Selection and Classification of Small Wind Turbines for Local Energy Systems: Balancing Efficiency, Climate Conditions, and User Comfort

Abstract

Micro and small wind turbines (MAWTs) are increasingly integrated into residential and prosumer hybrid energy systems. However, their real-world performance often falls short of catalog specifications due to mismatched wind resources, siting limitations, and insufficient attention to human comfort. This paper presents a comprehensive decision-support framework for selecting the type and scale of MAWTs under actual local conditions. The energy assessment module combines aerodynamic performance scaling, wind speed-frequency modeling based on Weibull distributions, turbulence intensity adjustments, and component-level efficiency factors for both horizontal and vertical axis turbines. The framework addresses three key design objectives: efficiency—aligning turbine geometry and control strategies with local wind regimes to maximize energy yield; comfort—evaluating candidate designs for noise emissions, shadow flicker, and visual impact near buildings; and climate adaptation—linking turbine siting, hub height, and rotor type to terrain roughness, turbulence, and built environment characteristics. Case studies from low and moderate wind locations in Central Europe demonstrate how multi-criteria filtering avoids oversizing, improves the autonomy of hybrid PV–wind systems, and identifies configurations that may exceed permissible limits for noise or flicker. The proposed methodology enables evidence-based deployment of MAWTs in decentralized energy systems that balance technical performance, resilience, and occupant well-being.

1. Introduction

Decarbonization in the building sector, particularly in the context of local and distributed energy systems, increasingly relies on integrated renewable sources that can operate independently of centralized grids. Although rooftop photovoltaics (PV) dominate residential applications, their strong diurnal and seasonal variability and lack of nighttime production limit the efficiency of standalone systems. In this context, small and micro wind turbines (MAWTs), with capacities up to approximately 20 kW, can be a valuable complement to PV—provided they are properly selected to suit local wind resources, spatial conditions, and user comfort requirements.

In practice, however, many MAWT installations fail to meet energy and social expectations. This is often due to a lack of thorough analysis of local wind speed distributions, a failure to consider terrain roughness, turbulence intensity, or occupant comfort criteria. This is particularly important in micro-scale locations—such as single-family housing estates—where airflow is disturbed by buildings and vegetation, and an incorrectly selected turbine can cause acoustic and visual disturbances [1].

Standard methods for estimating annual energy yield (AEP) are based on average values and simulations for open areas—which do not reflect conditions found in the built environment. A better fit is possible by using wind speed-frequency histograms, Weibull distributions, and corrections for turbulence intensity (TI)—especially when only short measurement campaigns are available [2,3]. This paper adopts this approach, developing a scalable model for assessing AEP and selecting MAWT technologies under conditions of limited input data availability.

Another challenge is the diversity of MAWT topologies. In the <20 kW class, there are both horizontal axis wind turbines (HAWT) and vertical axis wind turbines (VAWT [4,5]: Darrieus, Savonius) [6,7,8], duct and diffuser turbines (DAWT, WLT) [9,10,11], cascade systems (CAWT) [12,13], and building-integrated turbines (BIWT) [14,15,16]. Each type differs in terms of start-up characteristics, sensitivity to turbulence, noise emissions, vibrations, and visual impact—and these factors play a key role in the context of dense development [1,17].

There is growing interest in the use of MAWTs in distributed building systems in the scientific literature. Research focuses on adaptation to turbulent urban flows [18,19], optimization of MAWT geometry in confined spaces [14,15], and modeling acoustic–vibrational interactions [1,17]. Results suggest that some topologies—such as CAWTs and WLTs—are characterized by better resilience to local flow disturbances and a lower start-up threshold, but require precise adaptation to the site’s geometric data.

An important aspect is the integration of MAWTs with other building energy system components, such as PV, solar collectors, heat pumps, and energy storage. Hybrid systems [20] increasingly utilize predictive and fuzzy logic, which enables synchronization of energy production profiles with thermal and electrical demand profiles [20]. MAWTs can then significantly improve seasonal balance—especially in winter and at night [17,21].

Importantly, user comfort and social acceptance often play a greater role than energy yield alone. Research indicates that tonal and low-frequency noise, structural vibrations, and shadow flicker significantly influence residents’ opinions. It is recommended to incorporate aeroacoustic models [22,23] and urban analysis (neighbors’ location relative to the turbine) into the design phase—something that is still rarely comprehensively considered in the literature [1,17].

Despite progress, a simple, scalable, and practical approach is still lacking, one that—with a minimal dataset (histograms, TI, turbine geometry)—would allow for the simultaneous assessment of efficiency, comfort, and environmental resilience. In this paper, the authors propose an approach that considers three axes of analysis:

• Efficiency—maximizing energy yield in turbulent conditions using real-world wind data, taking into account TI and terrain features;
• Comfort—reducing noise, vibration, and flicker, improving social acceptance and aesthetics;
• Climate Adaptation—the resilience of MAWTs to climate change, extreme weather events, increasing urban density, and evolving local conditions.

The weights assigned in the MAWT-Score were selected through a combination of literature review, expert elicitation, and preliminary case-study modeling, ensuring balanced representation of aerodynamic performance, environmental impacts, and user acceptance. To ensure robustness, a sensitivity analysis of these weights is presented in Section 3.5, demonstrating how variations in weighting affect the ranking outcomes.

The scope of the analysis was limited to small turbines with a capacity of less than 20 kW, compliant with the IEC61400-2 standard [24], intended for prosumers and residential buildings. Typical mounting heights, spatial constraints, user needs, and regulatory compliance were taken into account. The result is a standardized method for MAWT evaluation and selection, enabling the selection of solutions most suited to specific locations, wind resources, and user expectations.

2. Material and Methods

This chapter presents the basics of wind resource analysis, a typology of MAWT microturbines, characteristics of typical installation locations, and the adopted multi-criteria design evaluation criteria. This chapter provides the analytical framework for the remainder of the article, which compares different types of MAWTs in terms of energy efficiency, integration with the environment, and adaptation to changing climatic conditions.

2.1. Meteorological Data and Wind Analysis

Wind is generated by differences in atmospheric pressure caused by uneven heating of the Earth’s surface by solar radiation. Its distribution and direction are also influenced by the Coriolis force, local terrain roughness, and orography. The kinetic power of the airflow impinging on the turbine rotor can be expressed as

$$P\left(v\right)={\displaystyle \frac{1}{2}}·\rho ·A·C_p\left(v\right)·v^3$$

(1)

where P(v)—instantaneous wind power [W]; ρ—air density [kg/m³], usually assumed ρ = 1.225 [kg/m³]; A—area swept by the rotor [m²]; Cp(v)—power coefficient, depending on the turbine power and its operating parameters; v—wind speed [m/s].

A key element in this formula is the dependence of power on the third power of wind speed (v³). This means that the instantaneous power available in the wind increases very rapidly with increasing speed. If the wind speed doubles (e.g., from 5 m/s to 10 m/s), the power increases 2³ = 8 times. If the wind speed triples (e.g., from 5 m/s to 15 m/s), the power increases 3³ = 27 times. This nonlinear relationship is fundamental to wind energy. Even a small increase in wind speed has a huge impact on the amount of energy harvested. Similarly, a calculation error, such as a small difference in wind speed, can lead to significant differences in the estimated available energy. For example, an error of 1 m/s at a speed of 6 m/s results in a power difference of around 40%. Based on multi-year measurement campaigns in Central Europe, the coefficient of variation CV of hourly wind speeds has been found to be in the range of 0.47–0.55. Daily components introduce fluctuations of ±1 m/s, while synoptic phenomena can maintain extreme conditions for several days. Therefore, using a single annual mean value can lead to significant errors when planning installations.

A practical tool for describing wind variability is the Weibull distribution. It is extremely useful because it allows for describing wind characteristics at a given location using two simple parameters. The Weibull distribution is described by the formula

$$f\left(v;k,c\right)={\displaystyle \frac{k}{c}}{\left({\displaystyle \frac{v}{c}}\right)}^{k-1}·exp\left(-{\left({\displaystyle \frac{v}{c}}\right)}^k\right),\hspace{1em}\hspace{1em}v>0$$

(2)

where

f(v)—The probability density function for wind speed v.

k—A shape parameter (typically 1.6–2.6) that describes wind stability. Low k values (below 2) indicate a wide dispersion of wind speeds, with frequent periods of gusty and light winds. The wind is then variable and less predictable. High k values (above 5–10) indicate stable winds, with speeds close to average, which favors regular turbine operation.

C—A scale parameter (Poland: 4–7 m/s in lowland areas) that reflects the “average wind speed level.” The higher its value, the higher the overall wind potential of a given location, which translates into more energy that can be generated by the turbine.

Air density was treated as ρ = 1.225 kg·m⁻³ (15 °C, 101.3 kPa) for all AEP integrations unless noted otherwise. For measured sites (B2P, B4P), a ±2% sensitivity to seasonal variability of ρ was tested and found to have a minor effect relative to wind speed uncertainty.

In the present study, the Weibull parameters (k,c) for each scenario were estimated from measured wind speed histograms using the maximum likelihood method, following standard approaches described in [25,26]. The goodness-of-fit between the measured and modeled Weibull distributions was quantified using the Kolmogorov–Smirnov (KS) and Anderson–Darling (AD) statistics [27,28]. For each parameter set (k,c), 95% confidence intervals were calculated using a non-parametric bootstrap with 10,000 resamples, ensuring robustness in the presence of skewed wind speed distributions. These metrics are presented in Figure 1 to complement the visual comparison.

The fitting was performed on hourly wind speed histograms covering a continuous 12-month period for measured datasets (B2P, B4P) and on hourly modeled series for the remaining locations.

All computations, including parameter estimation, KS/AD statistics, confidence interval calculation, and annual energy production (AEP) integration, were performed in Python 3.11.5 using the NumPy 1.26.0, SciPy 1.11.3, Pandas 2.1.1, and Matplotlib 3.8.0 libraries, with custom scripts for power-curve rescaling and Monte Carlo uncertainty propagation. This ensures reproducibility and transparency of the modeling workflow.

The Weibull distribution is commonly used in annual energy production (AEP) calculations, yield forecasts, and for fitting meteorological data. The distribution curve is most often fitted using the method of moments, the maximum likelihood method (MLE), or graphically. These parameters allow for easy assessment of the suitability of a given location for wind turbine installation.

In practice, the turbine’s mounting height significantly affects its energy yield. Wind speed scaling to the required height is achieved based on a logarithmic velocity profile:

$$v\left(h\right)=v_{ref}·{\displaystyle \frac{ln\left(\frac{h}{\alpha }\right)}{ln\left(\frac{h_{ref}}{\alpha }\right)}}$$

(3)

or power profile

$$v\left(h\right)=v_{ref}·\left({\displaystyle \frac{h}{h_{ref}}}\right)$$

(4)

where

v(h)—wind speed at altitude h;

v_ref—speed at reference altitude (e.g., for h_ref = 10 m);

h—target altitude at which we want to calculate the speed;

h_ref—reference height for which we know the wind speed;

α—Hellmann exponent depending on the type of terrain, usually 0.14 for suburban areas.

