i

THE FIFTH

INTERNATIONAL STARCH TECHNOLOGY

CONFERENCE

Program Proceedings

University of Illinois Urbana, IL, USA June 3-6, 2007

Acknowledgement Mary Schultze, University of Illinois

Edited by Kent Rausch Vijay Singh

Mike Tumbleson

ii

Agenda

Sunday, June 3, 2007 4:00 p.m. Registration 6:00 p.m. Casual Reception Monday, June 4, 2007 7:30 a.m. Registration and Continental Breakfast 8:45 a.m. Building a Prosperous Future Where Agriculture Produces and Uses Energy Efficiently and Effectively James R. Fischer, Research, Education and Economics, USDA 9:30 a.m. Break 10:00 a.m. Fuel Ethanol Policy in the United States: Overview and Policy Alternatives Kelly J. Tiller, University of Tennessee 10:45 a.m. Break 11:15 a.m. Disc Mill Energy Issues William R. Enterline, Andritz Sprout 11:30 a.m. Drying of Grain Residues and Sludges Using Biomass Fuels George Svonja, Barr Rosin 12:00 p.m. Lunch 1:30 p.m. Plate Heat Exchanger Design John Robertson, Alfa Laval 1:45 p.m. Ethanol as an Economic Competitor to Gasoline Andrew J. McAloon, Eastern Regional Research Center, ARS, USDA 2:15 p.m. Break

iii

Monday, June 4, 2007 2:45 p.m. Production of Ethanol and DDGS from Barley Containing Reduced Beta-Glucan and Phytic Acid Mian Li, Genencor International, A Danisco Company 3:00 p.m. Sources of Variation in Dry Grind Processing Streams Ronald L. Belyea, University of Missouri 3:15 p.m. An Overview of United States Sorghum Starch and Ethanol Production Jeff Dahlberg, National Sorghum Producers Association 3:30 p.m. Refreshments with Poster and Exhibit Review Tuesday, June 5, 2007 7:30 a.m. Registration and Continental Breakfast 8:45 a.m. Ethanol Reality Check Rodney J. Fink, Western Illinois University 9:15 a.m. Characterization of Glucoamylases for Conventional Simultaneous Saccharification and Fermentation Chee-Leong Soong, Novozymes North America 9:30 a.m. Break 10:00 a.m. Fuel Ethanol Life Cycle Energy Use and Greenhous Gas Emissions May M. Wu, Argonne National Laboratory 10:30 a.m. Continuing Evaluation of Low Conductivity Electrodialysis as an Alternative or Complement to Ion Exchange Resins in Starch Processing Daniel H. Bar, AMERIDIA, A Division of EURODIA INDUSTRIE 10:45 a.m. Break 11:15 a.m. Economics of Biomass Gasification and Combustion at Fuel Ethanol Plants Douglas G. Tiffany, University of Minnesota 12:00 p.m. Lunch

iv

Tuesday, June 5, 2007 1:30 p.m. Modification of Starch by Branching Enzyme From Rhodothermus Obamensis Anders Viksø-Nielsen, Novozymes A/S 1:45 p.m. A Primer for Lignocellulose Biochemical Conversion to Fuel Ethanol Bruce S. Dien, National Center for Agricultural Utilization Research, ARS, USDA 2:15 p.m. Break 2:45 p.m. Energy Savings Through Better Use of Separators Tristan Merediz, Westfalia Separator 3:15 p.m. Ultrasound Pretreatment of Corn Slurry to Enhance Sugar Release Samir Kumar Khanal, Iowa State University 3:45 p.m. Refreshments with Poster and Exhibit Review Wednesday, June 6, 2007 7:30 a.m. Registration and Continental Breakfast 8:45 a.m. Energy and Protein—A Global Perspective David A. Cook, Cargill 9:15 a.m. The Use of Biosolids to Generate Steam at Dry Grind Ethanol Production Facilities Gregory Coil, M. A. Mortenson Company 9:30 a.m. Break 10:00 a.m. Benchmarking Industrial Energy Performance: The Energy Star Approach Walt Tunnessen, US EPA, Energy Star 10:45 a.m. Break 11:15 a.m. Membrane Processes - Opportunities in Corn Pprocessing William J. Koros, Georgia Institute of Technology 12:00 p.m. Adjournment and Box Lunch

v

EXHIBITS

Ameridia, Division of Eurodia Industrie 732-805-4003 20 F Worlds Fair Drive Fax: 732-805-4008 Somerset, NJ 08873 Daniel Bar Andritz Sprout 678-947-1588 7225 Sheffield Place Fax: 678-947-1588 Cumming, GA 30040 William Enterline Barr-Rosin, Inc. 630-659-3980 255 39th Avenue, Suite G Fax: 630-584-4406 St.Charles, IL 60174 Kosta Kanellis C.W. Brabender Instruments, Inc. 201-343-8425 50 East Wesley Street Fax: 201-343-0608 PO Box 2127 South Hackensack, NJ 7606 Salvatore Iaquez Genencor International 585-256-5200 200 Meridian Centre Blvd. Fax: 585-256-6952 Rochester, NY 14618

Angela Blackwell

vi

Graver Technologies 302-731-3539 200 Lake Drive Fax: 302-731-1707 Newark, DE 19702

Scott Wittwer Gusmer Enterprises, Inc. 847-277-9785 1401 Ware Street Fax: 847-277-9789 Waupaca, WI 54981

Craig Panasy Larox, Inc. 708-974-2166 12 Cour Caravell Fax: 708-974-2166 Palos Hills, IL 60465

Joe Skafar Novozymes, N.A. 919-494-3000 77 Perry Chapel Church Road Fax: 919-494-3415 Franklinton, NC 27525

Steve Schnurrer Perten Instruments 217-585-9440 6444 S. 6th Street Road Fax: 217-585-9441 Springfield, IL 62712

Jim Powers Westfalia Separator, Inc. 901-751-0396 100 Fairway Court Fax: 901-751-0397 Northvale, NJ 7647

Tristan Merediz

vii

PLATFORM PRESENTATIONS

viii

Page

Building a prosperous future where agriculture produces and uses energy efficiently and effectively Jim Fischer

2

Fuel ethanol policy in the United States: overview and policy alternatives Kelly Tiller

13

Disc mill energy issues Bill Enterline

18

Drying of grain residues and sludges using biomass fuels George Svonja

21

Plate heat exchanger design John Robertson

27

Ethanol as an economic competitor to gasoline Andy McAloon

30

Production of ethanol and DDGS from barley containing reduced beta-glucan and phytic acid Mian Li

34

Sources of variation in dry grind processing streams Ron Belyea

42

An overview of United States sorghum starch and ethanol production Jeff Dahlberg

50

Ethanol reality check Rodney Fink

56

Characterization of glucoamylases for conventional simultaneous saccharification and fermentation Chee-Leong Soong

68

PLATFORM PRESENTATIONS

ix

A primer for lignocellulose biochemical conversion to fuel ethanol Bruce Dien

87

Energy savings through better use of separators Tristan Merediz

96

Ultrasound pretreatment of corn slurry to enhance sugar release Samir Khanal

102

Energy and protein --- a global perspective Dave Cook

112

The use of biosolids to generate steam at dry grind ethanol production facilities Greg Coil

117

Benchmarking industrial energy performance: the Energy Star approach Walt Tunnessen

123

Membrane processes --- opportunities in corn processing Bill Koros

132

Fuel ethanol life cycle energy use and greenhouse gas emissions May Wu

69

Continuing evaluation of low conductivity electrodialysis as an alternative or complement to ion exchange resins in starch processing

Dan Bar

73

Economics of biomass gasification and combustion at fuel ethanol plants Doug Tiffany

76

Modification of starch by branching enzyme from Rhodothermus obamensis

Anders Vikso-Nielsen

86

x

POSTERS

Page Deposition control in bioprocess equipment to increase plant efficiency Carol Batton

138

Oil, corn and ethanol: no more cheap food Ron Belyea, Joe Horner, Kent Rausch and Mike Tumbleson

140

Coproduction of fuel ethanol and new value added coproducts David Johnston

142

Starch components and properties in Triticale and other cereals John Lu, Byron Lee, Brian Beres, Andre Laroche, Denis Gaudet and Francois Eudes

144

Biofuels production in the Pacific Northwest: opportunities and challenges

G. S. Murthy

146

Improvements in corn gluten dewatering Dave Scheimann

148

Sorghum as a viable renewable resource for biofuels and biobased products

Xiao Wu, Renyong Zhao, Scott Bean, Paul Seib, Jim McLaren, Ron Madl, Mitch Tuinstra, Mike Lenz and Donghai Wang

150

Enzyme production by industrially relevant fungi cultured on coproducts from corn dry grind ethanol plants

Eduardo Ximenes, Bruce Dien, Mike Ladisch, Nate Mosier, Mike Cotta and Xin-Liang Li

152

1

PLATFORM PRESENTATIONS

2

BUILDING A PROSPEROUS FUTURE WHERE AGRICULTURE PRODUCES AND USES ENERGY EFFICIENTLY AND EFFECTIVELY

James R. Fischer*

Senior Scientific Advisor for Energy, Science and Education to the

Undersecretary of Agriculture for Research, Education and Economics, USDA 214W Whitten Building, 1400 Independence Avenue, Washington, DC 20250

(202-720-5923) james.fischer2@usda.gov INTRODUCTION

The current energy situation presents the US and the world with both challenges and

opportunities. Oil price levels and volatility and concerns about climate change caused by fossil fuel use are challenging the energy basis of America’s economy. But every coin has two sides. These challenges also represent opportunities to do things differently and represent a particular opportunity for agriculture, which is a significant consumer and producer of energy. In the future, agriculture will be an important part of the new directions in energy markets. Increased use and production of renewable energy in agriculture is one likely outcome. Increased efficiency in agricultural use of energy is another. The United States Department of Agriculture (USDA) already sponsors many energy related programs, from research and development to education. A congressionally mandated committee has advised USDA on how to develop its energy programs in the years to come, and the Department is responding to its recommendations. Investing in human talent and new technologies holds promise to take us in new directions. THE ENERGY SITUATION: CHALLENGES Oil Dependence

Many American adults have memories of the two oil price shocks of the 1970s, which contributed to high inflation and unemployment. Fears of similar supply related disturbances have led to a new facet of national security called energy security. This is the ability of the nation to obtain energy reliably and affordably. In practice, the term is used most often in connection with oil imports. While the economy is less vulnerable to oil supply disruptions or price spikes than it was three decades ago, geopolitical and oil market concerns are strong.

In 2005, the US imported about 60% (on a net basis) of the crude oil and petroleum products it used. Concerns are heightened because a significant share of current imports come from the Middle East. Most of the world’s long term supplies of less expensive crude oil deposits are in that region, so the share is expected to increase. After the terrorist attacks of 2001, concerns have grown. The war in Iraq, additional terrorist attacks on Western targets around the globe, and specific attempts to attack oil facilities in the Middle East, make markets and governments insecure about supply disruptions.

3

This is an even greater danger when markets are tight. Increasing demand from China and other countries has stretched production capacity and played a significant role in higher oil prices. With little spare capacity, supply disruptions could have more dramatic effects, and the risk of oil price volatility is greater than ever. The risk of terrorism and tighter crude oil markets have led to increased oil prices.

Oil is a finite resource; it was deposited in geologic processes during millions of years. There may be a large volume of it left but it is running out. The term peak oil refers to a kind of tipping point in world oil supply. The peak is the point where the maximum production is reached. After that, exploration to find new sources and new technologies to produce more from existing wells are insufficient to continue to increase production (Figure 1). The decline may be steep or gradual, but inevitable. Production in the US reached its peak in 1970 but the world as a whole has not yet reached that milestone.

Figure 1. Energy discovery and production.

Human use of oil has been outstripping our ability to extract it. The world consumes two barrels for every barrel discovered. It took approximately 125 years to use the first trillion barrels of oil; we are on pace to use the second trillion barrels in about 30 years. Production has exceeded new finds for the last two decades. Oil magnate T. Boone Pickens is not optimistic about continued increases in mankind’s oil use. He does not think global oil production can be increased much above its current level. He believes changes resulting from decreasing oil supply are not likely to be abrupt but there will be changes over time. Climate Change Global climate change is a growing concern with the use of fossil fuels, including oil. In recent years, the scientific consensus regarding anthropogenic warming of the earth’s climate has solidified. Human activities are warming the earth and there are serious resulting impacts.

Has World Oil Production Peaked?

4

Recent working group reports of the Intergovernmental Panel on Climate Change (IPCC) conclude there is “very high confidence” that human activities have resulted in warming (IPCC 2007a); there is “high confidence” these effects are taking place (IPCC 2007b).

Projections of possible effects are uncertain but many governments have initiated activity to limit the growth of, and eventually halt, concentrations of greenhouse gases (GHG) in the atmosphere. The Kyoto Protocol, ratified by 166 countries and other governmental entities, took effect on February 16, 2005. The Protocol has been criticized widely but is a significant step in global action to mitigate emissions. The US never ratified the Protocol and has not enacted mandatory emissions controls at the federal level. Instead, the US has emphasized the importance of scientific and technological advances in achieving similar goals.

State and local governments have taken steps to mitigate emissions. The state of California has passed legislation that requires a 25% reduction in carbon emissions by 2020 to reduce emissions to 1990 levels. In February 2007, California and four other western states agreed to set a cap for carbon emissions for their region before the end of 2007 and to set up an emissions trading system by August 2008. Seven Northeastern states agreed to mandatory limits on carbon dioxide emissions from power plants. This action aims at a target of stopping the increase in emissions by 2009 and reducing them by 10% from 2005 levels by 2019. Other states and cities have taken action and there is movement for federal leadership in this area.

A lawsuit filed by several states and environmental groups seeking to compel the Environmental Protection Agency (EPA) to regulate GHG emissions from motor vehicles reached the US Supreme Court. The Court decided on April 2, 2007, that EPA did have authority under the Clean Air Act and ordered EPA to reconsider regulation of GHG emissions from new cars and trucks. This bolsters activity already underway for federal regulation. Many legislative initiatives have been introduced to Congress; companies, including Shell Oil, have called for federal action to ensure consistent nationwide regulatory treatment of GHG emissions. Energy Use and Economic Development

Energy consumption is fundamental to modern economies and daily life in developed countries. Energy consumption and affluence are linked tightly. Some developed countries use energy more efficiently than others but these two variables track closely in a regression analysis. Developed economies will use more energy as their economies continue to grow and developing countries are expected to increase exponentially their energy use as their economies modernize during the next several decades. The two largest developing economies, China and India, will be the future world leaders in emissions. As the world’s population grows toward 10 billion or more this century, greater energy use and its resulting GHG gas emissions will be inevitable. How Well We Use Energy Where we get energy now and how well we use it is an indication of our current energy status and what direction we need to move in. Nonrenewable energy sources supplied 94% of US energy in 2001. Petroleum, natural gas and coal each supply about a quarter to a third of the

5

total; nuclear energy supplies under 10%. We use renewable energy for only 6% of our energy needs. More than half of this is biomass, mostly in the form of wood chips and other wastes and residues used in the forest products industries, like paper making. Hydroelectricity represents another large segment of the renewable share, with other sources like wind, solar and geothermal contributing smaller shares. Renewable energy, with its many positive attributes, could make up a significantly larger share of the total.

We also do not use energy as well as we could. Of the energy available in the resources

we tap, we lose 60% of it in the process of converting it to do work for us, whether in the form of mechanical work as in an automobile engine or in burning fuel to make electricity. We could reduce demand if we used energy more efficiently. Perhaps our best energy resource is reducing the energy we waste. THE ENERGY SITUATION: OPPORTUNITIES

Oil prices have remained elevated, US oil imports have increased and there is growing concern about carbon emissions. Policy makers are interested in paths to reduce petroleum consumption and reduce carbon emissions. Renewable alternative fuels and energy efficiency are top priorities. Technologies

There are reasons to be optimistic about the potential of some renewable technologies. Renewable energy sources, including wind, solar and geothermal, are likely to play an increasing role in our energy mix. Wind power in the US has grown to more than 11,600 Megawatts; its costs have decreased to a few cents per kilowatt hour, in a competitive range with fossil fired electric generation. Technological development may increase output and decrease costs for smaller wind turbines operating in lower wind speed environments, opening more potential markets. Solar power costs are higher but also have decreased dramatically. The worldwide solar industry has been booming, to the extent that prices for inputs such as silicon have increased with high demand. Geothermal energy, for electricity generation and direct supply of heat, also has increased, especially in the western US.

Similarly, energy efficiency will be more important, with technologies such as improved engines and zero energy buildings coming to market. Hybrid electric drive, already commonplace, can enhance vehicle power and performance while decreasing fuel use. Plug in hybrids could extend the use of electricity in vehicles. With enhanced batteries, cars and trucks could run solely on electricity or fuels, or on both in combination. Further in the future, fuel cells powered by hydrogen could replace the internal combustion engine altogether. Zero energy buildings are structures that integrate energy efficiency technologies and on site renewable electricity generation to produce at least as much energy as they use and sell power into the electricity grid. Such buildings might have highly insulating coated windows, LED lighting systems and other efficiency measures, along with solar panels integrated into roofing tiles and connected to the electrical system. The food system includes numerous opportunities to employ these renewable generation and efficiency opportunities. For example, wind turbines are present on many farms and

6

ranches, with more potential to supply land and wind resources from agriculture. Efficiency advances in buildings and vehicles also would benefit many parts of the agricultural value chain, both pre and postharvest. Also, there are opportunities unique to agriculture. For example, genomics could produce a greater array of nitrogen fixing crops, reducing the need for fertilizer. This can be understood as an energy efficiency technology, saving natural gas through agricultural science. Major Federal Initiatives

Public policy has been directed at energy challenges and opportunities. Federal legislation, beginning in the 1970s, was focused on energy directly or environmental issues that impacted energy. Below (Table 1, Figure 2) is some of the energy related legislation during this period. Table 1. Federal legislation and policies impacting energy.

One example can be used to illustrate the power of well constructed public policies to help achieve energy goals. A production tax credit (PTC) applicable to electricity generated by wind turbines was set at a level (1.5 cents/kWhr, with subsequent upward adjustments for inflation) that provided the incremental economic incentive to cause state of the art wind technology to compete with alternatives. The evidence of its power is the decrease in wind farm construction during lapses that occurred between expiration and renewal of this policy in 2000, 2002 and 2004.

Federal Energy Legislation and Other Developments 1978 – Public Utility Regulatory Policies Act (PURPA) 1978 – Energy Tax Act (ethanol blends $0.40/gal tax exemption) 1992 – Energy Policy Act (tax credit for renewable energy production) 1998 – Energy Conservation Reauthorization Act (included biodiesel credit) 1998 – Alternative Motor Fuels Act (Encouraged cars fueled by alternative fuels) 2000 – Biomass R&D Act (DOE/USDA joint R&D biobased industrial products) 2002 – Farm Bill (First energy title in Farm Bill history) 2004 – Job Bill – (included biodiesel fuel tax credit) 2005 – Energy Policy Act of 2005 (RFS, production tax incentive through 2007) 2006 – State of the Union – “addicted to oil” 2006 – Advanced Energy Initiative 2007 – State of the Union – Twenty in Ten 2007 – Biweekly Energy Briefings to USDA Secretary 2007 – Farm Bill – Increase Budgets for bioenergy R&D

Federal Environmental Policies Impacting Energy 1990 – Clean Air Act (CAA) – first major environmental policy to have an impact on renewable energy. 2006 – EPA requires the use of ultra low sulfur diesel fuel (15 parts per million sulfur) 2010 – Non-road diesel fuel regulations will take place

7

Figure 2. Production tax credit. Initiatives

In addition to the legislation, regulation, speeches and other notable activities listed in Table 1, a number of plans and goals have been initiated by the federal government and other groups in recent years. One of the most prominent is the Advanced Energy Initiative (AEI). Key components of the AEI include “chang[ing] how we power our automobiles” and “chang[ing] how we power our homes and offices.” One focus is on advanced battery technologies to improve hybrid automobiles, including plug in hybrids that could both draw from and contribute to the electric power grid. Another focus is on decreasing the cost of producing ethanol from cellulose. With lower costs for conversion, developments in feedstocks, and in infrastructure and vehicles, there is the potential for tens of billions of gallons of cellulosic ethanol in the next decade or decades. A third area in transportation is the development of hydrogen fuel cells. Practical, safe, powerful and cost effective hydrogen fuel cell powered vehicles, especially if hydrogen is generated from renewable sources, could be the future of transportation. The AEI also targets residential and commercial building energy use, focusing on clean coal, nuclear and renewable energy.

The Biofuels Initiative was developed to target a goal the President set in his 2006 State of the Union speech, to replace more than 75% of our oil imports from the Middle East by 2025. This initiative has been accompanied by a proposed near doubling (from $89.8 to 179.3 million) of the budget of the Biomass Program of the Department of Energy (DOE), in the FY 2008 budget request, compared to the FY 2006 appropriation, to accelerate cellulosic ethanol development and related technologies. The proposed budget for USDA includes a proposed increase of $50 million/yr for 10 yr for the Research, Education and Economics mission area, for energy.

Effect of Production Tax Credit (PTC) Effect of Production Tax Credit (PTC) on the US Marketon the US Market

Source: AWEA Wind Power Outlook 2005

Annual Megawatts Installed

8

The Biomass Research and Development Act of 2000 gave rise to the Biomass Research and Development Initiative. The vision is that by 2030, “a well established, economically viable, bioenergy and biobased products industry will continue new economic opportunities for the US, protect and enhance our environment, strengthen US energy security, provide economic opportunity and deliver improved products to consumers.” By 2030, the ambitious goals are that there will be 68 billion gallons of biofuels, constituting 20% of the market for liquid vehicle fuels, 10 quadrillion Btus (quads) of electricity generated from biomass sources and 55 billion lb/yr of bioproducts.

The most recent of these initiatives is the National Biomass Action Plan. This is an interagency effort of the federal government, led by USDA and DOE, to coordinate R&D activities across the government related to biofuels. Representatives met in a workshop in November 2006 to define agency roles and activities, identify gaps in R&D and synergies across agencies and assess budgets. A report of the workshop conclusions is forthcoming. Goals

Many goals have been set to motivate public and private sector efforts in developing biomass energy. Goals have been set by both governmental and nongovernmental groups. DOE set a goal, in response to the President’s 2006 State of the Union address, to displace 30% of 2005 gasoline usage with biofuels by 2030. This has been called the “30 by 30” goal, and envisions 60 billion gallons annually of biomass fuels in 23 yr.

In his 2007 State of the Union message, the President articulated a nearer term goal related to transportation fuel. Called the “20 in 10” goal, it calls for reducing US gasoline usage by 20% by 2017. Three quarters of this amount (15 percentage points of the 20%) would come from substituting biofuels, with the remainder from vehicle efficiency. A nongovernmental group has issued a call for “25 by ‘25”, the production of energy from the agricultural sector to account for 25% of all US energy needs by 2025. This includes renewable sources such as biofuels, wind and solar power, and represents an estimated 32 quads of bioenergy in 18 years.

The Aspen Institute, a nonpartisan, nonprofit organization, has set an ambitious goal of 100 billion gallons of ethanol produced in the US annually by 2025. The National Agricultural Biotechnology Council (NABC) also has set several goals. For liquid transportation fuels, it targets biofuels for 50 billion gallons by 2025 and 100 billion or more gallons by 2035. For organic chemicals, the aim is for glucose produced at $0.04/lb and competitively priced ethylene, most likely produced with genetically modified organisms. In the area of organic materials, the NABC envisions new fiber crops with functional improvements and higher yields, produced with genetically modified organisms. ENERGY AND AGRICULTURE The spectacular increase in agricultural productivity has been related to energy use in many ways. Mechanization with fossil fuels and the use of energy intensive fertilizer are the two main illustrations of this. Agriculture is a significant user of energy and it is increasingly a producer of energy. The recent rapid increase in ethanol production from grain starches and the

9

more common appearance of wind turbines on farm and ranch land are the most obvious examples.

One of the most followed energy inputs is nitrogen fertilizer. With natural gas prices trending upward in recent years and the North American supply basin apparently at maturity, natural gas is now joining oil as an imported commodity. Imports of fertilizer also have increased. On the positive side, US agriculture has been using energy more efficiently, essentially producing more with the same amount of energy. Energy intensity, the ratio of total energy inputs to total output, has decreased, from 2.1 in 1970 to 0.65 in 2004.

Agriculturally linked energy use is not limited to diesel fuel for tractors, natural gas used to make fertilizer and other obvious items. Energy relevant to our sector encompasses the entire food system, everything from agricultural production, through transport and processing of foods, to the electricity to operate refrigerators in consumer homes (Table 2.). Table 2. Energy use in the US food system.

Agriculture can increase its role as a producer of energy. A USDA and DOE sponsored study, “Biomass As Feedstock for a Bioenergy and Bioproducts Industry: The Technical Feasibility of a Billion Ton Annual Supply,” (USDA/DOE 2005) established the feasibility of a high level of biomass feedstock for energy and other nonfood uses. There is potential for enough sustainable production of cellulose to displace more than 30% of current US petroleum consumption, with no impacts on human food, animal food or export demand. This would come from a variety of biomass types. From agricultural lands, this includes: crop residues, such as corn stover, wheat straw and soybean residue; animal wastes, notably manure; and dedicated energy crops, such as switchgrass, poplar and willow trees. From forest lands, this could include other residues such as forest thinnings, fuelwoods, logging residues, wood processing and paper mill residues, and urban wood waste.

10

USDA AND ENERGY

The need for agricultural involvement is present, the potential for a significant energy presence has been established and USDA has been focused on energy to encourage the realization of the potential. USDA has a number of activities related to energy. The Biomass Research and Development Board (BRDB) is the body bearing the responsibility to implement the Biomass Research and Development Initiative. Cochaired by USDA and DOE, the Board is coordinating federal government R&D in pursuit of ambitious goals for biomass fuels, electricity and industrial chemicals and materials. Other Board members include representatives of the National Science Foundation (NSF), the EPA, the Department of the Interior (USDI), the Department of Transportation (USDOT), the Office of Science and Technology Policy (OSTP) and the Office of the Federal Environmental Executive. The Board is served by a Technical Advisory Committee in regard to technical focus.

