters can reduce fan energy consumption by 11–18%, which indicates that low-resistance electrostatic filters tend to hold considerable potential for reducing energy consumption and improving the overall performance of cleanroom air conditioning systems [20]. Current research on electrostatic filtration technology can be primarily classified into two categories: (1) filters utilizing charged media (electret filters) and (2) filters employing external electric fields.

5.3.1. Electret Filters

Electret materials integrate fiber filtration with electrostatic mechanisms [143], in which internal electrostatic fields are employed to improve particle capture efficiency [98,144]. Common methods for fabricating electret materials include corona charging [145,146,147], induction charging [22,148,149], and triboelectric charging [150,151]. A summary of related research is provided in

Table 13. Studies on electret filter materials.

Material	Charging Type	Air Flow Rate	Resistance	Filtration Efficiency	Particle Size	Ref.
PP resins	Corona charging	14.16 cm/s	22.45 Pa	98.6%	PM0.1–2	[145]
PVDF	Corona charging	5.3 cm/s	~25 Pa	>96%	PM0.15	[146]
PVDF	Corona charging	5.3 cm/s	<30 Pa	>94%	PM0.1	[147]
PP-BaTiO3	Corona charging	5.3 cm/s	95 Pa	99.97%	PM0.3	[154]
PVB/Si3N4-FPU	Induction charging	32 L/min	55 Pa	99.95%	PM2.5	[155]
PEI-SiO2	Induction charging	32 L/min	61 Pa	99.992%	PM0.3	[22]
PS/PAN/PS	Induction charging	5.3 cm/s	54 Pa	99.96%	PM0.3	[148]
PVDF/PTFE	Induction charging	-	57 Pa	99.972%	-	[21]
PVDF	Induction charging	30 L/min	66.7 Pa	99.985%	PM0.26	[138]
PVDF	Induction charging	5.33 cm/s	93 Pa	99.998%	PM0.3	[149]
PAN	Induction charging	0.21 m/s	-	>99.97%	PM2.5	[156]
PTFE/nylon	Triboelectric Charging	6 L/min	<125 Pa	96%	PM2.5	[150]
PVDF-TrFE	Triboelectric Charging	0.35 m/s	-	94%	PM1.0	[151]

. Li et al. developed a polyetherimide-silica electret fibrous membrane via induction charging, which could achieve a high filtration efficiency (99.992%) with a low pressure drop (60 Pa) [22]. However, the charge retention capability and environmental stability of electret materials during long-term operation still require further investigation to ensure their feasibility and reliability in cleanroom applications [152,153].

5.3.2. Electrostatic Precipitators

Electrostatic precipitators (ESPs) primarily consist of a discharge electrode and a collector plate. By applying a high-voltage electric field, microorganisms and particulates are electrically charged and subsequently removed through electrostatic attraction. The relevant research is summarized in Table 14. Traditional electrostatic precipitators are limited by the length and surface area of the collector plate, resulting in relatively low filtration efficiency for submicron particles [157,158,159]. However, improvements in the electrode material (e.g., non-metallic [160] and dielectric coatings [161]) and structural design (e.g., porous collector plate [162] and corrugated collector plate [163]) have enhanced the ability to capture fine particles, increasing the removal efficiency of 0.05–0.3 μm particles to 88–95%. While electrostatic precipitators exhibit extremely low filtration resistance, ozone and other byproducts may be generated during the operation, raising concerns about potential environmental and health impacts [157,164].

An intense field dielectric (IFD) filter is a novel electrostatic precipitator that utilizes dielectric materials as carriers to form honeycomb-like hollow microchannels [165]. It can achieve a PM_2.5 removal efficiency of 99.8%, with a low pressure drop (approximately 10–50 Pa), enabling 85% energy savings relative to conventional media filters [166,167,168]. The experimental results by Ren et al. further indicate that, compared with fan units equipped with PP filters, those using IFD filters exhibit a 31.8% reduction in resistance and a 9% improvement in system energy efficiency [169]. In addition, under identical operating conditions, IFD filters significantly outperform traditional ESPs with respect to ozone emission control and particle filtration efficiency [168]. Nonetheless, the long-term operational stability of IFD filters still requires further validation through application-oriented research.

