based butanol as at - Wiley Online Library

1204

Biotechnol. Prog. 2008, 24, 12041214

REVIEW: BIOSEPARATIONS AND DOWNSTREAM PROCESSING Assessment of Potential Life-Cycle Energy and Greenhouse Gas Emission Effects from Using Corn-Based Butanol as a Transportation Fuel May Wu, Michael Wang, Jiahong Liu, and Hong Huo Center for Transportation Research, Energy System Division, Argonne National Laboratory, Argonne, IL 60439 DOI 10.1021/bp.71 Published online December 2, 2008 in Wiley InterScience (www.interscience.wiley.com).

Since advances in the ABE (acetone-butanol-ethanol) fermentation process in recent years have led to significant increases in its productivity and yields, the production of butanol and its use in motor vehicles have become an option worth evaluating. This study estimates the potential life-cycle energy and emission effects associated with using bio-butanol as a transportation fuel. It employs a well-to-wheels (WTW) analysis tool: the Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) model. The estimates of V life-cycle energy use and greenhouse gas (GHG) emissions are based on an Aspen Plus simulation for a corn-to-butanol production process, which describes grain processing, fermentation, and product separation. Bio-butanol-related WTW activities include corn farming, corn transportation, butanol production, butanol transportation, and vehicle operation. In this study, we also analyzed the bio-acetone that is coproduced with bio-butanol as an alternative to petroleum-based acetone. We then compared the results for bio-butanol with those of conventional gasoline. Our study shows that driving vehicles fueled with corn-based butanol produced by the current ABE fermentation process could result in substantial fossil energy savings (39%–56%) and avoid large percentage of the GHG emission burden, yielding a 32%–48% reduction relative to using conventional gasoline. On energy basis, a bushel of corn produces less liquid fuel from the ABE process than that from the corn ethanol dry mill process. The coproduction of a significant portion of acetone from the current ABE fermentation presents a challenge. A market analysis of acetone, as well as research and development on robust alternative technologies and processes that minimize acetone while increase the butanol yield, should be conducted. Keywords: butanol, ABE fermentation, well-to-wheels, GHG emissions, fossil energy R

Introduction The use of liquid fuels accounts for the single largest share of crude oil consumption in the United States. In 2006, the United States consumed more than 20 million barrels of crude oil per day; 66% of this total was used in the transportation sector. Motor vehicles alone consumed 140 billion gallons of gasoline and 50 billion gallons of diesel in 2006. The use of gasoline has increased as a result of the growth in light-duty vehicle (LDV) travel in the past 20 years. The Energy Information Administration (EIA) projected that transportation fuel use will continue to grow by up to 30% by 2030 (Conti, 2007). On the petroleum supply side, the United States relies heavily on foreign oil—13.7 million barrels per day, according to EIA (2007). The world’s most oil-rich region has become

Correspondence concerning this article should be addressed to M. Wu at [email protected]. C 2008 American Institute of Chemical Engineers V

extremely unstable, which heightens concerns about energy security. Furthermore, competition for petroleum oil has increased dramatically as a result of rapid economic growth in the developing countries. Finally, the exploration, production, and use of petroleum-based fuels generate greenhouse gas (GHG) emissions, which are the primary cause of global warming, as confirmed in a recent report prepared by the Intergovernmental Panel on Climate Change (IPCC, 2007). Faced with these energy and security challenges, the nation is exploring alternatives to reduce its reliance on fossil fuel. Finding a liquid transportation fuel that can be produced from domestic resources and is carbon neutral would allow the United States to reduce its dependence on foreign oil and decrease environmental burdens. Recently, in the Energy Independence and Security Act (EISA, 2007), the U.S. Congress set a renewable fuel standard of using 36 billion gallons by 2022. Following dramatic growth in the ethanol industry, corn ethanol (EtOH) production reached a

Biotechnol. Prog., 2008, Vol. 24, No. 6

1205

Table 1. Key Properties of Products from Corn-Butanol Productiona Product

LHV (Btu/gal)b

Density (g/gal)b

Acetone Butanol Ethanol Gasoline DDGS

83,127 99,837 76,330 116,090 8703 (Btu/lb)

2,964 3,065 2,988 2,819 –

a Sources: Acetone and butanol: Guibet (1997); ethanol and gasoline: GREET; DDGS (distiller’s dried grains with solubles): Morey et al., 2006a,b. b Multiply Btu/gal by 0.00028 to obtain MJ/L for LHV; divide g/gal by 3.785 to obtain g/L for density.

record of 6.5 billion gallons in 2007, displacing 3.3% of gasoline supplies (in Btu equivalent). Still, a large gap remains to be filled by biofuels. Developments in feedstocks, processing technologies, and new biofuels are urgently needed. Among potential biofuels, butanol (BuOH) produced from starch has gained visibility in recent years as a replacement for gasoline. Butanol has unique properties as a fuel. Its energy content—99,840 Btu per gallon (low heating value [LHV])—is 86% of the energy content of gasoline (on a volumetric basis) and 30% higher than the energy content of ethanol (Table 1). And its low water solubility could minimize the cosolvency concern associated with ethanol, consequently decreasing the tendency of microbial-induced corrosion to occur in pipelines and fuel tanks during its transportation and storage. Butanol is much less evaporative than gasoline or ethanol, causing it to generate fewer volatile organic compound (VOC) emissions and be safer to use. The majority of butanol used as a chemical is produced from petroleum propylene through the Oxo process (in which synthetic gas [syngas] is reacted with propylene), and its ultimate end use is for surface coatings. The most dominant bio-butanol production process has been acetone-butanol-ethanol (ABE) fermentation (Jones and Woods, 1986). ABE fermentation by Clostridium acetobutylicum was used to produce butanol during World War II. That process was phased out when more economical petrochemical processes emerged. Now, almost all the butanol in the world is produced from petrochemical feedstocks. Interest in developing viable ABE fermentation processes has been rekindled recently as a result of the pursuit of nonfossil-based feedstocks. In the past 20 years, research and development (R&D) efforts have focused on various aspects of the ABE fermentation process. Research in molecular biology resulted in major breakthroughs in strain/mutant development that dramatically improved microbial tolerance to butanol toxicity, which resulted in a significant increase in the ABE solvent production. Experimental and computational engineering efforts have included designing new fermentor configurations, improved downstream processing, and process integration. Huang et al. (2004) reported on an experimental process that uses continuous immobilized cultures of Clostridium tyrobutyricum and Clostridium acetobutylicum to maximize the production of hydrogen and butyric acid and convert butyric acid to butanol separately in two steps. This process reportedly produced butanol at a rate of 4.64 g/L of fermentation medium per hour (g/L h) and used 42% glucose; up to 25% glucose is used in the traditional ABE fermentation process by Clostridium acetobutylicum alone. In the early 1990s, Clostridium beijerinckii BA101 was developed by using chemical mutagenesis together with selective enrichment; it can produce twice as much butanol as its parent strain (U.S. Patent 6358717). Extensive studies have been performed to characterize this strain, develop an ABE

