Introduction

Cement is the key construction material used for urbanization. With the development of the world economy, cement consumption has increased 2.7 times from 1.5 billion tons in 1998 to 4.2 billion tons in 20201. As an energy-intensive industry, cement production contributed about 8% of the total anthropogenic CO2 emissions and 25% of industrial CO2 emissions worldwide, making it the second-largest industrial CO2 emitter globally2,3. China’s cement industry has produced 50–60% of the world’s cement in recent decades4. Cement production has accounted for 6.8% of China’s total energy consumption and generated enormous quantities of air pollutants, including about 8% of China’s total CO2 emissions5,6,7. The cement industry is therefore regarded a key industry requiring CO2 and air pollutant emission reductions for climate change mitigation and clean air.

Cement production can be summarized as “two grinding processes and one burning process.” The raw materials are ground and mixed, the mixed materials are burned in a kiln to produce clinker, and the clinker is mixed and ground with other substances in the final step. Coal is the primary energy source for the “one burning process,” which makes that process the most energy-intensive, contributing a third of the total CO2 emissions discharged in cement production. The remaining two-thirds of the CO2 emissions come from the calcination reaction of carbonate materials such as limestone in the kiln8. This reaction is one of the main factors impeding CO2 emission reductions in the cement industry. Electricity is the primary energy source of the “two grinding processes”, the processes mainly responsible for the indirect emissions from the cement industry. For the reasons explained, therefore, the grinding and burning processes in cement production are the focal points of the emission reduction efforts in the industry3.

The Technical Guidance on Carbon Emission Reduction in the Cement Industry issued in China in November 2022 summarizes 16 energy-efficiency-technology categories and five intelligent control systems used to reduce CO2 and air pollutant emissions during clinker production in order to meet the tighter national energy consumption indicators stipulated by the national standard GB16780-2021. The advanced preheater technology, for example, can effectively save 3–3.5 kg oil equivalent (kgoe)/t of clinker9. The development of clinker substitutions (slag, fly ash, volcanic rock, etc.) and adoption of alternative fuels (refuse-derived fuels (RDF), biomass fuels, plastics, rubber, leather, waste tires, coke oven gas, hydrogen, etc.), moreover, are expected to save about 53% of CO2 emissions up to 205010,11. Replacing the limestone with calcareous industrial solid wastes such as calcium carbide slag, metallurgical slags, coal ashes, and concrete crusher residues, meanwhile, can significantly reduce the CO2 and air pollutants emitted from the decomposition of calcareous materials during the raw material combustion in the kiln12. Improving the electrification of the cement industry and achieving clean electricity by reducing the proportion of thermal power and increasing the ratio of renewable energies are effective ways to reduce indirect emissions and achieve a low-carbon cement industry. Carbon capture and storage (CCS) is a process closely linked to CO2 removal13,14. Post-and oxy-combustion captures are primarily to be assessed for cement industrial decarbonization, as these two methods can more effectively target CO2 emissions from both fuel combustion and production processes compared to the pre-combustion capture. CCS, however, merits closer attention, as a CCS scenario is expected to consume 6.7% more energy than a non-CCS scenario in 205015.

For several reasons, the reliability of the existing research needs to be improved at both the national and provincial levels. First, the future cement demand has usually been estimated based on empirical experiences or the correlations of demand with population and economic development. More detailed estimates of the future cement demand will have to be realized by categorically considering the end-user demand trends. Second, differences in the cement quality and the corresponding environmental effects will need to be fully considered when analyzing the low-carbon development of cement going forward. Detailed estimates of the potentials of energy and material substitutions will also be needed. Third, electricity consumption and related indirect emissions need to be accounted for when discussing the effects of decarbonization measures.

In the current study, we address research needs while exploring the greener transition pathways of the cement industry in China across all of the available options. We created a demand projection model and bottom-up integrated assessment model, AIM/Enduse [China-Cement], to analyze the future cement demand in detail, assuming different levels of cement quality and emission reduction potentials in China’s cement industry. Our models consider business-as-usual scenarios (BaU, “B” in the scenario code in Table 1 and Supplementary Table S1) and countermeasures (“C” in the scenario code in Table 1 and Supplementary Table S1). Three cement quality scenarios were considered “low”, “medium” and “high” quality (“L”, “M”, and “H”, respectively, in the scenario code in Table 1; for the detailed parameter settings, see Supplementary Methods and Supplementary Table S2). Cement demands were divided into “high” and “low” demand (only the high-demand scenario is tagged as “DH” in the scenario code in Table 1; see Supplementary results 1). In addition, the penetration rates of advanced technologies, different emission tax levels (“TAX” in the scenario code in Table 1), CCS (“CCS” in the scenario code in Table 1), energy and material substitutions, and clean electricity were also analyzed. After our simulation analyzes, we found that the industry selects CCS only when environmental taxes are imposed, as CCS is cost-competitive only if price incentives are in place. Correspondingly, we designed 13 scenarios (see Table 1) after testing the sensitivity of the environmental taxes (see Supplementary Table S1). The targets and novel points of our research can be summarized as follows: (1) Projecting province wise cement demand by demand categories (different usages) and considering differences in cement quality due to different C2Cr; (2) reducing emissions of CO2, air pollutants, and mercury and achieving cleaner production by fully promoting energy-saving technologies and CCS by constructing a provincial simulation model; (3) analyzing the effects of economic levers such as tax policy on energy transformation and production structure transformation and then quantifying the emission reduction effects; and (4) analyzing how the development of alternative fuels and alternative raw materials reaps co-benefit effects in reducing air pollutants and mercury.

Table 1 Overview of the scenario settings
Full size table

Results

Future cement demand in different cement user categories

The future cement demand is estimated on the basis of the development of downstream industries and the corresponding cement intensity. Given that construction-related sectors are the major sources of cement consumption, we classify the ultimate cement consumer into 13 categories: two in building construction and maintenance (residential building and commercial building), five in road construction and maintenance (expressway, first-class road, second-class road, other-class road, and substandard road), two in civil engineering (subway and sewage pipeline construction), and railway, agricultural infrastructure, transport loss, and other uses.