Calculation example for v₁₀ = 5 m/s, h = 11 m: v₁₁ ≈ 5.07 m/s; Δv ≈ 1.4%, ΔP ≈ 4.08%

Just 1 m of height difference can result in a 4% increase in power, which is significant given the low MAWT scale.

Planning and installing a wind turbine is a process that requires careful analysis and assessment of many factors, the key being the wind resource. A correct assessment of the energy potential at a given location is the foundation upon which the entire investment decision rests. This process is not a one-time measurement, but a four-step process leading from general understanding to detailed modeling, thus minimizing the project’s financial and technical risks. Although national wind atlases (e.g., with a resolution of 100–250 m) allow for a preliminary assessment of energy potential, in urban areas or areas with complex orography, they can significantly underestimate actual wind speeds—by as much as 10–30% [1]. Obstacles such as trees or buildings cause local speed drops of up to 40% at a 2D distance behind the obstacle. At the same time, turbulence intensity (TI) can exceed 20%.

The first step in the assessment process is macro-analysis, identifying potentially favorable locations based on regional data. Engineers use wind atlases, which are a compilation of long-term meteorological data or computer simulation results. These data, averaged over large areas, allows for the creation of a general wind speed map for a given region. Although such an analysis does not take into account local terrain obstacles, it allows for the quick selection of the most promising sites and the rejection of those that are unprofitable due to low wind speeds. This initial selection is extremely cost-effective and allows for the focus of further efforts on areas with the highest potential [1].

After the initial selection, it is necessary to collect precise data at the selected location. This stage involves in situ measurements, or “on-site”, to verify and refine the data from the atlases. Traditional anemometric masts, equipped with wind speed and direction sensors at various heights, are most often used. Measurements are conducted for at least 12 months to capture the full seasonal cycle of wind fluctuations. In recent years, modern, non-invasive technologies such as LIDAR and SoDAR have become increasingly important. These devices, which utilize laser beams and sound waves, allow for precise wind profile measurements without the need for tall masts, providing greater flexibility and mobility [2].

The collected measurement data becomes the basis for detailed statistical analysis. For this purpose, engineers use the Weibull distribution, a mathematical model that perfectly describes the statistical distribution of wind speed. This distribution is characterized by two parameters: the scale parameter (λ), related to the average wind speed, and the shape parameter k, which describes wind stability. After fitting the distribution to the collected data, it can be combined with the power curve of the selected turbine, i.e., a graph of the power versus wind speed relationship. This combination allows for a precise estimation of the AEP (annual energy production), which is a key parameter for assessing the project’s profitability [1].

The final, most advanced stage is CFD (Computational Fluid Dynamics) modeling. This technique utilizes advanced computer simulations to model airflow over complex terrain. Unlike simplified models, CFD takes into account detailed topography, terrain obstacles, and the influence of neighboring turbines. This modeling allows for precise determination of wind speed at the hub height of each planned turbine and identification of areas of increased turbulence. The use of CFD allows for optimal turbine layout within a wind farm, minimizing energy losses caused by aerodynamic shading (the so-called wake effect) while simultaneously increasing operational safety [29].

2.2. Wind Resource Classification for MAWTs

Unlike large wind turbines, for which standards such as IEC 61400-1 [30] clearly define wind resource classes (e.g., Class I—10 m/s, Class II—8.5 m/s, Class III—7.5 m/s at 10 m height), for small wind turbines (MAWTs—micro and small wind turbines), there is currently no strict, universally accepted wind resource classification based on the annual mean wind speed Vavg. The IEC 61400-2 standard [24]—design requirements for small wind turbines—covers turbines with a capacity of up to 50 kW and a rotor swept area not exceeding 200 m², but it focuses primarily on structural and safety requirements, not offering a similar division into wind resource classes as the IEC 61400-1 standard [30]. Therefore, several informal classification approaches have been proposed in the technical and scientific literature, dividing wind into classes based on its average speed [31]. Attempts have also been made to classify wind based on the use of building-integrated wind turbines (BIWTs). The authors indicated that sensible operation of these devices begins at an average wind speed above 4.5 m/s, while optimal performance is achieved at average annual speeds exceeding 6.0–6.5 m/s [31]. It is worth noting that even classifications originally developed for large wind turbines, such as those used by Global Wind Atlas, NREL, FINO, and NASA PowerLab, are increasingly being adapted to MAWTs in scientific publications.

This paper proposes a proprietary, functional classification of wind resources for MAWTs, which takes into account not only technical aspects (average speed, energy yields) but also urban conditions and operational requirements. The proposed classification is based on the following:

-

Analysis of actual annual energy production from 2 kW MAWTs, based on meteorological data from locations such as Gdynia Redłowo.

-

Operational requirements of various turbine types (e.g., cut-in values, rated values, power characteristics, aerodynamic efficiency).

-

Urban and landscape context (compact development, rooftops, suburban areas, open spaces).

Table 1. Proposed wind resource classification for MAWTs.

Class	Average Annual Velocity Vavg [m/s]	Typical Locations	Example of MAWT Application	Expected AEP for a 2 kW Turbine [kWh/Year]
A—high	>6.0 m/s	Open areas, coastlines, high roofs	MAWT HAWT/VAWT in independent installations	>3000
B—moderate	4.5–6.0 m/s	Suburbs, mid-rise roofs	MAWT CAWT, DAWT, BIWT with buffer	1500–3000
C—low	3.0–4.5 m/s	Urban roofs, compact development	MAWT WLT, Savonius, integrated systems	<1500
D—marginal	<3.0 m/s	Deep downtown development	Rarely profitable—only educationally	<500

presents the proposed wind resource classification for MAWTs.

The proposed classification is practical and can serve as a decision-making tool for designers, investors, and end users. It combines realistic energy yields with the technical capabilities of various turbine types, as well as spatial and social constraints. It can be used for, among other things, the following:

-

Preliminary site selection for MAWTs;

-

Turbine type selection (e.g., VAWT instead of HAWT in a high turbulence environment);

-

Estimated annual energy production (AEP) without the need for long-term measurements;

-

Economic cost-effectiveness assessment (e.g., LCOE—levelized cost of energy) in the context of low and medium wind speeds.

While the reference case and validation in this study are based on Central European conditions (terrain roughness class B4, neutral stability climatology, and moderate turbulence intensity bands as in Table 5), the framework can be applied to other regions by retuning these inputs. Outside Central Europe, site-specific roughness length z₀ (aerodynamic roughness length—a parameter describing the effect of surface obstacles on the vertical wind speed profile), seasonal and diurnal stability regimes, and turbulence intensity classes should be derived from local meteorological datasets or in situ measurements and substituted directly into the calculation chain. This ensures that predicted energy yield and comfort impacts remain representative of the target location.

2.3. Typology and Characteristics of MAWTs

Micro and small wind turbines (MAWTs) are a diverse group of devices that convert wind kinetic energy into electricity at relatively low unit power—typically up to 20 kW. Their primary applications are local systems: residential installations, powering autonomous facilities, supporting microgrids, and integrating them with buildings in urban environments (building-integrated wind turbines, BIWTs). The design diversity of MAWTs translates into significant differences in aerodynamic, acoustic, operational, and environmental characteristics, making their classification an essential element of both design and comparative analysis. MAWTs can be classified based on many criteria, the most common of which are as follows:

-

Rotor rotation axis: horizontal (HAWT) or vertical (VAWT);

-

Rotor type: classic, shrouded, with diffuser, multi-rotor;

-

Blade geometry: curved, straight, helical, composite blades, flexible;

-

Mounting type: freestanding, mast-mounted, roof-mounted, building-integrated;

-

Operating mode: off-grid, hybrid, with energy storage, or direct energy consumption.

Additionally, technical parameters such as nominal power, starting speed (cut-in), power factor (Cp), operating range (nominal speed, rated), efficiency at low wind speeds, acoustic emission, and environmental impact (vibration, safety, aesthetics) are taken into account. Depending on their intended use, MAWTs are also classified based on their driving mechanism (lift or drag), the method of flow acceleration (e.g., diffusers) and the degree of integration with the building, which is particularly important in urban environments.

2.3.1. Horizontal Axis Turbines (HAWT)

Horizontal axis wind turbines (HAWTs) are the most common type of small wind turbine (MAWT), whose design is directly modeled on large wind turbines. They are characterized by a rotor, usually with two or three blades, that is constantly facing the wind. High aerodynamic efficiency is their main advantage; the maximum power coefficient (Cp) can reach values between 0.45 and 0.50, meaning they are able to convert wind kinetic energy very efficiently.

To maintain optimal rotor alignment, these turbines require special mechanisms. Typically, this involves passive control using a directional tail or active control using electronic yaw systems. Furthermore, to operate properly, they require a large, clear space both in front of and behind the rotor to ensure a clean airflow zone. Their greatest advantage is their very high efficiency in stable, moderate wind speeds (5–9 m/s), which translates into a relatively low cost per kWh of energy. This makes them an ideal choice for locations with predictable wind conditions.

Unfortunately, they also have certain drawbacks that limit their application. They are intolerant of wind direction variations and turbulence, which are typical of urban environments, where wind often bounces off buildings. These variable loads generate high mechanical loads and vertical bending moments, requiring a rigid and durable structure. Furthermore, at high speeds, they can generate high levels of tonal noise, which can be problematic near residential buildings.

Due to their characteristics, HAWTs are best suited for open areas, rural locations, or locations far from obstacles that could disrupt wind flow.

2.3.2. Vertical Axis Turbines (VAWT)

Vertical axis wind turbines (VAWTs) are an alternative type of turbine that rotates around a vertical axis. Their key advantage is that they do not require wind alignment, making them ideal for operation in conditions of variable turbulence. VAWTs are divided into two main categories: Darrieus and Savonius.

Darrieus turbines are characterized by a design resembling an inverted “Ω” or spiral, often referred to as H-Darrieus or H-rotor. Their symmetrical blades operate on the principle of lift, similar to an airplane wing. They achieve relatively high efficiency (power coefficient Cp up to 0.35), but starting can be difficult and often requires the use of a starter motor. They are highly valued for their quiet operation and aesthetic design, making them attractive for installation in urban environments. However, cyclic stresses and material fatigue can lead to damage. Their main disadvantages are the need for blade buckling protection and lower energy efficiency compared to HAWTs.

Savonius turbines are distinguished by their simple design, in which the rotor consists of two or more curved cups. They operate on the principle of aerodynamic drag rather than lift, resulting in lower efficiency (Cp of 0.15–0.25). Their great advantage is their excellent start-up ability at very low wind speeds (below 2 m/s), making them independent of starter motors. They are also extremely quiet and stable. Their simple design and low cost make them easy to install. Due to their low efficiency, they are typically intended for educational, decorative, or low-power applications, where reliability and simplicity are more important than high efficiency.

Thanks to their tolerance for turbulence and operation with variable wind directions, VAWTs are ideal for installation on rooftops, in suburban educational farms, and in off-grid systems where access to stable wind is limited.

