For the building sector in China, about 50% of building energy consumption is used for heating [19], including water heating, space heating as well as some niche needs such as clothes drying [20] and dehumidification [21].

The buildings’ heating contains multiple heating modes as shown in Fig. 2. As is summarized in Table 2 [22], the buildings’ heating is usually with a relatively low temperature less than 80 °C,and the ambient air could serve as the heat source in most cases. The air source heat pump could reach a seasonal heating COP of 3.2 in Qingdao and 3.5 in Shanghai [23], and validated prototypes [24] with improved cycles [25,26,27] have achieved COP over 2.7 with the ambient temperature as low as -15 °C. Currently, there are more than 450 million units of air-conditioners in China [28,29], which could be also regarded as spilt-type air-source heat pumps if they are used for heating in winter and cooling in summer. If there are accesses to other sources with higher temperature, such as geothermal [30], waste heat [31], the efficiency will be further improved. For example, geothermal heat pump could gained a higher seasonal COP about 4.32 in Qingdao and 4.79 in Shanghai [23], yet there are simply not enough space for underground wells in most highly urbanized China cities [32]. Therefore, air-source heat pump, in particular small-scale air source heat pump, has accumulated its advantage in China’s heating market.

Moreover, through recovering waste heat from industrial process, large-scale heat pumps could achieve a better performance. As shown in Fig. 3, there are two validated cases based on waste heat recovery for district heating adopting absorption heat pump and compression heating pump. Figure 3 a shows the absorption heat pump absorbing heat from power station’s waste heat (Heat source: ∼35 °C with 0.24 MPa driven steam, Heat sink: 81.34 °C; Heating capacity:57 MW, COP:1.77) [33]. Figure 3 b shows a compression heat pump system recovering waste heat from a steel plant for district heating (Heat source: 32.5 °C waste water, Heat sink: 62.5 °C; Heating capacity:9.67 MW, COP:6.67) [34]. As the above two projects have utilized waste heat according to local conditions and achieved high-efficiency by waste heat recovery, their investment payback periods are about 2-3 years.

Water heating remains an increasing residential energy demand in China, given the urbanization rate and relatively lower living standard compared to developed countries [35] (water heating accounts for about 30% of Japan’s total residential energy consumption [36]). According to the national standards applicable for heat pump water heaters in Japan [37] and China [38], heat pump water heaters could gain a COP value of 3.2 to 3.7 under the nominal condition (based air-source with wet-bulb temperature of 15 °C for 55 °C hot water generation [38]) on and even reach 4.6 by excellent manufacturers [36].

Generally, most of current heat pump systems for buildings have a seasonal heating COP ranging between 3.0 and 4.5, further innovations could even increase the system COP to more than 6 [39], especially combined with waste heat recovery. For buildings’ heating, thermal comfort is also important, and some pilot research on small temperature difference (STD) fan coil unit [40] and desiccant-coated heat pumps(DSHP) [41] would be beneficial. As shown in Fig. 4, STD fan coil unit is suggested in an air source heat pump heating system, whereas the heat transfer area is expanded and the water supply temperature could be reduced to less than 35 °C. In that case, the COP of the heating system could be further improved. The desiccant coated heat pump’s heat exchangers are coated with solid desiccant, and thus the heat exchangers can independently handle sensible and latent loads at the same time. The DSHP is with an ultrahigh COP value of more than 6.2 without sacrificing any comfort or compactness. Such technology could also be applied into automobile [42] and high speed trains [43].

In China, heat accounts for about 50%-70% of industrial energy consumption, 30%-50% of which are in the range of 80 °C-160 °C [9]. As shown in Fig. 5, compared to buildings’ heat demands, the industrial heat demands are often with relatively higher temperatures. Due to a higher temperature demand in industry, industrial heat pumps often adopt industrial waste heat as the heat source. Validated cases have shown the COP of 120 °C heat pump is larger than 4.0 with a temperature lift value over 40 °C [44]. Newly-developed refrigerants could enable heating temperature up to 168 °C [45], and combining with absorption heat pumps could further achieve heating temperature to 180 °C [46]. Concerning heating temperature up to 100 °C, the corresponding refrigerants could be hydrocarbon (e.g., R290 and R600, yet with a heating capacity restriction), carbon dioxide, and synthetic refrigerants (e.g., R1234ze(E) and R1233zd(E)), and the corresponding compressors and other devices are also with a wide range of options. For heating output temperature less than 100 °C, ammonia is with market-available large-scale compressors and validated application cases [47], and could be considered for large-scale applications. However, there is still a lack of available commercial compressors with large capacity and high efficiency, especially the centrifugal compressor is considered to be only one choice for megawatt scale currently. The current good option could be ammonia vapor compression heat pump hybrided with LiBr-water absorption heat transformer for large scale industrial heating, as both systems are matured. Since water/steam also play an important role in the ammonia-water absorption cycle, it could be possible to find good combination of ammonia-water based vapor compression and absorption hybrid cycles suitable for various applications [48,49].