By analyzing the stock of cement consumed, the estimated historical cement consumption rose from 1332.5 million tons (Mton) in 2007 to 2484.8 Mton in 2014 and then held steady at ~2400 Mton from 2015 to 2020 (Supplementary Fig. S1). The differences between the estimated results and statistical data published by the National Bureau of Statistics (NBS) range between −5% and 7.8%. Two assumptions of future cement consumption are derived according to two trends: rapid growth before 2014, explained as high cement demand, and plateau development after 2014, explained as low cement demand (see “high” and “low” in Table 1). In the high-cement-demand scenario, cement production will increase to 3293.9 Mton in 2050 and decrease slightly to 3184.9 Mton in 2060. In the low-cement-demand scenario, cement production fluctuates up to 2635.4 Mton in 2060 (Supplementary Fig. S1).

The characteristics of cement demand vary by region. The Beijing‒Tianjin‒Hebei region (B–T–H) will account for 9.3% of the total cement demand in 2060 under low-cement demand scenarios. Shandong, Henan, and Hunan Provinces will grow rapidly, and cement consumption will rise from 5.6%, 5.8%, and 3.9% in 2020 to 6.1%, 6.4%, and 6.9% in 2060, respectively. However, the ratios of Guangdong and Zhejiang Provinces will decrease from 8.3% and 5.5% in 2020 to 6.3% and 2.5% in 2060, respectively (Fig. 1, Supplementary Fig. S1, and Supplementary Fig. S2).

Fig. 1: Provincial cement demand under the low-cement-demand scenario.
figure 1

Provincial cement demand with corresponding fuel consumption (unit: million tons of oil equivalent (Mtoe), blue bar on the left axis), electricity consumption (unit: Mtoe, orange bar on the left axis), and CO2 emissions (unit: Mton, gray bar on the right axis) in 2030 and 2060 in the CH-CCS-TAX6 scenario. Note: Beijing, Tianjin, and Hebei are combined into one region for analysis, referred to as Beijing–Tianjin–Hebei (B–T–H).

Full size image

Clinker production

Cement is produced by mixing clinker with gypsum and other materials. More clinker is consumed to produce cement of higher quality, which results in higher emissions and incurs higher costs for consumers. Turning to the cement quality improvement plan proposed by the Chinese government, we assume three conditions reflecting the cement quality improvements realized (see the Methodology section and Supplementary Methods for more detail): low-, medium-, and high-quality conditions (see “low”, “medium”, and “high”, respectively, in Table 1).

The clinker production demand under these three conditions is estimated according to the detailed infrastructure development information derived from cement demand analysis. Clinker production in the low-cement-demand scenario with low-quality cement grows from 1420.9 Mton to 1568.8 Mton in 2060 (Fig. 2). With improved cement quality, the production rises to 1681.2 and 1790.9 Mton in the medium- and high-quality-cement scenarios. Overall, 20.9% more clinker is produced in the high-cement-demand scenario than in the low-cement-demand scenario. The C2Cr, an important indicator of cement quality, increases from 59.3% in 2020 to 59.5%, 63.8%, and 68.0% by 2060 in the low-, medium-, and high-cement-quality scenarios, respectively. The cement quality is the only parameter affecting the C2Cr and the only parameter affecting the clinker production, regardless of technical upgrades, tax policies and other parameters in this research. Figure 2 demonstrates that a greater increase in C2Cr leads to a greater improvement of cement quality and a steeper rise in fuel consumption.

Fig. 2: Clinker production and clinker-to-cement ratio (C2Cr) under different cement quality scenarios.
figure 2

The linearly rising dot plots represent a positive correlation between clinker production (x-axis) and energy consumption (y-axis), that is, higher clinker production consumes more fuel. The different colors of the dot plots represent clinker production under different crude steel demands and different cement quality scenarios. The bubble size indicates the clinker-to-cement ratio (C2Cr) (ranging from 59.0% to 68.1%). The corresponding fuel consumption data are from the BaU scenarios.

Full size image

Energy consumption

Higher cement quality and more clinker production result in higher fuel consumption. In the BaU scenarios (BL, BM and BH), the cement industry’s fuel consumption increases from 118.7 Mtoe in 2020 to 131.1 Mtoe, 140.5 Mtoe, and 149.6  Mtoe in 2060 in the low-, medium-, and high-quality-cement scenarios, respectively (Fig. 3a). Compared with the low-cement-demand scenario (BL-DH), the high-cement-demand scenario (BL) increases fuel consumption by 20.9%. Fuel consumption is 40.5% lower in the countermeasure scenarios with technological development (CL, CM, and CH) than in the corresponding BaU scenarios in 2060. The fuel consumption gap between high-quality cement and low-quality cement in the countermeasure scenarios in 2060 is 11.0 Mton, 40.5% lower than the 18.6 Mton fuel consumption gap between BH and BL. This result shows that the best practices can effectively reduce fuel consumption when cement quality is improved. Although tax policy alone has less impact on energy conservation than do nontax scenarios and energy consumption remains unchanged when the environmental tax increases, the energy structure is optimized, with the ratio of coal consumption decreasing from 94.0% in 2020 to 72.9% and 40.3% in 2060 under the countermeasure scenarios with low-tax and high-tax conditions, respectively (Fig. 3c). Collaborative waste treatment under a high-tax policy is expected to replace 22.4% of the fuel consumption in the industry by 2060. Kilns and preheating devices are the major consumers of fossil fuels in cement production. Energy-saving technologies such as low-temperature cogeneration and resistance reduction optimization in preheaters improve energy efficiency by reducing fuel consumption. The widespread adoption of these technologies conserves 24–27 Mtoe of fuel in the BaU scenarios. Compared with the BaU scenario, the maximum fuel savings reach 60.6 Mtoe in the CH-CCS-TAX6 countermeasure scenario, with high C2Cr and high taxes in 2060.

Fig. 3: Fuel consumption, fuel intensity, and fuel structure analysis for different scenarios.
figure 3

A Fuel consumption: the blue bar represents fuel consumption in 2020. The gray bar indicates the fuel consumption in BL-DH scenario in 2060. The yellow, brown, and green bars represent fuel consumption under low-, medium-, and high-cement-quality scenarios, respectively, in 2060. B Fuel intensity: the legend remains consistent with (A), illustrating variations in fuel intensity across the same scenarios. C Fuel structure: the red, blue, grey, and green rings represent the consumption of coal, natural gas, oil, and bio-energy, respectively.