2.3.3. Diffuser-Augmented Wind Turbines (DAWT)

Diffuser-Augmented Wind Turbines (DAWTs) are an innovative type of turbine that utilizes a Venturi nozzle to increase local wind speed. Specially designed housing directs the airflow directly onto the rotor blades, significantly increasing its speed. Thanks to this solution, DAWTs can generate significantly more power (up to 2–4 times more energy) compared to conventional turbines with the same rotor area and at the same wind speed.

Their main advantage is the ability to miniaturize the turbine while maintaining high energy yield. This allows them to be installed in locations where large rotors would be impractical. Furthermore, the Venturi nozzle allows for better airflow control, which can contribute to noise reduction.

Unfortunately, DAWTs also have their drawbacks. Their design is typically heavier and more complex than that of traditional turbines, which increases production costs. The nozzle is also sensitive to dirt and duct resistance, which can reduce its efficiency. An additional challenge is the problem of cooling the generator, which, being enclosed in an enclosure, requires a suitable heat dissipation system.

Due to their specific characteristics, DAWTs are often used for integration into building facades, on university campuses, and as demonstration facilities. In these locations, where a compact form and higher efficiency within a given footprint are key, their unique design performs best.

2.3.4. Compact and Multi-Rotor Turbines (CAWT, WLT)

Among small wind turbines (MAWTs), a distinct category of innovative systems emphasizes compact design and architectural integration. This group includes CAWTs (Compact Axis Wind Turbines) and WLTs (Wall-mounted or Linear Turbines).

CAWTs are compact devices, often equipped with safety shields, designed for installation on balconies, facades, or other building elements. WLTs, on the other hand, are often mounted in linear arrays on roof edges, creating an aesthetically pleasing and functional element.

The main advantage of these solutions is their aesthetics and ease of installation, which translates into good social acceptance. Thanks to their compactness and location, they have minimal impact on the surrounding area.

Unfortunately, innovative turbines also have their drawbacks. They typically offer a low specific power, ranging from 200 to 500 W. Due to their low efficiency, the price per kWh of energy produced is relatively high. For this reason, their use is often more demonstrative or supplementary than intended to fully cover energy demand.

2.3.5. Building-Integrated Turbines (BIWT)

Building-integrated wind turbines (BIWTs) are a special category of devices designed from the outset with full architectural integration in mind. Instead of being separate, freestanding structures, they become an integral element of the façade, roof, or curtain wall. These solutions are particularly popular in modern sustainable construction and zero-energy building designs.

Their key advantage is reduced installation and infrastructure costs, as they do not require a separate tower or foundation. Additionally, they serve important demonstration and prestige purposes, making them attractive in university buildings, research centers, or the headquarters of innovation-focused companies. Their presence also has educational potential, raising awareness of renewable energy sources.

The main drawback of BIWTs is their strong dependence on architecture and variable wind turbulence in urban environments. Airflow around buildings is highly complex and unpredictable, leading to low energy yield predictability. Therefore, their design and optimization require advanced techniques, such as CFD (Computational Fluid Dynamics) modeling, to accurately estimate performance and avoid potential problems.

Table 2. (a) Technical parameters of different MAWT typologies, including aerodynamic class, axis of rotation, tip-speed ratio (TSR), maximum power coefficient (C_pmax), characteristic wind speeds (cut-in and rated), and turbulence tolerance. Data compiled from technical literature and manufacturer datasheets. (b) Functional and contextual characteristics of different MAWT typologies, covering noise levels, installation modes, aesthetics, integration potential, educational value, and indicative economic factors (Capex index and Installation complexity). Values synthesized from case studies, manufacturer information, and peer-reviewed literature.

(a)
Turbine Type	Aerodynamics	Axis of Rotation	TSR	Cpmax	vci [m/s]	vr [m/s]	Tolerance TI
HAWT 3-blade	Load-bearing	Horizontal	6–8	0.35–0.40	3	10–12	Low–Medium
HAWT multi-blade	Resistance	Horizontal	~1	0.25	1.5–2.5	8–10	High
VAWT—Darrieus	Load-bearing	Vertical	3–5	0.28–0.32	3–5	10–12	Medium
VAWT—Helical	Load-bearing	Vertical	2–5	0.26–0.30	2.5–4	9–11	Medium–High
VAWT—Savonius	Resistance	Vertical	<1	0.12–0.18	1–2	6–8	Very High
VAWT—Hybrid	Mixed	Vertical	1–3	0.20–0.30	1.5–3	8–10	High
DAWT	Diffuser	Horizontal	4–6	0.42–0.48	2	8–10	Medium–High
CAWT	Transverse	Horizontal	2–3	0.25–0.30	2	8–10	High
WLT (lenticular)	Load-bearing + lens	Horizontal	4–6	0.45–0.50	2	8–11	High
BIWT	Integrated	Any	1–4	0.28–0.35	2–5	Depending on location	Very High
(b)
Turbine Type	Noise LAeq [dB(A) @ 10 m]	Installation	Aesthetics	Integration	Educational Potential	Capex Index (1–5)	Installation Complexity
HAWT 3-blade	50–60 @50 m	Mast	Moderate	Low	Medium	4	High
HAWT multi-blade	~60	Mast/pumps	Low	Low	Medium	3	Medium
VAWT—Darrieus	42–48 @15 m	Low Mast	High	Medium	High	3	Medium
VAWT—Helical	<45	Roof/Mast	High	Medium	High	3	Medium
VAWT—Savonius	~38	Roof/floor	Medium	High	Very High	2	Low
VAWT—Hybrid	Different	Different	High	Medium	High	3	Medium
DAWT	~50	Mast or Roof	Low	Medium	Medium	5	High
CAWT	<45	Roof	High	High	High	3	Medium
WLT (lenticular)	46–52	Mast/Roof	Moderate	Medium	Medium	4	High
BIWT	Niesłyszalny	Elevation/Roof	High	Very High	Very High	4	Medium

a,b provide a structured comparison of the most popular types of micro and small wind turbines (MAWTs). To improve readability and thematic clarity, the characteristics were divided into two complementary views: Table 2a presents the technical and aerodynamic parameters (aerodynamics, axis of rotation, TSR, C_pmax, cut-in and rated wind speeds, turbulence tolerance), while Table 2b focuses on environmental, integration, and cost-related factors (noise, installation, aesthetics, integration potential, educational applications, as well as indicative Capex index and Installation complexity). This dual perspective facilitates a balanced assessment of MAWT typologies, supporting both engineering design decisions and broader considerations such as user comfort and economic feasibility. The values are indicative, synthesized from peer-reviewed literature, manufacturers’ data and case reports.

Noise values reported in Table 2b correspond to the A-weighted equivalent continuous sound level (LAeq), measured or estimated at a reference distance of 10 m from the rotor, unless otherwise indicated. This choice was made to ensure comparability between turbine types, as maximum instantaneous sound pressure levels (Lmax) are not consistently reported in the literature (Lmax denotes the peak noise level reached during short-term fluctuations, in contrast to LAeq, which represents the time-averaged level).

Table 2a,b jointly summarize the most important parameters of various types of small wind turbines (MAWTs), covering both technical performance and functional integration aspects. The analysis includes classic designs such as HAWTs and VAWTs, as well as innovative, architecturally integrated concepts such as BIWTs. Each type was characterized against multiple criteria that are crucial for energy efficiency, adaptability to local conditions, and suitability for hybrid systems.

Table 2a focuses on aerodynamic and operational parameters. The columns Aerodynamics and Axis of Rotation classify turbines by their drive principle and rotor orientation. Typical three-blade HAWTs operate on the lift principle with a tip-speed ratio (TSR) of 6–8, which results in high efficiency but also increased sensitivity to turbulence and noise. Turbine efficiency is described by TSR, maximum power coefficient (C_pmax), cut-in wind speed (v_ci), and rated wind speed (v_r). Advanced concepts such as DAWTs and lenticular WLTs achieve the highest efficiency (above 0.40), outperforming traditional HAWTs. By contrast, VAWT—Savonius and VAWT—Hybrid, despite lower efficiencies, operate at very low cut-in speeds (1–2 m/s), making them suitable for locations with weak winds. The Tolerance TI column illustrates the resilience of each design to turbulence, where resistance turbines (VAWT—Savonius, HAWT multi-blade) and BIWTs show the highest stability.

Table 2b complements this technical view with functional, social, and economic characteristics. The Noise column highlights that low-mounted VAWT—Savonius (≈38 dB(A)) and BIWTs (sub-audible) are the quietest, whereas classic HAWTs reach 50–60 dB(A). Installation options (Mast, Roof, Facade) and Aesthetics ratings indicate that architecturally integrated designs such as CAWTs and BIWTs stand out in urban contexts, while conventional mast-mounted HAWTs are less suitable. Integration and Educational potential emphasize the role of MAWTs in demonstration projects and learning environments, particularly VAWT—Savonius and VAWT—Hybrid, valued for their simplicity and safety.

Finally, the additional economic factors—Capex index and Installation complexity—provide an indicative comparison across typologies, synthesized from catalogs, case studies, and peer-reviewed literature. These columns highlight, for example, the relatively low cost and simple installation of VAWT—Savonius versus the higher capital intensity and complexity of DAWTs and HAWTs.

The Capex index (1–5) and Installation complexity (Low/Medium/High) values are indicative and harmonized across comparable MAWT sizes. They were compiled from manufacturer datasheets, peer-reviewed studies [4,31,32], and case reports. Actual project-specific costs and complexity depend on site conditions, mounting methods, and regulatory/permitting requirements.

The values reported in Table 2a,b are representative ranges derived from standard references, peer-reviewed literature, and case studies, rather than from proprietary manufacturer specifications. For example, HAWT three-blade TSR (6–8) and Cpmax (0.35–0.40) ranges follow IEC 61400-2 [24] and established handbooks [32,33]. Savonius and Darrieus values reflect experimental and numerical studies [7,8,12,13,16]. DAWT and CAWT values derive from validated CFD and tunnel tests [9,10,11]. Reported noise levels are consistent with acoustic studies of small wind [34,35]. Economic indicators (Capex index, Installation complexity) were harmonized across typologies using manufacturer datasheets and case-based studies [18,31]. Together, these references ensure transparency and reproducibility of the classification framework.

Taken together, the two tables offer a comprehensive comparison of MAWT typologies, enabling the selection of designs that balance efficiency, environmental compatibility, cost, and user acceptance.

2.4. Location and Architectural Context

Selecting the appropriate type of MAWT requires consideration of the spatial, location, and building context. These parameters influence not only energy efficiency but also operational safety, operational stability, and social acceptance.

MAWTs can be installed in four basic spatial contexts:

• Open areas—Sparse tree cover, minimal development, good wind exposure (e.g., rural areas, coastlines);
• Suburban areas and outskirts—Moderate exposure, presence of individual buildings, moderate turbulence;
• Urban areas—High development, high turbulence intensity, need for integration with buildings;
• Industrial and campus areas—Possibility of installing large masts, good technical conditions, moderate aesthetic requirements.