Table 14. Studies on filter with electrostatic fields.

ESP Type	Air Flow Rate/Exchange Rate	Filtration Efficiency	Particle Size	Ref.
Flat plate (FP) ESP	0.5 m/s	>96%	PM2.5	[170]
FP ESP	0.7 m/s	76.7%	PM0.2	[171]
FP ESP	944 L/s	70%	PM0.19–0.52	[172]
FP ESP	1.62 h−1	88%	PM0.3–0.4	[159]
FP ESP	1.62 h−1	99%	PM1–2	[159]
Bipolar ESP	0.8 m/s	99.1%	PM0.3–90	[173]
Non-metallic collection plates ESP	2 m/s	95%	PM0.1–0.3	[160]
Polymer-arrayed ESP	1 m/s	90%	PM0.3	[174]
ESP with dielectric coatings	1.6 m/s	92%	PM0.3–0.5	[161]
ESP with dielectric coatings	1.04 m/s	95%	PM0.3	[175]
Crenelated plate ESP	-	88%	PM0.05	[162]
Wavy Plate (WP) ESP	1 m/s	96%	PM0.5–2.5	[176]
WP ESP	1 m/s	~96% (24% higher than FP)	PM1	[177]
Cylindrical ESP	200 m3/h	94.6%	PM0.3	[178]
IFD	6 m3/min	90.7% (48.5% higher than ESP)	PM0.5–1	[179]
IFD	1 m/s	99.8%	PM2.5	[166]
IFD	2.5 m/s	78.6% (12.4–38% higher than ESP)	PM0.3–0.5	[168]
IFD	1.13 m/s	98%	PM2.5	[169]

Table 14. Studies on filter with electrostatic fields.

ESP Type	Air Flow Rate/Exchange Rate	Filtration Efficiency	Particle Size	Ref.
Flat plate (FP) ESP	0.5 m/s	>96%	PM2.5	[170]
FP ESP	0.7 m/s	76.7%	PM0.2	[171]
FP ESP	944 L/s	70%	PM0.19–0.52	[172]
FP ESP	1.62 h−1	88%	PM0.3–0.4	[159]
FP ESP	1.62 h−1	99%	PM1–2	[159]
Bipolar ESP	0.8 m/s	99.1%	PM0.3–90	[173]
Non-metallic collection plates ESP	2 m/s	95%	PM0.1–0.3	[160]
Polymer-arrayed ESP	1 m/s	90%	PM0.3	[174]
ESP with dielectric coatings	1.6 m/s	92%	PM0.3–0.5	[161]
ESP with dielectric coatings	1.04 m/s	95%	PM0.3	[175]
Crenelated plate ESP	-	88%	PM0.05	[162]
Wavy Plate (WP) ESP	1 m/s	96%	PM0.5–2.5	[176]
WP ESP	1 m/s	~96% (24% higher than FP)	PM1	[177]
Cylindrical ESP	200 m3/h	94.6%	PM0.3	[178]
IFD	6 m3/min	90.7% (48.5% higher than ESP)	PM0.5–1	[179]
IFD	1 m/s	99.8%	PM2.5	[166]
IFD	2.5 m/s	78.6% (12.4–38% higher than ESP)	PM0.3–0.5	[168]
IFD	1.13 m/s	98%	PM2.5	[169]