USDA also has an Energy Council. The makeup demonstrates the high level of

USDA’s commitment to energy issues: it is led by Tom Dorr, Undersecretary of Rural Development; the cochairs are the Chief Economist and the Undersecretary for Natural Resources and Environment. Other governmental departments are exofficio members, to ensure appropriate interagency consultation. The departments (and EPA) taking part in the Biomass R&D Board are members, with the addition of the Department of Commerce. The purposes of the Energy Council are: 1) oversight of implementation of President’s National Energy Plan including the EPA Act of 2005, 2) coordination of USDA energy related programs, 3) review and evaluation of key policy and program decisions on energy matters, 4) development of initiatives to transform and generate alternative energy sources and 5) assist and oversee continued implementation of Title IX of the 2002 Farm Bill.

The Bioenergy Bioproducts Coordinating Council (BBCC) is another tool USDA uses

to address energy developments. Its has the following purposes: 1) helping farmers and forest landowners to provide food, feed, fiber and fuels, 2) resolving technology and market barriers for biofuels and bioproducts, 3) leading biobased products development and Federal procurement and 4) providing information and education to support the bioeconomy and energy efficiency.

BBRC coordinates across the entire Department on these issues. It includes representation and participation from all agencies and programs at USDA; it works through committees. It develops a 5 yr plan and annual priorities for the Department and utilizes the different agency education and training programs to deliver its messages, providing information about policy implications within USDA. Where the Biomass R&D Board serves to coordinate research and development at the federal level, the BBCC at USDA links basic, applied and developmental research with commercialization of new technologies. The programs fully develop and use demonstration and pilot opportunities, provide formal and informal education and training across mission programs and promote energy conservation and efficiency in other programs.

The Agricultural Bioenergy and Bioproducts Research, Education and Economics (ABBREE) Task Force, under the Research, Education and Economics mission area, encompasses activities in several USDA organizations, including the Agricultural Research

11

Service (ARS), the Economic Research Service (ERS), the National Agricultural Statistics Service (NASS) and the Cooperative State Research, Education and Extension Service (CSREES). Some of the most important work directed by this task force concerns the promotion of energy science and education, covering fundamental and applied research and education of the next generation in energy and agriculture. FUTURE DIRECTIONS

Energy in the United States is moving in the direction of increasing renewable energy and energy efficiency, with USDA preparing for agriculture to play a big role. USDA’s Energy Science and Education mission area is working to make certain the Department is doing the right things in the right ways to lead in the transition to the energy future.

The National Agricultural Research, Extension, Education, and Economics Advisory Board (NAREEEAB) is a Congressionally-mandated Federal Advisory Committee, commissioned to examine USDA’s approach to science and education and provide advice to the Department. It held meetings to explore USDA’s approach to energy in March and October of 2006, and has developed recommendations for USDA for bioenergy development:

1. Take the lead on strategies for development of a bioenergy and bioproducts based economy. 2. Announce a holistic and coherent vision of its role and strategy in bioenergy and to convey the message to the public. 3. Undertake a focused effort to request the increased funding required to develop a nationally visible program. 4. Take a portfolio approach while identifying which new intermediates for current and new applications may hold the most promise for potential commercialization. 5. Develop a systems approach including economics, engineering and social management to evaluate research directions and alternatives. 6. Seek additional funding for new and enhanced research and education bioenergy and bioproducts initiatives.

The following vision will lead USDA in implementing the recommendations: “Building

A Prosperous Future Where Agriculture Produces and Uses Energy Efficiently and Effectively.” Goals under this vision include:

1. Develop comprehensive, integrated intramural and extramural research program that effectively explores the role of agriculture as both a user and producer of energy. 2. Establish energy science, education and extension activities related to agriculture with university and industry partners as well as other federal and state agencies. 3. Initiate comprehensive technology transfer programs for agriculture energy research to agriculture producers, suppliers and users. The focus of energy science and education programs include both renewable energy (biobased and other renewable energy) and energy efficiency, both preharvest and traditional agricultural, and postharvest, or including the rest of the food system. The Energy Science and Education mission area is planning a meeting to determine responses to the NAREEEAB recommendations and set the course for USDA’s future direction

12

in energy. The meeting is planned for September 2007, in Washington, DC. Its purposes will be to: 1) set vision and goals, 2) identify program areas of focus, 3) identify critical cross cutting issues, 4) establish agency responsibilities and 5) suggest processes to achieve goals. The expected outcomes include establishing a vision and goals for USDA’s energy activities, establishing program focus areas, identifying REE responsibilities and comparative advantages, accepting all the agency responsibilities, integrate cross cutting issues into program areas, identifying initial program priorities and presenting the process for moving forward. THE PATH FORWARD

In moving forward with renewable energy and energy efficiency, in meeting the challenges of oil dependence and global warming, we should remember that the greatness of the US has been our ability to cultivate human talents and apply them in developing new technologies. The history of American agriculture is an excellent example of this. Advances in crop and animal sciences have led to ever increasing yields, lower energy intensities and more abundant affordable food. Blessed with substantial agricultural lands, the US has made the most of our opportunities and fed a large and growing country and the world as well.

Our new challenge and our new opportunity is energy. The USDA has its sights set on applying its talent and technology to enhance energy production and energy efficiency in agriculture. Others should make a point of directing their talents and technologies to support this as well. This is the path forward for us. This is how we can move beyond a petroleum economy to make oil dependence a thing of the past, and safeguard our environment for future generations.

LITERATURE CITED

Intergovernmental Panel on Climate Change (IPCC). Climate Change 2007: The Physical

Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Summary for Policymakers, February 2007.

Intergovernmental Panel on Climate Change (IPCC). Climate Change 2007: Impacts,

Adaptation and Vulnerability, Working Group II Contribution to the Intergovernmental Panel on Climate Change Fourth Assessment Report, Summary for Policymakers, April 2007.

United States Department of Agriculture (USDA) and Department of Energy (DOE). Biomass

As Feedstock for a Bioenergy and Bioproducts Industry: The Technical Feasibility of a Billion-Ton Annual Supply, April 2005.

13

FUEL ETHANOL POLICY IN THE UNITED STATES: OVERVIEW AND POLICY ALTERNATIVES

Kelly J. Tiller*

Agricultural Policy Analysis Center, University of Tennessee, Knoxville, TN 37996

(865-974-7407) ktiller@utk.edu INTRODUCTION

The development of the US ethanol market has been influenced by a number of federal and state policy incentives and regulations. Primary policies encouraging rapid expansion in the domestic ethanol industry include: 1) an excise tax credit of 51 cents/gal, 2) an import tariff of 54 cents/gal and 3) a phased in requirement that a minimum annual quantity of ethanol, increasing to 7.5 billion gallons in 2012, be blended with gasoline consumed in the US. Many industry analysts suggest the fuel ethanol industry would not be competitive with the gasoline industry without these and other policy interventions encouraging ethanol production and consumption in the US. While these and other policies, along with favorable market conditions, have contributed to a more than doubling of corn based ethanol production capacity and consumption in recent years, the import tariff is scheduled to expire in 2008, the blender tax credit is scheduled to expire in 2010 and the renewable fuel standard is scheduled to flatline after 2012. There has been dialogue in US policy circles about the future direction of US policy regarding ethanol markets, as well as more general policy related to renewable and alternative fuels and energy. We will provide an overview of the complex set of policies and programs in place today that form the policy environment in which ethanol competes as a transportation fuel in the US. Also, we will examine some of the policy alternatives under consideration for future US energy policy.

CURRENT US ETHANOL POLICY

Policy support for the ethanol industry is not a new phenomenon. Following the 1970s oil embargo, US policymakers began to implement a number of policies and programs that would encourage development of a biofuels industry. The first major federal ethanol subsidy was introduced in the Energy Tax Act of 1978. At the time, the federal motor fuel excise tax was 4 cents/gal and the new ethanol policy provided a full exemption of the tax for ethanol. At blends of 10% ethanol and 90% gasoline (E10), the tax exemption was equivalent to 40 cents/gal of ethanol. As a result, the first 20 million gallons of commercial ethanol fuel production came online. Policies and subsidies, including policies designed to encourage and support the production of ethanol as well as expand the demand for ethanol, related to ethanol evolved during the next two decades (GAO 2005, Koplow 2006, Yacobucci 2007). Emphasizing the role of policy incentives in the history of the US fuel ethanol industry, a 1998 USDA report observed, “The fuel ethanol industry was created by a mix of Federal and State subsidies, loan programs and incentives. It continues to depend on Federal and State subsidies.”

14

Tax Credits

Prior to 2004, the primary federal incentive supporting the ethanol industry was a tax exemption that allowed blenders using 10% ethanol a 5.2 cents/gal exemption from the 18.4 cents/gal federal excise tax on transportation fuels. The 5.2 cent exemption was an effective subsidy of 52 cents/gal of ethanol, since it applied to 10% blends of ethanol with gasoline (E10). In 2004, the American Jobs Creation Act of 2004, comprehensive corporate tax overhaul legislation, replaced the partial tax exemption with a Volumetric Ethanol Excise Tax Credit (VETC) of 51 cents/gal of pure ethanol used in blending, and extended the sunset to December 31, 2010. One of the motivating factors for switching from a tax exemption to an excise tax credit was to shift the burden of the subsidy from a reduction in the Federal Highway Trust Fund to a reduction in the General Treasury. The tax credit is available for blenders of gasohol (ie, gasoline suppliers and marketers), up to a 10% blend of ethanol with 90% gasoline. The same 2004 American Jobs Creation Act established the first volumetric excise tax credit for producers of biodiesel. The tax credit is valued at $1.00/gal of agribiodiesel or 50 cents/gal of biodiesel from fats and recycled greases. The biodiesel tax credit was scheduled to expire in 2006 but was extended to December 31, 2008, in the Energy Policy Act of 2005.

In addition to the blenders tax credit, small ethanol producers with production capacity

below 60 million gal/yr are eligible for a producer tax credit. The credit is designed to encourage the expansion of small scale ethanol production facilities, which are owned locally by grower cooperatives. The credit, valued at 10 cents/gal of ethanol produced, may be claimed on the first 15 million gallons of ethanol produced annually. Many proponents and opponents of the ethanol industry agree survival is dependent on tax incentives (Yacobucci 2006b).

Renewable Fuel Standard

For the first time, a minimum required level of renewable fuel consumption was established in the Energy Policy Act of 2005. The Renewable Fuel Standard (RFS) requires the blending of renewable fuels in gasoline, beginning at 4.0 billion gallons of renewable fuel in 2006, increasing annual to a required level of 7.5 billion gallons of renewable fuel annually by 2012. While this was a benchmark, RFS was above production capacity at the time the policy was debated and passed, US production capacity already meets or exceeds the full RFS, well in advance of 2012. While the RFS does not function as a direct ethanol industry subsidy, it does stimulate industry growth and maturity by guaranteeing a minimum market demand for the product.

Import Tariffs Fuel ethanol imported into the US from most countries requires a most favored nation duty of 54 cents/gal and a 2.5% ad valorem tariff. The tariff offsets the domestic tax credit of 51 cents/gal for blending ethanol with gasoline. This fuel ethanol import tariff has significantly reduced US ethanol imports from countries with lower costs of production, most notably, ethanol derived from sugarcane produced in Brazil (Yacobucci 2006a). The import tariff was put in place in the Omnibus Reconciliation Act of 1980, amended in 1986 and extended by the Tax Relief and Health Care Act of 2006. The tariff is scheduled to expire December 31, 2008. A

15

trade agreement designed to promote growth and stability in the Caribbean region and Central America allows duty free imports of most products, including fuel ethanol, from the Caribbean Basin Initiative (CBI) countries, including Costa Rica, Jamaica and El Salvador. Given the exemption from the tariff, more than half of all fuel ethanol imports to the US come through CBI countries (Yacobucci 2006a).

Research and Development

Another important element of ethanol industry incentives is provided through a range of programs designed to expand the domestic ethanol industry, mostly working on the supply side to encourage research, development, demonstration and commercialization efforts. These take the form of grants, credit guarantees or other infrastructure or value adding factors that encourage the building of ethanol manufacturing plants. The Biomass Research and Development Act of 2000 led to a joint DOE-USDA program to coordinate biomass research, development and demonstration activities and developed a roadmap for future investments. Various related programs, currently authorized at $200 million/yr, provide grants for implementation. These programs take the form of biorefinery projects, loan guarantees for manufacturing facilities and basic research.

Other Programs and Incentives

There are a number of other programs and incentives that encourage ethanol industry expansion on both the supply side and the demand side. On the demand side, state mandates terminating the use of methyl tertiary butyl ether (MTBE) as an oxygenate in reformulated gasoline have contributed to increased demand for ethanol as a substitute for petroleum based MTBE. Government procurement preferences (either mandated or voluntary), designed to stimulate demand for renewable fuels and products, have become a standard demand enhancing policy tool for both federal and state governments. While they encourage the use of biofuels, often they are less effective than anticipated as they can be ignored in purchasing decisions if the renewable fuels are not readily available or are more expensive.

Various state and federal programs aimed at improving the availability of higher blend

ethanol fuels through infrastructure development also enhance demand. Manufacturers can use increases in the manufacture of flexible fuel vehicles designed to operate on higher blends of biofuels to offset higher corporate average fuel economy (CAFÉ) standards for cars and trucks sold in the US, further contributing to higher ethanol demand as more flexible fuel vehicles are produced. On the supply side, USDA programs that subsidize and support major program crops that serve as feedstock inputs to the ethanol and biodiesel industry help to increase the supply of ethanol. A number of individual states also support, either directly or indirectly, the development and expansion of the ethanol industry, through their own specific incentives, regulations and programs related to research, production and consumption of alternative fuels that supplements or exceeds federal incentives. Some of these state efforts aggressively promote the expansion of the biofuels market. For example, Minnesota established its own renewable fuels mandate prior to the federal RFS, requiring that all gasoline in the state contain at least 20% biofuels by 2013. In 2006, Iowa set a 25% petroleum fuels displacement requirement.

16

Table 1. Categorization and description of potential ethanol industry policies and incentives.

ETHANOL POLICY ALTERNATIVES

Future market conditions will influence the evolution of ethanol production and consumption. Production costs, market demand and prices will influence both the amount and rate at which the industry expands. Market behaviors will be influenced by consumers expectations and preferences regarding ethanol fuels. But given the strong role of policy in the development of the ethanol industry to date, and the significant level of attention and interest from a wide range of stakeholders, the future evolution of the domestic ethanol industry will continue to be largely policy driven. There are a range of policy instruments that may be applied in future ethanol industry policy. There are a number of ways to categorize and discuss policy alternatives which range from supply or capacity subsidies and policies to demand enhancing policies to regulatory requirements. Presented in Table 1 is a categorization of potential major policy directions and orientations. US policy changes marginally and incrementally over time. While the expectation is not that any of the particular “sets” of policies categorized below would move forward in isolation, they offer a framework for discussing various policy alternatives and the potential influence they may have on the future of the US ethanol industry.

Policy Approach Elements/ Orientation Potential Policy Vehicles Status Quo Extension of current programs

including tax credits, import tariffs, R&D, infrastructure

2007 Farm Bill; Energy Policy reauthorization; other legislative opportunities related to tax, commerce, trade or environmental policy

Market Stabilizing Variable rate market based subsidies, feedstock supply reserve

2007 Farm Bill; Energy Policy reauthorization; stand alone legislation

Free Market Expiration and/or elimination of market distorting subsidies and incentives

2007 Farm Bill (failure to include or attempt to water down)

Brazil Aggressive incentives and mandates to expand domestic production and consumption

2007 Farm Bill; Energy Policy reauthorization; stand alone legislation; other legislative opportunities related to tax, commerce, trade or environmental policy

Climate Change Carbon cap and trade, aggressive incentives toward next generation biofuels

Energy Policy reauthorization; stand alone legislation; other environmental reauthorization opportunities; other trade opportunities

National/ Energy Security

Domestic market protections, aggressive incentives to expand domestic production

Energy Policy reauthorization; 2007 Farm Bill; military reauthorizations/ appropriations; homeland security reauthorizations/ appropriations; stand alone legislation

17

LITERATURE CITED

Budny, D. 2007. The global dynamics of biofuels: potential supply and demand for ethanol and biodiesel in the coming decade. Woodrow Wilson International Center for Scholars, Brazil Institute Special Report, Issue No. 3. April.

Doering, O. 2006. Ethanol and energy policy. Purdue Extension Bioenergy Series ID-340, Purdue University Cooperative Extension Service. December. West Lafayette, IN.

Economic Research Service, USDA. 1988. Ethanol: economic and policy tradeoffs. Agricultural Economic Report No. 585, p. 2. April.

Government Accounting Office. 2005. Tax incentives for petroleum and ethanol fuels. RCED-00-301R. September.

Koplow, D. 2006. Biofuels – at what cost? Government support for ethanol and biodiesel in the United States. The Global Subsidies Initiative of the International Institute for Sustainable Development, Geneva, Switzerland. October.

Schnepf, R. 2007. Agriculture based renewable energy production. CRS Report RL32712, Congressional Research Service. January.

Tyner, W.E. and Quear, J. 2006. Comparison of a fixed and variable corn ethanol subsidy. Choices 21:199-202.

Yacobucci, B.D. 2006a. Ethanol imports and the Caribbean Basin Initiative. CRS Report RR21930, Congressional Research Service. March.

Yacobucci, B.D. 2006b. Fuel ethanol: Background and public policy issues. CRS Report RL33290, Congressional Research Service. October.

Yacobucci, B.D. 2007. Biofuels incentives: A summary of federal programs. CRS Report RL33572, Congressional Research Service. January.

18

DISC MILL ENERGY ISSUES

William R. Enterline*

Andritz Sprout, 7225 Sheffield Place, Cummings, GA 30040 (678-947-1588) william.enterline@andritz.com

Disc mills that incorporate replaceable cast grinding elements have been used in the corn wet milling process since early in the twentieth century. The Foos mill, a single disc attrition mill, developed by the Bauer Brothers Company (Springfield, OH) was applied to degermination circa 1910. Single runner attrition mills manufactured by Sprout Waldron Company (Muncy, PA) were applied to degermination around 1918. It was the Bauer Brothers development of the water tight double disc mill in the early 1930s that secured the use of disc mills as the preferred technology in the wet milling process. In terms of basic design and function, current models vary little from early machines, many of which are still operating today. A recent development is the use of large diameter single disc machines to replace the double disc machines in the third grind stage (fine grind) application. This trend began in the late 1960s and continues to the present. As with any rotating machinery, disc mills consume energy. Disc mills used for degermination applications operate at low speed and, because their purpose is to remove germ from the kernel without damaging or cutting individual germs, they are low energy consumers. Conversely, machines used for the third or fine grind application must rub starch and gluten from the fiber without reducing fiber size; therefore, more energy is required. In discussing energy issues, we will explore two basic areas: energy conservation and efficient energy utilization. Energy conservation encompasses items that directly affect the amount of energy consumed; we will discuss ways in which energy consumption can be reduced. Energy utilization is more process related and, for the purposes of this paper, may be defined as items that can improve performance by using the available energy in the most efficient manner. One of the largest energy consumers is water. Though corn solids typically are dewatered prior to each disc mill application in the process, we often see instances where additional flush water is added at the inlet to the disc mill to get material into the machine. In a new installation, this need for flushing water could be a matter of equipment orientation where the dewatered material must flow at a shallow angle into the mill and additional water is necessary to keep material moving towards the eye of the mill. Sometimes the problem is a lack of sufficient venting. The connection between the discharge of the disc mill and the tank below usually is tight to allow as little SO2 vapors as possible to escape onto the mill environment. The discharge tank must be vented to allow air displaced by the material to escape. Additionally, the disc mill acts as a fan, sucking air into the mill inlet and discharging it into the grind tank with the material. Depending upon size, speed and type of plates used in the disc mill, the air volume displaced can be significant. We have measured air volume and

19

static pressure generated by a 52 in disc mill equipped with conventional devil tooth plates operating at 1,750 rpm to be slightly in excess of 2,200 cfm at 2 in water gauge pressure. If the tank is not vented or if venting is insufficient, the tank and the refiner can become pressurized which can reduce ability of the disc mill to take the material. Often this situation is misdiagnosed as the problem and more flush water is introduced in an attempt to solve it. Instead of solving a problem, more energy consuming water is added to the process. Sometimes the reason for flush water is one of maintenance. Disc mills incorporate a flinger in the center of the rotating disc. These flingers are designed with radial blades that assist the material in “turning the corner” into the grinding area between discs. Though not a routine replacement item, they do wear; over time the blades wear to the point where the mill can no longer take material without use of flush water. Over worn grinding plates also can cause a need for flush water. When using flush water, either as an interim measure until maintenance can be performed or because of some other systemic circumstance, the absolute minimum necessary to keep material flowing will minimize energy use. The original Bauer Brothers and Sprout Waldron disc mill designs included integrally mounted motors where the motor rotor and the disc mill shaft were one piece. Later, when motor suppliers discontinued the manufacture of this type motor in smaller horsepower sizes, both companies converted to an overhead mounted motor (or two motors in double disc machines) with v-belt drive. To simplify this change, no additional design changes were made which meant the belt drive was mounted between inboard and outboard bearings of the mills. Since plate gap adjustment on these original designs is accomplished by moving the driving shaft axially and the driven sheave is mounted on this same shaft, the result is that the belt drive is never in alignment. Belt drives are less efficient than direct drives by 3%. With continual misalignment the belt drive is even more inefficient. Moreover, misalignment exacerbates belt wear, requiring more frequent belt changes and creating higher maintenance costs. Our modern degerminator design allows the belt drive to be mounted outboard of the bearings making belt changes easier but belt misalignment issues still exists. Fortunately, this design will accommodate a gear coupled direct drive that is more efficient, eliminating the need for frequent, costly belt changes and saving energy and maintenance cost. Until recently, disc mills used for degermination and third grind have been adjusted manually using a hand wheel and various types of mechanical linkages. This has meant that disc mill control could not be incorporated into the distributive control systems (DCS) found in most plants today. There are technologies available that make control and monitoring through a DCS possible. One such method is the Linear Voltage Distance Transducer (LVDT) System. This is a proven system, having been employed on disc refiners in the paper industry for many years and easily adapted to large diameter single disc mills on third grind and to modern degerminator designs. The system consists of an LVDT externally mounted to measure the travel of the adjusted rotating disc. A small, reversible electric gear motor equipped with a break and a built in variable frequency AC drive is attached to the adjusting linkage to provide movement necessary to adjust the plate gap. The reversibility of this drive allows the rotating disc to be

20

opened quickly in the event of a sudden overload, thus providing the same protection as the quick release lever on older Bauer mills. The LVDT is connected to a small display panel that is mounted on the machine or nearby. This panel provides a digital read out of the plate gap. Most importantly, a signal is provided by the panel that can be used in plant DCS systems to monitor operation, remotely adjust plate gap or, with additional programming, allow the disc mill to operate at a set horsepower with plate gap automatically adjusted to compensate for variations in the input stream. While we do not recommend operation to a set horsepower for degermination applications, there is a benefit in the energy intensive third grind application because control is taken away from the local operators, each of whom usually has his own ideas as to how the plate gap should be adjusted. By operating to a set point, available energy can be utilized to maximum benefit and overall performance enhanced. Applied to degermination, the LVDT system will allow remote monitoring and adjustment as well as allow historical data on plate gap and power to be accumulated and stored for future analysis. The last energy issue to be discussed relates to starting of large diameter direct coupled single disc refiners commonly used for third grind. These machines have a high inertial load that require a lot of energy to start the disc rotating and bring it to full speed. We recommend these machines be started using full voltage starters. In most instances, high voltage service is available, making full voltage starters a comparatively inexpensive means of starting (Walker 2006). On the negative side, there is a surge of current used in starting which can affect peak demand, thereby increasing power costs in some instances. High horsepower motors started across the line are hard on power distribution systems which can increase risk of premature transformer, breaker and bus failures due to excessive current draw. Where high voltage service is not available, an alternative may be the use of a digital soft start (Ghandi 2006). Many soft starts have current limit functions that flatten the acceleration curve of a motor during start up. Flattening the curve means the exposure to peak demand periods imposed by the utility can be reduced. This also results in reduced torque, which will help to reduce wear on mechanical components. Digital soft start starters are expensive and more involved compared to full voltage starters; however, they are alternatives and should be considered in instances where the situation warrants.

LITERATURE CITED Ghandi, C.I. 2006. Managing Director, Emco-Kimo Electronics Pvt. Ltd. Mumbai India.

E-mail exchange of October 31. Reference soft start starters for refiners. Walker, B. 2006. Electrical and Instrumentation Supt, Grain Processing Corporation,

Washington, IN. E-mail of October 13. Reference motor starting.

