Review on Noise Generation Issues and Noise Mitigation Methods in Electric Vehicle Charging Systems

Abstract

This paper presents an overview of issues related to noise generation in electric vehicle (EV) charging systems. It discusses the requirements for noise reduction in locations where charging stations are most commonly installed. The primary sources of noise in EV charging stations are identified, considering their design and configuration. The results of acoustic tests for specific noise sources and entire charging stations are presented, including measurements of sound pressure level (SPL), acoustic imaging, and the generated acoustic spectrum. The paper also describes noise reduction methods and proposes solutions aimed at minimizing the noise generated by charging infrastructure. Additionally, the results of tests illustrating the effectiveness of these methods are presented.

1. Introduction

Industrial development, public transport, the increase in the number of vehicles, and street traffic all contribute to the rise in environmental noise. In recent years, noise emitted into the environment, including from industrial equipment, has become a serious problem. Research shows that excessive noise negatively impacts the human body and can lead to cardiovascular diseases, sleep disorders, and stress [1,2,3]. Studies on the effects of noise on human health have been conducted for many years [4], and advancements in science and knowledge in this area now enable us to explain the mechanisms behind these effects and their negative consequences [5]. The increase in environmental noise, driven by industrial development [6], and its impact on human health has become so significant that the World Health Organization (WHO) has issued recommendations on this matter [7,8]. One of the most dynamically developing industries in recent years has been electromobility, which has been widely described in [9]. The electromobility market, not only in Europe but also in Asia and North America, continues to grow. This development is evident both in the electric vehicle (EV) sector and in charging infrastructure and equipment (EVSE—Electric Vehicle Supply Equipment) [10,11,12,13,14]. A particularly significant technological advancement can be observed in the area of Fast Charging stations [15]. Reports from organizations related to the electromobility market also provide quantitative data on how the charging infrastructure developed in 2024. The report from “EVMarketsReport.com” indicates that the number of charging points in China reached 3.2 million, including 1.78 million DC charging stations. In Europe, the number stood at 0.9 million, with 147,000 DC stations, and the total is projected to grow to 8.8 million by 2030. In the USA, the number of charging points was 180,000, including 44,000 DC charging stations. Additionally, various forecasts indicate that the number of electric vehicles will continue to rise in the coming years. A comprehensive analysis with forecasts up to 2040 concerning the electromobility market in Poland, and beyond, is presented in [16]. An interesting perspective on the development of electromobility is also offered in [17]. When analyzing different approaches to the development of electromobility and various forecasting models, a clear upward trend emerges. This is further confirmed by the results published in [18,19,20]. The development of charging infrastructure is also influenced by the increasing capacity and charging power of electric vehicle batteries, which require a corresponding increase in the available power at individual charging points (Figure 1). A separate class of vehicles, whose traction batteries have a capacity of over 500 kWh, includes heavy vehicles (trucks, buses, etc.) that require high charging power (>350 kW). The power classes of EV charging stations are defined by the CharIn organization, and their minimum requirements are shown in

Table 1. Minimum requirements for EV chargers with various power classes.

Name	Power Classes	${\mathit{U}}_{\mathbf{out}}$[V]	${\mathit{I}}_{\mathbf{out}}$[V]	${\mathit{I}}_{\mathbf{peak}}$[V]	Duration ${\mathit{I}}_{\mathbf{peak}}$[A]	Density [kW/${\mathbf{m}}^3$]
Low-Power Charging	LPC	200–920	≤80 A	<20	inf	–
DC Charging	DC	200–920	80–125	≥20	inf	25–50
Fast Charging	FC	200–920	125–250	≥125	≥30 min	50–150
Ultra-Fast Charging	UFC	200–920	250–500	≥250	≥20 min	247.9
High-Power Charging	HPC	200–920	≥500	≥500	≥10 min	>200
Megawatt Charging System	MCS	200–1250	≤3000	*  n.d.	* n.d.	* n.d.

* no data.

. The minimum required output voltage range is 200 to 920 VDC, and up to 1250 VDC for MCSs. The minimum output current is up to 500 amps in the case of Ultra-Fast Chargers (UFCs) and over 500 amps for High-Power Charging (HPC). Currently, HPC chargers can operate with an output current of up to 850 A (peak current). For MCSs, the current value is planned to reach up to 3000 amps [21,22,23]. Additionally, a minimum load duration time at the output peak current ($I_{peak}$) value is defined. After this duration, current derating may be applied to enable the system to cool down. However, during derating, the output current must not drop below 75% of the peak current [24]. Currently, the charging station standards include Fast Charging (FC), with output currents of up to 250 amps, Ultra-Fast Charging (UFC), with output currents of up to 500 amps, and High-Power Charging (HPC), with currents exceeding 500 amps. In 2018, CharIN initiated the Task Force “Megawatt Charging System (MCS)”, aimed at achieving rated power exceeding 1 MW, charging currents of up to 3000 amps, and output voltages of up to 1250 VDC [21,22]. As the rated power of EV chargers continues to grow, so does their volumetric power density, meaning that AC/DC converters with higher power are being installed in chassis of similar dimensions. During the charging cycle of an electric vehicle battery, power losses occur as alternating current (AC) is converted to direct current (DC) by power electronic converters (AC/DC) [25,26]. These losses, which are almost entirely converted into heat, amount to approximately 5% of the charging station’s operating power. For fast DC charging stations operating in the kilowatt (kW) or megawatt (MW) range, the amount of heat generated is substantial. Due to the limiting operational parameters of power semiconductors, AC/DC converters, and the overall charging station components, the heat resulting from power losses must be effectively dissipated to ensure reliable operation and prolong the service life of these components [27,28,29]. Electric vehicle charging stations with high power and increasing power density require significantly more intensive cooling. This necessitates high-intensity forced airflow through the AC/DC power block and other components of the charging station. The most commonly used cooling method for AC/DC converters is forced air cooling. However, the fans in these power converters, operating at high intensities, are a primary source of noise emitted into the environment [30,31].

