An inventory of primary air pollutants and CO2 emissions from cement production in China, 1990–2020

Direct emissions of air pollutants from the cement industry in China were estimated by developing a technology-based methodology using information on the proportion of cement produced from different types of kilns and the emission standards for the Chinese cement industry. Historical emissions of sulfur dioxide (SO 2 ), nitrogen oxides (NO X ), carbon monoxide (CO), particulate matter (PM) and carbon dioxide (CO 2 ) were estimated for the years 1990 e 2008, and future emissions were projected up to 2020 based on current energy-related and emission control policies. Compared with the historical high (4.36 Tg of PM 2.5 , 7.16 Tg of PM 10 and 10.44 Tg of TSP in 1997), PM emissions are predicted to drop substantially by 2020, despite the expected tripling of cement production. Certain other air pollutant emissions, such as CO and SO 2 , are also predicted to decrease with the progressive closure of shaft kilns. NO X emissions, however, could increase because of the promotion of precalciner kilns and the rapid increase of cement production. CO 2 emissions from the cement industry account for approximately one eighth of China ’ s national CO 2 emissions. Our analysis indicates that it is possible to reduce CO 2 emissions from this industry by approximately 12.8% if advanced energy-related technologies are implemented. These technologies will bring co-bene ﬁ ts in reducing other air pollutants as well. (cid:1) 2010 Elsevier Ltd. All rights reserved.


Introduction
China is the largest cement producing and consuming country in the world. Cement production in China was 1.39 billion metric tons in 2008 (CMIIT, 2009), which accounted for 50% of the world's production (USGS, 2009). Enormous quantities of air pollutants are emitted from cement production, including SO 2 , NO X , CO, and PM, and result in significant regional and global environmental problems. In China, the cement industry has been identified as an important source of pollution. For example, it is the largest source of PM emissions, accounting for 40% of PM emissions from all industrial sources (CEYEC, 2001) and 27% of total national PM emissions (Zhang et al., 2007a). Cement production also releases large amounts of CO 2 from both fuel combustion and the chemical process producing clinker, where calcium carbonate (CaCO 3 ) is calcined and reacted with silica-bearing minerals. According to the National Greenhouse Gas Inventory of China (NDRC, 2004), cement production contributed 57% of CO 2 process emissions (distinct from combustion emissions) from China's industrial sources in 1994.
There are two main kiln types in China: shaft kilns and rotary kilns. With higher productivity and efficiency, rotary kilns have dominated the cement industry in Western countries since the middle of the 20th century. Starting in the 1980s in China, however, small but easy-to-construct shaft kilns were built all over the country to meet the rapidly increasing demands of the construction industry. By the mid-1990s, they accounted for 80% of production (Lei, 2004). The extremely rapid increase in the number of shaft kilns resulted in poor operating practices within the Chinese cement industry. There were more than 7000 cement plants in China in 1997(Zhou, 2003, most of them small and releasing high emissions. At the end of the 1990s, China began to restrict construction of new shaft kilns and instead promoted precalciner kilns, which are the most advanced rotary cement kilns. Consequently, the production from precalciner kilns increased very rapidly and by 2008 they accounted for more than 60% of cement production (CMIIT, 2009).
Since China's cement industry is an important emission source of several types of air pollutants, the systematic and reliable estimation of its emissions is essential for atmospheric modeling and air pollution policy-making. From the perspective of criteria air pollutants, existing emission inventories for China usually treat the cement industry as a part of the industrial sector, roughly estimating its emissions based on coal consumption (Streets et al., 2003;Ohara et al., 2007). These emission inventories, however, are not capable of providing to the atmospheric modeling community reliable emission trends of China's cement industry. Moreover, there are shortcomings in future emission estimates because the effects from technology replacement and emission control measures were not taken into account. From the perspective of greenhouse gas (GHG) emissions, there have been some estimates made at a national level (He and Yuan, 2005;Liu et al., 2009;NDRC, 2004;Zhu, 2000) or as a part of a global analysis (Boden et al., 1995(Boden et al., , 2009WBCSD, 2002;Worrell et al., 2001). Most of these studies, however, have focused on a specific year and are not able to reflect changes in emissions due to technology replacement and energy efficiency improvement in China's cement industry. Our previous studies have addressed concerns over the technology-based emission estimates from the cement industry for specific air pollutants, such as PM (Lei et al., 2008) and NO X (Zhang et al., 2007b) or for a specific year (Zhang et al., 2009), however, a historical trend of emissions from China's cement industry is still missing. In this work, we developed a historical emission inventory of major air pollutants from China's cement industry for the period 1990e2008 to explore the effects of recent regulations and technologies on these emissions, and we predict future emissions up to 2020 in light of existing and possible future regulations.

