PV panels create unique waste-management challenges along with the increasing waste streams forecast. Apart from in the EU, end-of-life treatment requirements across the world for PV panels are set by waste regulations applying generically to any waste rather than dedicated to PV.
Waste regulations are based on the classification of waste. This classification is shaped according to the waste composition, particularly concerning any component deemed hazardous. Waste classification tests determine permitted and prohibited shipment, treatment, recycling and disposal pathways. A comprehensive overview of the widely varying global PV waste classification is beyond the scope of this report. Instead, this chapter characterises the materials contained in PV panels and corresponding waste-classification considerations. These determine the required treatment and disposal pathways for PV panels when other more specific waste classifications and regulations are not applicable.
3.1 PANEL COMPOSITION
To achieve optimal waste treatment for the distinct PV product categories, the composition of PV panels needs to be taken into consideration. PV panels can be broken down according to the technology categories shown in Table 7. The different technology types typically differ in terms of materials used in their manufacturing and can contain varying levels of hazardous substances that must be considered during handling and processing.
C-Si PV is the oldest PV technology and currently dominates the market with around 92% of market share (ISE, 2014). Multicrystalline silicon panels have a 55% and monocrystalline silicon panels a 45% share of c-Si technology respectively. Due to low efficiency ratios, a-Si products have been discontinued in recent years, and the market share nowadays is negligible.
The two thin-film PV panel technologies make up 7% of the PV market, 2% for CIGS panels, and 5% for CdTe panels. The following analysis will not pay any more attention to CPV and other technologies because it only has a low market share at less than 1%.
Although the market share of novel devices is predicted to grow, mainstream products are expected to retain market dominance up to 2030, especially c-Si panels (Lux Research, 2013). As shown in Table 7, silicon technology has great potential for improvement at moderate cost if new process steps are implemented into existing lines. For example, an increase in usage of hetero-junction cells is predicted, providing higher efficiencies and performance ratios. According to Lux Research (2013 and 2014), CIGS technology has great potential for better efficiencies and may gain market share while CdTe is not expected to grow. In the long term, CIGS alternatives (e.g. replacing indium and gallium with zinc and tin), heavy metal cells including perovskite structures, and advanced III-V cells, might take nearly 10% of market share. The same can be said of OPV and dye-sensitised cells (Lux Research, 2014). Recent reports indicate OPV has reached efficiencies of 11% and dye-sensitised cells 12% (IEA, 2014).
In line with a PV market heavily dominated by c-Si PV, all the main panel manufacturers except for First Solar rely on silicon-based PV panel technologies. In 2015, the top ten manufacturers for PV panels represented 32 GW per year of manufacturing capacity, which is around two-thirds of the global PV market, estimated at 47 GW (see Table 8).
The various components of major PV panel technologies will influence material and waste characterisation as well as the economics of treatment pathways. As shown in Boxes 4 and 5, the design of silicon-based and thin-film panels differs, affecting their composition accordingly.
A typical crystalline PV panel with aluminium frame and 60 cells has a capacity of 270 watt-peak (Wp) and weighs 18.6 kilogrammes (kg) (e.g. Trina Solar TSM-DC05A.08). For a standard CdTe panel, 110 Wp can be assumed on average for 12 kg weight (e.g. First Solar FS-4100). A CIGS panel usually holds a capacity of 160 Wp and 20 kg (e.g. Solar Frontier SF160-S).
Research on the PV components concludes that progress in material savings and panel efficiencies will drive a reduction in materials use per unit of power and the use of potentially hazardous substances (Marini et al. (2014); Pearce (2014); Raithel (2014); Bekkelund (2013); NREL (2011) and Sander et al., (2007)). On this basis, Figure 10 compares the materials employed for the main PV panel technologies between 2014 and 2030.
PV panels By weight, typical c-Si PV panels today contain about 76% glass (panel surface), 10% polymer (encapsulant and backsheet foil), 8% aluminium (mostly the frame), 5% silicon (solar cells), 1% copper (interconnectors) and less than 0.1% silver (contact lines) and other metals (mostly tin and lead) (Sander et al., 2007 and Wambach and Schlenker, 2006).