While the transition to electric transport offers numerous benefits, the charging infrastructure, particularly high-power stations, can contribute to noise pollution, potentially affecting the acoustic environment. The World Health Organization (WHO) recommends safe environmental noise levels of 55 dB(A) during the day and 45 dB(A) at night [7]. However, these values can vary slightly between European countries and even between regions within individual countries. Based on a directive of the European Parliament [31], the so called “Noise Directive”, permissible noise levels in the environment are determined. These levels depend on the nature of the environment (e.g., industrial or residential areas, sanatoriums, hospitals, schools, and health resorts) and the time of day. Examples of applicable noise levels in Germany [32] and Poland [33] are presented in

Table 2. Example values, in [dB(A)], of noise levels in the environment depending on the time of day and type of area.

Type of Area (Environment)	Day		Night
	PL	DE	PL	DE
Spas, hospitals, nursing homes	* 45	45	40	35
	** 50
residential, city center, agglomerations over 100 thousand inhabitants	55	60	45	45
residential, commercial, less than 100 thousand inhabitants		55		40

* out of town; ** in cities.

. According to Table 2, in urban areas, particularly near buildings intended for the permanent or long-term stay of children and adolescents (e.g., schools, universities, kindergartens, and dormitories), the permissible noise level—measured as sound pressure—should not exceed 50 dB(A) during the day and 40 dB(A) at night. For multi-family housing developments, collective housing, recreational and leisure areas, and residential-service areas, the permissible noise level is set at 55 dB(A) during the day and 45 dB(A) at night. Often, the ambient acoustic background is higher than the noise generated by devices, such as EV charging stations. In such cases, it must be demonstrated that the device does not increase the background noise level during operation. There is currently no scientific literature on the measurement of noise generated by charging infrastructure. To date, noise research has primarily focused on traditional sources, such as road, rail, and air transport, as well as industrial and construction activities. Meanwhile, noise generated by electric vehicle (EV) charging infrastructure remains relatively underexplored. The existing literature on this topic offers limited guidance, with recommendations regarding noise mainly provided by the CharIn organization [34]. Charging station manufacturers also do not provide detailed data on noise emissions. Taking the above into account, the discussed topic is not only current but provides a new perspective on the charging infrastructure for electric vehicles. The presented research results provide completely new knowledge on the impact of charging infrastructure on the environment in the area of noise emissions in terms of identifying noise sources, quantitative assessment of the emitted noise, and methods for its reduction.

One of the main motives for addressing the impact of electric vehicle charging infrastructure on environmental noise was the desire to initiate a discussion in this area and to highlight the fact that such devices can constitute significant sources of noise when installed in public spaces. Another motivation was the lack of research results on this topic in the existing literature. The main aim of this paper is to review the systems and technical solutions of EV charging infrastructure, identify noise sources, determine their acoustic characteristics, and evaluate the noise levels generated in the environment depending on the type of charging station. The primary objective of this study is to examine issues related to noise emissions from EV charging infrastructure, assess the acoustic parameters and characteristics of noise emitted by charging stations, and propose methods to reduce and limit noise generation—particularly in the case of Fast Charging stations. The rapidly developing charging infrastructure and the increasing number of Fast Charging stations may contribute to higher noise emissions into the environment. Consequently, in light of research on the effects of noise on the human body, this could have a negative impact on human health. Identifying noise sources in charging stations and implementing noise reduction methods will help to minimize noise emissions and mitigate their environmental impact. The paper is organized as follows: Section 2 identifies noise sources in EV chargers, presenting the results of acoustic measurements and acoustic imaging of specific noise sources. Section 3 provides information on noise emissions from EV chargers, while Section 4 discusses passive and active noise reduction methods and presents the implementation results of these methods in EV Fast Chargers. Finally, Section 5 presents the conclusions.