The above-mentioned heat pumps are with closed cycle, whereas the heat pump transfer heat in an indirect way through the heat exchanger (condenser). However, when considering the fair matching between heat pumps and certain industrial processes such as distillation, concentration and drying, that are widely integrated in petroleum pharmacy, and food industries [50], open cycle could be further adopted. Mechanical vapor compression/recompression [51] is a typical open cycle, an improved efficiency could be gained through recovering the latent heat from condensed steam [52]. For industrial heat pumps, another development trend would be large temperature lift. According to the statistical data, the amount of available heat sources increases dramatically when the heat source temperature decrease, while there is also increasing heat demands with higher temperature [53]. Therefore, a typical large temperature lift heat pump would be useful yet guaranteeing a high-efficiency is hard. Cascade cycles and multistage cycles are often adopted for large temperature lift, but the temperature lift value is hard to exceed 100 °C. As shown in Fig. 6, recent studies verified that air-source heat pump based water vapor compression system could generate 120 °C steam with COP over 130% at -18 °C ambient temperature, 210% at 35 °C ambient temperature [54]. With an average operating COP of 170%, the air-source heat pump boiler can directly satisfy the high temperature steam demand of small and medium-sized enterprises, and is an efficient driver for terminal electrification of high-temperature heating [13].

In many cases, industrial processes have heating demands across different temperature ranges [55], particular appling to industrial parks with various industry clusters [56]. In such cases, the benefits of coupling heat pump with other sources such as solar power would be flexible and efficient. In summary, industrial heat pumps based on different heat sources have already been applied in different cases, yet for the high temperature heating and large scale, high efficiency for compressors, improved cycle with less inverse loss and low-GWP refrigerants would be further investigated for much more competitive advantages in heating markets.

The buildings’ heating contains multiple heating modes as shown in Fig. 2 and Table 3, the buildings’ heating is usually with a relatively low temperature less than 80 °C, and the ambient air could serve as the heat source in most cases. According to the investigation, as for the primary energy consumption in China’s building sector, about half was used to provide heating [19]. According to our investigation as shown in Fig. 5 and Table 4, many industrial processes are with temperature demand between 80 °C and 160 °C, which could be satisfied by current industrial heat pump technology.

As for heat proportion, according to the investigation [9], for China’s industrial sector, 50% to 70% of energy is consumed in the form of thermal energy(heat),and 45% of the heat is with medium and low temperature (<250 °C). Meanwhile, taking EU industrial heating demand as an example, currently 45.80% of such process heating is within temperature below 200 °C, while 41.75%% is below 150 °C [57] as shown in Fig. 7. In terms of temperature match, the first part could be totally replaced by current industrial heat pump, while the heating demands between 150 and 200 °C is also expected to be replaced with the development of heat pump technology.

Conservatively, it is estimated that 60% of industrial energy consumption is consumed through heat, and 40% of this heat can be replaced by heat pump at temperatures below 150 °C. This alternative proportion by heat pump will further increase with the development of heat pump technology.

The economics of heating facilities can be affected by many factors include the energy price and operating boundary conditions and could be evaluated by many factors such the payback period, IRR and NPV. The Levelized Cost of Heating (LCOH) has been proposed as a more direct indicator to benchmark the economic performance of heat suppliers [4] and be calculated as follows:

where the amortized capital cost represents the nameplate capacity of the heating facilities; the Capacity Factor represents percentage uptime; the Cost of Energy represents energy price such as natural gas price and electricity price; And the Energy In and Energy input represents such natural gas and electricity input and output amount; the Heat Out represents the heat supplied by heating facilities; the O&M represents the Operating and maintain cost.

