Maximization of extraction of Cadmium and Zinc during recycling of spent battery mix: An application of combined genetic programming and simulated annealing approach
Introduction
Cleaner energy storage systems such as Lithium ion batteries have propelled society to become more mobile and portable (Nemecek, 1994). Despite their considerable advantages, they still pose significant environmental and health hazards. Most of these hazards arise from improper disposal and storage of end-of-life batteries. Global production and consumption of batteries has increased disproportionately to present waste management measures, causing both short and long term issues.
Battery packs are extremely flexible in their design and use, being composed of individual smaller and usually identical cells (Battery pack, 2018). They have longer life span of 2–3 years, and their usage in electric vehicles have resulted in decreased environmental impact when compared to traditional internal combustion (IC) engine-run vehicles when compared using a life cycle assessment (Notter et al., 2010). Regulations have also been passed limiting the amount of dangerous chemicals in batteries, especially mercury (U.S. Environmental Protection Agency, 1997). The end product has become more environmentally friendly than before while being more efficient and resistant to failure. Lithium ion batteries have a very small environmental cost to bear (Boyden, 2014). Other battery chemistries like Ni-Cd pose more significant environmental threats and rising production levels imply higher rates of their consumption. Materials like Cadmium and Cobalt have very adverse effects on both health and the environment (World Health Organisation, 2010; Leyssens et al., 2017). There are several regulations in place limiting the use of these materials in most products. Their disposal and repurposing after they have reached their end-of-life is severely lacking (Official Journal of the European Union, 2006).
In some countries, upwards of 250,000 tonnes of batteries were deemed as waste in 2014 (Eurostat, 2018). In 2016, worldwide consumption of lithium for battery use was 77,821 metric tons of lithium carbonate equivalent (Statista, 2018a). Demand for the metal is projected to reach 422,614 metric tons of lithium carbonate equivalent (Statista, 2018b) by the year 2025. Producers are not currently capable of meeting this demand. In some countries, <2% of all lithium batteries are recycled while the rest are put to landfill (Boyden, 2014). This represents a high threat to public health (Rall and Pope, 1995) and the environment via the leakage of dangerous chemicals (Andresen and Küpper, 2012). It also represents a waste of reusable resources. There is a 30% decrease in overall cost by using recycled materials (Rabah et al., 2008).
Rising production demands can be alleviated by using materials from spent batteries that have undergone a set of recovery and extraction processes. Each process must start with the sorting of various batteries based on their chemical or energy contents (Tonteri et al., 2000). This can be done manually or through some degree of automation (Bernardes et al., 2004). The mix so obtained must undergo extraction to obtain metals like Li, Cd, Ni etc. which are used in further production. These extraction procedures are usually hydrometallurgical for lower value metals, but can also by pyrometallurgical, physical, chemical or biochemical in nature (Li et al., 2009). The set of production treatments depends on the battery chemistry in question (Wang, 2014). One must factor in transport and energy requirements to see the economic feasibility of recycling spent batteries (Niu et al., 2014).
Metal extraction via hydro and pyrometallurgical methods involves a heavy energy intake, as well as high security and pollution risks (Rocchetti, 2013). An alternative to this is chemical and biochemical methods of metal extraction. Bioleaching is one biochemical technique which uses bacteria (like Acidithiobacillus ferrooxidans for iron pyrites (Zhang et al., 2008) and Penicillium citrinum for low grade manganese ores (Acharya et al., 2002)) to react with the metal to yield soluble products. These soluble products then undergo further filtration to extract metal. This technique provides high yields but requires significant improvements before it can be considered commercially viable (Olson et al., 2003). Metal solubilisation via H2SO4 can be performed in a single step leaching process with yields of up to 81% for Cadmium, 96% of Cobalt, 94% of Manganese, 68% of Nickel and 99% of Zinc from a mix generated from spent batteries.
