Comparison of HFC emission and bank modelling methods

Abstract

The abundance of HFCs with very high global warming potentials, particularly HFC-134a, HFC,125, HFC-32 and HFC-143a, are increasing in the atmosphere. Several studies have developed bottom-up methods for estimating global emissions using consumption or production data and emission factors. However, few studies have investigated the influence of different bottom-up methods on simulated emissions and banks. Quantification of these differences is particularly important when we consider future emissions as banks within equipment act as potential sources of future emissions. Therefore, as controls are introduced to phase down production and consumption, the contribution to emissions from banks will become more significant.

This thesis aims to explore how bottom-up modelling approaches may influence bottom-up emission or bank estimates of HFCs that are primarily used for cooling applications. By exploring two established bottom-up models, from Velders et al. 2022 and the IPCC Guidelines, the impact on emissions and banks of model architecture is quantified. Consideration is also given to the role of uncertain input parameter choices. The sensitivity of model outputs to the input parameters is explored using a sensitivity analysis and history matching.

The results presented in this thesis show that the two different models derive relatively consistent trends over the last three decades for all four gases. However, some differences were found for aspects of future emissions, such as the magnitude and timing of peak HFC emissions under a Kigali Amendment scenario. However, for some species, particularly HFC-143a, neither model could be brought into agreement with atmospheric observation-derived (“top-down”) emissions estimates for much of the recent observational record, suggesting that consumption estimates, which were not investigated here, may be the dominant source of uncertainty in emission and bank projections. Future work should focus on examining the sensitivity of these outputs to inputs and assumptions that feed into consumption models.

Date of Award	1 Oct 2024
Original language	English
Awarding Institution	University of Bristol
Supervisor	Matthew L Rigby (Supervisor)