6. Optimization Strategies Based on Intelligent Algorithms

To maintain the required cleanliness level, cleanroom air conditioning systems must operate efficiently and stably throughout the year. Traditional control methods, such as demand control filtration and pressure gradient adjustment, although capable of reducing fan energy consumption, largely depend on empirical formulas. This reliance limits their ability to respond to nonlinear disturbances and complex operating conditions, thereby constraining the potential for further energy efficiency optimization. In recent years, intelligent algorithm-based optimization strategies have gradually gained attention. They predict system responses through data-driven approaches, enabling dynamic and precise control. Jing et al. applied machine learning to ventilation systems, successfully controlling the supply air velocity error within 4.6% and achieving a 14.3% reduction in fan energy consumption [51]. Also, Ni et al. integrated AI models into semiconductor cleanroom environments, demonstrating significant potential for energy-savings [180]. Although the application of intelligent algorithm-based technologies in cleanroom air conditioning systems is currently limited, the promising potential for enhancing energy utilization efficiency offers valuable directions for the development of future energy-saving control strategies.

Research on intelligent algorithms for energy-saving control in HVAC systems can be categorized into two types: model-based and model-free (as summarized in

Table 15. Studies on optimization strategies of intelligent algorithms.

Classification	Modeling Method	Algorithm	Energy-Saving Efficiency	Ref.
Model based	ANN	MPC	37.8% and 40.8% (cooling and heating)	[182]
	ANN	Multi-objective Genetic Algorithm (MOGA)	18.2%	[181]
	Random forests	Not mentioned	15.1%	[183]
	Occupancy prediction model	MPC	8%	[23]
	Convolutional Neural Network (CNN) + Single-step Prediction Response Coefficient (SPRC)	Hybrid Model Predictive Control (HMPC)	-	[184]
	Feedforward Neural Network (FNN)	Deep Deterministic Policy Gradient (DDPG)	4.31%	[185]
Classification	Algorithm	Specification	Energy-Saving Efficiency	Ref
Model-free	Reinforcement Learning (RL)	Q-learning	26.48% (per day)	[186]
	RL	Q-Learning	7.7% (compared to rule-based control controller)	[187]
	RL	Proximal Policy Optimization (PPO)-Clip	22% (per week)	[24]
	RL	Deep Reinforcement Learning (DRL)-(Deep Q-network) DQN	4–5%	[188]
	RL	DRL-DQN	5–12%	[189]
	RL	DRL-DDPG	15%	[190]
	RL	DRL-DDPG	27–30%	[191]
	RL	DRL-DQN	13%	[192]
	RL	Hybrid-model-based DRL	26.99%	[193]
	RL	DRL-Decoupled Adversarial Long Short-Term Memory (DAL)-PPO	8% (compared to traditional RL)	[194]
	RL	Multi-task Learning (MTL)	20.9% (fan)	[53]
	GA + RL	GA + Multi-agent Deep Deterministic Policy Gradient (MADDPG)	6.7%	[195]

). The former approach establishes mathematical models from historical data (such as environmental parameters and occupancy) to predict energy consumption trends and optimize equipment operation, thereby reducing the overall energy consumption of the system [181]. Macintosh et al. developed an model predictive control (MPC) system utilizing an artificial neural network (ANN) model to improve the energy consumption in air conditioning systems [182]. It can reduce the cooling and heating energy consumption by 37.8% and 40.8%, while maintaining the temperature within ±0.5 °C deviations from the setpoint in 98% of cases [182]. However, the performance of this method highly depends on the model accuracy, and the computational processes tend to be time consuming when handling frequent variations in environmental loads and boundary conditions under practical operating conditions.

Model-free methods, by contrast, do not require the prior establishment of environmental models. By interacting directly with the environment and learning the input–output relationships from system operational data, it enables dynamic optimization of operational parameters [196]. DRL, as a combination of deep learning and reinforcement learning, enhances the optimization ability in dealing with complex air conditioning control problems. According to the study of Zou et al., it achieved optimal control over the AHU with a mean squared error (MSE) as low as 0.0015 and reduced energy consumption by 27–30% compared with conventional operating strategies [191]. Nevertheless, due to the variability in air conditioning system operational characteristics across different buildings, calibration for a specific building environment is necessary before implementation to ensure the effectiveness of the control performance [24]. Furthermore, the generalization ability of intelligent algorithms in energy consumption prediction is restricted by the quality and volume of data. Therefore, the prediction accuracy may decline when exposed to extreme weather conditions [197].

