794 Biotechnol. Prog.

1999, 15, 794−803

Process Design and Costing of Bioethanol Technology: A Tool for
Determining the Status and Direction of Research and Development
Robert Wooley,* Mark Ruth, David Glassner, and John Sheehan
Biotechnology Center for Fuels and Chemicals, National Renewable Energy Laboratory, Golden, Colorado 80401

Bioethanol is a fuel-grade ethanol made from trees, grasses, and waste materials. It
represents a sustainable substitute for gasoline in today’s passenger cars. Modeling
and design of processes for making bioethanol are critical tools used in the U.S.
Department of Energy’s bioethanol research and development program. We use such
analysis to guide new directions for research and to help us understand the level at
which and the time when bioethanol will achieve commercial success. This paper
provides an update on our latest estimates for current and projected costs of bioethanol.
These estimates are the result of very sophisticated modeling and costing efforts
undertaken in the program over the past few years. Bioethanol could cost anywhere
from $1.16 to $1.44 per gallon, depending on the technology and the availability of
low cost feedstocks for conversion to ethanol. While this cost range opens the door to
fuel blending opportunities, in which ethanol can be used, for example, to improve
the octane rating of gasoline, it is not currently competitive with gasoline as a bulk
fuel. Research strategies and goals described in this paper have been translated into
cost savings for ethanol. Our analysis of these goals shows that the cost of ethanol
could drop by 40 cents per gallon over the next ten years by taking advantage of
exciting new tools in biotechnology that will improve yield and performance in the
conversion process.

Introduction competitiveness on the basis of skyrocketing projections
for future petroleum prices. Dire projections for oil prices
The U.S. Department of Energy’s Biofuels Pro-
and availability, in the context of the harrowing experi-
gram. For roughly 20 years, the U.S. Department of
ence of long lines at gas stations and fears of OPEC-
Energy (DOE) has funded research on the development
driven supply shortages, seemed reasonable in the late
of renewable, domestically produced fuels for transporta-
tion. The program is driven by a number of important 1970s (2). Today, we are faced with a very different
national issues: national security, economic competitive- reality: cheap oil. Petroleum prices are now at histori-
ness in the global market, rural economic development, cally low levels, and DOE projects only slight increases
climate change, air pollution, and others. For a more over the next 20 years (3, 4).
detailed discussion of these issues, see the related article Why We Conduct Rigorous Process Design Stud-
in this journal issue by Sheehan and Himmel (1). ies. The moving target for making cost competitive
The Office of Fuels Development at DOE is responsible bioethanol has forced the Biofuels Program to constantly
for managing the Biofuels Program, which specifically rethink its approach and to push the potential for
targets transportation fuels made from biomass. The technology improvements to its limit. With this increas-
Biofuels Program has two major components: a feedstock ing demand for cost competitiveness comes a demand for
development program led by Oak Ridge National Labora- greater reliability and credibility in predicting the cost
tory (ORNL) in Oak Ridge, Tennessee, and a conversion of bioethanol production. In today’s fuel market, every
technology development program led by the National penny in cost savings makes a difference. Thus, today
Renewable Energy Laboratory (NREL) in Golden, Colo- more than ever, sound process design, modeling, and cost
rado. analysis are essential to the success of our research.
Over the years, the program has worked on a varied Our process design studies serve three purposes. The
portfolio of fuel products. Today, the Biofuels Program first one we have already alluded to. For a given feedstock
focuses the bulk of its resources on the development of cost, we need to be able to predict the absolute cost of
fuel grade bioethanol made from the cellulose and hemi- bioethanol. We use this information to judge the potential
cellulose components of trees, grasses, and many waste for market penetration of bioethanol. The second purpose
materials. The abundance of these biopolymers in nature of such studies is to understand the economic impacts of
makes them important resources for the production of proposed research strategies and consequently to guide
bioethanol. process development. By translating research goals into
One simple, but difficult, hurdle faces bioethanol as a process performance parameters that can be reflected in
commercially viable fuel substitute-cost. When DOE first our process models, we can predict the impacts of these
began its research on bioethanol, it set its goals for cost goals on the bottom line-the cost savings per gallon of
bioethanol produced. Finally, we use current and pro-
* Corresponding Author. jected technology costs as inputs to policy discussions.
10.1021/bp990107u CCC: $15.00 © 1999 American Chemical Society and American Institute of Chemical Engineers
Published on Web 09/15/1999

Vol. has established partnerships with entrepreneurs actively . This elimi- nates inconsistencies in approach. No. While a complex process can certainly be mod. Merrick Engineering provided similar input on the design of the wastewater treatment Methods section (11). which already includes all of the management and convergence of hundreds of streams and blocks and frees up our engineers to concentrate on the process data and configuration.ott. but this does require that other users also have a license to ASPEN PLUS. while estimation software (13) and from cost databases main- these studies often answered specific questions posed at tained by Delta-T Corporation.gov/biofuels/process•engineering.. and it lends In the past two years. and 70 We describe the basis for each of these assumptions in control blocks. The Choice of Process Technology and Configu- Thus.html). these necessary. It offers a capability for carrying out thermo. they did not necessarily serve as consistent ICARUS is commonly used by engineering and con- navigational tools for the program. a spreadsheet of this magnitude distinguishing feature among these platforms is the would require the developer and all users to be far too approach used to accomplish the first step. As a result. firm with extensive experience in the design and con- struction of ethanol facilities. The model uses built in physical property the following sections. We recognize that there are tradeoffs between com. various times. As with ASPEN PLUS. Four Aspentech’s ASPEN PLUS process simulator (9) to model overarching assumptions provide a fundamental frame- bioethanol process technology. There are several processes possible for convert- in our analysis. today that perform these steps by different means. 1999. We elsewhere (12). NREL has worked hard to credibility to our cost estimates. converting involved with the finest details to ensure that it was used carbohydrate biopolymers to sugar. •Residue utilization avoiding the need to build this sophistication from There are a number of technology platforms available scratch. Purchased and installed costs for and spreadsheets.doe. struction firms in the chemical industry. Unfortunately. The biofuels program’s approach to process design process model by Delta-T Corporation. in Figure 1. The results presented in this paper reflect a detailed review of the Figure 1. any changes to process flow diagrams or process design assumptions are made in one model. The process model described in this •Dedicated ethanol/electricity production versus the paper consists of 144 unit operations.Biotechnol. The ASPEN PLUS process simulator also serves as a central repository for all inputs to the process. we limit ourselves to processes that involve four basic plexity and rigor and accessibility in the development of steps: our process models. particularly for highly specialized units. Prog. it addresses the need for both rigor and credibility ration. aspects of the process. 5 795 ing a complex spreadsheet comprehensible. We utilize outside engineering expertise to further improve the integrity of our modeling. •The choice of feedstock dynamically rigorous material and energy balances for •The size of the proposed plant complex processes. For the purposes of this paper. 