Previous research focussed on the use of response surface methodology (RSM) for modelling and optimizing the metal yields (Tanong et al., 2017). RSM is based on assumption of model structure followed by an estimation of coefficients in the model using optimization methods. This method works satisfactorily if the information about the system behaviour is known. Actual engineering problems are often complex, multidimensional, and incomplete information. RSM is no longer suitable. Predictive modelling methods based on Artificial intelligence (AI) seems a better alternative. Among AI methods, evolutionary approach of genetic programming (GP) has the ability to automate the model structure and coefficients estimation resulting in the evolution of the best model (Woodward et al., 1999). The GP model has a free non-linear form that has the best fits. It can adapt to the system behaviour. A number of diverse applications for GP techniques have been found, which shows its effectiveness and efficacy to model the systems of any given complexity.
This study aims to propose a combination of GP and simulated annealing (SA) approach to maximize the recovery of Zinc and Cadmium. The specific works are listed as follows. Firstly, the effect of concentration of H2SO4, mass of Na2S2O5 as well as the solid-to-liquid ratio and time of retention is comprehensively studied. Secondly, experiments are firstly designed for the recovery of Zinc and cadmium from the spend Lithium-ion batteries mix. Thirdly, GP is used for the formulation of functional relationship between recovered metals Zinc and cadmium and the inputs (Solid/Liquid ratio, concentration of Sulfuric acid, mass of Sodium metabisulfite and retention time). A comparative study between GP, the Box-Behnken model and analysis of variance (ANOVA) analysis has also been performed. Then, the optimal input conditions are determined and validated using experiments. Finally, conclusions are then drawn upon the efficacy of the proposed approach, as well upon metal extraction.
Section snippets
Research problem undertaken
This section discusses the research problem statement for the combined GP and SA approach for the study of chemical metal extraction from a spent battery mix. A disproportionate amount of spent batteries is not recycled, in spite of various public programmes for the same. Recycling spent batteries to recycle valuable metals is one way to reduce rising demands on production. Recycling batteries consists of sorting, metal extraction and reprocessing. Existing pyrometallurgical and
Design of experiment
Samples of spent batteries were collected, manually disassembled and sorted according to the following concentrations (Tanong et al., 2017): 0.28% Li-ion, 0.80% lithium iron sulphide, 1.60% Ni-MH, 15% Zn-C, 14.3% Ni-Cd and 68% alkaline battery. The mix then underwent screening for alien particles including non-metallic components and other contaminants. Metallic composition of the resultant was then determined using inductively coupled plasma-atomic emission (ICP-AE) spectroscopy (Melville, 2014
Genetic programming approach
Genetic programming (Gandomi et al., 2015), an AI approach stem from the principle of Darwinian evolution i.e. “Survival of the fittest”. The procedure involves randomly initialising candidate solutions, which are probabilistically chosen to reproduce basing on their fitness on the output data. Each generation has a fixed population size, where each member is one model. During the initialisation, the initial input and output sets (terminal set), the function space with which to compose model
Statistical modelling using linear regression and GP
The experimental output is dependent on the four correlated parameters: Solid/Liquid ratio (x1), concentration of H2SO4 (x2), Mass of Na2S2O5 (x3) and the Retention time of the mixture (x4). The correlation matrix is given in Table 2. From Table 2, it can be seen that each parameter has some degree of correlation with each other. The linear modelling was unlikely to be successful. Multiple linear regression models were formulated from the data, and the results justify the findings of the
Conclusions
The present work proposes the comprehensive study to optimise the chemical metal leaching of valuable metals from a mix of spent batteries. The optimization of the chemical metal leaching process has been carried out using combined using Combined genetic programming and simulated annealing Approach. Experiments were conducted to validate this approach. The optimal conditions obtained are: Solid/Liquid ratio = 11.7%, molarity of H2SO4 = 0.86 M, g/g of Na2S2O5 = 0.56 g/g and 45 min of retention
Acknowledgement
Authors acknowledge Grant DMETKF2018019 by State Key Lab of Digital Manufacturing Equipment and Technology (Huazhong University of Science and Technology). Authors also like to acknowledge Shantou University Youth Innovation Talent Project (2016KQNCX053) supported by Department of Education of Guangdong Province. This study is also supported by Shantou University Scientific Research Fund (NTF 16002).
References (37)
- et al.
Recycling of batteries: a review of current processes and technologies
J. Power Sources
(2004) - et al.
Assessment of artificial neural network and genetic programming as predictive tools
Adv. Eng. Software
(2015) - et al.
An application of evolutionary system identification algorithm in modelling of energy production system
Measurement
(2018) - et al.