Figure 1. Schematic representation of WTW analysis system boundaries for butanol, ethanol, and gasoline flexible fuel vehicles (FFVs).

fermentation process with various feedstocks, and evaluate technologies for downstream product separation (Qureshi and Blaschek, 1999, 2001a,b; Parekh et al., 1999). Experimental and pilot-scale ABE fermentation processes by this organism resulted in up to 95.1% glucose utilization. Using in-situ gas stripping for solvent removal from the fermentor minimized product inhibition and enabled a higher feed concentration— up to 500 g/L of glucose (Ezeji et al., 2004). Solvent production in this process in a fed-batch mode reached 65%:35%:1.5% of butanol:acetone:ethanol by weight percent (or 43:23:1 by wt), which is a significant increase in butanol production from 6:3:1 (by wt) in the conventional ABE process. Recent studies have focused on the integration of fermentation and product removal through in-situ gas stripping and fermentation gas recirculation (Ezeji et al., 2004, 2005; Qureshi and Blaschek, 2001b). Liu (2003) and Liu et al. (2004, 2006) presented an optimized downstream processing scheme of acetone, butanol, and ethanol. The latest development includes a patent application (DuPont, 2007) describing a strain that produces single-product butanol from biological feedstocks. Cellulosic feedstock for butanol production has also been reported (Qureshi, 2007; Blaschek et al., 2008). (S&T)2 Consultants conducted a life-cycle energy and GHG assessment for the corn-to-butanol pathway on the basis of earlier work on conventional ABE fermentation processes with relatively low yield [(S&T)2 Consultants Inc. 2005 and 2007]. The study, which was done for Natural Resources Canada, relied on the GHGenius model to analyze butanol used as 10% in a gasoline blend to fuel LDVs. It found that life-cycle GHG emissions from Bu 10 were significantly higher than those from corn ethanol used as E10. Another widely used fuel life-cycle assessment model— the greenhouse gases, regulated emissions, and energy use in transportation (GREET) model—was developed in the 1990s at Argonne National Laboratory (Wang, 1995). GREET addresses the well-to-wheels (WTW) analytical challenges

1206


Table 2. Feed Input and Products Output for the Bio-Butanol Plant and Production Energy Share Parameter

Valuea

Corn grain feed rate (lb corn/h) Production (lb/h) Acetone Butanol Ethanol DDGS

237,338

Product Energy Allocation (%)

23,997 42,728 1,032 73,816

19 40 1 40

Table 3. Yields of Acetone, Butanol, and Ethanol from Bio-Butanol Plant Product

Yield (gal/bu corn)a

Energy (Btu/bu corn)b

Energy Output Share (%)

Acetone Butanol Ethanol Total

0.87 1.50 0.04 2.41

69,525 149,267 2,828 221,621

31.4 67.4 1.2 100.0

a Multiply gal/bu by 0.1487 to obtain L/kg. 0.04145 to obtain kJ/kg.

b

Multiply Btu/bu by

Source: Aspen Plus simulation (Wu et al., 2007). Divide lb/h by 2.2 to obtain kg/h.

a

Figure 2. Block diagram of corn-to-butanol production.

associated with transportation fuels and vehicle technologies (Figure 1). GREET estimates WTW energy (nonrenewable and renewable) use and GHG emissions from resource extraction to vehicle operation. It characterizes the production of various fuels (e.g., liquid, gaseous, and electricity) from a variety of energy feedstocks (e.g., coal, petroleum oil, natural gas, biomass, corn, soybean, and nuclear). To date, GREET can simulate more than 100 fuel production pathways and more than 80 vehicle/fuel systems. This study attempted to evaluate the potential of the recent ABE fermentation process, with the strain Clostridium beijerinckii BA101, from a life-cycle perspective. GREET was used to estimate the life-cycle energy use and GHG emission impacts of corn-based butanol (produced via the ABE fermentation process) when used to displace gasoline as a transportation fuel in LDVs. On the basis of the energy and mass V balance resulting from an Aspen Plus simulation of the advanced ABE production process (Wu et al., 2007), the study assessed the production and use of corn-derived bio-butanol. It also included a ‘‘cradle-to-user’’ analysis for bio-acetone (which is coproduced with bio-butanol) to address the impacts of displacing petroleum-based acetone with the bio-acetone coproduct. Merits and shortfalls of such process are discussed. R