Full size image

The fuel intensity of the cement industry increased slightly from 49.6 kgoe/ton cement in 2020 to 49.7 kgoe/ton, 53.3 kgoe/ton, and 56.8 kgoe/ton in the BaU scenarios with low-, medium-, and high-quality cement, respectively (Fig. 3b), and it decreases by 40.4% and 40.5% in the countermeasure scenarios without tax and with tax policy, respectively. The fuel intensity of the cement industry decreases to 29.6 kgoe/ton, 31.7 kgoe/ton, and 33.8 kgoe/ton under the best-practice scenarios with low, medium, and high cement quality, respectively (CL-CCS-TAX6, CM-CCS-TAX6, and CH-CCS-TAX6).

The fuel intensity of the provincial cement industry remains between 26.5–30.8 kgoe/ton and 33.1–34.4 kgoe/ton in the CL-CCS-TAX6 and CH-CCS-TAX6 scenarios, respectively, which is 21.4–40.9% lower than that in the BaU scenario (Supplementary Fig. S3). Yunnan, Xinjiang, Inner Mongolia, and Heilongjiang Provinces had higher fuel intensities than other regions did in 2020 and remained at high levels during the study period. Despite the smaller differences in energy intensity between provinces, there are large differences in energy consumption among provinces from 0.3 Mtoe in Tibet to 8.3 Mtoe in B–T–H in 2060 under the CH-CCS-TAX6 scenario. Zhejiang Province has the most significant energy-saving effect, reducing energy consumption by 65.1% and 69.6% during the research period under the best-practice scenarios with high and low cement quality, respectively (CH-CCS-TAX1/CH-CCS-TAX6 and CL-CCS-TAX1/CL-CCS-TAX6 scenarios, Supplementary Fig. S3). The fuel consumption of Jiangsu and Jilin Provinces will decrease by 55.8% and 47.1%, respectively, during the research period under the CH-CCS-TAX1/CH-CCS-TAX6 scenario. This is because, in addition to technological advances, the demand for cement in these provinces has rapidly decreased. The largest increase in fuel consumption is observed in Guizhou Province, with fuel consumption increasing by 69.0% from 2020 to 2060 under the CH-CCS-TAX1/CH-CCS-TAX6 scenario (Supplementary Fig. S3). Moreover, the energy consumption of Qinghai, Hunan, Chongqing, and Guangxi Provinces in 2060 will increase by 56.3%, 44.7%, 31.1%, and 9.0%, respectively, compared with that in 2020. This indicates that the effect of energy conservation measures is offset by the increased demand for cement, resulting in increased energy consumption.

Electricity consumption

The electricity consumption increases slightly from 13.7 Mtoe in 2020 to 15.1 Mtoe in 2060 under the BaU scenario with low cement quality (BL). In the BaU scenarios with medium and high cement quality (BM and BH, respectively), the industry consumes 15.7 and 16.2 Mtoe electricity, respectively, in 2060, or 3.6% and 7.0% more than it consumes under the low-cement-quality scenario. The industry consumes 20.9% more electricity in the high-cement-demand scenario with low-quality cement (BL-DH) than in the low-cement-demand scenario with low-quality cement (BL) in 2060 (Fig. 4). With the penetration of the best available technologies in the countermeasure scenarios without taxes (CL/CM/CH) (see Supplementary Results 3), electricity consumption decreases by 28% to 29% compared with that in the corresponding BaUs in 2060. With the penetration of CCS promoted by different tax levels, electricity consumption increases by 38.4% to 41.8% and by 37.9% to 41.3% in the scenarios with low taxes (TAX1) and high taxes (TAX6), respectively, in 2060 compared with the consumption levels in the corresponding BaUs. CCS adoption occurs later in the low-tax scenarios than in the high-tax scenarios, which results in lower cumulative electricity consumption in low-tax scenarios (Fig. 4). The electricity consumption of the kiln process will surge by 263.7% in the tax scenarios compared with the BaU scenarios in 2060, as the CCS is installed in the kilns. While the grinding process is also electricity intensive, the adoption of advanced technology reduces electricity consumption by 20% to 22% in the tax scenarios compared with the BaU scenarios in 2060. Geographically, the Beijing‒Tianjin‒Hebei region consumes 2.14 Mtoe in 2060, the most electricity for cement production, followed by Hunan and Henan Provinces, with 1.6 Mtoe and 1.5 Mtoe, respectively. Hunan, Guizhou, and Qinghai Provinces experience the fastest rates of growth in electricity consumption during the research period, at 222.5%, 277.0%, and 253.1%, respectively (Supplementary Fig. S4).

Fig. 4: Electricity consumption and electricity intensity of cement industry under different scenarios.
figure 4

Solid bar charts represent the cumulative electricity consumption of the cement industry from 2020 to 2060 (left axis). Shaded bar charts represent electricity intensity of the cement industry in 2060 (right axis). Different colors represent different demand and cement quality scenarios.

Full size image

The electricity intensity decreases by up to 27.8%, from 5.7 kgoe/ton in 2020 to 4.1 kgoe/ton in 2060, in the low-quality-cement countermeasure scenario without tax (CL). In the CH-CCS-TAX1 scenario, however, electricity intensity increases to a peak of 8.71 kgoe/ton during the research period because of high-quality cement and low taxes. The provincial electricity intensity ranges from 8.5 to 8.8 kgoe/ton. The electricity intensity in the CH-CCS-TAX6 scenario with a high carbon tax is ~0.4% lower than that in the CH-CCS-TAX1 scenario because of the lower degree of coal grinding caused by lower coal consumption (Supplementary Fig. S4). Geographically, Inner Mongolia and Heilongjiang Provinces have the highest power intensity at ~8.8 kgoe/ton. Tibet and Hainan Provinces have the lowest electricity intensity at 8.5 kgoe/ton.