21

DRYING OF GRAIN RESIDUES AND SLUDGES USING BIOMASS FUELS

George Svonja*

Barr Rosin, Ltd., 48 Bell St., Maidenhead, Berkshire, SL61BR, UK

(44-1628-641700) george.svonja@barr-rosin.co.uk

INTRODUCTION Convective dryers directly fired with natural gas or contact dryers using steam are used for drying of coproducts from grain processing plants. Typical products are corn gluten feed, corn gluten meal, wheat feed and distillers dried grains with solubles (DDGS). Natural gas prices have soared in recent years, thus driving up costs for drying and process heating. Some waste burning technologies have been adapted to burn residues directly and recover some energy but in these systems much of the fuel value of the residues is wasted in evaporating water into the exhaust stack without regaining energy. We will describe a method of drying to low moisture content such that the product can be sold dry as animal food or burned for its energy value. At the same time, latent heat used to evaporate water from the residue is made available as process heat. This is a system which minimizes CO2 generation, whether the fuel is biomass or fossil fuel. Barr Rosin and solid fuel combustion specialists are working together to develop biomass fired dryer systems. Process plants can integrate these systems to reduce energy consumption and carbon emissions. The concept is to use flue gas from biomass combustion in a gas to gas heat exchanger to heat the drying medium in the ring dryer. This will replace natural gas fired systems and reduce carbon dioxide emissions from fossil fuels. To provide the correct process conditions for drying and to ensure an economically sized heat exchanger, flue gas is required to be more than 800ºC. This furnace exit condition is ideal for efficient combustion of biomass where the combustion temperature must be high enough to eliminate emissions of volatile organic compounds (VOCs), dioxins, carbon monoxide and other products of incomplete or low temperature combustion. Details of combustor design, heat exchanger and operating conditions are being finalized for commercial application. CONVENTIONAL DRYING AND EXHAUST GAS RECYCLE Most large industrial drying is done with cocurrent convective dryers where wet material and heated gas pass together through the equipment. High inlet gas temperatures can be used without damaging the input material because the surface moisture on the moist input stream rapidly evaporates at the dryer inlet, lowering gas temperature and protecting the particle surface. Air is the most commonly used drying medium, but it has become common practice to recycle dryer exhaust gas to recover heat. In this case, the drying gas contains superheated water vapor and its oxygen content is reduced. Most commonly, rotary dryers or pneumatic conveying (flash) dryers are used. A ring dryer system (Figure 1) is a special flash dryer which has been used widely in the grain processing industry to dry coproducts of starch and bioethanol production and, particularly, high quality DDGS.

22

The material to be dried often consists of dewatered suspended solids (wet grains) which have been concentrated in a centrifuge and mixed with wastewater (thin stillage) which has been concentrated in an evaporator. These are mixed together and conditioned with recycled dry product from the dryer to form a friable nonsticky material which can be fed into the dryer. A high speed disperser throws the material to produce a finely divided stream that is combined with the hot drying medium at the input venturi of the dryer. Much of the moisture in the input stream is flashed off. The drying medium conveys product around to the manifold, which is a centrifugal classifier. Wetter and heavier particles are recycled to the input point for additional residence time in the dryer while drier and lighter particles are separated from the drying gas in a series of cyclonic separators.

Figure 1. Gas recycle ring dryer (Barr Rosin, Ltd). When directly firing a dryer where combustion products are mixed with the drying medium, the practical limit of recycling exhaust gas is 70%. This limit is reached because combustion products form the rest of the drying medium. The dryer exhaust thus contains combustion products and water vapor. In dryers which process fermentation and other organic residue, some VOCs are evaporated into the drying gas stream with water vapor; these must be removed from the bleed off gas stream before discharged to the atmosphere. The usual method is to thermally oxidize VOCs to CO2 and water vapor by heating the gas stream to more than 800°C (1500°F). Typically, a regenerative thermal oxidizer (RTO) or a thermal oxidizer (TO) is used.

23

SUPERHEATED STEAM DRYING A further development of this principle is to recycle 100% of the drying medium by heating the dryer through a heat exchanger. In this case, combustion products are on the shell side of the exchanger and the drying medium is on the tube side. Because there is now no combustion gas in the dryer process circuit, the drying medium consists of superheated steam with a small quantity of incondensable gases and VOCs. The net evaporation of the dryer is vented by a pressure control system and can be used as a heat source for other processes. Typically, multiple effect evaporator systems concentrating wastewater are heated by condensing the dryer exhaust so that latent heat in the dryer exhaust is recovered. Effectively, the system is both a dryer and a boiler. The small amount of incondensable gas left in the exhaust after condensation can be oxidized thermally in the air heater of the dryer, thus eliminating the need for a separate oxidation system. A system comprising an indirectly heated ring dryer operating in a superheated steam atmosphere and connected to a waste heat recovery evaporator is shown in Figure 2. To oxidize remaining VOCs, the remaining small flow of uncondensed gas is fed into the combustion chamber of the air heater.

Figure 2. Superheated steam ring dryer (Barr Rosin, Ltd).

24

A superheated steam ring dryer is shown in Figure 3. The ring dryer can be seen on the upper left with its two product collection cyclones and the product recycle and conditioning system below. In the foreground at the bottom is the main recirculating fan with the heat exchanger to the right. Evaporative capacity 37 ton/hr Heat input 29 MW Steam Output 34 ton/hr

FLUIDIZED BED COMBUSTOR The indirectly heated dryer can use a number of different fuels without contaminating the product because combustion products are not in contact with the material to be dried. In particular, solid biomass fuels can be used, but it will be possible to use any type of solid fuel to heat the dryer. To provide maximum flexibility, it is necessary to use a combustor which can handle a wide range of fuels; the most flexible is a fluidized bed combustor. Other types of combustors, such as chain grate stokers or pulverized fuel burners, can be used but the fluidized bed is recognized to be the most flexible (Figure 4).

Sand & Ash to screening & recycle

Fuel & AdditiveReagent hoppers

Start up Burner

Fluidising &Combustion air fan

Hot Gas to process & gas cleaning

Typical Fluid Bed Combustor

Figure 4. Fluidized bed combustor.

Figure 3. Superheated steam ring dryer. (Barr Rosin, Ltd)

25

The previously mentioned types of burners have been used to burn biomass but most biomass combustion systems have been built as waste reduction systems with energy production as a coproduct. Biomass in these systems is fed into the combustor at a high moisture content, typically 30 to 60% moisture. This means water in the fuel is evaporated in the combustion system and the latent heat to do this evaporation is lost in the combustor exhaust. Some sensible heat is recovered at the boiler exhaust but the water vapor must be condensed to make use of the latent heat. This is not practical when the vapor has been mixed with incondensable combustion products. The effect of moisture on caloric value of biomass is depicted in Figure 5.

Figure 5. Heat available in wet DDGS.

To maximize the heat available in the biomass and hence reduce the amount of CO2 produced per net unit of useful heat produced, biomass must be as dry as possible. Burning dry fuel has the following advantages: higher net caloric value and less fuel usage; dry fuel is easier to store and meter which gives better combustion control; fuel is stable and will not mold in storage; higher efficiency; lower CO2 emissions; smaller gas cleaning systems; lower exhaust volume and less visible exhaust plume. COMBINATION OF DRYER AND BOILER If we combine the process steps discussed above with the energy center of the plant, an energy efficient and flexible fuel system can be installed. A system for drying DDGS from an ethanol plant is depicted in Figure 6. DDGS consists of residual material from fermenting grain into ethanol. DDGS is dried to 10% moisture so it can be stored and metered. The dried material has a caloric value of 18 to 19 mJ/kg and, more importantly, a net caloric value of 17 to 18 mJ/kg.

0

2

4

6

8

10

12

14

16

18

20

0% 10% 20% 30% 40% 50% 60%

Moisture Content (wb)

mJ/

kg

26

Condensate

Exhaust to Atmosphere

Combustor

Ring Dryer

Wet Cake

DDGSTo storage

Syrup

Baghouse

HEX

Combair

SO2Adsorption

NOxReduction

WasteHeatEvap

Incondensable gases

DDGSfrom storage

Superheated Steam

Whole Stillage

Thin Stillage

Boiler

Energy Integration, DDGS-Fired Ethanol Plant

Process Engineering DivisionBarr-Rosin

Figure 6. Energy integration; DDGS fired ethanol plant.

One third of the fuel is used to dry DDGS and produce the vapor removed as steam which is used to preconcentrate the wastewater (thin stillage) from ethanol production into syrup. The syrup is mixed with dewatered solid waste (wet grains) prior to drying. The solid fuel combustor heats both the boiler and the dryer. At start up, when there is little or no material to dry, the boiler works at nearly maximum capacity to supply all steam requirements of the plant. However, as soon as the dryer starts to operate it provides steam and the boiler output can be turned down. The combustor does not turn down but supplies excess heat to the dryer as the boiler turns down. At normal operating conditions, the dryer will supply up to 40% of the steam demand of the ethanol plant. The small amount of incondensable gases coming out of the evaporator shell carry VOCs and are passed into the combustor to be oxidized thermally. This system also allows the process operator to minimize risk and maximize flexibility. If the market for DDGS as an animal food increases, then another biomass material can be used in the combustor, as can a wide range of fossil fuels in the event of short supply of biomass.

27

PLATE HEAT EXCHANGER DESIGN

John Robertson*

Alfa Laval, 5400 International Trade Dr., Richmond, VA (804-236-1259) john.robertson@alfalaval.com

The shell and tube heat exchanger (S&T) has been the standard heat exchanger within the dry grind and wet milling industries for many years. However, the plate heat exchanger (PHE) has replaced these units in a wide variety of applications. This is due mainly to its benefits of flexibility, compact footprint and high heat transfer capabilities. One major problem with converting customers from an accepted technology to a new technology is applying the old S&T standards to the newer PHE technology. For example, a standard velocity in a S&T for fouling application may be in excess of 6 ft/sec; whereas, 2 ft/sec may yield a better result in the PHE. In designing a PHE, mathematical formulae used for design purposes are the same as those used for the S&T. However, design considerations for a PHE are different. Therefore, we should design the PHE in a manner that allows us to take advantage of its unique capabilities. The physical properties of the fluids used in both channels of the heat exchanger are key factors in the design of the PHE. At a minimum, density, specific heat, thermal conductivity and viscosity of the fluids are required. It is best if this information is available at different points along a temperature profile. The primary objective of the PHE is heat transfer. Yet, one of the most critical factors is ensuring the media can pass through the channel. Suspended solids are common in a number of applications; PHE manufacturers have adapted their products to handle these solids. A PHE can pass solids that are 60 to 70% the size of the plate gap. Early PHE models with a narrow gap that averaged 2 to 3 mm were effective and efficient on clean fluids. Later, the plate gap increased to an average medium gap of 4 to 5 mm. Today, gaps can range up to 17 mm. Thus, there is a wide range of choices; selecting the most appropriate plate gap for specific applications leads to additional considerations. As a result of the laminar flow pattern in an S&T, any PHE will provide better heat transfer than an S&T. Turbulent flow generated by the corrugated plate pattern reduces laminar film thickness along the wall. Therefore, two of the three methods of heat transfer, conduction and convection, take place between the plates. A comparison of the same volume of water passing through a narrow gap PHE and a wide gap PHE demonstrates lower heat transfer in the wide gap PHE due to increased hydraulic diameter. The increased hydraulic diameter reduces velocity and turbulence created by the plate corrugation. However, this reduced performance can be offset by increasing the velocity of the water at the cost of pressure drop across the exchanger. This increased velocity will reduce laminar film thickness along the wall of the plate and provide more turbulent flow needed to allow increased convection within the plate channel.

28

Pressure drop across a PHE is the price we paid for the high heat transfer coefficients. Therefore, we must balance the pressure drop with the energy cost of the pump. Additionally, the pressure drop available needs to be used in an effective manner. In a PHE, there will be a loss of pressure in the port, plate neck and plate channel. Generally, the unit is designed to maintain 70% of available pressure drop through the plate port and neck. This ensures good fluid distribution across the plate pack. If the pressure drop in the ports and neck of the plate exceeds more than 30%, there is a risk of maldistribution and reduced heat exchanger performance. In essence, this would short circuit the exchanger by not using effectively the available heat transfer area. Thus, selecting a PHE with the correct port size is important and not as simple as adapting the pipe connection to the heat exchanger with a bell housing. In most clean services, a pressure drop of 10 to 15 psig should yield an effective design. Units in a fouling application should be designed with a higher pressure drop, as this will help keep the unit clean. As mentioned earlier, there is a correlation between high velocity and pressure drop. High velocity creates higher turbulence within the plate and helps reduce fouling. The force of the flow on the heat exchanger wall is measured as Tau, or shear stress. As a rule of thumb, we design the PHE to have a shear stress in excess of 50 Pa. This helps prevent precipitation of material on the plate and keeps solids in suspension. In regards to shear stress, a longer plate requires a higher pressure drop to achieve the same shear stress compared to a shorter plate. Also, when changing from one pass to two pass designs, the pressure drop needs to be doubled to achieve the same shear stress. Additionally, higher pressure drop allows us to design the PHE for a higher thermal duty; this is called thermal length. There are three ways to design the PHE for the required thermal duty. One involves the plate design. Most PHE plates come in two models: low theta and high theta. The low theta plate (Figure 1) will have a chevron angle less than 90 degrees. The high theta plate (Figure 2) will have a chevron angle greater than 90 degrees.

Figure 1. Plate design with low theta angle. Figure 2. Plate design with high theta angle.

Using low theta plates will result in a lower pressure drop, residence time and heat transfer. Conversely, using only high theta plates will result in a higher pressure drop, residence time, and heat transfer. It is possible to combine low and high theta plates in a single plate pack that results in a medium theta. These plate designs are available in almost all plate pressing depths (gaps) mentioned earlier. The overall plate length will affect the thermal duty of the heat exchanger.

29

If the design cannot be achieved using the various plate types, the final alternative is to use a multi pass plate configuration. An example is shown below (Figure 3). This is not uncommon, especially when using the wide gap PHE. To keep suspended solids in solution, a channel velocity in excess of 2 ft/sec is required. Therefore, a large thermal duty may require a multipass configuration.

Cold in Hot out Hot in

Cold out

Figure 3. Multipass plate.

Each application is unique and may create additional variables. However, these components (plate design, length and configuration) are consistent throughout most applications. Designing a PHE is a process of balancing all of these components and providing an economical solution that provides the required heat transfer. Each plate manufacturers’ product portfolio is different, consequently, it is difficult to compare competing PHE designs strictly on square footage. It is imperative these considerations be evaluated in the design process. As energy costs increase, the demand for heat recovery from processes will continue. The PHE is capable of closer approach temperatures than an S&T and has a smaller footprint. Additionally, as metal costs continue to reach new highs, reduced area requirements of the PHE will increase their demand.

30

ETHANOL AS AN ECONOMIC COMPETITOR TO GASOLINE

Andrew J. McAloon*

Eastern Regional Research Center, ARS, USDA,

600 E. Mermaid Lane, Wyndmoor, PA 19038 (215-233-6619) andrew.mcaloon@ars.usda.gov

ABSTRACT Fuel ethanol is one of the technology success stories of the 21st century. In less than one third of a century, it has gone from being a material produced rather inefficiently in small quantities to a major commercial product. This success can be attributed not only to the fact that ethanol is a renewable fuel and replaces imported oil, but also because it is capable of competing with petroleum based products on a cost basis under the right economic conditions. Ethanol’s current competitiveness is a result of improvements in the technology of producing it and conditions in the economic environment in which it competes. INTRODUCTION Fuel ethanol is produced from starch in the corn kernel. Technology improvements in the production of ethanol and changes in the automobile fuels market have helped transform ethanol from a material produced rather inefficiently in small quantities to a 15 billion l/yr energy source produced at more than 100 US facilities. Fuel ethanol is a commodity; one measure of its success is how well it compares on price with competing commodities. We will address this issue and how it arrived at its current status. The modern ethanol industry developed as a result of the attempted OPEC oil embargo in the early 1970s. These efforts by OPEC, which created long lines at gas stations, ultimately failed but did create awareness in this country that we were dependent on a fuel supply originating from other countries that may not have our best interests in mind. This resulted in the US government taking steps to encourage the development of an ethanol industry. The ethanol produced by this new industry came from ADM wet milling facilities and several dry grind plants. These early dry grind plants were small by today’s standards, some were farm based, most were inefficient and many of them had short production lives. Ethanol production costs in this early stage of the industry were reported to be about $1.50/gal in 1980 dollars or, adjusted by the Gross Domestic Product (Chained) Price Index Deflators (Budget of the US Gov't. 2006) $3.06/gal in 2005 dollars. Today the future of the ethanol industry is much brighter thanks to three major factors: 1) involvement of supporters of the ethanol industry, 2) reduction in the cost of producing ethanol and 3) changes in the ethanol marketplace.

31

INVOLVEMENT OF SUPPORTERS OF THE ETHANOL INDUSTRY Supporters of the ethanol industry are staff and members of organizations such as the Renewable Fuels Association (RFA) and the National Corn Growers Association (NCGA), the management of companies such as ADM and many others. They created a public awareness of ethanol and helped define in the public’s mind that ethanol is an oxygen additive for reformulated gasoline and a replacement for gasoline. Through lobbying for regulatory decrees, ethanol markets were created. Supporters successfully advocated for the replacement of MTBE with ethanol as an oxygenate additive to gasoline. In its early days, efforts succeeded in giving ethanol a cost advantage by obtaining the present $0.51/gal federal blenders tax credit as well as state tax credits for ethanol producers and blenders. These efforts bought time for the ethanol industry to become competitive with gasoline. REDUCTIONS IN THE COST OF PRODUCING ETHANOL Ethanol production costs can be grouped into cost categories for feedstock, energy, labor, capital and operating supplies. Each of these areas has experienced cost reductions which have helped lower the cost of ethanol. Feedstock Costs Corn is used to produce more than 95% of our ethanol. In the early 1980s we averaged 2.5 gal ethanol/bu corn, while today we are averaging 2.75 gal/bu. In the 10 year period bracketing 1980 (1976 to 1985), farmers were averaging yields of 101 bu/acre, but by 2005 yields had increased to 148 bu/acre (USDA 2007). In this same time period, national average corn prices were $5.13/bu for corn in 2005 dollars or $2.51/bu in 1980 dollars. In 2005, corn prices were $1.90/bu although 2006 prices ranged from $2.00 to 3.50/bu. Some agricultural economists believe $3.00/bu may be a new long term average price. Energy Costs The second largest cost in producing ethanol is energy, which has been reduced by using molecular sieves, automated process control systems and heat integration. Phil Madsen, President of Katzen Engineering, in an interview with Biofuels Journal, described being in an ethanol plant in 1986 that used 120 lb steam/gal ethanol produced; he went on to discuss a project that Katzen Engineering is designing that will use 18.5 lb steam/gal ethanol (Question and Answer 2006). Bryan and Bryan International (BBI) reported average utility costs in 1980 were $0.20/gal ($0.41/gal in 2005 dollars; Novozymes North America 2002). Utility costs were reduced to $0.25/gal by 2005 (Kwiatkowski et al 2006). Labor Costs Knowledge gained through operation of ethanol facilities has resulted in plants with less downtime and lower maintenance costs. In addition, average ethanol plant sizes have grown from less than 5 to more than 50 million gal/yr capacity with process control systems that require

32

only a minimal increase in the number of plant operators. Plant labor costs have decreased from $0.37/gal (2005 dollars; Novozymes North America 2002) in 1980 to less $0.05/gal in 2005 (Kwiatkowski et al 2006). Capital Costs Improvements in the amount of ethanol produced from a bushel of corn have resulted in smaller front end equipment sizing and increases in plant capacities have resulted in lower capital costs per gallon ethanol produced. BBI reported that ethanol capital costs were $2.45/gal in 1980 which is greater than $5.00/gal in 2005 dollars (Novozymes North America 2002). In 2005, ethanol plants could be built for less than $2.00/gal annual capacity (Kwiatkowski et al 2006). Operating Supplies Technical and operational successes have lowered the cost of operating supplies. One example is the cost for enzymes which, on a performance basis, has experienced a 70% cost reduction during the last 25 years (Novozymes North America 2005). Total Operating Costs Changes in the cost of producing ethanol in the last 25 yr are illustrated in Table 1. Ethanol production costs in 1980 were reported to be about $1.50/gal in 1980 dollars, or $3.06/gal in 2005 dollars. In 2005, when corn was selling at $2.00/bu, ethanol production costs were $1.20/gal. Today, with corn costs at $3.50/bu, it costs $1.80/gal in direct production costs (Kwiatkowski et al 2006). Table 1. Ethanol production costs 1980 to 2005.

CHANGES IN THE ETHANOL MARKETPLACE The automobile fuel market has changed during the last 25 yr and advances in technology and corresponding cost reductions in producing a gallon of ethanol have been greater than cost reductions in contemporary gasoline production. Approximately 75% of the retail cost of gasoline is attributable to production charges and the remaining 25% to distribution charges and taxes.

1980 Cost 2005 Cost Item Unit 1980 ($) 2005 ($) 2005 ($)

Capital costs $/gal annual capacity 2.45 5.00 2.00 Corn cost $/bushel 2.51 5.13 1.90 Corn cost $/gal 1.00 2.05 0.69 Utility costs $/gal 0.20 0.40 0.25 Labor costs $/gal 0.18 0.37 0.05 All other changes $/gal 0.12 0.24 0.21 Production costs $/gal 1.50 3.06 1.20

33

In 1980, the average retail price for unleaded gasoline was $2.95/gal in 2005 dollars ($1.25/gal in 1980 dollars) and the production cost of gasoline would have been $2.20/gal in 2005 dollars. In 2005, the average retail price for unleaded gasoline was $2.30/gal and production cost was $1.75/gal (US DOE 2006). In 1980, ethanol cost $3.06/gal to produce in 2005 dollars, and in 2005, based on $1.90/bu corn, these costs were about $1.20/gal. Even at today’s very high corn prices of $3.50/bu the cost of producing ethanol is $1.80/gal. CONCLUSIONS Ethanol produced from $3.00/bu corn could sell for $1.75/gal with a $0.25/gal allowance for overhead, profit and marketing charges and has a cost advantage if the cost of gasoline at the refinery gate is greater than this. The impact of the 30% lower energy level (and miles/gal driven) of ethanol compared to gasoline also should be considered and on a miles/gal of fuel driven ethanol must be priced at 70% of the cost of a gallon of gasoline. There were occasions in 2005 when average retail prices for gasoline were $2.30/gal, production costs were about $1.75/gal and average petroleum prices were $56/barrel. The cost of producing ethanol from corn was 68% of the cost of producing gasoline at $1.20/gal and corn was trading at less than $2.00/bu. Ethanol was less expensive to make than gasoline.

ACKNOWLEDGEMENTS The author wishes to acknowledge Kevin Hicks, David Johnston and Hosein Shapouri for their input into this paper.

LITERATURE CITED Budget of the United States Government. Fiscal Year 2006. Table 10.1 Gross Domestic

Product and Implicit Outlay Deflators. Washington, DC. Kwiatkowski, J.R., McAloon, A.J., Taylor, F. and Johnston, D.B. 2006. 40 MGY dry grind

ethanol process and cost model: modeling the process and costs of fuel ethanol production by the corn dry-grind process. Ind. Crops Prod. 23:288‑296.

Novozymes North America and Bryan & Bryan, Inc. 2002. Fuel Ethanol Production: Technological and Environmental Improvements. June.

Novozymes North America and Bryan & Bryan Inc. 2005. Fuel Ethanol, A Technological Evolution, BBI International. Page 10.

Question and Answer. 2006. Phil Madson, President, Katzen International Inc. 2nd Quarter Biofuels J. Pages 88-89.

US DOE. 2007. Energy Information Agency website. Accessed January 29. Annual Energy Review 2006, Table 5.24 Retail Motor Gasoline and On-Highway Diesel Fuel Prices, Selected Years, 1949-2005, http://www.eia.doe.gov/emeu/aer/txt/stb0524.xls.

USDA. 2007. National Agricultural Statistics Service website. Accessed January 29. US and All States Data-Corn Field. http://www.nass.usda.gov/QuickStats/PullData_US.jsp.

34

PRODUCTION OF ETHANOL AND DDGS FROM BARLEY CONTAINING REDUCED BETA-GLUCAN AND PHYTIC ACID

Mian Li*1, Pauline Teunissen1, Gerhard Konieczny-Janda1, Kevin B. Hicks2,

David B. Johnston2, John Nghiem2 and Jay K. Shetty1

1Genencor International, A Danisco Company, 2600 Kennedy Dr., Beloit, WI 53511 and 2Eastern Regional Research Center, USDA, ARS,

600 E. Mermaid Lane, Wyndmoor, PA 19038 (608-363-6467) mian.li@danisco.com

INTRODUCTION

Ethanol derived from renewable feedstocks has potential to meet one of the greatest challenges to today’s society as a sustainable replacement for fossil fuels, especially in the transport sector, with reduction in greenhouse gas emissions. In 2005, a record 4 billion gallons of fuel ethanol was produced in the US. Currently, there are 113 ethanol plants in operation with the capacity of 5.6 billion gal/yr; another 84 ethanol plants, under construction or expanding, could add 6.1 billion gal annual capacity (WSJ, Feb 14, 2007). In the US, corn is the primary feedstock for fuel ethanol production. In 2006, for example, 20% of the US corn supply was used to make fuel ethanol to replace only 3 to 4% of the gasoline supply. To avoid the “fuel vs food” issue, an alternative to corn feedstock is needed. Barley has potential as an alternative feedstock for ethanol production, especially in the Mid Atlantic region and states where it is a winter crop, allowing double cropping with soybeans. In North America, barley can provide at least 1 billion gal/yr, which was 20% of the total US ethanol production in 2006.

However, there is no plant in the US using barley as a feedstock because regular hulled

barley can not be processed in a conventional corn dry grind plant without modifications due to: 1) the abrasive nature of hulled barley which would damage grain handling and grinding equipment, increasing capital costs, 2) the lower starch content of barley (50 to 55%) would result in lower ethanol yields compared to corn, requiring barley plants to be built larger than corn plants for the same capacity, 3) the high viscosity of barley mashes due to beta-glucan and 4) the production of a distillers dried grains with solubles (DDGS) coproduct with high levels of beta-glucan cannot be used in poultry, swine and aquaculture diets, which limits the value of the coproduct. For a barley to ethanol process to be successful, these technical hurdles must be overcome. The objective of this paper is to review a barley based STARGEN™ process for ethanol production.

MATERIALS Hulled barley (Thoroughbred, Lot 1504-1, grown in 2005) was obtained from the Virginia Foundation Seed Center Farm at Mt. Holly, VA. Characteristics (Table 1) of hulled barley were determined by the USDA Eastern Regional Research Center (ERRC). Commercial Trichoderma reesei OPTIMASH™ BG (beta-glucanase), acid stable alpha amylase, STARGEN™ 001 (granular starch hydrolyzing enzymes), FERMGEN™ (protease) were from Genencor International, A Danisco Company (Beloit, WI).

35

Table 1. Chemical and physical characteristics of hulled barley.

