In recent years, the rapid growth of the photovoltaic industry has led to an increasing number of end-of-life solar panels, posing significant environmental challenges. As a researcher in sustainable energy, I have focused on evaluating the carbon footprint and environmental impacts of recycling technologies for waste photovoltaic panels. This study employs Life Cycle Assessment (LCA) to analyze three typical recycling processes: the traditional alkali-acid electrowinning method (M1), the improved reverse electroplating-alkali-acid method (M2), and the improved salt etching method (M3). The goal is to quantify their carbon emissions and identify reduction potentials, supporting the transition to a circular economy for solar energy systems.
The proliferation of solar panels as a key renewable energy source is undeniable, but their lifespan of 25-30 years results in substantial waste generation. Improper disposal of these photovoltaic panels can lead to soil and water contamination due to hazardous materials like lead and cadmium. Moreover, recycling valuable metals such as silver and silicon from retired solar panels not only conserves resources but also reduces carbon emissions associated with primary production. In this work, we establish a carbon footprint accounting system based on LCA, targeting the core metal recovery stages to assess environmental impacts and global warming potential (GWP). Our analysis covers normalization, sensitivity, material flow, and carbon flow analyses, providing a comprehensive view of the sustainability of these recycling technologies.
To conduct the LCA, we defined the system boundary from the physical disassembly of waste photovoltaic panels to resource recovery and waste emission. The functional unit is set as the treatment of 1 kg of waste solar panels, ensuring consistency in comparing inputs and outputs across the three processes. Data were sourced from laboratory-scale experiments and literature, with background processes like energy and chemical production derived from the GaBi database. For instance, China’s electricity grid mix, comprising 77.0% coal, 17.6% hydropower, and 5.4% other sources, was used to model energy-related impacts. Chemical inputs include KOH, HNO3, HF, CH3COOH, and NaOH, while outputs consist of recovered silver, silicon, and chemical waste. We applied the CML2001-Aug.2016 method in GaBi to evaluate 10 impact categories, such as ADP, AP, EP, FAETP, GWP, HTP, ODP, MAETP, POCP, and TETP. The GWP, a critical metric for carbon footprint, is calculated using the formula: $$GWP = \sum (m_i \times CF_i)$$ where \(m_i\) is the mass of substance \(i\) emitted and \(CF_i\) is its characterization factor for CO2 equivalence.
The three recycling processes for photovoltaic panels differ significantly in their approaches. M1 involves sequential leaching with KOH and HNO3, followed by electrowinning for silver recovery and ball milling for silicon powder production. M2 utilizes a diluted silver-plating solution in a cyclic electrochemical process to dissolve and recover silver, reducing chemical consumption. M3 employs a NaOH-KOH salt etching system to directly separate silver and silicon without toxic acids, minimizing secondary pollution. The material inputs and outputs for each process are summarized in Table 1, highlighting the variations in chemical use, energy consumption, and waste generation.
| Material/Energy | M1 | M2 | M3 | Unit |
|---|---|---|---|---|
| c-Si battery | 1000 | 1000 | 1000 | g |
| KOH | 896 | 0 | 58 | g |
| HNO3 | 1711 | 1111.5 | 0 | g |
| HF | 136 | 136 | 0 | g |
| CH3COOH | 79 | 79 | 0 | g |
| NaOH | 0 | 580 | 492 | g |
| Electricity | 2.65 | 2.14 | 0.813 | kWh |
| Output silver | 10.73 | 13.78 | 13.78 | g |
| Output silicon | 910 | 918 | 910 | g |
| Chemical waste | 2635 | 1794.9 | 550 | g |
| Hazardous gases | 15.7 | 13.33 | 0 | g |
Normalization analysis revealed the relative environmental impacts of the three processes. For M1, the order of impact categories from highest to lowest was MAETP > GWP > HTP > ADP > AP > POCP > EP > TETP > FAETP > ODP, with MAETP dominating due to the extensive use of nitric acid and hydrofluoric acid. In M2, the improved cyclic process reduced impacts in MAETP and HTP, but AP and POCP remained significant due to residual acid use. M3 showed the lowest overall impacts, with MAETP still being the highest but substantially lower than in M1 and M2, as it avoids toxic acids. The normalization results can be expressed mathematically for each impact category \(j\) as: $$N_j = \frac{I_j}{R_j}$$ where \(I_j\) is the impact value and \(R_j\) is a reference value. This allowed us to compare the processes on a unified scale, confirming that M3 is the most environmentally friendly option for recycling photovoltaic panels.
