Assessing application potential of clean energy ...

Journal of Cleaner Production 75 (2014) 11e19

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Assessing application potential of clean energy supply for greenhouse gas emission mitigation: a case study on General Motors global manufacturing Qiang Zhai a, Huajun Cao b, Xiang Zhao c, Chris Yuan a, * a b c

Department of Mechanical Engineering, University of Wisconsin Milwaukee, Milwaukee, WI 53201, USA School of Mechanical Engineering, Chongqing University, Chongqing 400044, China Manufacturing Systems Research Lab, Global Research and Development, General Motors Company, Warren, MI 48090, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 June 2013 Received in revised form 4 March 2014 Accepted 24 March 2014 Available online 1 April 2014

Greenhouse gas (GHG) emissions from global manufacturing companies are significant due to the high percentage of fossil fuels consumed in current energy structure. Clean energy technologies are widely recognized for their cleanliness in power generations and accordingly are promising as alternative energy supply for GHG mitigation from global manufacturing. However, deploying clean energy technologies at a global scale requires multi-attribute decision-making and currently there is a lack of information for assessing the application potential of various clean energy systems. This paper presents a mathematical approach based on cost benefit analysis to evaluate the application potential of such clean energy systems as solar photovoltaic, wind, hybrid solar-wind, and hydrogen-based fuel cells to partially supply the electricity needs of global manufacturing to reduce the facility GHG emissions. A case study is conducted on six selected production sites from GM’s global production locations, with the future trend of the results analyzed till 2035. The analysis results reveal that the optimal selection and deployment of a clean power system are dependent on such factors as location, time, technology and scale. The highest cost benefit result is obtained on wind power system deployed in Bochum, with 15.5 tons of CO2,eq mitigation potential per $1000 cost input. The number will be increased to 15.73 and 19.96 tons of CO2,eq per $1000 cost input in 2020 and 2035, respectively. The models and results presented in this study could be useful in decision support of optimal selection and deployment of clean energy system by global manufacturers in future. Ó 2014 Elsevier Ltd. All rights reserved.

Keywords: Greenhouse gas emission Global industrial production Clean energy supply Cost benefit analysis General Motors

1. Introduction Industrial productions such as automotive manufacturing require significant amount of energy inputs for the process operations (Unander et al., 1999; Yuan et al., 2006; Gutowski et al., 2006). As current global energy supply is mainly produced from fossil fuels, the energy consumed in industrial productions generates huge amount of greenhouse gas (GHG) emissions which are of grave concerns worldwide due to their global warming effects (the Intergovernment, 2007). In general, energies consumed in industrial productions are from two categories: direct consumption by burning such fossil fuels as coal, natural gas, oil, etc., on-site for supporting industrial process operations, and indirect consumption

* Corresponding author. Tel.: þ1 414 229 5639; fax: þ1 414 229 6958. E-mail address: [email protected] (C. Yuan). http://dx.doi.org/10.1016/j.jclepro.2014.03.072 0959-6526/Ó 2014 Elsevier Ltd. All rights reserved.

by using grid electricity which is usually generated from a mix of energy sources including fossil fuels and renewable energies. From both direct and indirect sources, the amounts of GHG emissions from industrial productions are significant. Take General Motors (GM) as an example, statistical data shows that the GM manufacturing facilities in U.S. alone consumed more than 5.54 1016 J of energy in 2007 which led to generations of 6.26 million metric tons of CO2 emissions (General Motors Company, 2008). As a whole, the industrial sector in the U.S. produced 28.3% of the total 6702 Million Metric Tons of CO2 equivalent GHG emissions in 2011, higher than any other economic sectors including transportation, commercial, residential and agriculture (U.S. EPA, 2013). In global scale, the CO2 emissions from global industrial sector will grow significantly in this decade from 15,165 Million Metric Tons in 2010 to 18,081 Million Metric tons in 2020 (IPCC, 2007). Concerned on the huge amounts and the associated adverse impacts on global warming, industrial GHG emissions are