Location, assembly and legal framework

Choosing a location for MAWTs (micro and small wind turbines) involves not only optimizing aerodynamic conditions, but also the legal framework and local building regulations. In Poland, the installation of small wind turbines with a capacity of up to 20 kW is regulated by the Building Law, which—under certain conditions—allows installation without the need for a building permit. Pursuant to Article 29, Section 2, Points 15–16 of the Act of 7 July 1994, mounting a device (e.g., a wind turbine) on a building does not require a permit if it does not exceed 3 m above the building’s ridge [36,37]. For freestanding masts, the height limit is 10 m from ground level. In practice, this means that installing a turbine on a mast attached to a single-family house—whose ridge typically reaches 8–9 m—can legally reach a total height of approximately 12 m. In such a configuration, a construction work notification is sufficient, without the need for a permit [1,36]. In the European Union, there are no uniform regulations regarding MAWTs; regulations remain the responsibility of national and local governments [38]. For example, in Germany, devices with a capacity of up to 10 kW can be installed in rural areas without a permit, provided that distance and height requirements are met [39,40]. In the United Kingdom, Permitted Development applies, which sets a height limit of 11.1 m for freestanding turbines and requires that the distance from the property line be equal to the height of the device [41].

In the United States, most states apply the “30-foot rule”, which stipulates that the minimum mast height should be 9 m above the highest obstacle within a 75 m radius. However, regulations can vary significantly—in some counties, full permits are required even for turbines up to 5 kW [42].

Building height and the influence of terrain roughness

The variation in wind conditions as a function of height can be determined using the Hellmann relation:

$${\displaystyle \frac{v}{v_{ref}}}={\left({\displaystyle \frac{h}{h_{ref}}}\right)}^{\alpha }$$

(5)

where

v—Wind speed at altitude h.

v_ref—Wind speed at reference height h_r. Usually 10 m is taken as the reference height because most meteorological measurements are made at this height.

α—Hellmann coefficient, depending on the roughness of the terrain (usually from 0.10 for sea surface to 0.30 for urban areas) [43].

The Hellmann equation describes how wind speed increases with height above ground. This phenomenon results from air friction with the Earth’s surface. The closer to the ground, the greater the resistance provided by trees, buildings, and other obstacles, which causes a decrease in wind speed. The higher the coefficient, the slower the wind speed increases with height, which is particularly important for rooftop installations. In urban areas, where the increase in wind speed with height is significant, MAWTs should be installed above turbulent buildings (usually >12 m).

In this study, the following Hellmann exponents were used when scaling wind speed between heights; α = 0.12 (open/exposed; B7), 0.14 (suburban/low-rise; B1–B4), 0.18 (light industrial/large roofs; B5), and 0.22 (dense urban/high-rise; B6), consistent with standard urban roughness ranges reported in the literature. Sensitivity to α was checked and found to be secondary compared to the Weibull scale parameter c.

Integration with the building

Roof-mounted MAWTs must be properly integrated into the building structure. Key aspects include the following:

-

Static roof strength;

-

Protection against vibration transmission;

-

Noise and structural resonance reduction;

-

Wind protection (e.g., minimizing edge vortices).

Building regulations must also be considered. For example, in Poland, according to [37], turbines > 3 m high mounted on roofs require a notification or a building permit.

Channeling and Aerodynamic Shading

The specific nature of urban development creates unique airflow conditions. On the one hand, wind can be directed between buildings (channeling effect), locally increasing its speed by up to 30%. On the other hand, each building creates an aerodynamic shadow, a zone of reduced wind speed and increased turbulence. To optimize MAWT performance, installation in an aerodynamic shadow should be avoided. It is recommended that the turbine be placed at least 2 m above the highest point of any obstacle within a 10 m radius and away from roof edges, where separation vortices form [4,17,44].

Optimal MAWT Rooftop Location

To ensure efficient and safe operation of a small wind turbine (MAWT) on a roof, several key principles must be followed. The optimal turbine location should be at least 2 m above the highest point of any obstacle within a 10 m radius. Furthermore, to avoid the negative impact of air vortices, it is not recommended to install it directly on the edges of flat roofs.

MAWT location requires not only technical correctness but also consideration of legal, environmental, and social aspects. Compliance with local noise limits, which can range from 40 to 55 dB at night, is crucial. Social acceptance is equally important, and factors such as aesthetics, noise, proximity to buildings, and the stroboscopic effect influence the installation. Moreover, although MAWTs rarely pose a significant threat to birds, the risk of collision is minimal for turbines with a low tip-speed ratio (TSR). It should also be noted that installations exceeding 40 kW or 30 m in height typically require an environmental impact assessment (EIA), although most MAWTs are not subject to these regulations [45,46].

2.5. Environmental and Social Factors in MAWT Assessment

The selection and location of MAWTs requires consideration not only of energy and technical parameters but also of their impact on the natural and social environment. This section presents a systematic analysis of environmental and social impacts, complementing the prior technical assessment.

Noise generated by wind turbines is one of the main factors affecting public acceptance of installations. MAWTs can emit from 35 dB(A) (resistance-type VAWTs) to over 60 dB(A) (multi-blade HAWTs) [34]. The impact of noise depends on the turbine type, rotational speed, rotor design, and mounting method (mast, roof, façade). In urban environments, it is recommended to use turbines with a noise level below 45 dB(A) @ 10 m.

The stroboscopic effect (sunlight flickering through rotating blades) can cause discomfort, especially when mounted near windows. In MAWTs, this effect is particularly significant for large-diameter HAWTs with a high TSR. VAWTs and BIWTs have a significantly lower stroboscopic potential [2].

Structural vibrations are transferred to the building structure primarily in rooftop installations without adequate insulation. Reducing these vibrations requires the use of vibration dampers and flexible spacers.

The literature indicates that MAWTs have a significantly lower impact on birds than large wind turbines. This is due to the smaller rotor diameter, lower ground clearance, and limited blade tip speed. Savonius and BIWTs with a low TSR and solid, opaque blades visible to birds are particularly advantageous [35]. The impact of a MAWT on the landscape depends on the location (urban vs. open), installation height, and turbine design. Types such as CAWT, WLT, and BIWT offer high architectural integration. Aesthetics play a significant role in social acceptance, especially in landscape-protected zones [47].

Studies indicate that the level of acceptance for MAWTs is significantly higher than for large HAWTs, especially if

-

Residents are involved in the project (e.g., as prosumers);

-

The installation serves a demonstration or educational purpose;

-

Noise and visual impact are minimal [48].

Turbines with high educational potential (Savonius, helical VAWTs, BIWTs) increase acceptance through a “technology familiarization” effect—they are often used in schools, universities, and science centers.

Table 3. Environmental and social impact of different types of MAWT.

Turbine Type	Noise	Vibrations	Strobe Effect	Risk to Birds	Visual Integration	Social Acceptance
HAWT 3-blade	Medium	Medium	High	Moderate	Low–Moderate	Medium
HAWT multi-blade	High	High	High	Moderate	Low	Low
VAWT—Darrieus	Low	Low	Medium	Low	High	High
VAWT—Helical	Low	Low	Low	Very Low	High	High
VAWT—Savonius	Very Low	Very Low	None	Very Low	Medium–High	Very High
VAWT—Hybrid	Low	Medium	Medium	Low	Medium–High	High
DAWT	Medium	Medium	High	Moderate	Moderate	Medium
CAWT	Low	Low	Low	Low	High	High
WLT (lenticular)	Medium	Low	Low	Low	High	Medium–High
BIWT	Inaudible	Very Low	None	None	Very High	Very High

presents a comparative overview of different types of MAWTs in terms of their environmental and social impact. The classification is derived from a synthesis of published studies [1,4,17,18,34,35,45,46,47,48], combined with the authors’ expert analysis. For less conventional concepts (e.g., CAWT, DAWT, WLT), where systematic acceptance studies are scarce, the assessment reflects the authors’ judgment based on turbine parameters and design-oriented reviews.

As shown in Table 3, the environmental and social impact of MAWTs varies considerably across design concepts. Traditional multi-blade HAWTs are generally associated with higher noise, vibrations, and lower visual integration, leading to lower acceptance levels. In contrast, Savonius, helical VAWTs, and BIWTs demonstrate higher social acceptance due to their low acoustic emissions, minimal stroboscopic effect, and stronger architectural integration. Hybrid and emerging designs (CAWT, DAWT, WLT) occupy an intermediate position, where technical potential is promising but acceptance levels remain less well documented and require further empirical validation.

3. Results

This chapter presents the results of a technical, environmental, and operational analysis of micro- and small-scale wind turbines (MAWTs) in the context of their application in local energy systems. Energy performance (AEP), user comfort, environmental impact, and integration with various building types were considered. The study was conducted for a representative location in northern Poland and a set of typical urban scenarios. Various MAWT designs (HAWT, VAWT, DAWT, CAWT, WLT, BIWT) were evaluated, compared in terms of efficiency, suitability to local conditions, and acoustic and visual impact.

3.1. Classification of Buildings and Locations for MAWT Integration

To reliably assess the feasibility of MAWT applications, it is essential to classify the building types and urban contexts in which the installations are planned.

Table 4. Building classification for MAWT integration.

Type	Building Description /Location	Sample Height	Urban Characteristics	Notes on MAWT
B1	Single-family home	6–8 m	Open or semi-open areas, low building density	Limited access to even wind conditions, but easy rooftop installation.
B2	Low-rise urban building	10–12 m	Denser development, apartment complexes, e.g., Gdynia Redłowo	Significant impact of wind disturbances, but typical urban conditions require analysis.
B3	Public building	12–15 m	Schools, clinics, municipal buildings—medium exposure	Stable energy demand, possible integration with PV and MAWT.
B4	4-5-story residential building	15–20 m	Medium or high-rise housing development	Height allows for better wind conditions (roof).
B5	Industrial/technical building	25–35 m	Warehouses, technical industrial buildings	Large roof area, good MAWT location.
B6	High-rise urban building (office, hotel)	30–60 m	Skyscrapers, hotels, commercial buildings	Very good wind exposure, but risk of noise and resonance.
B7	Autonomous/off-grid building	5–15 m	Passive houses, shelters, mobile homes in open spaces	Ideal wind conditions, MAWT as the main energy source.

presents seven representative building classes (B1–B7), taking into account their height, urban character, and relevant considerations for MAWT integration.

The following subsections present the results of the technical, environmental, and operational analysis of selected MAWT types in relation to the above building types and local conditions.

3.2. Local Scenarios and Reference Conditions

First, measurement data were collected for building B2 (Gdynia-Redłowo), where a full analysis of the wind speed distribution was performed based on actual measurements from a 12-month period. Based on these data, the Weibull distribution parameters for this location were estimated. The same distribution was then modeled using publicly available meteorological data from the Wunderground portal, Meteo-model.pl, and indirect information from reanalysis systems (ERA5, NOAA/NCEP), and compared with the measured distribution.

The next location was building type B4—a 4-5-story building. For building B4, an analysis was also conducted based on actual data, and then the appropriate Weibull distribution was modeled for the location. The accuracy and differences between the measured and modeled distributions were compared.

