Scenario input–output analysis on the diffusion of fuel cell vehicles and alternative hydrogen supply systems

Journal of Economic Structures, Feb 2019

Ratifying the Paris Climate Change Agreement of 2015, which is the new framework for global environmental measures for change after 2020 onward, Japan is proposing to reduce its greenhouse gas emissions by 26% by 2030 from 2013 levels. To achieve this target, it is indispensable to transcend the current fossil-fuel-based technologies (petroleum, coal, and natural gas) and shift to renewable energy systems. Neutral fuels or fuels free of carbon dioxide emissions must become the predominant source of energy, in addition to introducing energy conservation technologies in manufacturing, transportation, business, and households. Amid these developments, fuel cell vehicles and hydrogen production technologies are gaining much attention. Our research group is developing a new hydrogen-generating system that directly decomposes hydrogen from methane and separates carbon as a solid substance with zero carbon dioxide emissions. We estimate the carbon dioxide reduction effect of our new hydrogen-generating system and compare it with the current steam reforming method by applying scenario input–output analysis. Our new system is expected to lower carbon dioxide emissions to 14.1% of the conventional system in the industrial sector. With the replacement effect of gasoline vehicles to fuel cell vehicles, carbon dioxide emissions are expected to reduce for both hydrogen production technologies. The new system is more efficient and saves carbon dioxide emissions by 21.7% more than the conventional system, under the assumption that 800 thousand fuel cell vehicles will be available in Japan before 2030.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://link.springer.com/content/pdf/10.1186%2Fs40008-019-0137-3.pdf

Scenario input–output analysis on the diffusion of fuel cell vehicles and alternative hydrogen supply systems

Journal of Economic Structures December 2019, 8:4 | Cite as Scenario input–output analysis on the diffusion of fuel cell vehicles and alternative hydrogen supply systems AuthorsAuthors and affiliations Mitsuo YamadaKiyoshi FujikawaYoshito Umeda Open Access Research First Online: 06 February 2019 75 Downloads Abstract Ratifying the Paris Climate Change Agreement of 2015, which is the new framework for global environmental measures for change after 2020 onward, Japan is proposing to reduce its greenhouse gas emissions by 26% by 2030 from 2013 levels. To achieve this target, it is indispensable to transcend the current fossil-fuel-based technologies (petroleum, coal, and natural gas) and shift to renewable energy systems. Neutral fuels or fuels free of carbon dioxide emissions must become the predominant source of energy, in addition to introducing energy conservation technologies in manufacturing, transportation, business, and households. Amid these developments, fuel cell vehicles and hydrogen production technologies are gaining much attention. Our research group is developing a new hydrogen-generating system that directly decomposes hydrogen from methane and separates carbon as a solid substance with zero carbon dioxide emissions. We estimate the carbon dioxide reduction effect of our new hydrogen-generating system and compare it with the current steam reforming method by applying scenario input–output analysis. Our new system is expected to lower carbon dioxide emissions to 14.1% of the conventional system in the industrial sector. With the replacement effect of gasoline vehicles to fuel cell vehicles, carbon dioxide emissions are expected to reduce for both hydrogen production technologies. The new system is more efficient and saves carbon dioxide emissions by 21.7% more than the conventional system, under the assumption that 800 thousand fuel cell vehicles will be available in Japan before 2030. KeywordsScenario input–output analysis Fuel cell vehicles Hydrogen production technology Carbon dioxide emissions Direct decomposition of methane Methane steam reforming  JEL ClassificationC67 L62 P18 Q55  1 Introduction Ratifying the Paris Agreement1 of 2015, which is a new framework for global environmental measures for change after 2020 onward, Japan is proposing to reduce its greenhouse gas emissions by 26% by 2030 from 2013 level. To achieve this target, it is into transcend the current fossil-fuel-based (petroleum, coal, and natural gas) technologies and shift to renewable energy systems. Neutral fuels or fuels free of carbon dioxide (CO2) emissions must become the predominant source of energy, in addition to introducing energy conservation technologies in each sector of manufacturing, transportation, business, and households. Amid these developments, fuel cell vehicles and hydrogen production technologies2 are gaining much attention. The engineering team of our research group is developing a new hydrogen-generating system that directly decomposes hydrogen from methane and separates carbon as a solid substance with zero CO2 emissions. We evaluate the potential economic and environmental effects of reducing CO2 on the broader economy when the new system is introduced. Our simulation analysis estimates the impact of reducing CO2 in our new hydrogen production system and compares it with the current steam reforming method by applying scenario input–output analysis. In implementing this simulation analysis, we assume a certain volume of diffusion of fuel cell vehicles in the future and use it as a reference case for the simulation analysis. Broadly, several extended input–output models have been applied in the energy and environmental fields. According to Miller and Blair (2009), since the late 1960s, researchers theorized that the input–output framework could be extended to account for environmental pollution generation and abatement processes associated with inter-industry activity. Leontief (1970) provided a key methodological extension that has since been applied widely and extended further. In relation to CO2 emissions, there is accumulation of carbon footprints as Nansai et al. (2009), Wiebe et al. (2012), and Usubiaga and Acosta-Fernandez (2015). Cantono et al. (2008) present an assessment of the benefits of public transportation using hydrogen and fuel cell buses using environmental input–output analysis. The authors show that the process of producing hydrogen by steam reforming of methane does not reduce CO2 emissions completely, even though fuel cell buses do not emit CO2 during operation. They suggested the use of hydrogen in fuel cell buses as it is environmentally desirable, especially if accompanied by renewable sources, CO2 capture, or both. In contrast, using a framework of life cycle assessment (LCA), Bohnes et al. (2017) evaluated the environmental impact of passenger car fleet development in the City of Copenhagen for the years 2016–2030 and showed the relative environmental benefits from range-extended electric vehicles and fuel cell vehicles (FCVs) over conventional vehicles and battery electric vehicles (BEVs). Miotti et al. (2017) conducted a detailed LCA of the environmental impact of FCVs and other vehicles. They concluded that if fuel is sourced from renewable energy sources, as is the case of BEVs, FCVs have the advantage of lower greenhouse gas emissions over conventional vehicles. There are several methods to produce hydrogen.3 Dincer and Acar (2015) examined various potential methods of producing hydrogen using renewable and non-renewable sources, and compared the environmental impact, cost, and energy efficiency. Photonic energy-based hydrogen production is environmentally more beneficial compared to other methods in terms of emissions, although costs and efficiency are not attractive. On the other hand, fossil fuel reforming and biomass gasification produce cheaper hydrogen efficiently. They concluded that hybrid-energy-based hydrogen production methods, in which the energy sources are electrothermal, photo-biochemical, and electro-photonic, have higher rankings on average. Keipi et al. (2018) compared the costs of producing hydrogen using thermal decomposition of methane to steam reforming and water electrolysis in the current and potential future market environments. They estimated costs from engineering-based information and not from input–output