According to the French government-funded study, small leaks on refrigeration systems are very common but they have very little impact on total emission leakage figures. It’s the small number of very large leaks that account for over 90% of the annual leakage mass.
The report draws on research conducted by Cemafroid Irstea, the French air conditioning and refrigeration testing and certification company, on behalf of the French association Alliance Froid Climatisation Environnement (AFCE).
It reveals that leaks of around 5g/yr – the standard minimum detection level of electronic leak detectors – account for 57% of all leaks, yet they represent just 1.12% of the total mass of refrigerant potentially lost though leakage. In contrast, while large leaks of over 500g/yr account for just 4% of all leakage incidents, they contribute 91.6% to the total leakage mass.
The F-gas regulations and industry standards insist that electronic leak detectors should be able to detect a minimum leakage of 5g/yr. Based on the figures revealed in studies by Cemafroid Irstea, the reports author asks whether the 5g/yr detection threshold is relevant for leak testing?
Further, it argues that repairing a 5g/yr leak would actually be more harmful to the environment than allowing it to continue to leak.
Using the example of a 5g/yr leak on on a 100-litre liquid tank of a 70kW R404A refrigeration plant with evaporating temperature of -30°C and condensing temperature of 35°C, the total leakage over a 15-year period the leak would total 292kg CO2e, according to the report.
Repairing that leak and considering that the recovery unit stops when relative pressure is 0.2 bar, the residual mass of refrigerant in the plant at a temperature of 20°C would be 326g, an amount of R404A equal to 1,262kg CO2e that would inevitably be released to atmosphere.
The report points out the value of shut-off valves to avoid draining the entire plant to perform the repair. When there is no shut-off valve, it advises against the frequent draining of the plant to repair low-rate leakages, instead waiting until the next preventive maintenance operation to perform the repair.
Leak detectors having numerous detection thresholds could be a significant step forward in this area, it argues.
Leak detection regimes
During the study, laboratory testing was also used to determine how leak detection methods influenced the performance of measuring instruments. Results showed that the minimum flow rate detected varies according to how the operator uses it. For instance, a leak detector calibrated to 5g/yr can detect leaks with a flow rate under 5g/yr but fails to detect a 5g/yr leak if it is positioned too far away from the source.
The report emphasises the importance of leak detector verification certificates specifying the instruments’ ideal distance of use for efficient detection of leaks 5g/yr and over.
The report can be viewed and downloaded here.