Refuse and Rethink, Reduce (Low environmental impact materials, products, and processes)
Rationale for support
To identify the main areas of improvement for the environ- mental footprint of PV, it is necessary to regard the tech- nology’s entire lifecycle. Using LCA, important knowledge can be gained as to which processes and materials con- tribute most to the overall environmental footprint, and as such are key candidates in the first R-ladder strategies: Refuse/Rethink and Reduce.
For the strategy Refuse and Rethink, a good example would be the development of lead-free solder, avoiding the use of toxic components and their possible release to the en- vironment, and hence lowering the environmental impact in toxicity-related impact categories. LCA should ideally be applied early in the development of a technology, so that high-impact materials or processes are avoided when pos- sible.
The lifecycle thinking also aids in identifying key candidates for the strategy Reduce. Although it seems self-explanatory that reductions in material use lead to improved environ- mental impact, it is of course essential that these reduc- tions do not adversely affect the function of the technol- ogy.
State of the art PV LCA studies use life-cycle inventory data from 2020, assume manufacturing of PV in China, and module efficiencies of 20.5 % for mono- and 18 % for polycrystalline based PV systems. These systems, when transported to and installed in a region of moderate inso- lation (1700 kWh m-2 yr-1) have a carbon footprint of 23 to 25 gCO2-eq/kWh, respectively.41 The EPBT of these systems is around 0.8 years, while the ERoEIel are 11.5 and 10.5 and ERoEIpe are 38 and 35, respectively. Summarizing the latest state-of-the-art in LCA of PV, we can see:
Typical system energy payback time Southern Europe is around 0.8 years. The long-term potential is 0.25 years
non-cell general material intensity in 2018 (t/MW) (modules and BOS)
- Concrete (system support structures): 60.7
- Steel (system support structures): 67.9
- Plastic (environmental protection): 8.6
- Glass (substrates, module encapsulation): 46.4
- Al (module frames, racking, supports): 7.5
- Cu (wiring, cabling, earthing, inverters, trans- formers, PV cell ribbons): 4.6
Typical carbon footprint of a mono-Si PV ground-mounted system in Southern European insolation, 1700 kWh/m²/yr, performance ratio of 0.85, and lifetime of 30 years: 23-32 gCO2-eq/kWh
The lack of studies and poor details on the envi- ronmental performance of available processes for indium recovery from obsolete flat panel displays, semiconductors, and similar products leaves uncer- tain if EoL recycling of indium would actually result in net environmental benefits
China’s position as the key player along the solar PV supply chain (in raw materials, processed materials, components, and assemblies) creates the potential for supply risks and bottlenecks
Across the board, these latest results show a strong de- crease in carbon footprint compared to those using 2015 data. Key drivers here are:
- Silicon production and material use efficiency
- Thinner wafers
- Diamond wire wafering with much reduced kerf losses
- Reduced electricity consumption along the value chain
- Improved cell and module efficiency
Continuous technological improvements like these have resulted in a strong decline of the environmental impact of PV technology over the past decades. Due to improve- ments in material use (e.g. the amount of silicon in g/Wp) and improved module efficiency, both EPBT and carbon footprint have dropped by roughly an order of magnitude in the past 30 years. (46) This indicates a learning rate of around 12 % for CED of PV systems, and of 16.5-23.6 % for carbon footprint, depending on the type of system.
For a renewable energy technology to be successful, it needs to strongly reduce the carbon footprint of the sourc- es it will replace, while having a strong net positive ener- gy balance. This implies that the energy payback time of systems needs to be short, the carbon footprint needs to be reduced, the use of local materials to reduce transport costs in systems must be increased, the use of hazardous materials needs to be avoided, and systems and system components need to be designed in a way that encourages recycling and decreases material usage.
While the current environmental impacts of PV electrici- ty are very much (up to several orders of magnitude, de- pending on the impact parameters considered) smaller than those of incumbent (fossil fuel based) sources, the vast scale of projected PV deployments in the energy tran- sition requires that the technology keeps improving the technology, and, as Roadmap 5 also says, a harmonization and update of life cycle inventory databases is required to cope with this fast development of PV technologies and in- tegrated applications (BIPV, VIPV, AgriPV, FloatingPV, IIPV, etc.), so that accurate LCAs may be performed now and in the future.
Low environmental impact through dedicated life cycle en- gineering, focusing on low embodied carbon materials and low-carbon electricity for energy intensive manufacturing steps - thereby reducing energy payback time and carbon footprint / global warming potential of PV should be com- plemented with the incorporation of scope 3 emissions (related to the embodied carbon of the supplied system components) in public renewable energy tenders and com- mercial PPAs, with this a legal requirement if possible.
Targets, Type of Activity and TRL
Continuous improvements of PV technology are required for future success. Even incorporating steady learning rates, and thus continuous improvement of PV technolo- gy with reducing environmental impacts, up to 11 % of the remaining carbon budget needs to be used to install suffi- cient quantities of PV. (47) Thus, it is clear that PV technol- ogy needs to keep improving over the entire value chain.
Although the environmental impact of PV electricity is typ- ically assessed using metrics focusing on energy demand (EPBT) and greenhouse gases (carbon footprint), recent studies have been starting to focus more and more on the material constraints of the PV industry. The expected enor- mous growth of the PV industry, could result in it becom- ing the major consumer of several materials, including flat glass and silver. Furthermore, the consumption of alumin- ium for PV module frames results in high energy demand and associated carbon emissions, as well as posing possi- ble constraints.