tables. They found that thermal decomposition of methane is suitable for on-site demand-driven hydrogen production in small- and medium-scale operations and economically competitive with steam reforming. Thermal decomposition of methane has the advantage of feedstock availability via the current natural gas infrastructure, whereas electrolysis is highly dependent on the cost and availability of renewable electricity. Our study is an environmental input–output analysis, focusing on the economic effects of new hydrogen production technologies. As reference, we consider hydrogen demand from fuel cell passenger cars bought by consumers and not fuel cell buses for public passenger transport. Further, we compare the new hydrogen production technology from methane decomposition to steam reforming. Electrolysis and other production methods are beyond the scope of this research because they have little advantage in production costs compared to reforming fossil fuels using current technologies. Our input–output model comprises a wide range of technologies to produce hydrogen for a single commodity of hydrogen so that the row number differs from the column number of the input coefficient matrix, which requires certain arrangement to obtain the Leontief inverse matrix. Here, we introduce a weighted average of the plural technologies. The weights, which are given exogenously as a scenario, indicate the choice of technology. Such studies appear in Ikeda et al. (1996), Yoshioka and Suga (1997), Wang (2016), and Fujikawa and Wang (2017). In some studies, we find the same characteristic of input–output models based on a rectangular matrix in which the row–column sizes differ. Nakamura and Kondo (2002a, b, 2009) and Kondo and Nakamura (2004) developed waste input–output model that was extended to the conventional input–output model by including waste generation sectors in the rows and waste treatment sectors in the columns. Since the number of waste generation sectors (rows) is larger than waste treatment sectors (columns) in the waste industry input–output table, a suitable method was proposed and implemented to obtain a squared input coefficient matrix for calculating the Leontief inverse matrix. Klein (1983, 2003) proposed a flow-of-funds model that is similar to Leontief’s input–output model. The flow-of-funds model describes that each economic agent owns several financial assets and liabilities. Tsujimura and Mizoshita (2003) and Nishiyama (2008) extended the scope of Klein’s flow-of-funds model by developing new approaches that convert a rectangular table to a square table for analysis. In engineering studies, we found another type of application for the rectangular input–output model. Tsunoka et al. (2011, 2012) investigated environmental burdens associated with a complex production system with some feedback flows. They described process activities with material input–output in the system as a rectangular matrix, in which one technique was presented to obtain the matrix inversion. Fukuhara and Hondo (2011) proposed a generalized method to describe a production system as a geometrical figure and construct a regular coefficient matrix using graph theory. This paper is organized as follows. In the next section, we describe the study background. Section 3 outlines the input–output scenario analysis model, and Sect. 4 discusses the assumptions and analytical results of the scenario input–output analysis. The analysis results are summarized in Sect. 5. 2 Background research 2.1 Changes in carbon dioxide emissions Table 1 shows the trends in Japan’s recent greenhouse gas (GHG) and sectoral CO2 emissions. In FY2016 (preliminary figures), GHG emissions were 1322 million tons of CO2 equivalent, down 4.6% from FY2005 and 6.2% from FY2013, with CO2 emissions accounting for 1.222 million tons, or 92.4% of overall GHG emissions. Of the CO2 emissions, 93.4%, or 1.144 million tons, originated from energy sources, while the remaining 78 million tons had non-energy origins. The industrial sector (energy origin) accounts for 34.2% of total CO2 emissions, while business and other sectors 17.9%, transport sector 17.6%, household sector 14.6%, and energy conversion 9.2%. Table 1 Greenhouse gas emissions and sectoral CO2 emissions (Unit: Million t-CO2 equivalent) Source: Ministry of Environment (2017) Fiscal year 2005 2013 2016 (P) Growth rate (%, 2016/2005) Growth rate (%, 2016/2013) GHG 1386 1409 1322 − 4.6 − 6.2  Carbon dioxide (CO2) 1297 1316 1222 − 5.8 − 7.1  Methane and others 89 93 100 12.4 7.5 Carbon dioxide, total 1297 1316 1222 − 5.8 − 7.1 Energy origin 1206 1235 1144 − 5.1 − 7.4  Industrial sector 468 463 418 − 10.7 − 9.7  Transportation sector 245 224 215 − 12.2 − 4.0  Business and other sectors 217 244 219 0.9 − 10.2  Household sector 175 205 179 2.3 − 12.7  Energy conversion sector 100 100 113 13.0 13.0 Non-energy origin 91.8 80.9 78 − 15.0 − 3.6    Share (%)     GHG 100.0% 100.0% 100.0%     Carbon dioxide (CO2) 93.6% 93.4% 92.4%     Methane and others 6.4% 6.6% 7.6%    Carbon dioxide, total 100.0% 100.0% 100.0%    Energy origin 93.0% 93.8% 93.6%     Industrial sector 36.1% 35.2% 34.2%     Transportation sector 18.9% 17.0% 17.6%     Business and other sectors 16.7% 18.5% 17.9%     Household sector 13.5% 15.6% 14.6%     Energy conversion sector 7.7% 7.6% 9.2%    Non-energy origin 7.1% 6.1% 6.4%    2.2 Diffusion of next-generation vehicles Table 2 shows the number of vehicles owned and unit sales of next-generation vehicles sold in Japan after FY2011. Number of vehicles owned in FY2016 was 7.133 million units, accounting for 9.19% of total ownership of all vehicles. In addition, unit sales stood at 1.366 million, accounting for 26.91% of total sales of all vehicles. Next-generation vehicles include electric vehicles (EVs), plug-in hybrid vehicles (PHVs), fuel cell vehicles (FCVs), and hybrid vehicles (HVs). Table 2 Number of next-generation vehicles and number of units sold. Source: Japan Next Generation Automobile Promotion Center (2018), Japan Automobile Manufacturers’ Association (2018), and Automobile Inspection & Registration Information Association (2018) Fiscal year 2011 2015 2016 Number of units owned (units)  EV 22,262 80,511 89,844  PHV 4132 57,130 70,323  FCV   630 1807  HV 2,029,009 5,764,401 6,971,035  Next-generation vehicles, total 2,055,403 5,902,672 7,133,009  All vehicles (except motorcycle) 75,609,883 77,301,798 77,657,517  Share of next-generation vehicles (%) 2.72 7.64 9.19 Units sold (units)  EV 13,256 14,733 13,817  PHV 3753 14,997 13,847  FCV   494 1204  HV 633,708 1,146,164 1,337,497  Next-generation vehicles, total 652,799 1,176,388 1,366,365  All vehicles (except motorcycle) 4,753,273 4,937,734 5,077,903  Share of next-generation vehicles (%) 13.73 23.82 26.91 Hybrid vehicles are overwhelmingly large, both in number of vehicles owned and in sales volume. Nearly 1.337 million units of hybrid vehicles were sold in FY2016, while number of vehicles owned stood at 6.971 million units at the end of FY2016. Sales units of electric vehicles and plug-in hybrid vehicles were 13,800 units each, number of electric vehicles owned was 89,800 units, and number of plug-in hybrid vehicles stood at 70,323 units. Sales of FCVs started in 2014 and 1807 units were sold by the end of FY2016, and these vehicles are gaining popularity. 