Energy
2nd July 2021

Heat Pumps, Boilers, and Decarbonisation

Article by Tommy Isaac

Comparing heat pumps and boilers, and which is most suited for decarbonisation

In 2019¹ UK domestic combustion produced 66.5m tCO₂ and shared first place with passenger vehicles (67m tCO₂) as the largest single source of UK emissions. The majority of such combustion would have been used for heating purposes.

Tackling this challenge is critical and forms a central element of the recently-published Ten Point Plan² and Energy White Paper³ by the UK Government. Much debate naturally takes place around the "best" technology, or technologies, to achieve this goal. The two primary candidates in contention are heat pumps and gas (hydrogen) boilers. Although on the surface much separates these two visions of the world, in today’s energy system they are actually functionally identical with near identical carbon credentials. In outlining the supporting evidence for this equivalency, a spotlight will be placed on a systematic shortcoming within the decarbonised heat debate.

Marginal energy supply

The energy system is a highly integrated and complex system. Therefore, understanding the true relationship between cause and effect can be difficult. A key concept within the energy system is that of energy balancing, which is often referred to as "marginal supply". This concept is most relevant within the electricity market. The reason for this is that the electricity network does not have the ability to inherently store energy, therefore supply must be matched on a second-by-second basis with demand. To ensure security of supply, a generator can bid for "balancing services" to the UK’s primary electricity network operator, National Grid. This allows the network operator to call for more or less power from generators, as the system needs, to ensure sufficient supply is being maintained at all times. To enable generators to offer such balancing services they must be controllable in their output and able to respond rapidly. The generation technologies most suited to this operating mode are, pumped storage/hydroelectric power stations, and gas engines/open cycle turbines. It is the later that supplies the majority of the marginal supply to the UK electricity grid. It is the fast, controllable response times of gas engines and open cycle turbines that enables them to provide this service. However, the penalty for such responsiveness is a reduced thermal efficiency, relative to a closed cycle combined turbine.

Gas engines/open cycle turbines generate electricity with a typical thermal efficiency of 35%.

The concept of marginal supply is important as it provides insight into the emissions implications of additional demand, given that any increase in demand must be met with an equal increase in supply. This is one area within the energy system where the relationship between cause and effect is clear – if something causes additional demand, the effect will be an increase in required output from marginal generation.

Electricity generation data is reported in 30-minute intervals from National Grid⁴. Figure 1 provides the breakdown of electricity generation every 30 minutes within 2020, the data is shown in the four major generation types: nuclear, wind, gas and other (biomass, solar, hydro, coal, interconnectors).

Figure 1: 2020 UK electricity generation

2020 was the best year on record for low-carbon electricity generation; with nuclear generation accounting for 20%, wind generation 24%, other sources 15%, and natural gas 41%. Although great strides have been made to bring down the average carbon intensity of electricity generation, it can be seen from the data that natural gas is still the single largest source of electricity, and more importantly – provides almost all of the balancing services to the system. Within the 2020 data, there is not a single 30-minute period where natural gas was not the primary marginal supply source.

The equivalent picture for the gas network is much simpler, as it does not have the diversity of supplies that are seen in the electricity network, natural gas accounts for 99.5% of grid delivered gas⁵, with biomethane accounting for the other 0.5% of supply. Therefore, both the marginal and average supply source is natural gas.

Heating options

Today, 23m (85%) of British homes are heated by natural gas boilers⁶, the remaining 15% are primarily off-grid and therefore are heated by oil, LPG or electric heating. This dominant presence of natural gas boilers is largely the cause of the 66.5m tCO₂ of emissions in 2019; so low-carbon alternatives are critically needed to address these emissions. As outlined, the two primary candidates are heat pumps, and hydrogen boilers. District heating systems have not been considered in the analysis, as they ultimately still require a heat source, which would likely either be provided by gas combustion or a heat pump. Therefore, they can be considered to simply be an alternative delivery mechanism to having the heat generated within the home, but won’t change the macro balance substantially.

The shortcoming mentioned in the introduction is the difference in perception between a heat pump and a gas boiler. Gas boilers are identified as "high carbon" because when fuelled with natural gas they emit carbon dioxide at the point of use, whereas heat pumps are identified as "low carbon" because no carbon dioxide is emitted at the point of use. This is an incorrect distinction to make, as neither a gas boiler nor a heat pump is inherently low or high carbon. Both technologies are, in essence, plastic and metal boxes that provide heat on demand. Their carbon credentials relate solely to the fuel being provided to generate the heat. Therefore, to understand any difference in the carbon credentials between the two technologies, analysis must be undertaken of the fuel sources being used.

Emissions comparison

Given the status quo of domestic heating within GB, if a household converts from a gas boiler to a heat pump, additional electricity demand from the house will be created to provide the necessary heat. As outlined from the electricity generation data, marginal electricity supply is from natural gas combustion. Therefore, the additional supply that will be called upon to match the additional demand will almost certainly be a natural gas-fuelled gas engine or open cycle turbine. It is this relationship that yields the functional equivalency of the two technologies, as both are ultimately being fueled by natural gas. However, differences lie in the delivery supply chains. Over time, marginal supply should reduce in carbon intensity, as developments in alternative technologies such as batteries start to take effect. However, industrial-scale grid balancing via batteries is yet to be commercially and technically proven. Therefore for the time being, flexible natural gas technologies will continue to dominate this role within the supply mix.

Figure 2 demonstrates the ultimate quantity of natural gas that would be needed to provide 100 kWh of heat to the household in question. Both heat supply chains account for conversion efficiencies as well as network losses.

