Vertical vs. Horizontal Axis Small-Scale Wind Turbines Serving a Typical House in Italy and Norway: Numerical Energy, Environmental and Economic Assessments

This section of the paper describes the electric demand of the residential building assumed as reference (Section 2.1), the characteristics of the selected horizontal and vertical SWTs (Section 2.2), as well as the mathematical models adopted for simulating the SWTs as well as the climatic conditions (Section 2.3).

2.1 Building electric demand

It is commonly recognized that a wide range of factors influence how much electricity homeowners use. In order to predict the corresponding daily electric energy demand profiles linked to the operation of lighting systems and household appliances, an original tool created by Loughborough University based on a stochastic method [20] has been applied in this work. The maximum number of occupants, the day of the week or weekend, the month of the year, and the quantity and kind of household appliances can all be used to generate these daily profiles. The electric demand for cooking appliances, water heating, and air conditioning systems is not taken into consideration by this tool. The associated daily electric profile is then computed based on the actual number of people and the random activations of the lighting and household appliances. Profiles that may be regarded as representative of common homes can be obtained using this method [21]. In this study, in particular, a maximum of 4 people is assumed and the most commonly used domestic appliances have been considered. The annual electric demand of the house considered assumed as reference in this study has been defined by combining 365 different daily electric demand profiles obtained via the above-mentioned tool [20] with a time-step equal to 1 minute. Figure 1 reports both the annual electric demand profile as a function of the time (black curve) as well as the corresponding electric load-duration diagram with the values sorted in descending order (red curve). The annual electric energy required by the building assumed as reference is 2408.96 kWh.

1.png

Figure 1. Annual electric demand profile and load-duration profile of the residential building

2.2 Wind turbines

Three commercial SWTs (a Savonius SWT with vertical axis [13], and two SWTs with horizontal axis [14, 15]) have been investigated in this work. Table 1 reports the axis type (horizontal or vertical), manufacturer, model, power output at 12 m/s, rotor swept area, rotor diameter, blades height (for the VASWT only), number of blades, cut-in wind speed (i.e., the minimum wind speed at which the SWTs begin providing an usable electric power), cut-out wind speed (i.e., the highest wind speed at which the SWTs are intended to produce useable electric energy), and capital cost of the selected SWTs. The rotor swept area of the HASWTs is given as pR² (where R is the rotor radius), while for the Savonius VASWT it is calculated as DH (where D is the rotor diameter (equal to 0.80 m) and H is the blades height (equal to 2.0 m)) [12]. This table indicates that the rotor swept area of the HASWT FK-2000 is 3.8 and 4.3 times larger than that one of the VASWT FS-2000 and the HASWT ATO-WT-NE-300S5, respectively. Figure 2 reports the power curves of the selected SWTs according to the data provided by the manufacturers; these curves show the operating ranges of both wind speed (from the cut-in up to the cut-out) and power output of the SWTs. This figure shows that:

• for a given wind speed, the HASWT ATO-WT-NE-300S5 generates much less power than the other two SWTs;
• the HASWT FK-2000 produces more electric output than the VASWT FS-2000 in the wind speed range between about 6.5 and 11 m/s.

Table 1. Selected SWTs

2.png

Figure 2. Power curves of the selected SWTs [13-15]

2.3 Simulation model and climatic data

The TRaNsient SYStems simulation tool (TRNSYS) [19] (version 18) has been adopted in this paper in order to model and simulate the SWTs performance and related climatic conditions. Taking into account that it considers the part-load operation of the energy conversion systems as well as the interactions between electric demand and generation, the platform TRNSYS is usually utilized by the scientific community [22-24]. The TRNSYS library includes a number of mathematical models (called “Types”). In this work, the TRNSYS Type 90 has been considered to model and simulate the performance of the SWTs. This model needs the definition of some parameters (site elevation, data collection height, hub height, turbine power loss) and inputs (wind velocity, ambient temperature, site shear exponent, barometric pressure, control mode, rotor height, rotor diameter, sensor height, turbulence intensity, air density, power rated, speed rated, power curve).

