3.1.1. LCA Analysis of PV Systems
The LCA results were used for the evaluation of the environmental impacts of various types of PV technologies. Four different PV systems using crystalline and thin-film technologies (as described in Table 1
) were evaluated in this paper, all having the same nominal capacity of 3 kW. In this section, the detailed results from the LCA of the studied PV systems are presented in order to determine which technologies are more hazardous to human health and ecosystem quality in a comparative assessment, distinguish which lifecycle stage of the PV energy production represents the majority of these impacts, and finally evaluate their overall energy performance.
The LCA of a PV system starts with the extraction of raw materials and follows along the product to the end of its life and the disposal of the PV components. The first stage of the process entails the mining of raw materials, for example, quartz sand for silicon based PVs, followed by further processing and purification stages, to achieve the required high purities, which typically entails a large amount of energy consumption and related emissions. Other raw materials included are those for balance of system (BoS) components, for example, silica for glass, copper ore for cables, and iron and zinc ores for mounting structures. At the end of their lifetime, PV systems are decommissioned and the valuable parts and materials are disposed.
Although PV power systems do not require finite energy sources (fossil, nuclear) during their operation, a considerable amount of energy and emissions are released for their production. The environmental issues associated with this energy use for PV manufacturing will also affect the environmental profile of PV power systems. The environmental themes that are strongly related to the PV energy system are: Exhaustion of finite resources, human health implications, and climate change [25
The goal and scope of this LCA study was to evaluate over the lifecycle the impacts of the electricity produced by four different grid-tied 3 kW PV installations and the functional unit was the production of 1 kWh of produced electricity. The LCIA method used for the characterization of PV technologies was ReCiPe Midpoint, aiming to highlight the global warming potential and GHG emissions, fossil fuels, and climate change impacts related to each technology. The results were ranked from worst to best environmental performance and used to validate the environmental impacts of each PV system. The objective of conducting the LCA study was to make a comparative environmental analysis of different PV systems with a focus on comparing crystalline with thin film technologies.
The system boundaries account for all the impacts related to production, transportation, and system disposal of PV systems. The main parts of the studied systems are: i. The PV-panels, ii. the inverter, iii. the electric installation, and iv. the roof mounting structure. The process data for a 3 kW PV installation includes quartz reduction, silicon purification, wafer, panel and laminate production, and manufacturing of inverter, mounting, cabling, and infrastructure, assuming a 30 years operational lifetime. The following items were studied for each production stage as far as data were available:
Air and waterborne process-specific pollutants at all production stages (materials, chemicals, etc.);
Transport of materials, energy carriers, semi-finished products, and the complete power plant;
Waste treatment processes for production wastes;
Dismantling of all components;
Infrastructure for all production facilities with its land use.
The PV systems have the same nominal installed capacity (i.e., 3 kW) and differ according to the cell type (single- and multi-crystalline silicon, thin film cells with amorphous silicon, and CIS). All systems were assumed to be installed on existing buildings (slanted roof installation).
Life cycle inventory analysis involves creating an inventory of flows from and to nature for a product system. The Ecoinvent v3.4 database was employed for the inventories of PV systems, which can be assumed to be representative for typical PV installations. The Ecoinvent database provides detailed and transparent background data for a range of materials and services used in the production chain of photovoltaics. The delivery of the different PV parts to the final construction place was assumed as 100 km by a delivery van. This includes the transport of the construction workers. It was assumed that 20% of the panels are produced overseas and thus must be imported to Europe by ship. The lifetime of the inverter was assumed to be 15 years.
In Figure 1
, the process network for the studied mc-Si PV system is depicted for the cut-off threshold of 10% (similar figures represent the data for the other three PV types). The thick red line in the network trees is known as the elementary flow and indicates the environmental bottleneck or burden in each process.
For the CIS system, 64.2% of all total inflows and outflows are due to the production of the photovoltaic panel. The installation phase and the inverter require 23.3% and 9.5%, respectively, of the energy and materials inflow. The main environmental impacts include the panel and cell production, inverter, and installation/construction phases. There are also impacts associated with the electricity, transportation, and system disposal, which are taken into consideration. Similar values stand for the case of a-Si panel: 56.9% for the production phase, and 32.5% and 8% for the installation phase and the inverter, respectively. For the sc-Si and mc-Si panels, 77.6% and 72.5%, respectively, of all total inflows and outflows are due to the production of the photovoltaic panel, installation is 13.1% and 16.5%, respectively, while the inverter accounts for 7% and 8.3%, respectively.
From the process networks, it is evident that the production stage contributes the most important part of the environmental impacts in the life cycle of all studied PV technologies. The elementary flows indicate that most inflows of materials and energy for both thin-film and crystalline technologies occur during the cell and panel production phase. Subsequently, large emissions and impacts to the environment and human health follow this stage of the PV systems′ lifecycle. Based on the above, we can conclude that the cell and panel production phase are the most important inputs to the development of a 3 kW PV system, followed by the inverter and construction of the mounting systems.