Bio-Butanol Production ABE fermentation and downstream processing The bio-butanol plant produces butanol from corn via ABE fermentation by a hyper-butanol-producing strain (C. beijerinckii BA101), as reported by Qureshi and Blaschek (1999, 2001a,b). Corn is fed into a conventional corn dry mill retrofitted for butanol production. Corn grain is converted to glucose via liquefaction and saccharification under conditions similar to those in the corn dry mill process. Fermentation of glucose to acetone, butanol, and ethanol is carried out in a fed-batch and gas stripping system, as suggested by studies, to reduce butanol concentration and therefore its toxic effect in the fermentation stream and the inhibitory effect of a high concentration of glucose on the microorganisms (Ezeji et al., 2004, 2005). The acetone, butanol, and ethanol produced from ABE fermentation are removed from the fermentor by in-situ gas stripping; they

are further recovered through molecular sieve adsorption and three-stage distillation. Solids and biomass that are removed from grain processing and syrup from distillation undergo centrifugation and drying. Distiller’s dried grains with solubles (DDGS) generated from drying is used as animal feed. Acetone, DDGS, and ethanol are processed as coproducts. During bio-butanol production, several other chemicals are generated; these include fatty acids (butyric acid and acetic acid) and significant amount of hydrogen (H2) gas. This analysis assumed a corn-butanol plant with a production scale of 150,000 metric tons of butanol per year. Acetone is a major coproduct of the bio-butanol plant, with 82,000 metric tons produced per year. As seen in Table 2, the plant coproduces 0.56 pound acetone per pound of butanol. The bio-butanol plant also generates 253,600 metric tons of DDGS (11% moisture) per year. The yield of ethanol is relatively small (similar to the level of fatty acid mixtures) (Table 3). Ethanol yield from the ABE fermentation process accounts for 1.5% of total acetone, butanol, and ethanol by weight and only 1.2% by energy content (Btu/bushel of corn) (Table 3). H2 and carbon dioxide (CO2) were used internally for gas stripping and for maintaining anaerobic conditions for ABE fermentation. The fatty acids were not separated and purified, so they were not treated as coproducts. As illustrated in Figure 2, process steps—from grain receiving to saccharification for corn butanol—could be similar to those that are part of corn ethanol production. The major difference begins at the fermentation and subsequent downstream processing steps that remove products and separate acetone, butanol, and ethanol. A corn butanol production process model that was recently developed (Wu et al., 2007) describes the process with a hybrid approach. For the portion of the process upstream from ABE fermentation (i.e., grain cooking, liquefaction, and saccharification), it adopted a U.S. Department of Agriculture (USDA) corn dry mill Aspen Plus simulation model (Kwiatkowski et al., 2006; McAloon et al., 2004). For the ABE fermentation and downstream processing portion, it developed a separation model using ASPEN Plus. The two parts of the ABE process modeling were further integrated and scaled up to the desired feed inputs. Heat integration was considered in the integrated ASPEN model. Figure 3 depicts the ABE process simulated by the butanol ASPEN model. Compared with Figure 2, the simulated


1207

Figure 3. Block diagram of corn-to-butanol production process simulated by the corn butanol ASPEN model.

process has a concentrating step before ABE fermentation, and fermentation is followed by gas stripping, adsorption, and multistage distillation for product separation. The WTW assessment was based on the mass and energy balance from the corn butanol Aspen simulation model. Table 2 presents feed input and products output of the process. Table 3 shows the yield of acetone, butanol, and ethanol. A total of 2.4 gallons of solvents (acetone, butanol, and ethanol) could be produced from a bushel of corn, in which 67% of energy share by energy content is attributable to butanol (Table 3). Among the solvents produced, 1.5 gallons is butanol and 0.87 gallon is acetone. The process is able to yield 99.5% (wt %) pure butanol, 99.5% pure acetone, and 99.5% pure ethanol. Bio-butanol plant energy requirements The bio-butanol plant uses natural gas as the process fuel to generate steam and purchases electricity from the grid. Energy use for the butanol process was estimated on the basis of the corn butanol Aspen model, except for cooking heat use. There was a difference in the cooking heat requirement between the USDA dry mill model (McAloon et al., 2004) and an ethanol plant energy survey conducted by Mueller and Cuttica (2006). While the dry mill model provided a solid mass and energy balance for the conventional dry mill, Mueller and Cuttica (2006) offered results from current ethanol plants. We used the latter to reflect the status of dry mills. To estimate the amount of cooking heat required for corn butanol using Mueller and Cuttica’s data (corn ethanol), we converted the heat for corn ethanol to corn butanol via the amount of corn used for each of the process (i.e., 2.72 gallons of ethanol per bushel of corn and 1.50 gallons of butanol per bushel of corn). Table 4 summarizes process fuel requirements for corn butanol production. The demand for steam for processing is met by using a natural-gas-fired steam boiler with 80% efficiency to produce 150-psi (pounds per square inch), medium-pressure steam.

Bio-Butanol Life-Cycle Assessment Methodology and Assumptions System boundary Figure 1 depicts the GREET modeling boundary of the bio-butanol lifecycle framework. The life cycle of bio-butanol is divided into five stages: 1. Corn farming, 2. Corn transportation, 3. Bio-butanol production, 4. Bio-butanol transportation and distribution, and 5. Bio-butanol use in gasoline vehicles (GVs). The bio-butanol life cycle begins with the manufacture of fertilizer and farming machinery. Corn farming operations

Table 4. Natural Gas and Electricity Requirement for the Corn Butanol Production Process Process Steps Cooking Drying and RTOc ABE fermentation and gas stripping Distillation Total

Natural Gas (Btu/gal)a

Electricity (kWh/gal)b

18,216 19,703 77,195

1.23

115,114

1.76

0.53

a

Multiply Btu/gal by 0.00028 to obtain MJ/L. b Divide kWh/gal by 2.785 to obtain kWh/L. c RTO, regenerative thermal oxidizer.