The detailed assumptions for LCOH calculation are shown in Table 6. As mentioned in Section 2.1, the COP value of buildings’ heat pump is calculated by eq. (3), with the average ambient temperature as the heat source’s temperature [59].

where ΔT_lift represents temperature lift, T_sink is the outlet temperature, and the fitting parameters are as follows: a = 1.4480·10¹², b = 88.730, c = − 4.9460, d = 0.0000;

As mentioned in Section 2.2, the COP value of industrial heat pump is conservatively set as 2.5 that most waste heat based industrial heat pump could reach.

The CO₂ emission of heat-pump-based-heating is mainly from its input power’s CO₂ emission. Currently the CO₂ emission intensity of China’s power generation is around 581 g·kWh^− 1 due to coal-dominant power generation mix. In such cases, as shown in Fig. 9a, the total CO₂ emission intensity of heat pump would be comparable to that of natural gas-based heating, when the COP is above 3. Such indicator could be relatively achieved by air source heat pumps in most climate zones in China.

Figure 9 b further shows the CO₂ emission intensity of heat pumps under current and future scenarios. With the proportion increasing of renewable power in China, it is estimated that the power sectors’ average CO₂ emissions intensity will be reduced to 320 g·kWh^− 1 by 2030. In that case, the CO₂ emission of heat pumps will be less than that of natural-gas heating even with a COP less than 2. For scenarios where renewable electricity is used, due to the relatively lower CO₂ emission intensity less than 100 g·kWh^− 1, even the electrical resistance heating will have a less CO₂ emission than natural gas heating, let alone heat pump-based heating.

Based on the previous analysis, with progressive decarbonization of power grid, heat pumps would be one of the optimal clean-heating-facilities. Therefore, it is expected a huge emission reduction potential from heat pump will be fully unleashed and we provided the detailed methodology. We defined the heat pumps’ share in 2060 for industrial and buildings’ sector respectively. It should be addressed the share herein is a relative proportion, for example, based on current technology level, heat pump could cover ~ 100% buildings’ heating demands and ~ 40% industrial heating demands(there is a lack of accurate proportion of industrial heat demand by temperature level and herein 40% is adopted which would be reasonable in many research [9,50,60]), and these proportion that could be replaced by heat pumps are set as the maximum share.

By combining such statistics with the electricity grid emission projection as illustrated in method section (Heat pump decarbonization potential assessment), the decarbonization potential of heat pump is projected as shown in Fig. 8. As shown in Fig. 10, with an increasing share of heat pumps in buildings’ and industrial sectors, the maximum decarbonization potential would reach ~ 2232 Mton, with GHG emission offset potential reaching ~ 1532 Mton and ~ 670 Mton for buildings and industry heating respectively (The industrial energy consumption is predicted to be less than buildings’ energy consumption in 2060, and thus results in a less GHG emission reduction in 2060 for industrial sector). The overall mitigation potential is more than 20% of China’s current CO₂ emission. Yet the current heat pumps’ market share is around 5% for buildings sector [6] and even lower in industrial sector.

The economic factors have a more significant effect on the users’ willingness-to-pay for the heating facilities [59,61,62], especially with poor economic conditions and cold climate conditions [63]. Herein, two typical scenarios are set for buildings’ sector and industrial sector. In order to benchmark such economic comparison, the Levelized Cost of Heating (LCOH) has been proposed as a useful indicator; In such context, an economic comparison between heat-pump-based heating and other typical heating modes (using natural gas and electrical resistance as proxy) as shown in Fig. 11.

For buildings’ heating, herein a 20 kW heater is taken as a typical product. The LCOH presented in Fig. 11 a with ambient air as the heat source, and calculation is based on 15-year lifetime and 4000-hour annual operating time. The results show that the average LCOH of heat-pump-based-heating is around CNY 0.392·kWh^− 1, whereas LCOH of natural-gas-based-heating is around CNY 0.396·kWh^− 1, which shows the heat pump for residential heating has an overall economic advantage over natural gas in 20 provinces of China, and the rest provinces are mainly located in northeastern with relatively cheaper access to natural gas. Compared to electrical resistance heating, the LCOH values of heat pump are always lower as shown in Fig. 11 b.

For industrial heating, herein a 0.5ton·h^− 1 steam boiler(~ 350 kW) is taken as a typical product. The LCOH value is calculated on the basis of 15-year lifetime and 6000-hour annual operating time. The results in Fig. 11 c show that the LCOH of heat pump is lower than that of natural gas heating in 26 provinces of China, most of which are located in the country’s southeastern region with relatively higher natural gas price and meanwhile with good industrial base. Compared to electrical resistance heating, the LCOH values of heat pump are always lower as shown in Fig. 11 d.