The environmental impacts of PV systems were calculated through the conducted LCA. The typical operation of PV systems was taken under consideration. In Table 2
and Figure 2
, the aggregated LCA inventory results for the studied PV systems are presented. These are harmonized data representing the LCA results (for each impact category) per total electricity exported to the grid (in kWh) by each 3 kW PV system, thus providing a holistic evaluation indicator (i.e., environmental burden per total energy produced).
In Figure 2
the relative contributions to the impact categories (based on the ReCiPe 2016 midpoint evaluation) for the studied PV systems are shown. The cumulative CO2
-eq emissions per kWh over the whole life cycle of the PV systems vary between approximately by 3.9 × 10−2
and 5.2 × 10−2
During the lifecycle of a PV system, initially, the extraction of resources leads to emissions that affect human health, including carcinogens and respiratory inorganics, while at a second level, the use of fossil fuel during the production and manufacturing processes releases large amounts of greenhouse gases in the atmosphere, causing climate change. Processes occurring during the panel production phase can significantly affect air quality as hazardous substances are emitted into the atmosphere and biosphere.
According to this analysis, the most severe burdens seem to be gathered to the following categories: Global warming, fossil fuel resource scarcity, carcinogens, ecotoxicity, and land use. The crystalline technologies (mc-Si and sc-Si) have increased values in almost all impact categories. Thin-film CIS exhibits lower impacts in most categories and seems to be an optimum selection from an environmental perspective compared to its other counterparts. Results indicate that there are impacts in all indicators, especially those affecting human health from the substances released into the air and water. The manufacturing of a-Si PV cells and panels requires silicon and typically the energy intensive “Siemens process” [34
]. On the other hand, thin film PV systems have lower efficiencies and thus a 3 kW installation will require a larger number of cells and panels and more materials for the mounting systems. According to this analysis, thin-film technologies require less materials′ inflows for their construction and installation phases compared to crystalline systems and this coincides with reduced airborne pollutants, emissions, and energy (also connected with transportation, distribution, and mounting of the systems).
For the purposes of this study, two Monte Carlo analyses of the LCA results (repeated for 5000 iterations) were implemented for a comparison between the PV systems in each studied technology (i.e., crystalline and thin film). The aim of these analyses was to provide an additional validation (based on a statistical evaluation) for the credibility of the presented results. The first analysis was conducted between A: a-Si and B: CIS PV systems. During the Monte Carlo analysis, a stochastic variation of the parameters in the initial inventory database for each of the studied two cases (i.e., A and B) was performed, altering the LCA results and thus affecting the A−B outcome. A random variable was selected for each parameter within the specified uncertainty range and the impact assessment results were recalculated. The same process was repeated by taking different samples (within the uncertainty range) and all results were stored. After repeating the procedure for a set number of times (e.g., 5000), 5000 different results were obtained, thus forming the uncertainty distribution of the impacts (LCIA), with a confidence interval of 95%.
The results in a bar chart form are depicted in Figure 3
showing the percentage of times when system A has a greater impact than system B (A−B ≥ 0, in orange) and vice versa (A−B < 0, in blue). This is a balanced graph and, in general, we can conclude that A has increased impacts compared to B in most of the studied midpoint categories. This is quite evident for the human carcinogenic toxicity category, in which A has distinctively increased impacts compared to B for 96.6% of the completed iterations. Respectively, human non-carcinogenic toxicity and freshwater eutrophication are the two cases that A has a lower impact than B, for almost 80% of the completed iterations.
The second Monte Carlo analysis was conducted between A: mc-Si and B: sc-Si PV systems. Figure 4
presents the results in a bar chart form, showing the percentage of times when system A has a greater impact than system B (A−B ≥ 0, in orange) and vice versa (A−B < 0, in blue). In this case, it is evident that case A has lower impacts compared to B in most of the studied midpoint categories. The impact categories that a balanced result is observed are water consumption, land use, human non-carcinogenic toxicity, marine, freshwater, and terrestrial ecotoxicity.
It is very important to stress the fact that the results depicted in Figure 3
and Figure 4
refer to the comparison of the raw LCA data and not the harmonized results as mentioned in Table 2
and Figure 2
(i.e., LCA results for each impact category per total electricity exported to the grid for each PV system). Thus, these data do not include the provision for varying energy production for each of the studied systems.