Emission projections for CO2 and air pollutants

Both direct emissions from fuel combustion and production processes and indirect emissions from electricity consumption are considered in this research. In general, high-quality cement results in more emissions and high emission intensity. CO2 emissions increase slightly from 1205.1 Mton in 2020 to 1330.5 Mton in 2060 in the BaU scenario with low cement quality (BL). With the improved cement quality in the BaU scenario (BM and BH), CO2 emissions increase by 7.2% and 14.2%, respectively, over the levels in the low-quality-cement scenario (Fig. 5). High cement demand also leads to high CO2 emissions. With technological progress in countermeasure scenarios without taxes (CL, CM, and CH), CO2 emissions decrease by 19.2% from 2020 to 2060. With incentives from carbon and environmental tax policies in place, the widespread adoption of CCS leads to CO2 emission reductions of 71–74% and 75–78% during the research period in the low-tax scenarios (CL-CCS-TAX1, CM-CCS-TAX1, and CH-CCS-TAX1) and high-tax scenarios (CL-CCS-TAX6, CM-CCS-TAX6, and CH-CCS-TAX6), respectively. CO2 emissions are the lowest, at 250.1 Mton in 2060, under the low-quality-cement scenario with a high-tax policy (CL-CCS-TAX6 scenario), 35.4 Mton less than those in the high-quality-cement scenario (CH-CCS-TAX6 scenario). This gap is 81.2% smaller than the gap between the BL and BH, indicating that tax policies can balance low-carbon development and cement quality improvement. From the regional perspective, CO2 emissions in each province decrease by 53–90% from 2020 to 2060 under the CL-CCS-TAX6 scenario and CH-CCS-TAX6 scenario (Supplementary Fig. S5). The Beijing‒Tianjing‒Hebei region has the highest CO2 emissions, at 26.7 Mton, in 2060 under the CH-CCS-TAX6 scenario. Hunan and Henan Provinces have CO2 emissions of 19.7 Mton and 18.3 Mton, respectively.

Fig. 5: CO2 and air pollutants during the research period (unit: thousands of tons) from the cement industry under different scenarios.
figure 5

The line charts represent the trend of emissions for a time series. The color fill represents the difference in emission levels between different qualities of cement, and different colors represent different scenarios. The bar charts show cumulative emissions of CO2 and air pollutants during the research period from the cement industry under different scenarios.

Full size image

CO2 emission intensity increases slightly with improved cement quality in the corresponding BaU scenarios, whereas it decreases by up to 80.2%, from 503.2 kg/ton in 2020 to 94.9 kg/ton in 2060, in the CL-CCS-TAX6 scenario. Compared with the CL-CCS-TAX6 scenario, improving the cement quality in the CH-CCS-TAX6 scenario increases the CO2 emission intensity by 14.2%. The corresponding CO2 emission intensities in the CL-CCS-TAX6 scenario and the CH-CCS-TAX6 scenario range from 84.9 to 98.9 kg/ton and 105.1–110.2 kg/ton in 2060, respectively. Inner Mongolia, Heilongjiang, and Yunnan Provinces have higher emission intensity levels than other regions. Tibet has the lowest CO2 emission intensity.

To understand the cobeneficial and trade-off effects of decarbonization measures on China’s carbon-neutral target, we analyzed the emissions of six air pollutants, including methane (CH4), nitrous oxide (N2O), sulfur oxide (SO2), nitric oxide (NOX), carbon monoxide (CO), and particulate matter (PM, including PM2.5 and PM10) and one heavy metal, mercury (Hg) (Fig. 5 and Supplementary Fig. S6). As fuel consumption increases in the BaU scenarios (BL-HD, BL, BM, and BH), air pollutant emissions increase by 10.4–33.4% from 2020 to 2060. Owing to increased biomass consumption, CH4 emissions increased by up to 343.1% from 2020 to 2060 under the high-quality-cement-and-high-tax scenario (CH-CCS-TAX6). N2O, NOx, PM, and CO emissions decrease by 32.5%, 59.1%, 53.8%, and 53.6%, respectively, during the research period. SO2 and Hg emissions initially decrease but later increase slightly thereafter because of the higher biomass consumption and clinker production.

Indirect emissions from electricity consumption are also considered in this study, referring to the electricity production structure under a deep decarbonization scenario in the China Energy and Electricity Outlook16. Correspondingly, indirect CO2 emissions will decrease from 147.9 Mton in 2020 to net zero in 2060 in all countermeasure scenarios, as the proportion of thermal power will quickly decline and CCS will be widely available. The cumulative CO2 and air pollutant emissions in the countermeasure scenarios are 58.6% lower than those in the BaUs, at most, despite the extra power consumption caused by CCS.

Discussion

The carbon accounting system for cement clinker production is gradually being completed in China, and the cement industry is expected to become the next industry to be included in the national carbon trading market. This research develops a provincial AIM/Enduse [China-cement] model to focus on the barriers between low-carbon development pathways and quality improvement in the cement industry. Measures to reduce cement demand and promote energy substitution, material substitution, energy-saving technologies, and tax policies are set as key parameters of the model. The results show that building construction will still be the key cement consumer, whereas road and building maintenance and agricultural infrastructure will contribute to higher proportions of consumption. Compared with the BaU scenario, the best available measures reduce fuel consumption by up to 40.5% in 2060, and the corresponding fuel intensity of the industry decreases to 29.59 kgoe/ton. The proportion of coal decreases to 40.3%, whereas natural gas and biomass account for 37.7% and 16.4% of fuel consumption in 2060, respectively, under the high-tax scenarios. The extensively promoted collaborative waste treatment processes in the cement industry provide energy equivalent to 22.4% of fuel consumption in 2060 (see Supplementary Results 2). The electricity consumption increases to 67.3%, and the corresponding electricity intensity is 8.71 kgoe/ton by 2060. CO2 emissions are reduced by up to 78.2% in the low-quality-cement scenario with high taxes, and CCS contributes 75% to 77% of the CO2 emission reduction at most. Increased CH4, SO2, and Hg emissions caused by increased biomass consumption require emission control measures. The high energy intensity and emission intensity in Heilongjiang and Inner Mongolia also deserve our attention.

Improving the quality of cement according to government policy triggers a 13.7% increase in cumulative fuel consumption and emissions between the BL and BH scenarios. Higher taxes can narrow this gap between low-quality cement and high-quality cement in countermeasure scenarios. Therefore, increasing environmental taxes is one way to solve environmental problems in the short term: the tax should be set at no less than the tax6 level, i.e., 12.6 CYN per equivalent unit in 2030 and 46.8 CNY per equivalent unit by 2060, or an amount six times higher than the current pollutant treatment cost. Reducing cement demand is one of the most efficient ways to reduce CO2 emissions17. We conclude that the low-cement-demand scenario consumes ~17.3% less energy than the high-cement-demand scenario in 2060. Further deployment of advanced technologies can provide sustained incentives to reduce emissions from fuel combustion. Limestone substitutes have yet to be applied on a large scale; however, the proportion of production process emissions caused by limestone consumption will continue to rise.