Based on the conclusions from both locations (B2 and B4), fitted Weibull distributions were estimated for the remaining building types (B1, B3, B5, B6, B7), assuming height scaling and estimated corrections for turbulence intensity and building exposure.

This section presents the measured and predicted wind resource parameters for the reference sites (

Table 5. Weibull parameters (k,c), mean wind speed (Vavg), measurement/model reference height, Kolmogorov–Smirnov (KS) and Anderson–Darling (AD) goodness-of-fit statistics, and 95% confidence intervals for k and c for representative building locations (B1–B7).

Building Type	Location	Data Source	k (95% CI) [–]	c (95% CI) [m/s]	Vavg [m/s]	Height Measurement [m]	KS Statistic	AD Statistic
B1M	Northern Poland—single-family homes	Model	1.65 (1.52–1.78)	3.88 (3.74–4.02)	3.42	15	0.076	0.482
B2P	Gdynia–Rełowo	Measurement	8.15 (7.92–8.38)	3 (2.86–3.14)	2.7	12	0.082	0.501
B2M	Gdynia–Rełowo	Model	2 (1.88–2.12)	4 (3.87–4.13)	3.6	15	0.069	0.463
B3M	Luzino—school	Model	1.75 (1.63–1.87)	4.05 (3.92–4.18)	3.65	15	0.073	0.476
B4P	Kołobrzeg—apartment building	Measurement	9.08 (8.84–9.32)	4.61 (4.45–4.77)	4.1	15	0.081	0.498
B4M	Kołobrzeg—apartment building	Model	1.79 (1.67–1.91)	4.28 (4.14–4.42)	3.76	15	0.070	0.469
B5M	Gdynia—sports hall	Model	1.85 (1.73–1.97)	5 (4.85–5.15)	4.72	30	0.066	0.454
B6M	Gdańsk Wrzeszcz	Model	1.87 (1.75–1.99)	5.4 (5.25–5.55)	5.12	40	0.064	0.447
B7M	Northern Poland	Model	1.9 (1.78–2.02)	4.6 (4.46–4.74)	4	15	0.068	0.459

). For location B1, only modeled data (B1M) were available, as no dedicated measurement campaign was conducted at this site. As illustrated by buildings B2 and B4, the differences between BxM (measured) and BxP (predicted) values reflect the expected gap between atlas-based predictions and site-specific measurements. Such discrepancies are particularly evident under turbulent or obstructed urban conditions, and highlight the importance of local measurements to complement regional atlas data.

Table 5 summarizes the Weibull parameters (k, c), mean wind speed (Vavg), and reference measurement/model heights for the representative building locations B1–B7. For each dataset (measured or modeled), the Kolmogorov–Smirnov (KS) and Anderson–Darling (AD) goodness-of-fit statistics were computed to quantify the agreement between measured and modeled wind speed distributions. In addition, 95% confidence intervals (CI_k, CI_c) were calculated for the estimated k and c values, using maximum likelihood estimation with 10,000 bootstrap resamples.

The KS statistic measures the maximum absolute difference between the empirical and modeled cumulative distribution functions, while the AD statistic emphasizes discrepancies in the distribution tails. Lower KS and AD values indicate better agreement.

The reference heights reflect either the original measurement height or a modeled height correction to a standardized reference level (most commonly 10–15 m).

The site-specific context for each dataset is as follows: B1M refers to a semi-open terrain location with height correction applied to 10 m. B2P represents 12-month logger measurements with a turbulence intensity (TI) of 22%, while B2M is a modeled dataset derived from Meteomodel, Wunderground, and ERA5 sources, with height correction applied. B3M corresponds to a moderately exposed site with TI = 20%. B4P comes from a roof-mounted station with a north-west exposure, whereas B4M is modeled using TI values and local urban morphology. B5M is located on a flat roof at 30 m surrounded by low-rise buildings. B6M represents a high-rise building subject to interference from neighboring structures. Finally, B7M corresponds to an open area with no significant terrain obstacles.

These parameters provide the basis for subsequent AEP calculations and MAWT site suitability assessments presented in Section 3 and Section 4.

The parameters presented in Table 5 correspond to the same datasets whose Weibull fits are illustrated in Figure 1, allowing visual inspection alongside the statistical indicators. The KS and AD statistics, together with the 95% confidence intervals for k and c, were further used in Section 3 to interpret the robustness of AEP estimates and to evaluate the reliability of MAWT site suitability rankings.

Figure 1 illustrates the comparison between measured and modeled Weibull distributions for the seven representative buildings (B1–B7). For each site, the Weibull shape parameter k and scale parameter c were estimated using the maximum likelihood method, with 95% confidence intervals derived via non-parametric bootstrapping (10,000 resamples). Shaded regions denote the confidence intervals for both k and c, allowing a visual assessment of parameter uncertainty. Goodness-of-fit was evaluated using the Kolmogorov–Smirnov (KS) and Anderson–Darling (AD) statistics, with the results reported in Table 5. The high degree of overlap between measured and modeled curves indicates that the applied wind field-modeling framework reliably reproduces the observed wind speed distribution at each site.

3.3. Forecasted Annual Energy Production (AEP) for Selected MAWT Types

Six MAWT types with different aerodynamic characteristics and start-up thresholds were selected for electricity production forecasting: HAWT (Bornay Inclin 3.5, Bornay Aerogeneradores S.L., Castalla, Spain); VAWT Darrieus (Turby, Turby B.V., Delft, The Netherlands); VAWT Savonius (UGE 1K, Urban Green Energy Inc., New York, NY, USA); DAWT (Ogin 1.2, Ogin Inc., Waltham, MA, USA); CAWT (Windside WS-0.3B, Windside Production Ltd., Toivakka, Finland); and BIWT (IceWind CW-1000, IceWind ehf., Reykjavik, Iceland).

There are two complementary AEP views. To make rankings both fair (like-for-like aerodynamics) and realistic (installation footprint), we report AEP in two complementary ways:

(i)

Table 6a—2 kW-normalized curves: Each typology’s power curve is scaled to reach 2 kW at v_r while preserving the original cut-in/rated/cut-out thresholds and the (v) shape. This isolates aerodynamic behavior and start-up thresholds.

(ii)

Table 6b—Footprint-constrained curves: All typologies are evaluated for a common rotor footprint (D = 2.0 m, swept area $A={\displaystyle \frac{\pi ·D^2}{4}}\approx 3.14{\mathrm{m}}^2$). Here, no rescaling to 2 kW is applied; absolute power is limited by $P_r={\displaystyle \frac{1}{2}}\rho A{C}_{pmax}{v}_r^3$. This view reflects rooftop constraints and yields realistic absolute AEP at the same swept area.

The calculation of AEP follows directly from the wind power equation given in Equation (1), applied to each turbine’s specific rotor area, aerodynamic performance curve Cp(v), and operating wind speed range. To enable fair comparison, all AEP values were normalized to a common nominal electrical power of 2 kW at the rated wind speed v_r, while retaining the actual Cp(v) characteristics, rotor swept area A, and manufacturer-specified cut-in v_ci, rated v_r, and cut-out v_co speeds.

For each typology,

• The original power curve Porig(v) was obtained from manufacturer data or the literature, digitized where necessary, and rescaled so that P(v_r) = 2 kW, while retaining its original shape and the manufacturer-specified cut-in, rated, and cut-out wind speeds.
• The manufacturer- or the literature-based performance curve was applied without altering its shape except for the scaling in point 1.
• The probability density f(v) of wind speeds at the considered site was modeled using the Weibull parameters (k, c) in Table 5.
• Annual energy production was computed as

$$AEP={\int }_{v_{ci}}^{v_{co}}P\left(v\right)·f\left(v\right)·8760dv$$

(6)

where

P(v)—Power curve used in the AEP calculation after rescaling to P(v_r) = 2 kW with thresholds unchanged;

Porig(v)—Original (manufacturer- or literature-based) power curve before rescaling;

f(v)—Weibull distribution density;

v_ci, v_co—Turbine cut-in and cut-out speed;

v_r—Rated wind speed.

Consistent with Equation (1) and prior to evaluating the AEP integral in Equation (6), we define the following:

Rated power:

$$P_r={\displaystyle \frac{1}{2}}·\rho ·A·C_{pmax}·v_r^3$$

(7)

Operational power curve with thresholds:

$$P\left(v\right)=\left\{\begin{array}{c}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}0,\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}v<v_{ci}\\ {\displaystyle \frac{1}{2}}·\rho ·A·C_p\left(v\right)·v^3,\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}{v}_{ci}\le v<v_r\\ {\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}P}_r,\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}{v}_r\le v\le v_{co}\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}0,\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}v>v_o\end{array}\right.$$

(8)

where

ρ—Air density;

A—Rotor swept area;

C_pmax—Maximum power coefficient (attained near v_r);

C_p(v)—Typology-specific power–coefficient curve.

5.

Uncertainty ranges (±1σ) were estimated by propagating the 95% confidence intervals of k and c using Monte Carlo sampling (10,000 iterations).

Parametric Cp models by typology (for AEP integration)

To enable reproducibility without disclosing proprietary model data, we use typology-level parametric forms calibrated to the reference power curves employed in Table 6a,b. The fits preserve cut-in/cut-out thresholds (v_ci, v_co), match C_pmax and the vicinity of the rated region (v ≈ v_r), and are used solely to evaluate AEP (Equation (6)). For clarity, the adopted functional forms for each turbine class are given in Equations (9)–(13). Together, these parametric models provide a consistent framework for integrating AEP while ensuring that differences between typologies reflect aerodynamic suitability rather than disparities in nominal capacity.

HAWT (lift, horizontal axis):

$$C_p(\lambda ,\beta )\approx c_1\left({\displaystyle \frac{c_2}{{\lambda }_i}}-c_3\beta -c_4\right)exp\left(-{\displaystyle \frac{c_5}{{\lambda }_i}}\right)+c_6\lambda$$

(9)

with ${\lambda }_i^{-1}={\lambda }^{-1}+0.08\beta -{\displaystyle \frac{0.035}{{\beta }^3+1}}$ and β = 0 for fixed-pitch small HAWTs.

VAWT—Darrieus (lift, vertical axis):

$$C_p\left(\lambda \right)\approx a_1\lambda +a_2{\lambda }^2+a_3{\lambda }^3$$

(10)

VAWT—Savonius (drag, vertical axis):

$$C_p\left(\lambda \right)\approx b_1\lambda -b_2{\lambda }^2$$

(11)

DAWT (diffuser/wind-lens augmented rotor):

$$C_{peff}\left(v\right)\approx {\eta }_{shroud} C_{protor}\left(\lambda \right), {\eta }_{shroud}\in \left[\mathrm{1.3,1.8}\right]$$

(12)

where C_protor(λ) may be taken from (9) or (10) depending on rotor type; η_shroud is a weak function of operating point and installation quality.