Industry trend studies such as the International Technology Roadmap for Photovoltaic (ITRPV) suggest new process technologies will prevail, encouraging thinner and more flexible wafers as well as more complex and manifold cell structures. These will require new interconnection and encapsulation techniques. For example, bifacial cell concepts offer high efficiencies in double glass panels made of two glass panes each two millimetres thick. An encapsulant layer reduction of up to 20% is possible owing to thinner wafers. Cells with back-contacts and metal wrap-through technologies that reduce shadow and electrical losses (known as hetero-junction concept cells) are equally expected to gain significant market share (Raithel, 2014).
By 2030 the glass content of c-Si panels is predicted to increase by 4% to a total of 80% of the weight’s panel. The main material savings will include a reduction in silicon from 5% down to 3%, a 1% decrease in aluminium and a very slight reduction of 0.01% in other metals. Specific silver consumption is expected to be further decreased by better metallisation processes and replacements with copper or nickel/copper layers (Raithel, 2014).
In today’s market, the most efficient panels with back junction-interdigitated back-contacts have shown efficiencies of about 21%. Hetero-junction technologies have achieved 19%. The average efficiency of a c-Si panel has grown by about 0.3% per year in the last ten years (Raithel, 2014).
a-Si PV panels have lost significant market share in recent years and do not contain significant amounts of valuable or hazardous materials (see Figure 10). Thus, they will most likely not require special waste treatment in the future. This section and the rest of the report therefore does not cover a-Si panels.
In multi-junction cell design, two (tandem) or more cells are arranged in a stack. In all cases the upper cell(s) have to be transparent in a certain spectrum to enable the lower cells to be active. By tailoring the spectrum sensitivity of the individually stacked cells, a broader range of sunlight can be absorbed, and the total efficiency maximised. Such cell types are used in a-Si, c-Si and concentrator cells. The low cost of c-Si today allows cost-efficient mass production of high efficiency multi-junction cells. This can be combined, for example, with III-V alloys, chalcogenides and perovskites expected to perform extremely well even in non-concentrating tracker applications (Johnson, 2014).
Thin-film panels are technologically more complex than silicon-based PV panels. Glass content for c-Si panels is likely to increase by 2030. By contrast, it is likely to decrease for thin-film panels by using thinner and more stable glass materials. This in turn will encourage a higher proportion of compound semiconductors and other metals (Marini et al., 2014 and Woodhouse et al., 2013).
CIGS panels are today composed of 89% of glass, falling 1% to 88% in 2030. They contain 7% aluminium, rising 1% in 2030, and 4% polymer remaining stable. They will experience a slight reduction of 0.02% in other metals but a 0.2% increase in semiconductors. Other metals include 10% copper, 28% indium, 10% gallium and 52% selenium (Pearce, 2014; Bekkelund, 2013 and NREL, 2011).
CIGS panel efficiency is currently 15% and targeted at 20% and above in the long term (Raithel, 2014).
By 2030 the proportion of glass as total panel mass in CdTe panels is expected to decrease by 1% from 97% to 96%. However, their polymer mass is expected to increase by 1% from 3% to 4% compared to today. In comparison to CIGS panels, material usage for semiconductors as a proportion of panel usage will decline almost by half from 0.13% to 0.07%. However, the share of other metals (e.g. nickel, zinc and tin) will grow from 0.26% to 0.41% (Marini et al., 2014; Bekkelund, 2013 and NREL, 2011). The main reason for this increase in other metals is the further reduction in CdTe layer thickness (which brings down the semiconductor content of the base semiconductor). However, the efficiency improvements of the past couple of years were also related to ‘bandgap’ grading effects, which can be achieved by doping the semiconductor layer with other components. The addition of other components to the mix is reflected in the rise in other metals. Another reason for the increase in the proportion of other metals is the addition of a layer between back-contact metals and the semiconductor package. This reduces copper diffusion into the semiconductor and thus long-term degradation and leads to the thickening of the backstack of metals (Strevel et al., 2013).
The PV industry is aiming for 25% efficiency for CdTe panel research cells and over 20% for commercial panels in the next three years. This is substantially higher than the 15.4% achieved in 2015. New technologies are also expected to reduce the performance degradation rate to 0.5%/year (Strevel et al., 2013).
Chapter 6 provides additional details on panel composition, the function of various materials and potential future changes in panel design and composition.