12

Q. Zhai et al. / Journal of Cleaner Production 75 (2014) 11e19

always under a priority consideration for emission mitigation by industrialized countries and international communities. Within the international framework of Kyoto Protocol, many countries have adopted certain types of climate policy, agreements and incentive programs for GHG emission mitigation and monitoring, including UK Climate Change Agreement, German Agreement on Climate Protection, Finland’s Agreement on the Promotion of Energy Conservation in Industry, Dutch Long-Term Agreement, Danish Agreement on Industrial Energy Efficiency, etc (IPCC, 2007). These agreements, in nature, provide governmental pressure and regulatory threat for industrial companies to act on their GHG emissions mitigation and controls. In general, there are two effective ways in reducing the GHG emissions from industrial facilities: improving energy efficiency, or using clean energy supply. In the past decades, improving energy efficiency was mainly adopted by industry for reducing the GHG emissions because it can also reduce the utility costs from the industrial operations. For example, GM’s U.S. facilities have conducted a total of 1753 projects from 1991 to 2007 for energy efficiency improvements and conversions of energy sources by using lower GHG emitting fuels such as switching from coal to natural gas for the operation of boilers, which has led to a GHG reduction of over 17 million metric tons CO2 equivalent (General Motors Company 2008). However, after decades of continuous improvement, it is difficult to further reduce CO2 emissions of industrial productions through energy efficiency improvement and conservation, because industrial productions are energy-driven and a baseline of energy input is always required to operate a production system. A potential solution to this dilemma is to use clean energy technologies such as solar photovoltaic, wind, fuel cells, etc., to partially supply the power needs of industrial productions, so as to further reduce the GHG emissions from industrial facilities (Yuan and Dornfeld, 2009; Zhai et al., 2011). Clean energy technologies are recognized for their cleanliness during power generations and are promoted for use worldwide (Bilgen et al., 2004). In the past years, clean energy technologies were limited in electricity generations mainly because of their high economic costs (Brown, 2001). With continuous technology improvement and cost decreasing in the past decades, the application of clean energy power systems becomes increasingly feasible when compared with conventional grid power supply in terms of tradeoffs between economic cost and environmental benefits. However, a rigorous literature review found that past research were mostly focused on economic and environmental impact assessment of clean energy systems. For solar photovoltaic (PV), Branker et al. provided a detailed review of solar PV levelized cost of electricity (Branker et al., 2011); Azzopardi and Mutale conducted a life cycle analysis of solar PVs (Azzopardi and Mutale, 2010), and Tsoutsos et al. performed an investigation on the environmental impacts of solar energy systems (Tsoutsos et al., 2005). For wind, European Wind Energy Association published a guide on wind technology, economic performance and future wind power (European Wind Energy Asso, 2009); comprehensive environmental impacts of offshore wind energy are presented by Koller (2006); challenges of onshore wind power development are reported by Han et al. for China (Han et al., 2009). Besides solar and wind, fuel cell stationary power systems using hydrogen fuels were also researched as a clean and renewable technology. Zoulias and Lymberopoulos provided a techno-economic analysis of standalone hydrogen fuel cell based power system (Zoulias and Lymberopoulos, 2007). Neuhoff reported economic reviews of large-scale deployment of hydrogen energies for electricity generations (Neuhoff, 2005).

While the past research on the economic and environmental performance analyses have greatly enhanced the large-scale deployment of clean power systems for electricity generations, there are few studies on the tradeoffs of economic cost and environmental benefits of the clean energy supply for reducing GHG emissions from global industrial productions such as GM’ automotive manufacturing (Yuan and Dornfeld, 2009; Zhai et al., 2011). Currently there is a lack of methods and data support for decisionmaking in optimal selection and deployment of clean power systems for GHG mitigation of large scale industrial productions in a global scale. In this paper, a mathematical approach is developed for quantifying the cost benefit of clean energy supply, to support decision-making in assessing the application potential of various clean energy technologies for partially supplying the power needs of large-scale global industrial facilities. A case study is conducted on a generic assessment of the application potential of four clean energy supply systems including solar PV, wind, hybrid solar-wind, and fuel cell stationary power systems on the global production facilities of GM. It is expected that this study would be useful for global industrial manufacturers in decision-making and strategyplanning of employing clean energy technologies to mitigate their GHG emissions in future. 2. Method Using clean energy supply for GHG emission mitigation from global industrial productions requires complex decision-making processes. Since a global manufacturer such as GM usually has production sites at many locations around the world, and there are a number of clean energy technologies available on the commercial market for power generations, employing clean energy supply for GHG mitigation needs to consider multiple factors such as the feasibility of GHG mitigation for a specific production site, local geographic conditions for a clean energy deployment, cost of clean power systems, amount of GHGs to be reduced, scale of the production, future trend of the clean energy technology and cost change, etc. Considering the major factors of industrial concerns on the application of clean energy systems in industrial productions, the application potential of various clean energy power systems for greenhouse gas emission mitigation can be quantitatively assessed based on a cost benefit analysis. The cost is the unit cost of the electricity generated from the clean energy systems, and the benefit is the amount of GHG emissions reduced. In this study, four common clean energy systems including solar PV, wind, hybrid solar-wind and fuel cells are selected for analyzing and benchmarking their application potentials through connecting to local power grids to partially supply the electricity needs of industrial production facilities. Fig. 1 shows the generic scheme of the cost benefit analysis structure for assessing the application potential of these four clean energy supplies in manufacturing industry. For assessing the application potential of a clean energy supply in GHG emission mitigation of industrial productions, the first step is to identify the feasibility of the clean energy supply at a specific geographic location. Since clean energy technologies such as solar PV, wind, fuel cells, etc., all generates GHG emissions from their life cycles, application of the clean energy supply needs to consider the amount of GHG emissions from the life cycle of the clean energy power systems. For effective GHG emission mitigation, the amount of GHG emissions from the life cycle of a clean energy system must be lower than that from the grid power supply, based on the same amount of power delivery. Here we employ the GHG emission factors of grid power supply at each location and the life cycle GHG emission factors of each clean energy system for the feasibility analysis. Such GHG emissions as CO2, CH4, N2O, etc., from both grid


13

Regarding economic costs of clean energy power systems, there are different economic parameters currently used for describing the costs of the commercial clean energy power systems. Commonly used economic costs data for clean energy systems are overnight cost, fixed operating and maintenance (fixed O&M) cost, variable operating and maintenance (variable O&M) cost, and endof-life management cost (U.S.EIA, 2010). The overnight costs are the cost estimates to build a plant in a typical region of the country (U.S. EIA, 2010); operating and maintenance costs are those costs associated with operations and maintenance of clean power systems during their service life, with some costs fixed and some varying; end-of-life costs are those costs for dismantling the power system and recycling materials. For different products in the clean energy power market, these costs may be different on the same type of power system, but the difference becomes more and more negligible due to the globalization of the clean energy market. In general, the total cost of a clean energy power system can be expressed by the following equation:

i i i CTi ¼ P i Coi þ CfOM þ CRi þ Eactual;k t i CvOM

(2)

where

Fig. 1. Scheme for the application potential analysis of clean energy supply.

power supply and clean energy systems are included in the analysis and all converted to CO2 equivalent amount using the standard Global Warming Potential (GWP) metric (IPCC, 2007). The emission factor of a local grid power supply is the amount of GHG emissions per kWh electricity supplied, which is determined based on the energy sources of electricity generation for the local power grid. Due to the differences of energy sources and their proportions in the local electricity mix, the GHG emission factors from grid power supply are quite different in different regions. The feasibility of a clean energy supply for GHG emission mitigation from an industrial facility at a specific geographic location can be simply evaluated using following equation.