include irrigation, tillage, chemical (fertilizers, lime, herbicides, and pesticides) application, and corn harvest. Harvested corn grain is transported via barge, rail, and truck to fuel production facilities, where it undergoes biochemical (BC) processing for fuel production. Liquid fuel is then transported to refueling stations via rail, barge, and truck. The corn-derived butanol is assumed to displace gasoline in GVs. The gasoline life cycle, in contrast, begins with the recovery of crude oil in oil fields and ends in gasoline combustion in GVs. Coproduct credit allocation methodology We partitioned total energy and GHG emissions into bio-butanol and coproducts by using two methods: (1) product displacement and (2) product energy allocation. Coproduct displacement (also termed system expansion) is based on the concept of displacing the existing product with the new product. In this case, bio-acetone is regarded as a renewable chemical. The energy consumed and emissions released during bio-acetone production displace the energy and emissions associated with petroleumbased acetone production. Energy and emission credits from the acetone displacement are assigned to bio-acetone. DDGS produced from a corn ethanol plant is typically used to displace corn, soy meal, and urea as an animal feed in the diet of certain livestock, such as beef and diary cattle and swine. DDGS that is generated from corn butanol production contains 26.7% protein by weight, which is close to the amount of protein contained in the DDGS from the corn ethanol process. Based on the assumption that the nutritional values of DDGS from the two processes are similar, the DDGS from the bio-butanol plant displaces traditional animal feed (soy protein and corn). In the product energy allocation method, energy use and GHG emissions burdens are allocated among products according to their energy output shares from the bio-butanol plant, which are determined according to the heating value of each product. Energy use and associated GHG emissions from upstream feedstock production, butanol production, and transportation activities are partitioned among acetone, butanol, ethanol, and DDGS based on their corresponding energy shares. This approach treats all energy products from the

1208

production process as equal, regardless of the differences in form and quality among them. It also implies that all four products are energy products. The energy allocation method is applicable to this case because out of the acetone, butanol, and ethanol products, both butanol and ethanol are liquid fuels. Although acetone is normally regarded as a chemical solvent and feedstock, its energy content (in LHV) is, in fact, between that of butanol and ethanol (Table 1). DDGS has a LHV of 8,703 Btu/lb and can be used as a solid fuel. Use of DDGS combustion or gasification to provide heat and power for ethanol plants has been explored (De Kam et al. 2007; Morey et al. 2006a,b). Although the fuel quality of DDGS is lower than that of the liquid fuels, considering the large quantity of DDGS (Table 2), its use for process heat provides energy savings while relaxing pressure on the already stagnant DDGS market. Analysis cases This study analyzes a facility that produces 150,000 metric tons of bio-butanol per year, requiring 33 million bushels of corn—equivalent to a corn ethanol dry mill producing 89 million gallons per year. Because current bio-butanol production technology is still in the early stages of R&D, we assumed a 2010 timeframe for a large-scale demonstration plant. We conducted a full life-cycle (or WTW) analysis for biobutanol for a case in which corn-based acetone could displace petroleum-based acetone and in which DDGS produced from the bio-butanol process could displace animal feed. Two bio-butanol cases were established by using product displacement and product energy allocation approaches for acetone and DDGS credit. We also considered the impact of overproducing acetone. In this case, an overabundance of acetone produced by a large-scale butanol industry would flood the acetone market and consequently lose its commercial value. This case is similar to the above product displacement case, except that corn-acetone is not credited by petroleum acetone. We conducted a separate analysis for petroleum acetone and bio-acetone. A ‘‘cradle-to-user’’ approach was elected because of the limited amount of data on acetone use available in the open literature. In the bio-acetone case, bio-butanol and DDGS are coproducts that displace petroleum butanol and animal feed, respectively. For comparison, corn ethanol life-cycle cases (with displacement and product energy allocation for coproduct DDGS) are presented. Finally, petroleum gasoline is used as the baseline case. Thus, we established a total of seven cases: • Case 1: Conventional gasoline (baseline fuel). • Case 2: Bio-butanol with natural gas as the process fuel, where bio-acetone, DDGS, and ethanol are credited on the basis of the product displacement method. • Case 3: Bio-butanol with natural gas as the process fuel, where bio-acetone, DDGS, and ethanol are credited on the basis of the energy allocation method. • Case 4: Bio-butanol with natural gas as the process fuel, where acetone is regarded as waste and therefore no acetone credit is assigned. DDGS and ethanol are credited on the basis of the product displacement method. • Case 5: Corn ethanol from dry milling with natural gas as the process fuel, where DDGS is credited on the basis of the product displacement method. • Case 6: Corn ethanol from dry milling with natural gas as the process fuel, where DDGS is credited on the basis of the energy allocation method.

Biotechnol. Prog., 2008, Vol. 24, No. 6 Table 5. Corn Butanol WTW Analysis: Corn Farming, Transportation of Corn and Butanol, and Vehicle Operation Parameters

Assumptions a

Corn yield (bu/harvested acre) Ratio of harvested acreage to planted acreage Fertilizer use (g/bu)b

158 0.9 N ¼ 420, P2O5 ¼ 149, K2O ¼ 174e 1,202

Lime use (g/bu) Corn transportation mode and distance (mi)c Truck (100%) Butanol transport and distance (mi) Barge (40%) Rail (40%) Truck (20%) On-road fuel economy of LDVs fueled with bio-butanol (mi/gal gasoline equivalent)d

50 520 800 110 24.8

a Multiply bu/acre by 62.8727 to obtain kg/hectare. b Multiply g/bu by 0.00004 to obtain g/g. c Multiply mi by 1.6 to obtain km. d Multiply mi/ gal by 0.4227 to obtain km/L. e N, nitrogen; P2O5, phosphorus fertilizer; K2O, potash fertilizer.

Table 6. U.S. Average Electricity Generation Mix Used in This Studya Source

Percent of Total

Residual oil Natural gas Coal Nuclear power Biomass Others

2.7 18.9 50.7 18.7 1.3 7.7

a

Source: GREET, year 2010.