Bottom-up methodology
The emission inventory developed here includes four gaseous air pollutants (SO 2 , NO X , CO and CO 2 ) and PM in three different size ranges: PM 2.5 (particulates with diameter less than 2.5 mm), PM 10 (particulates with diameter less than 10 mm) and TSP (total suspended particulates). Only direct emissions from cement production were considered in this inventory. The indirect emissions, such as which from power consumption during manufacture and the transportation of raw materials and end products are not included. Kilns are the major source of most air pollutants in the cement production process. In this study, cement kilns are classified into three groups: shaft kilns, precalciner kilns, and other rotary kilns.
Emissions from the cement industry of each province in China were calculated based on province-level cement outputs and emission factors (EFs), and then summed to give estimates at a national level. The approaches used for estimating emissions of different pollutants are listed in Table 1. The burning of fuel in the kilns is the sole source of SO 2 , NO X and CO emissions. The emissions of these pollutants were estimated based on coal consumption as almost all cement kilns in China use coal as fuel. In addition to fuel combustion, calcination of carbonate such as limestone is the other important source of CO 2 emissions, which we calculated separately in our analysis. PM emissions were more complicated to estimate because: (1) besides kiln emissions, there are several other emission sources such as quarrying and crushing, raw material storage, grinding and blending, and packaging and loading; and (2) abatement efficiency varies a lot between the different PM emission control technologies. Taking the multiple sources and control technologies into account, a dynamic methodology was developed to estimate the inter-annual emissions of PM.

Activity rates
Total cement production by province from 1990 is available from the China Statistical Yearbook (NBS, 1991e2008). A breakdown of national cement production by kiln type was estimated from the historical capacity of precalciner kilns and other rotary kilns (Kong, 2005;Lei, 2004;Zeng, 2004), as shown in Fig. 1. There are no statistical data for clinker production or coal consumption for the cement industry in China as a whole. We therefore estimated these by using typical clinker to cement ratios and energy efficiency data of the Chinese cement industry.
In general, cement is produced by mixing auxiliary materials with milled clinker. In China in the 1990s, the national average clinker to cement ratio varied within the range 0.701e0.738 (Zhou, 2003) and so for this study we used a value of 0.72. The energy efficiency of China's cement industry has improved considerably in last two decades. The average energy intensity of the whole industry dropped from 5.27 MJ tonne À1 -clinker in 1990 (Liu et al., 1995) to 4.77 MJ tonne À1 -clinker in 1998 (Zhou, 2003). And the current energy efficiency of China's precalciner kilns is around 3.51 MJ tonne À1 -clinker, a value that is recommended as the basic level for clean production of cement (MEP, 2009). Based on this information, the historical energy efficiency of different kiln types was interpolated, which enabled coal consumption to be calculated, as shown in Table 2.