2. Identification of Noise Sources

2.1. General Construction of DC EV Chargers

Due to their construction, EV charging stations can be divided into two types: (i) monolithic chargers, also called standalone, where user interface with charging cables and connectors and power converter blocks are installed in the same casing; and (ii) split-type, where casing, with AC/DC converters—called power units (PUs)—and user interface with HPC cables and connectors—called user units (UUs)—are separate units in two separate casings. Monolithic chargers are usually with charging power up to 400 kW, while split-type charging stations are constructed usually with a charging power of over 400 kW. Regardless of the rated power and type of charging station, each charger is equipped with a converter power block, in addition to the necessary AC and DC switchgear with electrical protections, as well as a control system, HMI, and other devices (Figure 2). The power block, depending on its rated power, typically contains several to a dozen or more AC/DC converters (Figure 2). Each converter is equipped with multiple fans (usually two to four) to force airflow for cooling. For example, a 400 kW charging station equipped with 10 converters rated at 40 kW each, with each converter having 4 high-speed forced-air fans, would have a total of 40 fans. The cooling fans of AC/DC converters operate at high rotational speeds. Depending on the ambient temperature and the internal temperature of the converter, the rotational speed can range from a few hundred to several thousand RPM. A single fan can generate noise, with a sound pressure level (SPL) of up to 70–80 dB(A). Additionally, High-Power Charging (HPC) stations with liquid-cooled charging connectors are equipped with liquid–air heat exchangers, which include additional fans. The number of fans depends on the design of the heat exchanger and can range from one to several units. Liquid-cooled AC/DC converters are also beginning to enter the market. These liquid-cooled charging stations feature dedicated cooling systems that include liquid–air heat exchangers (chillers).

2.2. Main Noise Sources

In charging stations, several main sources of noise can be identified depending on the type and construction of the station (Figure 3). For air-cooled EV chargers, which account for nearly all installed devices, the primary source of noise is from AC/DC converters, specifically the fans used to force cooling airflow. In the case of liquid-cooled EV chargers, the primary source of noise is from the fans of the liquid–air heat exchanger. Additionally, if the charger is equipped with a liquid-cooled HPC cable, an additional source of noise is the chiller associated with the HPC cable and charging connector.

2.2.1. AC/DC Converter

The main sources of noise in DC electric vehicle chargers with forced air cooling systems are the AC/DC converters. The acoustic characteristics of an AC/DC converter—including parameters such as sound pressure level, acoustic spectrum, dominant frequencies, and directionality—significantly influence the noise emitted by the charging station into the environment. By conducting acoustic tests on the AC/DC converter, it is possible to determine key acoustic parameters during the design stage of the charging station.

Test Method

All acoustic tests described in this paper were conducted in accordance with relevant standards [35,36,37,38]. The research was carried out using a 4-channel class 1 vibration and sound meter (SVAN 958A) with a set of 4 microphones, acoustic camera with a hexagonal array of 128 microphones (Nor848B/HEX-K1), and sound intensity tracking and visual mapping system—equipped in class 1 dual microphone sound intensity probe (MEZZO I-Track System Kit) (

Table 3. List of measuring equipment and their functions.

Measuring Device	Type	Function/Measurement
4-channel sound analyser	SVAN 958A	Sound pressure level, acoustic spectrum
Acoustic camera Norsonic	Nor848B/HEX-K1	Acoustic imaging, acoustic spectrum
Acoustic imaging system	MEZZO I-Track system	Acoustic imaging, sound intensity

).

The measurement methods and procedures for determining acoustic parameters, as outlined in these standards, are well established and widely discussed in the literature [39,40]. An exemplary acoustic test setup for an AC/DC converter, utilizing a semi-spherical measuring surface above a reflecting plane in accordance with EN-ISO 3745 [36], is shown in Figure 4. During acoustic testing, the AC/DC converter operates at its rated power and maximum charging current under an ambient temperature of 20 ^∘C. Before initiating the acoustic measurements, the AC/DC converter undergoes a warm-up period (at nominal charging power and maximum charging current) until its internal temperature stabilizes, defined as a temperature variation of $\Delta T<1$ ^∘C, over 10 min. The arrangement of measurement points on the semi-spherical surface, as specified by the standard [36], is illustrated in Figure 5. Noise measurements at each point are recorded for no less than 20–30 s and are averaged over time to determine the equivalent continuous sound level ($L_{A_{eq}}\left(i\right)$). According to [36], the average time-averaged sound pressure level is determined by following Equation (1):

$$\begin{array}{c}\hfill \overline{{L^{\prime }}_{A_{e}q}}=10log\left({\displaystyle \frac{1}{N}}\sum _{i=1}^N{10}^{0.1L_{Aeq}\left(i\right)}\right).\end{array}$$

(1)

The average time-averaged sound pressure level generated by the tested device is determined taking into account the background noise corrections ($K_1$) and the environmental correction ($K_2$).

$$\begin{array}{cc}\hfill \overline{L_{Aeq}}& =\overline{{L^{\prime }}_{A_{e}q}}-K_1-K_2,\hfill \end{array}$$

(2)

$$\begin{array}{cc}\hfill K_1\hfill & \hfill =10log\left(1-{10}^{-0.1(L_{Aeq}\left(i\right)-L_{bg})}\right).\hfill \end{array}$$

(3)

Because the measurements were conducted in an anechoic chamber and the background noise level ($L_{bg}$) is much lower than the noise generated by the tested converter, and there is no reverberation phenomenon, these corrections take values equal to zero ($K_1=K_2=0$). Based on the measurements of the sound pressure level, the sound power level of the noise source is determined (4)

$$\begin{array}{c}\hfill L_w=\overline{L_{Aeq}}+10log\left({\displaystyle \frac{S}{S_0}}\right)+C_1+C_2+C_3,\end{array}$$

