Under the Kigali Amendment to the Montreal Protocol, new controls are being implemented to reduce emissions of HFC-23 (CHF33 ), a by-product during the manufacture of HCFC-22 (CHClF22 ).
Starting in 2015, China and India, who dominate global HCFC-22 production (75% in 2017), set out ambitious programs to reduce HFC-23 emissions. Here, we estimate that these measures should have seen global emissions drop by 87% between 2014 and 2017. Instead, atmospheric observations show that emissions have increased and in 2018 were higher than at any point in history (15.9 ± 0.9 Gg yr−1± 0.9 Gg yr−1 ).
Given the magnitude of the discrepancy between expected and observation-inferred emissions, it is likely that the reported reductions have not fully materialized or there may be substantial unreported production of HCFC-22, resulting in unaccounted-for HFC-23 by-product emissions. The difference between reported and observation-inferred estimates suggests that an additional ~309 Tg CO2CO2 -equivalent emissions were added to the atmosphere between 2015 and 2017.
Due to their influence on climate, the major hydrofluorocarbons (HFCs) were regulated under the Kyoto Protocol of the United Nations Framework Convention on Climate Change (UNFCCC) and have recently been included in the Kigali Amendment to the Montreal Protocol to ensure that their radiative forcings does not offset climate gains provided by the phaseout of chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons (HCFCs)1,2,3.
With a long atmospheric lifetime (∼∼ 228 years1,4), and high radiative efficiency, HFC-23 has the highest global warming potential (GWP) among HFCs (100-year GWP, 12,6901). It is primarily produced as an unwanted by-product during the manufacture of the refrigerant HCFC-22, via the over-fluorination of chloroform (CHCl33 )5. Smaller emissions of HFC-23 are associated with its use as a feedstock for halon-1301 (CBrF33 ) production, plasma etching and chamber cleaning in the semiconductor industry, very low-temperature refrigeration and specialty fire suppression1,5,6,7.
Previous studies, based on in situ atmospheric observations and firn air measurements, have shown an increase in the global annual mean mole fraction of HFC-23 from near zero in early 1960 to 28.9 ±± 0.6 pmol mol−1−1 by the end of 20161,5,6. These data, when combined with a model of atmospheric transport and chemistry can be used to infer global emissions.
Such top-down methods have previously shown that global HFC-23 emissions grew from 4.2 ±± 0.7 Gg yr−1−1 in 1980 to 13.3 ±± 0.8 Gg yr−1−1 in 2006, before declining rapidly to 9.6 ±± 0.6 Gg yr−1−1 in 2009 in response to emission reductions from developed countries and as a result of the UNFCCC Clean Development Mechanism (CDM)6,8. As the CDM period ended, HFC-23 emissions grew to 14.5 ±± 0.6 Gg yr−1−1 in 2014, before declining again, to 12.7 ±± 0.6 Gg yr−1−1 , in 20166.
Here, we present an update of global HFC-23 emissions derived from atmospheric observations, based on new data from 2015 to the end of 2018. In addition, we compile a new inventory-based HFC-23 emissions estimate through to the end of 2017 that includes reported emission reductions by China and India. We find that in 2018, observation-based HFC-23 emissions are higher than at any point in history (15.9 ± 0.9 Gg yr−1± 0.9 Gg yr−1 ), whilst inventory-based emissions are at the lowest in the past 17 years (2.4 ± 0.9 Gg yr−1± 0.9 Gg yr−1 in 2017) when reported emission reductions were included.
Due to the magnitude of the discrepancy between reported emissions reductions and emissions inferred from the atmospheric data, it is highly likely that developing countries have been unsuccessful in meeting their reported emissions reductions. Alternatively, or additionally, there may be substantial unreported production of HCFC-22 at unknown locations resulting in unaccounted-for HFC-23 by-product being vented to the atmosphere.