Emission factors (EFs)
SO 2 mainly comes from the oxidation of sulfur in coal. In precalciner kilns, approximately 70% of SO 2 is absorbed by reaction with calcium oxide (CaO) (Liu, 2006), while much less is absorbed in other rotary kilns and in shaft kilns (Su et al., 1998). Utilization of ef j;m;y ¼ EFRm Â Fm;y Â P n C m;n;j Â ð1 À h n;y Þ a i, j, m and n represent the province, year, type of kiln or emission source and type of PM control technology respectively; y represents the three size categories of PM: PM 2.5 , PM > 2.5 mm but < 10 mm (PM 2.5e10 ), and PM > 10 mm (PM >10 ); E is the total emissions; CC is the coal consumption; CP denotes the clinker production; P is the cement production; EF is the emission factors from coal combustion; EFC denotes the emission factor of CO 2 from calcination during clinker production; ef is the final emission factor of PM; EFR is the unabated emission factor of PM prior to the effect of PM control devices; F y is the fraction of all PM in size range y with P y Fy ¼ 1; C n is the penetration rate of the PM control technology n with P n Cn ¼ 1; h is the PM removal efficiency of the control technology.
baghouse filters, as required with new precalciner kilns, can further reduce SO 2 emissions. Assuming that SO 2 absorption is 80% for the entire precalciner kiln process and 30% for other types of kilns (Liu, 2006), we estimated SO 2 emissions by province using a mass balance approach and based on the average sulfur content of coal in each province (Liu, 1998). Generation of NO X and CO is highly dependent on temperature and oxygen availability. Compared to shaft kilns, rotary kilns produce much more NO X and less CO because of their higher operation temperature and stable ventilation. Based on local test results, our previous studies have estimated the EFs of NO X (Zhang et al., 2007b) and CO (Streets et al., 2006) from different types of kilns in China. These EFs were used in this study, and the average NO X and CO EFs of the cement industry as a whole were calculated based on the historical balance of kiln type, as listed in Table 3.
CO 2 EFs were given in quite a few studies estimating CO 2 emissions from the cement industry in China (Boden et al., 1995;Cui and Liu, 2008;He and Yuan, 2005;NDRC, 2004;Wang, 2009;Worrell et al., 2001;Zhu, 2000). In this work, we used CO 2 EFs from the study by Cui and Liu (2008) who followed the approach recommended by the International Panel on Climate Change (IPCC, 2006) and calculated the EFs based on the typical practices of the cement industry in China. Their estimates of the CO 2 EFs were 0.55 kg kg À1 clinker from the calcining process, and 1.94 kg kg À1 coal from coal combustion. The CO 2 EFs from different published works are compared in Sect. 3.3 to assess the reliability of our CO 2 emission estimates.
The EFs of PM is dependent on both the characteristics of unabated emissions from the overall production process and the effectiveness of PM emission control devices. Our previous study (Lei et al., 2008) reviewed prior research and calculated the unabated PM EFs for different types of kilns and other emission sources, such as quarrying, crushing and other mechanic processes, as summarized in Table 4. PM control devices can reduce PM emissions by 10e99.9%, depending on the type of control technology employed and the size distribution of PM in the raw flue gas (Lei et al., 2008). Although more efficient PM control technologies require higher investment and have higher operational costs, improving emission standards is driving the promotion of these technologies within the industry. The standard value for the PM concentration in kiln flue gas dropped from 800 to 50 mg m À3 in 20 years, according to progressive editions of the air pollutant emission standards for the cement industry (SEPA, 1985(SEPA, , 1996b(SEPA, , 2004. Based on the PM concentration requirements of the three successive emission standards, penetration rates of the different PM control technologies in newly built cement plants are estimated for four periods: before 1985, 1985e1996, 1997e2004, and after 2004. Although the emission standards published in 1996 and 2004 allow 3e10 years for existing plants to reduce their PM emission rates, reduction is not likely to be significant in existing plants as there are few measures to enforce the standard. In this work, we assume that all plants retrofit their whole production line every 15 years, and in doing so meet the present standards for new plants. We then calculate the penetration rates of PM control technologies across the cement industry for the period 1990e2008, and estimate the corresponding PM EFs, as shown in Fig. 2. Over the 18-year period, the EF of TSP from the cement industry dropped by 88%, from 27.93 g kg À1 to 3.31 g kg À1 . The reduction in PM 10 and PM 2.5 EFs is also considerable: 18.08e2.53 g kg À1 and 10.65e1.61 g kg À1 , respectively.