Figure 2: Heat supply chains

If heat is provided by the gas boiler supply chain, ultimately 112 kWh of natural gas is required to deliver 100 kWh of heat, which accounts for the 90% conversion efficiency within the boiler and the 1% of losses from the gas network.

The equivalent supply chain from a marginal heat pump would require 105 kWh of natural gas to supply 100 kWh of heat. This accounts for a heat pump coefficient of performance of three, the 8% losses typically experienced through an electricity network, a 35% thermally efficiency gas engine supplying the marginal demand and finally 1% of losses from the gas network.

Therefore, it can be seen that the heat pump route does represent a marginal carbon emissions benefit of 7% to deliver the same quantity of heat. A typical household requires 9,700 kWh of heat each year and the emissions intensity of natural gas is 184 gCO₂/kWh. Therefore, each year, the boiler supply chain would produce 2.0 tCO₂ compared to 1.9 tCO₂ for the heat pump supply chain. To put the emissions difference into context, a typical car in the UK produces 140 gCO₂/km, therefore the same emissions would be saved by forgoing a round trip from London to Newcastle once a year.

It is clear that both options are incompatible with a net-zero energy system. Reducing the carbon emissions associated with either technology would be contingent upon developing greater low-carbon supply capacity to allow a lower carbon energy source to be utilised. Within the gas boiler supply chain this would be hydrogen production capacity, and within the heat pump supply chain this would either be wind/solar with batteries or hydrogen-fuelled flexible electricity generation.

Technology equivalency

A natural gas-fuelled boiler and a heat pump currently provide heat for essentially the same carbon emissions, both ultimately through the combustion of natural gas. The gas boiler requires combustion at the point of use within a home, and the heat pump requires combustion within a centralised power station. The carbon credentials of both technology routes are constrained by the current availability of low-carbon energy supply. There is simply not enough low-carbon electricity or gas currently to enable either technology route to use a low-carbon alternative to natural gas. An advantage of heat pumps is that they are already "low-carbon ready". In that once sufficient low carbon electricity supply is available, they will be able to use this supply to provide low-carbon heat. A natural gas boiler today would not be able to use a hydrogen supply beyond 20 vol%, but there is a growing demand from the appliance manufacturer community for the government to mandate all domestic gas appliances to be "hydrogen-ready" by 2025. Working, hydrogen-ready prototypes of all domestic gas appliances have been developed. Therefore, it would be reasonable to assume that by 2025 any new domestic gas appliance would be capable of using a hydrogen supply once available.

Both technologies have similar expected asset lifetimes of around 15 years. Therefore to understand if any tangible difference in lifetime emissions would be expected, low-carbon supply projections must be studied for both gas and electricity supplies.

Energy supply projections

There are inherent difficulties in projecting anything in life, and the energy system is no different. That is why National Grid issues annual Future Energy Scenarios, where the potential band of future energy system configurations is outlined through four scenarios. These scenarios are often used as the basis of capacity planning for energy networks and therefore should be sufficiently credible as the basis of projection.

The broad trends of energy supplies within all four scenarios are the same – a reducing role for fossil fuels and an increasing role for low-carbon technologies to deliver a net-zero energy system by 2050. The key differences between the four scenarios relate to the pace of change and the final mixture of technologies achieved by 2050.

Within the electricity supply projections, the role of natural gas generation is shown to reduce over time. As more wind and other low-carbon generation comes online, the operating mode of natural gas plants will very likely change from supplying both baseload and marginal supply to just supplying marginal supply to balance the grid. All scenarios reach a non-fossil fuel supply mixture by 2050, however the point at which natural gas generation is shown to account for a negligible proportion of generation (<1 TWh/y) is between 2040–2045. Therefore, it is likely that for the next 20 years natural gas generation will still play the marginal supplier role within the electricity generation mix, as the average supply becomes ever lower carbon intensity through the deployment of wind, solar and nuclear generation.

Within the gas supply projections, the role of natural gas also declines over time with lower carbon alternatives replacing the supply capacity. As hydrogen is projected to become the primary supplied gas to users, the role of natural gas becomes a precursor to hydrogen production through reformation-based production methods combined with CCS, alongside green hydrogen from electrolysis. Across the four scenarios, the conversion from natural gas supplies to hydrogen is shown to be on the same timescale to the role of natural gas with electricity generation, ie 2040–2045. Therefore, energy supply projections for both boilers and heat pumps can be seen to be on the same time frame. Over the course of either asset’s lifetime, eg 15 years, it is unlikely that a wholesale change to the energy supply dynamics outlined will change. However, over the 2020s it would seem reasonable to install equipment that is capable of receiving low-carbon supplies in time for when a greater low-carbon supply capacity becomes available.

Conclusions

The carbon credentials of any technology that uses energy to provide a useful function, such as the provision of heat, relates solely to the fuel source being supplied. Although on the surface, a heat pump and a gas boiler offer distinctively different options to provide heat to a home, they are both constrained by the current availability of low-carbon energy supplies. This creates a functional equivalency between the two options, where both ultimately rely upon natural gas combustion to provide heat. If installed today, it is unlikely that this dynamic would materially change over either asset’s lifetime. Neither a gas boiler nor a heat pump should be seen as a genuinely low-carbon heating option today, as both will still be constrained by low-carbon energy supplies by the time they are replaced. Therefore, from the perspective of a consumer, the current decision between the two options should be made on personal preferences and relative economics, instead of a perception of relative carbon credentials.