2.3 Basic structure of next-generation vehicles Table 3 shows the basic structure of an FCV compared to hybrid and electric vehicles. A hybrid vehicle adds a motor and an auxiliary battery to an engine-driven conventional vehicle to increase energy efficiency by assisting the engine power. Conversely, an electric vehicle has a simple structure that does not have an engine/fuel tank, charges a large-capacity battery, and is driven by a motor, but needs to seek electric power from other sources. Fuel cell vehicles are driven by motors as well as electric vehicles, but power is generated by the process of reacting hydrogen atoms with oxygen atoms using a fuel cell stack. Therefore, the battery does not need to have a large capacity; however, in addition to the FC stack, a hydrogen tank is required. For these next-generation vehicles, expensive parts such as high-performance motors, high-power density batteries, FC stacks, and hydrogen tanks are required, whereas major automotive parts such as engines, fuel tanks, and transmissions are no longer needed. Table 3 Basic structure of next-generation vehicles. Source: Influence of Next Generation Mobility Promotion on the Chubu Region Industries, CRISER (2015)   Engine fuel tank Motor Auxiliary battery FC stack Hydrogen tank Large-capacity battery Conventional vehicle ○       Hybrid vehicle (HEV) ○ ○ ○     Fuel cell vehicle (FCV)   ○ ○ ○ ○   Electric Vehicle (EV)   ○     ○ Figure 1 represents well-to-wheels-based CO2 emissions for each vehicle. For gasoline vehicles, it is 147 g-CO2/km, diesel vehicles are slightly lower at 132 g-CO2/km, and hybrid vehicles are at 95 g-CO2/km. Gasoline refueling emissions from PHVs are almost equal to that of a hybrid vehicle, 102 g-CO2/km, and drop to 55 g-CO2/km when charging. In electric vehicles, refueling emissions depend on the mix of power sources, which was 55 g-CO2/km in 2009 and 77 g-CO2/km in 2012 when nuclear power plants were shut down due to the Great East Japan Earthquake in March of 2011. In contrast, when electricity generated from photovoltaic power is used almost no CO2 is generated as 1 g-CO2/km. Open image in new window Fig. 1 Comparison of CO2 emissions (well-to-wheels JC08 mode). Source: “Report on overall efficiency and GHG emissions by type of vehicles,” Japan Automobile Research Institute, March, 2011. “The roadmap of the hydrogen and fuel cell strategy, Revised,” Hydrogen and fuel cell council, METI, March, 2016. Note The CO2 emission of EV (the power source mix in FY2012) is estimated suitable to the power source mix in FY2012 by the Hydrogen and fuel cell council, METI Further, FCVs depend on hydrogen production technologies. It is 79 g-CO2/km when hydrogen is used by on-site reforming of gas, and 78 g-CO2/km for off-site reforming of natural gas. These amounts are not different from that of EVs depending on a mix of power sources in 2012. Gas-reforming technologies are currently established to produce hydrogen. In on-site alkaline water electrolysis with solar power, it is considerably lower at 14 g-CO2/km. The chemical formula for steam reforming of methane, primary component of city gas and natural gas in Japan, for producing hydrogen is given by: $$ {\text{CH}}_{4} + 2{\text{H}}_{2} {\text{O}} \to {\text{CO}}_{2} + 4{\text{H}}_{2} $$ (1) It generates CO2 in the hydrogen production process. If hydrogen is produced without generating CO2 from the same gas, CO2 emissions from FCVs can be significantly reduced. This is possible through an alternative hydrogen production technology, namely direct decomposition. Similarly, when using methane, the chemical formula is given by: $$ {\text{CH}}_{4} \to 2{\text{H}}_{2} + {\text{C}} $$ (2) In this case, instead of CO2, solid carbon (C) is generated. Once this technology is established, there is a possibility that FCVs will approach the same amount of CO2 emissions as on-site alkaline water electrolysis with solar power. 3 Method: Scenario input–output analysis model When there is more than one activity (production technologies) for one product in the input–output analysis, there is difficulty in handling technology selection among the several activities. One approach is solving the equation with additional constraints on the input coefficient matrix.4 The electric power generation sector is a typical example, but is just one product. However, in Japan’s input–output table, there are three activities: (a) nuclear power, (b) fire power, and (c) hydro power and other activities. Figure 2 shows the input–output table. Open image in new window Fig. 2 Input–output table with non-square input transaction matrix Source: Illustrated by Authors If we change the composition of these activities, the environmental load and the economic effect also change. In our study, the technology for hydrogen production consists of two methods: conventional and direct decomposition of methane. If the input structure and energy utilization structure differ for each hydrogen production technology, the environmental load and economic effect will also change by altering the composition. This input–output model is expressed as follows. $$ \begin{aligned} {\mathbf{A}}_{{{\mathbf{11}}}} {\mathbf{x}}_{{\mathbf{1}}} + {\mathbf{A}}_{{{\mathbf{12}}}} {\mathbf{z}} + {\mathbf{f}}_{{\mathbf{1}}} = {\mathbf{x}}_{{\mathbf{1}}} \hfill \\ {\mathbf{A}}_{{{\mathbf{21}}}} {\mathbf{x}}_{{\mathbf{1}}} + {\mathbf{A}}_{{{\mathbf{22}}}} {\mathbf{z}} + {\mathbf{f}}_{{\mathbf{2}}} = {\mathbf{x}}_{{\mathbf{2}}} \hfill \\ \end{aligned} $$ (3) Here, \( {\mathbf{x}}_{{\mathbf{i}}} \) is the product vector, \( {\mathbf{f}}_{{\mathbf{i}}} \) is the final demand vector, and \( {\mathbf{A}}_{{{\mathbf{ij}}}} \) is the input coefficient matrix. Suffix 1 denotes the usual sectors, and suffix 2 shows a sector with plural activities of hydrogen. Since there are plural activities, \( {\mathbf{z}} \), and one product, \( {\mathbf{x}}_{{\mathbf{2}}} \), the input coefficient matrix,\( {\mathbf{A}}_{{{\mathbf{22}}}} \), does not become a square, and it is difficult to obtain the conventional Leontief inverse matrix. Therefore, the following scenario (restriction) is added. $$ \begin{aligned} & \left[ {\begin{array}{*{20}c} {z_{1} } \\ {z_{2} } \\ \end{array} } \right] = \left[ {\begin{array}{*{20}c} {\alpha_{1} } \\ {1 - \alpha_{1} } \\ \end{array} } \right]{\mathbf{x}}_{{\mathbf{2}}} \\ & {\mathbf{z}} = {\mathbf{c}}\,{\mathbf{x}}_{2} \\ \end{aligned} $$ (4) Here, vector “c” on the right-hand side represents the scenario. This ratio will be given exogenously. Typically, if vector “c” changes, the required production volume also changes. Substituting Eq. (4) into Eq. (3), we obtain $$ \left[ {\begin{array}{*{20}c} {{\mathbf{A}}_{{{\mathbf{11}}}} } & {{\mathbf{A}}_{{{\mathbf{12}}}} {\mathbf{c}}} \\ {{\mathbf{A}}_{{{\mathbf{21}}}} } & {{\mathbf{A}}_{{{\mathbf{22}}}} {\mathbf{c}}} \\ \end{array} } \right]\left[ {\begin{array}{*{20}c} {{\mathbf{x}}_{{\mathbf{1}}} } \\ {{\mathbf{x}}_{{\mathbf{2}}} } \\ \end{array} } \right] + \left[ {\begin{array}{*{20}c} {{\mathbf{f}}_{{\mathbf{1}}} } \\ {{\mathbf{f}}_{{\mathbf{2}}} } \\ \end{array} } \right] = \left[ {\begin{array}{*{20}c} {{\mathbf{x}}_{{\mathbf{1}}} } \\ {{\mathbf{x}}_{{\mathbf{2}}} } \\ \end{array} } \right] $$ (5) The input coefficient matrix becomes a square, and, with the identity matrix \( {\mathbf{I}} \) for appropriate orders, the Leontief inverse matrix can be obtained as follows. $$ \left[ {\begin{array}{*{20}c} {{\mathbf{x}}_{{\mathbf{1}}} } \\ {{\mathbf{x}}_{{\mathbf{2}}} } \\ \end{array} } \right] = \left[ {\begin{array}{*{20}c} {{\mathbf{I}} - {\mathbf{A}}_{{{\mathbf{11}}}} } & { - {\mathbf{A}}_{{{\mathbf{12}}}} {\mathbf{c}}} \\ { - {\mathbf{A}}_{{{\mathbf{21}}}} } & {{\mathbf{I}} - {\mathbf{A}}_{{{\mathbf{22}}}} {\mathbf{c}}} \\ \end{array} } \right]^{{ - {\mathbf{1}}}} \left[ {\begin{array}{*{20}c} {{\mathbf{f}}_{{\mathbf{1}}} } \\ {{\mathbf{f}}_{{\mathbf{2}}} } \\ \end{array} } \right] $$ (6) In the hydrogen sector, vector “c” represents the composition of hydrogen supply for two different technologies. By changing the proportion of the conventional method for direct decomposition of methane, it is possible to see the influence exerted in CO2 emissions. In addition, it is also possible to simulate how CO2 emissions differ at the same final demand level (i.e., with the same GDP). 4 Assumptions and results of analysis We examine the impact of the following three aspects. 1. By introducing FCVs, there is fuel substitution effect from gasoline to hydrogen.   2. Conventional vehicles and FCVs have different economic consequences in the automobile production process because the input structure for manufacturing each automobile is different.   3. By considering two hydrogen production technologies, that is, methane steam reforming and direct decomposition of methane, different economic and environmental effects are expected because these technologies have different input structures.   For the hydrogen production sector, there are different activities for one product, and it is necessary to generate the composition ratios externally as a scenario.5 According to The Strategic Road Map for Hydrogen and Fuel Cells, released in March 2016 by the hydrogen-fuel cell strategy council of Japan’s Ministry of Economy, Trade and Industry (METI), fuel cell vehicles (stock base) are projected at 800 thousand units in 2030. The price of FCVs to be realized would be equivalent to the price of a hybrid vehicle by 2025. In addition, the Japanese government plans to establish 900 hydrogen fueling stations by 2030. The price of hydrogen is equal to or less than the fuel cost for hybrid vehicles. 