Results and discussion
3.1. Emissions from 1990 to 2008  Table 5 show emissions of gaseous air pollutants and PM from China's cement industry for the period 1990e2008. The emissions in 2005 are also compared with China's total emissions from all anthropogenic sources in Table 5. The cement industry is a major source of PM in China, contributing more than a quarter of PM 2.5 and PM 10 in 2005. As a significant contributor to GHG emissions in China, the cement industry produces approximately one eighth of China's total anthropogenic CO 2 emissions. The    of 12.6%). The decline in SO 2 emissions in 2008 is attributed to two factors. First, the global economic recession suppressed the construction industry and saw the annual rate of increase in cement production drop from 10% in the previous year to 2% in 2008. Second, nationwide replacement of shaft kilns with precalciner kilns from 2007 to 2008 led to a 20% of reduction in cement production from shaft kilns, which emit several times more SO 2 per mass unit of cement.
The trend observed for CO emissions is similar to that of SO 2 . In contrast, NO X emissions increased much faster than any other pollutant. During the 1990s, NO X emissions doubled, and the 2000 emissions were three times higher by 2008. With the recent rapid expansion of precalciner kilns in China, the average annual increase in NO X emissions from the cement industry from 2003 to 2008 was over 220 Gg. This accounts for about 20% of the incremental NO X emissions seen for China as a whole according to the INTEX-B emission inventory (Zhang et al., 2009). As awareness grows of China's increasing NO X emissions and its consequences for ozone pollution and acidification (Zhao et al., 2009), policies to combat acid rain pollution will inevitably have to specifically address the cement industry.
Emissions of CO 2 increased 5.8 times, from 153 Tg in 1990 to 892 Tg in 2008. The proportion of CO 2 emissions from fuel combustion compared to that from calcination of carbonates decreased from 46.0% in 1990 to 38.6% in 2008, representing improved energy efficiency in the cement industry. Our estimates of CO 2 emissions are lower than the results delivered by some researchers such as Boden et al. (2009), Liu et al. (2009) and Worrell et al. (2001. The main reasons of the differences are discussed in Section 3.3. Emissions of PM rose rapidly from 1990 to 1995, when cement production developed with an average annual increase of 17.8%. In the second half of the 1990s, expansion of China's cement industry slowed and the new emissions standard released in 1996 promoted the application of electrostatic precipitators (ESPs) in shaft kilns, resulting in an industry-wide decrease in PM emissions. After 2000, although the average annual increase in cement production was greater than 12%, PM emissions gradually decreased due to the replacement of shaft kilns by precalciner kilns and the application of high-performance PM removal technology, especially after 2004. Over the whole period, PM emissions reached a peak in 1997, with 4.36 Tg of PM 2.5 , 7.16 Tg of PM 10 and 10.44 Tg of TSP.

Spatial distribution of emissions
The spatial distribution of emissions changed year-by-year. Using 2005 as a base year, cement production from 5294 plants was collected from the China Cement Association (CCA), including the capacity of 612 clinker production lines installed with precalciner kilns. These plants accounted for almost all precalciner kilns and more than 95% of cement production in China in that year. The location of these plants and production lines is determined at county-level from cement plant registration information (CCA, 2006). Thus emissions of PM 2.5 , SO 2 and NO X from the cement industry in 2005 are mapped onto an 18 0 Â 18 0 grid of China, as shown in Fig. 4.
In different ways, the distribution of PM 2.5 , SO 2 and NO X emissions reflect regional operational differences and kiln combinations within China, a point illustrated by the following examples. The grid cells indicating high PM 2.5 emissions show a greater   concentration in north China. The provinces of Shandong, Hebei and Henan account for 31.5% of total PM 2.5 emissions. By consuming coal with much higher sulfur content than other provinces, Sichuan has the second highest provincial emissions of SO 2 , after Shandong. Shandong, Zhejiang, Jiangsu and Anhui are the largest contributors to NO X emissions, which is due to a number of clinker-producing centers using large precalciner kilns in these provinces. The highest PM 2.5 , SO 2 and NO X emissions are all located in the same grid cell in Shandong province, where the city of Zaozhuang is found.

Comparison of CO 2 emissions with other studies
Some studies have been conducted to quantify CO 2 emissions from cement production in China, but few of them have observed the effects of developments in technology on CO 2 emissions in China. We compare our estimates with some results from other studies in Table 6. All CO 2 emission estimates were converted to CO 2 EFs in the comparison.
Generally speaking, estimates of China's CO 2 emissions as a part of global studies (e.g., WBCSD, 2002;Boden et al., 2009) are much higher than estimates made from domestic studies because some parameters used in global studies doesn't fit the real situation of Chinese cement industry. For example, a higher clinker to cement ratio was used in global studies (83e100%) than in domestic ones (72e75%). Higher energy intensity was assumed in global studies as well, which led to higher EFs of CO 2 from energy consumption. The other factor leads to higher results in global studies is that indirect CO 2 emissions from electricity consumption are usually included in those analyses. The national average electricity intensity of China's cement industry is 110e115 kwh/tonne cement (Liu et al., 1995;MEP, 2009). Therefore electricity consumption during cement production leads to additional indirect CO 2 emissions of 0.102e0.107 kg CO 2 /kg cement.
Our estimates of CO 2 emissions are generally comparable with most domestic studies. However, the recent studies by Cui and Liu (2008) and Wang (2009) indicate much lower values of CO 2 emission than our study, as their estimations were based on the working practices of advanced precalciner cement plants. These lower EFs indicate that China's cement industry shows promising potential to reduce CO 2 emissions.