15. (12) that is available on our laboratory’s website (http:// tency and rigor. The cornerstone of this effort is the use of Overall Basis for the Process Design Study. an engineering and costing using ASPEN PLUS. Over the years. whether they come from outside engineering firms used to provide design data on specific aspects of the process or they come from research results generated by the Biofuels Program. A comprehensive description of the process model and input from Delta-T Corporation is available How We Conduct Process Design Studies. data as well as property data developed at NREL (10). have a long history of using process design to set direction Process flow diagrams and ASPEN-generated material for our research (5-8). We have chosen a commercial •Conversion of carbohydrate biopolymers to sugar simulation package to model the complete biomass to •Fermentation of sugars to ethanol ethanol process because it builds much of the technical •Ethanol recovery sophistication into the basic architecture of the software. we have developed and energy balances are used to generate detailed cost estimates using a variety of in-house modeling tools equipment designs. This process simulator is work for our process design: widely accepted and used in the chemical and petroleum •The choice of process technology and configuration industries. All of the detailed upgrade its process analysis capability and to develop an cost data and assumptions are published in a report approach to process design that ensures greater consis. 57 components. process design and economics are an integral part input on current practices in the industry for many of the planning process. Thus. ing biomass to ethanol. The eled using a spreadsheet. This approach greatly speeds our model development and simplifies our ability to transfer the model to others. we chose to use the ASPEN PLUS simulator. We have also relied on engineering equipment are obtained from vendor quotes. This approach is outlined schematically www. We various studies were often hard to relate to each other. wherever firms to develop cost estimates. So rather than spend considerable effort mak. They provided invaluable Thus. 668 streams (462 biorefinery concept material and 206 heat or work). The Biofuels Program correctly. obtained cost data on standard pieces of equipment (such differing as they did in assumptions and approaches and as pumps and heat exchangers) from the ICARUS cost the degree of explicit documentation.

No.0 able sugars from biomass.5 are based on concentrated acid hydrolysis (14) and two. Their efforts will lead to the establishment Poplar) of “pioneer plants” that will demonstrate the commercial component hardwood yellow poplar corn stover viability of producing ethanol from cellulose and/or cellulose 42. research progress. while some equipment (like the production fermen- process analysis is typical yellow poplar (20).S. 15.. affect the nature of the feedstock handling field of biotechnology. We can (and have) disaggre- available estimates for performance of an enzyme-based gated feedstock and conversion costs to allow more process. This may mean that pretreatment of stover (the residue left in the field after harvest) and stover could be easier and could result in a more readily switchgrass. This design case reflects the current status ethanol based on biomass supply cost curves. we envision the is less than 1. acetatea 4. established inorganic acid catalysts.9 0. and agricultural residues (such as bagasse. These costs can range from $15 to $40 per U. For a plant use in our ongoing integrated process development work. niche opportunities. The point at which this diminishes is For purposes of convenience and continuity.0 the isolation of genes expressing enzyme components a Acetate is the acetate groups present in the hemicellulose from different microbes for the purpose of producing artificial enzyme consortia that exhibit better synergism polymer. we expect to be work. to linear scaling. 5 engaged in exploiting some of these technologies in niche Table 1. Corn stover is one of the largest sources of activity could lead to higher prices for the feedstock. The ethanol cost results design. Plant size is important in the design reflect current laboratory and pilot plant results overall cost of production. They are generally converted to acetic acid in the prehydrolysis reactor. rice which allows for economies-of-scale when the exponent straw. arabinan 0. 16) to produce ferment. More recently we have known components is available. there are still economies-of-scale available in the ASPEN PLUS model. in lieu of more feedstocks to accommodate differences in composition. for ex- offered by the explosion in innovations occurring in the ample. By virtue of its “less Composition is not the only factor that will affect our trodden” nature and the potential for improvements process design. systems. Collection distance for a plant is highly site specific.50 per MT). In the long run. Vol. The com. fuel market of our research efforts. Furthermore. mannan 2.6 16.7. It is a benchmark for evaluating demand data. as well (21). that lignin and acetate levels in stover are plant size on collection distance for two feedstocks: corn significantly lower. depending on the level lowest potential cost and it is our first priority for process of demand and type of material. since the late 1980s. turned our attention to protein engineering of the indi- vidual enzyme components (18). we have when the equipment is as large as it can be built and selected yellow poplar sawdust as a “model” feedstock for multiple pieces of equipment are necessary. the more feedstock a plant demands. An exponent of 1 would translate grass.0 1. While it is our goal to model all of the technology feedstock supply cost curves for a number of feedstocks platforms supported by the program. focused on the develop. New Size* exp The Choice of Feedstock. as it would look if it were to be designed and sophisticated analysis of market penetration for bio- built soon.50 to $44 per MT). Corn stover is very similar in feedstock collection costs. threshold of cost competitiveness. Oak Ridge National Laboratory has developed (19). waste.5 0. It is worth noting. and conversion costs. The downside to this 2000 ton per day plant is reasonable for both corn stover lower lignin content is that less fuel is available for steam (which has a collection radius of 23 miles) and switch- and electricity production in the plant. These include softwoods collected as part * size could be a characteristic linearly related to the size.8 cellulase enzyme systems. and corn stover).4 40. the farther out it ing with industrial partners who are focused on specific must go to get it. These companies are using processes that xylan 18.9 hemicellulose. The biggest impact. presented here use a fixed feedstock cost of $25 per U. The results of the analysis suggest that a digested feedstock for enzymes. analysis to date (22). A variety of feedstocks can be used to produce bioethanol.0 We have. Generally. Our industrial partners New Cost ) Original Cost (Original Size*) are looking to waste materials that offer low-cost. . enzyme technology.7 ment of a new generation of genetically engineered ash 1.6 1. we find A typical composition for yellow poplar used for this that. Prog.S. b No quantification of the remaining un- than natural enzyme systems (17). in many parts of the process.9 lignin 26.796 Biotechnol. stock that we believe will be of commercial interest to we have done preliminary analyses to get a better future partners. Still. we include data on its composition understanding of this tradeoff between plant size and in Table 1. tors and the pretreatment equipment) has reached their position is translated into components that are tracked maximum size. Thus. capacity of 2000 tons per day of dry feedstock. municipal solid The size factor is adjusted using an exponential term. Put quite simply. ton shows that the enzymatic conversion technology has the (dry) (from $17. These enzymes can be our analysis of it) will have to be adapted to specific used to release sugars from cellulose. nent are normalized to 100% as shown in Table 1. Performance parameters used in the The Size of the Plant. The important assumptions for composition to our model feedstock. Even Savings due to economies-of-scale are offset by in- before the economics of the enzyme process have met the creased costs for feedstock collection. Feedstock Composition (Hardwood Yellow applications.1 21. however. in our view. is in feedstock offers the greatest opportunity for future cost reduction cost.2 stage dilute acid processes (15. this analysis are shown in Figure 2 shows the effect of however. This work initially involved otherb 3. of forest health management activities. increased collection feedstocks. available biomass for ethanol production and is a feed.0 6. exponent of about 0. Differences in feedstocks will. The process (and grass (which has a collection radius of 40 miles). galactan 0.9 11. 1999. The values for each compo. equipment costs scale with an use of energy crops such as hybrid poplar and switch. The process design presented here reflects the best ton (dry) ($27. We use the following scaling and sometimes expected laboratory results over the equation to adjust equipment costs for different sizes: course of the next year or two.