EFFECT OF BETA-GLUCANASE ON VISCOSITY REDUCTION

Using barley for ethanol production, barley mash viscosity would become a major issue at higher solids levels due to its beta-glucan content. The high viscosity of barley mash makes agitation, liquefaction, saccharification and fermentation difficult and adds to operating costs. Therefore, nonstarch hydrolyzing enzymes, such as cellulase and beta-glucanase, are required for reduction of viscosity to acceptable levels. The beta-glucanase tested was OPTIMASH™ BG, which contains a combination of enzymes that modify and digest nonstarch carbohydrates.

Barley mash was made at 30% dry solids content and adjusted to pH 3.6. After mixing,

slurry was transferred to the measuring tube of the Haake Viscotester VT550. The Viscotester was preheated to 57°C. OPTIMASH™ BG was added directly at the start of viscosity measurement. The control was run with no OPTIMASH™ BG. The Viscotester was started and allowed to run for 90 min at 57°C. After 90 min, the temperature was lowered to 32°C (fermentation temperature) and the Viscotester was kept running for an additional 30 min. OPTIMASH™ BG helped reduce viscosity of barley mash (Figure 1). For the control, the Viscotester could not reach 57°C test temperature, as the rotor stopped at 54°C, indicating the control mash was too viscous for measurement.

Figure 1. Effect of OPTIMASH™ BG on barley viscosity reduction (30% dry solids, pH 3.6).

Moisture, % (ground kernels) 7.85 Ash, % 2.32 Oil, % 1.92 Starch, % 59.89 Protein, % 7.60 Beta-glucan, % 3.90 Acid detergent fiber, % (ADF) 5.47 Neutral detergent fiber, % (NDF) 17.22 Crude fiber, % 4.66 Bulk density, lb/bu 52.94

0100200300400500600700800900

1000

0 2000 4000 6000 8000

Time (s)

Visc

osity

(mP

asS

)

0

10

20

30

40

50

60

Tem

pera

ture

(C)

36

GRANULAR STARCH HYDROLYZING ENZYME (GSHE) FOR BARLEY FERMENTATION

In a typical dry grind ethanol process, the entire grain is ground (Falling Number AB, KT-30, Stockholm, Sweden) and processed without separating the various components of the grain. The milled grain is slurried with water. After alpha-amylase and beta-glucanase are added, the slurry is cooked at 58 to 60°C to reduce viscosity of the barley mash. The slurry is cooked at high temperature (85 to 88°C) to gelatinize and liquefy starch in a process called liquefaction. Higher temperatures also reduce microbial levels in the resulting mash. After liquefaction, mash is cooled and a secondary enzyme (glucoamylase) is added to convert liquefied starch to fermentable sugar (glucose) in a process called saccharification. Yeast (Ethanol Red, Fermentis, Marcq-en-Baroeul, France) is added to the mash to ferment sugar to ethanol and carbon dioxide. This process is called fermentation. The conventional barley to ethanol production process is illustrated in Figure 2. In general, this is an energy intensive process that requires the addition of heat energy to starch granule slurries until the gelatinization temperature of the starch is exceeded.

Fresh water

Milled Barley

Evaporation condensate

Alpha Amylase

Beta glucanase

30-33°C

Steam

Saccharification enzyme/Protease

58 - 60°C

Thin stillageAcid

Beta Glucanase

Urea

Steam

Conventional Barley Process

LiquefactionMixing

Fermentation

85-88°C

1.5 hour

85-88°C

1.5 hour

Figure 2. The conventional barley dry grind ethanol process.

Granular starch hydrolyzing enzymes (STARGEN™) were used in a low energy process that hydrolyzed starch in the granular (uncooked) state. The technology has potential to eliminate the need for high energy processing of starch and provide more cost effective production of glucose for conversion to ethanol and other bioproducts and biomaterials. Because of the ability to conduct several grain processing steps simultaneously (liquefaction, saccharification and fermentation) in the same vessel, the process also has the ability to lower equipment and capital costs in an ethanol facility. The STARGEN™ line of products includes blends of enzymes that have synergistic activities on granular starch. The blend includes an

37

alpha-amylase and a glucoamylase that can drill holes in starch granules or peel starch granules, depending on the substrates. We applied this new enzyme technology to barley fermentation.

Ground hulled barley slurry (27 to 30% dry solids) was prepared and adjusted to 3.6 pH using sulfuric acid. OPTIMASH™ BG was added to the slurry at a dosage equivalent to 0.2 kg/ton of grain and acid stable alpha-amylase at 0.13 kg/ton (57°C, pH 3.6) (Table 2). The slurry was placed in a 57°C water bath for 1.5 hr. During incubation, the slurry was stirred gently with an overhead mixer. Barley starch was not gelatinized at 57°C, which is below the gelatinization temperature for barley. Viscosity problems were not observed in this step. HPLC analyses, supernatant Brix and percent solubilization of hulled barley are listed in Table 3.

Table 2. Viscosity reduction conditions.

Table 3. HPLC profile, Brix and % solubilization of hulled barley mash.

Simultaneous saccharification and fermentation (SSF) was carried out with addition of 400 ppm urea. At each dosage, fermentations were run in triplicate. The enzymes added were 1.56 kg/ton STARGEN™ 001 and 0.1 kg/ton FERMGEN™. At various time intervals, samples of the beer were removed for HPLC analyses.

Fermentation finished in 45 to 50 hr, producing 11.80% v/v ethanol (Figure 3). In

another experiment, hulled barley (30% dry solids) was used (Figure 4). Again, there was no viscosity issue. Glucose concentrations were low (0.048 to 0.067%) during fermentation (Table 4) which would enhance the active yeast population and limit growth of undesirable microorganisms. Direct conversion of granular starch using STARGEN™ enzymes allowed high gravity fermentation with low soluble solids. This reduced osmotic stress on the yeast and resulted in higher ethanol concentrations and higher throughput in the final distillation step. Lower osmotic pressure exerted also resulted in yeast producing lower levels of waste byproducts such as glycerol; reduced glycerol production enables more glucose to be converted to ethanol.

Enzyme Dose Acid stable alpha amylase kg/ton 0.13

OPTIMASH™ BG kg/ton 0.2

DS% % DP1 % DP2 % DP3 % HS Brix % Solubilization 27 19.78 20.90 8.80 50.53 6.5 28.4 30 11.43 15.00 12.36 61.21 7.0 27.1

38

Fermentation of 27% DS Hulled Barley: 1.56 kg/MT STARGEN, 0.1 kg/MT FERMGEN

02468

101214

0 10 20 30 40 50 60

Hours

% V

/V E

than

ol

Figure 3. Production of ethanol at 27% dry solids hulled barley (1.56 kg/ton STARGEN, 0.1 kg/ton FERMGEN)

02468

10121416

0 10 20 30 40 50 60 70 80

Hours

% V

/V E

than

ol

0246810121416

% D

P1

Figure 4. Production of ethanol at 30% dry solids hulled barley (1.56 kg/ton STARGEN, 0.1 kg/ton FERMGEN).

39

Table 4. HPLC analyses during fermentation with STARGEN™ 001.

In addition to STARGEN™ and FERMGEN™ at the above dosages, addition of

OPTIMASH™ BG at 0.1 kg/ton during the fermentation had no effect on ethanol yield (Figure 5). However, adding OPTIMASH™ BG in the SSF step may have the benefit of further reducing mash viscosity, thus improving downstream processing.

Figure 5. Effect of OPTIMASH™ BG on ethanol yield at 30% dry solids hulled barley (1.56

kg/ton STARGEN, 0.1 kg/ton FERMGEN, with and without 0.1 kg/ton OPTIMASH BG). A process using STARGEN™ enzyme technology, capable of hydrolyzing insoluble

granular (uncooked) starch into fermentable sugars by enabling depolymerization of starch to glucose in a SSF process, offers several potential benefits for ethanol production (Figure 6). With STARGEN™ enzyme, jet cooking is eliminated which would result in significant energy savings. In addition, STARGEN™ process resulted in higher ethanol yield than conventional process (Table 5). For the STARGEN™ process, 0.538 kg ethanol/kg starch can be obtained, corresponding to a fermentation efficiency of 95.8%.

Hours % W/V % W/V % W/V % W/V % W/V % W/V % V/V DP>3 DP3 DP2 Glucose Lactic

Acid Glycerol Ethanol

17 2.209 0.344 0.588 0.067 0.044 0.444 7.65 24 2.038 0.301 0.587 0.066 0.045 0.551 9.67 40 1.850 0.239 0.571 0.067 0.057 0.705 13.08 48 1.844 0.236 0.595 0.065 0.053 0.744 13.28 65 2.080 0.000 0.572 0.052 0.039 0.726 13.79 70 2.062 0.000 0.559 0.048 0.019 0.733 13.75

02468

10121416

0 10 20 30 40 50 60

Hours

% V

/V E

than

ol

No OPTIMASH BGwith OPTIMASH BG

40

Fresh water

Milled Barley

Evaporation condensate

Acid alpha amylase

Beta Glucanase

30-33°CpH 3.3

Steam

57°CpH 3.71.5 hour

STARGEN™ 001

57°C30-40 min

pH 3.7

Thin stillage

Protease

Urea

STARGEN™ Process

Figure 6. Low energy ethanol production process.

Table 5. Comparison of hulled barley fermentation.

CHARACTERIZATION OF DDGS FOR RESIDUAL STARCH, BETA-GLUCAN AND PHYTIC ACID

After fermentation, the beer was dried in a forced air oven to obtain DDGS. Residual starch, beta-glucan and phytic acid contents were determined (Table 6). The conventional process resulted in less than 1% residual starch, while STARGEN™ process resulted in 2.5% residual starch, indicating excellent conversion of the starch during SSF. Beta-glucan content of hulled barley was 3.90% and residual beta-glucan level after SSF was 0.3 to 0.4%. More than 95% of beta-glucan was hydrolyzed resulting in DDGS with low levels of beta-glucan.

Table 6. Residual starch and beta-glucan contents in DDGS.

Ethanol % v/v Standard Deviation %

Conventional Process 14.60 0.08

STARGEN™ Process 14.87 0.06

Residual Starch % beta-glucan %

STARGEN™ Process 2.51 0.37

Conventional Process 0.96 0.39

41

Dry grind fermentation of corn normally results in DDGS containing high levels of phytic acid. This is undesirable from an animal diet formulation point of view because the phosphate present in phytate is unavailable due to the limited digestibility by nonruminant animals. Therefore, a significant amount of phosphorus is deposited on soil from the unused phytate carried in the manure. This has been a concern in some countries due to environmental pollution from animal waste, especially from swine and poultry. Interestingly in the barley STARGEN™ process, due to the hydrolysis of phytic acid, presumably by endogenous barley phytase during the viscosity reduction step at 57°C for 1.5 hr, the resulting DDGS from yeast fermentation is free from phytic acid (the phytic acid in hulled barley samples was 0.36%). Therefore, the barley STARGEN™ process is able to produce DDGS with reduced beta-glucan and no phytic acid.

CONCLUSIONS

Treating barley slurries with OPTIMASH™ BG can reduce viscosity problems associated with slurries containing high levels of beta-glucan. The reduction in viscosity can resolve problems with pumping and processing the mash. Advantages of using nonstarch hydrolyzing enzymes and STARGEN™ enzyme technology for barley fermentation were demonstrated: higher ethanol yields, DDGS containing reduced beta-glucan and no phytic acid, elimination of jet cooking, less capital equipment and less energy. Lower concentrations of fermentable sugars in the fermenter resulted in enhancing active yeast population and, along with lower pH during SSF, limited growth of undesirable microorganisms. Use of STARGEN™ enzymes with nonstarch viscosity reducing enzymes provide ethanol producers more tools to help in processing grain to ethanol while increasing total plant yield and throughput.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the assistance from Gerry Senske, Karen Kohout and Mike Kurantz from Eastern Regional Research Center, ARS, USDA and Oreste Lantero, Kees-Jan Guijt, Brad Paulson, Jim Miers and David Bates from Genencor International.

42

SOURCES OF VARIATION IN DRY GRIND PROCESSING STREAMS

Ronald L. Belyea1*, Kent D. Rausch2, M. E. Tumbleson2 and Ganti S. Murthy3

1University of Missouri, Columbia, MO, 65211(573-882-6354) belyear@missouri.edu 2University of Illinois at Urbana-Champaign, Urbana, IL 61801 and

3Oregon State University, Corvallis, OR 97331

ABSTRACT Distillers dried grains with solubles (DDGS) contain high protein and fat concentrations, making them a valuable ingredient for ruminant production diets. DDGS are associated with inherent variation, which detracts from quality and makes accurate diet formulation difficult. The causes of variation have not been well documented.

From a series of studies, we examined potential sources of variation. We concluded that much of the variation is among fermentation batches, due to subtle differences in processing equipment or methods. Within batch variation appears to be less than among batch variation, although within batch variation could vary from plant to plant. Corn and growing conditions contributed little to variation in composition of DDGS. INTRODUCTION The high nutritional value of coproducts from corn processing has been recognized since at least the mid 1900s. Morrison (1948) summarized analytical and feeding trial data for a vast array of feedstuffs, including distillers dried grains with solubles (DDGS). He found DDGS to be a valuable ingredient in ruminant production diets because of high protein and fat contents. He also issued the caveat that DDGS composition could vary markedly and this variation could impact animal productivity. Starting in the 1980s, inclusion of DDGS, as well as other coproducts, in production diets began to increase. Consequently, considerable amounts of analytical data have been reported (Akayezu et al 1998, Arosemena et al 1995, Bath et al 1981, Belyea et al 1989, Belyea et al 1998, Belyea et al 2004, Belyea et al 2005, Belyea et al 2007, DePeters et al 1997, Shurson et al 2001). Data from these studies documented magnitudes of variation in composition associated with DDGS. Examples from several sources are summarized in Table 1. Table 1. Variation in DDGS composition (g/100 g db) from various sources.

1Lignocellulose (Goering and Van Soest 1970). 2N × 6.25.

Source Fiber1 Protein2 Fat Ash Shurson et al (2001) 14-19 28-32 10-12 5.2-6.7 Akayezu et al (1998) 14-17 29-32 9-12 3.0-5.2 Belyea et al (2004) 16-19 28-33 11-13 4.3-5.0 DePeters et al (1997) 16-22 29-30 9-11 4.1-4.8 Arosemena et al (1995) 16-22 15-30 9-11 4.1-4.8 Belyea et al (2007) 18-33 26-37 5-15 2.6-5.1

43

While there are extensive data in the literature on the extent of variation in DDGS composition, causes were not addressed. Partly, this is because much of the data represented small numbers of samples (sometimes only one sample) and because processing conditions either were not known or not documented. In addition, processing technologies have changed. Prior to the mid 1990s, most corn was processed by wet milling, in which the corn kernel is fractionated into components. In the mid 1990s, dry grind processing came on line; in this process, the entire corn kernel is ground and fermented into ethanol. (For a more detailed discussion of corn processing technologies, see Rausch and Belyea 2006). Processing streams from dry grind processing could differ substantially from streams produced by earlier technologies. In dry grind processing, fermentation essentially is a batch process; variation in composition of streams could be due to variation within and/or among batches. There were no published data to examine the role of fermentation batches on composition of streams in dry grind processing. The objective was to evaluate potential sources of variations in composition of dry grind processing streams. Standard statistical methods were used for analysis of variance and mean comparisons in all three experiments.

MATERIALS AND METHODS Experiment 1. Multiple Plant Study Samples of DDGS and other processing streams were obtained from nine dry grind plants during four periods (roughly, seasons) of three weeks each. This resulted in 108 samples of each stream (12 per plant). Each sample represented one specific fermentation batch and processing condition. Samples were processed and analyzed for fiber, protein, fat, ash and elements. Experiment 2. Multiple Year Study

Corn and DDGS samples were collected during a 5 yr period at one dry grind plant; there were more than 200 samples which represented a variety of growing conditions and corn varieties. Samples were analyzed for nutrient concentrations; relationships among nutrients were determined. Climatic and growing conditions also were included in the analysis.

Experiment 3. Within Plant Study

We collected DDGS samples and four other processing streams from two dry grind plants. Samples were obtained every 15 min during a 4 hr period (10 samples per stream or 50 samples per plant). These were processed and analyzed for fiber, fat, protein and ash concentrations. The impetus of this project was to address within batch variation. RESULTS AND DISCUSSION Experiment 1. Multiple Plant Study There were significant effects of processing plant on concentrations of most nutrients in DDGS and parent processing streams. This is illustrated in Figure 1, which shows variation in protein and fat concentrations for 9 plants × 3 weeks within each plant. DDGS protein concentrations of samples varied from 27 to 37% across plants and weeks; within plants, variation was somewhat smaller. DDGS fat concentrations varied considerably; the extreme range was 5 to 15%, with most plants varying from 10 to 14%. Most other major nutrients had similar amounts of variation. DDGS mineral concentrations also varied among plants; for

44

example, DDGS phosphorus concentrations varied from 5,000 to 9,000 mg/kg (Figure 2). Because of environmental concerns about phosphorus in animal wastes and phosphorus loading of cropland, the variation in DDGS phosphorus presents additional challenges to diet formulation.

Figure 1. DDGS protein and fat concentration by plant × week.

Figure 2. DDGS phosphorus concentration by plant × week. In the dry grind process, whole stillage is the residue after ethanol has been removed; it is characterized by high water content. Whole stillage is centrifuged to form wet grains (suspended particulate matter) and thin stillage (soluble liquid matter). Thin stillage is dewatered partially to form syrup (condensed distillers solubles). Syrup is added to wet grains, which are dried to form DDGS. Both wet grains and syrup can have variable nutrient concentrations, as illustrated for protein content in Figure 3, probably due to incomplete separation at the centrifuge step. In addition, when the two streams are recombined to form the DDGS stream, proportions are not

0

5

10

15

20

25

30

35

40

45

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9Plant * week

g/10

0 g

dm .

Protein Fat

0100020003000400050006000700080009000

10000

1 2 3 4 5 6 7 8 9

Plant * week

mg/

kg d

m

P

45

controlled consistently. Thus, variation in DDGS composition can be due to several sources and not easily managed.

There were significant variations in DDGS nutrient composition and parent streams

both within and across processing plants. While these effects were confounded with batches, batches appeared to play a significant role. The processing plants in this study were constructed quite similarly and used similar equipment and processing methods. However, there can be subtle differences in processing conditions from plant to plant, as well as from batch to batch within a plant. Thus, the variation observed was not surprising. This is complicated by lack of data in the literature to identify sources of variation either within or among fermentation batches.

Experiment 2. Multiple Year Study

In this experiment, the relationships (correlations) among nutrient concentrations in corn and corresponding nutrients in DDGS were low (less than 0.30) and not statistically significant (Table 2). There also did not appear to be any effect of year (agronomic/climatic conditions) on corn or DDGS composition (data not shown; Belyea et al 2004).

Figure 3. Protein concentrations of wet grains and syrup by plant for three weeks (SR =

syrup; WG = wet grains; WK=indicate weeks 1, 2 and 3).

0

5

10

15

20

25

30

35

40

45

1 2 3 4 5 6 7 8 9Plant

g/10

0 g

dm

SR WK1 SR WK2SR WK3 WG WK1WG WK2 WG WK3

46

Table 2. Correlations among corn and DDGS constituents.1 -------------------------------------------------------------------------------------- DDGS constituents ------------------------------------ Corn constituents Fat Protein Starch --------------------------------------------------------------------------------------- Fat -0.15 -0.06 0.11 Protein 0.04 0.15 Starch -0.21 --------------------------------------------------------------------------------------- 1Belyea et al (2004).

Experiment 3. Within Plant Study

In this experiment, there were differences in compositions of streams between the two processing plants. Wet grains solids concentrations were higher for plant 2 than plant 1 (Figure 4); variation also was greater for plant 2 than plant 1. The protein content of wet grains and DDGS were uniform for plant 1; for plant 2, wet grains had greater variation in protein content than DDGS (Figure 5). Fat contents of wet grains and DDGS were similar and uniform for plant 1; for plant 2, fat content of wet grains was lower and more variable than for DDGS (Figure 6). Other nutrients displayed similar patterns. Plant 1 had consistent solids and nutrient concentrations within batch, while plant 2 had more variation. In plant 2, wet grains had more variability than DDGS. This suggests variation at the centrifugation step; the cause is not known. Variation within batches may be dependent on conditions in the processing plant and the amount of batch to batch variation could vary from plant to plant.

Figure 4. Wet grains solids concentration for two dry grind plants (WG=wet grains; PL1, PL2 indicate Plant 1 and 2, respectively).

05

101520

253035

4045

1 2 3 4 5 6 7 8 9 10Sample no.

g/10

0 g

sam

ple

WG PL1

WG PL2

47

2022242628303234363840

1 2 3 4 5 6 7 8 9 10Sample no.

g/10

0 g

dm

WG PL1 DG PL1

WG PL2 DG PL2

Figure 5. Protein concentrations of DDGS and wet grains from two dry grind plants (WG=wet grains; DG=DDGS; PL1, PL2 indicate Plant 1 and 2, respectively).

Figure 6. Fat concentrations of wet grains and DDGS for two dry grind plants (WG=wet grains; DG=DDGS; PL1, PL2 indicate Plant 1 and 2, respectively).

0

2

4

6

8

10

12

14

1 2 3 4 5 6 7 8 9 10Sample no.

g/10

0 g

dm

WG PL1 DG PL1

WG PL2 DG PL2

48

Figure 7. Protein concentrations in corn by plant.

Figure 8. Phosphorus concentrations of corn and DDGS by plant.

0100020003000400050006000700080009000

10000

1 1 1 2 2 2 3 3 3 4 4 4 5 5 5 6 6 6 7 7 7 7 8 8 9 9 9Plant * week

mg/

kg d

m

DDGSCorn

0

2

4

6

8

10

12

2 4 6 8Plant*week

g/10

0 g

dm

49

CONCLUSIONS

Variation in DDGS composition is due, in part, to variation in composition of corn, which varies with corn variety and/or growing conditions. We found no evidence to support either of these concerns. In experiment 1, the protein (Figure 7) and phosphorus (Figure 8) contents of corn were not different among processing plants; however, phosphorus (Figure 8) and protein (Figure 1) DDGS concentrations DDGS varied significantly. Most nutrient concentrations in corn were similar to published data (NRC 1982), suggesting the composition of commercially grown corn has not changed during the past several decades. In experiment 2, the relationships among concentrations in corn and nutrient concentrations were low. Corn or growing conditions contribute little to variation in DDGS composition.

Variation in DDGS composition occurs within and among processing plants, strong evidence for variation among fermentation batches as a major source of variation. Within batches, variation appears to depend on the plant; some plants have more within batch variation. DDGS composition is unrelated to constituents in corn, hybrids or growing conditions.

LITERATURE CITED

Akayezu, J-M., Linn, J.G., Harty, S. and Cassady, J.M. 1998. Use of distillers grains and

coproducts examined. Feedstuffs 70:11-13. Arosemena, A., DePeters, E.J. and Fadel, J.G. 1995. Extent of variability in nutrient

composition within selected by-product feeds. Anim. Feed Sci. Technol. 54:103-120. Bath, D.L. 1981. Feed by-products and their utilization in ruminant diets. in: Upgrading

Residues and By-products for Animals. CRC Press, Boca Raton, FL. Belyea, R.L., Clevenger, T.E., Singh, V., Tumbleson, M.E. and Rausch, K.D. 2006. Element

concentrations of dry grind corn processing streams. Appl. Biochem. Biotechnol. 134:113-128.

Belyea, R.L., Eckhoff, S.R., Wallig, M.A. and Tumbleson, M.E. 1998. Variation in the composition of distillers solubles. Biores. Technol. 66:207-212.

Belyea, R.L., Rausch, K.D. and Tumbleson, M.E. 2004. Composition of corn and distillers dried grains with solubles from dry grind processing. Biores. Technol. 94:293-298.

Belyea, R.L., Steevens, B.J., Restrepo, R.R. and Clubb, A.P. 1989. Variation in composition of by-product feeds. J. Dairy Sci. 72:2339-2345.

Belyea, R.L., Rausch, K.D. and Tumbleson, M.E. 2007. Unpublished data. DePeters, E.J., Fadel, J.G. and Arosemena, A. 1997. Digestion kinetics of neutral detergent

fiber and chemical composition within selected by-product feeds. Anim. Feed Sci. Technol. 67:127-140.

NRC. 1982. United States-Canadian Tables of Feed Composition. 3rd rev. Nat. Res. Council. Nat. Acad. Press, Washington, D.C.

Rausch, K.D. and Belyea, R.L. 2006. The future of coproducts from corn processing. Appl. Biochem Biotechnol. 128:47-86.

Shurson, J., Spiehs, M., Whitney, Baidoo, S., Johnston, L., Shanks, B. and Wulf, D. 2001. The value of distillers dried grains with solubles in swine diets. Pages 22-52. in: Minn. Nutr. Conf. and Minn. Corn Growers Assoc. Symp. Bloomington, MN.

50

AN OVERVIEW OF US SORGHUM STARCH AND ETHANOL PRODUCTION

Jeff Dahlberg*

National Sorghum Producers Association, 4201 N. Interstate 27, Lubbock, TX 79403

(806-749-3478) jeff@sorghumgrowers.com INTRODUCTION Sorghum [Sorghum bicolor (L.) Moench] is the second most important animal food crop in the US after maize (Zea mays L.). The production value of sorghum was estimated to be $871 million in 2006 and was planted on 6.52 million ac in the US with an average yield of 1.58 ton/ac (USDA 2007). Sorghum is used primarily as an animal food grain in the US; 12% of the current domestic crop is used in grain to ethanol production. In 2005/06, 50% of the domestic crop went to the export market, 37% to domestic animal food and residual uses and 13% to food, seed and industrial uses (US Grains Council 2007). REVIEW OF SORGHUM STARCH As with all cereal grains, starch is the major component of sorghum. On a weight basis, 50 to 75% of the sorghum grain is starch (Rooney and Serna-Saldivar 2000). Starch is located in endosperm (both vitreous and floury) and kernel pericarp, which is an unique feature of sorghum (Rooney and Serna-Saldivar 2000). Sorghum starch granules range from 2 to 30 μm in diameter. Starch granules in the corneous endosperm are polygonal and smaller than those in the floury endosperm, which are more round in shape (Serna-Saldivar and Rooney 1995). Amylose and amylopectin molecules are held together by hydrogen bonds in a highly organized manner and are arranged radially in spherical granules (Waniska and Rooney 2000). The kernel vitreous part with the contents of an endosperm cell are shown in Figure 1. Also, there are polygonal starch granules, protein bodies (p) and lack of air spaces in the structure.