Sensitivity analysis was conducted by varying key inputs by ±5% to identify critical factors. For M1, electricity consumption and HNO3 usage were the most sensitive, contributing over 30% to changes in GWP. In M2, the sensitivity was more distributed, with electricity, acetic acid, and HF acid each affecting results by up to 22%. M3 exhibited low sensitivity, with electricity and alkaline reagents contributing less than 4% to variations, indicating its robustness. This analysis underscores the importance of optimizing energy and chemical use in recycling solar panels to minimize environmental footprints. The sensitivity coefficient \(S\) for a variable \(x\) can be defined as: $$S = \frac{\Delta GWP / GWP}{\Delta x / x}$$ where \(\Delta GWP\) is the change in GWP due to a change \(\Delta x\) in the variable.
Material flow analysis illustrated the pathways of inputs and outputs. In M1, large amounts of chemicals like HNO3 and KOH led to high waste generation, with no material recycling, resulting in inefficient resource use. M2 incorporated electrolyte recycling, reducing chemical consumption but still involving HF acid, which poses health and environmental risks. M3 demonstrated a simplified flow with minimal chemical use and waste output, as the NaOH-KOH system allows for salt recovery and reuse. The material balance for each process can be represented as: $$\sum Inputs = \sum Outputs + Accumulation$$ For M3, this balance shows nearly complete recovery of silver and silicon with negligible accumulation, highlighting its efficiency in handling end-of-life photovoltaic panels.
Carbon footprint analysis focused on GWP, with results showing that electricity and HNO3 consumption were the primary contributors, accounting for over 75% of emissions in M1 and M2. M1 had the highest carbon footprint at approximately 4.5 kg CO2 eq per kg of solar panels treated, while M2 reduced this to around 3.0 kg CO2 eq due to improved efficiency. M3 achieved the lowest carbon footprint of 0.83 kg CO2 eq, attributed to the absence of nitric acid and reduced energy use. The carbon emissions \(C\) from electricity can be calculated as: $$C_{elec} = E \times EF_{elec}$$ where \(E\) is electricity consumption and \(EF_{elec}\) is the emission factor for the grid. Similarly, for chemicals, \(C_{chem} = m \times EF_{chem}\), where \(m\) is the mass and \(EF_{chem}\) is the chemical-specific emission factor. The total GWP for each process is the sum of these contributions, emphasizing the need for low-carbon energy and chemicals in recycling photovoltaic panels.
Carbon flow analysis traced the movement of carbon through the processes. In M1, CO2 emissions were significant across all stages, particularly from electricity and acid use. M2 showed similar patterns but with reduced flows due to recycling. M3 had minimal carbon flows, concentrated in the etching stage, with no toxic gas emissions. This analysis helps identify hotspots for intervention, such as replacing HNO3 with greener alternatives in solar panel recycling.
Carbon reduction potential was assessed under two scenarios: low-carbon energy and low-carbon materials. In the low-carbon energy scenario, replacing the grid electricity with nuclear power reduced carbon emissions by up to 50% for M1 and M2, and 82% for M3, as M3’s simpler process made energy a larger proportion of its footprint. Using biomass or photovoltaic-based electricity also offered significant reductions. In the low-carbon materials scenario, switching to higher-purity chemicals had minimal impact, with less than 5% reduction across processes, due to the limited variety and quantity of chemicals involved. The combined scenario achieved the highest reductions, with M3’s carbon footprint dropping to 0.15 kg CO2 eq, demonstrating its superior potential for decarbonizing photovoltaic panel recycling. The reduction potential \(RP\) can be quantified as: $$RP = \frac{C_{base} – C_{scenario}}{C_{base}} \times 100\%$$ where \(C_{base}\) is the baseline carbon footprint and \(C_{scenario}\) is the footprint under the alternative scenario.
| Process | Baseline GWP | Low-Carbon Energy GWP | Low-Carbon Materials GWP | Combined GWP | Overall Reduction % |
|---|---|---|---|---|---|
| M1 | 4.50 | 2.25 | 4.28 | 2.03 | 54% |
| M2 | 3.00 | 1.50 | 2.85 | 1.35 | 57% |
| M3 | 0.83 | 0.15 | 0.79 | 0.15 | 82% |
In conclusion, this study demonstrates that the improved salt etching method (M3) is the most sustainable technology for recycling waste photovoltaic panels, with the lowest carbon footprint and environmental impacts. The LCA-based carbon footprint accounting provides a rigorous framework for evaluating and optimizing recycling processes, crucial for achieving carbon neutrality in the photovoltaic industry. Future work should focus on scaling up M3 and integrating renewable energy sources to further enhance its benefits. As solar panel deployment continues to grow, adopting such advanced recycling methods will be essential for minimizing the ecological footprint of photovoltaic systems and promoting a circular economy.