Various additional technical components, the so-called balance of system (BoS) elements, can also play an increasingly important role for the comparison of different types of PV technologies with different efficiencies and thus different sizes of mounting systems for the same electric output. These BoS elements can have a significant share of 30% to 50%. On the one hand, this is due to the improvements, which could be observed for the production chain until the output of the final photovoltaic cell. On the other hand, now a more detailed investigation of these additional elements is available, which, for example, also includes the electronic components of the inverter. The low efficiency systems need larger amounts of the mounting structure and cabling, which partly outweighs the better performance per kWp of the module alone [26
]. Overall, in the entire life cycle of both types of PV technologies, it was observed that the magnitude of environmental impacts of crystalline was greater than that of the thin film.
3.1.2. Energy and Economic Assessment of PV Systems
The first step in a pre-feasibility study of a solar (i.e., PV) project is to define the solar energy potential of the region in which the PV systems will be installed. This serves as a planning tool to quantify the anticipated electricity production and plant costs. The evaluation of these PV technology costs require in-depth analysis of site-specific solar energy potential; costs of solar technologies; customer types; meter types; utility types; physiographic conditions; local, regional, and national laws and regulations; feed-in-tariffs and financial mechanisms; etc. The techno-economic analysis carried out in this part of the paper quantifies the energy output and the economic income associated with each of the studied 3 kW PV power plants. The proposed area for installation of the PV systems is the island of Crete located in the southern part of Greece, which was selected as a typical representation of regions with a mild climate and high average insolation that lasts almost throughout the year (with greater intensity from April to October). These climatic conditions render Crete as one of the best available locations in Greece for installation of solar systems. The island is not interconnected to the mainland distribution grid and the necessary electricity is produced via diesel burning conventional thermal stations, thus increasing the cost (environmental and economic) per produced energy unit. In addition, Crete presents extreme variations in energy demand throughout the year, with significant peaks during the summer due to the tremendous increase of the population due to visiting tourists and increased air-conditioning needs. Thus, the need for decentralized production of electricity is more than obligatory as the solar grid parity in non-interconnected islands can already be considered as a fact [22
]. On the other hand, the deficiencies in the existing electricity grid and local supporting schemes/governmental rules for renewables have created a vague scenery for potential investors. The economic and energy assessment of PV systems was carried out using the RETScreen software. The completed study involves quantifiable results for energy—economic impacts and savings for the chosen PV system. The site location for the installation of the PV systems was chosen to be the Acrotiri area in Chania, while all meteorological data (in the form of the annual time series of average climate conditions) were extracted from RETScreen referring to a weather station of Souda Bay, Chania.
The results of the RETScreen economic analysis provide a reliable and comprehensive evaluation of the anticipated technology, the energy production, potential emissions reduction, necessary investment cost, financial viability, and risks associated with the specific project. The accuracy of RETScreen is considered to be more than sufficient for preliminary feasibility studies and a small reduction in accuracy due to the use of monthly rather than hourly solar radiation data is more than compensated for due to the ease-of-use of the software.
After selecting the location area, the complete RETScreen analysis for each one of the studied PV systems was conducted. This analysis comprised four discrete steps: i.
Selection of the technology (i.e., sc-Si, mc-Si, CIS, a-Si) and specification of the technical parameters, ii.
energy analysis (see results in Table 3
emissions analysis), and iv.
For all financial calculations, the electricity price was set to 0.10 €/kWh and we considered that the installation was funded by own means (no bank loan). For Greece, the employed feed-in-tariff for roof top PV will decline to 0.8 €/kWh by the end of 2019, but residential installations up to 10 kWp can benefit from a net-metering scheme, which can allow for compensation at prices up to 0.15 €/kWh [35
]. In Table 3
, the main results of the RETScreen analysis for all studied PV systems are presented. The cell efficiencies of the PV systems vary (from 6.1% to 17%), but this parameter does not play an important role as the nominal capacity of all systems is set to 3 kW. On the other hand, the larger the efficiency of the panel, the less the area needed for the installation (from 17.7 m2
to 49.2 m2
). The simple payback period is 8.8 to 10 years (for regions with same insolation, i.e., Andalucía in Spain, the corresponding values for residential PV are 7.6 to 12.1 [38
]) and IRR values vary from 11.5 to 13.1. The a-Si system seems to have a higher annual energy yield, and this is practically due to the ability of these systems to produce more electricity under hazy or cloudy conditions and thus their capacity factor is increased (21.8%) compared to their counterparts. The electricity produced allows for the mitigation of ~4 tons of CO2
-eq annually for all PV systems.
According to the comparison of the different PV technologies, the anticipated energy production, emissions reduction, investment cost, financial viability, and risks associated with the four technologies are approximately the same. All technologies portray relatively equal cost benefit ratios and financial parameters. This is mainly due to the fact that our selection of comparing 3 kW systems harmonizes the influence of all technical advantages amongst technologies. On the other hand, the sc-Si system is the most efficient per cell, thus needing less area per installation compared to the other cases.