3.2 WASTE CLASSIFICATION
PV panel waste classification follows the basic principles of waste classification. This also considers material composition by mass or volume and properties of the components and materials used (e.g. solubility, flammability, toxicity). It accounts for potential mobilisation pathways of components and materials for different reuse, recovery, recycling and disposal scenarios (e.g. materials leaching to groundwater, admission of particulate matter into the soil). The overall goal of these classification principles is to identify risks to the environment and human health that a product could cause during end-of-life management. The aim is to prescribe disposal and treatment pathways to minimise these threats. The risk that materials will leach out of the end-of-life product or its components to the environment is very significant, and assessment of this threat helps define necessary containment measures. However, this is just one possible risk. Other examples assessed through waste characterisation include flammability, human exposure hazards through skin contact or inhalation. Risks assessed may differ by country and jurisdiction.
Depending on national and international regulations such as the Basel Convention on the Control of Transboundary Movements of Hazardous Wastes and Their Disposal (UN, 2016), waste can be classified into various categories such as inert waste, non-hazardous waste and hazardous waste. To some extent, the origin of the waste is also taken into consideration, defining subcategories such as industrial waste, domestic waste and specific product-related categories such as e-waste, construction waste and mixed solid wastes. The different categories of classified waste then determine permitted and prohibited shipment, treatment, recycling and disposal pathways.
In 2015 two-thirds of PV panels installed across the world were c-Si panels. Typically, more than 90% of their mass is composed of glass, polymer and aluminium, which can be classified as non-hazardous waste. However, smaller constituents of c-Si panels can present recycling difficulties since they contain silicon, silver and traces of elements such as tin and lead (together accounting for around 4% of the mass). Thin film panels (9% of global annual production) consist of more than 98% glass, polymer and aluminium (nonhazardous waste) but also modest amounts of copper and zinc (together around 2% of the mass), which is potentially environmentally hazardous waste. They also contain semiconductor or hazardous materials such as indium, gallium, selenium, cadmium tellurium and lead. Hazardous materials need particular treatment and may fall under a specific waste classification depending on the jurisdiction.
Key criterion for PV panel waste classification:
Table 9 summarises typical waste characterisation leaching test methods in the US, Germany and Japan. The overview provides one of the most important characterisation metrics used in PV waste classification across the world at this time.
The key criterion for determining the waste classification is the concentration of certain substances in a liquid which has been exposed to fragments of the broken PV panels for a defined period of time in a particular ratio. This leachate typically dissolves some of the materials present in the solid sample and hence can be analysed for the mass concentration of certain hazardous substances. Different jurisdictions, such as Germany, the US or Japan provide different threshold values for the allowable leachate concentrations for a waste material to be characterised as nonhazardous waste. For instance, the threshold for leachate concentration for lead allowing a panel to be classified as hazardous is 5 milligrammes per litre (mg/l) in the US and 0.3 mg/l in Japan. For cadmium, the hazardous threshold is 1 mg/l in the US, 0.3 mg/l in Japan and 0.1 mg/l in Germany. These compare to publicly available leaching test results in the literature (summarised in Sinha and Wade, 2015) for c-Si and CdTe PV panels. They range from non-detect to 0.22 mg/l for cadmium and non-detect to 11 mg/l for lead. Thus, in different jurisdictions, CdTe and c-Si panels could be considered either non-hazardous or hazardous waste on the basis of these test results.
Regulatory classification of PV panel waste
From a regulatory point of view, PV panel waste still largely falls under the general waste classification.
An exception exists in the EU where PV panels are defined as e-waste in the WEEE Directive. The term ‘electrical and electronic equipment’ or EEE is defined as equipment designed for use with a voltage rating not exceeding 1,000 V for alternating current and 1,500 V for direct current, or equipment dependent on electric currents or electromagnetic fields in order to work properly, or equipment for the generation of such currents, or equipment for the transfer of such currents, or equipment for the measurement of such currents (EU, 2012).
Hence, the waste management and classification for PV panels is regulated in the EU by the WEEE Directive in addition to other related waste legislation (e.g. Waste Framework Directive 2008/98/EC). This comprehensive legal framework also ensures that potential environmental and human health risks associated with the management and treatment of waste are dealt with appropriately. By establishing a List of Wastes (European Commission, 2000), the EU has further created a reference nomenclature providing a common terminology throughout the EU to improve the efficiency of waste management activities. It provides common coding of waste characteristics for classifying hazardous versus nonhazardous waste, transport of waste, installation permits and decisions about waste recyclability as well as supplying a basis for waste statistics.
Some codes from the EU’s List of Wastes applicable to PV panels are given in Table 10.