Keywords

1 Introduction

The promotion of Renewable Energy Sources (RES) is a pillar of the European strategy for climate and energy (EC—COM/2014/015) and in general of European sustainable development.

Among RES, photovoltaic is, in Italy, the one with the highest growth in the last decade. Due to strong national promotion policy, only in four years (2008–2011) the average increase of photovoltaic production was around 300%. In 2016 photovoltaic with a production of more than 22 TWh covered almost the 8% of the total national electricity production.

In this framework it is essential to understand if RES, and in particular photovoltaic solutions, whereas contributing to climate change mitigation have the potential to reduce or to increase the contribution of energy system to other environmental impact categories (e.g. air acidification, particulate matter formation potential). A suitable methodology to face this problem is of course the LCA—Life Cycle Assessment—(Sumper et al. 2011). LCA has been widely applied in the field of photovoltaic systems assessment as demonstrated by several literature review studies (Peng et al. 2013). However, the same review studies have underlined the need for further research. In particular there is the need to enlarge the system boundaries because often end of life impacts are not investigated. Moreover there is the need to increase the number of the considered environmental impact categories since, usually, they are only limited to greenhouse effect and to energy payback time (Gerbinet et al. 2014).

In the following paragraphs, after a brief description of the Italian photovoltaic system evolution, we will discuss the LCA of different photovoltaic solutions following the ISO 14040 scheme.

2 Evolution of Photovoltaic Production in Italy

The production of electricity from photovoltaic increased in Italy in the last decade (Fig. 2.1). Due to strong national promotion policy, only in four years (between 2008 and 2011) the average increase of photovoltaic production was around 300%. Photovoltaic has covered 18.5% in 2014 and 20% in 2016 of the electricity produced by renewables (GSE 2015), compared to less than 1% in 2009. In 2016 with a production of more than 22 TWh (almost the 8% of the national electricity production) it was the second renewable energy sources, after hydropower.

Fig. 2.1
figure 1

Source GSE (2015)

Production from photovoltaic plant.

Full size image

As regards the size of the plants, recent years show a trend towards smaller plants. During 2014 for example, new installations were essentially residential with an average power of 8.1 kW, considerably lower than the past years. The average plant size, in fact, was three times higher in 2012 and six time higher in 2011. As regard the installation options, almost 40% of the installed power is ground mounted, almost 50% is on buildings (mainly on roof) and 6% is mounted on greenhouses or canopies (such as in car parking). The remaining 4% covers different installation options, sometime really interesting like those on the highway noise barriers (GSE 2015).

Concerning cells technologies and materials, single-crystalline silicon panels have decreased in favour of multi-crystalline panels. In all the Italian regions the multi-crystalline silicon panels cover the majority of the installed power, followed by single-crystalline silicon. Other technologies like thin film cover a little percentage of the installed power. According to GSE (2015), in 2014 more than 72% of the installed power at national level is in multi-crystalline silicon panels, 21% in single-crystalline silicon panels, while remaining technologies accounts for only 7% of the installed power.

From a technology point of view a decisive evolution towards more efficient and low impact solutions is likely in the coming years.

For example the results of APOLLON, a recent EU research project on concentrated photovoltaic outline that this solution could emit per kWh only the 5% of CO2 eq emitted by a natural gas combined cycle plant (RSE 2014). Similar results come from other studies (Fthenakis and Kim 2013). Also without switching to concentrated photovoltaic, other solutions are available to increase photovoltaic systems performance. For example silicon heterojunction (SHJ) cells offer high efficiencies and several advantages in the production process compared to conventional crystalline silicon solar cells (Louwen et al. 2015). The use of bifacial modules can reduce up to 38% the life cycle CO2 emission of a single-crystalline system (Gazbour et al. 2016). Shifting from the conventional cell technology to the state-of-the-art PERC (Passivated Emitter and Rear Cell) technology will reduce the energy payback time and greenhouse gas emissions for photovoltaic electricity generation (Luo et al. 2018). Finally, one-axis tracking installations can improve the environmental profile of photovoltaic systems by approximately 10% for most impact categories (Leccisi et al. 2016).

3 Goal and Scope

The goal of the present study is to compare, from an environmental point of view, different existing technologies and installation options for electricity production from photovoltaic panels.