Fki ¼

n X j¼1

n X j j GWPj fgrid;k GWPj fi

(1)

j¼1

where Fki : feasibility of clean energy supply i at location k; if Fki >0, feasible; otherwise, not feasible. GWPj: Global Warming Potential of the jth GHG, j fgrid;k : The jth GHG emission factor of a local grid power supply at location k, kg/kwh fij : The jth GHG emission factor from the life cycle of ith clean energy power system, kg/kwh

CTi : The total cost of ith clean energy power system in US$, Pi: The rated power of the ith clean energy power system in kW, Coi : The overnight cost of the ith clean energy power system, in US $/kW, i CfOM : The fixed operating & maintenance cost of the ith clean energy system, in US $/kW, CRi : The recycling cost of the ith clean energy system, in US $/kW i Eactual;k : The actual amount of electricity generated by the ith clean energy power system annually at location k, kwh/year i CvOM : The variable operating & maintenance cost of the ith clean energy system in US $/kWh, Ei: The total power output during the life time of the ith clean energy power system, in kWh. ti: The life time of the ith clean energy power system, in years Using the economic cost metric in equation (2), a mathematical model based on classic cost benefit approach can be developed for assessing the application potential of clean energy stationary power supply for GHG mitigations from global industrial facilities at a specific location. Here the model is developed with an aim to provide a simplified mathematical approach for the global manufacturing industry such as GM to support the decisionmaking and strategy-planning during the early stage of project assessment, budget-planning and goal definition for GHG mitigation from their global production facilities. In this mathematic model, the benefit is defined as the total amount of GHGs which can be reduced based on certain amount of economic input. From the perspective of applied industrial economics, the cost associated with the selected clean energy supply pattern should be minimal to achieve a pre-defined strategic goal of GHG reduction, or the reduced amount of GHG should be maximal based on a fixed amount of economic cost input. In this cost benefit model, all kinds of GHG emissions from electricity generations and supply are considered, including CO2, CH4, N2O, etc. The GHG emissions are characterized as the total amount of CO2 equivalent. Those non-CO2 emissions are transformed into CO2 equivalent through their IPCC global warming potential (GWP) metric (IPCC, 2007). In this model, the conversion of D.C. power output from the clean energy power systems to conventional A.C. power supply is considered in the actual power output. Overall operating efficiency of each clean energy power system corresponding to local

14


geographical and environmental conditions is also considered in the analysis. The mathematical expression of the cost benefit model is shown in the following:

" i Eactual;k

Bik

¼

ti

P n j j j j j ¼ 1 GWP fgrid;k j ¼ 1 GWP fi

Pn

#

i i i P i Coi þ CfOM þ CRi þ Eactual;k t i CvOM (3)

where Bik : cost benefit of the ith clean energy power system at location k, kg/$, or ton/$1000. i Eactual;k : The actual amount of electricity generated by the ith clean energy power system annually at location k, kwh/year ti: The life time of the ith clean energy power system, in years Using equation (3), the amount of GHG reduction per unit cost investment can be quantitatively determined for each clean power system at a specific location of industrial production facilities. But calculation of the actual power output from a clean power system, i Eactual;k , is complicated which is determined by the working principle of the selected clean energy technology, the scale of the power system to be deployed, and geographic factors influencing the power density and output of such clean power systems as solar and wind. For solar, the actual amount of electricity generated is determined by the local solar insolation level and the solar PV panel selected. In general, the actual power output of a solar PV system in terms of AC electricity for connecting to the local grid network can be calculated using the following equation: solar Eactual;k ¼ Nm Iave;k Am em fDCAC

(4)

where solar : the actual power output of solar PV system annually, AC Eactual;k electricity (kWh/year) Nm: number of PV modules Iave, k: the average annual solar insolation (kWh/m2/yr) at location k. Am: surface area of one PV module (m2) em: module efficiency (%) fdc-ac: DC-AC conversion efficiency

For wind power systems, the output is dependent on the local wind energy density which is usually a function of wind speed and air density in the selected geographic location for the power system deployment. The wind speed and air density vary at different heights above ground which are usually available from long-term meteorological survey database through statistical approaches. The actual wind energy power output at a selected location can be calculated through: wind Eactual;k ¼

lþ3 1 3 rk Vz;k G As dk l 2

(5)

where wind : the actual power output of wind power system annually Eactual;k (kWh/year) rk: air pressure (in units of Pascals or Newtons/m2) at location k. Vz,k: wind speed at z m height above ground at location k. l: the dimensionless Weibull shape parameter

AS: the Sweeping area of turbine blades, m2.