(4)

where S—measurement surface area [${\mathrm{m}}^2$], $S_0$—reference surface, $C_1$—reference quantity correction, $C_2$—acoustic radiation impedance correction, and $C_3$—correction taking into account sound absorption in the air

$$\begin{array}{cc}\hfill C_1& =-10log\left({\displaystyle \frac{p_s}{p_{S_0}}}\right)+5log\left({\displaystyle \frac{273+\theta }{314}}\right),\hfill \\ \hfill C_2& =-10log\left({\displaystyle \frac{p_s}{p_{S_0}}}\right)+15log\left({\displaystyle \frac{273+\theta }{294}}\right),\hfill \\ \hfill C_3& =A_0{\left(1.0053-0.0012A_0\right)}^{1.6}.\hfill \end{array}$$

Results of Acoustic Measures of AC/DC Converters

Below in Figure 5 are shown exemplary acoustic measure results of AC/DC converter with rated power of 40 kW. The exemplary acoustic spectrum in octave bands recorded at the measurement points (M1–4, M2–4, M3–4, and M4–4) is shown in Figure 6. Average time-averaged sound pressure level from air intake side is 73.6 dB(A), and 71.1 dB(A) from the air outlet side. The left and right side results are very similar and are, respectively, 71.2 dB(A) and 71.0 dB(A). The acoustic spectrum and identification of dominant frequencies are crucial factors in understanding noise emission and propagation. The conditions of noise propagation in the environment depend on several factors, including the attenuation coefficient of the atmosphere, which is influenced by temperature, humidity, and sound frequency. Generally, higher frequencies correspond to greater attenuation coefficients. As shown in Figure 6, the dominant frequency in octave bands is 1000 Hz (1 kHz). However, it should be noted that relatively high-frequency components are also present in the range of 500–4000 Hz. The speed of cooling fans depends on the ambient temperature and the internal temperature of the AC/DC converter, and the tests are conducted at an ambient temperature of 20 ^∘C. Figure 7 illustrates the sound pressure level as a function of fan speed for a single AC/DC_6 converter with a nominal power of 40 kW (see

Table 4. Sound pressure levels (SPLs) emitted by various types of AC/DC converters for EV chargers.

Model of AC/DC Converter	Rated Power	Power Density	Sound Pressure Level	Number of Fans
	[kW]	${\varrho }_{\mathbf{V}}\left[\mathbf{kW}/{\mathbf{m}}^3\right]$	$\overline{{\mathbf{L}}_{\mathbf{Aeq}}}$ [db(A)]
AC/DC_1	30	2135.9	64.0	3
AC/DC_2	30	1796.4	64.1	4
AC/DC_3	30	1642.9	67.6	3
AC/DC_4	30	2752.3	70.0	3
AC/DC_5	30	2727.3	78.2	3
AC/DC_6	40	3491	71.9	2
AC/DC_7	40	2847.9	76.0	3
AC/DC_8	40	3624	74.1	3

). Currently available AC/DC converter modules for EV charging stations have a rated power of 20–30 kW. From the point of view of generated noise, individual converter modules differ not only in rated power but above all in geometric dimensions and the number of fans pushing cooling air. The high power and small dimensions of the AC/DC converter cause large amounts of heat to be released, and the high degree of packing causes difficulties in the cooling airflow. As a consequence, it is necessary to force intensive airflow through the converter in order to dissipate the heat energy. The high rotational speed of the fans and the airflow through the converter generate a high level of noise. To compare various types of AC/DC converters, for each converter, power density is specified, understanding it as a relation between rated power to geometric volume, and quantity of air forcing fans. In Table 4, exemplary acoustic measurement results of various types AC/DC converters for EV chargers are collected. Results have been obtained at test condition with ambient temperature 20 ^∘C.

In Table 4, the lowest SPL is emitted by converters with lowest power density. Higher power density causes higher value of SPL emitted by converter. Manufacturers often increase the nominal power of the converter, leaving the dimensions of the converter housing unchanged. In the case with the AC/DC_4 and AC/DC_8 converters, the nominal power was increased from 30 kW to 40 kW, with similar dimensions of the housing. As a result, the volumetric power density increased from 2752.3 to 3624 kW/m³. Sound pressure level, because of more intense airflow, increased from 70 dB(A) to 74.1 dB(A).

2.2.2. HPC Cable Chiller

The noise emitted by the chiller of the HPC cable depends on the type (size) of the heat exchanger, the fans used, their type, quantity, and control method. Below (Figure 8) are the exemplary noise measurement results emitted by a user unit (UU) with an HPC liquid-cooled cable and a chiller equipped with seven fans. The measurements were conducted at a distance of 1 m from the surface of the UU housing on each of its four sides. Depending on the ambient temperature and the temperature of the cooling liquid, either 2, 5, or 7 fans may operate. As shown, the highest sound pressure level is generated from the air outlet side, reaching levels of up to 63 dB(A). It should be highlighted that, in case of tested UU, at ambient temperature 22 ^∘C, the noise emitted by UU is lower than 55 dB(A).