CAWT/BIWT (compact/shrouded and building-integrated designs):

$$C_p\left(\lambda \right)\approx d_1\lambda +d_2{\lambda }^2+d_3{\lambda }^3, 0 \le C_p \le C_{pmax}$$

(13)

Notes and calibration: λ—tip-speed ratio; β—pitch angle (deg). Coefficients (c_i, a_i, b_i, d_i) are obtained by least-squares fitting to digitized reference curves used in Table 6a,b; bounds enforce 0 ≤ C_p ≤ C_pmax and a smooth rise between v_ci and v_r. The fitted sets (per typology) and scripts are available on request. These parametric forms are only a vehicle to integrate AEP (Equation (6)) and do not replace certified manufacturer data.

The least-squares fitting was applied to representative Cp(λ) or P(v) data available in the literature and manufacturer catalogs, with weighting emphasized around the rated wind speed (v_r) to ensure accurate representation of the transition towards rated operation.

The fitted coefficients for Equations (9)–(13) were obtained by least-squares adjustment to digitized reference curves (Table 6a,b) and are available on request. These forms are intended solely as auxiliary models for AEP integration and do not replace certified manufacturer performance data.

This procedure ensures that differences in the reported AEP values (Table 6 and Table 7) reflect aerodynamic suitability and alignment with local wind statistics rather than disparities in nominal capacity. Calculations were performed using manufacturer power profiles and simplified P(v) models dependent on Cp, rotor area, and air density. The results are summarized in Table 6.

The determined AEP values refer to a uniform rated power of 2 kW for all turbine types, maintaining their actual Cp, rotor area, and cut-in/rated speed thresholds. Uncertainty values shown as ±1σ in Table 6 result from 10,000 Monte Carlo draws of (k, c) around the reference Weibull parameters (B4M; k = 1.79, c = 4.28 m/s) and propagation through the AEP integral; details are provided in the Methods section.

Table 6. (a) AEP from 2 kW-normalized power curves at the reference site (B4M; k = 1.79, c = 4.28 m/s); original (v_ci, v_r, v_co) and C_p(v) shape preserved; values are mean ±1σ from Monte Carlo on (k, c). (b) AEP for a fixed rotor footprint (D = 2.0 m, A ≈ 3.14 m²) under the B4M; ±1σ from Monte Carlo on (k, c). Thresholds (v_ci, v_r, v_co) and C_pmax are retained as typology-specific.

(a)
Turbine Type	Reference Model	Rotor Area [m2]	Cut-in [m/s]	Rated [m/s]	Cpmax	AEP [kWh/Year, ±1σ]
HAWT	HAWT—Ref. 2 kW	9.62	3.5	10.5	0.34	1254 ± 232
VAWT (Darrieus)	VAWTD–Ref. 2 kW	3	2.5	10	0.28	1033 ± 152
VAWT (Savonius)	VAWTS—Ref. 2 kW	2.2	1.5	7.5	0.22	812 ± 70
DAWT	DAWT—Ref. 2 kW	4	2	9	0.4	1476 ± 176
CAWT	CAWT—Ref. 2 kW	1.2	1.5	6.5	0.21	642 ± 44
BIWT	BIWT—Ref. 2 kW	2.5	2.2	9	0.31	1142 ± 141
(b)
Turbine Type	Reference Model	Rotor Area [m2]	Cut-in [m/s]	Rated [m/s]	Cpmax	AEP [kWh/Year, ±1σ]
HAWT	HAWT – Ref. D = 2.0 m	3.14	3.5	10.5	0.34	166 ± 31
VAWT (Darrieus)	VAWTD – Ref. D = 2.0 m	3.14	2.5	10	0.28	172 ± 27
VAWT (Savonius)	VAWTS – Ref. D = 2.0 m	3.14	1.5	7.5	0.22	202 ± 16
DAWT	DAWT – Ref. D = 2.0 m	3.14	2	9	0.4	305 ± 38
CAWT	CAWT – Ref. D = 2.0 m	3.14	1.5	6.5	0.21	192 ± 11
BIWT	BIWT – Ref. D = 2.0 m	3.14	2.2	9	0.31	232 ± 30

Note: “Rotor area” denotes the swept area normal to the wind. For HAWT $A={\displaystyle \frac{\pi ·D^2}{4}}$ ; for VAWT A = H⋅D. Equalizing A ensures each typology intercepts the same flow; differences in AEP arise from C_pmax, v_ci, and the wind speed distribution.

Table 6. (a) AEP from 2 kW-normalized power curves at the reference site (B4M; k = 1.79, c = 4.28 m/s); original (v_ci, v_r, v_co) and C_p(v) shape preserved; values are mean ±1σ from Monte Carlo on (k, c). (b) AEP for a fixed rotor footprint (D = 2.0 m, A ≈ 3.14 m²) under the B4M; ±1σ from Monte Carlo on (k, c). Thresholds (v_ci, v_r, v_co) and C_pmax are retained as typology-specific.

(a)
Turbine Type	Reference Model	Rotor Area [m2]	Cut-in [m/s]	Rated [m/s]	Cpmax	AEP [kWh/Year, ±1σ]
HAWT	HAWT—Ref. 2 kW	9.62	3.5	10.5	0.34	1254 ± 232
VAWT (Darrieus)	VAWTD–Ref. 2 kW	3	2.5	10	0.28	1033 ± 152
VAWT (Savonius)	VAWTS—Ref. 2 kW	2.2	1.5	7.5	0.22	812 ± 70
DAWT	DAWT—Ref. 2 kW	4	2	9	0.4	1476 ± 176
CAWT	CAWT—Ref. 2 kW	1.2	1.5	6.5	0.21	642 ± 44
BIWT	BIWT—Ref. 2 kW	2.5	2.2	9	0.31	1142 ± 141
(b)
Turbine Type	Reference Model	Rotor Area [m2]	Cut-in [m/s]	Rated [m/s]	Cpmax	AEP [kWh/Year, ±1σ]
HAWT	HAWT – Ref. D = 2.0 m	3.14	3.5	10.5	0.34	166 ± 31
VAWT (Darrieus)	VAWTD – Ref. D = 2.0 m	3.14	2.5	10	0.28	172 ± 27
VAWT (Savonius)	VAWTS – Ref. D = 2.0 m	3.14	1.5	7.5	0.22	202 ± 16
DAWT	DAWT – Ref. D = 2.0 m	3.14	2	9	0.4	305 ± 38
CAWT	CAWT – Ref. D = 2.0 m	3.14	1.5	6.5	0.21	192 ± 11
BIWT	BIWT – Ref. D = 2.0 m	3.14	2.2	9	0.31	232 ± 30

Note: “Rotor area” denotes the swept area normal to the wind. For HAWT $A={\displaystyle \frac{\pi ·D^2}{4}}$ ; for VAWT A = H⋅D. Equalizing A ensures each typology intercepts the same flow; differences in AEP arise from C_pmax, v_ci, and the wind speed distribution.

Footprint-constrained AEP (Table 6b). Complementary to the normalized view, we compute AEP for a fixed rotor diameter D = 2.0 m (equal swept area for all typologies). Power is given by $P\left(v\right)={\displaystyle \frac{1}{2}}\rho A{C}_p(v)v^3$ with a smooth ramp from v_ci to C_pmax at v_r and a rated-power regime beyond v_r; P=0 for v>v_co. Uncertainty bands (±1σ) come from Monte Carlo propagation (10,000 draws) of the Weibull (k, c) confidence intervals (B4M). This table enforces the same swept area (same “air stream”) and shows how typology-specific C_pmax, v_ci, and v_r translate into absolute energy yield at a shared footprint.

To visualize the two reporting conventions, we overlay the typology-specific power curves under the following: Figure 2a the 2 kW-normalized view (like-for-like aerodynamics; Table 6a), and Figure 2b the fixed-footprint view (D = 2.0 m, A ≈ 3.14 m²; Table 6b). Vertical dashed lines mark cut-in v_ci and cut-out v_co for each typology; the shaded band indicates the rated-power regime (v ≥ v_r). Together, these overlays explain why rankings differ between Table 6a,b.

At lower wind speeds (≲6 m/s), drag-based typologies (VAWT-S, CAWT) begin producing earlier due to low v_ci. As wind speed increases, lift-based designs with higher C_pmax (DAWT, HAWT, BIWT) dominate until curves clip at P_r for v ≥ v_r. Under the 2 kW-normalized view, differences reflect aerodynamic shape and thresholds; under the fixed-footprint view, equalized swept area shifts emphasis to v_ci, v_r (via P_r ∝ v_r) and sub-rated C_p(v), explaining the shifts in AEP rankings between Table 6a,b.

Agreement between measured and modeled wind speed distributions was quantified using KS and AD statistics (Table 5). AEP uncertainty reflects Monte Carlo propagation of (k, c) confidence intervals (Table 6) and a ±0.2 m·s⁻¹ instrument accuracy band for measured wind speed; additional sensitivity to α and ρ was tested and found to be secondary.

In this section, energy production (AEP) calculations were performed for each turbine based on the local wind speed distribution and the rotor swept area. Calculations were performed according to Formula (6).

Findings for the 2 kW-normalized view (Table 6a) are as follows:

-

DAWT reaches the highest AEP under reference winds due to high C_pmax and low v_ci;

-

BIWT and VAWT-D perform well;

-

VAWT-S and CAWT remain competitive at low speeds (low v_ci) despite lower C_pmax;

-

HAWT is disadvantaged in low/turbulent winds because of higher v_r, despite a high C_pmax.

Findings for the footprint-constrained view (Table 6b, D = 2.0 m) are as follows:

-

Absolute AEPs are lower (no 2 kW rescaling);

-

DAWT leads at rooftop-relevant winds, followed by BIWT and VAWT-S;

-

HAWT approaches the leaders only at higher speeds (benefiting from large v_r);

-

Rankings shift because A is equalized and typology-specific v_ci, v_r, C_pmax govern the usable part of the distribution.

The shift between the 2 kW-normalized view (Table 6a) and the fixed-footprint view (Table 6b) arises because equalizing swept area removes nominal-power effects and makes AEP primarily governed by typology-specific cut-in v_ci, rated speed v_r (via P_r ∝ v_r), and sub-rated Cp(v) under the site’s Weibull distribution.

Table 7 extends the 2 kW-normalized view (Table 6a) across locations (B1M–B7M), highlighting the role of height/exposure and turbulence; differences between B2P (measured) and B2M (modeled) reflect height reference, TI, and local morphology.

Table 7. Annual energy production (AEP [kWh/year]) of selected MAWTs depending on location.

Location	HAWT	VAWT-D	VAWT-S	DAWT	CAWT	BIWT
B1M	841	697	524	1065	421	679
B2P	978	811	599	1227	486	790
B2M	1022	861	648	1355	525	845
B3M	1104	921	693	1463	561	904
B4P	1166	972	732	1545	591	954
B4M	1215	1012	764	1609	617	992
B5M	1350	1125	847	1792	687	1102
B6M	1428	1190	896	1895	727	1165
B7M	965	805	605	1280	491	780

Table 7. Annual energy production (AEP [kWh/year]) of selected MAWTs depending on location.