Figure 1. Sorghum kernel (Courtesy Scott Bean, USDA).

51

Gelatinization temperatures of sorghum starch have been reported to vary from 71 to 80°C (Sweat et al 1984) with starch isolated from corneous endosperm of sorghum having a higher gelatinization temperature than that from the floury endosperm (Cagampang and Kirleis 1985). Corneous endosperm starch has a higher intrinsic viscosity and lower iodine binding than floury endosperm (Cagampang and Kirleis 1985). Starch from normal grains contains 23 to 30% amylose while that from waxy sorghum has less than 5% amylose. Waxy sorghum starch differs in its properties compared to normal starch and has higher peak viscosity, higher water binding and more resistance to gel formation and retrogradation (Serna-Saldivar and Rooney 1995). Digestibility of waxy sorghum starch is higher than normal sorghum starch (Rooney and Pflugfelder 1986). SORGHUM USE IN THE ETHANOL INDUSTRY Sorghum traditionally is grown in what is referred to as the “Sorghum Belt” (Figure 2). Typically, Kansas and Texas have been the largest production states. Sorghum has been used regularly and interchangeably with corn in ethanol production in Nebraska and Kansas facilities. One ton of maize produced 387 L of 182 proof alcohol, while the same amount of sorghum produced 372 L (Coble et al 1981). With refinements in technologies and enzymes, a bushel of sorghum now produces an equivalent amount of ethanol to a bushel of maize. As demand for ethanol has increased in the US, plants have begun to move into the sorghum belt (Figure 3).

Figure 2. Sorghum planted acres by County (USDA 2005).

52

Figure 3. Ethanol biorefinery locations within the US (RFA 2007).

Approximately 80% of sorghum production lies within a 50 mile radius of an ethanol plant (Table 1). Sorghum grain is brought into a plant and goes through a typical dry grinding process before conversion to ethanol. DDGS typically have higher protein and less fat than corn DDGS. Currently there are 114 ethanol plants producing 5.6 billion gal/yr ethanol with 80 plants under construction or expansion which will produce an additional 6.4 billion gal/yr, for a total of 12.0 billion gal/yr. As these plants move into the sorghum belt, researchers have begun to evaluate how differences in sorghum starch may effect ethanol production. Kansas State University scientists reported that starch content in sorghum varies considerably but even within high starch sorghums, variability exists in starch content vs ethanol yield and fermentation efficiency (Figures 4 and 5). Table 1. Locations with an ethanol plant and sorghum grown within a 50 mile radius. Location Bushels

Liberal, KS 37,917,300 Garden City, KS 42,728,700 Hastings, NE 22,752,500 York, NE 12,575,100

53

Figure 4. Sorghum starch content and ethanol yield (D. Wang, Kansas State University).

Figure 5. Sorghum starch content and ethanol conversion efficiency (D. Wang, Kansas State University).

FUTURE RESEARCH NEEDS The National Sorghum Producers Association is working with various universities within the sorghum belt and the USDA/ARS to coordinate an effort to evaluate starch content of various commercial sorghum hybrids grown in state yield trials under varying environments and irrigation schemes to develop a database of sorghum starch. As more ethanol plants are moving into the sorghum belt, which traditionally represents harsh environmental conditions, a database of hybrids that show good starch and ethanol conversion efficiencies under various environmental conditions will be valuable in assisting producers in their choice of sorghum hybrids as an ethanol feedstock. Initial research at Kansas State University (KSU) was indicative that even within sorghum starch there is variability in ethanol efficiency and conversion rates. This research is being expanded to understand factors that make these starches different. Sorghum proteins are

54

known to cross link when heated. Researchers at KSU and the USDA/ARS have shown the extent of these cross links formed during ethanol production is related to fermentation efficiency of sorghum. Future research into altering cross linking during fermentation may lead to improvements in the time required to produce ethanol from sorghum. Likewise, improvements in processing the grain prior to fermentation may lead to higher ethanol yields from sorghum. For example, removing the outer layer (bran) from sorghum before it undergoes fermentation can result in up to 20% increases in ethanol yield due to higher starch loading in the fermenter. In addition, DDGS produced from such sorghum is high in protein (up to 50% db) and lipids which may improve food values (Corredor et al 2006). Fermentable sugars can be liberated from sorghum bran, possibly providing a means for ethanol plants to process both grain and bran for ethanol in the future (Corredor et al 2007). Research will need to be expanded in the area of irrigation effects and drought on starch and proteins of grain sorghum and how this will effect ethanol production. In conjunction with this, the development of rapid methods to predict fermentation performance of sorghum hybrids will enable scientists to screen breeders’ samples for the purpose of identifying high yielding hybrids and breeding specifically for these properties. With the addition of 500 million gal/yr ethanol production in the Texas Panhandle, these research issues will be critical to ensure the continued use of sorghum in the ethanol industry.

LITERATURE CITED

Cagampang, G.B. and Kirleis, A.W. 1985. Properties of starches isolated from sorghum floury and corneous endosperm. Starch 37:253-257.

Coble, C.G., Hiler, E.A., Sweeten, J.M., O’Neal, H.P., Reidenback, V.C., LaPori, W.H., Schelling, G.T. and Kay, R.D. 1981. Small scale ethanol production from cereal feedstocks. Page 611. in: Cereals: A Renewable Resource. Y. Pomeranz and L. Munck eds. American Association of Cereal Chemists, St. Paul, MN.

Corredor, D.Y., Bean, S.R., Schober, T. and Wang, D. 2006a. Effect of decorticating sorghum on ethanol production and composition of DDGS. Cereal Chem. 83:17-21.

Corredor, D.Y., Bean, S.R. and Wang, D. 2007. Pretreatment and enzymatic hydrolysis of sorghum bran. Cereal Chem. (In press).

Renewable Fuels Association. 2007. http://www.ethanolrfa.org/objects /documents/plantmap070129.pdf Accessed Mar 20. Rooney, L.W. and Pflugfelder, R.L. 1986. Factors affecting starch digestibility with special

emphasis on sorghum and corn. J. Anim. Sci. 63:1607-1623. Rooney, L.W. and Serna-Saldivar, S.O. 2000. Sorghum. Pages 149-176. in: Handbook of

Cereal Science and Technology. K. Kulp and J. G. Ponte, eds. 2nd edition. Marcel Dekker, New York, NY.

Serna-Saldivar, S.O. and Rooney, L. W. 1995. Structure and chemistry of sorghum and millets. Pages 69-124. in: Sorghum and Millets: Chemistry and Technology. D.A.V. Dendy, ed. American Association of Cereal Chemists, St. Paul, MN.

Sweat, V.E., Faubion, J.M., Gonzalez-Palacios, L., Berry, G., Akingbala, J.O. and Rooney, L.W. 1984. Gelatinization energy and temperature of sorghum and corn starches. Trans. ASABE. 27:1960-1963, 1969.

55

USDA-NASS. 2005. http://www.nass.usda.gov/Charts_and_Maps/Crops_ County/2005/Maps/as-pl.asp Accessed March 20.

USDA-NASS. 2007. http://www.nass.usda.gov/Data_and_Statistics/Quick_ Stats/index.asp#top Accessed Mar 20.

US Grains Council. 2007. http://www.grains.org/page.ww?section=Barley%2 C+Corn+%26+Sorghum&name=Sorghum Accessed March 20.

Waniska, R.D. and Rooney, L.W. 2000. Structure and chemistry of the sorghum caryopsis. Pages 649-688. in C.W. Smith and R.A. Fredericksen, eds. Sorghum: Origin, history, technology and production. John Wiley & Sons, Inc. New York, NY.

56

ETHANOL REALITY CHECK

Rodney J. Fink*

Western Illinois University, 1 University Circle, Macomb, IL 61455 (309-833-5192) rjfink@macomb.com

ABSTRACT

Ethanol is an alcohol, the same as in beer and wine. It is made by fermenting biomass high in carbohydrates through a process similar to beer brewing. Today, ethanol is made from starches and sugars, but National Renewable Energy Laboratory (NREL) scientists are developing technology to allow it to be made from cellulose and hemicellulose, the fibrous material that makes up the bulk of most plant matter. Ethanol is used primarily as a fuel additive for vehicles to increase octane and reduce carbon monoxide and other smog-causing emissions (NREL 2007). The industry outlook for 2007 reports 4.9 billion gallons of ethanol produced in 19 US states from 110 biorefineries. This exceeded the previous year’s production by more than one billion gallons (RFA 2007). Fifteen new ethanol plants came on line in 2006; the addition of these biorefineries, including completion of expansion projects, added 1.05 billion gallons of capacity for the year. In addition, 2006 ended with 73 biorefineries under construction and 8 expanding that will add 6 billion gallons of new production capacity by 2009 (RFA 2007). Much of the industry growth is focused on corn producing regions; however, considerable new construction in 2006 was in states like Arizona, Oregon, Texas and New York.

The US and Brazil are leading producers of ethanol while China, India and other nations are expanding their domestic ethanol industries. In Brazil, ethanol is added to gasoline at a level of 25% compared to 10% in the US. More than 90% of US ethanol is made from corn; in Brazil, sugar cane is the main feedstock for ethanol production. In addition to corn and sugar cane, alcohol, hydrogen, methanol, vegetable oils and various other combinations are being tested or are already in use for blending into Brazil’s gasoline supply (BBI 2004). Other sources of biomass are being considered for making ethanol; Fink and Fink (2005) showed biomass availability in Missouri. They measured Btu components of numerous feedstocks, on a county basis, and included crop residues from corn and other crops (Figures 1 and 2). The largest total amount of residues came from timber thinnings; however, acquisition of these residues requires extensive equipment and may be expensive. Corn residues constituted the largest output of crop residues available in the state. Similar data are available for other states and should be important considerations when cellulosic ethanol becomes more common. The net energy balance of corn ethanol has been questioned, although USDA scientists report the net energy balance of corn ethanol, adjusted for coproduct credits is 27,729 and 22,196 Btu/gal for wet milling and dry grinding, respectively, and 30,528 Btu/gal for the industry. The study suggests that corn ethanol is energy efficient, as indicated by a positive energy output/input ratio of 1.67 (Shapouri and McAloon 2004). David Pimentel, a Cornell agricultural ecologist reached other conclusions and suggested that making a gallon of ethanol from corn required 29% more energy, from fossil fuels, than a gallon of ethanol can provide. At the same time, he said that ethanol has only two thirds of the energy content of a gallon of gasoline (Pimentel 2003).

57

17%

36%

7%

1%2%

30%

3%

4%

0% Total Crops

Switchgrass

Landfill Gas

Animal Wastes

Timber Harvest

Timber-Thinning

Timber Mill

Tires

Municipal Solid Waste

Figure 1. The breakdown of BTU components in crop residue and processing feedstocks available in Missouri (clockwise in order listed). Missouri Department of Natural Resources

1%0%3%28%

4%13%

46%

5%

CornSorghumWinter WheatRiceSoybeansCottonCotton Gin WasteFescue Seed

Figure 2. Ranking of biomass availability from crop residues in Missouri (clockwise in order listed). Missouri Department of Natural Resources. Grain ethanol is produced by wet milling or dry grinding processes with the main difference being the initial grain processing of the plant. In wet milling, corn is steeped and separated into its component parts. Only the starch component is sent to fermentation. Construction and operating costs of a wet mill are higher than a dry grind plant (often 2 to 5 times more initial capital) and vary due to the coproducts produced. The added coproduct stream can justify higher capital investment as the marketable coproducts are ethanol or sweeteners, corn oil, corn germ, corn gluten meal, corn gluten feed (wet or dried), CO2 and yeast. The common conversion from wet milling is 2.5 gal

58

ethanol/bu of corn and from dry grinding the yield is 2.7 to 2.8 gal/bu. Dry grind production constitutes 82% of the US ethanol production; wet mills produce the remaining 18%. Farmer owned ethanol plants account for about half of the fuel ethanol plants and produce 40% of fuel ethanol. Ethanol has these advantages:

• Ethanol burns cooler than gasoline. • Ethanol is an oxygenate and when added to gasoline creates better combustion and

reduces carbon monoxide in the exhaust. • Ethanol production has reduced US surplus corn supply and contributed to an increase in

corn prices. • Ethanol is a renewable resource. • Ethanol’s coproduct, distillers dried grains with solubles (DDGS), is high in protein and

useful in animal diets.

RFA estimates 2006 ethanol uses are formulated gasoline (68%), discretionary blending (26%) and winter oxygenate program (6%). In 2006, American corn farmers produced 10.74 billion bushels of corn of which 1.8 billion went to the production of ethanol, representing 17% of total US corn production. Ethanol represents the third largest market, after animal food use and exports, for domestically produced corn. Also, ethanol production utilized 26% of the nation’s sorghum crop in 2006.

DDGS is an important coproduct of the ethanol industry. Used primarily in beef and

dairy cattle markets, growing percentages are used in swine and poultry production (Commodity Specialists Co.). Inclusion rates are 46, 42, 3 and 10% DDGS for dairy, beef, poultry and swine. In addition to domestic livestock use, DDGS is a component of exports to countries such as China and Japan. The production of DDGS has increased from 2.7 million metric tonnes in 2000 to more than 12 million metric tonnes in 2006. Continued increases in production of ethanol will have a major impact on disposition of DDGS and may facilitate more integration of livestock operations with ethanol plants to reduce the high cost of drying and transporting DDGS. Corn use is changing rapidly as its use for ethanol production increases. Some livestock producers have contacted the US Secretary of Agriculture about their concerns. The Earth Policy Institute warned that ethanol is on track to consume half of the US corn crop as early as 2008. They warned about a merging of the food economy and the energy economy. Corn consumed in the US in 2001 was 9.8 billion bushels compared with 11.8 billion bushels in 2006. The pattern of corn grain use shows this trend during the past 5 years.

Source: USDA Economic Research Service

US Corn Utilization % Consumed, 2001 % Consumed, 2006 Ethanol 7% 18% Food 13% 12% Livestock Feed 61% 51% Exports 19% 19%

59

INTEGRATION OF OPERATIONS There has been a great deal of creativity in the development and implementation of business plans for ethanol plants. The Sunshine Energy Cooperative of Blairstown, IA, is a good example of such an operation (Fink 2001). The Cooperative was formed in 1995 and began operating in 1999. The original plan, developed by the parent organization (Iowa Beef Cooperative), was to operate an integrated business that included an ethanol plant as the hub of an integrated energy farm to produce ethanol, heat, carbon dioxide and animal food. Heat was to be used in an aquaculture facility for raising commercial fish such as tilapia. Carbon dioxide was to be furnished to a commercial greenhouse and high protein coproduct (DDGS) for cattle in an adjacent feedlot. Methane was to be captured from anaerobic digestion of cattle manure and used as a supplemental source of energy for ethanol production. Digested manure was to be sold on site or used as a fertilizer or as a cellulosic feedstock for ethanol production. The directors of the Cooperative used the success of Reeves Agri-Energy Corporation of Garden City, KS (integrated ethanol, feedlot and tilapia operation) as an example. The Sunshine Energy Cooperative elected not to employ a process design company as they felt they had the expertise to initiate and manage the plant properly without added expense. Even though the plant had support from the Iowa Department of Natural Resources and the Iowa Department of Economic Development, they encountered many problems on startup including some early licensing problems, a dispute with a general contractor ending up in arbitration and a judgment against the cooperative. The end result was a startup cost well above the general average for new construction costs (more than $2.00/gal of capacity). Financial difficulties caused the plant to go through one bankruptcy, one change of ownership and, finally, a foreclosure before being acquired by Xethanol Biofuels, LLC (Close 2004). Integration of operations creates complex management challenges to owners and operators. The integrated version of the Blairstown plan had appropriate planning, with input of private and public entities, but overall management problems hindered the success of the operation. It may have been an operation ahead of its time. After the recent fluctuations of ethanol and corn prices, brought about by increased use of ethanol, and phased out MTBE, some synergies exist for current integration potential. With ethanol production projected to reach 15 billion gal/yr by 2015, the amount of DDGS to be dried and transported increases, causing a potential decline in price and increased transportation costs. Using Iowa as an example, 2005 ethanol production in the state was 900 million gallons which would produce enough DDGS to feed roughly 1.08 million dairy cows (20% of dry matter intake of DDGS). Iowa currently has 190,000 dairy cows in the state (Babcock and Hart 2006). This increase is within the realm of possibility as 1.08 million dairy cows would produce 15% of total US milk production. At least three synergies could occur as a result of bringing dairy (or beef cattle) into Iowa to consume the DDGS (Babcock and Hart 2004). These synergies are:

• Locating cattle close to an ethanol plant so DDGS would not have to be dried; this would save a 50 mmgy ethanol plant $5.0 million/yr in drying costs.

• 1.08 million dairy cattle would generate vast amounts of manure for crop fertilization. • Working together, the dairy farmer and ethanol plant could capture methane from the

manure before it is applied to farm fields. The manure from one dairy cow can generate 3,170 kWh of energy/yr. Capturing methane could reduce odor problems.

60

As the nation places more reliance on ethanol as an alternative fuel, a greater amount of DDGS will be produced which will need to be dried and transported to the point of use. As more reliance is placed on ethanol, for example, the price of corn will become more sensitive to changes in quantity produced (bigger decreases in long crop years and larger increases in short crop years). As more ethanol plants are built, the synergies of transporting grain (corn) to the site of DDGS utilization vs transporting DDGS to the site of consumption will need to be evaluated. MARKETING ETHANOL AND COPRODUCTS As more corn is used for fuel, the amount of grain available for food use decreases. As demand rises, so will the price. Countries that will suffer are poor countries, such as Mexico, who import large amounts of grain. This trend has been apparent in the past year as even in a bumper crop year, the price of corn has increased (Figure 3). The price of ethanol has had major fluctuations, as the year 2006 reveals. In January, 2006, the CBOT daily cash ethanol was $1.80/gal, rising to a peak of $4.25 in July, 2006. The July peak accompanied the purchase of ethanol by major formulators when legislation to “hold them harmless” from any problems associated with MTBE failed in Congress. In October, 2006, ethanol was as low as $1.70/gal and closed the year at $2.25/gal. Thus a fluctuation of $2.50/gal occurred in 2006. Corn as well fluctuated from $2.00 to $4.00/bu. These fluctuations in corn and ethanol prices create challenges for ethanol producers. The break even prices of ethanol production at a range of prices for both corn and ethanol are shown in Tables 1 and 2.

Figure 3. Price of corn at Chicago Board of Trade.

61

Table 1. Potential profit of ethanol sales based on feedstock costs (corn) and direct processing costs of $0.70/gal ($0.20 credit for coproduct sales) and 2.8 gal ethanol/bu.

Table 2. Potential profit of ethanol sales based on feedstock costs (corn) and direct processing cost and related expenses of $l.00/gal ($0.20 credit for coproduct sales) and 2.8 gal ethanol/bu.

The sales prices shown are within the ranges of each commodity (corn and ethanol) during the preceding 12 mo and show the variation possible in a short commodity supply situation. With nearly half of the US gasoline now containing ethanol, the other half of the market remains to be captured. According to the RFA, 73 new plants are under construction and 8 are being expanded. According to a recent popular magazine report, 200 additional plants have been proposed which will cover a great deal of market expansion (Lavelle and Schulte 2007).

Sale Price Ethanol at Plant per gallon Cost of Corn ($/Bushel) $1.50 $2.00 $2.50 $3.00 $3.50

Breakeven Sale price of Ethanol

$2.00 $0.29 $0.79 $1.29 $1.79 $2.29 $1.21 $2.25 $0.20 $0.70 $1.20 $1.70 $2.20 $1.30 $2.50 $0.11 $0.61 $1.11 $1.61 $2.11 $1.39 $2.75 $0.02 $0.52 $1.02 $1.52 $2.02 $1.48 $3.00 -$0.07 $0.43 $0.93 $1.43 $1.93 $1.57 $3.25 -$0.16 $0.34 $0.84 $1.34 $1.84 $1.66 $3.50 -$0.25 $0.25 $0.75 $1.25 $1.75 $1.75 $3.75 -$0.34 $0.16 $0.66 $1.16 $1.66 $1.84 $4.00 -$0.43 $0.07 $0.57 $1.07 $1.57 $1.93 $4.25 -$0.52 -$0.02 $0.48 $0.98 $1.48 $2.02 $4.50 -$0.61 -$0.11 $0.39 $0.89 $1.39 $2.11 Breakeven Cost of Corn $2.80 $4.20 $5.60 $7.00 $8.40

Sale Price Ethanol at Plant per gallon

Cost of Corn ($/Bushel) $1.50 $2.00 $2.50 $3.00 $3.50

Breakeven Sale price of Ethanol

$2.00 -$0.01 $0.49 $0.99 $1.49 $1.99 $1.51 $2.25 -$0.10 $0.40 $0.90 $1.40 $1.90 $1.60 $2.50 -$0.19 $0.31 $0.81 $1.31 $1.81 $1.69 $2.75 -$0.28 $0.22 $0.72 $1.22 $1.72 $1.78 $3.00 -$0.37 $0.13 $0.63 $1.13 $1.63 $1.87 $3.25 -$0.46 $0.04 $0.54 $1.04 $1.54 $1.96 $3.50 -$0.55 -$0.05 $0.45 $0.95 $1.45 $2.05 $3.75 -$0.64 -$0.14 $0.36 $0.86 $1.36 $2.14 $4.00 -$0.73 -$0.23 $0.27 $0.77 $1.27 $2.23 $4.25 -$0.82 -$0.32 $0.18 $0.68 $1.18 $2.32 $4.50 -$0.91 -$0.41 $0.09 $0.59 $1.09 $2.41 Break even Cost of Corn $1.96 $3.36 $4.76 $6.16 $7.56

62

MARKET ACCEPTANCE OF ETHANOL Ethanol use runs the gamut from a 10% gasoline additive that works without engine modifications to E85 (15% gasoline) that works in flex fuel vehicles (FFVs). General Motors reported estimated sales of 400,000 FFVs in 2006. Consumers like the idea of FFVs as they can operate on gasoline or E85. This should make the transition from gasoline to ethanol easier. Ethanol has a reduced energy content compared to regular gasoline (27% lower energy), so the price of E85 should remain lower than gasoline. A major problem with the expansion of E85 is the low number of stations selling the product. The number of sites selling E85 is changing with reports of 600 to 1,000 of the nation’s 170,000 gas stations selling E85. The web site of the National Ethanol Vehicle Coalition (www.e85refueling.com) reports 1,064 locations in the US with the majority located in the Midwest. Texas had 21 refueling sites (not all publicly accessible) and Illinois had 149. According to the San Antonio Business Journal, of the 5 million FFVs nationwide, an estimated 450,000 are in Texas. Minnesota has the most (306) refueling sites. Thus the availability of FFVs without refueling sites will do little to promote use of E85. The increase of FFVs is motivated by generous fuel economy credits that auto makers receive for each FFV built, even if it never runs on E85 (Consumer Reports 2006). They report that this enables manufacturers to turn out more SUVs and pickups resulting in consumption of more gasoline than E85 may replace. For example, Texas, with nearly 0.5 million FFVs and only 21 refueling sites, has little chance of reducing dependence on gasoline supplies using vehicles burning E85. The National Taxpayers Union (2006) stated that because of the lower energy content of ethanol, drivers using E85 would require 1.4 gallons of E85 to equal the energy content of 1 gallon of gasoline and FFV owners can expect a 5 to 15 % reduction in fuel economy (which may be partially offset by lower pump prices). Table 3 compares the cost of driving a Chrysler Sebring and a Dodge Durango in the “ethanol belt” and in the 4 most populated states in the country. Drivers in all regions will spend more using E85 than gasoline.

Table 3. Annual cost of driving a flexible fuel vehicle using E85 vs gasoline in various states. State Cost of E85 Cost of Gasoline Difference

Chrysler Sebring "Ethanol Belt" $1,677.51 $1,382.08 $295.43 California $2,458.58 $1,534.23 $924.35 Florida $1,828.40 $1,369.40 $459.00 New York $2,414.20 $1,445.48 $968.72 Texas $2,165.68 $1,388.42 $777.26 Dodge Durango "Ethanol Belt" $2,892.86 $2,479.15 $413.71 California $4,239.80 $2,752.08 $1,487.72 Florida $3,153.06 $2,456.41 $696.65 New York $4,163.27 $2,592.87 $1,570.40 Texas $3,734.69 $2,490.52 $1,244.17 Source: US DOE FFV Cost Calculator. DOE baseline numbers are as of June 22, 2006.