4.1 Input structure of the automobile sector According to the Strategic Road Map for Hydrogen and Fuel Cells mentioned above, we assume that the spread of FCVs will proceed at a pace of 53,333 vehicles per year (cumulative total of 800,000 vehicles over 15 years). Table 4 shows the estimated decline in the price of components used in FCVs, based on the research on next-generation vehicles by the Chubu Region Institute for Social and Economic Research. The cost of hydrogen tanks, fuel cell stacks, and batteries, which currently increase the price of FCVs, are expected to drop by 2030 due to advances in manufacturing technology and mass production effects. By 2025, the goal is to fix FCV prices as high as hybrid vehicles. We assume that the prices for major FCV components shown in Table 4 are realized and the purchaser’s price of an FCV6 declines to 3.696 million yen (51.1% of the current price, 7.236 million yen in the case of the Toyota Mirai hydrogen fuel cell vehicle). Table 4 Costs of FCV components (Unit: Yen). Source: Influence of Next Generation Mobility Promotion on Chubu Region Industries, CRISER (2015) Sectors 2013 2030 Ratio Parts name Other ceramics and non-mineral products 2,000,000 549,250 0.275 Hydrogen tank Industrial electric machine 116,100 48,549 0.418 Driving motor, generator motor, motor inverter (for driving), inverter (for generator), inverter, DC–DC converter, reactor Electronic appliances and electric measuring instruments 9388 6398 0.682 Battery management unit, electric current sensor for inverter, electric current center for battery Other electrical equipment 1,330,764 408,229 0.307 Nickel metal hydride battery, lithium ion battery, fuel cell stack Electronic parts 3667 2591 0.707 Electric current sensor (for inverter) Auto parts − 241,294 − 241,294 1.000 Engine, fuel tank, transmission, etc. Total 3,459,919 1,015,017 0.293   Producer’s price 4,984,800 2,546,398 0.511 Purchaser’s price 7,236,000 3,696,384 0.511 Year “2013” is set as an estimation year in the original table of CRISER (2015), though FCV “Mirai” of Toyota Motor Co. released in 2014 As shown in Table 5, from the 2011 input–output table, it can be seen that the average price of an ordinary size car is 2.942 million yen (purchaser’s price), and the producer’s price is 2.022 million yen, excluding the commercial and transport margins, although the average prices in the automobile sector are 2.406 million yen and 1.653 million yen for purchaser’s price and producer’s price, respectively. Compared to these prices, the FCV purchaser’s price of 7.236 million yen in 2014 is exorbitant, which is one reason that FCVs are not accepted in the market. Therefore, as shown in Table 4, we assume that the price of FCVs is reduced to 3.696 million yen, replacing gasoline vehicles sold at the same price.7 Table 5 Relationship between car purchaser price and producer price. Source: Authors’ calculation   Automobile sector average Ordinary size vehicle FCVs (2014) Gasoline vehicle FCVs (2030) Unit price (Yen) Unit price (Yen) Unit price (Yen) Unit price (Yen) Unit price (Yen) Producer’s price base 1,653,468 2,021,748 5,995,755 2,539,753 2,539,753 Wholesale margins 48,407 59,188 79,728 74,353 74,353 Retail margins 665,039 813,164 1,095,356 1,021,510 1,021,510 Railway freights 26 32 43 41 41 Road freights 30,646 37,472 50,476 47,073 47,073 Coastal and inland water freights 1242 1519 2046 1908 1908 Port service fees 4266 5216 7026 6552 6552 Domestic air freights 0 0 0 0 0 Handling service fees 1651 2018 2719 2536 2536 Warehouse fees 1731 2116 2850 2658 2658 Purchaser’s price base 2,406,475 2,942,474 7,236,000 3,696,384 3,696,384 We estimate the input structure of FCVs as follows. First, we obtain the input values for the gasoline vehicle with a purchasing price of 3.696 million yen, after conversion to the producer’s price of 2.540 million yen, by multiplying the input coefficient of the automobile sector. To obtain input values for the FCV, we modify the costs of the gasoline vehicle. We break the total cost down into major input expenses and indirect expenses and obtain each sectoral cost for the indirect expenses by multiplying their total value by the input coefficient of headquarters’ activity sector.8 The sectoral costs for the major inputs are obtained by subtracting the indirect costs from the originally estimated costs. Thus, major material costs to produce FCVs are modified, based on the costs of parts in Table 4. The difference in the total input cost between the two types of vehicle is absorbed in the value-added sectors of the FCV not to change the price.9 We finally estimate sectoral inputs by adding two estimated costs: major costs and indirect costs. We refer to a gasoline vehicle at the same price in comparison with an FCV. The difference between them is as follows. Table 6 partly shows the estimated input coefficients of gasoline vehicles and FCVs in the input–output table, which is integrated into 38 sectors to make it easier to see the characteristics.10 In FCVs, inputs for ceramic, stone and clay products, electric machinery, and electronic components have increased, while input for transportation machines (automobile parts) is decreasing. Total input ratio in FCVs is larger than that in gasoline vehicles, so that the value-added ratio for each stands in opposite relation. Table 6 Estimated input coefficients in gasoline vehicles and FCVs. Source: Authors’ calculation No. Name of sector Gasoline vehicles FCVs 1 Transportation equipment 0.5373 0.3327 2 Ceramic, stone, and clay products 0.0198 0.2303 3 Electrical machinery 0.0417 0.1824 4 Iron and steel 0.0475 0.0378 5 Plastic and rubber products 0.0416 0.0331 6 Education and research 0.0408 0.0325 7 Commerce 0.0287 0.0229 8 Business services 0.0281 0.0224 9 Information and communication electronics equipment 0.0208 0.0166 10 Transport and postal services 0.0192 0.0153 11 Nonferrous metals 0.0074 0.0059 12 Electricity, gas, and heat supply 0.0062 0.0050 13 Chemical products 0.0060 0.0048 14 Finance and insurance 0.0043 0.0034 15 Metal products 0.0041 0.0032 16 Textile products 0.0030 0.0024 17 Information and communications 0.0026 0.0021 18 Miscellaneous manufacturing products 0.0023 0.0018 19 Real estate 0.0014 0.0011 20 Electronic components 0.0000 0.0010   Other inputs 0.0043 0.0034   Total input ratio 0.8670 0.9599   Value-added ratio 0.1330 0.0401 Table 7 shows the overall estimates of fuel purchases for gasoline vehicles and FCVs in the market. The number of units purchased is 53,333 units per year, assuming that the target number of FCVs is 800,000 units over 15 years. The price is 3.696 million yen for each, and the annual purchase amount is 197.14 billion yen. Since it is assumed that the price of conventional vehicles, replaced by FCVs, is the same, sales value is 197.14 billion yen. Table 7 Comparison of gasoline vehicles and FCVs. Source: Authors’ calculation   Gasoline vehicles FCVs Unit Year Stock of FCV 800,000 800,000 unit 2030 Sales per year 53,333 53,333 unit/year   Unit price 3.696 3.696 mil. Yen   Annual purchase amount 197,140 197,140 mil. Yen   Average mileage 8000 8000 km/year 2017 Fuel efficiency 10.0   km/l 2015 Gasoline consumption per year unit 800.0   l/year unit   Gas price 137.8   Yen/l 2015 Gasoline consumption amount 5878   mil. Yen   Tank capacity of FCV   5 kg   Cruising distance of FCV   650 km   Hydrogen fuel efficiency   130 km/kg   Hydrogen consumption per year unit   61.54 kg/year unit   Hydrogen price   1080 Yen/kg   Hydrogen consumption amount   3545 mil. Yen   Gasoline for 800 thousand vehicles 640,000   kl   Hydrogen for 800 thousand FCV   49,230,769 kg   Fuel consumption amount for 800 thousand vehicles 88,173 53,169 mil. Yen   CO2 emissions 1,486,080   t-CO2   CO2 emissions coefficient 23.591   t-CO2/mil. Yen   If the average mileage11 of one vehicle is 8000 km/year and fuel efficiency12 is 10 km/l, then annual gasoline consumption for a vehicle will be 800.0 l/year. If gasoline price13 is 137.8 yen/l, then annual gasoline consumption value is projected at 5.878 billion