Analysis of the evolution of the carbon footprint over the years (from 40-100 g CO2-eq/kWh to- wards lower values) (48) for PV modules and sys- tems with a clear definition of the boundaries of the calculation
Determine the environmental benefits of recy- cling precious metals
c-Si PV module BOM without harmful substanc- es (Pb, F etc.) but without limitation in terms of quality and reliability
Development of Cu-based contact systems
Use of recyclable polymers (PET, PP, PE, etc.)
Reduced kerf loss in sawing
Favour the consumption of raw material pro- duced in Europe, for example:
- Silver produced in Poland is about 20 % of global production share, together with Peru and Australia
- Silicon produced in Norway (6 % of global pro- duction)
Analysing the historical development of key PV technology and designs to aid in setting KPI tar- gets for 2030 and beyond
- Using experience curves and prospective mar- ket developments to estimate the improve- ment of key design parameters (efficiency, wafer thickness, etc)
- Modelling the environmental impact using these values and future estimates of electric- ity grid mixes
Digitalisation can be used to track the resources needed for a PV project or service.
The role of digitalisation
Although an LCA is an environmental approach which brings a global vision of environmental impacts of a ser- vice or a product, the incorporation of LCA in companies is most of the time slow and costly. Digitalisation offers the opportunity to both improve on this time-consuming pro- cess, as well as to improve the availability, quality of data and to ensure it is up to date.
1. Flexible data update:
1.1 Challenge: it takes a lot of time and energy for LCA experts to gather the data needed in order to produce a complete and reliable analysis of a product (life cycle in- ventory). One of the reasons why reliable data is difficult to find for LCA experts is that the information among an organisation is hard to find, either because it can be “lost” among this organisation or it isn’t transmitted well from one service to another.
- Digital cooperation on an international platform to help in creating a data commons. Data commons also decrease the workload of requesting each in- dividual organization for data by making legitimate data open
- LCA could be associated with Computed-Aided De- sign (CAD) software. Thus, LCA would be linked to the value chain of the company: some data entered into the CAD software would be reused into LCA thanks to an Application Programming Interface. Or LCA could be a plug-in added to a CAD software. Data would thus be visualised directly through the CAD software:
- The amounts of materials used for the product designed would be more easily found because they will be directly integrated into the model designed. The environmental impact calculation would intervene at the beginning of the design process;
- The visualisation of the product itself could be in- teresting so LCA experts could easily see if some data is missing: the drawing of the product could change its colours if the data corresponding has been entered into the LCA analysis.
- Modular, flexible and transparent LCA software: LCA software, no matter to which categories they be- long to, are seen as too inflexible in their processes. This could be explained by the lack of open-source software competitiveness. A modular software can be thus a solution to avoid cognitive overloading and calculations issues. Such a software could be developed as independent bricks. Designing a scal- able LCA software could help small and medium sized companies (SME) to adapt to environmental norms, especially companies who don’t have the financial capital for an LCA expertise. Some ques- tions remain unsolved. For example: How modular an LCA software should be developed? Can a LCA software be compatible with Industrial Ecology (IE) or Material Flow Analysis (MFA) tools? Those issues should be tackled in future research.
2. Validation of LCA results
the validation of an LCA is the assurance that the model matches the actual system identified. Currently, the validation phase is seen more as an “additional” phase of an already well-established tool. However, LCA lacks empirical validation. Perhaps the digital transformation would allow the development of techniques capable of confirming or invalidating the results of LCA models in reality. This would allow LCA to be further linked into concrete experiences.
KPIs envisaged for this roadmap are:
|KPI||Target Value (2030)||Current value|
|Energy required to produce MGS|
< 8 kWh/kg
|Electricity for SoG silicon|
<10 kWh/kg for advanced processes
|Electricity for Cz ingot|
< 25 kWh/kg
|Wafer thickness||150-160 µm depending on wafer size||170 µm2,12|
|Kerf losses||< 50 µm12||60-65 µm2,12|
|Primary raw material usage for|
BOS i.e., concrete and steel
|Reduction by at least 3 % (4 % reduction by 2030 and further 6-7 % by 2050)|
|Primary raw material usage||Reduction of Plastic, glass, Al, and Cu, by at least 3 % (respectively 3 %, 4 %, 4 %, 2 %|
reduction by 2030 and further 7 %, 6 %, 6 %, 7 % by 2050)
|Acquisition of PV materials from|
|Increase silicon metal by 20 % (Norway, 6 % global share in 2019), and silver by 30 %|
(Poland, 20 % global share in 2019)
|Environmental KPIs||PV system in S. Europe:|
|PV system in N. Europe:|
|PV system in S. Europe:|
Values for 2030 are based on
current production electricity
mixes. As the energy transition
evolves, EU mix gradually moves
toward Norway’s electricity mix
Thin film, single junction:
Thin film, single junction:
IEA PVPS Task
Mono-Si 20.5 %:
Multi-Si 18 %:
Thin film, single junction:
|Process/technical KPIs||Target Value (2030)||Current value|
|CED (MJ/Wp)||<9.5 MJ/Wp mono||IEA PVPS Task|
Mono 20.5 %:
13.6 MJ/Wp mono
Multi 18 %:
14.8 MJ/Wp multi
|EPBT||<0.55 years||<0.93 years||IEA PVPS Task|
Mono 20.5 %:
13.6 MJ/Wp mono
Multi 18 %:
14.8 MJ/Wp multi
|EROI electricity kWhel/kWhPE||>16||>10||IEA PVPS Task|
Mono-Si 20.5 %: 11.5
Multi-Si 18 %: 10.5
|EROI primary energy|
|>54||>32||IEA PVPS Task|
Mono-Si 20.5 %: 38
Multi-Si 18 %: 35