7. Discussion and Future Perspectives

7.1. Discussion

As the primary source of energy consumption in cleanroom operation, cleanroom air conditioning systems are characterized by large air supply volumes, considerable cooling and heating loads, significant system resistance, high fan pressures, and continuous high-efficiency operation [2,76]. The energy-saving strategies discussed in this review were developed based on these characteristics. Existing research has achieved notable progress in improving the energy efficiency of cleanroom air conditioning systems, particularly in the areas of temperature–humidity control and air delivery optimization. By reducing cooling and heating loads and minimizing fan energy consumption, these strategies have effectively mitigated overall system energy demand and demonstrated respectable energy-saving potential [14,16,36,40]. However, despite the encouraging outcomes, current measures have yet to fundamentally alleviate the energy burden, and improvements in overall system efficiency remain challenging. Further analysis reveals that fan, as the core driving components responsible for maintaining air circulation, contributes substantially to total energy consumption. Especially under high cleanliness requirements, fan must overcome considerable internal resistance to maintain stable airflow organization and pressure differentials, leading to significant increases in electricity consumption [17]. Among these resistance sources, the air filtration system tends to be the primary contributor to the persistently high energy usage of fan, accounting for 25.3% to 46.6% of total air conditioning system energy consumption, and represents a key bottleneck in advancing system efficiency [2,17]. Although recent advances in filter media such as nanofibers and electret materials have achieved promising reductions in initial pressure drop, the resistance will still increase inevitably over time due to particulate accumulation during continuous operation. This in turn escalates the fan load and degrades the overall energy performance of the system [37].

7.2. Future Perspectives

7.2.1. Optimization of Non-Consumable Filtration Technology and Its System Construction

Conventional fiber-based filters are associated with high resistance, elevated energy consumption, and limited service life, leading to significant increases in system operating costs during long-term operation [198]. Although existing studies have sought to optimize filter performance by improving filter media materials and fiber structures to reduce pressure drop and extending lifespan, they have not fundamentally addressed the issue of progressively increasing resistance during continuous use [122,124,127,139,199]. Hence, the development of non-consumable filtration technologies featuring low resistance, high efficiency, and long-term sustainability is of critical importance. Electrostatic filtration technologies represent an effective approach to achieving non-consumable air filtration. ESPs have been widely applied in industrial waste gas treatment and air conditioning systems. According to research, ESPs can achieve high filtration efficiency for 2.0–5.0 μm particles [168]. But, the performance in capturing smaller particles remains limited. More recently, advancements in IFD technology have significantly enhanced the overall performance of electrostatic filtration. It expanded the effective particle size range (0.4–5.0 μm) and offered theoretical potential to replace sub-HEPA filter in conventional cleanroom air conditioning systems [166,168].

In the future, through in-depth research on IFD technology, especially enhancing its capture capacity for large-sized particles, a comprehensive replacement from primary filter to medium efficient particulate filter is expected to be achieved. Based on this, the filtration architecture of cleanroom air conditioning systems could shift from a traditional multi-stage mechanical filtration system to a novel system centered on non-consumable technologies. In the corresponding system design, an optimized IFD filter (with broad particle size coverage, high filtration efficiency, and large dust-holding capacity) would only require a pre-filter (e.g., wire mesh) upstream to intercept large airborne contaminants (e.g., insects, catkins, etc.) to prevent electrode short circuits caused by foreign object bridging. Meanwhile, a HEPA filter is retained at the terminal stage of the system as a safety backup to manage emergency scenarios such as power outages or incomplete particle capture by upstream filters. At this stage, the surface resistance of the HEPA filter is significantly lower than that in conventional configurations, which helps extend the service life and further reduces the overall energy consumption of the system. According to initial resistance values reported by Zhao et al. and Wang et al., this proposed system is estimated to reduce initial resistance by approximately 380 Pa, which accounts for 58% of the resistance in conventional cleanroom filtration systems [2,168]. Before deploying IFD technology in real-world cleanroom air conditioning systems, it is essential to ensure that its filtration efficiency, operational stability, and system safety fully meet the required cleanroom standards.

