Toward a life cycle approach for classifying the toxicity of refrigerants

Abstract

The American Society of Heating Refrigerating and Air-Conditioning Engineers (ASHRAE) classifies the safety of refrigerants based on their flammability and toxicity. Toxicity classifications are based on Occupational Exposure Limits (OEL), which estimate industry workers' exposure conditions and frequency (ASHRAE, 2013, 2019). Using these exposure limits and acute toxicity exposure limit values set to prevent danger to life or health, the toxicity classifications are based on a threshold, where Class A (lower toxicity) is assigned when the OEL is higher than 400 ppm while Class B (higher toxicity) refrigerants have OELs below this threshold (ASHRAE, 2013). In general, refrigerants are not considered to be highly toxic compounds. Table 1 shows that the most commonly used hydrofluoroolefin (HFO) refrigerants are in Class A1, which is an indication of lower toxicity for mammals (“A”) and no flame propagation (“1”) (ASHRAE, 2013). Nevertheless, it is important to point out that this toxicity classification only pertains to the parent compound and not necessarily to the precursors used in refrigerant manufacturing or the degradation products resulting from refrigerant emissions or use. Furthermore, the fully fluorinated methyl group (-CF3) in HFOs has prompted their classification as per- and polyfluoroalkyl substances (PFAS) in the United States and Europe (Table 1).

The newest classes of refrigerants, hydrofluorocarbons (HFCs) and HFOs or halogenated olefins are currently in use due to their low global warming potentials (GWPs) and negligible ozone depletion potentials (ODPs). The addition of hydrogen in HFCs and a double bond in HFOs have helped lower their GWPs. For example, the double bond in HFOs is highly reactive with atmospheric hydroxyl (OH) radicals, which leads to their short atmospheric lifetimes and low GWP. However, because these compounds degrade quickly, they have the potential to create significant yields of various degradation products. One of the most well-known degradation products, particularly from HFCs (e.g., R-227ea) and HFOs (e.g., R-1234yf), is trifluoroacetic acid (TFA), whose classification as an ultrashort PFAS is under considerable debate (Table 1). This classification has policy implications as both the European Commission and the USEPA have signaled their commitments to systematically decrease the usage of PFAS compounds (Glüge et al., 2020). Scientific arguments have been made to manage all PFAS compounds together as a chemical class because of their common characteristics of being highly persistent, bioaccumulative, and potentially hazardous (Kwiatkowski et al., 2020). Trifluoroacetic acid is the simplest of the perfluorocarboxylic acid (PFCA) group of substances (Burkholder et al., 2015) and is generally regarded to be highly persistent in the environment, toxic at elevated concentrations, and bioaccumulative depending on its dispersion in the environment. Although TFA can be taken up from contaminated soils by plants and translocated within the plant (Boutonnet et al., 1999), there is not enough evidence of potential for bioaccumulation in the food chain due to its very low K_ow (Xu et al., 2022). For this reason, some argue that TFA and its anthropogenic precursors (e.g., HFOs) should be excluded from any future regulatory action (Singh & Papanastasiou, 2021). One of the greatest uncertainties associated with halogenated olefin refrigerants is whether they can degrade to produce enough TFA to increase the nominal value of 239 ng TFA sodium salt L⁻¹ estimated in the global oceans (Frank et al., 2002; UNEP, 2022a).

Regulatory bodies, such as the USEPA and European Chemicals Agency (ECHA), have indicated their desire to begin analyzing any PFAS and related compounds through a life cycle assessment (LCA) framework (ECHA, 2023). This entails analyzing how potentially harmful and persistent compounds are used and could influence both environmental and human health throughout their production, usage in consumer products, and end-of-life disposal. An iteration of the LCA framework has been already proposed for PFAS to better characterize their environmental impacts (Holmquist et al., 2020). The ecotoxicity life cycle impact assessment (LCIA) framework proposed by Holmquist et al. (2020) integrated PFAS and their transformation fractions, human toxicity, and marine and freshwater aquatic ecotoxicity to predict the fate and accumulation of PFAS in aquatic systems. One of the major findings of this work was that even low emissions of PFAS can have large effects on LCA results (Holmquist et al., 2020). The LCA framework may also be a valuable tool for expanding the scope of the study, focusing not only on the final products, that is, commercially available refrigerants, but also their precursors and degradation products.

In general, an LCA framework includes the goal and scope definition, inventory analysis, impact assessment, and interpretation (ISO:14040, 2006). The inventory analysis phase of the LCA involves data collection and calculation to quantify the inputs and outputs of a product system. The production of refrigerants involves the use of chemicals and even some refrigerants with high GWPs such as HFCs. For example, the refrigerant R-134a with a 100-GWP of 1530 is used as a feedstock to make several refrigerant blends, including R-467A and R-470B, which contain 52.4% and 3% mass of R-134a, respectively (UNEP, 2022b). Even though R-134a will be phased down under the Kigali Amendment due to its high GWP, its use in blend composition remains a point of ongoing discussion; therefore, the use of feedstocks should be included in the inventory phase as part of the manufacturing process of refrigerants. The next phase of the LCA is the impact assessment, which involves associating inventory data with specific environmental impact categories and their corresponding indicators. Hydrofluoroolefins were introduced into refrigeration systems to address GWP, ODP, safety, and toxicity concerns from international legislation and the public. The short atmospheric lifetimes (days) and low 100-GWPs (<20) have contributed to the deployment of HFOs in a variety of equipment. However, the -CF3 group in HFOs and some of their degradation products (e.g., TFA) have raised concern about their classification as PFAS (OECD, 2022), and also their association with harmful health effects in humans and animals. The environmental indicators of this phase should not only include the impacts from parent compounds but also from their degradation products, to account for all relevant impacts. The results from these two phases can be incorporated into the LCA framework to identify overlooked environmental impacts or health concerns from precursors and degradation products according to their refrigerant parent compound. This information can then be fed into models to predict the impacts of new or alternative refrigerants with similar chemical compositions following an LCA approach.

Regardless of whether the newest classes of refrigerants are defined as PFAS and some of their degradation products as PFCA, their environmental footprint can comprehensively be assessed using an LCA approach as a prerequisite for the development of strategies to better address the impact of refrigerants in the environment. Given the finite number of chemicals effective for refrigeration, the use of halogenated refrigerants will likely continue to be used in the foreseeable future. Although the addition of hydrogen and a double bond in the newest classes of refrigerants have contributed to lower GWPs, the unintended impacts of the degradation products of refrigerants remain understudied or unknown for some aquatic systems (e.g., marine ecosystems). With new regulations and definitions for contaminant classes, it is recommended to adopt an LCA approach, which can present a comprehensive picture of potential impacts of production, usage, consumer products, and degradation products of in-use, new, and alternative refrigerants. The LCA information can be utilized by industry agencies like ASHRAE to develop a broader approach for assessing refrigerant production's overall impact. This requires expanding the information incorporated in the inventory phase of the LCA to include the use of raw materials or chemical precursors with high GWP in producing some of the newest classes of refrigerants. These data can then be used to better characterize and classify the impacts of refrigerants in the environment throughout their life cycle. Finally, LCA phases are iterative: Each section's results support the others and can accommodate new information as it is presented, ensuring the comprehensiveness and consistency of the approach.

Federico Sinche Chele: Conceptualization; investigation; visualization; writing—original draft; writing—review and editing. Louise Stevenson: Investigation; writing—review and editing. Christian Mark Salvador: Investigation; writing—review and editing. Fred Dolislager: Investigation; writing—review and editing. Teresa Mathews: Conceptualization; investigation; writing—review and editing.

The authors declare no conflicts of interest.