According to the above-mentioned parameters/inputs, the model allows to calculate 3 outputs (power output, power coefficient and time of continuous wind turbine operation). Table 2 summarizes the values assumed in this study for the parameters and inputs of the TRNSYS Type 90 for the SWTs. The SWTs power loss has been considered equal to 0, while a site shear exponent of 0.26 has been adopted with the aim of modelling the scenario corresponding to obstructed airflows (according to the study [25]). The option “P” (i.e., pitched-control) has been selected as control mode for both the HASWTs FK-2000 and ATO-WT-NE-300S5 taking into account that the initial outward pitching manoeuvre serves to reduce the blade’s effective angle of attack, delay the moment when the critical static stall angle is exceeded, and reduce the maximum effective angle of attack [26]. The turbulence intensity is a dimensionless number defined as the standard deviation of wind speeds within a simulation time step divided by the average wind speed over that simulation time step; in this study a value of 10% [27] has been assumed as turbulence intensity, whatever the SWT model is. The weather conditions (site elevation, wind velocity, ambient temperature as well as barometric pressure) required by the TRNSYS Type 90 have been set based on the Typical Meteorological Year version 2 weather database (TMY2) [28] via the TRNSYS Type 15-6. This Type is a weather data processor reading an external weather data file and providing climate conditions representative of the selected cities. In this paper, 5 cities in Italy (Alghero, Milan, Naples, Palermo, Rome) and 3 cities in Norway (Bergen, Karasjok, Tromsø) have been considered as installation sites for the SWTs to take into account the influence of different climatic scenarios on SWTs performance. These cities are characterized by the following coordinates: Alghero (latitude 40° 33' 55.48" North, longitude 8° 19' 15.31" East); Milan (latitude 45° 27' 51.1596'' North, longitude 9° 11' 28.9788'' East); Naples (latitude 40° 51' 11.8584'' North, longitude 14° 18' 20.0628' East); Palermo (latitude 38° 7' 0.0084'' North, longitude 13° 22' 0.0012'' East); Rome (latitude 41° 54' 10.0152'' North, longitude 12° 29' 46.9176'' East); Bergen (latitude 60° 23' 34.4736" North, longitude 5° 19' 27.7788'' East); Karasjok (latitude 69°28' 18" North, longitude 25° 30' 40'' East); Tromsø (latitude 69° 39' 0.9108'' North, longitude 18° 59' 42.4140'' East).

Table 2. Parameters and inputs of the TRNSYS type 90

Figure 3 reports the annual wind velocity-duration diagrams upon varying the city (according to study [28]), with the data in descending order. This plot indicates that Palermo is characterized by the largest average annual wind speed (4.24 m/s), while the minimum average annual wind velocity (1.14 m/s) corresponds to the city of Milan.

3.png

Figure 3. Annual wind velocity-duration diagrams

In this paper the performance of the proposed scenarios where the building is served by both the SWTs and the central grid (as back-up) have been compared with those corresponding to a baseline scenario corresponding to the case where the electric demand of the same building is totally covered by the central grid only (without the SWTs). This comparison has been performed with the aim of determining the potential benefits in terms of electric energy imported from the grid (Section 4.1), equivalent global CO₂ emissions (Section 4.2) and operating costs (Section 4.3); the simple payback period has been also calculated (Section 4.3).

4.1 Energy comparison

The percentage difference $\Delta \mathrm{E}_{\mathrm{el}}$ between the values of annual electric energy imported from the grid $E_{{el,imp }}^{P S}$ in the cases of the proposed scenarios (including the SWTs) with respect to the case of the reference scenario (without the SWTs) $E_{e l, i m p}^{R S}$ has been calculated according to the following formula:

$\Delta \mathrm{E}_{\mathrm{el}}=\frac{\mathrm{E}_{\mathrm{el}, \mathrm{imp}}^{\mathrm{PS}}-\mathrm{E}_{\mathrm{el}, \mathrm{imp}}^{\mathrm{RS}}}{\mathrm{E}_{\mathrm{el}, \mathrm{imp}}^{\mathrm{RS}}}$                   (1)

Figure 8 reports the values of ΔE_el as a function of both the wind turbine model and the city. All the values reported in this figure are negative; according to Eq. (1), this means that the proposed scenarios with the SWTs allow to reduce the electricity imported from the grid with respect to the reference scenario, whatever the wind turbine model and the city are. In greater detail, this figure indicates that, for a given city, the adoption of the SWTs allows to reduce the electric energy imported from the grid from a minimum of –0.4% in the case of the HASWT ATO-WT-NE-300S5 operating in Milan up to a maximum of –37.5% in the case of the VASWT FS-2000 operating in Palermo.