2.2.3. Liquid-Cooled AC/DC Converters

Liquid-cooled AC/DC converters for EV chargers do not have fans forcing airflow, and, consequently, they do not contain noise-generating components. However, EV charger systems with liquid-cooled AC/DC converters are equipped with air–liquid heat exchangers that include fans to enhance cooling performance. These air–liquid heat exchanger fans are the primary source of noise in liquid cooling systems. Figure 9 shows the results of acoustic imaging (Figure 9a) of an air–liquid heat exchanger (Figure 9b) for a liquid-cooled Ultra-Fast Charger (UFC). The tested heat exchanger is equipped with two fans, each with a 400 mm diameter and an adjustable airflow range of 0 to 4000 m³/h. Acoustic spectrum in octave bands, measured from each side of heat exchanger, is shown in Figure 10. The dominant frequency band is 500 Hz. However, the highest-frequency components are in range from 500 to 2000 Hz (0.5–2 kHz). As is visible (Figure 9 and Figure 10), the highest value of SPL is obtained at air outlet side. For volumetric flow intensity equal 1600 m³/h (it is about 40% of maximum speed of fan), SPL is 63.7 dB(A). The characteristic of SPL as a function of volumetric airflow through heat exchanger is shown in Figure 11. However, it should be highlighted that the useful and real operation range is up to 2000 m³/h (up to 50% of fan speed). This value is sufficient to maintain temperature of AC/DC converters at required level.

2.2.4. Converter Power Block

Each charging station is equipped with a power block that contains, depending on the charger’s nominal power, anywhere from a few to several AC/DC converters. Each individual AC/DC converter installed in the power block can be considered as a single source of noise. Since the power block contains n identical noise sources, the total sound pressure level can be described by Equation (5). In the case of the tested power converter block with a rated power of 400 kW, containing ten AC/DC_6 converters (see Table 4), the resulting SPL value exceeds 85 dB(A).

$$\begin{array}{c}\hfill \overline{L_A}=10log\left(\sum _{i=1}^n{10}^{0.1L_i}\right),\end{array}$$

(5)

where $L_i$ is an SPL emitted by a single AC/DC converter.

Figure 12 shows the sound pressure level as a function of fan speed emitted by a 400 kW power block, which contains 10 × 40 kW AC/DC_6 converters. As shown, the highest sound pressure level is emitted from the air intake side, similar to the case of a single AC/DC converter (see Figure 7).

3. Noise Emission from EV Chargers

Depending on the internal construction of the charging station, the noise emitted by the power converter block with AC/DC converters can be limited to some extent. Examples of the internal construction of the charging station, considering airflow pathways and noise propagation routes, are shown below (Figure 13). The first type of construction (Figure 13a) is very simple. Air flows straight through the casing, for example, from the front to the back. This type of casing is used when chargers need to be installed side by side or in split-type chargers in the power unit (PU). For standalone chargers, due to the user interface at the front, a casing where the airflow passes between the sides of the casing is more useful, as shown in Figure 13b. In the first case (Figure 13a), airflow between the sides of the casing can be achieved by rotating the AC/DC converters 90 degrees. However, in the construction shown in Figure 13b, a longer air intake path may help to reduce the noise generated by the converter block. For EV chargers with liquid-cooled AC/DC converters, a housing like the one shown in Figure 13c can be used, with air intakes on both sides of the housing and the air outlet at the back. In Figure 14, measuring results are presented of sound pressure level for casing shown in Figure 13b. Inside the casing, the power block described in Section 2.2.4 is installed, containing ten AC/DC_6 converters. Comparing the sound pressure levels shown in Figure 12 and Figure 14, it can be seen that lower SPL values were obtained after closing the power block in the housing. Noise from the air intake side of the AC/DC converters and from the rear was significantly reduced, but there was little change on both sides of the housing, where the air intakes are located. This phenomenon can be explained by the fact that sound waves are reflected inside the housing and spread sideways through the air intakes (see Figure 12). The noise emitted by this solution was further reduced using various methods described in the next section. Exemplary acoustic images and spectra emitted by the 400 kW charging station with the internal construction shown in Figure 13b are presented in Figure 15.

As is visible in acoustic spectra (Figure 15, Figure 16, Figure 17 and Figure 18), the dominant frequency component is over 0.9 kHz. The results obtained by acoustic imaging confirm the results obtained during measurements with the acoustic analyzer. The acoustic spectrum is typical for noise generated by airflow, with the dominant frequency (0.9 kHz) resulting from the rotation of the fans. The obtained parameters, such as SPL and the acoustic spectrum, are very important for evaluating noise cancellation methods and analyzing noise propagation. Additionally, acoustic imaging makes it very easy to assess the noise characteristics. Currently, in city centers, due to the available power at the connection point, EV chargers with power ratings up to 400 kW are installed, with most typically ranging from 60 to 180 kW. These are usually monolithic EV chargers (standalone type) (Figure 19a). Along highways, transit routes, city ring roads, and in some cities, EV chargers with a rated power of up to 600 kW are installed. These are typically split-type constructions with power ratings of 350, 400, or 600 kW (Figure 19b). In shopping centers, airport parking lots, and along main roads (where sufficient power is available at the connection point), high-power EV chargers, known as charging hubs, are installed. This type of EV charger is equipped with several user units, and the rated power of the power unit can exceed 1 MW (Figure 19c). Noise propagation conditions depend on many parameters [41,42], such as the acoustic power of the source, the frequency (source noise parameters), the number of noise sources, e.g., the number of charging stations, PU, UU, and environmental factors like air temperature, humidity, terrain, substrate, etc. However, the most important parameter is the distance from the noise source. The way in which the sound pressure level changes as a function of distance from the source is predictable and can be determined by knowing the sound pressure level $L\left(r_1\right)$ at a distance $r_1$ from the source. By adopting simplifying assumptions and assuming that sound is propagated from a point source in all directions equally, taking into account only the effect of distance on sound propagation and assuming the presence of a free field, the sound pressure level at the reception point ($L\left(r_2\right)$) can be determined by following equation:

$$\begin{array}{c}\hfill L\left(r_2\right)=L\left(r_1\right)-\left|20log\left({\displaystyle \frac{r_1}{r_2}}\right)\right|.\end{array}$$

(6)

By transforming above equation, one can determine the distance ($r_2$) from the sound source at which the required sound pressure level ($L\left(r_2\right)$) will be achieved (7) (Figure 20). Because sound pressure (p) decreases inversely proportional to distance ($p\sim 1/r$), it is easy to show () that the sound pressure level changes by 6 dB (Figure 20) with each doubling of distance from the sound source ($\Delta L_P\left(2r\right)$ = 6 dB).

$$\begin{array}{cc}\hfill r_2& =r_1{10}^{\left({\displaystyle \frac{|L\left(r_1\right)-L\left(r_2\right)|}{20}}\right)},\hfill \end{array}$$

(7)

$$\begin{array}{cc}\hfill \Delta L_p& =L\left(r_2\right)-L\left(r_1\right)⟶\Delta L_p=\left|10log\left({\displaystyle \frac{r_1}{r_2}}\right)\right|.\hfill \end{array}$$

(8)

For a simplified analysis and estimation of noise propagation, Equation (6) can be used. In order to determine the sound pressure level at a distance ($r_2$) from the sound source more precisely, i.e., to determine the sound propagation in an open space, the method described in the PN-EN-ISO 9613-2 standard [38] should be used and many additional factors should be taken into account, such as the influence of the atmosphere, distance, ground, and any terrain obstacles (9).

$$\begin{array}{c}\hfill L_P\left(R\right)=L_W+D_C-(A_{atm}+A_{div}+A_{bar}+A_{gr}),\end{array}$$

(9)

where

• $L_P\left(R\right)$—sound pressure level at a distance R from the sound source;
• $L_W$—sound power level of the source (4);
• $D_C$—correction for the solid angle of radiation emission and directional interaction (10);
• $A_{atm}$—the influence of air on sound attenuation (11);
• $A_{div}$—the effect of distance on sound attenuation (12);
• $A_{bar}$—the influence of screens and terrain obstacles on sound propagation (13);
• $A_{gr}$—the influence of screens and terrain obstacles on sound propagation (14).

$$\begin{array}{c}\hfill D_C=10log\left(1+{\displaystyle \frac{d_p^2+{(h_s-h_r)}^2}{d_p^2+{(h_s+h_r)}^2}}\right),\end{array}$$

(10)

where $d_p$—distance of the receiving point from the sound source, $h_s$—height of the sound source above the ground, and $h_r$—height of the receiving point above the ground.

$$\begin{array}{c}\hfill A_{atm}={\displaystyle \frac{\alpha d}{1000}},\end{array}$$

(11)

where $\alpha$—atmospheric attenuation coefficient and d—distance from the source.

$$\begin{array}{c}\hfill A_{div}=20log\left({\displaystyle \frac{d}{d_0}}\right)+B,\end{array}$$

(12)

where $d_0$—reference distance and B—constant resulting from the number of reflecting planes.

$$\begin{array}{c}\hfill A_{bar}=10log\left(3+{\displaystyle \frac{\lambda }{\delta }}\right),\end{array}$$

(13)

where $\lambda$—wavelength and $\delta$—coefficient resulting from the height of the obstacle.

$$\begin{array}{c}\hfill A_{gr}=4.8-\left({\displaystyle \frac{2h_m}{d}}\right)\left(17+\left({\displaystyle \frac{300}{d}}\right)\right).\end{array}$$

(14)

When analyzing noise emissions from several sources—several charging stations—e.g., from several monolithic chargers (Figure 19a) or several PUs or UUs, as in case shown in Figure 19b,c, the influence of each noise source should be considered separately.

Noise propagation in open space, as described above, depends on many factors (9). As can be seen from the sample results of in situ measurements performed with an acoustic camera (Figure 21), determining the contribution of the charging station to the noise generated in the environment under real-world conditions is difficult due to other noise sources commonly found in urban environments, such as electric vehicles, which are generally considered quiet. The sound pressure level (SPL) generated by them, e.g., by battery cooling systems, is often similar or even higher than that generated by the charging station, which, at a distance even of a few meters from the charging station, makes it very difficult to reliably assess the impact of this station on the resultant noise level at the measurement site, as shown in Figure 21. As can be seen, the SPL generated by a charging vehicle is 0.2 dB(A) higher than that emitted by the charging station.