63

Consumer Reports tests show that consumers will get cleaner emissions but poorer fuel economy using E85 vs gasoline in FFVs. The fuel economy of a 2007 Tahoe decreased 27% when running on E85 as compared with gasoline, from 14 mpg to 10 mpg. With the retail pump price in August (2006) averaging $2.91/gal (according to the Oil Price Information Service 2006), a 27% fuel penalty means drivers using E85 would have paid $3.99 for the energy equivalent of a gallon of gasoline. According to the San Antonio Business Journal, the Texas grocer H-E-B has pledged to price its E85 at 30 cents per gallon less than regular unleaded gasoline regardless of market fluctuation. H-E-B is a blender of record and receives the $0.51 federal tax incentive and is passing this credit along to consumers. With more ethanol production coming on line, blenders see the price of E85 improving and the price per gallon becoming more competitive. If, however, E85 was readily available tomorrow, it could be pumped only into the 2.5% of the nations cars that are flex fuel equipped (Lavelle and Schulte 2007). IMPORTING ETHANOL FROM BRAZIL Some US companies planning to import Brazilian ethanol into the US use dehydration plants in Central America and take advantage of the Caribbean Basin Initiative (CBI) which allows up to 7% of the previous years ethanol production to be exempt from import duties. The April, 2005, issue of Ethanol Today stated that up to 10 ethanol dehydration plants were underway in CBI countries. The Central American Free Trade Agreement (CAFTA) locks in tariff free access to the US market for foreign ethanol. Congress has instituted a $0.54/gal tariff on ethanol imports to promote development of domestic renewable fuels. Typically, the cost of producing ethanol in South America would be at least one third less than the cost of production in the Midwest. If ethanol prices remain strong, Brazilian ethanol could become a major factor in the ethanol market. Ships carrying 136,000 barrels ethanol each, heading to the US might deliver ethanol to the east coast at shipment costs no higher than by rail or truck from the Midwest. Import tariffs of $0.54 have been a deterrent to Brazilian ethanol but, if prices remain high, imported ethanol (even with tariffs) can be a viable alternative. FEEDSTOCKS FOR ETHANOL Feedstock refers to any basic starch or cellulose containing material (eg, corn, grain sorghum, barley, wheat, agricultural residue) used to produce ethanol. Grain is used to produce more than 90% of the US ethanol although emerging cellulosic technologies could present other feedstock possibilities. Alternatives include stover (cornstalk, husk, cob), switchgrass and municipal waste. Biotechnology Industry Organization (2006) summarized cellulosic biorefinery projects in development which includes Iogen Corporation (www.iogen.ca), Abengoa Bioenergy (www.abengoabioenergy.com), DuPont (www.dupont.com), Mascoma (www.mascoma.com ) and Broin (www.Broin.com). They suggest that crop residues are the most likely near term feedstocks for commercial scale production of cellulosic ethanol. Graham et al (2007) reported 30% of the corn stover (58 million dry metric tonnes) produced in the US could be collected with existing equipment and used without undue increases in soil erosion. If no till corn production technology was applied to all US corn production, the harvestable corn stover would expand enough to produce more than twice the ethanol currently produced with grain. Their study used USDA county level statistics on corn production between 1995 and 2000 on tillage practices

64

from the National Crop Residue Management Survey. The cost of collecting, harvesting and baling stover was $30/ton, including nutrient replacement costs. Most collectable stover resided in Iowa, Minnesota and Illinois. Collectable stover was less than expected in Nebraska because of the high risk of wind erosion. Thus, as the study by Fink and Fink (2005) revealed a large amount of biomass in corn stover residues is available. The study by Graham and coworkers supports the utilization of this stover if good soil conservation management (no till) practices are used. ROLE OF THE GOVERNMENT State and federal governments have been major players in promoting and supporting the development of ethanol producing facilities. Government has supported the $0.51 federal tax incentive and applied a tariff of $0.54 on foreign ethanol. The governments can both give and take away. Assume that in future years the price of oil was to decline and ethanol tax credits and ethanol import taxes were eliminated! Such a move would have a major impact on ethanol plant profit but such a scenario is possible. Examples of state subsidies to ethanol producers are shown in Table 3.

Table 3. State subsidies to ethanol producers.

State Incentive Indiana $0.125/gal to facilities that increase production by at least 40 million gal/yr

Kansas $0.07/gal Maryland $0.20/gal for small grain sources, $0.05/gal for all other sources Minnesota $0.20/gal for the first 15 million gal/yr; capped at $3 million per plant per

year Mississippi $0.20/gal Missouri $0.20/gal for the first 12.5 million gallons, then $0.05/gal for the next 12.5

million gallons, for a plant's first five years Montana $2 million/plant per year for plants using Montana grown grains North Dakota $0.40/gal through 2007 Oklahoma $0.20/gal until 2011, then $0.075/gal for the next three years Pennsylvania $0.05/gal, up to 12.5 million gallons South Dakota $0.20/gal up to 416,667 gal/mo Texas $0.20/gal for ethanol and biodiesel for the first 18 million gal/yr for 10

years Wisconsin $0.20/gal, $3 million a year for the first 15 million gallons Wyoming $0.40/gal, up to $4 million a year Renewable Fuels Association, Grand Forks Herald, March 11, 2006.

Support for production of ethanol has been received at all levels of government (including local) and is unlikely to stop. Grassroots support for ethanol comes from the increased jobs and improved air quality and safety when compared to other oxygenates. The

65

federal government’s trend of investing in ethanol appears to be continuing. Thus, based on government support and a strong corn growers’ lobby, ethanol production should continue to receive a great deal of government help. FRACTIONATION In corn, fractionation involves the separation of the three primary grain components which are the endosperm, germ and pericarp. Corn fractionation can produce three higher value corn fractions: corn germ, pericarp and endosperm. The endosperm fraction is called degermed, debranned corn and is utilized in the existing dry grind ethanol plant instead of whole corn. The germ and bran are then processed into corn germ, corn oil, deoiled corn germ and processed bran products. Corn fractionation can produce the following benefits:

• Increase in volume and value of coproducts. Coproducts could include corn oil, deoiled corn germ, higher protein DDGS, corn pericarp products and neutraceuticals. Potential coproducts could benefit companies developing hybrids, especially those high in oil.

• Potential improvement in plant profitability with additional profit from coproducts. • Potential reduction in plant energy requirements. Since germ and pericarp reduce the

amount of material going through the system, DDGS volume and the thermal energy required for drying will be reduced. DDGS, although lower in volume, will be lower in fiber and higher in protein (more value per unit).

• Increased output of ethanol, because unfermentables do not enter the fermentation process.

Wet degermination, dry degermination and downstream centrifugation are used to

fractionate corn. Some commercial companies offer a patented process of corn dry fractionation and many entities are working to identify new coproducts in the dry grind process. Work at the University of Illinois, which includes quick germ, quick germ and quick fiber and enzymatic milling processes has shown promising results (Singh et al 2005). The QG, QGQF and E-Mill processes increased ethanol concentration by 8 to 27% relative to the conventional dry grind process. These process modifications reduced fiber content of DDGS from 11 to 2% and increased the protein content of DDGS from 28 to 58%. These high protein and low fiber DDGS could be used as a nonruminant poultry and swine foodstuff and expand DDGS markets. The DDGS would have protein content greater than the protein content of soybean meal. ENERGY CONSIDERATIONS Currently, 85% of ethanol plants in the US use natural gas. Exceptions occur when limitations of supply exist or where other sources, such as coal, are available. Coal is a less costly energy source than natural gas but has some associated problems such as a more expensive boiler and environmental challenges and associated costs. Several plants are planning to use coal or coal slag as a fuel for creating steam either for their own boiler or by using waste water from another power plant. When considering coal as an energy source, the cost of building a coal fired power plant is high and must be considered along with environmental challenges. Considerations for developing a coal fired system are: capital cost, coal supply, environmental issues and manpower required to operate the facility.

66

SUMMARY Fuel ethanol serves an important role as a gasoline oxygenate and has served rural areas as a “value added” product for farmers. All grain farmers have benefited as a result of an increase in corn prices due to ethanol use. In 2006, more than 18% of US corn went to ethanol production. The ethanol fuel market is a “high risk market” and investors should recognize this when they make a commitment. If the price of corn is high, profit from ethanol production may be low and conversely, if corn price is low, ethanol profit can be good. Two potentially worrisome factors for the ethanol market are the large numbers of plants being built or expanded and the risk of low cost ethanol being imported from Brazil or through Caribbean Basin nations. The cost of food for human consumption and the price of corn for animal use is being impacted, as is the cost of exports to developing nations. Since much of ethanol’s growth is related to government programs, it remains possible that even though the government has given incentives to ethanol producers, it can also take them away. As with all major businesses, good financial, management and marketing are vital if financial success in the ethanol business is to be attained.

LITERATURE CITED

Babcock, B.A. and Hart, C.E. 2006. Do ethanol/livestock synergies presage increased cattle

numbers? Iowa Ag Review. Spring. www.card.iastate.edu/iowa_ag_review/spring_06/article2.aspx

BBI International. 2004. Ethanol Plant Development Handbook, 4th Ed. In cooperation with the Renewable Fuels Association.

Biotechnology Industry Organization. 2006. Achieving Sustainable Production of Agricultural Biomass for Biofinery Feedstock. www.bio.org/ind.

Close, D. 2004. Blairstown ethanol plant to be sold at auction. October 13. Cedar Valley Times. www.cedarvalleydailytimes.com/articles/2004/10/05/news/news01.txt

Consumer Reports. 2006. The Ethanol Myth. October. www.consumerreports.org/cro/search.htm?query=The+Ethanol+Myth&header_health_search.x=13&header_health_search.y=3

Fink, R.J. 2001. Sunshine Energy Cooperative Case Study. Illinois Institute for Rural Affairs. www.iira.org/pubsnew/index.asp

Fink, R.J. and Fink, R.L. 2005. An assessment of biomass feedstock availability in Missouri. Missouri Department of Natural Resources and the University of Missouri Office for Special Programs. www.dnr.mo.gov/energy/renewables/biomass-inventory2005-07.pdf.

Graham, R.L., Nelson, R., Sheehan, J., Perlack, R.D. and Wright, L.L. 2007. Current and potential US corn stover supplies. Agron. J. 99:1-11. www.agron.scijournals.org/content/vol99/issue1.

Lavaelle, M. and Schulte, B. 2007. Is ethanol the answer? US News and World Report. February 12.

National Renewable Energy Laboratory. 2007. Learning about renewable energy and energy efficiency. www.nrel.gov/learning/re_biofuels.html.

National Taxpayers Union. 2006. Policy Paper No. 121. July 20. Pimentel, D. 2003. Ethanol fuel: energy balance, economics, and environment impacts are

negative. Int. Assoc. Math. Geo. Nat. Resources Res. 12(2).

67

Renewable Fuels Association. 2007. Building New Horizons – Ethanol Industry Outlook for 2007. www.ethanolrfa.org/objects/pdf/outlook/RFA_Outlook_2007.pdf

Shapouri, H. and McAloon, A. 2004. The 2001 net energy balance of corn ethanol. USDA/ARS, Eastern Regional Research Center. Wyndmoor, PA. www.ethanolrfa.org/objects/pdf/net_energy_balance_2004.pdf

Singh, V., Johnston, D.B., Naidu, K., Rausch, K.D., Belyea, R.L. and Tumleson, M.E. 2005. Comparison of modified dry-grind corn processes for fermentation characteristics and DDGS composition. Cereal Chem. 82:187-190.

68

CHARACTERIZATION OF GLUCOAMYLASES FOR CONVENTIONAL SIMULTANEOUS SACCHARIFICATION AND FERMENTATION

Chee-Leong Soong*, Guillermo Coward-Kelly and Kevin Wenger

Novozymes North America, Inc., 77 Perry Chapel Church Road, Franklinton, NC 27525

(919-494-3195) soon@novozymes.com In the conventional dry grind process, corn mash is cooked at high temperature with

addition of a thermostable α-amylase which randomly cleaves internal α-1-4 bonds of starch to

shorter water soluble dextrins. Liquefied mash is cooled to 32°C, adjusted to pH 4.5,

glucoamylase and yeast are added, and subjected to a simultaneous saccharification and

fermentation process (SSF). Glucoamylase is an exoacting enzyme that catalyzes hydrolysis of

α-1-4 linkages; this will further break down dextrins and soluble starch into glucose. It also

hydrolyzes α-1-6 bonds, but at a much slower rate than α-1-4 bonds. Glucose produced will be

fermented continuously to ethanol by yeast.

We are characterizing various glucoamylases for conventional SSF. The performance

of glucoamylases from our enzyme collection was evaluated in SSF and benchmarked with

existing commercial products. Potential candidates were studied further for pH sensitivity,

ethanol stability and other physicochemical properties. Enzyme kinetics for dextrin and starch

degradation also were determined.

69

FUEL ETHANOL LIFE-CYCLE ENERGY USE AND GREENHOUSE GAS EMISSIONS

May M. Wu* and Michael D. Wang

Argonne National Laboratory, 9700 S. Cass Ave., Argonne, IL 60439

(630-252-6658) mwu@anl.gov

INTRODUCTION

President Bush announced in his 2006 State of the Union Address a goal to increase the use of biofuels and other alternative fuels in the transportation sector: replacement of 20% of the gasoline demand in the US by 2017. The Energy Information Administration (US Department of Energy) forecasts the demand for gasoline in the transportation sector will continue to grow from the current level of 140 to 175 billion gallons by 2017. A 20% reduction thus translates to 35 billion gasoline equivalent gallons of fuel. Growing at a record speed, the ethanol industry produced 4.9 billion gallons of ethanol in 2006, which is slightly more than 3% of the total gasoline supply. While the amount of ethanol produced from corn will continue to grow, the biofuels industry and the research and development (R&D) community face many new opportunities and challenges, such as producing ethanol from new feedstocks, implementing technologies in conventional corn ethanol plants to reduce fossil fuel use during production and building large scale biorefineries. Ethanol can be produced from many sources other than corn: crop residues, farmed cellulosic biomass, such as switchgrass and forest residue, and many others. In addition, sugar cane ethanol from Brazil could play a role. The energy and greenhouse gas (GHG) emissions benefits of fuel ethanol have been examined extensively by Argonne National Laboratory and other organizations. While researchers generally agree that corn ethanol achieves moderate benefits in terms of reducing both fossil energy and GHG emissions, ethanol produced from other feedstocks could offer greater benefits. In addition, recent technology deployment trends in corn ethanol plants could further increase or, in some cases, reduce corn ethanol’s energy and GHG emission benefits. We will summarize our recent studies on the life cycle (or well to wheels [WTW]) energy use and GHG emissions associated with biofuels. We used the Greenhouse Gases Regulated Emissions and Energy Use in Transportation (GREET) model developed at Argonne to examine ethanol production from various feedstocks including corn, cellulosic materials and sugar cane. We included analyses of different operation stages to determine their contributions to life cycle energy use and emissions. We further compared the energy impacts of corn ethanol plants fueled with conventional and/or renewable sources. Sugar cane ethanol imported from Brazil for use in the US transportation sector was examined to determine its energy and GHG emission effects.

70

AGRICULTURAL AND FOREST WOOD RESIDUE DERIVED CELLULOSIC ETHANOL At Argonne we examined the life cycle of ethanol produced from cellulosic feedstocks, including forest wood and agricultural residue such as corn stover (Wu et al 2006). The analysis included feedstock farming, feedstock transportation, ethanol production, ethanol transportation and ethanol use in light duty vehicles (LDVs). Ethanol is used as E85 (a mixture of 85% ethanol and 15% gasoline by volume) in flexible fuel vehicles (FFVs). For the farming stage, we considered fertilizer manufacturing and application, farming operations, lime production and application to corn, farm machinery manufacturing and feedstock harvesting. Corn farming requires additional fertilizers as a result of corn stover removal; application of these fertilizers contributes to additional nitrous oxide (N2O) emissions. On the other hand, a fraction of the nitrogen in corn stover is converted to N2O through microbial activity and emitted when corn stover is left in the fields. When corn stover is collected, this portion of N2O emissions is avoided. We took into account the net N2O emissions that result from corn stover harvesting. The energy embedded in farming machinery, including energy used in steel and rubber making, fabrication, assembly and in manufacturing parts to repair the machinery, also is included in this analysis. Feedstocks are transported via truck, rail and barge to ethanol plants, where they are converted to ethanol by means of selected biochemical and thermochemical processes. For this study, we estimated consumption of petroleum oil and fossil energy, emissions of GHGs (N2O, carbon dioxide [CO2] and methane [CH4]) and emissions of criteria pollutants (carbon monoxide [CO], volatile organic compounds [VOCs], nitrogen oxide [NOx], sulfur oxide [SOx] and particulate matter with diameters smaller than 10 micrometers [PM10]) during the fuel cycle. We compared life cycle energy use and GHG emissions of ethanol produced from corn grain, corn stover and forest wood residue with those produced from petroleum reformulated gasoline (RFG). We considered technology progress for conventional petroleum gasoline and ethanol in the near and long terms. FFVs fueled with corn stover ethanol blends offer substantial energy savings (94 to 95%) relative to those fueled with RFG (Wu et al 2006). For each Btu of corn stover ethanol produced and used, 0.09 Btu of fossil fuel is required. The amount required for forest wood residue is 0.16 Btu. The cellulosic ethanol pathway avoids 85 to 89% of GHG emissions which is slightly higher than those avoided by the sugar cane ethanol pathway (Macedo et al 2004). Fossil energy consumption for a cellulosic ethanol plant is close to that of a sugar cane ethanol mill, but lower than a corn based ethanol plant. Cellulosic ethanol plants can supply heat and power from lignin residue to meet their own needs. Excess electricity generated from these ethanol plants could be exported to the power grid to alter the US electricity generation mix. In that regard, cellulosic ethanol plants not only save a portion of fossil based (natural gas and/or coal) fuels during processing but also become a net energy exporter. Similarly, during sugar cane ethanol production, bagasse (a dry, fibrous residue remaining after the extraction of juice from the crushed stalks of sugar cane) is combusted to generate steam and electricity to meet the demand for heat and power in sugar cane mills (Macedo et al 2004). The ability of these plants

71

to export excess electricity results in a marked difference in life cycle fossil energy between cellulosic/sugar cane based ethanol and corn ethanol. Unlike the life cycle of corn grain based ethanol, in which the ethanol plant consumes most of the fossil fuel, in the life cycle of corn stover based ethanol, farming consumes most of the fossil fuel. Farm machinery manufacturing energy accounts for less than 2% of total WTW fossil energy use. The fossil energy distribution in the cellulosic ethanol life cycle appears closely related to the choice of ethanol production process. ETHANOL PRODUCTION IN DIFFERENT PLANT TYPES In 2006, 4.9 billion gallons of corn ethanol were produced and used in the US; 80% of this amount was from dry grind plants and 20% from wet milling plants. All new corn ethanol plants that have come online in the past several years and those that will be constructed in the next few years are dry grind plants. Dry grind facilities have been powered primarily by natural gas. Because natural gas prices have skyrocketed in recent years, new ethanol plants are designed to reduce process fuel demand or employ process fuels other than natural gas. We evaluated several dry grind plant types, together with the aggregate ethanol production from all ethanol plants (Wang et al 2007). These ethanol plant types include the following: Ethanol Plants with Natural Gas. These new gas fueled ethanol plants have lower natural gas consumption compared with some older gas fueled ethanol plants.

Ethanol Plants with Natural Gas and Wet Distillers Grains with Solubles (DGS). Some new ethanol plants are sited near animal feedlots so that wet DGS can be moved directly to the feedlots, allowing operators to eliminate drying of DGS and resulting in large energy savings for ethanol plants.

Ethanol Plants with Natural Gas and Combined Heat and Power (CHP) Systems. The US Environmental Protection Agency (EPA) has been working with several ethanol plants to install CHP systems in those plants.

Ethanol Plants with Coal. One hurdle with coal fueled ethanol plants is these plants may become major emission sources under current EPA classifications; because of this, they will be required to go through a longer process to obtain emission permits.

Ethanol Plants with Coal and Wet DGS. This type of coal fired ethanol plant design moves wet DGS to nearby animal feedlots to avoid drying of DGS.

Ethanol Plants with Coal and CHP Systems. Adding CHP systems to coal fueled ethanol plants helps reduce the overall energy use in these plants.

Ethanol Plants with Wood Chips as Process Fuel. Two corn ethanol plants in Minnesota are adding wood chip gasifiers to produce syngas from wood chips; steam from the syngas will be used for ethanol plant operation.

72

Ethanol Plants with Natural Gas and Syrup. Corn syrup (or condensed distillers solubles) from the ethanol distillation process can be burned to provide a portion of the process heat. The remaining heat demand can be met by natural gas. This technology has been installed at the Corn Plus Ethanol Plant in Winnebago, MN.

Ethanol Plants with DGS. In dry grind plants, for each gallon of ethanol produced, 6 lb of distillers dried grains with solubles (DDGS) is produced. DDGS contains 53,760 Btu. Thus, the energy contained in DDGS exceeds the energy needed by an ethanol plant. Of the 11 corn ethanol options evaluated, each achieves a positive fossil energy balance. A close examination of the energy use of these ethanol options shows that all the options reduce petroleum use relative to gasoline at the expense of increasing natural gas use (when natural gas is the process fuel in corn ethanol plants) or coal (when coal is the process fuel). A switch from natural gas to coal in corn ethanol plants could eliminate the GHG reduction benefits of corn ethanol. A switch from fossil fuels to biomass based process fuels (such as wood chips and DGS) helps to increase corn ethanol’s energy and GHG benefits. Installing CHP systems in ethanol plants offers smaller energy and GHG emission benefits because the amount of electricity used in corn ethanol plants is small. Considering resource supply, as well as energy and GHG benefits, cellulosic ethanol represents a long term, sustainable ethanol production pathway. ACKNOWLEDGMENTS This work was sponsored by the US Department of Energy, Office of Energy Efficiency and Renewable Energy. The manuscript was prepared by Argonne National Laboratory to present the results of analyses conducted, in part, for the 30×30 study in 2006.

LITERATURE CITED

Macedo, I.D.C., Leal, M.R.L.V. and Silva, J.E.A.R.D. 2004. Assessment of Greenhouse Gas

Emissions in the Production and Use of Fuel Ethanol in Brazil, prepared for the State of Sao Paulo, Brazil, March.

Wang, M., Wu, M.M. and Huo H. 2007. Life-cycle energy and greenhouse gas emission

impacts of different corn ethanol plant types. Preprint, Environmental Research Letters. Wu, M.M., Wang, M. and Huo, H. 2006. Fuel-Cycle Assessment of Selected Bioethanol

Production Pathways in the United States. Argonne National Laboratory. Report ANL/ESD/06-7. Argonne, IL.

The submitted manuscript has been created by UChicago Argonne, LLC, Operator of Argonne National Laboratory (“Argonne”). Argonne, a US Department of Energy Office of Science laboratory, is operated under Contract No. DE-AC02-06CH11357. The US Government retains for itself, and others acting on its behalf, a paid up nonexclusive, irrevocable worldwide license in said article to reproduce, prepare derivative works, distribute copies to the public and perform publicly and display publicly, by or on behalf of the Government.

73

CONTINUING EVALUATION OF LOW CONDUCTIVITY ELECTRODIALYSIS AS AN ALTERNATIVE OR COMPLEMENT TO ION EXCHANGE RESINS IN

STARCH PROCESSING

Daniel H. Bar*, Florence Lutin and Mathieu Bailly

AMERIDIA, Division of EURODIA INDUSTRIE 20F Worlds Fair Drive, Somerset, NJ 08873

(732-805-4003) dbar@ameridia.com At the Fourth International Starch Technology Conference, a novel electrodialysis (ED) technology was introduced aimed at low conductivity applications in starch processing (Bar and Lutin 2005). Commercialized under the trade name AQUALYZER® LCD (LCD), it allows the demineralization of low conductivity products without polarization at the surface of the membranes, a common problem when the product conductivity is low (<1 mS/cm). The purpose of this technology is to minimize the polishing resin volume in the demineralization process line of syrups (starch hydrolyzates, glucose syrups, specialty sugars and polyol syrups) and the related consumption of chemicals and pollution load. We will focus on industrial tests of ED during the last two years and on the determination of its optimum combination with activated carbon for decolorization and ion exchange resins in cost effective and environmentally friendly processes for the demineralization of dextrose syrups (Figure 1). Since it is difficult and expensive to discharge large volumes of salt effluents from ion exchange (IEX) resins, ED has been the technology of choice for many applications as an attractive demineralization alternative to ion exchange resins because it does not consume large quantities of chemicals and generates lower volumes of salt effluents compared to the regeneration of resin beds. Until recently, ED has found limited uses in the starch industry because of the relatively low conductivity of raw syrups (20 to 800 μS/cm). Due to this low conductivity range, it was not possible that ED could be economically attractive for demineralization of dextrose syrups.

Figure 1. Semi-industrial LCD unit.

74

LCD technology was developed with the main objective of providing an LCD stack with the same design as commercial ED stacks. The new stacks are expected to operate at higher current densities (compared to conventional ED) for complex solutions with low conductivities to reduce the required installed membrane area and investment cost. The new technology is different from Electro-Deionization (EDI) that was developed for ultrapure water applications; it incorporates a proprietary stack component that makes it well adapted to high viscosity and high solids streams in the agrofood and fine chemical industries for conductivity reductions between 100 and 1,000 μS/cm. Ion exchange resin beds are required for polishing treatment but their size is minimized, leading to lower consumption of chemicals and lower effluent volumes. For existing plants, the new technology would allow capacity increase without generating additional effluents. At this time, several semi industrial units have been designed and installed in starch factories using full size ED stacks (Figure 1). Several months of stable operation have generated sufficient data to confirm the performance of this new technology. From these industrial results, operating and investment costs have been calculated for larger plants. As an example, for 96 DE (dextrose equivalent) syrup at 30 to 35 Bx and 50 to 55°C, the production curve has been generated depending on outlet conductivity of the product (Figure 2).

Production (L/h.m2 ) function of outlet conductivity of product

050

100150200250300350400450500550

0 50 100 150 200 250 300 350 400 450 500 550 600 650

Outlet conductivity (µS/cm)

Prod

uctio

n (l/

h.m

²)

Figure 2. Production of permeate of syrup (96 DE, 30 to 35 Bx and 50 to 55ºC) as input conductivity vanes.

The optimum demineralization range for LCD would be 50 to 90%, leading to an equivalent reduction of the chemicals used for IEX regeneration and of the corresponding salt effluents. The volume of water consumed by a process combining LCD and IEX resins is lower than with IEX resins alone. Additional tests were performed with 96 DE Syrup (33.3 Bx) at a conductivity of 320 μS/cm and 105 ICUMSA (Figure 3). The goal was to compare the demineralization of this syrup by four different processes: 1) separate anion and cation exchange resins beds, 2) LCD plus anion and cation exchange resins beds, 3) LCD plus mixed resins beds, 4) LCD plus decolourization with activated carbon beds and 5) mixed resin beds. The target syrup conductivity is 3 μS/cm and 5 ICUMSA.