yen. On the other hand, since the tank capacity of an FCV is 5 kg of hydrogen and its cruising distance is 650 km, then hydrogen fuel efficiency is 130 km/kg, assuming the same average mileage of 8000 km/year. Hydrogen consumption is 61.54 kg/year. Therefore, if the price of hydrogen is 1080 yen/kg,14 hydrogen consumption value is estimated at 3.545 billion yen. Total gasoline consumption for 800,000 units is 640,000 kl, and consumption value is 88.173 billion yen. Hydrogen consumption is 49,230,769 kg, and consumption value is 53.169 billion yen. Thus, CO2 emissions due to gasoline consumption are 1,486,080 t-CO215 and CO2 emissions coefficient is estimated as 23.591 t-CO2/million yen, considering the commercial margin and transport cost. 4.2 Hydrogen production and input structure We compare the two hydrogen production technologies: steam reforming and direct decomposition of methane. The upper half of Table 8 shows the amount of material methane, heating methane, CO2 generated, and carbon in mol when producing 1000 mol of hydrogen. Table 8 Material balance of hydrogen production (by technology). Source: Authors’ calculation   Steam reforming of methane Direct decomposition of methane Unit Methane for material 250.000 500.000 mol Methane for heating 46.213 41.951 mol Methane, total 296.213 541.951 mol CO2 emissions 296.213 41.951 mol Carbon 0.000 500.000 mol Hydrogen 1000.000 1000.000 mol Methane for material 4.000 8.000 kg Methane for heating 0.739 0.671 kg Methane, total 4.739 8.671 kg CO2 emissions 13.033 1.846 kg Carbon 0.000 6.000 kg Hydrogen 2.000 2.000 kg In steam reforming of methane, according to Eq. (1), 250 mol of material methane and 46.213 mol of heating methane (in total 296.213 mol) are required for producing 1000 mol of hydrogen with 296.213 mol of CO2 as emissions. By contrast, in direct decomposition of methane, according to Eq. (2), 500 mol of material methane and 41.951 mol of methane for heating (in total 541.951 mol) are required for producing 1000 mol of hydrogen.16 In this process, 41.951 mol of CO2 is generated due to combustion of methane for heating, and 500 mol of (solid) carbon is generated. For producing the same hydrogen, in direct decomposition of methane, methane required for material and heating is 1.83 times compared with steam reforming of methane, but the amount of CO2 generated is only 14.2%. The lower half of Table 8 shows the relationship on a mass kg basis. Table 9 shows the input structure for both manufacturing methods.17,18 Hydrogen production amounts to 53.169 billion yen, a value obtained from annual consumption of hydrogen at 49,230.8 t for 800 thousand units of FCVs and hydrogen price of 1080 yen/kg (see Table 7). Methane material costs can be calculated from the relationship in Table 8. Capital depreciation is calculated assuming the establishment of 900 hydrogen stations, at a construction cost of 500 million yen per site, with a service life of 20 years. Indirect expenses were obtained from the annual expenses of 20 million yen per station, which is a METI estimate. In both the technologies, transportation margin for hydrogen is required in the case of off-site production but not for on-site production. This transport margin is set as 1% of the total costs, considering the corresponding value for gasoline production. Table 9 Input structure of hydrogen production (by technology). Source: Authors’ calculation   Steam reforming of methane Direct decomposition of methane Unit Methane input 6694 15,879 mil. Yen Indirect expenses 18,000 18,000 mil. Yen Transport costs 539 539 mil. Yen Other 5976 − 3210 mil. Yen Capital depreciation allowance 22,500 22,500 mil. Yen Hydrogen production 53,169 53,169 mil. Yen CO2 emission coefficient 6.034 0.855 t-CO2/mil. Yen Carbon (by-product)   59,062 mil. Yen Methane input 0.1259 0.2987 – Indirect expenses 0.3385 0.3385 – Transport costs 0.0101 0.0101 – Other 0.1124 − 0.0604 – Capital depreciation allowance 0.4232 0.4232 – Hydrogen production 1.0000 1.0000 – Input ratio is shown in the lower half of Table 9. The coefficient of CO2 emissions is 0.855 t-CO2/million yen for direct decomposition of methane and 6.034 t-CO2/million yen for the methane steam reforming method. Based on this, the input structure of hydrogen production was estimated. First, total cost is divided into direct expenses such as methane for materials and other indirect expenses. The former is estimated from Table 9, and the latter is introduced by referring to the input structure of the headquarters sector as in the case of FCV in 4.1. Finally, we summed up both and fixed them as sectoral inputs for hydrogen production. The main difference in the input structure of both technologies is the amount of methane used as the raw material and source of heat. Direct decomposition of methane needs approximately double the methane input as in the methane steam reforming method. However, the former has a bigger advantage when CO2 emissions are lower than the latter. The presence or absence of a transport margin also depends on whether production is on-site or off-site. 4.3 Simulation results Simulation is conducted for the following cases. 1. Purchase of gasoline vehicle: Purchase of 53,333 gasoline vehicles per year, at the same price as FCVs   2. Purchase of FCVs: 53,333 FCVs per year   3. Gasoline purchase: Annual purchase of gasoline for gasoline vehicles in case 1   4. Hydrogen purchase: Annual purchase of fuel hydrogen for FCVs in case 2   5. Substitute gasoline vehicles by FCVs (subtract case 2 from case 1)   6. Substitute gasoline with hydrogen (subtract case 4 from case 3)   In the analysis below, we used an input–output table aggregated into 188 sectors based on the 2011 national input–output table (benchmark table) and the employment table. As for CO2 emissions, we obtained the sectoral CO2 emissions given by the National Institute for Environmental Studies (3EID) corresponding to the 2011 input–output table.19 Table 10 shows the final demand, induced production amount, gross added value, number of workers, and CO2 emissions for cases 1 to 6, when hydrogen is produced by the methane steam reforming method. Further, CO2 emissions indicate these industries (endogenous sectors) and household sectors. Table 10 Hydrogen production by steam reforming of methane method. Source: Authors’ calculation   (1) Gasoline vehicle purchase (2) FCV purchase (3) Gasoline purchase (4) Hydrogen purchase (5) Vehicle substitution (2)–(1) (6) Fuel substitution (4)–(3) Unit Final demand 197,140 197,140 5878 3545 0 − 2334 mil. Yen Production 509,121 497,773 7384 4742 − 11,347 − 2642 mil. Yen Gross added value 165,861 156,998 2788 3288 − 8863 500 mil. Yen Employment 25,374 25,743 339 110 369 − 229 Persons CO2 emissions 509,445 685,733 114,074 22,891 176,288 − 91,183 t-CO2  Industrial sector 509,445 685,733 15,002 22,891 176,288 7889 t-CO2  Household sector 0 0 99,072 0 0 − 99,072 t-CO2 Although final demand for gasoline vehicles and FCVs is the same, the economic ripple effect of FCVs is relatively small compared to gasoline vehicles. Production of gasoline vehicles requires more automobile parts, and the sector has a greater ripple effect. On the other hand, FCVs use more electrical components, such as electric machinery, electronic parts, and industrial machinery. For this reason, the effect of substitution from gasoline vehicles to FCVs, in case 5, is negative for production, added value, although a positive effect is obtained for employment. It has the impact of increasing CO2 emissions by 17,288 t-CO2. Gasoline consumption value in using the vehicle is higher than hydrogen consumption, under the assumed prices. Therefore, replacing energy from gasoline with hydrogen reduces final demand, resulting in a negative impact on production and employment, even though the effect on value-added is positive. However, in this case, hydrogen production induces more CO2 emissions than gasoline production; thus, CO2 emissions increase in the industrial sector by 7899 t-CO2. Additionally, in the household sector, gasoline consumption directly results in 99,072 t-CO2 emissions, but if replaced with hydrogen, the same amount of CO2 reduction is achieved. Overall, these amounts are reduced by 91,183 t-CO2. Table 11 shows similar simulation results when hydrogen is produced by direct decomposition of methane. The direction