Location	HAWT	VAWT-D	VAWT-S	DAWT	CAWT	BIWT
B1M	841	697	524	1065	421	679
B2P	978	811	599	1227	486	790
B2M	1022	861	648	1355	525	845
B3M	1104	921	693	1463	561	904
B4P	1166	972	732	1545	591	954
B4M	1215	1012	764	1609	617	992
B5M	1350	1125	847	1792	687	1102
B6M	1428	1190	896	1895	727	1165
B7M	965	805	605	1280	491	780

Based on the annual energy production (AEP) calculations, the following conclusions can be drawn:

• AEP increases with mounting height and exposure. In the lowest site (B1M), most typologies produce < 950 kWh/year, with DAWT ≈ 1065 kWh as the only exception (Table 7). This confirms the limited feasibility of MAWT on low-rise buildings without access to free flow.
• Top yields occur in B5M–B6M. The highest AEP values are delivered by DAWT, reaching ≈ 1.8–1.9 MWh/year (e.g., 1895 kWh in B6M). HAWT peaks around 1.43 MWh/year (B6M), i.e., does not reach the 1.8–2.0 MWh range (Table 7). For the 2 kW class, this corresponds to capacity factors ≈ 0.06–0.11, not ~0.25.
• DAWT is the top performer across all locations (B1M–B7M). HAWT typically ranks second in stronger-wind sites (e.g., B5M–B6M), while VAWT-D is consistently moderate, above VAWT-S and CAWT but below HAWT/DAWT (Table 7).
• VAWT-S and CAWT yield the lowest AEP, especially in B1–B3. These low-speed designs trade peak efficiency for reliability and low acoustic impact (Table 6a,b and Table 7).
• BIWT sits in the mid-range; generally above VAWT-S and CAWT and below HAWT and DAWT, with a relative advantage in mid-to-strong wind settings (e.g., B4–B6) (Table 7).
• At the reference site B4M, the 2 kW-normalized ranking is DAWT > HAWT ≈ BIWT > VAWT-D > VAWT-S > CAWT (Table 6a), whereas under the fixed-footprint view (D = 2.0 m) DAWT also leads, followed by BIWT and VAWT-S (Table 6b).
• B2P vs. B2M show consistent, modest differences (≈4–10%) with the modeled series tending higher, reflecting differences in height reference, turbulence intensity, and local morphology (Table 7).
• Design/placement implications: Turbine choice should follow site characteristics. Tall buildings (B5–B6) enable efficient use of HAWT/DAWT; in space- or noise-constrained settings, VAWT-S, BIWT, or CAWT may be preferable despite lower AEP, owing to compactness and lower acoustic impact.

3.4. The Impact of Turbines on User Comfort and the Environment

The assessment of the environmental and user comfort impact of micro and small wind turbines (MAWTs) combines both quantitative and qualitative parameters that are critical for installations in residential and urban contexts. The most relevant include the following:

• Noise and acoustic profile—Expressed as sound pressure level ranges in dB(A) at a standard reference distance of 10 m, measured under conditions close to rated wind speed. The ranges reflect typical variability with wind speed, blade geometry, and mounting configuration. Data are derived from manufacturer documentation, field measurement reports, and peer-reviewed literature on small turbine aeroacoustics [18,19,24,30,31,34,44].
• Vibrations and resonances—Particularly relevant for rooftop and building-integrated installations. Categorization (“Very Low” to “High”) is based on rotor mass, rotational speed, bearing type, and mounting stiffness [19,31,44].
• Stroboscopic effect and shadowing—Linked to the periodic shading caused by rotor blades intercepting sunlight, evaluated qualitatively as “High” to “Negligible” depending on rotor geometry, tip-speed ratio, and solar incidence [17,45,46].
• Visibility and aesthetics—Influencing social perception and acceptance, particularly in dense urban areas, as reported in resident surveys and urban planning case studies [45,46,47].
• Compliance with urban planning regulations—Covering setbacks from property lines and windows, compliance with day/night noise limits, and integration with building architecture [19,21,49].

In particular, noise and vibration are among the most relevant limitations for MAWT integration in urban areas (evaluated as LAeq at 10 m for noise, unless otherwise noted).

Table 8. Environmental and operational impact summary for MAWT types; noise metrics follow IEC 61400-11 (where applicable).

Turbine Type	Noise Level [LAeq dB(A) @ 10 m]	Stroboscopic Effect	Vibration	Social Acceptance	Installation Notes
HAWT	55–65	High (horizontal rotor)	Moderate (mast- mounted)	Low in densely built-up areas	Requires free space, visually dominant
VAWT-D	45–58	Low	Moderate	Medium—beneficial for flat roofs	Height may affect resonance
VAWT-S	42–55	None	Low	High in off-grid locations	Low efficiency but quiet profile
DAWT	50–60	Medium (rotor cover)	Low–Medium	Medium	Precise tunnel installation required
CAWT	45–52	Negligible	Very Low	High	Ideal for green or urban roofs
BIWT	48–55	Configuration dependent	Medium	High in urban areas	High aesthetics and modularity

summarizes the comparative environmental and operational impact of different MAWT configurations, integrating the above factors. All noise values are standardized as L_Aeq (equivalent continuous A-weighted sound pressure level, in dB(A) at a reference distance of 10 m) for operation near rated wind speed; vibration, flicker, and acceptance ratings are based on literature and field evidence. Actual values may differ depending on the specific model, local wind regime, and installation method [34,50,51].

Noise values are A-weighted equivalent continuous sound pressure levels (LAeq) at a reference distance of 10 m, near rated wind speed. Vibration is assessed qualitatively (Very Low–High) with emphasis on rooftop transmission risk. Stroboscopic effect (shadow flicker) is rated qualitatively based on rotor geometry and tip-speed ratio. Social acceptance reflects syntheses from the literature and field experience in urban contexts. Other qualitative ratings (vibration, shadow flicker, social acceptance) draw on literature, manufacturer data, and field measurements. Actual values may vary with turbine model, site conditions, and installation method.

3.5. Multi-Criteria Evaluation of the Effectiveness and Usefulness of MAWTs

A multi-criteria assessment (MCDM) was used to comprehensively evaluate micro- and small-scale wind turbines (MAWTs). Three main evaluation axes were considered:

• Energy efficiency (E)—Based on annual energy production (AEP).
• Environmental impact and user comfort (C)—Including noise, vibration, stroboscopic effect, and social acceptance. In urban contexts, this axis is dominated by three constraints: LAeq noise at 10 m, structural vibration, and shadow flicker.
• Suitable for local conditions (L)—Taking into account the impact of turbulence, installation height, openness, and regulatory compliance.

The partial values of each axis were normalized to a scale of [0, 1], where 1 represents the most desirable result. A composite evaluation index—the MAWT-Score—was then calculated using the following formula:

$$MAWT-Score=0.4·E+0.3·C+0.3·L$$

(14)

where

E—Normalized value of energy efficiency (e.g., AEP of a given turbine type related to the maximum);

C—Normalized value for comfort and environmental impact (lowest noise, no vibration, etc., receive a higher score);

L—Normalized value of the fit to local conditions.

The evaluation was conducted for six turbine types (HAWT, VAWT-D, VAWT-S, DAWT, CAWT, and BIWT) based on averaged data from sites B1M–B7M.

The final result was provided as a summary efficiency index (MAWT-Score), which allows for ranking technologies in the analyzed scenarios. The results are presented in a table (

Table 9. Multi-criteria MAWT evaluation (MAWT-Score).

Turbine Type	AEP Score	Comfort Score	Environmental Score	MAWT-Score
HAWT	0.65	0.3	0.2	0.38
VAWT-D	0.53	0.65	0.55	0.57
VAWT-S	0.42	0.85	0.8	0.72
DAWT	0.81	0.65	0.75	0.75
CAWT	0.73	0.85	0.75	0.73
BIWT	0.63	0.75	0.85	0.71

) and as a bar chart enabling quick identification of the most advantageous solutions.

Table 9 presents a synthetic evaluation of six types of micro- and small-scale wind turbines (MAWTs), based on three main criteria: energy efficiency (E), user comfort and environmental impact (C), and local suitability (L). Each criterion was normalized on a scale of 0 to 1 and then incorporated into the final MAWT-Score according to the adopted weights: 40% for efficiency, 30% for comfort, and 30% for local suitability. The DAWT achieved the highest MAWT-Score (0.75), indicating its good energy efficiency, relatively favorable environmental impact, and universal suitability for the analyzed locations. Second place was taken by the BIWT (0.71), which, despite its moderate efficiency, distinguished itself by high ratings for comfort and social acceptability. The VAWT-D turbine (0.57) also received a high score, demonstrating a balance between the three criteria.

Figure 3 visualizes the MAWT-Score values for each technology in a column chart, facilitating comparison of the relative differences between turbine types. This comparison indicates that the choice of a specific MAWT solution should be determined not only by energy efficiency but also by environmental factors and local installation conditions.

Multi-criteria radar chart—qualitative assessment

Additionally, a radar chart (Figure 4) was developed presenting the distribution of scores for each MAWT type across nine qualitative criteria:

• Energy efficiency (AEP): Mapped from the relative ranking and spread established in Table 6a,b (reference-site and fixed-footprint views).
• Acoustic comfort (noise): Informed by the small-wind noise and annoyance literature, including low-frequency effects. Reported values are based on LAeq levels at 10 m distance; higher scores reflect quieter designs and lower tip speeds [34].
• Shadow flicker: Qualitative risk reflecting rotor diameter/TSR and urban geometry, following recent syntheses on urban wind environmental effects [17].
• Structural vibration and shake: Based on feasibility evidence for building-mounted small-wind and rooftop transmission considerations [31].
• Operational safety: Referenced to small-wind design and safety guidance [24] (IEC 61400-2).
• Visual acceptability and social acceptance: Informed by reviews on social acceptance and wind-energy landscapes in urban contexts; compact/shrouded forms generally rate higher [45,48].
• Architectural and technical integration: Guided by interdisciplinary reviews of wind-powered building skins and integration aspects [21].
• Educational potential: Expert judgment (public-facing safety/visibility, demonstrability) where formal benchmarks are sparse.

Scores were assigned on a scale of 0 to 10 based on expert knowledge and a review of the scientific literature. Scoring protocol and sources are as follows. Expert scores (0–10) were obtained via a structured elicitation with three domain experts (renewable energy systems, wind engineering, environmental/user impact). Each expert scored typologies against nine criteria; scores were reconciled to a consensus and linearly normalized to [0, 1] for aggregation in Equation (15). Anchors were drawn from the quantitative AEP ranking (Table 6a,b), small-wind noise and annoyance literature [34], urban environmental/flicker syntheses [17], building-mounted small-wind vibration evidence [31], small-wind safety guidance [24], and social/visual acceptance studies in urban settings [45,48], with architectural integration informed by building-skin reviews [21]. The results are presented in the chart below:

Mathematical relationship for the radar chart:

$$MAWT=\sum _{i=1}^9{w}_i·x_i$$

(15)

where

x_i—Normalized value of the partial score on a scale [0, 1] (higher is better);

w_i—Weight assigned to a given criterion.