dk: Wind power system efficiency at location k. Due to seasonal changes of wind and solar energy density, hybrid solar-wind system is also popularly used as a clean power system to generate a more stabilized output throughout the year. The actual power output of a hybrid solar-wind system can be characterized as a combination of solar and wind energy system output with above equations (4) and (5), respectively, based on the proportion of solar and wind power rating in the hybrid system. Besides solar and wind, fuel cell stationary power system using hydrogen as fuel inputs is also widely recognized as a promising clean energy system due to its cleanliness in operations (only producing water emission) and also stable output (Sorensen, 2012). The actual output of a fuel cell stationary power system depends on the power rating of the system and the actual operating time. So the actual power output can be simply calculated as a product of the power rating and the actual operating time of the fuel cell system in a year. In this study, the above four types of clean energy systems are analyzed and benchmarked for their application potentials as a case study on GHG emission mitigation from GM’s global manufacturing facilities. 3. Case study Here details are presented on the case study for assessing the application potential of representative solar PV, wind, hybrid solarwind, and fuel cell stationary power systems for GHG mitigation from GM’s global automotive manufacturing facilities at the following six selected representative locations: Detroit (United States), Mexico City (Mexico), Sao Paulo (Brazil), Shanghai (China), Cairo (Egypt) and Bochum (Germany). With the case study, the application potential of each clean power system at each location is quantitatively determined and benchmarked, aiming to provide decision support in the early stage of strategic-planning and priority-setting for employing clean energy supply in automotive manufacturing and similar large-scale global productions. 3.1. Selection of clean energy power systems On the global commercial market, there are a variety of clean energy technologies available for power generations and supply. These different products have different technical specifications which can lead to different power output and different cost benefit in the same application scenario. In order to make the analysis results comprehensive and representative, in this case study a number of the most popular clean energy power systems are selected and their average technical specifications are used as representative specifications of current technologies on the commercial market. For this study, a total of five multi-crystalline silicon solar PV modules are selected, which are the top five modules in terms of both production volume and installed capacity in the world (Renewable Energy World, 2010). The selected five PV modules are Suntech STP210-18/Ud, Sharp ND-224uC1, Qcells Q.BPPARROSOE 225, YingLi 210 P-26b/1495x990 and Trina Solar TSM-PC05. For wind power system, four wind turbines at 1.5 Megawatt are selected, including GE 1.5XLE, Sinovel SL1500/77, Suzlon S82 1.5 MW and Nordex S77 1.5 MW. These four wind power systems have the largest installation capacity throughout the world (My Wind Power System, 2009). In current stage, fuel cells are mainly developed as mobile energy sources for transportation applications. For stationary power generations, there are only a few models available in the United States. In this analysis, the selected


15

fuel cell power system is Nedstack PS100 power system which uses hydrogen as fuel and is rated at 100 KW for each system. The economic cost data and LCA GHG emission factors of solar, wind and fuel cell power systems are collected from U.S. EIA and published literature, as shown in Table 1. The hybrid solar-wind power system is assessed with an equal share of solar and wind power rating in the hybrid system. The economic cost and LCA GHG emission factors of the hybrid solar-wind power system is calculated according to the solar and wind data corresponding to their proportions in the hybrid power system. The data demonstrates that among these four clean energy power systems, wind has the lowest overnight cost and lowest life cycle GHG emission factor, while solar PV and fuel cell power systems are much more expensive in the system deployment and have much higher GHG emission factors due to their complex system structure and deployment processes. 3.2. Feasibility analysis Using equation (1), the technical feasibility of each clean energy system is analyzed for the six selected global locations. Here the GHG emission factors of local conventional grid power systems in the six selected locations, as compiled by U.S. EIA, are 680 g CO2eq/ kWh for Detroit (United States), 594 g CO2eq/kWh for Mexico City (Mexico), 93 g CO2eq/kWh for Sao Paulo (Brazil), 845 g CO2eq/kWh for Shanghai (China), 437 g CO2eq/kWh for Cairo (Egypt) and 542 g CO2eq/kWh for Bochum (Germany) (U.S. EIA, 2007a; U.S.EIA, 2007b). Using the life cycle GHG emission factors in Table 1 and equation (1), the feasibility of each clean energy power system for GHG mitigation at each of these six selected locations are quantitatively investigated. The calculated results are shown in Fig. 2 below. Fig. 2 demonstrates that solar PV, wind, hybrid solar-wind, and hydrogen-based fuel cell technologies all have certain potentials for actual applications as clean energy supply for GHG mitigations at all these six locations. Among all these six locations, wind power supply has the highest potential in terms of absolute amount of GHG mitigation from grid power supply, when compared with the other three clean power systems. Solar PV and hydrogen-based fuel cell power systems are roughly at the same level because of the small differences of the life cycle GHG emission factors from the two power systems. The Fig. 2 results also demonstrate that, per

Table 1 Economic and environmental data of the representative clean power systems. Clean energy

Overnight cost Fixed O & M cost Recycling cost Life cycle GHG ($/kW) ($/kW) ($/kW) emission factor

Solar PV 4697a Wind 2409a Hybrid 3553f Fuel cell 10735g

25.73a 27.73a 26.73f 2147g

160.55b 399.00d 279.78f 373.24h

72.4c 10.84e 41.62f 83i

a Multicrystalline Si Solar PV, onshore wind and natural gas based fuel cells, costs data from reference (U.S. EIA, 2011a). b The recycling cost for solar PV is calculated based on (Fthenakis, 2000; McDonald and Pearce, 2010). c Multicrystalline Si Solar Cell, LCA result from reference (Azzopardi and Mutale, 2010). d The recycling cost data for Wind is from reference (Kabir et al., 2012). e 1.5 MW onshore wind Turbine, LCA result from reference (Pehnt, 2006). f Hybrid solar PV-wind system data are calculated based on 50%:50% composition from stand-alone solar PV and wind system. g The cost data for hydrogen PEM fuel cells, from reference (Gerboni et al., 2008). h The recycling cost data for fuel cell is calculated based on (James and Spisak, 2012). i Stationary Fuel Cell (hydrogen), LCA result from reference (Viebahn and Krewitt, 2003).