of the effect in each case is almost the same as in Table 10, except CO2 emissions in the fuel substitution of the industrial sector. Carbon dioxide emissions increase by 176,288 t-CO2 when the impact of vehicle substitution decreases by 9372 t-CO2 in the industrial sector due to fuel substitution, and by 99,072 t-CO2 in the household sector, resulting in total reduction of 108,444 t-CO2. Compared with Table 10, even though the effect of the household sector remains dominant, additional reduction is achieved in the industrial sector. Table 11 Hydrogen production by direct decomposition of methane method. Source: Authors’ calculation   (1) Gasoline vehicle purchase (2) FCV purchase (3) Gasoline purchase (4) Hydrogen purchase (5) Vehicle substitution (2)–(1) (6) Fuel substitution (4)–(3) Unit Final demand 197,140 197,140 5878 3545 0 − 2334 mil. Yen Production 509,121 497,773 7384 5653 − 11,347 − 1731 mil. Yen Gross added value 165,861 156,998 2788 2977 − 8863 189 mil. Yen Employment 25,374 25,743 339 136 369 − 202 Persons CO2 emissions 509,445 685,733 114,074 5630 176,288 − 108,444 t-CO2  Industrial sector 509,445 685,733 15,002 5630 176,288 − 9372 t-CO2  Household sector 0 0 99,072 0 0 − 99,072 t-CO2 In Tables 10 and 11, we evaluate the production effect of replacing gasoline vehicles by FCVs and the changing effect of fuel purchase per year required for using vehicles. However, cars, as consumer durable goods, can be used for a certain period of time, during which fuel purchase is required. According to the statistics given by the Ministry of Land, Infrastructure and Transport, a passenger car’s average life span in 2017 was 12.9 years. Thus, we assume that the car purchased will be used for a slightly longer period of 15 years. We obtained the effect of CO2 emissions for 15 years of vehicle and fuel substitution, as summarized in Table 12. Table 12 Hydrogen production by direct decomposition of methane: Indicative of accumulated effect only (Unit: t-CO2). Source: Authors’ calculation Year (5) Vehicle substitution (6) Fuel substitution (industrial) (6) Fuel substitution (household) Total 1 176,288 − 9372 − 99,072 67,844 2 176,288 − 18,744 − 198,144 − 40,600 3 176,288 − 28,116 − 297,216 − 149,044 4 176,288 − 37,488 − 396,288 − 257,488 5 176,288 − 46,860 − 495,360 − 365,932 6 176,288 − 56,232 − 594,432 − 474,134 7 176,288 − 65,604 − 693,504 − 582,820 8 176,288 − 74,976 − 792,576 − 691,264 9 176,288 − 84,348 − 891,648 − 799,708 10 176,288 − 93,720 − 990,720 − 908,152 11 176,288 − 103,092 − 1,089,792 − 1,016,596 12 176,288 − 112,464 − 1,188,864 − 1,125,040 13 176,288 − 121,836 − 1,287,936 − 1,233,484 14 176,288 − 131,208 − 1,387,008 − 1,341,928 15 176,288 − 140,580 − 1,486,080 − 1,450,372 Table 12 shows the calculation for hydrogen production using the method of direct decomposition of methane. According to this table, CO2 emissions increased by 176,288 t-CO2 only in the first year due to vehicle substitution; however, CO2 reduction for fuel substitution, which occurs when using the car for 15 years, is 9372 t-CO2 per year in the industrial sector, 99,072 t-CO2 in the household sector, and the cumulative effects of 15 years show 140,580 t-CO2 and 1,486,080 t-CO2, respectively, amounting to 1,450,372 t-CO2. This effect varies with the choice of hydrogen production technology. Table 13 shows the kind of change that occurs depending on the ratio of the two hydrogen production technologies. This table shows values for on-site production. Thus, when the ratio of hydrogen production for direct decomposition of methane is 0% (hydrogen produced by methane steam reforming method), 20% case, 40% case, 60% case, 80% case, and 100% (hydrogen produced by direct decomposition of methane only), the cumulative effect of CO2 emissions in the 15th year, as shown in Table 12, is compared. Table 13 Hydrogen production rate and changes in CO2 emissions by direct decomposition of methane (on-site) (Unit: t-CO2). Source: Authors’ calculation   0% 20% 40% 60% 80% 100% Vehicle substitution 176,288 176,288 176,288 176,288 176,288 176,288 Fuel substitution (industry) 118,331 66,549 14,767 − 37,015 − 88,798 − 140,580 Industry, total 294,619 242,837 191,054 139,272 87,490 35,708 Fuel substitution (household) − 1,486,080 − 1,486,080 − 1,486,080 − 1,486,080 − 1,486,080 − 1,486,080 Total − 1,191,461 − 1,243,243 − 1,295,026 − 1,346,808 − 1,398,590 − 1,450,372 Vehicle substitution 1.000 1.000 1.000 1.000 1.000 1.000 Fuel substitution (industry) 1.000 0.562 0.125 − 0.313 − 0.750 − 1.188 Industry, total 1.000 0.824 0.648 0.473 0.297 0.121 Fuel substitution (household) 1.000 1.000 1.000 1.000 1.000 1.000 Total 1.000 1.043 1.087 1.130 1.174 1.217 In Table 13, the total CO2 reduction effect of hydrogen production by methane steam reforming method only (0%) was 1,191,461 t-CO2, whereas in the case of hydrogen production by direct decomposition of methane method only (100%), it was 1,450,372 t-CO2. The latter reduces about 21.7% more CO2 emissions than the former. This effect can be divided into the impact of vehicle substitution and effects of fuel substitution. The effect of fuel substitution can be further divided into industrial sector and household sector. Among them, the most effective CO2 reduction method is the effect of fuel substitution in the household sector, and this effect will be constant at 1,486,080 t-CO2, regardless of the hydrogen production technology. Furthermore, the effect of vehicle substitution is constant but increasing at 176,288 t-CO2 for the choice of hydrogen production technology. The effect varies strongly for fuel substitution in the industrial sector, from 118,331 t-CO2, increasing in the case of hydrogen production only with methane steam reforming (0%), to 140,580 t-CO2, decreasing in the case of hydrogen production by the decomposition method (100%). Table 14 shows the accumulated effect of the selection of on-site or off-site hydrogen production, as well as the selection of two production technologies. Changes in CO2 emissions from vehicle substitution and fuel substitution occurring in the industrial and household sectors are shown by the selection of hydrogen production technology (0% or 100%). The values for the three rows at the bottom of the column for 100%, which shows the case of direct decomposition of methane (on-site production), correspond to CO2 emissions in the 15th-year effect in Table 12. Table 14 Differences in CO2 emissions between on-site and off-site hydrogen production (Unit: t-CO2). Source: Authors’ calculation   On-site production Off-site production Steam reforming of methane (0%) Direct decomposition of methane (100%) Steam reforming of methane (0%) Direct decomposition of methane (100%) Vehicle substitution Fuel substitution Vehicle substitution Fuel substitution Vehicle substitution Fuel substitution Vehicle substitution Fuel substitution Agriculture, forestry, and fishery − 1 − 2 − 1 − 0 − 1 − 2 − 1 − 0 Mining 3357 − 628 3357 − 528 3357 − 628 3357 − 527 Hydrogen 0 320,822 0 45,436 0 320,822 0 45,436 Petroleum products 1575 − 161,062 1575 − 160,244 1575 − 160,968 1575 − 160,150 Gasoline vehicles − 7903 0 − 7903 0 − 7903 0 − 7903 0 FCVs 7903 0 7903 0 7903 0 7903 0 Other manufacturing industries 151,759 − 1805 151,759 − 368 151,759 − 1767 151,759 − 330 Electricity 15,803 − 26,949 15,803 − 22,374 15,803 − 26,801 15,803 − 22,226 City gas − 99 4863 − 99 11,709 − 99 4863 − 99 11,710 Other tertiary industries 3894 − 16,907 3894 − 14,211 3894 − 14,819 3894 − 12,123 Industrial sector total 176,288 118,331 176,288 − 140,580 176,288 120,700 176,288 − 138,211 Household sector 0 − 1,486,080 0 − 1,486,080 0 − 1,486,080 0 − 1,486,080 Total 176,288 − 1,367,749 176,288 − 1,626,660 176,288 − 1,365,380 176,288 − 1,624,291 It is evident here that CO2 emissions in the hydrogen production sector for both fuel on-site and off-site are about seven times more in the methane steam reforming method (320,822 t-CO2) than in the methane direct decomposition method (45,436 t-CO2). Although fuel substitution in the industrial sector augments 118,331 t-CO2 in the methane steam reforming method, it saves 140,580 t-CO2 in direct decomposition method. In off-site hydrogen production, hydrogen has to be transported to the hydrogen refueling station, so that induced production increases and CO2 emissions also rise. As a result, the saving effect of CO2 emissions will lower, according to the results in Table 14. 