This accomplishes two things: the hydroly- sis of most of the hemicellulose to its constituent sugars and a loosening of the lignin-hemicellulose-cellulose complex. The plant University of Arkansas introduced the important concept is designed to accommodate 136 trucks per day. remaining water.15b percent of surrounding acres 50% 10% participating in collection percent of surrounding acreage 75% 75% available for crop production a Corn stover yield based on reported yields from an ongoing Figure 3. components. In this case. discussed briefly in the following sections. The design also reflects costs coproducts. for the purposes cellulose sugars. which allows a little design opens up a myriad of possible products and process over two weeks per year for downtime. 5 797 Table 2. A vapor support the bulk of a large-scale plant’s output by making phase molecular sieve system is used to remove the fuels. distillation column are dewatered and sent on to residue diates. No. Coproducts are critical to carrying the bottom line residue. 15. Prog. we can use our analysis antees and other costs related to reducing the risk of to identify key technical issues for the core technology introducing new technology. we prefer to leave these choices up to the and performance associated with an “nth” plant. Gulf Oil Company and the Area 100: Feed Handling and Storage. been improved through the use of organ- isms capable of fermenting the glucose from cellulose and the hemicellulosic sugars together.67a 6. Vol. A small portion of the biomass slurry is diverted to the This assumption is not only unrealistic but also seriously cellulase production fermentors. 1999. Readers sumed that the enzymatic hydrolysis of cellulose and the interested in a more detailed discussion of the design can fermentation of glucose would be carried out in two request copies of our published report (12). Major Assumptions for Feedstock Collection Analysis assumption corn stover switchgrass yield (MT biomass per acre) 1. 27). b Switchgrass yield based on an analysis of the ORNL ORECCL county level ars are continuously produced by the enzymes and database for energy crops (24). Petroleum refiners. Similarly. more recently. Thus. Overview of the SSCF Process. ment step is a reactor similar to pulp digesters that Estimates of the incremental transportation cost (not permits co-current contacting of biomass and acid at presented in detail here) of going from 1000 to 2000 MT elevated temperatures. Even separated from the liquid hydrolysate. The unreacted solid phase is per day are less than $2 per MT of biomass (12. wet mill corn ethanol plants typically processing. each of simultaneous saccharification and fermentation (SSF) carrying around 47 MT of wet material. and washers designed to handle an average of 176 MT . The SSF process has. but it enhances its profitability through the pro. the lignin is combusted to view ethanol as among the lowest-value products leaving produce steam for use in the plant. It is the basis for our process design. Details of the Design Basis for a Near-Term The plant design is broken down into nine process Bioethanol Plant Design. The lessons of today’s petroleum stream from the SSCF reactors contains a relatively low refineries and today’s wet mill corn ethanol plants are concentration of ethanol as well as unreacted lignin clear. Process flow diagram for the enzymatic process corn stover collection demonstration project. solid biomass can be recombined and sent to subsequent Dedicated Ethanol/Electricity Production Versus biological processing steps. This yield allows for evaluated in this study. solids conveyor systems. The handling (26. The latter renders the remaining cellulose more Figure 2. for example. may ethanol concentration up to the azeotrope level. introducing coproducts into our We assume an on-line time of 96%. A traditional two-column distillation brings the for these facilities. which contains the 4000 MT per day plant size may be workable. Monomeric sug. After size reduction. Highlights of the design issues in each area are for using enzymes in the production of bioethanol as. Rather than tying our analysis to specific from Delta-T Corporation. while the bulk of the limiting in the picture it paints for economic viability of material is sent on to the SSCF vessels. Solids from the bottoms of the first duction and sale of higher-value petrochemical interme. This new variant of SSF. and fermenting microbes are combined. An implicit assumption we stock and microorganisms allow elimination of these make in our design studies is that we limit the range of “conditioning” steps. cellulase enzyme area includes weigh stations. The heart of the pretreat- distance. the plant gate. An early proposed process areas. products made at the facility to ethanol and electricity. Later. we chose to stay with the 2000 MT per day using “overliming” and ion exchange before the liquid and capacity used in many of our previous studies. The product bioethanol technology.. Some combinations of feed- the Biorefinery Concept. In other future entrepreneurs who will make this technology a words. converted by fermentative organisms to ethanol avoiding product inhibition of the enzymes. known as SSCF for Simultaneous Saccharification and CoFermentation. The effect of plant size on feedstock collection amenable to enzymatic attack. we do not account for cost of engineering guar- commercial reality. is shown schematically in Figure 3. quantities left behind to control soil erosion (23). In that way. However. as long acetic acid and other inhibitors in addition to the hemi- as economies-of-scale still apply. sequential steps (25). based on input scenarios. 23). The acid and inhibitors are removed of this study. the biomass is pretreated using dilute acid at high temperature. In the SSF process scheme.Biotechnol.