• Case 7: Cradle-to-user bio-acetone analysis to examine the displacement of petroleum acetone by bio-acetone. Biobutanol, DDGS, and ethanol are credited on the basis of the energy allocation method.

Butanol life-cycle assumptions Bio-butanol in this study is produced from corn. It would be economical for the bio-butanol plant to be built near corn farms to reduce feedstock transportation costs. Operations associated with corn farming and feedstock transportation would be similar to those for corn ethanol. Similarly, transporting, blending, and distributing bio-butanol can be modeled after those same activities for ethanol. Table 5 presents key assumptions and GREET inputs for the bio-butanol lifecycle analysis. To clearly show the energy and emission effects of biobutanol versus conventional gasoline, we assumed in this study that bio-butanol is used in pure form in GVs and that butanol could achieve the same fuel economy per gallon of gasoline equivalent as gasoline-powered GVs (Table 5). WTW results are expressed in a per-million-Btu fuel matrix for comparison. In reality, bio-butanol may be used in various gasoline blends. The per-million-Btu base system is more appropriate from a fuel supply perspective. Because energy and emission results are presented in Btu or grams per million Btu of fuel produced and used, the effects of differences in fuel economy are removed. Electricity for the bio-butanol plant is supplied from the U.S. electric generation mix (Table 6).


1209

Figure 4. Pathways and their data sources for cradle-to-user assessment of fossil-based acetone and corn-based acetone.

Cradle-to-User Assessment of Petroleum Acetone Petroleum acetone production Acetone is primarily used as a chemical intermediate in the manufacture of acetone cyanohydrin for methyl methacrylate, biphenol A, and adol chemicals. Direct solvent use accounts for about 30% of world demand for acetone. Acetone is almost exclusively produced via cumene peroxidation, as a coproduct of phenol. A synthetic process based on hydrocarbons or derivatives now accounts for all acetone production in the United States and nearly all production in other countries. In this process, producing a pound of acetone requires 2.21 pounds of cumene (SRI Consulting 2005). Cumene is produced via alkylation of benzene with propylene under an elevated temperature and pressure in the presence of a catalyst. Upstream from cumene production, benzene is produced principally through catalytic reforming from naphtha in a BTX (benzene, toluene, and xylene) process. The other feedstock for cumene, propylene, is one of a half dozen products from crackers that receive raw hydrocarbon (naphtha) from oil refineries, as well as natural gas. Figure 4 presents a schematic of cradle-to-user petroleum acetone production pathway. It encompasses resource extraction, feedstock production, acetone production, and acetone transportation to the user site. In a parallel path, bio-acetone cradle-to-user encompasses feedstock farming and transportation, acetone production via the ABE fermentation process, and acetone transport to the user site. At each stage of the pathways, energy inputs and GHG emissions were estimated.

Assumptions and data sources The petroleum acetone production life-cycle data used in this study are from the life-cycle inventory (LCI) of EcoProfile (Boustead, 2005), which contains extensive industry data and detailed environmental data for various petrochemical processes. We used Eco-Profile LCI to characterize the petroleum acetone production process (Figure 4). Benzene and propylene (cumene feedstock) were assumed to be captive (i.e., the benzene and propylene are produced and consumed on site) because on average, 85% of total cumene in this region is produced from captive benzene and propylene, according to SRI Consulting (2005). Cumene is transported to acetone production facilities. Acetone is produced from cumene for captive use (i.e., petroleum acetone is produced and consumed on site in the same facility). The facility further converts acetone to acetone cyanohydrin, biphenol A, and adol chemicals. The GREET database energy values and associated GHG emissions were used for upstream fuel production (oil extraction, refining, and transportation) and transport of feedstocks.

We assumed that bio-acetone produced from the corn butanol plant would displace petroleum acetone produced in the region of Petroleum Administration for Defense District (PADD) II (EIA), which includes 14 Midwestern states and overlaps with the Corn Belt, where most corn-based fuel plants would operate. The cumene is purchased by and subsequently transported to the acetone producer. Thus, major assumptions for cradle-to-user analysis of petroleum acetone are listed below: • Benzene and propylene are captive in the refinery, so no transportation is associated with these feedstocks. • Cumene required for the acetone production is estimated on the basis of a weight ratio at 1:2.21 for acetone: cumene (SRI Consulting 2005). • The acetone production facility purchases 100% of feedstock cumene. • Cumene is transported to the acetone production facility an average distance of 323 miles in a 1,028-Btu/ton-mile (multiply Btu/ton-mile by 0.7268 to obtain KJ/metric ton-kilometer) Class 8 diesel truck. The hypothetical corn butanol plant is located in the Corn Belt to reduce corn transportation costs. We assumed that bio-acetone would usually be transported to acetone consumers via train (500 miles) and truck (100 miles) within PADD II (Figure 4). The GREET model is used for estimating bioacetone transportation to users. In the coproduct displacement approach (Case 2), petroleum acetone is displaced by corn-based acetone from cradle to user. The displacement is carried out in the following sequence: the cradle-to-gate results of petroleum acetone (as illustrated in Figure 4) were first subtracted from the WTW results of bio-butanol, then the energy or emission values associated with bio-acetone transportation from the bio-butanol plant to the acetone user were added.

Results and Discussion Energy consumption and GHG emissions Using bio-based feedstock for butanol results in a closed carbon cycle. On a life-cycle basis, the use of corn butanol could achieve substantial fossil energy savings (39%–56%) when compared with the use of gasoline (Cases 2 and 3, Figure 5). In both cases, bio-butanol production contributes to a positive energy balance (defined in terms of Btu of biofuel minus Btu of fossil fuel used to produce that amount of biofuel). When acetone is over produced, as in Case 4, in which acetone is regarded as a waste, bio-butanol production is not an option for fossil energy savings. The primary fuel consumed in the bio-butanol life cycle is natural gas; consumption of petroleum oil and coal follow

1210


Figure 5. Well-to-wheels fossil energy breakdown for bio-butanol and corn ethanol compared with gasoline, using different coproduct allocation methods. (Negative value indicates fossil energy avoided. Multiply by 1 to obtain J/MJ.)