To summarize, heat pumps have obvious economic advantages when compared to electrical resistance heating due to higher efficiency, while heat pumps could be economic comparable to natural gas in most cases. For buildings’ heating, heat pump has obtained overwhelming advantage over natural gas in most region of China. For industrial heating with relatively higher heating temperature and lower efficiency, the predominance area of heat pump is concentrated in the southeast region of China, which accounts for more than 70% of China’s total industrial revenue.

As a climate-sensitive technology, the energy performance of heat pump is location specific. For example, the south and southeast regions is with good industrial base, relatively warmer climate condition and relatively higher natural gas price. Such conditions would be better for heat pump-based heating, especially for some pilot heat pump projects such as air-source heat pump boilers [13].

Comparatively, increasing heat pump heating market share in north regions of China is more challenging: on one hand, there already exists fossil-fuel based central heating in many parts of north China, which creates barriers for heat pump penetration. In such cases, strategic use of existing fossil fuel heating capacity, together with heat pumps, as a so-called “dual-source system” could potentially create a win-win situation where heat pumps provide most of the heating energy demand with fossil fuel shaving the winter peak. With increase of heat pump efficiency and decrease of cost, heat pump will be more prepared to expand in north regions in China.

Participation of heat pumps in demand response programs harbors large potential on aiding variable renewable energy integration. Moreover, for newly added heating demand in both north and south regions of China, fossil fuel-based heating through furnaces and boilers, should be suspended in almost all situations. Such infrastructure will be obsolete in a carbon neutral world, and related investors will face stranded asset risk. Divestments of fossil fuel heating infrastructure must be part of the energy banking strategy in a carbon-constrained future.

After some initial experiences are accumulated through demonstration projects, both technology R&D and novel business model are further needed to facilitate large-scale application of heat pump technologies. In terms of technology R&D, the heat pump manufactures should focus on improving energy efficiency. Compressor is regarded as the heart of heat pump system, which also bears most of the energy loss. Currently, the adiabatic efficiency of compressors is less than 70%, whereas the adiabatic efficiency of permanent-magnet synchronous frequency-convertible centrifugal compressor could reach 75% ~ 88%. However, such high efficiency compressors mostly are applied in large-size unit, thus associated with high capital cost. Continuous R&D on small-size heat pump compressor is regarded as the first principle for improving heat pumps’ energy efficiency.

For residential heat pump products, chasing better thermal comfort, especially in cold climate regions, is also very important. Measures including air distribution design and noise reduction in residential heat pump are important. For industrial heat pump products, chasing high temperature heat supply using low temperature heat source (e.g., the temperature difference between heat source and heat sink is more than 100 °C) through large temperature lift heat pump is a future direction. Such exploration has not been covered enough so far [13]. Expanding the working temperature of industry heat pumps (vapor compression heat pumps, absorption heat pumps, adsorption heat pumps, and chemical heat pumps) still needs extended R&D through both learning by research and learning by doing. Given the huge market potential brought by the elimination of coal-fired boilers, investment on such high temperature industrial heat pump should attract more attention from related companies.

Business models also play an indispensable role here. In terms of market regulation, novel business models should be explored from the perspectives of purchase, after-scale, recycle. In north China, utility companies could work on heat pump-based district heating. Such business models rely on big utility companies to supply heat as an energy commodity similar to electricity. It has been proven that heat pump-based heating could create even lower energy price than the conventional coal-based district heating if based on waste-heat recovery [34]. In such cases, heat purchase rather than heat pump purchase could be a suitable business mode [64]. For industrial heating, the success of waste heat-based heat pumps for industrial processes and steam generation on a large-scale, also proves the effectiveness of such a business modes, such pilot demonstrations for such possibilities should be designed and tested to kick off the scalability process. Moreover, the carbon tax/price will further strengthen economic advantages of heat pumps.

Customer experience is also an important dimension that should not be overlooked during the scalability pathway. Compared to traditional utility companies, which have established a quick and mature customer response system over years, the heat pump players have not found an integrated manner to provide same-level customer service. Such concerns become a social barrier for heat pump acceptance among potential users. Digitalization could be a powerful assistance in such cases, if remote intelligent diagnosis could solve most of the issues, customer service level could be largely improved with cheaper cost and quick response. Meanwhile, considering the market sharing proportion of heat pumps’ further increases, reproduction and recovery on heat pumps should be considered, especially for the heat pumps’ substitution caused by the high-GWP refrigerants elimination.