5 Conclusions Japan is moving toward its target of reducing GHG emissions by 26% by 2030 from the 2013 levels. To attain this target, it becomes inevitable to introduce energy-saving technologies in industries, transportation, business, and household sectors to transit from a society dependent on fossil fuels to one based on renewable energy, and obtain fuels that do not emit CO2. Fuel cells with hydrogen fuel are emerging as a viable solution and attracting wider attention. We analyze the economic and environmental impact of defusing FCVs using hydrogen fuel with the selection of several production technologies. The overall effect of production on the economy, value-added, employment, and CO2 emissions is obtained by the scenario input–output analysis model. As for the hydrogen production technology, we compared the steam reforming method, which is currently considered mainstream, and our newly developed methane direct decomposition method. The findings obtained are as follows. 1. Substituting conventional vehicles with FCVs has a negative effect on production value, added value, and employment, because the ripple effect of producing FCVs is relatively small compared to conventional vehicles. However, CO2 emissions increase by 176,285 t-CO2 because carbon products are used in FCV production.   2. Fuel substitution from gasoline to hydrogen has the dominant effect of reducing CO2 emissions, 1,486,080 t-CO2, in the household sector.   3. In the industrial sector, the effect depends on the selection of the hydrogen production technology. Both technologies have CO2 emissions directly in their production. However, the methane direct decomposition method lowers CO2 emissions to 14.1% by the methane steam reforming method.   4. In addition, substitution of fuels in the industrial sector augments 118,331 t-CO2 in the methane steam reforming method, although it saves 140,580 t-CO2 in the methane direct decomposition method.   5. The effect on the broader economy is the reduction of CO2 emissions for any hydrogen production technology because the saving effect in the household sector is dominant for any method. However, hydrogen production by the methane direct decomposition method saves CO2 emissions by 21.7% more than that by the methane steam reforming method.   In our analysis, we compared gasoline vehicles and FCVs of the same price and same volume. However, there is another possibility in comparison of cheaper but fuel-inefficient conventional vehicles with expensive but fuel-efficient FCVs. This might be important criteria of comparison. Also, we do not consider the impact of the construction of hydrogen refueling station and the capital investment effect due to expansion of automobile production. These effects increase production, added value, employment, and CO2 emissions and, therefore, reduce the effect evaluated in this research. Effective utilization of the solid carbon, obtained as by-product in the hydrogen production by the direct decomposition of methane, is another remaining issue. Furthermore, we focused on diffusion of FCVs and selecting hydrogen production technology, but hydrogen use is not limited only to vehicles. The potential of using hydrogen for fuel combustion and generating electric power exists. Issues concerning the economic and environmental impact of the technology choice will form the focus of our future research. Footnotes 1. According to this agreement, each country sets an individual greenhouse gas (GHG) emission targets called nationally determined contributions (NDC). 2. Hydrogen, a source of energy for fuel cells, can be generated in various ways. Representative examples include extraction from fossil fuels or electrolysis of water. Currently, there are two practical ways to generate hydrogen: steam reforming of natural gas during petroleum refining and gasification of coal. These methods, however, are at a disadvantage as they emit CO2 during the process of generating hydrogen. 3. Valente et al. (2017) conduct a literature review of the methodological choices made in LCA studies of hydrogen energy systems. 4. Input–output analysis with a single product produced by multiple activities appears in Yoshioka and Suga (1997), Wang (2016), and Fujikawa and Wang (2017). 5. Substitution from conventional vehicles to FCVs can be regarded as having two activities for one automobile sector, although we treat these sectors as two independent entities because the automobile, which is the final good, has no intermediate demand. This calculation results in the same values as the case with the given composition of vehicles. 6. Purchasing an FCV has a subsidy of 2 million yen, but the high manufacturing costs impede market diffusion. Low accessibility to hydrogen fueling stations is another obstacle. 7. We assume that gasoline vehicle has the same price of FCV, 3.696 million yen. This is regarded as the condition that any user of the conventional vehicle replaces it easily to an FCV. However, in the 2011 input–output table, the purchaser’s price of the normal-size car with more than 2000 cc engine capacity was 2.942 million yen, which might be another option for the price of a gasoline car. The choice of expensive but fuel-efficient FCVs and cheaper but fuel-inefficient conventional vehicles is one of the important issues. We would like to examine this issue as a future challenge, although conclusions in our paper are at least quantitatively invariant against the variation of vehicle prices. 8. The input coefficient of headquarters’ activity sector is obtained from the 2011 Tokyo metropolitan input–output table. 9. In estimation of the input structure of FCV, we set the future costs of major parts, by referring the report on the next-generation mobility (CRISER 2015). However, the future costs of the other parts and services are indirectly estimated by multiplying the future price of FCV by the current input coefficients of the corresponding sector. In that sense, our estimation includes partly some kind of errors, which is one of the remaining issues. 10. Input coefficient is estimated based on the input–output table of 188 sectors, as described later. 11. The average mileage of conventional vehicles is obtained from a report by Next Generation Vehicle Promotion Center on the diffusion of clean-energy vehicles in 2017. 12. Fuel efficiency of gasoline vehicles in 2015 is calculated from the Fuel Consumption Statistics of Japan’s Ministry of Land, Infrastructure and Transport. 13. Value as given by Ministry of Resources and Energy, 2015. 14. JX Nippon Oil and Energy Corporation started to sell hydrogen for 1000 yen/kg (excluding consumption tax), according to a newspaper article from Nikkei Inc. dated December 26, 2014. 15. Gasoline CO2 emission coefficient is 2.322 kg-CO2/l, which is calculated by using the 2005 revised data of Resource and Energy Agency, Japan. 16. It requires 41.2 kJ/mol-H2 for steam reforming of methane and 37.4 kJ/mol-H2 for direct decomposition of methane. Methane calorific value is 39.8 MJ/Nm3 and the molar volume is 22.4 l and the volume of heating methane is calculated. To evaluate the constant pressure specific heat of each species, thermodynamic data were cited from JANAF Table (see Stull and Prophet (1971)). 17. In our analysis, we adopted the simplest input structure for hydrogen production and used methane as an input, as a preliminary approach. Methane is a dominant input, although other materials can be used as the energy source of energy. 18. In Table 9, the value of the by-produced carbon in the methane direct decomposition method is calculated as reference by using the price of carbon rods, 399.9 yen/kg, which is the lowest price among carbon products in the 2011 input–output table of Japan. The by-product carbon is not used in the following simulations. 