(Figure 5). The amount of electricity required is relatively small (Table 5). Therefore, even though coal-based electricity constitutes 50% in U.S. average electricity generation mix (Table 6), only 9%–17% of total fossil energy use is from coal (Cases 2 and 3). Figure 6 shows the bio-butanol life-cycle fossil energy breakdown, which indicates that nearly three-fourths of the fossil fuel is used in the butanol production plant (73%). Corn cultivation accounts for 23% of the fossil energy (12% for agricultural chemical and fertilizer manufacturing and 11% for farming operations). In the butanol plant, the majority of natural gas demand comes from the process steam requirement (Table 4), which leads to a total demand of 115,000 Btu per gallon of butanol produced—approximately three times the amount demanded for corn ethanol production. GHG emission benefits are similar to energy benefits. The majority of CO2 reduction results from the displacement of fossil feedstocks with renewable feedstock. Vehicles fueled by bio-butanol achieve reductions of 32%–48% in GHG emissions relative to gasoline-fueled vehicles on a WTW basis in Cases 2 and 3 (Figure 7). For every million Btu of bio-butanol used in place of the energy-equivalent amount of gasoline, 32–48 kg of GHG emissions could be avoided. Of total GHG emissions in the bio-butanol WTW cycle, more than 58% by weight of the GHG emissions is from CO2, 38% is from nitrous oxide (N2O), and the remaining less than 4% is from methane (Case 2). Most of the N2O emissions are associated with fertilizer application at the corn farming stage.

Effect of acetone coproduct credit The energy partitioning for acetone has a significant effect on overall bio-butanol energy use and associated GHG emissions. Three cases—2, 3, and 4—present different options for acetone. When acetone credit is allocated on the basis of product energy content (Case 3), implying it is a fuel, the WTW analysis yields moderate fossil energy and GHG emission benefits for bio-butanol (Figures 5 and 7). When acetone is used as a renewable chemical to displace petroleum acetone via the displacement method (Case 2), the results of

Figure 6. Breakdown of fossil energy use in various stages of fuel life cycle for corn-based butanol (Case 3).

the analysis show increased savings in fossil fuel and decreased GHG emissions (Figure 7). Finally, treating acetone as a waste stream (Case 4) resulted in an increase of almost 370,000 Btu in fossil energy use and 34,000 g of GHG emissions (compared with gasoline), for each million Btu of bio-butanol produced. Acetone plays a critical role in the WTW of bio-butanol analysis because of several factors. First, petroleum acetone production is a fossil-energy-intensive process. Propylene and benzene—feedstocks for cumene (Figure 4)—are derived from naphtha and natural gas, which are 100% fossil feedstocks (SRI Consulting 2005). Another factor is the high yield of acetone from the ABE fermentation process: 0.56 pound of acetone per pound of butanol produced. These two factors suggest that when bio-acetone displaces petroleum acetone, a significant portion of fossil will be displaced by renewable source, which leads to a large coproduct credit in bio-butanol life cycle and eventually results in pronounced energy use and emission benefits (Case 2, Figure 5).


1211

Figure 7. Well-to-wheels GHG emissions of bio-butanol and ethanol compared with gasoline, using different coproduct allocation methods. (Multiply g/million Btu by 0.00095 to obtain g/MJ.)

Figure 8. Fossil energy use based on the production of bio-acetone compared with fossil energy use based on the production of petroleum acetone, case 7. (Multiply Btu/lb by 0.00232 to obtain MJ/kg.)

Cradle-to-user comparison of bio-acetone with petroleum acetone As seen in Case 7, using bio-acetone to displace petroleum acetone shows promising energy and environmental benefits (Figures 8 and 9) from a life-cycle perspective. By producing acetone from corn, manufacturers could save a significant amount of total fossil energy—up to 71% relative to petroleum-based acetone (which consists of 91% petroleum and 58% of natural gas savings, Figure 8). Figure 9 further indicates the cradle-to-user fossil energy distribution between the two. The majority of fossil energy use shifts from feedstocks stage (57%) in the petroleum acetone production process to acetone production stage (70%) in the corn acetone production process. Cradle-to-user bio-acetone results demonstrated substantial GHG benefits. Bio-acetone in this study is produced from corn—a renewable agricultural resource that absorbs atmospheric CO2 during its growth. The carbon in CO2 is con-

Figure 9. Cradle-to-user fossil energy consumption breakdown for (a) bio-acetone and (b) petroleum acetone.

1212


ment of bio-butanol through the ABE fermentation process could take limited feedstocks away from conventional corn ethanol production. R&D efforts in bio-butanol may need to be directed toward identifying new feedstocks, such as lignocellulosics, for ABE fermentation. In fact, Blaschek (2008) recently reported promising yields and toxicity tolerance of cellulosic bio-butanol production using Clostridium beijerinckii BA101. We anticipate an increase in R&D effort and a breakthrough in the production of this third-generation biofuel.

