Parametric study on a swiss location for alpine PV systems

^{Created by R.G., ZHAW, on 07.05.2024}

^{Last updated by R.G., ZHAW, on 09.09.2024}

This page gives an overview about different simulation results about bifacial PV systems carried out for an alpine location in Switzerland. For that, parameters such as Ground Cover Ratio (GCR), azimuth, module tilt, Slope of Ground (SoG) of a system are changed and analyzed. The simulations are done with the software tool PVsyst [1].

Structure of the simulations

The simulations use a representative system, with azimuth, GCR and SoG values that can be changed. This illustration shows an example with a vertical system, including tables of 10 modules, with eight tables in a row and 40 rows. Each tile of this cone represents an individual orientation. At each one of them, the module tilt and GCR are varied.

The GCR is a value used to define the distance between the rows relating to the height of the used module tables. The following examples illustrate how this value is measured and how the proportions of 0.5, 1 and 1.3 look.

Illustration of the simulation system boundary, which was taken into account in the following assessment. Only the yield up to the inverter was included, as the subsequent losses are dependent on the respective system.

The representative location for Davos Toalp is chosen due to the existing experimental PV plant, which has been installed and measured since 2017. The location is only relevant for weather data and the horizon profile. The realistic topography is not considered for the simulations.

The yield of an Alpine PV system was simulated using meteo data from Meteonorm. The data for a typical year of Davos Totalp and Winterthur was compared, showing that global irradiation on a horizontal plane is approximately 50% higher in this alpine region.

The parametric study involves changing the parameters as shown in the table below. Each combination is simulated for one year in hourly resolution, resulting in 675 simulations per batch. The batch is then calculated a second time without a horizon profile for more general statements. The modules are connected in a string via two tables, as shown in the figure below the table.

Using bifacial modules in alpine PV systems the reflected irradiation and direct irradiation on the back can be used. The albedo value describes how well the light is reflected back (1 – full reflection, 0 – no reflection). According to the table shown here new snow can have a high albedo of at least 0.8 and grass a low albedo of 0.2. Depending on the snowheight measurement in Davos the monthly values of the albedo can be estimated.

Simulation results

Annual specific yield

The simulation results show that for a complete east or west orientation the 90° tilted modules have the highest specific yield. With the increase in the SoG the GCR could be increased to and reach the same amount of specific yield as the lower GCRs. According to the simulation results the vertical system with high SoG has the highest annual specific yield.

Orienting the system in south-east or south-west direction the 60° module tilt for lower SoG and 75° module tilt for higher reach highest specific yield.

Around south orientation the module tilt of 60° is leading before the other angles of inclination. With a small GCR one can reach the highest annual specific yield for this azimuth.

Winter specific yield

According to recommendations of the SFOE [4] a specific yield of at least 500 kWh/kWp has to be reached over the winter half year (october – march) at the measurement point in average over three measurement years. Since the results of the simulations were carried out without a horizon profile and are only considered on the DC side, only system designs that achieve more than 700 kWh/kWp are presented here.

Overview of the systems with highest and smallest share of specific yield during winter months. The first line presents the shares of the lowest share over all simulations. The middle line is showing the shares of the lowest values for the vertical system and the third line shows the best share over all simulations.The closer a system is orientated towards the south, the higher the winter share of the specific yield. Steeper module angles also achieve higher winter shares, as the rear side is better illuminated.

With steep SoG, one can arrange the rows up to a high GCR depending on the module inclination. However, as the SoG decreases, the GCR must also decrease in order to remain above 700 kWh/kWp. Systems which have a SoG of 0° cannot reach the limit.

For the south-east/west systems all three module tilt have the same GCR limit to stay above 700 kWh/kWp. But it is interesting to see that for lower GCRs the 60° tilt is slightly better then the other two and with increasing GCR the 90° tilt gets better then the rest. Here as well the horizontal system cannot reach the boundary of 700 kWh/kWp.

For south orientations horizontal systems can reach the threshold but only with a GCR of 0.5. With this orientation, larger GCRs can be selected for steep slopes. Furthermore, it has been shown that the 90° module inclination can achieve a higher specific yield with narrower systems.

Annual area specific yield

In order to qualify for the subsidy for the large-scale PV system, a second target must be achieved in accordance with the SFOE’s recommendations [4]. On average over the three measurement years, at least 10 GWh of energy must be fed into the grid. Optimising the system based on the specific yield could result in a significant proportion of the energy that could be generated in a given Alpine region being lost. Consequently, the analysis was also conducted on a yearly basis, taking into account the area-specific yield.