In addition. production vessels. Material exiting the pretreatment reactor is flash. by the Tennessee Valley Authority (28). This optimization included mass-transfer calcu- acetic acid losses in the vapor phase due to dimerization. These design inputs are based on experiments con. The hydrolyzate is then adjusted to pH 4. Conditions and yields for the pretreatment reactor are shown in Table an adequate volume is built up to feed the three continu- 3. The large vessels are costed as exist for an “nth” plant (29).5 for the whole area more practical and representative of best fermentation step and filtered to remove gypsum. Key Design Inputs for Pretreatment Reactor Table 4. in lieu of sending this water to the Area 300: Simultaneous Saccharification and waste treatment system.. due to the lack of eleven 1000000-L (264000-gal) aerated fermentors oper- experience in fabricating these units for our operating ated in batch mode (eight are operating at all times). ous trains of 3600000-L (950000-gal) production vessels. a known Area 500: Product Recovery and Water Recycle. design is based on bench scale performance data (32) for Our design assumes that the reactor is built from NREL’s genetically engineered Zymomonas mobilis (33). Trichoderma reesei. Design inputs for this part of the process are sum- cooled from 12. They are not pres- Delta-T Corporation. Its design was productivity 75 FPU/liter‚h done with input from a manufacturer of digester equip. Though this addition to the CoFermentation. tal cost of this area. Specifications for the production-scale reactor. Delta-T recommended accounting for reactions that may occur in the overliming the use of a water scrubber to recover ethanol from all step is done in the process model. They are not pressure-rated for may yield similar or superior results. seed trains with five stages for building up the inoculum Merrick’s analysis demonstrated that the savings in to the production fermentors. A requirement for telloy C reactor that is 50% higher than the cost for the steam sterilization would significantly increase the capi- 304 stainless reactor.25% HMF yield from mannan and galactan 15% theoretical yield of ethanol from glucose 92% theoretical yield of ethanol from xylose 85% per hour (wet basis). The continuous ion Distillation and molecular sieve column designs in this exchange system removes 88% of the acetic acid and study represent a significant improvement over previous 100% of the sulfuric acid in the hydrolyzate and mini. Prog. . design studies. Modeling of this flash vaporization in optimization was done to determine the minimum cost ASPEN PLUS using an equation of state specific for when vessel aspect ratio and air flow rate were allowed acetic acid allowed us to more accurately account for to vary. This cost differential sized to provide 5% inoculum to each subsequent fer- most likely reflects a high degree of risk that would not mentor in the train.5% residence time 10 min temperature 30 °C temperature 190 °C initial solids level (soluble and insoluble) 20% solids in reactor 22% residence time 7 days hemicellulose sugar yields 75% cellulase level 15 FPU/g of cellulose furfural yield from xylan and arabinan 10% corn steep liquor 0. syrup that is burned. Overliming requires ASPEN PLUS for the rigorous design of the distillation reducing the pH of the ion exchanged material to pH 2 columns. lations constrained to provide a minimum oxygen- A continuous belt filter is used to separate the liquid transfer rate in the vessels of 80 mmol/L‚h. inhibitor in the fermentation (30). Hastelloy C. cellulase requirement for SSCF 15 FPU/g of cellulose Area 200: Pretreatment and Hydrolyzate Condi. we settled on a cost for the Has. showed a Three trains of three seed fermentors each are used to factor of 3 or more difference in cost between stainless provide inoculum to the production vessels. This flash step removes much of the furfural and data for the production of enzyme (34) by one of the most hydroxymethyl furfural (HMF) and a portion of the acetic commonly used types of industrial fungus for cellulase manufacture. hydrolyzate prior to removal of acetic acid.2 to 1 atm in a vessel made of Carpenter marized in Table 5. The fermentation step includes two process design was not an obvious way to reduce cost. On the basis of input from stirred 304 stainless steel reactors. conditions. This area includes quotes for this unit are not available. We assume that steam sterilization. Key Design Inputs for SSCF Production Fermentors (32) acid concentration 0. A 10% volume of inoculum wastewater treatment system costs outweighed the added is added to subsequent fermentors in each stage until cost of the evaporator (11). many practical tips from Delta-T with sulfuric acid. The pretreatment reactor is one of the most cellulose + xylose) unique pieces of equipment in this plant. For example. 1999. No industry practices. followed by addition of lime to a pH Corporation contributed to making the design of this of 10. The model assumes acid. Merrick Engineering’s recommendations for waste this step to better understand what is happening (31). The cost of the reactor turns the production fermentors are shown in Table 4. 15. The evaporator produces a efficient alternatives. Vol. The vessels are costed as stirred 304 term improvements (20). particularly given the ability to use mizes dilution with no loss of sugars. A treatment led to the addition of a multiple effect evapo- better understanding may allow us to identify more rator to treat stillage water. Reliable vendor Area 400: Cellulase Production. based on corrosion experiments conducted which is capable of fermenting glucose and xylose only. sure-rated for steam sterilization. initial cellulose concentration 4% ment for the pulp and paper industry. for example. for a total of 18 Sunds reactor with adjustments made for expected near. A preliminary quote. The design is based on bench scale 20. No. yield of enzyme 200 FPU/(g of tioning. Other conditions or equipment stainless steel reactors. 5 Table 3. Bulldozers maintain 40-foot piles Table 5: Key Design Inputs for Cellulase Production in an area designed to hold 7 days of material. Removal of these heterocyclic aldehydes is beneficial that 47% of the soluble sugars go to cell mass production since these compounds can be detrimental in the fer- and 53% go to enzyme production. ducted on yellow poplar hardwood in NREL’s pilot scale There are six vessels in each train. We are still studying vents. An iterative cost mentation step. The out to be one of the biggest uncertainties in the process. They are steel and Hastelloy C construction. This design is consistent with current the results of the Sunds pilot unit will scale directly to practice in the corn ethanol industry.798 Biotechnol.