On this basis, future research could further focus on the integrated optimization of non-consumable filtration systems, such as the development of self-cleaning filter modules. These innovations will support achieving truly integrated, low-resistance, and high-reliability clean-air delivery systems, laying a technological foundation for the energy efficient and sustainable development of cleanroom air conditioning systems.

7.2.2. The Intelligent Evolution of Control Strategies

In cleanroom air conditioning systems, strategies such as DCF and pressure gradient control largely rely on empirical formulas or preset rules, which limits the capability to effectively respond to dynamic system changes and to achieve precise control under complex environmental conditions [47]. To address this issue, current control methods can be enhanced by integrating intelligent algorithms that leverage historical operation data and real-time monitoring information [191,192]. This integration can uncover latent correlations between cleanroom parameters and energy consumption, enabling the development of predictive and adaptive control models, thereby achieving the transformation from “passive response” to “active prediction and decision making” and promoting the development of cleanroom air conditioning systems towards greater efficiency and intelligence [180,195,200]. For example, combining DCF strategies with machine learning allows for dynamic optimization of fan operation parameters based on indoor particle concentrations and occupancy levels. This enables real-time adaptation of control strategies, enhancing both energy utilization efficiency and operational stability. In the future, distributed intelligent control architectures can be adopted to enable coordinated optimization of multi-zone cleanroom air conditioning systems [188]. Compared to single-module optimization approaches, distributed architectures facilitate information sharing and collaborative regulation across multiple zones. This enables real-time adjustment of air flow rate and pressure control, improves energy distribution, and ensures that each zone meets cleanliness requirements while maximizing the overall system efficiency. Overall, the integrated optimization provides an effective solution for the intelligent and efficient operation of multi-zone cleanroom air conditioning systems.

7.2.3. The Introduction of Renewable Energy

Despite continuous progress in energy-saving technologies, current cleanroom HVAC systems still depend heavily on high-intensity electricity consumption to maintain required temperature, humidity, airflow organization, and cleanliness levels [201]. This energy structure not only imposes sustained operational cost burdens but also exacerbates carbon emissions, being in conflict with the goals of green and low-carbon transitions [202]. Under the global framework of The Paris Agreement, energy consumption has become a central constraint to the sustainable development of the cleanroom industry, highlighting the urgent need for more efficient, flexible, and cleaner energy utilization pathways [203,204,205]. Therefore, in addition to optimizing filtration systems and integrating intelligent control strategies, the incorporation of renewable energy tends to be a necessary supplement to enhance system efficiency and support low-carbon transformation. In conventional air conditioning systems, studies have already explored the potential of partially replacing traditional electricity with renewable energy (particularly solar energy) [206,207]. For instance, photovoltaic (PV) arrays deployed on large rooftop areas can supply power for fans, chillers, and other HVAC components [208,209]. Moreover, solar thermal collectors can be used to preheat outdoor air or assist in reheating air after dehumidification, thereby reducing the energy demand from traditional electric heating [210,211]. These indicate that integrating solar energy into cleanroom air conditioning systems holds considerable promise for reducing reliance on conventional energy sources and lowering carbon emissions, thereby promoting greener system transitions [212]. Building upon this foundation, future research may also explore the integration of other clean energy (such as wind and geothermal energy) into cleanroom HVAC systems to achieve cost savings and reduce the lifecycle carbon footprint of cleanroom facilities.

