Depending on the refrigerant used the level of benefit from using a heat pump is at least about an order of magnitude greater than the emissions associated with refrigerant loss. There are three main factors to consider for the emissions related to heat pumps.
The comparison with direct heating: For each kW of electricity consumed by a heat pump, about 4kW of thermal energy is generated. This corresponds to a 300% efficiency whereas all direct heating methods are less than 100% efficient. The source of primary electricity generation and direct heating fuel will affect the relative emissions. Where nuclear or renewable primary sources are used to generate the grid electricity, then heat pumps have essentially no emissions associated with the electricity use.
The efficiency of the heat pump: If a refrigerant is used properly, the efficiency due to various refrigerants that are suitable for use at the applicable conditions, will typically have differences in the range +/-5%. For example, where appropriate for safety requirements, the switch from R-410A to R-32 improves energy efficiency typically in the range 5-10% and reduces refrigerant charge by about 15 to 30%. Heat pump design, including controls, compressors (eg inverter variable speed), heat exchangers, fans or water pump performance (for air to water heat pumps) and reducing heating loads can have a larger effect on energy consumption.
The direct emissions due to the refrigerant used: Refrigerant selection takes into account energy efficiency, safety, reliability and technical performance, which is why non-flammable HFCs such as R-410A and R-134a are widely used in heat pumps. Direct refrigerant emissions occur during the equipment lifetime (annual losses due to gradual leaks for non-hermetic systems), end-of-life disposal losses and failure losses (rare event). The consequences of refrigerant loss due to failure help determine the refrigerant safety classification required for the application. The refrigerant GWP will affect the benefit gained to a limited extent from using a heat pump assuming equivalent energy performance and the same refrigerant losses.
Illustration, comparison will contribution will depend on climatic conditions, emissions intensity for electricity generation and application. Assumes heat pumps with 4% annual emission rate, 15% end of life emissions and 15 years lifetime. Hermetic (monobloc) heat pumps have essentially no annual leakage except due to failures, but do have end of life emissions, reducing the contribution of the refrigerant to the total heat pump emissions.
For this example, HFC-410A contributes about 9% of the emissions due to the heat pump for the EU CO2 emissions intensity for electricity generation (but depending on climatic conditions and application). The switch to HFC-32 and A2L HFC/HFO refrigerants with similar GWPs reduces the refrigerant emissions contribution to about 3%. The use of R-454C with a GWP of 148 or similar HFC/HFO blends reduces the refrigerant contribution to about 0.7%. For larger heat pumps that may be used for district heating or industrial applications with compressors that can use lower pressure HFO refrigerants with a GWP <1 the contribution due to refrigerant emissions is insignificant (<0.005% at GWP=1). Reducing annual leak rates and end of life emissions will reduce these contributions further.
IIR 32nd Informatory Note on Refrigeration Technologies / October 2016 Harmonization of Life Cycle Climate Performance Methodology.
RTOC 2018 Assessment Report Section 2.2.4 Refrigerants choice and energy efficiency.
Overview of electricity production and use in Europe — European Environment Agency https://www.eea.europa.eu/data-and-maps/indicators/overview-of-the-electricity-production-2/assessment-4
For more information see https://www.fluorocarbons.org/heat-pumps-and-hfcs/ and