Cobalt toxicity in humans—a review of the potential sources and systemic health effects
Toxicology
(2017) - et al.
Study of spent battery material leaching process
Trans. Nonferrous Metals Soc. China
(2009) - et al.
Recovery of nickel, cobalt and some salts from spent Ni-MH batteries
Waste Manag.
(2008) - et al.
Genetic programming as an analytical tool for non-linear dielectric spectroscopy
Bioelectrochem. Bioenerg.
(1999) - et al.
Metallurgical and mechanical methods for recycling of lithium-ion battery pack for electric vehicles
Resour. Conserv. Recycl.
(2018) - et al.
Bioleaching of pyrite by A. ferrooxidans and L. ferriphilum
Trans. Nonferrous Metals Soc. China
(2008) - et al.
Optimal sustainable life cycle maintenance strategies for port infrastructures
J. Clean. Prod.
(2017)
Bioleaching of low grade manganese ore with Penicillium citrinum. 197
Eur. J. Miner. Process. Environ. Prot.
Cadmium Toxicity in Plants. Cadmium: from Toxicity to Essentiality Metal Ions in Life Sciences
The Environmental Impacts of Recycling Portable Lithium-Ion Batteries (Unpublished Master's Thesis)
Linear Genetic Programming
Generation of Waste by Waste Category, Hazardousness and NACE Rev. 2 Activity
An evolutionary framework in modelling of multi-output characteristics of the bone drilling process
Neural Comput. Appl.
Cited by (10)
Boosted capture of trace Cd(II) with a magnetic dual metal-organic-framework adsorbent
2023, Journal of Solid State ChemistryExploring recycling potential of rare, scarce, and scattered metals: Present status and future directions
2022, Sustainable Production and ConsumptionCitation Excerpt :Although it is an important and meaningful project, the current research is still limited, and the recycling of not all RSSMs has been studied. Overall, the information supporting our results mainly includes beryllium (He et al., 2020), cadmium (Yun et al., 2019), rhenium (Shen et al., 2021), indium (Wang et al., 2019; Silveira et al., 2015; Savvilotidou and Gidarakos, 2020), selenium (Venkatachary et al., 2020), tellurium (Yu and Tong, 2021; Kavlak and Graedel, 2013; Yue et al., 2019; Zhang et al., 2020), REEs (Schulze and Buchert, 2016; Li et al., 2020b; Wang et al., 2020a), etc., the information for each metal is provided in the Supporting Information S1. Whether end-uses are identified as potentially recycling sources depends on a series of factors.
Integrating offline logistics and online system to recycle e-bicycle battery in China
2020, Journal of Cleaner ProductionCitation Excerpt :In order to improve the utilization efficiency of used batteries, optimizing the recovery of materials from the batteries is a promising research in the future (Li et al., 2019). Yun et al. pointed out that by optimizing battery material extraction technology, the harm of lithium-ion batteries can be alleviated and the development of clean energy storage systems can be promoted (Yun et al., 2019). Therefore, based on the proposed recycling system, more attention should be paid to the recycling technology of battery materials.
Evaluation of batteries residual energy for battery pack recycling: Proposition of stack stress-coupled-AI approach
2019, Journal of Energy StorageCitation Excerpt :The evolutionary system can mimic the behaviors of the targeted complicated system [17,18]. A lot of researches related to modeling problems in energy storage system verifies the practicality of the modeling method of GP [19–23]. Compared to other modeling methods, GP has advantages from different aspects:
A coupled electrochemical-mechanical performance evaluation for safety design of lithium-ion batteries in electric vehicles: An integrated cell and system level approach
2019, Journal of Cleaner ProductionCitation Excerpt :Battery pack installation and placement An important research direction could be to re-design the battery packs and its components which can be easily placed in a vehicle to optimize its space, have a minimum impact from the crash, and can be easily dismantled, disassembled and replaced for user-friendly and efficient recycling (Yun et al., 2019a; Yun et al., 2019b). Topology design optimization of the electric vehicle and its integrated components including battery packs could be studied and explored in detail.
A Coupled Genetic Programming Monte Carlo Simulation-Based Model for Cost Overrun Prediction of Thermal Power Plant Projects
2022, Journal of Construction Engineering and Management