this design is more capital. transportation and commercial investors for is $234 million. Steam is pulled off the turbine at 12. this amounts to an annual production capacity that an after tax discount rate of 10%. For this level of accuracy.0 Area 700: Product and Chemical Storage. the net present value of the project is zero. Chemical storage ranges from four to 7 days. insurance.2 quirements for this part of the plant were based on diesel 0. Thus.2 million Btus of a medium-Btu methane/CO2 fuel that is burned for Table 7.2 Area 800: Burner. The price of equipment costs. b Megawatts (10. The discounted cash installation factors from Walas (35). site preparation. on the basis of arguments presented applied for warehouse. Breakdown of installed equipment cost for near-term includes an equalization basin that feeds into an anaero. of which 32 MW is used in the process and Operating Costs.” be used. The overall yield of ethanol from biomass investments with risks similar to those of conservation is 68 gallons per dry U.59 1. Figure 4 shows the percentage breakdown of installed “Best of Industry” Near-Term Case. from the multiple-effect evaporator. Once the capital and operating accurate to around (25%. Table 8 summarizes the economics of a near-term intensive than today’s corn ethanol facilities.44 per gallon. They are based on cooling water. plant and instrument air. nutrients 3. and a concentrated syrup grid at 4 cents per KWh. which typically sells for around our design. and other project. No.600.5 streams: residual lignin solids.000 site development $6. $4. At assume a 20-year plant life and 100% equity financing. This is not surprising.000 home office & construction fee $35.000 warehouse $2. Table 6 summarizes the capital require.. Total project absence of statistical data on discount rates used by investment for the near-term enzyme-based plant design industrial.0 10%. His view is that “In the ments for a 2000 MT per day facility. ethanol has a minimum selling price of capacity (36).500. standard rules of thumb.48 per gallon of capacity.2 million gallons) of ethanol.700. 5 799 Table 6.48 61. This area makeup water 0. a grassroots facility based on the Biological processing (enzyme production and SSCF) is enzyme process as tested in our lab is approaching (but . 1999. Area 900: Utilities. Ninety percent of the organic loading in Process ($1997) the wastewater is removed in the anaerobic digester.000.S.. Boiler.48 0. medium-Btu methane gas a $25 per MT. as we have discussed.20 in today’s fuel market.45 0. of electricity. as well as solid waste The Cost of Ethanol for the Near-Term Enzyme. disposal costs and a credit for excess electricity sold to Based Plant Design. Operating Costs for a Near-Term Enzyme-Based process energy. We assume a ence in constructing ethanol facilities. Variable costs include Results feedstock and other raw materials. Other factors were discount rate of 10%.31 37. ton. The table presents total annual cost and costs per gallon of ethanol.22 6. Re.9 recommendations from Delta-T Corporation. An item MM $/yr cents/gal ethanol aerobic digestor is used to remove most of the remaining biomass feedstock 19. bic digestion system. Vol.1 atm (510 °C) steam that is fed to a cogenerating turbine. Fixed and variable costs are sum- 6 MW is sold to the grid.44 per gallon for bioethanol is.942) of excess electricity sold to from the anaerobic digestors.68 -7. or nearly so. we costs have been established. representing atm for use in the process. Delta-T Corporation flow analysis iterates on the selling cost of ethanol until reviewed these factors for consistency with their experi. process water. Capital investment estimates are The Bottom Line. Fixed costs include labor. All of the costs reported here are the grid. which produces 4.600. marized in Table 7. a discounted cash flow felt it was reasonable to use a factored analysis for analysis can be used to determine the minimum selling determining the installed cost of equipment.3 and 4.Biotechnol.4 the next largest contributor to capital cost.61 1.2 product.S. overhead. Prog.9 includes storage capacity for 7 days of denatured ethanol utility chemicals 0. maintenance.000 indirect costs field expenses $28. The boiler produces 103. For the 2000 MT per day and renewable energy investments.000 total installed cost $143. A fixed costs 7. We of 198 million liters (52. indexed to 1997 dollars.000 project contingency $4. and Turbogenerator. The turbine generates 38 MW 22% of the total installed equipment cost. At a feedstock cost capital costs of around $2 per gallon of annual ethanol of $25 per ton.2 electricity credit -3. chemicals 4. $1. enzyme-based process. Waste treatment Figure 4.900. evaluations of renewable energy technologies for the U.000 total project investment $233.000 other costs $21. which have enzyme-based process for bioethanol.000 Area 600: Wastewater Treatment. The power plant and pretreatment $1.000. Sludge produced in this last step is burned. Total Project Investment ($1997) total equipment costs $135. making it a higher-cost alternative to cogeneration and enzyme production capability built into corn-derived ethanol. estimates provided by Delta-T Corporation and from and clean-in-place solution. sections are the two largest ticket items and together based on the current status of research that is complete represent 52% of the capital investment for equipment.000 total capital investment $212.50 13. solid waste disposal 0. considering the $1.300. This area includes chilled water. it is recommended capacity.300. Capital Cost. Department of Energy (37). 15. by Short in his description of how to perform economic related costs. and taxes.800.. We used price per gallon of ethanol produced.3 fluidized bed combustor burns three available waste fuels total 32.0 8.