Fig. 2. Feasibility analyses of clean energy technologies at the six selected locations.

unit amount of electricity supply, wind power system in Shanghai (China) can reduce as high as 834 g CO2 eq per kwh of electricity consumed, due to the high GHG emission factor from grid electricity supply in Shanghai (845 g/kwh). The least mitigation potential from the feasibility analysis is obtained on using solar PV system in Sao Paulo, with only 20 g CO2 eq to be reduced per kwh of electricity consumption. But for the actual GHG mitigation effects, it also depends on the actual power system outputs which rely on the scale of the power plants and the local geographic conditions for solar PV and wind technologies, which will be demonstrated in the following cost benefit analysis.

3.3. Cost benefit analysis of clean energy power systems In this section, the cost benefits of each clean energy supply pattern at the six selected locations are assessed for quantifying and benchmarking the application potential of the four clean power systems to partially supply the power needs of GM’s production facilities at the selected locations. The amounts of the GHG reduction are quantified individually for the solar PV, wind, hybrid solar-wind, and hydrogen-based fuel cells at the six selected locations, with the same amount of economic investment. In this project, each clean power system is designed with a 3 GWh of power rating. In the calculations, the average solar insolation data, Iave,k, are the statistical data collected by NASA on each location during a 22-year time period (from July 1983 to June 2005) (NASA, 2008). The representative conversion efficiency of solar PV modules, em ¼ 14.79%, is the average of the five efficiency values of the selected solar PV modules as mentioned above. A representative surface area is also calculated as the average of the five surface areas of the selected solar PV modules, Am ¼ 1.578 m2. The conversion efficiency from DC to AC power supply is taken as the typical value of 77% (Marion et al., 2005). The wind speed data for the selected geographic locations are the statistical data collected by NASA during a 10-year time period (from July 1983 to June 1993) (NASA, 2005). The selected wind turbines have an average rotor diameter of 79.73 m and a swept area of 4992.05 m2, while the NASA wind speed data are collected only for 50 m height above the ground. Considering the differences of wind speed at different heights above ground, in this analysis the wind speeds at each location are converted between these two

16


different heights using the following equation (Simiu and Scanlan, 1978):

vz ¼ v0

k z z0

(6)

where vz: Wind speed at z m height above the ground v0: Wind speed at specified height of z0 Z0: specified height, (m) k: Hellman exponent. In equation (6), the Hellman exponent value is a key parameter. The k value depends on the location and the shape of the terrain on the ground and the stability of the air (Kaltschmitt et al., 2007). Since large scale industrial facilities such as GM’s automotive manufacturing facilities are usually located in or near city areas, in this analysis, the k ¼ 0.34 value is selected for the condition of neutral air above human inhibited areas (Kaltschmitt et al., 2007). The fuel cell stationary power systems are dependent on continuous supply of hydrogen fuels for power generations. In terms of economic cost, current hydrogen-based fuel cell power systems are much more expensive than other clean power systems. As demonstrated in Table 1, the overnight cost of hydrogen-based fuel cell power systems, per kw power rating, is more than four times that of wind power system, while the operating and maintenance costs are more than 85 times due to the high-cost processes associated with hydrogen production and storage using current technologies. Using the data in Table 1, the cost benefit analysis is conducted on these clean energy supply patterns for each location using equation (3). The calculated cost benefit results, in terms of GHG emission reductions per $1000 cost investment on these clean energy systems at the six selected locations are shown in Fig. 3. The analysis results demonstrate that among these four clean power systems; wind energy has the highest cost benefit performance at all the six selected geographical locations except in Mexico City. The reason is that Mexico City has the lowest wind energy density among all these six locations, only 76.5 W/m2 based on our calculations. In comparison, the wind energy density is 356.4 W/m2 in Detroit, 116.6 W/m2 in Sao Paul, 352.1 W/m2 in Shanghai, 213.0 W/ m2 in Cairo, and 591.8 W/m2 in Bochum, respectively. As calculated on the amount of CO2 reduction per unit amount of economic investment, Fig. 3 shows that the high cost benefits of

Fig. 3. Cost benefit results of GHG mitigation through clean energy supply.

wind power systems are obtained at Bochum, Shanghai, and Detroit, for a reduction of GHGs at 15.5, 14.5, and 11.8 tons CO2,eq per $1000 cost. In comparison, the cost benefits of solar PV and fuel cells power systems are only between 0 and 5 tons CO2,eq reduction per $1000 cost at the six selected locations with feasible application potential. The cost benefit performance of the hybrid solar-wind system is between stand-alone solar and wind power systems, with the highest mitigation potential in Shanghai at an amount of 9.5 tons CO2,eq mitigation per $1000 cost input. In this study, it is found that even though the costs associated with current hydrogen-based fuel cell power system are high, its average cost benefit performance is quite comparable to that of solar power system at the six selected locations. This is because hydrogen fuel cell power system only generates H2O as the byproduct, without any CO2 emission during the power generation process. In this analysis, the cost benefit performance of hydrogenbased fuel cells power system is a little bit higher than that of solar PV in Shanghai (5.04 vs. 4.56 tons/$1000), Bochum (3.13 vs. 1.94 tons/$1000), and Detroit (3.95 vs.3.23 tons/$1000). While in such locations as Cairo which has high solar energy intensity, solar PV has better cost benefit performance than hydrogen fuel cells power system (2.94 vs. 2.34 tons/$1000). Based on the statistical data, Cairo has an averaged solar insolation at 1929.23 kwh/m2/year (NASA, 2008), which is the highest among these six selected locations.