If the energy output is insufficient for an east- or west-facing system, it may be beneficial to increase the GCR. As illustrated in the graph, the output per square metre of system area could be significantly enhanced. In this orientation with a low GCR, the inclination of the modules has minimal impact, but as the GCR increases, the vertical installation of the modules becomes more pronounced.

The system orientations in question have resulted in a stagnation of the higher yields at high GCRs. Therefore, it is not advisable to position the modules closer than a GCR of 1 on flat undergrounds. It is also not possible to make a clear statement regarding the module inclination, as the optimal angle would depend on which GCR is chosen and which inclination of the ground is available. However, the 60° inclined systems tend to achieve higher area specific yields.

Flat south-facing systems show a marginal increase in area-specific yield even with small GCRs. Nevertheless, the maximum remains at a GCR of 1.3 for all substrates. Even with this orientation, the stagnation of the area specific yields begins around a GCR of 1; depending on the inclination of the underground, this limit can vary.

Impact of the horizon

With regard to the realistic horizon at the selected location, the horizon profile appears to be as shown on this slide. It indicates that there is a mountain in the west that significantly impairs direct sunlight throughout the year. The individual images show how this profile of Meteonorm is generated and which orientations have to deliver lower yields.

The graphs illustrate the impact of the horizon on specific yield. In this case, an eastern orientation of the system is recommended to minimise losses, with a maximum of 2.5% compared to a 5%-26% loss with a western orientation.

If the systems are orientated south-east or west, the maximum losses increase in both directions. The western system would therefore be directed exactly in the direction of the local high mountain peak, where the maximum losses occur.

The pure south orientation still achieves losses of 5% – 20% depending on the available substrate. These losses can be minimised if the orientation of the modules is selected using the horizon profile. What can be seen in these graphs is that the vertical systems lose more specific yield than systems with tilted modules.

Reliability of the simulation results

The reliability of these results is assessed by comparing them with measurements from the experimental plant of ZHAW Wädenswil [7] over the last five years, which are located in Davos, Switzerland. It should be noted that these are free-standing modules, so the values can only be compared with a system where the self-shading caused by individual rows is minimised. In contrast to the previous data, the simulated energy was not used at the inverter input (DC side) for this comparison but at its output (AC side). However, the inverter used in the simulation differs from the one used in reality, which is likely to be a contributing factor to the difference. It should also be noted that the same weather data was not available as in the measurement years. Only a typical year was used. The simulations were based on the assumption that the systems are not affected by snow cover. In reality, according to the experience of the ZHAW Wädenswil, this can amount to 1-2% for steep module angles.
The graph shows that the simulation of a 60° tilt system with a SoG of 40° leads to a higher yield than the actual measurements. Basically, the summer months (55.2%) are overestimated, while winter (44.8%) is underestimated. In fact, the simulation with 75° inclined modules is also below the actual measurements at a SoG of 40°. Here, the simulation underestimates the achievable specific yields. The same applies to the vertical system. It also simulates less specific yield in winter than was measured in Davos. Furthermore, the vertical system demonstrates a greater tendency to underestimate the actual specific yield in comparison to the 60° and 75° inclined modules. The aforementioned differences can be attributed to two key factors: the discrepancies already discussed and the inherent uncertainties associated with PVsyst modelling. These include the selected albedo, the limitations in calculating the bifacial component and other special effects, such as edge effects. It is important to note that this comparison is only intended to serve as a preliminary assessment of the simulation and is not intended to be a comprehensive verification.

Reliability assessment of the simulated system with GCR 0.5 mean value (bars) of all SoG with the measurements (lines) of the experimental free standing plant of ZHAW Wädenswil [6]. The errorbars are showing the minimum value for the SoG of 0° and the maximum value for a SoG of 40°.

Reliability assessment of the simulated system with GCR 0.5 mean value (bars) of all SoG with the measurements (lines) of the experimental free standing plant of ZHAW Wädenswil [6]. The errorbars are showing the minimum value for the SoG of 0° and the maximum value for a SoG of 40°.

Conclusion

Horizon

All the parameters which were changed during this study have a cuucial impact in how the PV system on a Alpine location should be designed. To understand how the horizon profile is impacting the course of the sun and therefore the specific yield of a system is a key to find the optimal solution. According to the example used here it would make sense to orient the modules to the east (-90°), of course this could be different for other location. The example provided here indicates that orientating the modules to the east (-90°) may be a suitable approach, although this may vary depending on the location.