In such conditions, even the measurement with an acoustic camera may be burdened with an error resulting from sound reflections from other sources that appear on the measurement surface. To confirm noise emissions from the charging station, a more advanced measurement—sound intensity—should be performed, taking into account the directionality of sound emission on a specific surface using a measurement system, e.g., Mezzo I-track imaging system. However, this type of measurement does not provide information about noise levels in areas farther away from the measured surface—specifically in locations where exposure occurs. An example of such a measurement is shown in Figure 22. The issue of assessing noise generated by charging stations in real-world environments is therefore complex and requires an individualized approach each time. Above all, it demands precise determination of the location where the exposure assessment is to be conducted. Regardless of the assessment method adopted—whether predicting noise levels at a specific location based on laboratory SPL measurements or performing actual environmental measurements—the authors believe it is essential to encourage charging station manufacturers to reduce noise emissions. This can be achieved by introducing appropriate noise emission limits, uniform assessment methods, and requirements for providing reliable information about the levels of generated sound power on nameplates or in the technical documentation of the charging station.

4. Noise Reduction Methods

Noise reduction in the environment can be achieved in two ways. The first is through administrative methods, such as legal regulations, ordinances, standards, and directives. The second is through technical methods, which can be further divided into two types: active and passive methods [43,44]. The method choice depends on the frequency of the sound source being reduced (Figure 23). In general, passive methods are more effective for higher frequencies, while, at low sound frequencies, passive methods are less effective compared to active methods. Additionally, active methods can be further classified into two levels: Level 1 (ANC L1) and Level 2 (ANC L2). By knowing the noise characteristics emitted by EV chargers (such as dominant frequency or frequencies, sound pressure level, acoustic power, etc.), it is possible to implement passive, active, or combined methods for noise reduction.

4.1. Passive Methods

Passive noise reduction methods can be implemented during the design and construction stages of EV charging stations. These methods include using materials that absorb acoustic energy (absorbers); selecting components (such as AC/DC converters, fans, etc.) with lower noise emissions to limit noise generation at the source; employing appropriate mechanical designs that enable sound energy to dissipate before being emitted into the environment (e.g., proper air ducting to absorb acoustic wave energy); and designing devices to minimize housing vibrations (e.g., optimizing housing stiffness and sheet metal thickness). The use of sound-dampening materials is one of the most popular passive noise reduction methods. The ability of a material to absorb sound energy is determined by its absorption coefficient ($\alpha$). The values of the absorption coefficient for exemplary materials are shown in

Table 5. Absorption coefficients of various types of materials.

Type of Material	Absorption Coefficient $\mathit{\alpha }$
	125 Hz	250 Hz	500 Hz	1 kHz	2 kHz	4 kHz
Air intake/outlet grille 50% clearance	0.30	0.50	0.50	0.50	0.50	0.50
Mineral wool thickness 100 mm	0.351	0.647	0.653	0.732	0.821	0.893
Mineral wool PT80 thickness 50 mm	0.20	0.60	1.00	0.90	0.85	0.85
Polyurethane acoustic foam 50 mm	0.30	0.65	0.95	0.92	0.98	0.98
Absorber Rolsound SGP VL80 30 mm	0.32	0.65	0.87	0.94	0.93	0.70

.

$$\begin{array}{c}\hfill \alpha ={\displaystyle \frac{E_i-E_r}{E_i}}={\displaystyle \frac{E_a}{E_i}},\end{array}$$

(15)

where $\alpha$—absorption coefficient ($0\dots 1$), $E_i$—incident energy, $E_r$—reflected energy, and $E_a$—absorbed energy.

As it is visible (Table 5), polyurethane acoustic foam has a high absorption coefficient in the frequency range occurring in the EV charging station (500–2000 Hz). Also, hybrid materials, e.g., polyurethane + rubber, as “Rolsound”, have sufficient absorption coefficients in range 0.5–2 kHz. The use of this type of material inside the charging station reduces sound wave reflections inside the housing and reduces noise emissions into the environment by 3–5 dB(A). Exemplary results of implementation of passive methods of noise reduction, such as dedicated internal construction and acoustic foams (absorbers) in air intake and air outlet chambers, in 180 kW (6 × 30 kW AC/DC 2-type) monolithic-type EV charger are shown in Figure 24. As shown in Table 5, polyurethane acoustic foam has a high absorption coefficient in the frequency range typically found in EV charging stations (500–2000 Hz). Hybrid materials, such as polyurethane combined with rubber (e.g., “Rolsound”), also exhibit good absorption coefficients in the 0.5–2 kHz range. The use of these materials inside the charging station helps to reduce sound wave reflections within the housing and lowers noise emissions into the environment by 3–5 dB(A). Exemplary results of implementing passive noise reduction methods, such as dedicated internal construction and the use of acoustic foams (absorbers) in the air intake and air outlet chambers, are shown in Figure 24 for a 180 kW (6 × 30 kW AC/DC_2) monolithic EV charger. The achieved noise reduction is over 9 dB(A) at the front side and approximately 5–6 dB(A) at the air intake sides. Another passive noise reduction method is the use of liquid cooling systems, which also help to reduce noise, especially systems with heat recovery and liquid–liquid-type heat exchangers. These systems enable the elimination of fans in the cooling systems.