With the deployment of energy recovery technology, up to 59.1 Mtoe of energy is recovered and reused. While China’s cement industry falls short of achieving zero carbon emissions in 2060 in our models, it still has a chance to contribute to China’s carbon-neutral target, as some room is allowed for the implementation of CCS and energy and material substitution (Supplementary Fig. S7). Moreover, second-generation technology for dry-process cement production, now years in development, will improve the productivity of the cement industry through process reforms, equipment upgrades, evolving low-carbon technology, energy conservation, emission reduction, and so on. As costs decline and the technology matures and is more widely adopted, the reduction potential in the cement industry will grow. CO2 absorption during concrete production and use can also offset certain emission that have not been considered in this study.

As CO2 emissions from the cement industry come mainly from coal combustion and calcined limestone, the high CO2 concentrations discharged from cement plants make it easier to capture the CO2. Cost, however, is a major factor limiting the popularization of CCS. The process of carbon capture and sequestration has no economic value to companies. One way to break through this barrier and possibly gain economic benefits is to adopt any of the various chemical and biological utilization technologies evolving alongside the methods to reuse captured carbon18. The chemical-utilization technologies of focus at present include processes to use CO2 for production in chemical industries such as organic carbonate and coal chemical. The bio-utilization technologies under development so far widely focus on the use of CO2 for food and beverage production. The time lag between the carbon capture and utilization needs to be considered when carbon utilization is encompassed within the research boundaries.

Fuel substitution and material substitution are effective ways to reduce CO2 and air pollutant emissions from coal combustion and production processes. Biomass with CCS is considered as an effective negative-emission technology, which is expected to replace 30% of coal combustion in the kiln19. Collaborative waste treatment, an approach to reducing and recycling municipal waste, has attracted attention in China’s cement industry in recent years, as the high temperature and negative pressure in cement kilns provide suitable conditions for harmless waste disposal20,21. Low investment costs for government subsidies and cost savings through reduced coal consumption are the two main forces driving the adoption and development of waste co-treatment in the cement industry. The Chinese government, moreover, has promoted a series of policies to increase the penetration rate of waste co-processing. The National Development and Reform Commission has been providing large cement enterprises near big or central cities with incentives and support to encourage the adoption new dry-process cement kilns for the treatment and disposal of industrial waste sludge and municipal solid waste, since 2006. In 2013, the State Council expanded its targeted number of waste co-processing lines to no less than 10% of all cement production lines operating in the country22. According to the assumptions in the 14th Five-year Plan, the comprehensive utilization rate of solid waste will be increased to 60% by 2025 by means of cement collaborative disposal and incinerator construction23. With the improvement of systems for the sorting and recycling treatment of municipal waste, the development prospects for collaborative waste treatment in cement kilns are expected to further improve.

Conclusion

The cement industry is the second largest CO2-emitting manufacturing sector in China. The CO2 emissions from the cement industry are produced from both fossil fuel combustion and the cement production processes themselves. In this study, we developed a detailed cement demand model to analyze the emission reduction potentials of the industry. We estimated 13 categories of cement consumption in 31 provinces over the period from 2020 to 2060, to explore the factors influencing cement demand and improve the credibility of future projections. The AIM/Enduse[China-cement] model is a bottom-up optimization model developed to analyze the emission reduction effects of advanced energy-efficient technologies, fuel and material substitutions, environmental tax policy, and CCS deployment. The model takes account of detailed production and energy-saving technologies such as the second generation of new dry cement production technology and solid waste pretreatment technology. The model also factors in the effects of the cement quality, given its important bearing on energy consumption. To offset the high fuel consumption caused by improved cement quality, a high environmental tax of no less than 12.6 CNY/equivalent units must be in place in 2030. The fuel intensity, meanwhile, decreases to 29.59 kgoe/ton. According to our estimates of CO2, six air pollutants, and mercury emissions from fuel combustion, the production process, and electricity consumption, the minimum emission intensity becomes 90.9 kg CO2/ton cement. Our next challenge will be to analyze the impacts of waste classification on collaborative treatment and the utilization of recovered carbon. Another research challenge will be to estimate the emission-reduction effects achieved by replacing limestone with calcareous industrial solid waste in cement production.

Methodology

Estimation of future cement demand

According to the input–output table in 2017, 60.1% of the products in the cement industry were consumed by construction-related sectors, and 29.0% were consumed by the gypsum and cement products sector. In total, 90.8% of the products in the gypsum and cement products sector went to construction-related sectors6. Given that construction-related sectors account for most cement consumption, we classify the ultimate cement consumers into 13 categories: two in the building sector (residential building and commercial building), five in road construction (expressway, first-class road, second-class road, other-class road and substandard road), two in civil engineering (subway and sewage pipeline construction), and railway, agricultural infrastructure, transport loss, and other uses. We also consider cement consumption for infrastructure maintenance. We do not account for cement imports, exports, or stocks in this study, as cement is poorly suited for long-distance transportation and long-term storage. China, therefore, has exported less than 1% of its cement products in recent years and imported none. The relevant data are obtained from the NBS6. The total cement demand in China is calculated via Eq. (1).

$${D}_{t}={C}_{{BD}}+{C}_{{RD}}+{C}_{{SP}}+{C}_{{RW}}+{C}_{{AgI}}+{C}_{{TL}}+{C}_{{OT}}$$
(1)

where \({D}_{t}\) represents the total cement demand of year t; \({C}_{{BD}}\) represents the cement consumed by two building sectors; \({C}_{{RD}}\) represents the cement consumed by five kinds of road construction; \({C}_{{SP}}\) represents the cement consumed by two types of civil engineering (subways and pipelines); and \({C}_{{RW}},{C}_{{AgI}},{C}_{{TL}}\) and \({C}_{{OT}}\) represent the cement consumed by railways, agricultural infrastructure, transport loss, and other uses, respectively.