75

Figure 3. Demineralization of syrup (96 DE, 33.3 Bx) test apparatus. EXAMPLE OF A DEMINERALIZATION PROCESS INCLUDING LCD The investment in a LCD system must be decided based on chemical costs to regenerate resins and effluent disposal costs that are specific to each site. For a new site, when it is beneficial to consider the LCD technology, the lowest operating and investment costs to meet the required conductivity and color targets are obtained with a process combining LCD and two separate resin beds. However, for an existing facility, activated carbon beds and mixed beds are already in place and addition of LCD after the decolorization step leads to the most attractive process to meet required conductivity and color targets.

LITERATURE CITED

Bar, D. and Lutin, F. 2005. Low conductivity electrodialysis, a cost-effective and environmentally friendly alternative to ion exchange resins. Page 15 in: Intl Starch Technol. Conf. (Rausch, K.D., Singh, V. and Tumbleson, M.E., eds.). Urbana, IL.

76

ECONOMICS OF BIOMASS GASIFICATION AND COMBUSTION AT FUEL ETHANOL PLANTS

Douglas G. Tiffany*, R. Vance Morey and Matthew DeKam

University of Minnesota, St. Paul, MN 55108

(612-625-6715) dtiffany@apec.umn.edu

ABSTRACT Dry grind corn processing plants have the potential to reduce operating costs and improve net energy balances by using biomass as the source of process heat and electricity. By replacing natural gas and coal based electricity, usage of renewable biomass reduces the carbon imprint of the ethanol produced, which may have enhanced value in the developing marketplace. We identified various technology bundles of equipment, fuels and operating activities capable of satisfying emissions requirements for dry grind plants of 50 and 100 million gal/yr capacity using distillers dried grains with solubles (DDGS), syrup or corn stover. The technology bundles were chosen based on available combustion/gasification technology, integration with plant processes (eg, drying) and emission control. An electronic workbook was constructed to model relative competitiveness of competing technology bundles by testing parameters of ethanol price, DDGS cost, corn stover cost, natural gas cost and electricity price for estimated capital and operating costs. This framework allows ethanol plant managers and boards to consider use of biomass to produce process heat and electricity. INTRODUCTION Production of fuel ethanol by the dry grind process is expanding rapidly in the US; annual production capacity is expected to exceed 12 billion gal/yr by the end of 2008 (RFA 2007). Typically, natural gas has been the fuel used to produce process heat at these plants. Coal is sometimes used for fuel, especially in plants with capacities greater than 100 million gal/yr. Dry grind plants typically yield 2.75 gal anhydrous ethanol and 17 lb distillers dried grains with solubles (DDGS). Drying DDGS requires one third of the natural gas used by the plant. Consideration of DDGS as a biomass fuel reveals there is sufficient energy to supply process heat and electricity for the facility with some additional energy available for electrical power generation for sale to the grid. As a team with expertise in analysis of engineering and economic topics, we sought to identify the leading methods of thermal conversion of coproducts or field residues that would be technically feasible and financially prudent under a range of economic conditions Technical data related to characteristics of distillers dried grains (DDG), DDGS, syrup and corn stover were collected and analyzed to model the potential of extracting energy derived from these biomass fuels. This data will be presented in other reports in more detail and only be presented here as necessary to establish conditions to analyze the economic aspects of biomass fuel utilization.

77

KEY ECONOMIC DRIVERS FOR THIS RESEARCH Natural gas prices are the second largest operating cost of dry grind ethanol plants following the cost of corn. At this time of rapid expansion of dry grind ethanol production in the US Corn Belt, demands for natural gas are expanding rapidly, which exacerbates supply issues, especially on natural gas lines of limited capacity in certain rural areas. The price history of natural gas prices in Iowa, the heart of the US Corn Belt, are shown in Figure 1.

$-

$2.00

$4.00

$6.00

$8.00

$10.00

$12.00

$14.00

$16.00

Feb-

01

May-01

Aug-01

Nov-01

Feb-

02

May-02

Aug-02

Nov-02

Feb-0

3

May-03

Aug-03

Nov-03

Feb-

04

May-04

Aug-04

Nov-04

Feb-

05

May-05

Aug-05

Nov-05

Feb-

06

May-06

Aug-06

Nov-06

Dol

lars

per

Dek

athe

rm

Commercial Price

Industrial Price

Figure 1. Commercial and industrial natural gas prices in Iowa from 2001 to 2006. Electricity Prices Electricity costs are not as important to plant economics in magnitude, but plants have a self interest to produce enough power on site to maintain uninterrupted operation of computers, process control and other vital systems. In addition, there may be improving incentives available to ethanol plants and other factories to produce power for the grid from biomass as individual states establish goals of increasing the percentage of renewable power used within their borders. DDGS Quantities and Likely Prices Revenues from DDGS sales have represented 20% of the total revenue stream of dry grind plants; however, in the past year the percent of total revenues from this coproduct has decreased to about half that amount. Given the rapid expansion of ethanol capacity underway in the US, it will be improbable for US livestock populations to consume the burgeoning production of this coproduct. The reasons stem from the maximum potential inclusion rates for this midlevel protein food when fed to certain classes of livestock. DDGS contain nutritional energy, but it is in the form of fat that not all animals can tolerate with favorable performance. Dairy cows experience milk fat depression when fed diets too high in fats found in DDGS. Swine and poultry have lower abilities to utilize DDGS in their diets due to adverse effects of dietary fat on carcass quality and poor balances of amino acids, respectively.

78

As a foodstuff, DDGS have been hampered by issues of quality variability due to differences in corn quality (year to year) as well as ethanol plant operation issues involving the amount of condensed distillers solubles (syrup) dried with the other portions of stillage. Control of DDGS dryers can cause one problem in food quality when syrup balls are formed in DDGS. Solubles composition in DDGS and the manner in which they are dried or handled can affect issues such as caking when DDGS are shipped. DDGS prices, which historically have been correlated with corn prices, are shown in Figure 2. Challenges of feeding the US production of DDGS produced by 2009 when 2006 US livestock populations are fed maximum dietary inclusion rates are depicted in Table 1. Maximum dietary inclusion rates will be required to be fed to 75% of the livestock populations to consume the 28 million metric tonnes of DDGS expected in 2009.

$-

$20

$40

$60

$80

$100

$120

$140

Oct-01

Jan-

02

Apr-02

Jul-0

2

Oct-02

Jan-

03

Apr-03

Jul-0

3

Oct-03

Jan-

04

Apr-04

Jul-0

4

Oct-04

Jan-

05

Apr-05

Jul-0

5

Oct-05

Jan-

06

Apr-06

Jul-0

6

Oct-06

Jan-

07

Dol

lars

per

Ton

Figure 2. Historical prices of distillers dried grains with solubles at Lawrenceburg, IN. (Source: USDA, ERS Feed Grains Database). Table 1. Consumption of available DDGS (28 million metric tons) by percent of market penetration based on annual ethanol production of 10 billion gal/yr.

Species Grain Consuming

Animal Units (millions) Maximum Rate

of Inclusion, (%) (Metric tonnes, thousands)

Market Penetration %

50 75 100

Dairy 10.2 20 1,887 2,831 3,774 Beef 24.8 40 9,176 13,764 18,352 Pork 23.8 20 4,348 6,521 8,695 Poultry 31.1 10 2,877 4,315 5,754 Total 89.9 18,288 27,431 36,575 (Cooper 2006)

79

FRACTIONATION There is interest in deriving greater value from the coproduct stream of ethanol plants. Up front technologies that fractionate the corn kernel result in enhanced process economics and have potential to produce a more profitable portfolio of products. Back end technologies, such as those that separate corn oil from thin stillage, currently are utilized to produce corn oil for biodiesel production while at the same time lowering fat in DDGS for the benefit of certain livestock classes. NET ENERGY BALANCE Use of coproducts (DDGS, DDG or syrup) from ethanol production or corn stover as fuels to operate the plant can improve the net energy balance of the entire process of making fuel ethanol from corn. This occurs because fossil energy sources are replaced by renewable sources. Morey et al (2006a) estimated net renewable energy values for corn ethanol with biomass to operate the plant comparable to estimates for cellulosic ethanol. EMERGING TRENDS FOR LOW CARBON FUELS The efforts of California and other states to reduce the carbon imprint of their fuel supply should establish higher prices for ethanol produced by methods that result in lower emissions of carbon. As California’s AB-32 legislation is implemented, firms selling fuels in that state should be willing to pay more for ethanol produced with a low carbon imprint whether due to the feedstock used or the source of the imbedded energy in the fertilizer used. Well to wheels studies that compare cellulosic ethanol with ethanol produced from conventional dry grind corn plants (Wang et al 2005) suggest cellulosic ethanol will result in 3 fold reductions in greenhouse gas emissions. Thus, a California fuel blender would need to purchase and transport one third as much ethanol to blend if it has a greater propensity to lower GHG emissions. Ethanol produced at plants using biomass fuels also will have a smaller carbon imprint than ethanol produced at plants using natural gas and should be able to command a price premium for their production in this market. OUR WORK PLAN With these economic drivers in mind, our team set out to evaluate use of biomass to provide process heat and electricity at dry grind plants. First we characterized ethanol plant coproducts and corn stover as fuels. Second, we modeled the ethanol plant process incorporating the required additional equipment to use biomass as a fuel source. Finally, we performed an economic analysis reflecting the additional capital and operating costs associated with biomass as a fuel. CHARACTERIZATION OF BIOMASS FEED STREAMS Process streams sampled at five cooperating dry grind ethanol plants included: 1) DDGS, 4 plants; 2) DDG, 1 plant; 3) distillers wet grains (DWG), 5 plants and 4) syrup, 5 plants. One sample of corn stover was obtained. Details of this work are found in Morey et al (2006b). Key results are summarized below.

80

Moisture content averaged 10, 65 and 67% (wet basis) for DDGS, DWG and syrup, respectively. DWG and syrup moisture contents were too high to sustain gasification and combustion without augmentation with some other fuel or drying of the material before use as a fuel. Other results are compared on a dry matter basis. Ash content contents were 3.9, 2.6, 7.0 and 6.7% for DDGS, DWG, syrup and corn stover, respectively. Higher heating values were 9350, 9440, 8480 and 7710 Btu/lb dry matter for DDGS, DWG, syrup and corn stover, respectively. Nitrogen and sulfur occur in significant quantities in the three ethanol coproducts. Nitrogen contents were 4.8, 5.4 and 2.6% for DDGS, DWG and syrup, respectively. Sulfur contents were 0.77, 0.66 and 0.96% for DDGS, DWG and syrup, respectively. The alkali metal content (potassium and sodium) of the ash was high (22 to 34%) for coproducts and corn stover. Such high levels of alkali metals can lead to ash fouling in combustion and steam generation units and to potential agglomeration of bed material in fluidized bed systems. Ash fusion temperatures are low, particularly for DDGS and DWG, providing further challenges for design of combustion/gasification systems. We used a computational fluid dynamics approach to model combustion and gasification. Major components of the models were developed by RMT, one of our project partners, and are proprietary; however, results of the modeling using fuel characteristics of ethanol coproducts and corn stover are presented (Morey et al 2006b). The primary use of the model was to predict emissions of key pollutants such as carbon monoxide (CO), nitrogen oxides (NOx), sulfur dioxide (SO2) and unburned carbon. The primary emission of concern will be NOx. CO can be controlled by adjustments to combustion or gasification conditions, or at the stack by catalytic oxidation. SO2 can be controlled by adding a chemical (usually limestone) to the combustion or gasification unit; this technology is used in industry. NOx can be a greater control challenge, since it arises not only from combustion air but also from the fuel. Ethanol coproducts (DDGS, DWG and syrup) have high levels of nitrogen (2.6 to 5.4% db) compared to fossil fuels or even corn stover; these have a high potential for NOx generation. Thermal NOx (NOx from the air) increases as the temperature of the combustion unit increases; for this reason, gasification, which operates at a lower temperature than combustion, is attractive for minimizing NOx emissions. NOx from fuel nitrogen also is decreased by operating in a reducing atmosphere, which further suggests an advantage for gasification. Technologies are available for NOx reduction at the stack, if acceptable NOx levels cannot be achieved from the combustion or gasification processes. An example is selective catalytic reduction which involves injecting urea in stack gases to react with the NOx to produce nitrogen gas (N2), carbon dioxide and water. Fairly aggressive NOx control may need to be applied to the coproducts for the combustion mode.

81

PROCESS MODEL USING BIOMASS AS A FUEL The modeling of these systems was based upon an Aspen Plus model ver. 2004.1 (Aspen Technology 2005) of the complete dry grind process developed by ARS, USDA. We simulated components for biomass thermal conversion technologies, emissions control, process heat utilization and combined heat and electric power production. The model allows us to compare systems by keeping track of mass and energy flows throughout the system as we evaluate various alternatives. Using this information we can specify equipment components and develop estimates for capital and operating costs associated with these biomass alternatives. The various technology alternatives were grouped into technology “bundles” based primarily upon the method of thermal conversion (combustion or gasification) and included variations for drying and thermal oxidation methods. Different scenarios were simulated for each technology bundle under the following criteria: providing process heat demand only, providing process heat and generating electricity with a backpressure turbine or using all coproduct fuel available and generating a surplus of electricity to sell to the grid. Various scenarios were modeled based on the fuel or mixture of fuels used. Technology alternatives were grouped into three broad areas: 1) conversion systems: combustion vs gasification, 2) energy use: process heat only, process heat plus electricity for the plant or process heat plus electricity for the plant and grid and 3) drying technology: steam tube, direct heat or possibly superheated steam. The primary conversion system is a fluidized bed. So far we are evaluating DDGS gasification, combustion of corn stover and combustion of mixtures of syrup and DDG or syrup and corn stover. We plan to evaluate alternatives involving an updraft gasifier and a stoker/grate system since examples of these either are in use or proposed for use at ethanol plants. ECONOMIC ANALYSIS At the economic modeling stage, we describe competing technology bundles with each assigned a specific worksheet. Proforma budgets are constructed for each bundle and a common Menu page is established to orchestrate various economic conditions to determine the economic viability of various options. The proforma budgets were developed (Tiffany and Eidman 2003) to analyze ethanol plant economic sensitivity. Factors with greatest impact are natural gas and DDGS prices. The tension between these two has much to say about the particular technology bundle likely to produce the highest rate of return. Other key factors are capital equipment cost as well as expected operating costs needed to implement a particular technology bundle. A preliminary example of the workbook’s menu page and how the rates of return for competing technology bundles are compared in the workbook are shown in Figure 3. Equipment capital costs were $72, 84, 88 and 90 million, for Base Case, Case 1, Case 2 and Case 3, respectively. Cases differed in DDGS amounts and value of the available to be sold and the price received for electricity generated. An example page of an individual proforma spreadsheet that describes the operations of a typical dry grind plant operated with natural gas and purchased electricity is depicted in Figure 4. Individual worksheets are used to describe each of the technology bundles.

82

Ac

tive

Valu

esBa

se V

alue

sDe

bt-E

quity

Ass

umpt

ions

Fact

or o

f Equ

ity40

%40

%Fa

ctor

of D

ebt

60%

60%

Inte

rest

Rat

e Ch

arge

d on

Deb

t9%

9%In

vest

or R

equi

red

Retu

rn o

n Eq

uit y

12%

12%

Depr

ecia

tion

base

d on

ass

et lif

e (y

ears

)15

15Ba

se C

ase

Case

#1

Case

#2

Case

#3

8.04

%11

.87%

13.6

4%16

.82%

Out

put M

arke

t Pric

esEt

hano

l Pric

e (d

enat

ured

pric

e) $

/gal

.$1

.75

$1.7

5DD

GS

Price

$/T

$70.

00$7

0.00

Elec

tricit

y Pric

e (P

lant

is S

elle

r)

($ p

er k

Wh)

$0.0

9$0

.06

CO2

Price

($

per T

on liq

. CO

2)$8

.00

$8.0

0

Gov

ernm

ent S

ubsi

dies

Fede

ral S

mal

l Pro

duce

r Cre

dit (

$/ga

l.)$0

.10

$0.1

0RF

S Et

hano

l Tra

dabl

e Cr

edit

($/g

al.)

$0.1

0$0

.10

Feed

stoc

k De

liver

ed P

rices

Pai

d by

Pro

cess

orCo

rn P

rice

($ p

er b

u.)

$3.5

0$3

.50

Stov

er P

rice

($ p

er d

t)$6

0.00

$60.

00

Ener

gy/W

ater

Pric

esNa

tura

l Gas

Pric

e ($

per

1,0

00,0

00 B

tu)

$8.0

0$8

.00

Elec

tricit

y Pric

e (P

lant

is B

uyer

) ($

per

kW

h)$0

.06

$0.0

6LP

(Pro

pane

) Pric

e

($

per g

allo

n)$1

.10

$1.1

0W

ater

Pric

e ($

per

gal

lon)

$0.0

0005

$0.0

0005

Ope

ratin

g Co

sts/

Inpu

t Pric

esAs

h Di

spos

al ($

/ dr

y ton

)$2

5.00

$25.

00W

aste

wate

r Tre

atm

ent C

ost,

$/10

00 g

al$2

.00

$2.0

0O

xyge

n ($

/den

atur

ed g

allo

n of

eth

anol

)$0

.10

$0.1

0De

natu

rant

Pric

e / g

al$1

.80

$1.8

0De

nat/1

00 g

al A

nhyd

.5

5Ba

se C

ase

Case

#1

Case

#2

Case

#3

Feed

stoc

k-to

-Eth

anol

Con

vers

ion

Yiel

ds8.

04%

11.8

7%13

.64%

16.8

2%Co

rn (a

nhyd

rous

gal

per

bus

hel)

2.75

2.75

Rate

s of

Ret

urn

for A

ltern

ativ

e He

at a

nd P

ower

Te

chno

logi

es U

sing

DDG

S

0.00

%2.

00%

4.00

%6.

00%

8.00

%10

.00%

12.0

0%14

.00%

16.0

0%18

.00%

Base

Cas

eCa

se #

1Ca

se #

2Ca

se #

3

Figu

re 3

. Ass

umpt

ions

com

mon

acr

oss a

ll pr

oces

ses.

83

; y

Cost/Denat. Gal. Ethanol

Ranges for Column C Plant Totals

Nameplate Ethanol Prod. (Denat. Gal.) 40,000,000Investment per Nameplate Gallon $1.80 $1.00- $2.00 Plant Cost 72,000,000$ Factor of Nameplate Capacity 1.25 (80%- 150%)Debt-Equity AssumptionsFactor of Equity 0.40Factor of Debt 0.60 Initial Debt 43,200,000$ Interest Rate Charged on Debt 0.09Rate of Return Reqd. by Investors on Equity 0.12

Rate of Return 8.04%Conversion Efficiency Assumptions Annual Production

Anhydrous Ethanol Extracted (Gal. per Bu.) 2.750 2.5-2.9 gal/buBushels Ground Denat. Gallons

DDGS per Bushel (lb. per Bu.) 17 15-19 lb./bu 17,355,372 50,000,000CO2 extracted per Bushel (lb. per Bu.) 17 15-19lb./bu

Establishment of Gross Margin Price per UnitRevenue/Bu.

GroundRevenue/Gal.

Denatured Sold Plant TotalsEthanol Price (denatured price) $/gal. $1.75 $.80 to $2.50 $5.0417 $1.7500 87,500,000$ DDGS Price $/T $70.00 $55-$130 $0.5950 $0.2065 10,326,446$ CO2 Price ($ per Ton liq. CO2) $8.00 $2- $12 / liq.Ton $0.0680 $0.0236 1,180,165$ Electricity SoldFederal Small Producer Credit $0.10 $0.0864 $0.0300 1,500,000$ RFS Ethanol Tradable Credit $0.10 $0.2750 $0.0955 4,772,727$

-$ Revenue per Unit $6.0661 $2.1056 105,279,339$ Corn Price Paid by Processor ($ per bu.) $3.50 $1.60---$3.25 $3.5000 $1.2149 60,743,802$ Gross Margin $2.5661 $0.8907 44,535,537$

Operating Expenses Per Bushel Price per UnitCost /Bushel

GroundCost /Gal.

Denatured Sold Plant TotalsNatural Gas Price ($ 1,000,000 Btu) $8.00 ($3-$15/Dtherm)LP (Propane) Price ($ per gallon) $1.10 $.80- $2.00/gal.Factor of Time Operating on Propane 0.02 0-.12BTU's of Heat fr Fuel Req./ Denat. Gal. 35,000 28,500-55,000Combined Heating Cost $0.8148 $0.2828 14,140,306$ Electricity Price ($ per kWh) $0.06 $.025-$.090/kWhKilowatt Hours Required per Denat.Gal. 0.75 .70 -1.25 kWh/denat. gal.Electrical Cost $0.1296 $0.0450 2,250,000$ Total BTU's of Fuel and Electricity 42,500Total Energy Cost $0.9444 $0.3278 16,390,306$

Cost/Denat. Gal. Ethanol

Enzymes $0.0480 $0.1383 $0.0480 2,400,000$ Yeasts $0.0220 $0.0634 $0.0220 1,100,000$ Other Proc.Chemicals & Antibiotics $0.0200 $0.0576 $0.0200 1,000,000$ Boiler & Cooling Tower Chemicals $0.0050 $0.0144 $0.0050 250,000$ Water $0.0060 $.005-.010 $0.0173 $0.0060 300,000$ Denaturant Price per Gal. $1.8000 Denat/100 gal Anhyd. 5 $0.2357 $0.0818 4,090,909$ Total Chemical Cost $0.5267 $0.1828 9,140,909$

Depreciation based on C49 asset life 15 Years $0.2766 $0.0960 4,800,000$ Maintenance & Repairs $0.0125 $0.0360 $0.0125 625,000$ Interest Expense $0.2240 $0.0778 3,888,000$ Labor $0.0450 $.04--$.06 $0.1296 $0.0450 2,250,000$ Management & Quality Control $0.0136 $.010-$.022 $0.0392 $0.0136 680,000$ Real Estate Taxes $0.0020 $0.0058 $0.0020 100,000$ Licenses, Fees& Insurance $0.0040 .0030-.0050 $0.0115 $0.0040 200,000$ Miscellaneous Expenses $0.0135 $.01-$.03 $0.0389 $0.0135 675,000$ Total of Other Processing Costs $0.7616 $0.2644 13,218,000$ Total Processing Costs $2.2327 $0.7750 38,749,215$ Net Margin Achieved Per Unit $0.3334 $0.1157 5,786,322$ Investor Reqd. Return on Equity 12.00% $0.1991 $0.0691 3,456,000$ Increment of Success/Failure to Meet Required Return $0.1343 $0.0466 2,330,322$

Ethanol Plant Profits for Shareholders and Principal Reduction $5,786,322 $5,786,322 5,786,322$

Figure 4. Base case; natural gas and electricity.

84

KEY ECONOMIC PARAMETERS TO BE TESTED Using the workbook, we examined the economic performance of the competing technology bundles for ethanol plants of 50 and 100 million gallons of capacity with particular attention paid to the effects of the following economic variables: DDGS, corn stover, natural gas, ethanol, corn, electricity purchase and selling prices and price premiums for ethanol with low carbon imprint. Economic Conclusions The advantages of using the coproducts, or some portion of them, as fuel for the ethanol production process, are that these materials are available at the plant, they are produced in sufficient quantity to meet process heat and electricity needs and during periods of low coproduct prices the task of marketing them is reduced or eliminated (Morey et al 2005). Since ethanol plants are usually located in corn growing areas, corn stover is available in close proximity. However, since under current production practices most corn stover is left in the field, there are costs associated with collection, transportation, storage and preprocessing to make the material into a convenient fuel. The cost of the corn stover per unit of energy may be higher than the ethanol coproducts at coproduct prices that might emerge. (Sokhansanj and Turhollow 2004, Morey et al 2006a, Petrolia 2006). As of March, 2007, it is premature to comment on relative rankings of the competing technology bundles because our economic modeling is to start in May, 2007. SUMMARY DDGS coproducts are attractive for biomass fuel because they exist in sufficient quantities to provide process heat and power for a dry grind plant as well as additional electricity. Stover is an attractive biomass fuel because it is plentiful in the vicinity of corn dry grind ethanol plants and may be easier to gasify than DDGS. However, there are logistical issues with the use of this material as a fuel. Computational fluid dynamics modeling performed with proprietary methods and software have been used to help predict emissions challenges of NOx and SOx from biomass fuels. In addition, issues of ash fusion caused by the alkali metals in the biomass have been studied to help identify combustion and gasification strategies that will have operational reliability. After performing modeling of thermal requirements, we have specified capacities for various components of discrete choices that dry grind plants might pursue. We have sought estimates of capital costs and estimated operating costs of implementing biomass options for process heat and electrical production at dry grind ethanol plants from an engineering firm. Capital costs will change but hopefully our analysis will be helpful to ethanol plants considering the utilization of coproduct or residue biomass as a source of process heat or electrical generation. We sought to uncover key issues that would be encountered by ethanol producers considering use of biomass to reduce production costs and produce ethanol with a lower carbon imprint. While we had engineering firms as part of our research team, our analysis does not

85

represent the final details of implementing utilization of biomass energy at dry grind plants. Retaining appropriate engineering and economic expertise is the best path for ethanol plants interested in pursuing an individual solution. We show our work, but acknowledge that we do not have all the answers necessary to implement such a plan. We are aware of some thoughtful work done by various engineers seeking to implement this concept. In many cases, innovative firms have settled on a particular technology bundle and made other accommodations to make it work. We have taken a broader, conceptual approach and sought to identify and address the challenges for a broader audience. We live in exciting times of technical change and policy initiatives such as the low carbon fuel imprint. Our workbook was built with the flexibility to capture effects of further policy initiatives such as premiums for ethanol produced due to California legislation. Policy signposts which may affect feasibility of this analysis include state and federal incentives for distributed power that would hasten the time to secure a power purchase agreement for an ethanol plant and preferential payment rates for cogeneration facilities. ACKNOWLEDGEMENT Supported in part by a grant entitled “Generating Electricity with Biomass Fuels at Ethanol Plants” funded by the Xcel Energy Renewable Development Fund. More information can be found at www.biomassCHPethanol.umn.edu.