Future emissions and mitigation potential
Since emissions from China's national cement industry contribute significant levels of several air pollutants, accurate emission projections are necessary to inform Chinese national strategies on air pollution control and GHG mitigation. In this study, the future emissions from the cement industry for the period 2010e2020 were estimated, and the potential of mitigation technologies to reduce the emission is analyzed.

Emissions projection
Cement production in China exceeded 1.63 billion metric tons in 2009, and the available statistical data show another 19% increase in the first 5 months of 2010 (http://www.stats.gov.cn/tjsj/), in comparison with the same period of 2009. Expert opinion from the CCA indicates that cement production may reach approximately 1.8 billion tons in 2010 (personal communication), but projecting as far ahead as 2020 reveals large differences between the available predictions: the Chinese Academy for Environmental Planning predicted production to be 2.1 billion metric tons based on future investment in fixed assets; Ho and Jorgenson (2007) modeled China's economy with a computable general equilibrium (CGE) economic model that included 33 production sectors and one household sector, and predicted production to be 1.7 billion metric tons in 2020; Wei and Yagita (2007) coupled cement production with future rates of urbanization and building area and predicted production to be 1.2 billion metric tons by the same date. Considering the large uncertainties involved in predicting the development of China's cement industry over the next 10 years, our emission projections are based on these three studies' predictions of cement production by 2020, representing scenarios of high, medium and low production, respectively.
The key technological features of the cement industry are projected based on the existing policies on industry structure (NDRC, 2006), energy saving (MEP, 2009) and emission control (SEPA, 2004). Assuming no new policy will come into effect before 2020, future EFs of air pollutants from the cement industry were estimated using the same methodology described in Sect. 2, as listed in Table 7. Generally speaking, the continued replacement of shaft kilns by precalciner kilns will lead to a decrease in both coal intensity and in CO 2 EFs. Higher penetration of precalciner kilns will also result in a decrease in SO 2 and CO EFs, and an increase in NO X EFs. PM EFs are predicted to fall with the progressive construction of new cement plants which meet the requirement of the latest emission standards.
According to our estimates (Table 8), the emissions of SO 2 , CO and PM from China's cement industry will decrease in all of the three scenarios; however, emissions of NO X and CO 2 will increase until 2020 in the high-production scenario. Emissions of NO X will be considerable, compared with SO 2 and PM emissions. The ratio of NO X to SO 2 will increase from 2.07 in 2010 to 3.14 in 2020, indicating that greater focus should be given to NO X emission control in order to prevent pollution from acid rain. The differences in CO 2 emissions between 2010, 2015 and 2020 are less than those of other pollutants. This is because under current policies the EF of CO 2 will  Zhang et al., 2009. c Boden et al., 2009 Lei et al., 2010. not change as much as other pollutants. However, as CO 2 emissions mitigation is of serious concern, more policies should be considered to reduce CO 2 EFs from China's cement industry.

Potential for CO 2 mitigation
Several studies have addressed potential reductions in CO 2 emissions from China's cement industry. The World Business Council for Sustainable Development (WBCSD, 2009a) has summarized the recent studies and pinpointed four areas of technology for the reduction of CO 2 emissions: thermal and electric efficiency, alternative fuels, clinker substitution, and carbon capture and storage (CCS). Following the technology roadmap laid out by WBCSD (2009a), we analyzed the potential for CO 2 mitigation by China's cement industry in 2020, as listed in Table 9. Thermal efficiency could be improved by replacing small shaft kilns by large precalciner kilns. On a global level, top 10% advanced kilns can currently operate at a average thermal efficiency of 3.10 MJ kg À1 -clinker (MEP, 2009;WBCSD, 2009a), and although the clinker to cement ratio in China is already lower than the global average (WBCSD, 2009b;Worrell et al., 2001), the predicted increase of power plants and iron and steel plants will increase the availability of by-products which can be used as substitutes for clinker. It is estimated that the clinker to cement ratio could drop to 65% (CCA, personal communication). A few cement plants in China have begun to use solid waste as a fuel in kilns as a substitute for coal. WBCSD (2009a) projected that the share of alternative fuel in clinker fuel use in Asia could increase at an annual rate of 0.6%. CCS technologies are long-term approaches to carbon management and are not likely to be accessible by 2020; therefore we do not include them in our analysis.
Our analysis indicates that the potential for CO 2 mitigation by clinker substitution is likely to be much larger than that possible by improvements in thermal efficiency and from the use of alternative fuels. If these three technologies are implemented together, CO 2 emissions from cement production could be reduced by 12.8% by 2020. Using the mid-level scenario of predicted increase in cement production, mitigation of CO 2 emissions will be 130 Tg of CO 2 , equivalent to the emissions of CO 2 from the usage of 67 Tg of coal.