The cement consumption in each category is basically estimated by multiplying the construction unit by the cement intensity, or what we call the cement utilization rate, as shown in Eq. 2.

$${C}_{i}={P}_{i}* {{CUR}}_{i}$$
(2)

where i represents the 13 categories of cement consumer; \({P}_{i}\) is the construction unit of category i; and \({{CUR}}_{i}\) is the cement utilization rate of category i; where i represents the 13 categories of cement consumption; \({P}_{i}\) is the construction unit of category i; and \({{CUR}}_{i}\) is the cement utilization rate of category i.

Previous studies have used two standard analysis methods for estimating future cement demand. One assumes that China has production and consumption trends similar to those of developed countries. Nevertheless, per capita cement consumption in China reached 1.68 tons in 20216, far exceeding the peak consumption levels recorded in developed countries such as Japan and the United States. The other assumes various relationships between cement demand and socioeconomic parameters such as gross domestic product (GDP) growth and the urbanization rate24,25,26. In this study, the national and provincial cement demand is analyzed in detail in the context of development plans and socioeconomic trends.

  1. (1)

    Cement consumption for buildings

    The building sector is separated into two sub-types, residential use and non-residential use, according to the different construction methods applied. The latter includes commercial and service use; office use; scientific research, education, and medical use; culture, sports, and entertainment use; workshop and building use; storage; and other uses.

    The building sector, especially the residential area, has a significant correlation with the indicators related to changes in the population and per capita living area. Economic development, the reconstruction of old urban areas, and changes in living habits from multi-generation households to single-generation households largely drove the growth of the building sector before 2015, especially from 2006 to 2014. Completed floor space, for example, drastically rose from 807.14 million m2 in 2000 to 4233.57 million m2 in 2014. The completed floor space has held steady at about 4000 million m2 in recent years, however, as the population growth rate began to slow from around 2015. As the building sector accounts for about 50% of the total cement consumption and has great impacts on both cement production and consumption, we carefully take two different approaches to estimating the future cement demand from the building sector. The first considers the inertia of the rapid growth trend of the building sector between 2004 and 2014. The second considers the steady phase after 2014. We consider the SSP2 (Shared Socioeconomic Pathways: middle-of-the-road challenges to climate mitigation and adaptation) scenario to model the future population changes of each province in China. Finally, we estimate the change in the future floor area based on the correlation function and information on the future population.

    $${C}_{{BD}}= {C}_{B{D}_{r}}+{C}_{B{D}_{{nr}}}={\sum}_{k}\left[\left({A}_{B{D}_{r}}+{{AM}}_{B{D}_{r}}\right)* {{CS}}_{k}* {{CUR}}_{B{D}_{r}}\right] \\ +{\sum}_{k}\left[\left({A}_{B{D}_{{nr}}}+{{AM}}_{B{D}_{{nr}}}\right)* {{CS}}_{k}* {{CUR}}_{B{D}_{{nr}}}\right]$$
    (3)

    where r and nr represent residential and non-residential buildings, respectively; \({A}_{{BD}}\) is the construction area, including newly constructed and redeveloped areas beyond the building service lives, based on data collected from NBS; \({{AM}}_{{BD}}\) is the area of existing buildings in need of repair, estimated based on the maintenance periods and maintenance funds of the buildings; \({{CS}}_{k}\) represents the consumption proportion of cement of different quality, i.e., low-quality (compressive and flexural strength: 32.5), regular-quality cement (compressive and flexural strength: 42.5), and high-quality cement (compressive and flexural strength: 52.5 and 62.5). The proportion of different cement quality in different construction types was also estimated through the survey for the subsequent analysis of cement low-carbon development; and \({{CUR}}_{{BD}}\) is the cement utilization rate of residential and non-residential buildings.

  2. (2)

    Cement consumption for road construction

    Roads in China are divided into five classes: expressway, first-class road, second-class road, other-class road, and substandard road. Some provinces have issued road development plans of their own. The latest National Expressway Network Planning states that the length of expressways will reach 162,000 km by 203527. Beijing city has built 22,264 km of road, including 1173 km of expressway, as of 2020, the final year of its 13th Five-year Plan. The city’s 14th Five-year Plan from 2021 to 2025 calls for an increase in the road network density by 0.4 km/km2 versus that at the end of the 13th Five-year Plan, and the total mileage in the city is forecasted increase to 23,150 km by 203528. The increase in road mileage is directly related to car ownership, which in turn is influenced by population and economic conditions. In China overall, the correlation coefficient between the road length and GDP per capita (GDPP) after 2005 is as high as 0.994. Hence, the future GDPP of China under the SSP2 scenario is used as an important driver for road mileage analysis in provinces that have not announced relevant plans. The road service life is assumed to be 100 years, with major maintenance carried out every decade. The road maintenance is correspondingly considered as shown in Eq. 4.

    $${C}_{{RD}}={\sum}_{j}{C}_{{{RD}}_{j}}={\sum}_{k}{\sum}_{j}\left({A}_{R{D}_{j}}+{{AM}}_{R{D}_{j}}\right)* {{CS}}_{k}* {{CUR}}_{{{RD}}_{j}}$$
    (4)
    $${A}_{R{D}_{j}}={{RM}}_{R{D}_{j}}* {{RW}}_{R{D}_{j}}$$
    (5)

    where j represents the five road classes; \({A}_{{RD}}\) is the construction area of each road class; \({{AM}}_{{RD}}\) is an area maintained every 10 years after construction; \({{CUR}}_{{{RD}}_{j}}\) is the cement utilization rate of each road class; \({{RM}}_{R{D}_{j}}\) is the road mileage; and \({{RW}}_{R{D}_{j}}\) is the road width.

  3. (3)

    Cement consumption for railway construction

    Based on the “medium- and long-term railway network planning,” the railway network will reach about 175,000 km, including about 38,000 km of high-speed railways, by 2025. In another 5 years, by 2030, the railway network will reach about 200,000 km, including 45,000 km of high-speed railway29. The railway mileage of each province is estimated based on the transportation development plan and socio-economic indicators. The construction of railway lines, bridges, and tunnels makes up the greater part of the cement consumption, which is calculated using Eq. 6.

    $${C}_{{RW}}={\sum}_{k}\left[\left({A}_{{RW}}+{{AR}}_{{RW}}+{{AM}}_{{RW}}\right)* {{CS}}_{k}* {{CUR}}_{{RW}}\right]$$
    (6)

    where \({A}_{{RW}}\) is the railway construction area; \({{AR}}_{{RW}}\) is a refurbished railway that has passed its service life; and \({{AM}}_{{RW}}\) is the area maintained after construction.