4.2 Environmental comparison

This work evaluated the environmental impacts by means of the energy output-based emission factor method proposed by Chicco and Mancarella [29]; it allows to estimate the global equivalent mass m_x of a given pollutant x emitted while producing the energy output E according to the following formula:

$\mathrm{m}_{\mathrm{x}}=\mathrm{u}_{\mathrm{x}}^{\mathrm{E}} \cdot \mathrm{E}$                   (2)

where, $u_x^E$ is the energy output-based emission factor of x per unit of E. CO₂ emissions generally show to be quantitatively more significant than emissions of other pollutants. The CO₂ emission factor associated to the electricity generation in Italy $u_{\mathrm{CO}_2}^{E_{\mathrm{el}}}$ depends on the location, the day of the year as well as the time of the day. According to the values suggested in the studies [30, 31], this factor ranges between 41.3 gCO₂/kWh_el and 827.3 gCO₂/kWh_el for the Italian cities and between 0 and 56.1 gCO₂/kWh_el for the Norwegian cities. The percentage difference $\Delta \mathrm{CO}_2$ between the values of global equivalent CO₂ emissions in the case of the proposed scenarios (including the SWTs) with respect to the values of the reference scenario (without the SWTs) has been derived as follows:

$\Delta \mathrm{CO}_2=\frac{\mathrm{CO}_2^{\mathrm{PS}}-\mathrm{CO}_2^{\mathrm{RS}}}{\mathrm{CO}_2^{\mathrm{RS}}}=\frac{\sum_{\mathrm{i}} \mathrm{u}_{\mathrm{CO}_{2, \mathrm{i}}}^{\mathrm{E}_{\mathrm{el}}} \cdot \mathrm{P}_{\mathrm{el}, \mathrm{imp}, \mathrm{i}}^{\mathrm{PS}} \cdot \mathrm{STS}-\sum_{\mathrm{i}} \mathrm{u}_{\mathrm{CO}_{2, \mathrm{i}}}^{\mathrm{E}_{\mathrm{el}}} \cdot \mathrm{P}_{\mathrm{el}, \mathrm{imp}, \mathrm{i}}^{\mathrm{RS}} \cdot \mathrm{STS}}{\sum_{\mathrm{i}} \mathrm{u}_{\mathrm{CO}_{2, \mathrm{i}}}^{\mathrm{E}_{\mathrm{el}}} \cdot \mathrm{P}_{\mathrm{el}, \mathrm{imp}, \mathrm{i}}^{\mathrm{RS}} \cdot \mathrm{STS}}$                 (3)

where, STS is the simulation time step (equal to 1 minute), and $u_{C O_{2, i}}^{E_{e l}}$ is the i-th CO₂ emission factor associated to the i-th electric power imported from the grid in the case of the proposed scenario ($P_{e l, i m p, i}^{P S}$) or in the case of the reference scenario ($P_{e l, i m p, i}^{R S}$) at the same simulation time. Figure 9 underlines the results in terms of ΔCO₂ upon varying both the city as well as the SWT model. The data are always negative and, therefore, the proposed scenarios including the SWTs allow to decrease the global equivalent CO₂ emissions with respect to the reference scenario, whatever the city of the SWT model is. In more detail, this figure demonstrates that, for a given city, the utilization of the SWTs reduces the global equivalent CO₂ emissions in comparison to the reference scenario from a minimum of –0.4% in the case of the HASWT ATO-WT-NE-300S5 in Milan up to a maximum of –37.8% associated to the VASWT FS-2000 in Palermo.

8.png

Figure 8. ΔE_el upon varying the SWT model and city

9.png

Figure 9. ΔCO₂ upon varying the SWT model and city

4.3 Economic comparison

The percentage difference ΔOC between the values of operating costs (associated to the electricity imported from the grid) in the cases of the proposed scenarios (including the SWTs) with respect to the values of operating costs OC^RS of the reference scenario (without the SWTs) has been calculated via this formula:

$\Delta \mathrm{OC}=\frac{\left(\mathrm{OC}^{\mathrm{PS}}-\mathrm{REV}_{\mathrm{el}, \text { sold }}\right)-\mathrm{OC}^{\mathrm{RS}}}{\mathrm{OC}^{\mathrm{RS}}}=\frac{\sum_{\mathrm{i}}\left(\mathrm{UC}_{\mathrm{el}, \mathrm{imp}, \mathrm{i}} \cdot \mathrm{P}_{\mathrm{el}, \mathrm{imp}, \mathrm{i}}^{\mathrm{PS}}-\mathrm{UC}_{\mathrm{el}, \text { sold }, \mathrm{i}} \cdot \mathrm{P}_{\mathrm{el}, \text { sold }, \mathrm{i}}^{\mathrm{PS}}\right)-\sum_{\mathrm{i}} \mathrm{UC}_{\mathrm{el}, \mathrm{imp}, \mathrm{i}} \cdot \mathrm{P}_{\mathrm{el}, \mathrm{imp}, \mathrm{i}}^{\mathrm{RS}}}{\sum_{\mathrm{i}} \mathrm{UC}_{\mathrm{el}, \mathrm{imp}, \mathrm{i}} \cdot \mathrm{P}_{\mathrm{el}, \mathrm{imp}, \mathrm{i}}^{\mathrm{RS}}}$                          (4)

where, REV_el,sold is the annual revenue obtained thanks to the electric energy sold to the grid, UC_el,sold,i and UC_el,imp,i are, respectively, the unit price of electric energy sold to the grid and the unit cost of electric energy imported from the grid. In the case of the Italian cities, the values of UC_el,imp,i have been obtained according to the values indicated in the study [32] by considering the zone of Italy where the city is located (Central-South, North, Sicily, Sardinia). With reference to Norwegian cities, the values of UC_el,imp,i have been defined according to those suggested by the federation of the European electricity industry based on the zone of Norway (Norway_4 and Norway_5) [33] where the considered cities are located. The values of UC_el,sold,i for the selected Italian cities has been adopted according to the study [34]. In the cases of the Norwegian cities, the unit price of electricity sold to the grid UC_el,sold,i has been assumed equivalent to the unit cost of electricity purchased from the grid UC_el,imp,i when the electrical energy production exceeds the demand of the individual residential building [35]. According to studies [32-35], the maximum UC_el,imp,i and UC_el,sold,i for the Italian cities is equal to 400 €/MWh and 170 €/MWh, respectively, while for the Norwegian cities they are both equal to 332 €/MWh. Figure 10 shows the results in terms of $\Delta$OC upon varying both the city and the SWT model. This figure reports only negative values (whatever the city or the SWT model is), so the proposed scenarios with the SWTs always allow to decrease the operating costs in comparison with the reference scenario. In further detail, this plot underlines that, for a given city, the utilization of the SWTs decreases the operating costs in comparison to the reference scenario from a minimum of –0.4% in the case of the HASWT ATO-WT-NE-300S5 operating in Milan up to a maximum of –112.0% in the case of the VASWT FS-2000 operating in Tromsø.

The utilization of SWTs require a larger investment cost in comparison to the reference scenario. The simple payback period (SPB) represents the period that is needed to recover the extra investment thanks to both the savings in terms of operating costs as well as the revenues corresponding to the electricity sold to the grid; it can be obtained as follows:

$\mathrm{SPB}=\frac{\mathrm{WT}^{\mathrm{CC}}}{\left(\mathrm{OC}^{\mathrm{PS}}-\mathrm{OC}^{\mathrm{RS}}\right)+\mathrm{REV}_{\mathrm{el}, \text { sold }}}$                    (5)

where, WT^CC is the capital cost of SWTs.

10.png

Figure 10. ΔOC upon varying the SWT model and city

Figure 11 reports the values of SPB upon varying both the city and the SWT model. This plot highlights that SPB ranges from a minimum of 2.8 years (obtained when the HASWT FK-2000 operates in Palermo) up to a maximum of 241.8 years (that is achieved for the HASWT ATO-WT-NE-300S5 installed in Milan). The HASWT FK-2000 and the VASWT FS-2000 are characterized by similar SPBs, that can be assumed as suitable from an economic point of view (except in the case of Milan) taking into account that the expected SWTs lifetime is about 20÷25 years; the HASWT ATO-WT-NE-300S5 is characterized by much larger SPBs in contrast to the other SWTs, with acceptable values only when it is installed in Palermo, Rome and Alghero.

11.png

Figure 11. SPB upon varying the SWT model and city