we anologens. Simply by increasing the verting cellulose to ethanol. colocation with existing facilities (bio. 40) that will ferment known as site-directed mutagenesis. ethanol production 52. evolution” (45). Summary of Economics for a Near-Term “Best the insoluble surface of the biomass. 15. Certainly. The question of risk is further exacerbated by surface charge will lead to an increase in enzyme the high capital investment required. and academia to determine what types of improvements Substantial improvement in biomass conversion can in enzyme production and performance offer the greatest be achieved by making the following additional improve- potential for success in the short term (41). This could be feedstock targets for research are being pursued: costs (very low or negative for environmental wastes). There are other modifications of the enzyme will be geared toward technologies available to the project developer. 5 Table 8. at least for the on lignin is no longer available for hydrolysis. higher conversions of hemicellulosic sugars (38) and that We have identified two approaches for achieving these the current cellulase industry could provide a microor.. performance of 10 fold or more.44/gal •Improved Cellulose-Binding Domain. Cellulase en- ethanol yield 68 gal/ton zymes contain a catalytic domain and a binding domain. Research and development is needed Research Strategies and Goals for Improved Eth- to lower the cost of producing sugars. The first is a rational design approach there are other ethanologens (39. Specifi. It combines advanced genetic engineering Taking advantage of the best technology reported to be techniques with highly automated laboratory robotics to available today brings the cost of bioethanol down to randomly evolve new enzymes with the features required. In addition to modifying the binding domain. Genetic assumptions about feedstock costs. temperature at which these enzymes can operate. cellulase-producing cellulases are used. Our process stability of the enzymes should yield a 3-fold improve- model assumes an “nth” plant that does not carry with ment in specific activity. the development of Higher efficiency microorganisms for use in submerged higher value products from sugars would contribute to culture fermentors should be available by 2005. with high efficiency. In this section. producing microorganisms has yielded microorganisms ics of bioethanol production over the next 10-15 years. The biomass hydrolysis for every 20 °C increase in temper- genetic pool available in our labs and in others around ature of saccharification can be expected if T. There was a ments in ethanol-producing microorganisms: clear consensus in these discussions that the prospects •Ethanol-producing microorganisms capable of produc- for enzyme improvement through protein engineering are ing 5% ethanol at temperatures greater than or equal to very good. The second is a more recent Table 9 summarizes these “best of industry” improve. ganisms ferment xylose and glucose mixtures to ethanol gies. the active site. •Genetically engineered crops harvested as feedstock. 40). galactose and 3-D modeling tools to identify specific amino acids in the mannose) to ethanol. Research over the past 10 years on ethanol- describe briefly our strategies for improving the econom. This represents a major advance in Research Strategies and Targets for Cellulase technology. We identified the following targets for protein 50 °C and engineering: •Ethanol-producing microorganisms capable of con- •Increased Thermal Stability. By 2010. 1999. Summary of Economics for Near-Term organisms that represent a good starting point for Enzyme-Based Bioethanol Process ($1997) engineering new enzymes. Vol. Enzyme that adsorbs is not quite ready for) commercialization. enhancements in the it any allowance for the added risk of a “first-of-its-kind” cellulose binding domain.800 Biotechnol.7 MM gal/yr can lead to dramatic improvements in catalytic activity capital reduced 12% to $205 MM of the enzyme. protein sequence that can affect the enzyme properties listed above (42-44). production cost of ethanol $1. Also. It uses sophisticated the other hemicellulose sugars (arabinose. both representing the state-of-the-art in biotech- ganism to produce the enzyme more efficiently. •Improved microbial organisms genetically engineered used equipment (utilization of related equipment that has for high productivity of enzymes been shutdown). we can We have recently shown that a doubling of the rate of dramatically improve the rate of cellulose hydrolysis. Prog. nology research. 39. which contain high levels of cellulase enzymes bination of all of these. we plan to modify amino acid sequences production cost of ethanol reduced from $1. of Industry” Enzyme-Based Bioethanol Process ($1997) •Improved Active Site. the research strategies. Because of the importance of cellulase en. strategy known among biotechnologists as “directed- ments over the near-term case we have already described. $1.16/gal at the active site. adjusting its surface charge to minimize such unwanted cally. Further- zymes in the process. improvements in thermo- technology is commercially “unproven”. plant profitability. •Reduced Nonspecific Binding. mass burners and waste treatment facilities) or a com. it is been suggested that pretreatment can produce binding. Two term for a small market segment. By 2005. capable of converting hexose and pentose sugars to We also show what kinds of cost reductions can be ethanol (33.44/gal to $1. Even minor modifications of the enzyme yield increased 12% to 76 gal/ton production increased 12% to 58. In parallel with the protein engineering work. reesei-like the world includes thermo-tolerant. as previous conversion of pentose sugars by Enzymes. The development of ethanologens . DOE and NREL sponsored a series more.16 per gallon. These ethanol-producing microor- expected from the successful completion of these strate. This suggests that an enzyme-based The enzyme performance goals that are indicated in the facility could be built in the near term that competes future cases are based on the projected progress for these successfully in today’s fuel ethanol market. goals. However. The Impact of Proposed Research Activities. these new ethanologens have eliminated the need of colloquies with experts and stakeholders in industry for separate pentose and hexose fermentation trains. No. and protein facility. natural yeasts was not industrially attractive.2 MM gal/yr Improvements in the latter will lead to more efficient total project investment (capital) $234 MM interaction between the soluble cellulase enzymes and Table 9. our Facilities that are being engineered today all have some program plan calls for research aimed at improving the niche that allows them a special advantage in the short productivity of the enzyme expression systems.