19. National Institute for Environmental Studies (2018). In our analysis, we apply a domestic input–output model to evaluate domestically economic and environmental effects of the choice of hydrogen production technologies in the production, though footprint and global allocation of the hydrogen production are another important issues. Notes Authors’ contributions All authors have contributed equally to designing the research, the process of data collection and calculation, and drafting and revision of the manuscript. All authors have read and approved the final manuscript. Acknowledgements This paper is an outcome of our research, themed “Development of methane direct decomposition hydrogen production system,” which is a part of the “Project E: Technological development for building a hydrogen-energy society” conducted at “Knowledge Hub Aichi,” the government of Aichi Prefecture. The original version of the paper was presented at the Third International Conference on Economic Structures held in Nagoya, Japan, from March 28–29, 2018. The authors express their gratitude for the support by “Knowledge Hub Aichi,” the government of Aichi Prefecture, and are deeply grateful to the anonymous referees who provided constructive comments and warm encouragement. Competing interests The authors declare that they have no competing interests. Availability of data and materials Various sources were used to collect data for the analysis. Each source of data and materials is available and referred to in the paper, where appropriate. Consent for publication Not applicable. Ethics approval and consent to participate Not applicable. Funding Aichi Science & Technology Foundation, government of Aichi Prefecture, supports this research financially. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Automobile Inspection & Registration Information Association (2018) Car ownership trend in Japan (in Japanese). http://www.airia.or.jp/publish/statistics/trend.html. Accessed 1 July 2018 Bohnes FA, Gregg JS, Laurent A (2017) Environmental impacts of future urban deployment of electric vehicles: assessment framework and case study of Copenhagen for 2016–2030. Environ Sci Technol 51:13995–14005CrossRefGoogle Scholar Cantono S, Heijungs R, Kleijn R (2008) Environmental accounting of eco-innovations through environmental input–output analysis: the case of hydrogen and fuel cells buses. Economic Systems Research 20(3):303–318CrossRefGoogle Scholar Chubu Region Institute for Social and Economic Research (2015) On the Influence of Next Generation Mobility Promotion on Chubu Region Industry (in Japanese) Google Scholar Dincer I, Acar C (2015) Review and evaluation of hydrogen production methods for better sustainability. Int J Hydrogen Energy 40:11094–11111CrossRefGoogle Scholar Fujikawa K, Wang J (2017) Economic and environmental impact of renewable energy: an application of input output analysis. Ritsumeikan Econ Rev 65(4):619–630 (in Japanese) Google Scholar Fukuhara H, Hondo H (2011) A generalization of matrix-based life cycle inventory analysis by the construction of a regular coefficient matrix using the graph theory. J Life Cycle Assess 7(1):59–71 (in Japanese) CrossRefGoogle Scholar Hydrogen and Fuel Cell Strategy Council, METI (2016) Hydrogen and fuel cell strategy roadmap: accelerate efforts toward realization of hydrogen society (in Japanese). http://www.meti.go.jp/press/2015/03/20160322009/20160322009.html. Accessed 1 July 2018 Ikeda A, Suga M, Shinozaki M, Hayami H, Fujiwara K, Yoshioka K (1996) Input–output table for environmental analysis, KEO Monograph Series No. 7 (in Japanese) Google Scholar Japan Automobile Manufacturers’ Association (2018) Sales statistics of Japanese vehicles (in Japanese). http://www.jama.or.jp/index.html. Accessed 1 July 2018 Keipi T, Tolvanen H, Konttinen J (2018) Economic analysis of hydrogen production by methane thermal decomposition: comparison to competing technologies. Energy Convers Manag 159:264–273CrossRefGoogle Scholar Klein LR (1983) Lectures in econometrics. North-Holland, AmsterdamGoogle Scholar Klein LR (2003) Some potential linkages for input–output analysis with flow-of-funds. Econ Syst Res 15(3):269–277CrossRefGoogle Scholar Kondo Y, Nakamura S (2004) Evaluating alternative life-cycle strategies for electrical appliances by the waste input–output model. Int J Life Cycle Assess 9(4):236–246CrossRefGoogle Scholar Leontief W (1970) Environmental repercussions and the economic structure: an input-output approach. Rev Econ Stat 52(3):262–271CrossRefGoogle Scholar Miller RE, Blair PD (2009) Input–output analysis: foundations and extensions, 2nd edn. Cambridge University Press, CambridgeCrossRefGoogle Scholar Ministry of Environment (2017) Emissions of greenhouse gases in Japan (preliminary figures for (2016)) (in Japanese). https://www.env.go.jp/press/104900.html. Accessed 1 July 2018 Miotti M, Hofer J, Bauer C (2017) Integrated environmental and economic assessment of current and future fuel cell vehicles. Int J Life Cycle Assess 22:94–110CrossRefGoogle Scholar Nakamura S, Kondo Y (2002a) Input–output analysis of waste management. J Ind Ecol 6(1):39–63CrossRefGoogle Scholar Nakamura S, Kondo Y (2002b) Recycling, landfill consumption, and CO2 emission: analysis by waste input-output model. J Mater Cycles Waste Manag 4(1):2–11Google Scholar Nakamura S, Kondo Y (2009) Waste input–output analysis, concepts and application to industrial ecology. Springer, BerlinCrossRefGoogle Scholar Nansai K, Kagawa S, Kondo Y, Suh S, Inaba R, Nakajima K (2009) Improving the completeness of product carbon footprints using a global link input–output model: the case of Japan. Econ Syst Res 21(3):267–290CrossRefGoogle Scholar National Institute for Environmental Studies (2018) Embodied energy and emission intensity data for Japan using input–output tables (3EID) (in Japanese). http://www.cger.nies.go.jp/publications/report/d031/index_j.htm. Accessed 1 July 2018 Next Generation Vehicle Promotion Center (2018) Statistics of the next generation vehicle (in Japanese). http://www.cev-pc.or.jp/chosa/. Accessed 1 July 2018 Nishiyama S (2008) A financial macro econometric model of the United States 1977–2002. J Appl Input Output Anal 13–14:1–31Google Scholar Stull DR, Prophet H (1971) JANAF thermochemical tables, 2nd ed, NSRDS-NBS 37. https://www.gpo.gov/fdsys/pkg/GOVPUB-C13-bfd606acc2525ccef2762b19002a6d4f/content-detail.html. Accessed 1 July 2018 Tsujimura K, Mizoshita M (2003) Asset–liability–matrix analysis derived from the flow-of-funds accounts: the Bank of Japan’s quantitative monetary policy examined. Econ Syst Res 15(1):51–67CrossRefGoogle Scholar Tsunoka S, Ychiyama Y, Okajima K (2011) An analytical method for environmental burdens on a production system including feedback flows. J Jpn Soc Energy Resour 32(4):9–17 (in Japanese) Google Scholar Tsunoka S, Uchiyama Y, Okajima K, Murata K (2012) Analysis of environmental burdens on by-product hydrogen in oil refinery using a matrix method. J Jpn Soc Energy Resour 33(1):23–31 (in Japanese) Google Scholar Usubiaga A, Acosta-Fernandez J (2015) Carbon emission accounting in MRIO models: the territory vs. residence principle. Econ Syst Res 27(4):458–477CrossRefGoogle Scholar Valente A, Iribarren D, Dufour J (2017) Life cycle assessment of hydrogen energy systems: a review of methodological choices. Int J Life Cycle Assess 22(3):346–363CrossRefGoogle Scholar Wang J (2016) Economic and environmental effects of introduction of renewable power sources: an application of scenario input–output analysis to China. Input Output Anal 24(1):35–48 (in Japanese) CrossRefGoogle Scholar Wiebe KS, Bruckner M, Giljum S, Lutz C (2012) Calculating energy-related CO2 emissions embodied in international trade using global input–output model. Econ Syst Res 24(2):113–139CrossRefGoogle Scholar Yoshioka K, Suga M (1997) Applications of input–output approach in environmental analysis: a study of scenario Leontief inverse, Economic Analysis. Economic and Social Research Institute, Cabinet Office, Japan, 154 (in Japanese) Google Scholar Copyright information © The Author(s) 2019 Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Authors and Affiliations Mitsuo Yamada1Email authorView author's OrcID profileKiyoshi Fujikawa23Yoshito Umeda41.School of EconomicsChukyo UniversityNagoya CityJapan2.Applied Social System Institute of AsiaNagoya UniversityNagoya CityJapan3.Institute of EconomicsChukyo UniversityNagoya CityJapan4.Toho Cryogenics Co., LtdNagoya CityJapan


This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1186%2Fs40008-019-0137-3.pdf

Mitsuo Yamada, Kiyoshi Fujikawa, Yoshito Umeda. Scenario input–output analysis on the diffusion of fuel cell vehicles and alternative hydrogen supply systems, Journal of Economic Structures, 2019, 4, DOI: 10.1186/s40008-019-0137-3