To better understand the role of photovoltaic in sustainable development of the national electric system, the different photovoltaic (PV) solutions are also compared with a Natural Gas Combined Cycle (NGCC) power plant. As matter of fact, NGCC is (in terms of efficiency, greenhouse gas and particulate matter emissions) the best fossil fuel technology for electricity production. Moreover it is the main technology pushed out of the market by the new photovoltaic installed power (GME 2014).

The functional unit is 1 kWh delivered to the Italian distribution network. System boundaries include all life cycle phases, including end-of-life.

The considered technologies are: amorphous silicon (a-Si), copper-indium–gallium-selenium thin film (CIS), cadmium telluride thin film (CdTe), single-crystalline silicon (single-Si), multi-crystalline silicon (multi-Si), ribbon-silicon (ribbon-Si).

The considered installation options are: on roof-integrated, on roof-not integrated, ground mounted. The combination between technologies and installation options gives the eleven solutions, illustrated in Table 2.1.

Table 2.1 Solutions taken into account in the analysis
Full size table

Concerning impact categories, we followed the indications of the working group of European Commission on Product Environmental Footprint Category Rules (PEFCR)Footnote 1 for photovoltaic.

In particular, although other documents and updates have followed, we referred to the interim report of July 2014 (Frischknecht and Itten 2014) in which the impact categories are classified by relevance. Impact categories classified as highly relevant have been selected. Among them we excluded the “water scarcity” since it is too much dependent on local conditions for the scope of the present analysis. Moreover the characterization method selected by JRC’s guidelines for life cycle impact assessment (EC-JRC 2011) is far from being reliable (it has “to be applied with caution”).

In order to compare the PV system with NGCC we have included two impact categories (acidification potential, photochemical ozone formation potential), that, even though with medium relevance for PV systems, are relevant for conventional fossil fuel power plant. Finally we included the Cumulative Energy Demand (CED) both renewable and non-renewable as it is relevant for estimating the energy payback time of PV solutions.

4 Life Cycle Inventory

As regard the inventory analysis, the data used for background systems came from Ecoinvent version 3 (Wernet et al. 2016). When dealing with PV systems, environmental impacts are mainly related to the materials used for module production and installation and to efficiency in electricity production.

Given an installed power (at peak), the production depends on several factors such as solar radiation, modules inclination and cell efficiency. This factors can be expressed by the so-called yield factor (in hours) which is the ratio between the annual production of a PV plant and its installed peak power.

In this study, on the basis of GSE’s statistics (GSE 2014) an average yield factor of 1140 h was estimated.

Life time of photovoltaic system has been assumed equal to 30 years.

The PV system includes the panel, the mounting structure and the inverter.

5 Life Cycle Impact Assessment

As discussed above the impact assessment has been carried out using eight impact categories (see Table 2.2) and the CED (both renewable and not). Besides being compared to each others, the PV solutions impacts are compared to the electricity production from a NGCC plant. All impacts are referred to the functional unit, that is 1 kWh delivered into the distribution network, with the hypothesis that the PV panels are connected to the distribution network, while the NGCC power plant is connected to the high voltage grid. For this reason the impacts of the kWh delivered by the NGCC include 6.3% of grid losses.

Table 2.2 Selected impact categories
Full size table

In Tables 2.3 and 2.4 the results of impact assessment are reported for the eleven PV solutions. Values higher than the average are highlighted in red.

Table 2.3 Impact assessment results of PV solutions and NGCC
Full size table
Table 2.4 Impact assessment results of PV solutions and NGCC
Full size table

As regards the comparison among the different PV solutions, results show that there is not a better or a worse solution for all the impact categories taken into consideration (Fig. 2.2).

Fig. 2.2
figure 2

LCIA comparison among PV solutions. Functional unit 1 kWh

Full size image

On the other hand it can be noticed that the single-Si, slanted roof installation mounted (i.e. not integrated) shows impacts higher than average for six categories and it is the solution with the worst environmental performance for four impact categories out of eight.