3.4. Strategic plan development and policy-making with the cost benefit results The above cost benefit results are calculated on the amount of GHGs which can be reduced using a clean power system at the six selected representative locations from GM’s global production sites. These results can be used effectively for strategic plan development and policy-making on GHG reductions by GM and similar global manufacturing companies based on its current production volume and investment plan at global scale. To support strategic plan development and policy-making, here we also calculated and benchmarked the economic costs of using such clean energy power systems for GHG mitigation on GM’s manufacturing facilities at the selected global locations. The cost is calculated for a strategic reduction of 5%w30% of GHG emissions from GM’s manufacturing facilities in the base year of 2009 which has a total amount of 6.75 million metric tons of GHG emissions (General Motors Company, 2010). The calculated results are shown in Fig. 4 below. The analysis results demonstrate that for reducing a fixed amount of CO2 emissions among the six selected locations, the lowest cost is with wind power system in Bochum. The economic costs of wind power supply for GHG reduction in Bochum is $21.8 million for 5% GHG reduction, and $131 million for 30% GHG reduction, followed by the economic costs of using wind power supply for GHG mitigations in Shanghai and Detroit at $23.3 and $28.7 million respectively for 5% GHG reduction from GM’s 2009 baseline global emissions. The analysis results reveal that the highest cost of clean energy supply for GHG reductions is in Sao Paulo because Sao Paulo has the lowest GHG emission factor from grid power supply (only 93 g CO2eq/kWh) among the six selected locations. In our analysis, the highest cost for reducing 5% of GHGs from the GM’s global facilities on the base year of 2009 is obtained at $5.11 Billion on hydrogen fuel cell power supply in Sao Paulo, which is about 102 times the lowest cost ($21.8 million) on wind power supply in Bochum. The second most expensive option is using solar PV for GHG reduction in Sao Paulo, at a cost of $2.36 billion for 5% GHG emission reduction from GM’s 2009 global facility emissions.


17

Fig. 4. Costs of clean energy supply for reducing 5%w30% of GM’s GHG emissions.

Overall, the economic costs of using wind power systems for reducing 5% of GHGs from GM’s 2009 global facility emissions range between $21.8 million in Bochum and $1.03 billion in Sao Paulo, while solar PV between $74.1 million in Shanghai and $2.36 billion in Sao Paulo, hybrid solar-wind power system between $35.5 million in Shanghai and $1.43 billion in Sao Paulo, hydrogen fuel cells between $67 million in Shanghai and $5.11 billion in Sao Paulo. These cost data can be appropriately used to develop GHG reduction strategies and mitigation goals based on the available budget and investment plan for optimal selection of an appropriate clean power system and location of a manufacturing facility by GM and similar global manufacturing companies. 3.5. Cost benefit trend of clean energy supply for GHG emission mitigation As the clean energy supply is targeting a future deployment, a trend analysis will be useful for decision support in the strategic planning of clean energy supply for GM’s automotive manufacturing and similar large-scale global manufacturing facilities. Here we present our analysis results on the future trend of cost benefits for solar PV, wind, hybrid solar-wind, and hydrogenbased fuel cell power systems till year 2035. The cost benefit projections were calculated in the virtual scenario of reference case following the cost changes of clean energy systems, as defined by U.S. EIA (U.S.EIA, 2011b). For the reference case, initial overnight costs for all technologies were updated to be consistent with costs estimates for 2010 (U.S.EIA, 2011b). Based on the scenario, a cost adjustment factor based on the projected producer price index for metals and metal products is then applied throughout the forecast, allowing the overnight costs to decrease in the future with the decreasing of this index or to increase in the future with the increasing of this index (U.S.EIA, 2011b). Besides the cost benefit results shown in Fig. 3 above for year 2010, the future cost benefit trends are calculated for the year 2020 and 2035 based on the U.S. EIA cost prediction data for solar PV and wind power systems (U.S.EIA, 2011b), while the cost data for the hybrid solar-wind power system are calculated based on the

proportion of the individual power system using the data in (U.S.EIA, 2011b). The cost data for the hydrogen fuel cell power system in the year 2020 and 2035 are converted from the projected cost data in (Gerboni et al., 2008). The calculated cost benefit trend results are shown in Fig. 5 for each location with different types of clean energy supply systems. Overall, the cost benefits of all these clean energy supply for GHG mitigation are increasing from 2010 to 2035. In terms of the amount of GHG mitigation through unit economic input (tons CO2,eq/$1000), wind technology is the best option at five selected locations, except in Mexico City where solar PV is the best option. In year 2020, the expected cost benefits of wind supply for GHG mitigation can reach 14.69 and 15.73 tons CO2,eq/$1000 in Shanghai and Bochum, respectively, while in 2035, the cost benefits can be expected to be increased to 18.64 and 19.96 tons CO2,eq/$1000, respectively. It is interesting that in Mexico city, the solar PV power system has the best cost benefit performance while the hydrogen fuel cells will catch up by year 2035. In Cairo, although the four clean power systems currently have different cost benefit performance, they will become quite similar in 2035 year. 4. Conclusions This paper presents a mathematical approach for cost benefit analysis of using clean energy systems to partially supply the electricity needs of global industrial productions in the case of GM to reduce GHG emissions. The cost benefit model is developed for quantifying the amount of GHG reduction per unit amount of economic input, by incorporating a list of relevant parameters to support multi-attribute decision-making for optimal selection and deployment of a clean power system in global scale. A case study is conducted on assessing and benchmarking the application potential of four clean power systems including solar PV, wind, hybrid solar-wind, and hydrogen fuel cell power systems at six selected representative locations from GM’s global production sites. The quantitative analysis results reveal that the cost benefit performance of clean power systems are dependent on multiple factors including time, scale and location of the power system deployment. In this study, the analysis is conducted on a