7.2.4. Integration and Collaborative Optimization of Multiple Energy-Saving Strategies

A large number of existing studies focus on the local optimization of individual modules, lacking systematic exploration of the interaction relationships among different technical paths and their potential for collaborative enhancement. In practical applications, the optimal energy-saving performance of cleanroom air conditioning systems often depends on the coordinated implementation of multiple measures [51,190]. Therefore, future research should place greater emphasis on multi-strategy integration, taking a holistic perspective to examine the interdependence and synergy among airflow distribution, temperature and humidity regulation, filtration performance, and intelligent control strategies. In addition, the application of advanced data-driven approaches and multi-objective optimization algorithms can contribute to the identification of optimal combinations of energy-saving measures under diverse operating conditions and performance goals, laying the foundation for achieving truly intelligent and low-carbon cleanrooms [187,195].

8. Conclusions

As one of the most energy-intensive subsystems in cleanroom operation, cleanroom air conditioning systems are not only crucial for improving technical performance but also for advancing industrial sustainability and achieving carbon reduction goals. This article reviewed the existing energy-saving methods for cleanroom air conditioning systems, covering system air supply, air handling process, air filtration modules, and intelligent algorithm-based strategies, aiming to provide a systematic framework and theoretical foundation for future energy.

In terms of system air supply, optimizing key parameters such as air velocity and redistributing airflow can significantly improve operational efficiency and reduce energy consumption. Also, the optimization of fan structure and operating characteristics has been proven crucial for reducing air resistance and enhancing energy efficiency. For the air handling process, the mismatch between cooling and heating loads is the key factor contributing to energy waste. In recent years, researchers have proposed strategies such as demand reduction, heat recovery integration, and temperature–humidity decoupled control, which effectively mitigate cold–heat offset and enhance the system’s thermal regulation capability.

Improvements in air filters are also essential for energy conservation. Innovations in filter media materials, structural design, and the integration of electrostatic technologies can effectively reduce filtration resistance, extend service life, and enable coordinated reductions in energy consumption while maintaining air purification performance. However, most existing research on filter optimization has focused on conventional air conditioning systems, with limited studies addressing the unique requirements of cleanroom air conditioning systems, such as high cleanliness and operational stability. In future applications of low-resistance and long-lifespan filtration technologies in cleanroom air conditioning systems, particular attention should be given to their long-term operational stability, maintenance of filtration efficiency, and system compatibility, to ensure that energy-saving measures remain effective and reliable under stringent environmental conditions.

Meanwhile, the integration of intelligent algorithms has provided a new direction for dynamic optimization in cleanroom air conditioning systems. Leveraging machine learning and deep learning algorithms, the systems can achieve real-time perception of environmental conditions, load forecasting, and adaptive control, enabling precise regulation and energy efficiency optimization under complex operating scenarios. However, current research on intelligent algorithm-based strategies, also largely confined to conventional building air conditioning systems, and their applicability in cleanrooms, where higher reliability, safety, and response accuracy are required, has not been fully validated. Future studies should prioritize the adaptation of intelligent algorithms to cleanroom scenarios, enhancing system fault tolerance and real-time responsiveness to meet the strict requirements of cleanroom air conditioning systems for control stability.

However, cleanroom air conditioning systems still face multiple challenges in achieving high-efficiency operation, including persistently high system resistance, lagging response of traditional control strategies, and strong reliance on conventional energy sources. Future research should focus on optimizing non-consumable filtration technologies, developing novel low-resistance filtration systems, deeply integrating intelligent control strategies, and incorporating renewable energy sources, in order to establish a multidimensional and systematic pathway for energy efficiency optimization.

Overall, improving the energy efficiency of cleanroom air conditioning systems is not only a reflection of technological innovations but also a proactive response to global carbon reduction strategies. This review provides a solid theoretical basis and technical reference for promoting the development of cleanroom air conditioning systems toward greener, low-carbon, intelligent, and high-efficiency operation, thereby laying an important foundation for the construction of sustainable and environmentally friendly cleanrooms.