  4. (4)

    Cement consumption for municipal construction

    Two sources of cement consumption in municipal construction are considered: urban rail transit and sewage pipelines (see Eqs. 7–9).

    Urban rail transit includes subways, light rail monorail, and maglev trains. As of 2022, 247 urban rail transit lines were operating in 48 cities of 31 provinces, spanning a total length of 7970 km. Shanghai ranked first in total length, at 766 km, followed by Beijing and Chengdu6. In this study we compile statistics on the future development of urban rail transit in each province based on the short- and medium-term development plan with supplemental data on the historical construction conditions. The numerous tunnels and small numbers of bridges along railway lines consume cement, in addition to the main parts of the lines.

    Cement pipes are commonly used for sewers, another cement-consuming infrastructure to be considered. The length of China’s sewage pipelines in 2020 totaled 803,000 km. This statistic is directly related to the urban area and the correlation coefficient between the urban area and population, namely, 0.992. Correspondingly, the future length of sewage pipelines and the future urban construction area are estimated based on the population in SSP2 and the sewage pipeline density of the urban area in each province (see, Eq. 9 and Eq. 10).

    $${C}_{{CE}}={C}_{{MT}}+{C}_{{SP}}$$
    (7)
    $${C}_{{MT}}={\sum}_{k}\left[\left({A}_{{MT}}+{{AR}}_{{MT}}+{{AM}}_{{MT}}\right)* {{CS}}_{k}* {{CUR}}_{{MT}}\right]$$
    (8)
    $${C}_{{SP}}={\sum}_{k}\left[\left({L}_{{SP}}+{{LR}}_{{SP}}+{{LM}}_{{SP}}\right)* {{CS}}_{k}* {{CUR}}_{{SP}}\right]$$
    (9)
    $${L}_{{SP}}=\alpha * {A}_{{UC}}$$
    (10)

    where \({C}_{{MT}}\) is the cement consumed by urban rail transit construction; \({C}_{{SP}}\) is the cement consumed by sewage pipelines; \({A}_{{MT}}\) is the construction area of urban rail transit; \({{AR}}_{{MT}}\) is refurbished railway that has passed its service life; \({{AM}}_{{MT}}\) is the area maintained after construction; \({L}_{{SP}}\) is the length of sewage pipelines; \({{LR}}_{{SP}}\) is the length of refurbished sewage pipelines; \({{LM}}_{{SP}}\) is the length of pipes to be repaired; \(\alpha\) is the sewage pipeline density, a parameter that shows the correlation coefficient between the length of the sewage pipelines and the urban construction area (\({A}_{{UC}}\)); and \({{CUR}}_{{MT}}\) and \({{CUR}}_{{SP}}\) are the cement utilization rate of urban rail transit and sewage pipelines.

  5. (5)

    Cement consumption for agricultural infrastructure

    The construction of agricultural facilities includes diversion channels, dams, embankments, reservoirs, etc. Lacking information on actual construction, we estimate the cement consumption of agricultural infrastructure based on the agricultural investment and GDP (see Eq. 11). Agricultural investment is closely correlated with GDP and has accounted for 2.3% of the total GDP, on average, in recent years. As agricultural infrastructure has entered a stage of steady progress, we set the proportion of future investment in agricultural infrastructure to GDP at the same level, 2.3%.

    $${C}_{{AgI}}={\sum}_{k}\left[{I}_{{AgI}}* {{CS}}_{k}* {{CUR}}_{{AgI}}\right]={\sum}_{k}\left[\beta * {{GDP}}_{t}* {{CS}}_{k}* {{CUR}}_{{AgI}}\right]$$
    (11)

    where \({I}_{{AgI}}\) is agricultural investment; \(\beta\) is the correlation coefficient between agricultural investment and GDP; and \({{GDP}}_{t}\) is the GDP in year t. The cement utilization rate (\({{CUR}}_{{AgI}}\)) is expressed in tons/billion CNY in this equation.

  6. (6)

    Transport loss and other cement consumption

    Several forms of cement consumption other than the above are uncounted, such as park construction, overdue cement, sightseeing facility construction, high-voltage line bases, and other equipment. Altogether, these are expected to account for about 17.4% of the total cement consumption going forward. And based on our references, an estimated 1.5% of the total consumption will be caused by cement loss during transportation30,31.

Low-carbon development pathway analysis for the cement industry

Mitigation analyses by using the AIM/Enduse[China-cement] model

The AIM/Enduse model is a bottom-up integrated assessment model that simulates flows of energy and materials from primary energy and material supplies to secondary energy and material supplies. The model is chiefly used to satisfy the final energy and material services in the demand sectors by considering the detailed technology selections. It can be considered a linear optimization framework with the logic of cost minimization32,33,34. Energy and service flows are modeled to reflect key industrial production processes. Energy consumption and the corresponding emissions of greenhouse gases, air pollutants and mercury are calculated via the results of the technology selections. The annual emissions of CO2 and air pollutants are calculated via Eqs. (12–14).

$${{{ Q}}}_{i}^{m}={\sum}_{j}{\sum}_{(l)\in {W}_{j}}{X}_{{l},{i}}\cdot {e}_{{l},{i}}^{m}$$
(12)
$${e}_{{l},{i}}^{{m}}=[{f}_{{{0}},{l},{i}}^{m}+{\sum}_{k}{f}_{{k},{l},{i}}^{m}\cdot ({1}-{\varepsilon }_{{l},i})\cdot {E}_{{k},{l},{{p}},i}\cdot ({1}-{U}_{{k},l})]\cdot ({{1}}-{d}_{{l},{p},i}^{m})$$
(13)
$${X}_{{l},i}=\frac{{S}_{{l},i}}{{1}+{\alpha}_{{l},i}}$$
(14)

where \({Q}_{i}^{m}\) is emissions of gas type m in industry or region i (Unit: ton); \({X}_{l,p,i}\) is operating quantity of production equipmentlin region/industry i; \({e}_{l,i}^{m}\) is emissions of gas m per unit operation of production equipment l in region/industry i (Unit: ton); \({f}_{0,l,i}^{m}\) is emissions of gas m per unit operation of production equipment l in region/industry i (emissions from production processes in the cement industry, unit: ton); \({f}_{k,l,i}^{m}\) is gas m emissions from per unit consumption of energy k by production equipmentlin region/industry i (unit: ton/ton-energy); \({\varepsilon }_{l,i}\) is the energy efficiency improvement rate of production equipmentlfor region/industry i; \({E}_{k,l,p,i}\) is consumption of energy k by per unit operation of production equipmentlin region/industry i (unit: ton-energy); \({U}_{k,l}\) is the proportion of energy k in production equipmentlused for noncombustion operations; \({d}_{l,p,i}^{m}\) is the removal rate of gas m by end-of-pipe technologies per unit operation of production equipmentlin region/industry i; \({S}_{l,i}\) is the quantity remaining of production equipmentlin region/industry i that existed in the base year; and \(1+{\alpha }_{l,i}\) is the operating rate of production equipmentlin region/industry i.