18


Fig. 5. Cost benefit trend of clean energy supply for GHG mitigations from 2010 to 2035.

rated 3 GWh power capacity for each clean power system, and the time is analyzed from 2010 into 2020 and 2035 year. Among the six selected locations, currently wind deployment at Bochum has the highest cost benefit performance, with 15.5 tons of CO2,eq to be reduced per $1000 cost input. The number is expected to be increased to 15.73 and 19.96 tons of CO2,eq per $1000 cost input in 2020 and 2035, respectively. For validation, we also calculated the GHG mitigation through wind power in Bochum using the economic cost and information provided in Sims et al. (2003), with values obtained between 10.6 and 17.7 tons of CO2,eq per $1000

input. Also, a report from McKinsey estimated that the average GHG mitigation is 40 Euros per ton of CO2 mitigation which corresponds to approximately 18 tons of CO2,eq per $1000 input (Enkvist et al., 2007). In this study, it is also noted that the life cycle GHG emissions from each clean power systems are considered by using the representative values from published literature and database (Table 1). These life cycle GHG values may have uncertainties from such attributional factors as time, technology, location and scale of a clean power system in actual deployment. These uncertainties


could be systematically analyzed in future through a detailed uncertainty analysis. Also, scale of deployment may have some effects on the cost benefits of the selected clean power systems, which could be analyzed through a sensitivity analysis in future. Acknowledgment The financial supports from General Motors company and the U.S. Department of Energy’s Industrial Assessment Center at University of Wisconsin Milwaukee (Award number: DE-EE0005537) are gratefully acknowledged. References U.S. EPA. Inventory of U.S. Greenhouse Gas Emissions and Sinks, 1990e2011. http:// www.epa.gov/climatechange/Downloads/ghgemissions/US-GHG-Inventory2013-Main-Text.pdf (accessed 11.22.13) U.S. EIA (Energy Information Administration), 2007a. Domestic Electricity Emission Factors, 1999e2002. http://www.eia.doe.gov/oiaf/1605/excel/electricity_ factors_99-02region.xls (accessed 09.10.10.). U.S. EIA (Energy Information Administration), 2007b. International Electricity Emission Factors by Country, 1999e2002. http://www.eia.doe.gov/oiaf/1605/ excel/electricity_factors_99-02country.xls (accessed 09.10.10.). U.S. EIA (Energy Information Administration), 2010. Cost and Performance Characteristics of New Central Station Electricity Generating Technologies. http:// www.eia.doe.gov/oiaf/aeo/excel/aeo2010%20tab8%202.xls (accessed 09.12.13.). US EIA (Energy Information Administration), 2011a. Levelized Cost of New Generation Resources in the Annual Energy Outlook 2011. http://205.254.135.24/oiaf/ aeo/pdf/2016levelized_costs_aeo2011.pdf (accessed 15.04.12.). U.S. EIA (Energy Information Administration), 2011b. Assumptions to the Annual Energy Outlook 2011. http://205.254.135.7/forecasts/aeo/assumptions/pdf/ 0554%282011%29.pdf (accessed 12.12.11.). Azzopardi, B., Mutale, J., 2010. Life cycle analysis for future photovoltaic systems using hybrid solar cells. Renew. Sustain. Energy Rev. 14, 1130e1134. Bilgen, S., Kaygusuz, K., Sari, A., 2004. Renewable energy for a clean and sustainable future. Energy Sources 26 (12), 1119e1129. Branker, K., Pathak, M.J.M., Pearce, J.M., 2011. A review of solar photovoltaic levelized cost of electricity. Renew. Sustain. Energy Rev. 15 (9), 4470e4482. Brown, M.A., 2001. Market failures and barriers as a basis for clean energy policies. Energy policy 29 (14), 1197e1207. Enkvist, P.A., Naucler, T., Rosander, J., 2007. A cost curve for greenhouse gas reduction. The McKinsey Quarterly, Online Journal of McKinsey & Co. http:// www.epa.gov/air/caaac/coaltech/2007_05_mckinsey.pdf (accessed 11.29.13.). European Wind Energy Association, 2009. Wind Energyethe Facts: a Guide to the Technology, Economics and Future of Wind Power. Routledge, New York: USA. Fthenakis, V.M., 2000. End-of-life management and recycling of PV modules. Energy Policy 28, 1051e1058. General Motors Company, 2008. Public Policy Center Environment and Energy. Voluntary reporting of General Motors corporation United States green house gas emissions for calendar year (1990e2007). http://www.gm.com/corporate/ responsibility/environment/reports/greenhouse_gas_emissions/ghgreport_ 2007.pdf (accessed 11.05.11.). General Motors Company, 2010. Environmental Commitment. http://www.gm.com/ corporate/responsibility/environment/facilities/index.jsp (accessed 11.02.10.). Gerboni, R., Pehnt, M., Viebahn, P., Lavagno, E., 2008. Final Report on Technical Data, Costs and Life Cycle Inventories of Fuel Cells. New Energy Externalities Developments for Sustainability (NEEDS). NEEDS Project, Rome, Italy. Available online. http://www.needsproject.org/RS1a/RS1a%20D9.2%20Final%20report% 20on%20fuel%20cells.pdf (accessed 15.09.11.). Gutowski, T., Dahmus, J., Thiriez, A., 2006. Electrical energy requirements for manufacturing processes. In: Proceedings of 13th CIRP International Conference of Life Cycle Engineering, Lueven, Belgium, May 31 e June 2, 2006, pp. 623e628.