tion. ature ethanologen by 2005. The year 2000 price shown is for the “best of Strategies for Process Development and Integra. The improvements in enzyme and with cellulase enzymes to emerge over the next five years. can be carried out to various extents by a number of microorganisms.16 per gallon. They are based on the application of engineering of agricultural crops holds great promise. also known as direct microbial conversion (DMC) or Consolidated Bioprocessing (CBP). Our program plan calls for introducing a high-temper. Figure 5. This new organism should be able to operate at 50 °C.S. industry” scenario. research on genetic conservative. and in the SSCF step improvements are major factors in •Optimal yields and operating conditions for the cocur- reducing cost. production. Thus improvements in yield following kinds of improvements: that occur in pretreatment. Figure 5 shows the decline in bioethanol pricing In the meantime.44 and $1. bound is a price projection for $15 per U. 5 801 capable of fermentation at temperatures greater than 50 °C can potentially reduce the cost of cellulase enzyme by one-half. we plan to continue to improve the based on these research targets. The cellulase enzymes and the fermenting organ- isms are the major thrusts of our applied research efforts. and the upper bound is a price goal is readily attainable. which provides a direct improve. (3) enzyme expression in the biomass.Biotechnol. 15. We research advances that build on the same process con- see tremendous opportunities to integrate biomass char. projection for $40 per U. targets. such as Fusarium oxysporum and bac- teria. Price trajectory for enzyme-based process technology. However. acteristic enhancement with the conversion process to lower the cost of ethanol production. though the the base established by “pioneer plants” working in niche timing for this last item needs to be determined more markets. precisely. we look at market. known DMC strains often exhibit relatively low ethanol yields and have not yet been shown to be effective in handling high concentrations of biomass. and biomass sugar fer- mentation) occur in a single microorganism (46). production of bioethanol. figuration. such as Clostridia sp. Vol. Conversion technology improvements could provide a Integrating these into a complete process is critical to 34 cents per gallon cost reduction over the next 10 years. while maintaining the best characteristics of the current ethanologens. As we continue to improve 2015 case study. with near-term cost projections bracketed between ment in the yield of ethanol. be used to understand the value of proposed research Strategies and Goals for Feedstock Improvements. in enzyme production. Prog. become available We show these projections not so much to claim any An optimized and integrated process using cellulase accurate forecast of the technology’s future over the next enzymes and cocurrent pretreatment technology should 10 years but to demonstrate how our process model can be available by 2005. The lower thermotolerant ethanologen capable of operating at tem. One company bounds on the error bars reflect the results of sensitivity has already disclosed early development work on a studies to assess the effect of feedstock price. 1999. by their nature. ton (dry) peratures as high as 70 °C (47).S..g. carbohydrate model is an improvement over previous work and shows composition increase.. the dramatic improvements rent pretreatment step in enzyme activity help to reduce the high cost of enzyme •Optimal yields and operating conditions for the si. The upper and lower core of the model and to expand our process modeling . (2) plant structure modification to We have presented the results of two years of effort to facilitate pretreatment in milder conditions or with upgrade our biomass-to-ethanol process model and to hemicellulases. commercial success. enzymatic cellulose hydrolysis. Combining these improvements with geneti- multaneous saccharification and co-fermentation step cally engineered feedstocks brings the savings to 40 cents •Testing of improved organisms and enzymes as they per gallon. utilize the model and the latest biotechnology tools to and (4) combinations of these suggestions or other ideas. performance and industry demonstrates commercial Price Trajectories for Bioethanol Based on Re. that. This process. We have also shown that future improved this as a long-term improvement. the enzymatic conversion processes offer low-cost sugar and fruits of such a research program are reflected in our year ethanol production capability.50 per MT) feedstock. including fungi. Pilot and bench scale optimization Feedstock costs represent the largest contributor to of an integrated process is expected to lead to the operating costs (see Table 7). establish future technology performance targets. somewhat Although still in its early stages. ton (dry) ($44 per MT) feedstock. we expect biomass conversion search Targets. In our analysis. which suggests that our ($17. In addition. ethanologen performance will impact the process in 2005 The introduction of low-cost cellulase enzyme technology and 2010. Plant characteristics Conclusions that could lower ethanol cost include (1) carbohydrate composition increase. The most advanced processing option is one in which all biologically mediated steps (e. These targets are. this technology (with on-site details of a research program targeted at higher carbo. Genetically engineered feedstocks with higher will expand the growth of bioethanol production beyond carbohydrate content might happen in 2015. production of enzymes) is competitive in today’s ethanol hydrate content have not been worked out. No. Our We have analyzed the impact of one change. enzyme production. This is because the current industrial etha- nologens can only meet desired performance at temper- atures of 30-33 °C. Because the feasibility and $1.

R. U. Advanced Preston. Symp. Allison. S. Glassner. Himmel.000. Schechinger. O.gov/ (35) Wales. von Stockar. J. NREL Subcontract AXE-8-18020-01. June 30. 5 activities beyond the process configuration presented performance in the mini-pilot biochemical conversion system here.. April 1996. Washington. Adney. 71-78. In BioEnergy ‘98-Expanding Bioenergy ronment: Cellulase Development in the Emerging Bioethanol Partnerships: Proceedings. Close. M. Riley. Cam. J.. pp 23-27. December 20. January 1998.. Eddy. A.. Golden CO. sion of Biomass to Ethanol. personal communication. ACS Symposium Series #666. M. Philadelphia. CO. pp 208. J. Sheehan. Hayward. Depart- (15) Anonymous. aggio. J. D. R... Laymon. U. September. Golden. Science 1998. A Manual for the Waste Into Energy to Fuel Cars. American Chemical Society: Washington. The overliming of dilute acid pretreated 639-649. F. (6) Wright. J.802 Biotechnol.. J. Development of an ASPEN PLUS (32) McMillan. C. Prog. Vinzant. Becker. Butterworth Publishing.S. tion by Escherichia coli strains coexpressing Zymomonas PDC American Chemical Society. November 9. (20) Farmer. W. T. Finkelstein. 1976... Tucker. (http://www. P. Bioethanol Production: Status and Prospects. 34/35. V. Sewell. Results of experimental work conducted at TVA.. J. Biotechnol. 8 (2). 