Potential for emission control of other air pollutants
Current emission standards have promoted the use of advanced emission control technologies; however, the level of emission control in China's cement industry is still lower than that of advanced countries. If state-of-the-art control technologies are used, there is potential to substantially further reduce the emission of air pollutants from the cement industry in China.
PM emissions have been the major focus of air pollution control from the cement industry for years. As of 2010, all new cement plants are required to meet an emission standard of 50 mg m À3 flue gas (SEPA, 2004), which equates to a PM EF for the whole production process of approximately 1 g kg À1 cement (CRAES, 2003). This emission level could be even lower, however, when baghouse filters replace PM control devices that are relatively inefficient. If all of China's cement plants used baghouse filters in major production processes, the average PM EF could drop to 0.7 g kg À1 cement.
Deployment of baghouse filters will benefit SO 2 emission control as well. However, emissions of NO X would not be reduced unless selective catalytic reduction (SCR) or selective non-catalytic reduction (SNCR) technology is used. Although BREF documents state that NO X emissions could be as low as 200e500 mg m À3 if the best available technologies (BAT) are used, actual NO X emissions from most European cement plants using SNCR technology are 500e800 mg m À3 (Jiao, 2007). If China's cement plants could reduce the average NO X emission level to 500 mg m À3 in 2020, the corresponding NO X EF would be 9.7 k kg À1 coal.
The technologies used to mitigate CO 2 emissions are also effective in reducing emissions of PM, SO 2 and NO X . Based on midscenario emissions of these pollutants in 2020, we estimated the potential for emission reduction from each mitigation technology, as listed in Table 10. Our estimates indicate that there is the potential for China's cement industry to reduce air pollutants substantially. Baghouse filters and SCR/SNCR technologies are likely to be the most effective in controlling PM and NO X emissions, respectively, and technologies to abate CO 2 emissions will also bring significant benefits in terms of SO 2 emission control.

Conclusions
The cement industry plays an important role in emissions of many air pollutants in China. This study estimates the direct emissions of major air pollutants from cement production based on information on the development of production technologies and rising emission standards in China's cement industry. Our analysis shows that with the replacement of old shaft kilns by precalciner kilns, there is an opportunity to reduce PM emissions through the implementation of stricter emission standards and the promotion of high-performance PM control technologies. Other air pollutants such as CO and SO 2 will also decrease as shaft kilns are gradually retired. However, the promotion of precalciner kilns within China and a rapid increase in cement production has led to greatly increased NO X emissions. Future NO X emission could be reduced if SCR or SNCR technologies are introduced within the cement industry, although the cost of introduction is likely to be considerable.
Although energy-use efficiency in China's cement industry has improved significantly in recent years, the industry still contributes approximately one eighth of the nation's CO 2 emissions. Our analysis indicates that it may be possible to reduce CO 2 emissions by 12.8% by 2020 if advanced energy-related technologies are implemented, and that the substitution of clinker with other material is likely to be the most effective technology in this regard. These energy-related technologies are likely to bring additional benefits by reducing the emission other air pollutants as well.  Table 9 Estimated emission reduction potential from major CO 2 mitigation technologies in 2020.
Technology category Improvement of performance CO 2 reduction (%) Thermal efficiency Thermal intensity drops from 3.22 MJ kg À1 clinker to 3.10 MJ kg À1 clinker 1.3 Clinker substitution Clinker to cement ratio drops from 72% to 65% 9.7 Alternative fuels Share of alternative fuel increases by 6% 2.1