19

Han, J., Mol, A.P.J., Lu, Y., Zhang, L., 2009. Onshore wind power development in China: challenges behind a successful story. Energy Policy 37 (8), 2941e2951. James, B.D., Spisak, A.B.. Mass Production cost Estimation of Direct H1 PEM Fuel cell Systems for Transportation Applications, 2012 Update”. http://www1.eere. energy.gov/hydrogenandfuelcells/pdfs/sa_fc_system_cost_analysis_2012.pdf (accessed 11.10.13.). Kabir, M.R., Rooke, B., Dassanayake, G.D.M., Fleck, B.A., 2012. Comparative life cycle energy, emission, and economic analysis of 100 kW nameplate wind power generation. Renew. Energy 37, 133e141. Kaltschmitt, M., Streicher, W., Wiese, A., 2007. Renewable Energy: Technology, Economics, and Environment. Springer, New York, p. 55. Koller, J., Koppel, J., Peters, W., 2006. Offshore Wind Energy: Research on Environmental Impacts. Springer, New York: USA. Marion, B., Anderberg, M., Gray-Hann, P., 2005. Recent Revisions to PVWATTS. NREL/CP-520-38975. In: 2005 DOE Solar Energy Technologies Program Review Meeting, November 7e10, 2005. Golden, CO: National Renewable Energy Laboratory, Denver, Colorado. McDonald, N.C., Pearce, J.M., 2010. Producer responsibility and recycling solar photovoltaic modules. Energy Policy 38, 7041e7047. MWPS(My Wind Power System), 2009. The 10 major Wind Power Companies in the World. http://www.mywindpowersystem.com/2009/04/the-10-major-windpower-companies-in-the-world/ (accessed 10.23.13.). NASA Atmospheric Science Data Center, Surface meteorology and Solar Energy, 2005. A renewable energy resource web site (Release 5 Data Set). http:// eosweb.larc.nasa.gov/sse/global/text/10yr_wspd50m. NASA, Atmospheric Science Data Center, Surface Meteorology and Solar Energy, 2008. A renewable energy resource web site (release 6.0), accessed on 10/12/ 2010. http://eosweb.larc.nasa.gov/cgi-bin/sse/sse.cgi?naþs01þs06#s01. Neuhoff, K., 2005. Large-scale deployment of renewables for electricity generation. Oxf. Rev. Econ. Policy 21 (1), 88e110. Pehnt, M., 2006. Dynamic life cycle assessment (LCA) of renewable energy technologies. Renew. Energy 31, 55e71. Renewable Energy World, 2010. Top 10: Ten Largest Solar PV Companies. http:// www.renewableenergyworld.com/rea/blog/post/2010/06/top-10-ten-largestsolar-pv-companies (accessed 10.23.13.). Simiu, E., Scanlan, R.H., 1978. Wind Effects on Structures. Wiley, New York, p. 47. Sims, R.E.H., Rogner, H.H., Gregory, K., 2003. Carbon emission and mitigation cost comparisons between fossil fuel, nuclear and renewable energy resources for electricity generation. Energy Policy 31, 1315e1326. Sorensen, B., 2012. Hydrogen and Fuel Cells: Emerging technologies and applications, second ed. Academic Press, New York: USA. IPCC (the Intergovernmental Panel on Climate Change), 2007. IPCC Fourth Assessment Report. Climate Change. http://www.ipcc.ch/publications_and_data/ar4/ wg3/en/contents.html (accessed 11.28.13.). Tsoutsos, T., Frantzeskaki, N., Gekas, V., 2005. Environmental impacts from the solar energy technologies. Energy Policy 33 (3), 289e296. Unander, F., Karbuz, S., Schipper, L., Khrushch, M., Ting, M., 1999. Manufacturing energy use in OECD countries: decomposition of long-term trends. Energy Policy 27 (13), 769e778. Viebahn, P., Krewitt, W., 2003. Environmental and Ecological Life Cycle Inventories for present and future Power Systems in Europe. Eclipse. http://www.ist-world. org/ProjectDetails.aspx?ProjectId¼f7de03bf412649a682c6e41085cdfb26 (accessed 03.20.13.). Yuan, C., Dornfeld, D., 2009. Reducing the environmental footprint and economic costs of automotive manufacturing through an alternative energy supply. Transaction NAMRI/SME 37, 427e434. Yuan, C., Zhang, T., Rangarajan, A., Dornfeld, D., Ziemba, B., Whitbeck, R., 2006. A decision-based analysis of compressed air usage patterns in automotive manufacturing. J. Manuf. Syst. 25 (4), 293e300. Zhai, Q., Cao, H., Zhao, X., Yuan, C., 2011. Cost benefit analysis of using clean energy supplies to reduce GHG emissions of global automotive manufacturing. Energies 4, 1478e1494. Zoulias, E.I., Lymberopoulos, N., 2007. Techno-economic analysis of the integration of hydrogen energy technologies in renewable energy-based stand-alone power systems. Renew. Energy 32 (4), 680e696.