The model also sets some basic constraints according to the actual production situation. For example, the energy and electricity demand should not exceed the maximum value of the total supply (after considering imports and exports), and the cement demand should be less than the total supply (after considering imports and exports). The ultimate goal of the model is to minimize the total cost, which includes the initial investment cost, the maintenance cost, the cost of raw material and energy consumption, and the cost arising from the implementation of the policy, such as the carbon tax or carbon trading. The model takes cost minimization as the objective function and constructs a linear optimization framework (Eq. 15).

$${{TC}}=\sum \left({{{C}}}_{{{i}}}+{{{C}}}_{{{o}}}+{{{C}}}_{{{t}}}\right)\to {{{\rm{Minimization}}}}$$
(15)

where \({C}_{i}\) is the annualized investment cost; \({C}_{o}\) is the annual operating cost, including the maintenance cost and material and energy cost; and \({C}_{t}\) is the cost of environmental policies, including the carbon tax and carbon trading.

AIM/Enduse is a recursive dynamic model that can calculate multiple years and analyze various scenarios, including policy countermeasures. The detailed formulations of AIM/Enduse are described in a manual report35. This model has been applied in the residential sector36,37,38, building sector39,40, service sector41, and transport sector42 and on a global scale43,44,45, national scale37,38, and regional scale46.

The process of cement production includes a limestone crusher, a grinding mill for raw material, a preheating system, a kiln, a clinker cooler, and a grinding mill for cement. The coal-based energy structure and production process emissions from limestone calcination leave the cement industry with no choice but to promote energy substitution and raw-material substitution technologies to achieve cleaner production. The cement industry is also working towards collaborative disposal of solid wastes as an important means of reducing fossil fuel consumption and meeting its social responsibility. In light of these trends and the industry’s plans for low-carbon development, our group developed a provincial AIM/Enduse [China-cement] model to analyze the energy and material flows in the cement industry under different technology and policy conditions. Cement is mostly sold within a radius of 200–300 km around a cement facility because of its short shelf life. With this in mind, we divide China into 31 provinces as separate research objects and assume that the cement production of each province only supplies its own demand.

Scenario settings

We select 13 scenarios to study (see Table 1) after testing the carbon price sensitivity of the model (see the complete 27 scenario settings in Supplementary Table S1 and Supplementary Fig. S8). Factors such as the range of future cement demand, the response of cement consumers to policies related to improving cement quality, the penetration of advanced technologies, tax levels, electricity decarbonization, and CCS deployment are considered in different combinations.

Two levels of future cement demand are set on the basis of two different approaches used to estimate floor areas in the building sector: “high” cement demand, which is sustained by the inertia of the trend of rapidly increasing building construction areas between 2004 and 2014, and “low” cement demand, which is sustained by the inertia of the steady phase of building construction areas after 2014 (Section 5.1 and Supplementary Results 1). The cement quality is classified into three levels: high-quality cement (compressive and flexural strength: 52.5 and 62.5, respectively), regular-quality cement (compressive and flexural strength: 42.5), and low-quality cement (compressive and flexural strength: 32.5). A higher cement quality generally means a higher C2Cr and more significant environmental impacts due to higher energy consumption and corresponding emissions and a higher price. We estimate the historical consumption of cement of different qualities in different construction types at the provincial and national levels. In 2018, the Chinese government proposed a phase-out of low-quality Portland composite (P.C.) 32.5 cement, which accounted for ~50% of China’s total cement production. We assume three different C2C ratio scenarios in this study considering the response of cement consumers to this cement quality improvement policy. The first scenario is the low-C2C ratio scenario (“BL” and “CL” in Table 1), in which low-quality P.C.32.5 cement shifts to other types of low-quality 32.5 cement, such as ordinary Portland (P.O.) 32.5 and fly ash Portland (P.F.) 32.5. The future share of high-, medium-, and low-quality cement production remains unchanged from the base year, 2020, in this scenario. The second scenario is the medium-C2C ratio scenario (“BM” and “CM” in Table 1), in which 50% of the P.C.32.5 low-quality cement shifts to other types of low-quality 32.5 cement, and the remaining 50% is upgraded to regular-quality 42.5/42.5R cement. The third scenario is the high-C2C ratio scenario (“BH” and “CH” in Table 1), in which 100% of low-quality P.C.32.5 cement is upgraded to regular-quality 42.5/42.5R cement rather than 52.5 and 62.5 cement, in light of the higher production costs for 52.5 and 62.5 cement than for the medium- and low-quality alternatives (Supplementary Notes, Supplementary Methods, and Supplementary Table S2).

The CO2 emission factor of regional power grids (EFEs) in China is published by the Ministry of Ecology and Environment of China every year47. With respect to the electricity production structure under a deep decarbonization scenario in the China Energy and Electricity Outlook16, our models project that the proportion of thermal power will quickly decline and that the power sector will achieve net zero emissions, as there are no emissions from electricity production by renewable energy. The calculation method of the annual EFE during the research period was described in our previous study48. The carbon tax applied for CO2 emissions and the environmental tax applied for air pollutants refer to the carbon trading price and the current environmental taxation system, respectively. More detailed information on the parameter settings in the different scenarios is provided in Supplementary Notes and Supplementary Methods. Parameters such as energy prices, raw material prices, and technology stocks are collected at the national and provincial levels for multidimensional analysis. Finally, the corresponding energy-saving effects and emission reductions of CO2, air pollutants, and mercury in the cement industry driven by the penetration of advanced production technologies under different clinker production scenarios can be estimated.