15. lignocellulosics. National Renewable Energy Laboratory. Enzymatic Degradation of (39) Ingram. R. B. 67. Muscle Shoals Laboratories. R.. Madison. R. Energy Crop County Level Database.... S. December 1998. 63/65... (33) Zhang.. 14. Tucker. T. Vol. CO. 267. 1999. Patent 3990945. Final Report. Putsche.. NREL/TP-580-2615. 236. W. July 17.. Aurora. T. 1987. CO. Renewable Energy 1997.000. Hettenhaus.com. 1997. Issued March 19. The Oak Ridge Forum for Applied Research and Public Policy.C. D. C. 281. D. J. (5) Wright. Inc. 1997. D. D. Galvez. Biotechnol. L. R. J. 53. Glassner. Energy Prog.ott. Team Internal Report. Version 4. Process. N. G. National Renewable Energy Labora- Report No. Spring 1992. NREL/TP-462-5173. Biotechnol. Golden. pp (41) Hettenhaus.. Milestone Completion Re- 2-45. A New Bacterium Helps Turn Agricultural (37) Short.. Y. Newman. Conway. (36) Grethlein. process technology meeting or exceeding process qualifier Biochem. J. ICARUS Corpora- tion... U. J. D. E. August. F.. National Renewable Energy Laboratory Internal Draft). Schell.. Physical Property Database for Biofuels Components. In bridge. Himmel. Inc. D. 1997. Biotechnol. 1997. 1992.2-1. R. Science 1995. 2420-2425.C. Process. D. US Patent 5. The Evolution of the Fuel Ethanol sis and fermentation are carried out by one organism Industry: Feedstock Availability and Price. M. H. Hydrolysis Processes for Alcohol Fuel Production. 1976. 185-198. Release 9. personal communication.. Enzymes. Process Qualifier Milestone. Ruth. 3990944. (13) ICARUS Process Evaluator. Aspen Technology. Baker.. Picat- Merrick & Company. 1999. A. Part 1. Final Report to U.doe. M.. 1998. McMillan. CO.html). Nelson... The New York Times. J. Metabolic Engineering of a Pentose Metabolism (12) Wooley. Wooley. F. D. Tennessee. Alterthum. T. and Chemicals from Biomass. MA..html. Identification of Inhibitory Components Toxic tute. March T. P. National Renewable Energy Laboratory. 315-325. J. Inc. Hayward. Rockville. Collection Project. Energy Technologies. NY. 1997 December 1998. L. R. Golden.....ornl/bfdp/oreccl/database. Fuels Escherichia coli.. Oak Ridge. ACS Symposium Series No. 6. Helm... Packey. (29) Platz..C. (7) Technical and Economic Evaluation of Wood to ethanol (30) Ranatunga. Hatzis. McMillan. .. Lignocellulosic Biomass to Ethanol Process Design and (34) Hamilton. Chemical Process Equipment Selection and biofuels/process•engineering. O. 1999.. Internal Report.. D. Appl. Future Oil Supplies: Is Wolf Really at Door? (24) Graham. T. Biotechnol. and ADH Genes. 1995. 1998.. (38) Katzen. A report prepared for Solar Energy Research Insti. Biochem. T..S.. 1100-1110.. NY. Thomas.. (18) Himmel. Economic Evaluation and Energy Efficiency and Renewable October 25. Ethanol ment of Agriculture North Atlantic Area Eastern Regional Report. 1998. S.. MD. D. C. Golden. Karplus. Majdeski. Appl. (16) Wald. S. M. Mielenz. Helm. U. Tucker. A. Na- (17) Thomas. et al. Pathway in Zymomonas Mobilis. January 8. NREL/ C. Energy and the Envi. Oak Ridge National Laboratory. D. Internal report •Advanced power generation technologies to the Office of Transportation Technologies. Biotechnol. 1996. No.. Environ. Appl. Internal Report. •Consolidated bioprocessing in which cellulose hydroly- (22) Walsh. 15. No. (27) Huff et al. L. Sakon. Terrytown. J. J. J. xxxx-xxxx. C. M. (40) Ingram. R. Bagasse-to-Ethanol Plant Proposed. S. P. Yang. R. 1998. 1976. (10) Wooley. MP-425-20685. M. F. Washington. T. D’Agincourt. Golden.. W. April. Enzymatic Process Development and Enzymatic Hydrolysis Current and Futuristic Scenarios. Prog. Nieves. 1997. J. D. Toward Zymomonas mobilis CP4(pZB5) Xylose Fermentation. D. T... J. Bioeng. 10. Evaluation of Sulfuric Acid (28) Platz. and ether soluble organics. 1991. W. Technol.. M. M. 36: Air/ Economics Utilizing Co-Current Dilute Acid Prehydrolysis Oxygen 22 Factorial Design. 1998. November 9. 1128-1131. M. R. Ethanol Production from Corn. Enzyme Microb.. 1997. 155-175. J.. Hatzis. Bioeng. Microbiol. J. Baker. Deanda. CO.. P. J. Report.. and without the use of enzymes) (21) McMillan. 295-302. (3) Annual Energy Outlook 1999 with Projections to 2020. Clark.. Initial Approaches to Artificial Cellulase Systems for Conver. National Renewable Energy tory. WI. R. Bergeron. M. Future potential technologies to be evaluated and assess continuous SSCF performance to recommend include future work. J.esd. CO. J. D. DOE/EIA. L. 31 (11) Wastewater Treatment Options for the Biomass-To-Ethanol August 1996.. July 1999. McMillan. Laboratory.. 1995. 618. Ethanol Produc- Insoluble Carbohydrates. Available at www. Baker. Patent No. Jervis. uronic acids (9) ASPEN Plus. (8) Hinman. H. preparation. D. Date of tests unknown. National Renewable Energy •Counter-current pretreatment technology (both with Laboratory. M. 1988. 1992.arkenol. M. 105-123. P. Colorado. Biochem.S. Personal Communication. 1984. PA. Department of Energy. (25) Wilke. Design. Preliminary Cost Energy Information Administration. port: Enzyme Hydrolysis of Cellulose: Short-Term Com- (19) Bioethanol Strategic Road map: A planning framework for mercialization Prospects for Conversion of Lignocellulosics to development of biomass-to-ethanol technology (Preliminary Ethanol. (2) Lynch. B. M. Research Center. Cellulase Production Experiment No. Ruth. pp Industry. Sunds Defibrator. J.. June 5. Adney. T.. L. Report No. O. R. Sunds Defibrator. 1988. tional Renewable Energy Laboratory. 240-243..0. Chem Systems. (4) Kerr. References and Notes (23) Glassner. R.. Walter.. The Next Oil Crisis Looms Large-and Perhaps (26) Gauss et al. Preliminary Estimate of the Cost of Ethanol Production for (31) Ranatunga. K. Milestone Completion Report: Demonstrate Diversity of structure and Convergence of Function. October 1995. Holt. Report No. Projected Process Economics for (14) See http://www.. Analyses for Enzymatic Hydrolysis of Newsprint. Elander. Adney. 0383(99). No.. Chou. SSF Technology. Golden. The fate of inorganics.. 2. Polysaccharide Hydrolase Folds M. M. Vol. Corn Stover (1) Sheehan. 1990. D.S.. (42) Himmel. 10/20/98. W. Genetic Engineering of Ethanol Production in Bioethanol Production Technologies: A Perspective. Ethanol from Lignocellulose: An Overview. Appl. Conway.. Jervis. K. C. 10 (2/3)..

4-Glyca. Costs of Mature Biomass Ethanol Technlogy. 741-761. 1999. reesei and Other Microor. Himmel. Moore... 5712142. 1999.. Elander. C.. Vol. Appl. L. W. Royal Society of Chemistry: Cambridge. K. Method for Increasing Thermostability in Cellulase Enzymes. ganisms. U.. Likely Features and nases. Nerinckx. 15. (44) Thomas.. F./Biotechnol.. BP990107U . (46) Lynd. 5 803 (43) Warren.K. Prog. W. Optimizing Industrial Enzymes Accepted August 9. E. (47) AGRE-0063: High Temperature Ethanol Fermentation of 123. Wyman. Adney. 1997. A. R. Baker. S. R. Patent No. See www. Structure and Function in β-1. Biochem. R. S. J. H. Chou. Eds. (45) Arnold. July 1.nf2000. 1998. M. 58... Eng.Biotechnol. Biochem. Piens. NF-2000 Online Database Information. 1997.-C. No. U.S. 1 July 1992 to 30 June 1993. Lignocellulosic Waste. Claeyssens.. M. 1996. by Directed Evolution. Y.... 57/58. J. Adv. pp 115. O. J. The Biotechnol. E. 1-14.org. In Carbohydrases from T.