Uncertainties

Figure 10. GHG emissions generated or avoided by bio-acetone production compared with petroleum acetone production, case 7. (Multiply g/lb by 2.2 to obtain g/kg.)

verted to corn and is further transformed and embedded in acetone through the ABE fermentation process, which leads to a net carbon sequestration. As illustrated in Figure 10, nearly 300 g of CO2 equivalent per pound of acetone were sequestered from cradle-to-user. Nevertheless, several issues and concerns need to be addressed. In 2005, a total of 1.9 million metric tons of acetone was produced in the United States (SRI Consulting, 2005). PADD II, which overlaps with the U.S. Corn Belt, would most likely be the location for corn-based butanol plants. Production of petroleum acetone in this region in 2005 was 505,000 metric tons, representing 26% of total U.S. production. If bio-acetone was produced from a single bio-butanol plant located in this region, the volume of bioacetone would be about 16% of total petroleum acetone production in the region and 4% in the nation. Furthermore, assuming that total bio-butanol production levels could reach a quarter of the production level of corn ethanol (4.9 billion gallons in 2006), the amount of corn feed used for bio-butanol production could generate 676 million gallons of biobutanol and 391 million gallons (1.2 million metric tons) of bio-acetone. At this production scale, the acetone market would be flooded with both bio-acetone and petroleum acetone, dramatically decreasing the market value of acetone. Thus, one could argue that acetone should be treated as a waste stream (as we did in Case 4). Unfortunately, the outlook for replacing petro-chemicals with bio-chemicals is far from clear. Petroleum acetone is currently generated as a by-product during phenol production from cumene. On average, 0.61 pound of acetone is obtained per pound of phenol produced (SRI Consulting, 2005). The United States consumed 2.1 million metric tons of phenol in 2004 (SRI Consulting, 2007). Worldwide, the consumption of phenol grows at a rate of 4.1% per year. As demand for phenol continues to rise, petroleum acetone will be produced with or without the presence of bio-acetone. Feedstock for bio-butanol The feedstock supply for bio-butanol is another major concern. The ABE fermentation process shares the same starchbased feedstocks with conventional corn ethanol. Develop-

A parameter that has not been fully evaluated is DDGS from the corn-to-butanol ABE fermentation process. Its yield and quality were often not reported. Since DDGS accounts for 40% product/coproduct output by weight (Table 2), a change in the DDGS composition would likely result in a large variation in the displacement credit. Field tests of the value of DDGS from such a process are needed to refine the analysis. Another uncertainty is the gap between process simulation and large-scale production. This analysis was based on an Aspen simulation that built on laboratory pilotscale testing data. Process feasibility and scalability have not been addressed. Moving from lab to pilot testing to full-scale commercial production is complex; the fermentation performance, production yield, energy consumption, etc. can vary significantly. Furthermore, process integration has not been demonstrated in the pilot scale. These factors affect the overall bio-butanol production and energy use estimates. Sensitivity tests should be conducted with various input parameters for each life-cycle stage or production process step. Finally, data validation is an important part of the analysis. Ideally, we would like to rely on reproducible results and/or data verification from separate laboratories and pilot testing to improve the fidelity of the modeling and analysis.

Conclusions The use of corn-based butanol, produced by means of the current ABE fermentation process using Clostridium beijerinckii BA101, could result in substantial fossil energy savings relative to the use of conventional petroleum gasoline. It could avoid a large amount of GHG emission burdens. From a liquid fuel production standpoint, however, the ABE fermentation process examined might not be as effective as conventional corn ethanol production because it produces less liquid fuel (on an energy basis) per bushel of corn than does the corn ethanol process. The key challenge to bio-butanol production is bio-acetone that is coproduced in the ABE fermentation. Given a yield of 0.56 pound per pound of butanol produced, bio-acetone has significant effects on bio-butanol life-cycle energy use and GHG emissions. There is a need for R&D on robust, alternative technologies, and processes that minimize acetone and increase butanol yield; they could dramatically increase life-cycle energy and GHG emission benefits and improve petroleum oil displacement. New technologies that would significantly reduce the acetone that results from phenol production are needed. The impacts of bio-acetone on the acetone market under large-scale bio-butanol development should be carefully examined. In addition, new applications for acetone should be developed to prevent or minimize the devaluation of acetone.


1213

Acknowledgments

Literature Cited

This work was sponsored by the U.S. Department of Energy’s Office of FreedomCAR and Vehicle Technologies, which is part of the Office of Energy Efficiency and Renewable Energy. We would like to thank Professor Hans Blaschek of the University of Illinois at Urbana-Champaign and Dr. Nasib Quresh of the U.S. Department of Agriculture (USDA), Agricultural Research Service (ARS), National Center for Agricultural Utilization Research (NCAUR), for providing process data and insights on the ABE fermentation process with Clostridium beijerinckii BA101. We also thank Andrew McAloon of the USDA Economic Research Service (ERS) Eastern Regional Research Center (ERRC) for providing the corn-to-ethanol dry mill Aspen Plus model.