Specific yield and area specific yield

It is not recommended to optimize a System only on the parameter of specific yield because that mor energy could be produced within the available space. According to this the best System would be the one with les self-shading, therefore with e small GCR. Depending on the orientation and inclination of the substrate, however, up to twice as much energy could be obtained per available square metre of system area. With a higher GCR, more modules can therefore be placed on the same area, which leads to a higher area-specific yield. If the module prices are not the cost-driving factor, it makes sense to design the system in such a way that the optimum between the specific and the area-specific yield is achieved. These two KPIs can therefore be used to ensure that the system sufficiently meets/exceeds the federal government’s targets of 500kWh/kWp (winter half-year) and 10 GWh.

Module tilt and orientation

Azimuth of -90°/90°

This parametric study results show that the module tilt for the optimal system is dependent of the chosen or available orientation in the alpine region. Orienting the system -90°/90° the simulations show that a vertical system will have highest specific and area specific yield. With Increasing SoG it is possible to place the rows with a short distance between each and reach a specific yield in winter that will be more than 700kWh/kWp. The lower SoGs are shifting the boundary for reaching this value during winter more to smaller GCRs. In the case of a horizontal underground the system will not be able to reach this aim. For this orientation, the optimal winter production share is 39.8% of the annual specific yield, achievable with a SoG of 40° and 90° tilted modules. The minimum winter share is observed with the horizontal system and 60° tilted modules, at 37.9%.

Azimuth of -45°/45°

Annual specific yield has been observed to decrease with the rotation of the systems to an azimuth of -45°/45°. However, this reduction is offset by an increase in yield during the winter half year. The 60° tilted modules have been found to perform better than the others, with the exception of a SoG of 40°, where the 75° tilted modules could potentially offer a more favourable outcome in the simulation environment. Consequently, the discrepancy between the module tilt angles is reduced when the specific yield at the winter half year is considered. It can be seen that the horizontal systems will not reach the threshold, and with a low SoG of 10°, only GCRs up to 0.6 could be used. This would increase the size of the overall system to reach 10 GWh annually. For steeper SoGs, it is possible to design systems with higher GCRs than 1 in order to reach the winter half-year aim.

Azimuth of 0°

The use of an azimuth of 0° for the PV system allows for the maximisation of the winter half-year share for all types of module inclination. This is evidenced by the fact that vertical modules on a SoG of 40° have the highest share at 54.1%. The horizontal system with vertical modules can achieve up to 52.2%, while the horizontal 60° system has the lowest share of 47.5% across all module inclinations. In the south direction, even the horizontal system could reach 700 kWh/kWp if it has a low GCR and self-shading is reduced to a minimum. However, the graph shows that in winter, all slopes produce similar specific yields, with the differences in the relative proportion mainly due to the different specific yields in summer. From an annual perspective, a module inclination of 60° stands out and achieves the highest specific yield.

Winter share

The parametric study results show that the module tilt for the optimal system is dependent on the chosen or available orientation in the alpine region. Orienting the system at -90°/90° results in the highest specific and area-specific yield in the simulations. As the SoG increases, it becomes possible to place the rows with a short distance between each, allowing the system to reach a specific yield in winter that will be more than 700 kWh/kWp. The lower SoGs shift the boundary for reaching this value during winter to smaller GCRs. In the case of a horizontal underground system, this aim cannot be reached. The optimal share of winter production for this orientation is 39.8% of the annual specific yield, which can be achieved with a 40° and 90° tilted module system. As a minimum, 37.9% of the annual yield can be achieved with a horizontal system and 60° tilted modules.

Adaption to other locations

The simulations were conducted without a horizon to ensure independence from the location in question. However, this is only one of several parameters that describe the selected location, Davos Totalp. The weather conditions in each alpine location are unique, so it is important to ensure that the results are accurate. One way to do this is to compare the annual and seasonal irradiation to identify any differences in global horizontal irradiation. If the value is higher, it can be assumed that the PV system will yield more. Conversely, if the yield is lower, the opposite would be the case.

References

• [1] PVSyst V7.4.6.
• [2] Google Maps
• [3] Meteonorm V7.3
• [4] SFOE, «Requirements for calculating the energy yield for large-scale photovoltaic systems in accordance with Art. 71a EnG» (German), Ittigen 2023
• [5] Harvey S.,»Weissfluhjoch – 2536 m.a.s.l.», whiterisk.ch, https://whiterisk.ch/en/conditions/measurements/station/IMIS/WFJ2 (accessed 25.04.2024)
• [6] Bucher Ch., Photovoltaikanlagen, Faktor Verlag 2021
• [7] Anderegg, D., Strebel, S., & Rohrer, J. (2023). Alpine Photovoltaik Versuchsanlage Davos Totalp: Erkenntnisse aus 5 Jahren Betrieb. ZHAW Zürcher Hochschule für Angewandte Wissenschaften, IUNR Institut für Umwelt und Natürliche Ressourcen. https://digitalcollection.zhaw.ch/handle/11475/28797