4.2. Active Methods

4.2.1. ANC Level I

Active noise reduction Level 1 (ANC L1) methods involve appropriate control of AC/DC converters in order to reduce noise emissions and the generated sound pressure level. In the conventional approach, the charging power demand is met in two ways (Figure 25). Typically, the demand is fulfilled by the smallest sufficient number of converters, with each operating converter loaded to its maximum power. If the demand increases, an additional converter is connected. This method is called “sequential control”. In the second approach, the charging power demand is divided equally among all available AC/DC converters, with each converter loaded to the same extent. This method is called “simultaneous control”. As mentioned, the noise generated by the AC/DC converter fans depends on temperature inside the converter. An increase in charging power results in greater heat generation, which in turn increases the noise emitted by the fans. Research on various AC/DC converters has shown that reducing the charging power by just a few percent can decrease the intensity of fan operation and, consequently, reduce the emitted noise. The introduced AC/DC converter control method consists of limiting the charging power ($\Delta P_{ac/de\_lim}$) of each converter by approximately 15% ($\Delta P_{ac/de\_lim}\approx 4.5$ kW) in both scenarios—in “sequential control” and in “simultaneous control” (Figure 25). In Figure 26, results of sound pressure level measurements are shown for both control methods with reduced power—simultaneous and sequential. The tests were carried out on an EV charger with a rated power of 360 kW, equipped with 12 converters (12 × 30 kW) of the AC/DC_2 type (see Table 4). To compare with conventional control method, obtained results are compiled with value of SPL measured for 300 kW output power PDC (10 × 30 kW). As can be seen, for lower value of charging power (PDC), better results (lower value of SPL) are obtained for sequential control method. For higher value of charging power (PDC), better results (lower value of SPL) are obtained for simultaneous control method.

4.2.2. ANC Level II

Active noise canceling Level 2 (ANC L2) methods are methods of reducing existing noise by introducing controllable secondary noise sources that influence the production, radiation, transmission, and reception of the original primary noise source. The theoretical foundations and methods of L2 ANC have already been well known and well established for the past 30 years [44,45]. However, successful industrial and civilian applications of this technology are still limited to some specific cases, such as headsets, aircraft, helicopters, and cars. Full use of ANC L2 methods requires the use of acoustic channels (waveguides) in which noise reduction occurs by summing noise from main and secondary sources. This in turn entails the need to increase the dimensions of charging stations, which is at odds with the current trend of reducing the dimensions of charging stations while increasing installed power—increasing the volumetric power density.

5. Conclusions

Electric vehicle charging infrastructure is increasingly integrating into our surroundings, and intensive cooling systems—especially at Fast Charging stations—can generate high noise levels. This article addressed the noise issues and requirements associated with electric vehicle charging infrastructure, particularly fast DC charging stations. Through the identification and analysis of the noise sources within charging stations, it was determined that the main source of noise comes from the cooling systems of AC/DC power electronic converters. Depending on the mechanical construction and the nominal power of the converter (volumetric power density), the noise emitted by a single converter can reach up to 75 dB(A). In the case of a charging station with 10 such converters, the total sound pressure level can exceed 85 dB(A). Additional sources of noise in Fast Charging stations include liquid cooling systems for HPC cables and connectors, with noise levels that can exceed 65 dB(A). A key feature of these charging stations is the high acoustic directionality of the emitted noise. This means that the level of noise varies depending on the side from which it is emitted, which may be important when installing stations in highly urbanized areas. In addition to identifying and characterizing these noise sources, the article explored both passive and active methods for reducing the noise emitted by Fast Charging stations. The presented research findings indicate that identifying noise sources in charging stations, selecting appropriate components, and implementing noise reduction methods, such as employing appropriate internal station structures, acoustic absorbers, and control algorithms for converters, reduce noise emissions and mitigate their environmental impact. As the research results have shown, noise can be reduced by up to 15%. For a 360 kW charging station, this results in a noise reduction from 61 dB(A) to 51.8 dB(A) at a distance of 1 m from the station. Taking this into account, one of the primary goals of electromobility development is to reduce the noise generated by the transport sector, particularly in urban environments. However, the considerations and research results presented in this article indicate that, without a proper approach to the design and construction of Fast Charging systems, the intended effects of electromobility development in this area could have the opposite result. Although electric vehicles are known for lower noise emissions compared to combustion vehicles, their Fast Charging infrastructure can become a significant source of noise in urban settings. Therefore, further research is needed to develop methods and procedures for assessing noise levels during the implementation of charging stations and hubs, as well as strategies for noise reduction. Developing new, advanced AC/DC converter control algorithms appears to be a promising direction for achieving this goal. Another intriguing avenue for further development would be the implementation of control strategies based on real-time measurements of the noise generated by the charging station. IoT (Internet of Things) and AI (Artificial Intelligence) technologies appear to be particularly interesting in this context. Their implementation potential should be analyzed further in future studies. Moreover, despite the technical challenges associated with such implementation, research into active noise reduction methods of the second level (ANC Level 2) should not be abandoned as these methods remain the most effective among those described.