Blaschek H, et al. ‘‘Bio-Butanol from Lignocellulose,’’ presented at Biomass 2008 Conference, Alexandria, Va; 2008. Boustead I. Eco-Profiles of the European Plastic Industry: Acetone, Plastics Europe, March, Available at http://www.plasticseurope.org/Content/Default.asp?PageID¼392#, accessed Dec. 2006. Conti J. Annual Energy Outlook 2007, presented at Energy Information Administration (EIA) Energy Outlook, Modeling, and Data Conference, Washington, D.C., March; 2007. DuPont. Fermentive Production of Four Carbon Alcohols, World Patent Application WO/2007/041269; USPTO PCT/US2006/ 038007; 2007. EIA. Annual Energy Outlook 2007, DOE/EIA-0383(2007), U.S. Department of Energy, Available at http://www.eia.doe.gov/oiaf/ aeo/pdf/0383(2007).pdf, accessed 23 May 2007. EIA. PADD Definitions, http://tonto.eia.doe.gov/oog/info/twip/padddef.html, accessed 5 August 2008. Ezeji TC, et al. Acetone Butanol Ethanol (ABE) production from concentrated substrate: reduction in substrate inhibition by fedbatch technique and product inhibition by gas stripping. Appl Microbiol Biotechnol. 2004;63:653–658. Ezeji TC, et al. Improving performance of a gas-stripping-based recovery system to remove butanol from Clostridium beijerinckii Fermentation. Bioprocess Biosyst Eng. 2005;27:207–214. Guibet J-C. Carburants et Moteurs: Technologies, Energie, Environnement, Publication de l’Institut Français du Petrole; 1997. Huang WC, et al. Continuous Production of Butanol by Clostridium acetobutylicum immobilized in a fibrous bed bioreactor. Appl Biochem Biotechnol A: Enzyme Eng Biotechnol 2004;113–116:887– 898. IPCC (Intergovernmental Panel on Climate Change), Working Group I: The Physical Basis of Climate Change, IPCC WG1 AR4, Available at http://ipcc-wg1.ucar.edu/wg1/wg1-report.html, accessed 23 May 2007. Jones DT, Woods DR. Acetone-butanol fermentation revisited. Microbiol Rev. 1986;50:484–524. Kwiatkowski JR, et al. Modeling the process and costs of the production of fuel ethanol by the corn dry-grind process. Ind Crops Prod 2006;23:288–296. De Kam M, et al. ‘‘Integrating Biomass to Produce Heat and Power at Ethanol Plants,’’ American Society of Agricultural and Biological Engineers (ASABE) Presentation Paper 076232, Minneapolis, Minn., June; 2007. Liu J.Downstream Process Synthesis for Biochemical Production of Butanol, Ethanol, and Acetone from Grains, Ph.D Dissertation, Kansas State University, Manhattan, Kansas; 2003. Liu J, et al. Downstream process synthesis for biochemical production of butanol, ethanol, and acetone from Grains: generation of optimal and near optimal flowsheets with conventional operating units. Biotechnol Prog. 2004;20:1518–1527. Liu J, et al. Holistic approach to process retrofitting: illustration with downstream process for biochemical production of organics. Ind Eng Chem Res. 2006;45:4200–4207. McAloon AJ, et al. A model of the production of ethanol by the dry grind process, in Proceedings of the Corn Utilization and Technology Conference, Indianapolis, Ind., June; 2004. Morey V, et al. Characterization of feed streams and emissions from biomass gasification/combustion at fuel ethanol plants, American Society of Agricultural and Biological Engineers (ASABE) Presentation Paper 064180, Portland, Ore., July; 2006a. Morey V, et al. Biomass for electricity and process heat at ethanol plants. Appl Eng Agriculture 2006b;22:723–728. Mueller S, Cuttica J. Research Investigation for the Potential Use of Illinois Coal in Dry Mill Ethanol Plants, prepared by the Energy Resources Center at University of Illinois at Chicago for the Illinois Clean Coal Institute and the Illinois Department of Commerce and Economic Opportunity, Oct; 2006. Parekh M, et al. Pilot-scale production of butanol by Clostridium beijerinckii BA 101 using a low-cost fermentation medium based on corn steep water. Appl Microbiol Biotechnol 1999;51:152– 157.

Notation ABE BC BTX BuOH CEH CO2 DDGS EIA EISA EtOH FFV GHG GREET

= = = = = = = = = = = = =

GV H2 IPCC K2 O LCI LDV LHV N NG N2 O PADD P2O5 R&D RTO USDA VOC WTW

= = = = = = = = = = = = = = = = =

acetone, butanol, and ethanol (fermentation process) biochemical benzene, toluene, xylene butanol Chemical Economics Handbook carbon dioxide distiller’s dried grain with solubles energy information administration energy independence and security act ethanol flexible-fuel vehicle greenhouse gas greenhouse gases, regulated emissions, and energy use in transportation gasoline vehicle hydrogen intergovernmental panel on climate change potash fertilizer life-cycle inventory light-duty vehicle low heating value nitrogen natural gas nitrous oxide petroleum administration for defense districts phosphorus fertilizer research and development regenerative thermal oxidizer U.S. department of agriculture volatile organic compound well-to-wheels

Units of Measure Btu bu g gal h kg kJ km kW kWh L lb mi MJ psi wt %

= = = = = = = = = = = = = = = =

British thermal unit bushel(s) gram(s) gallon(s) hour(s) kilogram(s) kilojoule(s) kilometer(s) kilowatt(s) kilowatt-hour(s) liter(s) pound(s) mile(s) mega Joule pound(s) per square inch weight percent

1214 Qureshi N, Blaschek HP. Economics of butanol fermentation using hyper-butanol producing Clostridium beijerinckii BA 101. Trans Inst Chem Eng. C 1999;78:139–144. Qureshi N, Blaschek HP. ABE production from corn: a recent economic evaluation. J Ind Microbiol Biotechnol 2001a;27:292– 297. Qureshi N, Blaschek HP. Recovery of butanol from fermentation broth by gas stripping. Renewable Energy 2001b;22:557– 564. Qureshi N. Bioconversion of agricultural residues to butanol: simultaneous saccharification, fermentation, and product recovery. Presented at 29th Symposium on Biotechnology for Fuels and Chemicals, Denver, Colo; 2007. (S&T)2 Consultants Inc. Ethanol GHG Emissions using GHGenius: An Update, prepared for Natural Resources Canada, Ottawa, Ontario, Canada, March; 2005. (S&T)2 Consultants Inc. The Addition of Bio-Butanol to GHGenius and a Review of the GHG Emissions from Diesel Engines with

Biotechnol. Prog., 2008, Vol. 24, No. 6 Urea SCR, prepared for Natural Resource Canada, Ottawa, Ontario, Canada, Feb; 2007. SRI Consulting. Chemical Economics Handbook (CEH) Marketing Research Report, Menlo Park, Calif; 2005. SRI Consulting. Chemical Economics Handbook (CEH) Marketing Research Report, Menlo Park, Calif; 2007. Wang M. The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) Model, 1999, Version 1.5, Available at http://www.transportation.anl.gov/software/GREET/publications.html, accessed May 2008. Wu M, et al. Life-Cycle Assessment of Corn-Based Butanol as a Potential Transportation Fuel, ANL/ESD/07-10, Argonne National Laboratory, Argonne, Ill., Nov; 2007. Manuscript received Jan. 16, 2008, and revision received May 28, 2008. BTPR080021D