Policy design will also play important roles at various points along such a technology roadmap. Intensification of existing heat pump support policy, including but not limited to building energy performance regulation, heating emission standards, carbon taxation/financial incentive are needed to support the sustainable increase of heat pump sales and make heat pumps more competitive with their fossil fuel counterparts, eventually dominating the heating decarbonization market. Waste heat recovery heat pump for industry applications is truly worthy to be supported, thus that industry heat pump users can develop their own heat pump for specific heating application, which is more complicated than building heating and cooling. Regardless of technology advancement, public awareness of heat pump technology should be raised through active public education as well (the fact that heat pump could have COP over 100% still seems unreasonable for many policy makers who are not familiar with the technology in China). The experience from other sectors has also shown that technologies that are initially unfamiliar to the public can become mainstream options over time, if attention is paid to product standards, information dissemination and installer training. In short, design of heat pump support policy (in both residential and industry sector) should be integrated into China’s carbon neutral legalization and carefully considered in the energy transition pathways.

Considering the three aforementioned aspects, a provincial heat pump development roadmap is proposed in the perspective (Fig. 12). In near-term, heat pump could accumulate development momentum from China’s southern provinces. On one hand, ambient air temperature is one of the most important factors for air-source heat pump energy performance. Given the relatively higher ambient temperature, heat pump in the southern provinces tend to have higher COP and thus better economic competitiveness. Therefore, heat pumps in these provinces are more likely to be developed in larger scale in near term. Heat pump is essentially an electricity driven heating facility, thus its decarbonization potential is largely determined by the local power emission intensity. In China, lots of southern regions like Jiangxi, Hunan, Guangxi, Yunnan, are rich in hydropower, as a result, the CO₂ emission reduction potential from heat pump will be more obvious in those provinces. In middle term, heat pump could penetrate to mild climate provinces, like Hebei, Shanxi, Shandong, Shaanxi, and Hubei. The power generation mix in these provinces are more fossil fuel dominant at the current stage, so the heat pump decarbonization potential would be greater with progressive power sector decarbonization in these provinces. In long term, heat pump heating will be necessary in northeastern and northwestern provinces of China, like Xinjiang, Qinghai, Gansu, Ningxia. These provinces have abundant fossil fuel resources, thus natural gas price is relatively cheaper. The population density of these regions is smaller, and there is usually a lack of large industrial base. Therefore, it is more suitable for these regions to use natural gas in near-term at least, until in long-term heat pumps with both higher efficiency and lower cost are available.

Comparatively, in South China region with relatively higher ambient air temperature in winter (mainly over 0 °C), it is more suitable to develop distributed heat pump modes. For the northeast provinces, the main barriers come from the low ambient air temperature. Multi-stages heat pump cycle and geothermal heat pumps for residential heating might be a potential solution, yet such heat pump configuration will increase the complexity of system and undermine its reliability. Sequentially, large-scale heat pump application in such provinces needs further R&D and demonstrations.

LCOH :Levelized Cost of Heating;GHG :Greenhouse Gas;η:Heating efficiency;HP :heat pump;NG :natural gas;ER :electrical resistance

This research was supported by the Key Project of the National Natural Science Foundation of China (52036004). The authors gratefully acknowledge the financial support from the Research Council of Norway and user partners of HighEFF (Centre for an Energy Efficient and Competitive Industry for the Future, an 8-year Research Centre under the FME-scheme).

As the corresponding author, RZW proposed the concept and guided the whole research. HZY and CZ conducted the data analysis and drafted the manuscript, CZ suggested the decarbonization potential calculations, provided draft on energy structure, policy implement, and provided key data on the power sector and decarbonizations. BH and ZYX provided draft on compression heat pumps and absorption heat pumps respectively, ZS provided draft figures on heat pump technology roadmap. All authors contributed to the interpretation of the findings. The authors read and approved the final manuscript.

Open access funding provided by Shanghai Jiao Tong University. Sponsored by National Natural Science Foundation of China: Study on cycle construction and application of large temperature lift heat pump and steam generation system (52036004).

The availability of any original data is presented in Section 3 with detail methodology, while the detailed data in our database and we have the right to freely share or copy and redistribute the material in any medium or format.