From the point of view of the cells technologies, thin film technologies, CdTe and CIS, show always impact below the average. For climate change they show impacts respectively 50 and 30% lower than single-Si. Among “traditional” technologies the a-Si slanted roof installation integrated has on average lower impact than ribbon-Si, multi-Si and single-Si, with the noticeable exception of the impact categories human toxicity-cancer effect. Among crystalline silicon technologies, multi-Si solutions show lower impacts than single-Si solutions (the differences range from 2 to 16%) for almost all impact categories. One exception is human toxicity (both cancer and non-cancer effects) for which the impacts of this two technologies are almost equal.

Tables 2.5 and 2.6 report the comparison between PV solutions and the NGCC. The results are reported in percentage respect to NGCC impacts.

Table 2.5 Impact assessment results of different PV solutions compared to NGCC (values in percentage)
Full size table
Table 2.6 Impact assessment results of different PV solutions compared to NGCC (values in percentage)
Full size table

For climate change, acidification, photochemical ozone formation and freshwater ecotoxicity impact categories, all PV solutions show far better performance than NGCC. As regards climate change, the maximum potential impact among the analyzed PV solutions is only 14% of the NGCC impact. For acidification this ratio raises to 27% while for photochemical ozone formation and freshwater ecotoxicity it is respectively 36 and 37%. Also for particular matter, PV solutions show better performance than NGCC, but in this case the values are closer and the worst PV solution (single-Si, not integrated) reaches the 66% of the NGCC value. This means that PV solutions have the potential to contribute to this impact category with the same order of magnitude of a NGCC power plant.

On the contrary PV solutions show worse environmental performance for what concern human toxicity impact categories. All the PV solutions generate higher impacts than NGCC power plant in the case of non-cancer effect. For toxicity-cancer effect, five out of eleven solutions entail higher impacts, and the remaining show values very close to those of NGCC. Finally, as foreseeable, for the land use impact category, PV open ground installation is by far the worst solution not only among PV solutions, but also compared to NGCC.

Besides the discussed environmental impact categories, dealing with photovoltaic it is important also to look at the so called CED. CED is in fact linked to the Energy Pay Back Time, which is one of the most used indicator to evaluate PV solutions (Gerbinet et al. 2014). Differences among PV solutions are little and related only to non-renewable energy use. The worst performer under this point of view is again the single-Si, slanted roof installation mounted, while the best one is the roof-integrated installation of CdTe modules (Fig. 2.3).

Fig. 2.3
figure 3

LCIA comparison of PV solutions and NGCC. Impact category: cumulative energy demand. Functional unit 1 kWh

Full size image

All the PV solutions have a CED which is almost the half of the NGCC CED value.

6 Conclusions

Taking into account the rapid growth of the photovoltaic electricity production in Italy, this study was aimed at comparing the environmental performance of the different PV solutions installed in Italy. Life cycle environmental impacts of the different solutions were, also, compared with the impacts of a natural gas combined cycle plant which is, in Italy, the main technology replaced by new photovoltaic installed power.

Life cycle environmental performance of a PV power plant depends on both the cell technology (a-si, multi-si, thin film, etc.) and the installation type (on roof integrated or not integrated, open ground mounted).

Results show that there is not a better solution (i.e. combination of cell technology and installation type) for all the analyzed impact categories. Single-Si, slanted roof installation mounted (i.e. not integrated) is the solution that most frequently has the worst environmental performance over the considered impact categories.

On the other hand, photovoltaic solutions lead in general to potential environmental impacts definitely lower than natural gas combined cycle plants. Nevertheless it is important to underline that all PV solutions show criticalities for what concerns human toxicity. As matter of fact all the PV solutions, in the case of non-cancer effect, and five out of eleven, in the case of cancer effects, show higher impacts than a NGCC plant. Concerning particular matter, PV solutions have in general better performance than NGCC, but in this case the values are closer and the worst PV solution reaches the 66% of the NGCC impact. This means that PV solutions can potentially contribute to this impact category with the same order of magnitude of a NGCC power plant. Moreover, open ground solution, which cover in Italy almost the 40% of the photovoltaic installed power, shows a huge amount of land use, when compare to fossil fuel power plants.

The results discussed above refer to the nowadays Italian situation, while the photovoltaic technology is evolving rapidly. Subject of further investigations could be the assessment of technology innovations such as silicon heterojunction, bifacial modules or one-axis tracking installations, especially for impact categories, such as particular matter and human toxicity, for which PV systems and NGCC have similar performance.