The adopted weights for each of the nine criteria:

w_E → Energy efficiency (AEP) → 0.20

w_N → Acoustic comfort (noise) → 0.15

w_F → Shadow flicker → 0.10

w_V → Vibrations and structural vibrations → 0.10

w_S → Operational safety → 0.10

w_A → Visual acceptance (aesthetics) → 0.10

w_SA → Social acceptance → 0.10

w_I → Architectural and technical integration → 0.10

w_ED → Educational potential → 0.05

These weights reflect the relative importance of each criterion in qualitative MAWT assessment. Energy efficiency carries the highest weight (0.20), followed by acoustic and societal impact (0.15–0.10). Educational potential receives the lowest weight (0.05), being less critical for private use though valuable in demonstration projects. The set of criteria and weights can be adapted to the application context, local conditions, and stakeholder preferences. Below are three example scenarios to illustrate how priorities change with context; they also serve as a basis for sensitivity check (±25% for each weight separately).

• Set 1 (balanced)
• w_E = 0.25, w_N = 0.12, w_F = 0.06, w_V = 0.10, w_S = 0.12, w_A = 0.07, w_SA = 0.15, w_I = 0.08, w_ED = 0.05

Rationale: Energy and safety are prioritized; social acceptance receives meaningful emphasis; integration is non-negligible, while acoustic, visual and vibration criteria jointly address user comfort.

• Set 2 (urban, noise-sensitive)
• w_E = 0.20, w_N = 0.20, w_F = 0.10, w_V = 0.12, w_S = 0.10, w_A = 0.06, w_SA = 0.13, w_I = 0.06, w_ED = 0.03

Rationale: Stricter acoustic and social constraints typical of dense urban sites; integration weight is moderate; educational potential is least critical.

• Set 3 (off-grid, energy-first)
• w_E = 0.35, w_N = 0.08, w_F = 0.04, w_V = 0.08, w_S = 0.12, w_A = 0.05, w_SA = 0.17, w_I = 0.06, w_ED = 0.05

Rationale: Energy efficiency dominates; acoustic constraints are relaxed; social acceptance remains relevant due to visual impact and siting; integration kept moderate.

Both the quantitative (AEP-based ranking) and the qualitative (radar chart) perspectives capture complementary aspects of MAWT evaluation. Depending on the investor’s priorities (e.g., efficiency, comfort, urban integration), one can (i) adjust weights to the local context, (ii) select the technology that best meets site needs, and (iii) build an acceptability profile for the affected community. The assessment suggests that DAWT and BIWT have high potential for sustainable deployment in residential environments, while HAWT—despite high efficiency—may require compromises in acoustic comfort and social acceptance.

4. Discussions

The results presented in Section 3 provide important information on the energy efficiency and suitability of individual types of micro- and small-scale wind turbines (MAWTs) in various urban and climatic contexts. This chapter interprets them, comparing them with literature results and highlighting limitations, practical applications, and design recommendations derived from the analysis.

The Importance of Location and Mounting Height

Both the measured data and the modeled Weibull distributions for the analyzed buildings (B1–B7) clearly confirm the strong variation in energy potential depending on the installation height and surrounding environment. The lowest AEP values (below 850 kWh/year) were obtained for locations B1 (single-family homes) and B2 (low-rise urban buildings) (Table 7), which is due to the limited installation height and high turbulence intensity (TI > 20%). The highest yields (up to 1900 kWh/year for DAWT and HAWT) were observed for locations B6 and B5 (Table 7), i.e., for tall commercial buildings and industrial halls with roof heights exceeding 30 m. According to observations from the literature [1,17,21], installation height and avoiding slipstreams are key factors enabling the efficient use of MAWTs, especially in built-up environments. For urban rooftops, acoustic emissions (evaluated as LAeq at 10 m) and structural vibration often become binding constraints and must be verified alongside predicted energy yield.

Differences in MAWT Efficiency Depending on Turbine Type

AEP analysis (Table 7) revealed significant differences between turbine types (Table 7). Tunnel-type structures (DAWTs) achieved the highest annual energy production values, confirming their effectiveness in locations with limited access to free wind. Similar observations were presented in [10], where DAWTs achieved 2–3 times higher yields than conventional turbines with the same rotor area. HAWT designs achieved relatively good results in stable wind locations (B5–B6) (Table 7), but their efficiency decreased significantly in built-up environments. This is consistent with the z-scores (Table 7), which indicate HAWTs’ high sensitivity to directional variations and turbulent flows.

VAWT-D (Darrieus) and BIWTs achieved moderate yields but demonstrated good aesthetic and social acceptability in urban locations. Resistance turbines (VAWT-S, CAWT) exhibited the lowest AEP, but their reliability, low noise level, and resistance to turbulence make them an attractive choice for educational and off-grid applications.

User comfort and environmental impact are key selection factors

In urban environments, where users remain in the immediate vicinity of turbines, factors such as noise, stroboscopic effects, structural resonance, and aesthetics often play a greater role than energy production itself [17,22]. The results (Table 8) showed that

• BIWT and CAWT turbines were characterized by the highest user comfort ratings, with minimal noise and good architectural integration.
• HAWT and DAWT, despite their high efficiency, generated significant tonal noise (50–65 dB, evaluated as LAeq at 10 m), which could lead to neighbor conflicts if inappropriately located.
• VAWT-S (Savonius) demonstrated the lowest noise levels (evaluated as LAeq at 10 m) and the highest social acceptance—confirming the data from [48].

In the multi-criteria assessment (MAWT-Score, Table 9), the DAWT achieved the best result (0.75), demonstrating its balanced efficiency and comfort characteristics. BIWT and CAWT achieved almost equally high scores, thanks to their high architectural integration and low acoustic impact. HAWT performed the least well, as their effectiveness is limited to open locations.

Applying the MAWT-Score Methodology in Design Practice

The developed MAWT-Score indicator (Table 9), along with a radar chart with nine qualitative assessment criteria, enables quick comparison and selection of the appropriate turbine type for a given location. Importantly, the adopted weighting system can be adjusted depending on priorities:

• A commercial investor can prioritize energy efficiency.
• A municipality or school can focus on social acceptance, aesthetics, and educational potential.
• A designer can assess compliance with building regulations and noise restrictions.

The flexibility of the method allows for its application in a variety of contexts: from single-family homes to large-scale educational campuses. This model can also be used as a decision-support tool in construction notification and public consultation procedures.

Model Limitations and Directions for Further Research

Despite its high level of agreement with real-world data, the MAWT-Score model has certain limitations:

• Meteorological data were based on annual wind speed histograms and Weibull parameters, which do not account for short-term variability and extremes.
• The influence of adjacent dynamic obstacles (e.g., deciduous trees that seasonally change their aerodynamic structure) was not considered.
• User comfort was assessed based on the literature and expert data, without formal public perception studies.
• Turbine acoustics depend on the specific manufacturer’s model and can vary significantly even within the same technology class.

In the future, it is worthwhile to expand the model to include the following elements:

• Incorporating hourly and seasonal data;
• Integration with PV and heat pump models in hybrid systems;
• Assessment of LCOE (levelized cost of energy) in the context of the economic viability of different MAWT types;
• Development of a decision-making application (e.g., in the form of an interactive online platform for prosumers).

The discussion of the results confirms that the effective selection and implementation of a MAWT requires a multi-criteria approach, taking into account both technical parameters, user comfort, and local installation conditions. The proposed methodology allows for adapting the technology to actual wind and urban conditions. This increases the chances of investment success and public acceptance of the installation. DAWTs and BIWTs demonstrate the greatest potential, in selected scenarios, combining good energy performance with high social acceptance and urban adaptability. At the same time, the analysis highlights the importance of an individual design approach—there is no one-size-fits-all turbine suitable for all locations. From a practical perspective, the synthesis of performance, acoustic, and structural criteria suggest that not all MAWT concepts are equally suited for dense urban environments. Vertical axis turbines of the Savonius type, while less efficient, provide high turbulence tolerance and low acoustic impact, making them viable for rooftop or façade integration. Compact Darrieus or ducted concepts (DAWT, CAWT) may also be considered where building codes permit, provided vibration isolation is ensured. By contrast, horizontal axis turbines (HAWT) and larger blade-integrated designs (BIWT) achieve higher yields but are more appropriate for peri-urban or open-field locations, where acoustic and visual impacts are less restrictive.

5. Conclusions

This paper presents a comprehensive methodology for selecting micro- and small-scale wind turbines (MAWTs) for local energy systems, taking into account energy efficiency, user comfort, and adaptation to urban and climatic conditions. The originality of the work lies in the integration of technical yield assessment with user comfort and environmental acceptance criteria, providing a structured framework for real-world applications in urban energy systems. Based on the conducted analyses, the following key conclusions can be drawn:

• MAWT efficiency strongly depends on location, mounting height, and turbulent airflow characteristics. Locations with high building height and good wind exposure (B5–B6) enable achieving annual energy yields (AEP) exceeding 1800 kWh for 2 kW turbines, while in single-family homes (B1), these values fall below 900 kWh/year (Table 7).
• The turbine type should be closely matched to the environmental conditions. DAWTs and HAWTs achieve the highest AEP values in locations with stable airflow, but require good exposure (Table 7). In the built environment, BIWTs, CAWTs, and VAWTs are the better choice, demonstrating greater tolerance to turbulence and improved user comfort (Table 8).
• Environmental impact and user comfort play a key role in the acceptance of MAWT technology in residential environments. Low noise levels, lack of shadow flicker, and architectural integration increase public acceptance, especially in densely built-up urban areas. Savonius, BIWT, and CAWT turbines achieved the highest comfort and aesthetic ratings (Table 8).
• The proposed MAWT-Score assessment methodology, supplemented by a radar chart for nine qualitative criteria (Figure 4), enables a transparent and scalable assessment of various turbine types. This framework allows tailoring technology selection to the investor’s needs, local conditions, and social and legal requirements.
• The assessment model can be expanded to include additional aspects, such as levelized cost of energy (LCOE), integration with PV and storage, and life-cycle assessment (LCA) for a comprehensive sustainability assessment. This opens up opportunities for integrating the method into planning tools and decision-making applications for prosumers and local government units.
• The MAWT industry requires further standardization and research into micro-scale environmental impacts. This particularly applies to acoustics, structural vibrations, and the stroboscopic effect in urban environments, which are currently not clearly regulated but can significantly impact the perception of installations (Table 8).
• Although the methodology was developed using wind and urban conditions typical of Central Europe, it is applicable in other climatic regions provided that key input parameters—such as terrain roughness, turbulence intensity (TI), and atmospheric stability—are localized to reflect site-specific conditions.

In summary, the presented approach offers a practical and scalable tool for supporting decision-making in the selection and implementation of MAWTs in distributed systems. Taking into account actual wind conditions, comfort aspects, and urban integration can increase the efficiency, safety, and acceptance of MAWTs in the building environment. The presented model is a step towards sustainable and socially acceptable integration of wind energy in urban space.