LITERATURE CITED

Aspen Technology. 2005. Aspen Plus modeling software. Version 2004.1. www.aspentech.com. Aspen Technology, Inc. Cambridge, MA.

Cooper, G. 2006. National Corn Growers Association. Distillers Grains Quarterly, 1st Quarter. Morey, R.V., Tiffany, D.G. and Hatfield, D.L. 2006a. Biomass for electricity and process heat

at ethanol plants. Appl. Engr. Agri. 22:723-728. Morey, R.V., Hatfield, D.L., Sears, R. and Tiffany, D.G. 2006b. Characterization of feed

streams and emissions from biomass gasification/combustion at fuel ethanol plants. ASABE Paper No. 064180. ASABE, St. Joseph, MI.

Petrolia, D.R. 2006. The economics of harvesting and transporting corn stover for conversion to fuel ethanol: a case study for Minnesota. Staff Paper Series, Applied Economics, University of Minnesota. http://agecon.lib.umn.edu/mn/p06-12.pdf.

RFA. 2007. Ethanol Biorefinery Locations. March 13. Renewable Fuels Association. http://www.ethanolrfa.org/industry/locations/. Sokhansanj, S. and Turhollow, A.F. 2004. Biomass densification – cubing operations and

costs for corn stover. Appl. Engr. Agri. 20:495-499. Tiffany, D.G. and Eidman, V.R. 2003. Factors associated with success of fuel ethanol

producers. Staff Paper P03-7. Applied Economics, University of Minnesota. August. Wang, M. 2005. Energy and greenhouse gas emissions impacts of fuel ethanol, center for

transportation studies. Argonne National Laboratory, August 23.

86

MODIFICATION OF STARCH BY BRANCHING ENZYME FROM RHODOTHERMUS OBAMENSIS

Anders Viksø-Nielsen*, Per Linå Jørgensen and Sven Pedersen

Novozymes A/S, Krogshøjvej 36, 2880 Bagsværd, Denmark

(+45 4442 4747) avni@novozymes.com During the last decade, cloning and characterization of a large number of new glycosyltransferases have been published. One of these is the branching enzyme (EC 2.4.1.18) which has been cloned from a variety of bacteria, plants and mammals (Kawabata et al 2002, Shinohara et al 2001, Takata et al 1994 and 2003). A branching enzyme is essential in glycogen and starch biosynthesis, being responsible for producing α-1,6-branch points by a transglycosylation of α-1,4-glucans in glycogen and starch. Until recently, a branching enzyme has been classified into the α-amylase family GH13, but Murakami et al (2006) discovered a novel branching enzyme from T. kodakaraensis belonging to family GH57. The branching enzyme is well characterized from a number of species. A number of publications and patent applications describes potential applications of the enzyme with respect to use in food and nonfood industries (Fuertes and Petitjean 2003, Fuertes et al 2005). We will describe basic characteristics of a highly thermostable branching enzyme from R. obamensis (Shinohara et al 2001). Analysis of the modified starch product showed content of α-1,6 branch points was increased to 17% by a combined branching enzyme and β-amylase treatment. The resulting starch dextrin product was highly soluble.

LITERATURE CITED Fuertes, P. and Petitjean, C. 2003. Patent application: EP 1 316 614 A2, Procédé continu de

modification de l’amidon et de ses dérivés par enzymes de branchement, Roquette Freres. Fuertes, P., Roturier, J.M. and Petitjean C. 2005. Patent application: EP 1 548 033 A2, Polymées

de glucose hautement branches, Roquette Freres. Kawabata, Y., Toeda, K., Takahashi, T., Shibamoto, N. and Kobayashi, M. 2002. Preparation of

highly branched starch by glycogen branching enzyme from Neurospora crassa N2-44 and its characterization. J. Appl. Glycosci. 49:273-279.

Murakami, T., Kanai, T., Takata, H., Kuriki, T. and Imanaka, T. 2006. A novel branching enzyme of the GH-57 family in the hyperthermophilic archaeon Thermococcus kodakaraensis KOD1. J. Bact. 188:5915-5924.

Shinohara, M.L., Ihara, M., Abo, M., Hashida, M., Takagi, S. and Beck, T. 2001. A novel thermostable branching enzyme from an extremely thermophilic bacterial species, Rhodothermus obamensis. Appl. Microbiol. Biotechnol. 57:653-659.

Takata, H., Ohdan, K., Takaha, T., Kuriki, T. and Okada S. 2003. Properties of branching enzyme from hyperthermophilic bacterium Aquifex aeolicus, and its potential for production of highly-branched cyclic dextrin. J. Appl. Glycosci. 50:15-20.

Takata, H., Takaha, T., Kuriki, T., Okada S., Takagi, M. and Imanaka, T. 1994. Properties and active center of the thermostable branching enzyme from Bacillus stearothermophilus. Appl. Env. Microbiol. 1994:3096-3104.

87

A PRIMER FOR LIGNOCELLULOSE BIOCHEMICAL CONVERSION TO FUEL ETHANOL

Bruce S. Dien*

National Center for Agricultural Utilization Research, ARS, USDA,

1815 N. University St., Peoria, IL 61604 (309-681-6270) dienb@ncaur.usda.gov

INTRODUCTION

Developing a commercial industry for using lignocellulose feedstocks to produce fuel ethanol has taken greater urgency with increasing concerns regarding US dependency upon imported petroleum. The President has referred to the need for cellulosic ethanol, the popular press has touted its advantages (usually over corn ethanol) and this spring the Federal Government announced it will support 6 commercialization efforts up to $385 million; the specific amounts will depend upon completion of milestones. I will emphasize the technical aspects for converting lignocellulose to ethanol.

Is lignocellulose needed as a feedstock for ethanol? The US already produces 5 billion

gal of grain ethanol per year and should be able to increase this to 15 billion gal in the future. The answer still is yes because the US annual consumption of gasoline is 140 billion gal/yr and growing. However, 15 billion gal/yr represents only 10% of our current usage. Further options, including conservation, are needed to meet our future oil needs.

Recently, a joint panel of USDA and Department of Energy (DOE) experts met to

estimate the amount of fibrous biomass available for ethanol production. They concluded that enough biomass could be produced and collected to meet one third of oil needs (on an energy basis). Using more conservative numbers cited in the report, total bioethanol production could substitute for 25% of our oil (Farrell et al 2006). Sources of biomass include agricultural and forestry wastes, municipal solid wastes and perennial energy crops.

Using lignocellulose will benefit the entire industry. It will lower net greenhouse gas

(GHG) emissions, expand production without further increasing cost pressure on food markets, extend crop production to land unsuitable for row crops and continue ethanol related development of rural economies.

STRUCTURE AND CHEMICAL COMPOSITION Plant carbohydrates mostly appear in cell walls. Mature plant cell walls (secondary walls) contain cellulose, hemicellulose and lignin. Cellulose is a polymer of glucose, attached by α-1,4- and ß-1-4 glycosidic linkages as opposed to the α-1,4- and α-1,6-linkages found in starch. In mature plant cell walls, cellulose strands interweave to form microfibrils. Filling fiber intercellular spaces are hemicellulose and lignin. Hemicellulose consists of a xylose backbone (termed xylan). In fact, xylose is the second most common sugar found in nature, after glucose. Unlike cellulose, the xylan backbone has combinations of neutral sugars and organic acids

88

attached as short side chains. Lignin is a complex, amorphous network of hydrophobic aromatic compounds interconnected by highly resilient ether bonds. Typically cell walls consist of 30 to 40% cellulose, 20 to 30% hemicellulose and 20 to 30% lignin, with the remainder including organic and inorganic compounds.

Plant cell walls differ in composition and structure from the starch granules processed

for ethanol. Three major differences are that the plant cell walls have a rigid structure, after all they are holding up the plant, that is recalcitrant to hydrolytic enzymes; it is water insoluble. The carbohydrates are a complex mixture of sugars that includes: glucose, xylose, galactose, mannose, arabinose and uronic acids. In particular, the presence of xylose and arabinose are problematic from a processing perspective because they are pentoses or 5 carbon sugars, which are not fermented by distillers yeasts. Given that xylose can account for up to 40% of the available sugars, ignoring it is not a realistic option. As a consequence of these differences, plant cell walls need to be treated initially under much harsher conditions than corn starch, usually processed as a wet solid and using specially engineered microorganisms in place of commercial yeast strains.

FuelEthanol

Milling

Steam ExplosionReactor

Solids/Liquid

HydrolysateConditioner

Fermentor Distillation

BiomassWater Recycle

Wash

Fig. 1. SSF Configuration (see text)

FuelEthanol

Milling

Steam ExplosionReactor

Solids/Liquid

HydrolysateConditioner

Fermentor Distillation

BiomassWater Recycle

Wash

Fig. 1. SSF Configuration (see text)Figure 1. Biomass process configuration for simultaneous saccharification and fermentation (SSF).

MODEL PROCESS FOR CONVERTING BIOMASS TO ETHANOL

As benefits a nascent industry, there are many ideas being championed as the best process for converting fibrous biomass to ethanol. The most well characterized process is dilute acid pretreatment coupled with simultaneous saccharification and fermentation (SSF). This process (Figure 1) is instructive as an introduction to basic unit operations needed for processing fibrous biomass. After arriving at the facility, biomass is cleaned of foreign objects and milled for size reduction. It is next pretreated, which consists of mixing solids with dilute sulfuric acid solution to decrease pH to 1.0 and heating acidified biomass briefly in a steam explosion reactor. A steam explosion reactor is used by the pulp and paper industry to remove lignin. As its name implies, the reactor consists of a small chamber where high pressure steam is mixed with biomass followed by explosive decompression. Steam explosion reactors have the advantages of

89

allowing for rapid steam heating and evaporative cooling of high solids streams in a continuous manner. Reaction conditions that may be used for corn stover are pH 1.0 and 190°C for 60 sec.

Pretreated material is pressed and washed prior to fermentation. Washed solids contain

cellulose and lignin. Syrup consists of monosaccharides released from starch (a minor component of fibrous biomass) and hemicellulose, as well as extractables and acid digestible lignin. Syrup is neutralized with lime and further processed to remove or neutralize organic byproducts formed during the harsh pretreatment that may interfere with subsequent fermentation. Processed syrup is remixed with the solids and available recycled process water before entering the bioreactor. Hydrolytic enzymes (eg, cellulases), biocatalysts and required fermentation nutrients are added to the bioreactor; mash is allowed to ferment for 72 to 144 hr. It is assumed that genetically modified microorganisms will be used as a typical catalyst, those that can ferment xylose, possibly arabinose and other 5 carbon sugars. As the hydrolytic enzymes release glucose from cellulose, the biocatalyst ferments it immediately to ethanol. The hydrolytic enzymes used to digest cellulose stop working when too much glucose is present; therefore, keeping the glucose concentration low increases their efficiency. It also helps to prevent contamination. As in the corn ethanol industry, this type of fermentation is termed SSF. Along with cellulose, the biocatalyst will ferment the other sugars (ie, from hemicellulose) to ethanol. Following fermentation, the solids are separated from the beer and the beer distilled to ethanol. The solids, which contain the high heating valued lignin, are combusted to generate steam and power to run the process; burning lignin is one reason why ethanol produced from lignocellulose generates more “net energy” than that from corn.

If solids were hydrolyzed partially with enzymes prior to entering the bioreactors, higher

solids could be fermented in the bioreactor, thereby increasing the final ethanol concentration. Therefore, it is not uncommon to include a partial hydrolysis step prior to inoculating. This also has the advantage of allowing the enzymes to begin hydrolyzing the biomass at a higher temperature (eg, 50 to 60°C) than suitable for the fermentation (35°C). Alternately, solids could be saccharified completely prior to fermenting to ethanol. This would have the further advantage of removing nonfermentables from the bioreactor. However, running a separate hydrolysis and fermentation (SHF) would depend upon using sufficient enzymes to compensate for lower hydrolysis rates incurred at higher glucose concentrations. There is considerable research being directed toward developing hydrolytic enzymes that would operate at higher temperatures (>80°C) with reaction rates unaffected by glucose concentrations.

DOE recently published a detailed technoeconomical model simulating a plant capable

of converting 2000 tonne (db) corn stover per day to ethanol (Aden et al 2002). While process parameters and reduced enzyme costs are beyond what is feasible, the study is useful in giving an overview of the dilute acid process and insight into cost sensitivities. They proposed a target yield of 90 gal/ton dried corn stover with overall conversion yields of 85% for cellulose and 77% for xylose to ethanol.

The above scheme is one of many proposed for converting biomass to ethanol. However, all processes share the same processing steps: pretreatment, enzymatic hydrolysis, inhibitor abatement and fermentation. In the remainder of this review, each of these will be discussed in more detail.

90

PRETREATMENT OF BIOMASS

The goal of pretreatment is to open up the structure of the cell wall to enable enzymes to convert cellulose to glucose. The success of pretreatment is measured by treating the washed solids with cellulase in a dilute solution and measuring glucose release. Despite this simple operational definition, there is no universal theory on what makes pretreated biomass highly digestible by enzymes. This deficiency is more indicative of the complexity of the cell wall structure and enzyme interactions than shortcomings on the part of researchers.

A number of traits have been correlated with digestibility for pretreated biomass. These

relate to allowing the cellulase enzymes quick and easy access to individual cellulose strands. They include reduced particle size and increased porosity, removal of the surrounding hemicellulose, displacement or removal of the lignin away from cellulose fibers, swelling of microcellulose fibers and breaking apart individual glucan strands. This is just a partial list of identified factors. Most pretreatments do not achieve all these goals and those that do often use multiple mechanisms.

Pretreatments rely on a variety of chemical mechanisms to achieve the aforementioned

goals. Hemicellulose can be solublized by partial hydrolysis with dilute acid or strong alkali. Water can act directly as a weak acid when heated to high temperatures (>180°C), which increases its dissociation constant. Organic acids disassociating from the xylan continue to increase acidity of the water. On the other hand, employing a mineral acid catalyst has the important advantage of saccharifying xylan to monosaccharides, which readies them for fermentation.

Lignin is much harder to transform because of its resilient ether bonds. The ether bonds

can be broken by oxygen radicals which are introduced by treating biomass with ozone, hydrogen peroxide or wet oxidation. Wet oxidation consists of pretreating the wetted biomass at high temperatures under a pressurized oxygen atmosphere. Alternately, heating the biomass to high temperatures (>140°C) will displace the lignin by melting it and treating with dilute acid at high temperatures will partially solubilize it. Finally, in the special case of warm season grasses, ester bonds formed between ferulic acid and arabinose join the lignin and hemicellulose together; these bonds can be saponified directly by treating under alkaline conditions.

Cellulose microfibrils are held together by a tight network of hydrogen bonds that form

within individual glucan strands and among adjacent strands. The strands are crystalline and exclude enzymes from binding. Pretreating cellulose breaks apart these bonds and swells the fibers. Hydrogen bonds can be broken thermally or by using solvents that interfere with hydrogen bonding. Examples of the latter are strong acid and specific room temperature ionic solutions that act as solvents for cellulose.

Most pretreatments contain multiple mechanisms. Consider the dilute acid pretreatment described earlier. The mineral acid catalyzes the hydrolysis of hemicellulose, partially breaks apart and melts the lignin, swells the cellulose and increases porosity while steam explosion reduces particle size. Another popular pretreatment is alkaline peroxide. Here, the alkaline catalyst solubilizes the xylan, saponfies ester bonds and disrupts the cellulose hydrogen bonding.

91

Peroxide disrupts the lignin network by breaking apart ether bonds. Pretreatment conditions, and often even the type of pretreatment, need to be tailored to the source of biomass. For example, alkaline peroxide is effective against grasses but not woody materials.

In 2000, major US laboratories investigating pretreatment processes organized the

Biomass Refining Consortium for Applied Fundamental Innovations (CAFI) to coordinate research efforts. This group first selected corn stover as a target substrate. The technologies evaluated were flow through dilute acid, AFEX or ammonia fiber expansion (formally explosion), liquid hot water and high pressure liquid ammonia percolation. They produced a joint publication that can serve as a useful guide in evaluating the state of the art in pretreatment technology (Wyman et al 2005). Other pretreatments worth mentioning include alkaline peroxide, organosolv, ozone and room temperature ionic solution. Further pretreatments for herbaceous biomass also are described in a recent review (Dien et al 2005). INHIBITOR MITIGATION

In the process of breaking down the cell wall structure, compounds are released that are detrimental to subsequent fermentation. Major sources of these chemicals can include salts from neutralization of mineral acid or base catalyst, some types of extractables, xylan associated organic acids, lignin related aromatics and sugar degradation products. Even when biomass is treated with acid or base, salts may not be a concern. Examples of this are sulfuric acid which can be neutralized with lime to form gympsum and ammonia which can be evaporated off. The degree to which organic acids and lignin aromatics are concerns vary with pretreatment condition and feedstock source. Acetic acid, because of its higher concentration, is the most troublesome of the organic acids. Sugar degradation products are the most troublesome. Once glucose, xylose and other sugars are released under harsh pretreatment conditions they can undergo further reactions to form furfural or hydroxyl methyl furfural (HMF). Both of these compounds are aldehydes and inhibit microbial activity even at low concentrations. Furfural is more of a concern with dilute acid than other pretreatments because it converts xylan directly to monosaccharides. The presence of inhibitors needs to be viewed holistically, meaning that inhibitors act in concert on the cells and also interact with other stresses, such as higher temperatures (needed to raise cellulase activity) and increasing ethanol concentration, within the fermentation. Also, the toxicity of the organic acids will vary with culture pH.

There are several strategies for dealing with these byproducts. One of the oldest and

still most popular is over liming, which consists of incubating the hydrolysate at an elevated temperature after adjusting to pH 10 with lime. This method reduces furfural and HMF as well as having numerous other beneficial effects. Other methods include absorption, ion exchange, solvent-solvent extraction, and biochemical and biological processing. Alternatives to removing inhibitors are to adapt the biocatalyst to grow in the hydrolysate, increasing the beginning titer of cells (most microorganisms reduce the aldehyde site on furfural and HMF to the less reactive alcohol form) and to dilute the inhibitory chemicals; the latter generally being a bad idea because it also dilutes final ethanol concentrations. One newer method being pursued is to isolate genes related to furfural reduction and/or stress tolerance and to over express these genes.

92

ENZYMES FOR BIOMASS CONVERSION

Enzymes are classified with what they react with on the cell wall (Himmel et al 1993). Thus the relevant enzymes to biomass pretreatment are cellulase, pectinase, hemicellulase and ligninases. However, commercial enzyme preparations usually contain a wide variety of unreported activities. For example, preparations marketed as cellulases contain hemicellulase activities.

Cellulase is by far the most important of these because of its role in converting

cellulose to glucose. Four separate enzyme activities are required to saccharify cellulose. Endocellulases cut strands of glucans randomly along their lengths. Exocellulases bind at the ends of strands and travel down the attached strand, progressively releasing cellodextrins. Exocellulases are directional; separate enzymes progress from the reducing and nonreducing ends. Finally, ß-glucosidase completes the process, saccharifying cyclodextrins and cellobiose to glucose. The usual source of cellulase for biomass conversion is produced by T. reesei. This cellulase works best at pH 4.5 and temperatures of 55 to 60°C. Unfortunately, cellulases are sensitive to end product inhibition. Excess glucose formation inactivates ß-glcuosidase, leading to excess cellobiose, which in turns inhibits endo and exocellulase. Often extra ß-glucosidase is added to alleviate this inhibition.

Hemicellulases are as complex as hemicelluloses because each single chemical bond

requires its own unique enzyme and specific properties for each can be expected to vary with its family classification. The basic enzyme components are endoxylosidase and ß-xylosidase. Endoxylosidase splits apart xylan strands and ß-xylosidase hydrolyzes shorter chains (X2 to X5) to xylose. However, the situation is complicated because side chains substituted on the xylan backbone protect it from these enzymes. Therefore, complete digestion of xylan requires further enzymes for removing side chain groups. Side chains include neutral sugars, carboxyl organic acids, linked by ester bonds, and uronic acids. Commercial cellulases have hemicellulase activities with the notable exception of feruloyl esterase. Xylanases are prepared commercially from T. reesei and Aspergillus cultures. All pretreatments, other than dilute acid, will depend on hemicellulases for complete saccharification of the xylan; no commercial preparations are marketed for this application.

Pectinases are used to digest pectin, a linear chain of α-(1-4)-linked D-galacturonic

acid that forms the pectin backbone. Like xylan, pectins also have multiple side groups. Commercial pectins are produced from A. niger cultures and contain high levels of contaminating activities, especially esterases. Supplementing commercial cellulases with pectinase can increase cellulase effectivenss on pretreated biomass, no doubt the increase being related to side activities.

Ligninases include lignin peroxidase, manganese peroxidase and laccase. These

enzymes have not been applied to pretreatment except indirectly in biological pretreatments. Laccases slightly increase the efficiency of cellulases and aid in detoxifying hydrolysates, presumably by condensing lignin related aromatics.

Perhaps no area of research has more potential to revolutionize biomass conversion than work on enzymes. On the processing side, there is increasing dependence for hydrolytic

93

enzymes that will work on pretreated biomass at high solids content to produce concentrated sugar streams for fermentation. Likewise, enzyme costs still need to be lowered; the target value is $0.10/gal; enzyme producers appear to be focusing on selectively supplementing T. reesei cellulases with addition enzymes to favor higher specific activities and processing at higher temperatures. With the public release of the T. reesei genome, it should be possible to selectively over express individual enzymes by genetic manipulation. Enzyme preparations may need to be customized for individual pretreatments and sources of biomass to ensure the lowest possible protein loading and cost. In particular, this allows for integrating pretreatment conditions and enzymatic saccharification. FERMENTATION OF HYDROLYSATES

Only two known microorganisms are considered suitable for commercial ethanol production: Saccharomyces cerevisiae (distillers yeast) and Zymononas mobilis. Both share exceptional ethanol tolerance (>15% v/v), yields (>90% of maximum theoretical) and high productivities (>2.5 gal/L/hr). S. cerevisiae is preferred because it is more robust to industrial fermentations and easier to limit contamination by opportunistic bacteria. However, neither ferments xylose which represents 30 to 40% of carbohydrates found in herbaceous plants and hard woods.

There are a few yeasts that are able to ferment xylose to ethanol with significant yields.

Unfortunately, despite much research, these yeasts have productivities and yields that are too low to sustain commercial interest (van Maris et al 2006). Therefore, microbiologists have turned to molecular biology to develop new strains that will ferment glucose and xylose (Dien et al 2003, Jeffries et al 2004). Two approaches have been undertaken. The first is to construct Z. mobilis and S. cerevisiae capable of metabolizing xylose (and sometimes arabinose). Xylose metabolism has been introduced by borrowing the pathway from native xylose fermenting yeast or by introducing a functional xylose isomerase. Research on Saccharomyces is intense and is being pursued globally in scores of laboratories. The other approach is to make gram negative bacteria, which normally ferment xylose and other sugars, to selectively produce ethanol. Specifically, this has meant expressing the two terminal enzymes in ethanol production from Z. mobilis and eliminating genes responsible for production of other fermentation products. Microorganisms engineered with the second approach include E. coli and K. oxytoca. Strains representing all four of these species are being pursued for commercialization; research is continuing to construct superior versions of each. Several scientists are developing yeasts and gram negative microorganisms capable of producing their own hydrolytic enzymes for biomass hydrolysis.

In addition to the above mentioned species, others are being developed for ethanol

production. More important ones include gram positive bacteria, thermophiles and Corynebacterium. Each has unique properties but all are still in the precommercialization stage. Work appears to be progressing more rapidly for thermophiles and Corynebacterium.

While strain development is not the bottleneck to commercialization that it was 5 years ago, it is still an area deserving considerable attention. Saccharomyces strains require improved yields and productivities for fermenting xylose. Current work using gene chips has lead to better

94

understanding the remaining obstacles and hopefully will lead to further breakthroughs. It is doubtful the strains will have the desired hardiness when cultured in industrial scale fermentations unless their physiology when growing on xylose is balanced. Work on developing biocatalysts that express their own hydrolytic enzymes is a promising avenue for reducing enzyme costs. SUMMARY

Growth of an industry for producing ethanol from fibrous biomass is right around the corner, as it has seemed for the past 50 yr. Two factors suggest that beyond the rhetoric, biochemical conversion of fibrous biomass will be evaluated at commercial scale. The first is a confluence of events and geopolitical concerns that has convinced large corporations and institutional investors the technology is worthwhile. Few believe that petroleum will become inexpensive, our sources will grow more secure, or given the increasingly stringent warnings from climate scientists that it will continue to be burned with the same abandon as in the previous century. No doubt, this trend has been aided by the investments and large profits from the corn ethanol industry and is made possible by the willingness of the government to make vast investments in energy. The second factor is that since the oil crisis of the 1970s, scientists have been working to enhance technology to make it feasible, lower technological risks and reduce operating costs. The ability of Iogen to produce 1 million gal/yr ethanol from wheat straw at their demonstration plant is powerful evidence for this statement. There is little doubt investments in cellulosic ethanol are risky and investors are motivated at this point by the opportunities, if successful, of licensing the technology to others as opposed to profiting directly from the produced ethanol.

LITERATURE CITED

Aden, A., Ruth, M., Ibsen, K., Jechura, J., Neeves, K., Sheehan, J. and Wallace, B. 2002. Lignocellulosic biomass to ethanol process design and economics utilizing co-current dilute acid prehydrolysis and enzymatic hydrolysis for corn stover. NREL/TP-510-32438 (available on web).

Dien, B.S., Cotta, M.A. and Jeffries, T.W. 2003. Bacteria engineered for fuel ethanol production: current status. Appl. Microbiol. Biotechnol. 63:258-266.

Dien, B.S., Iten, L.B. and Skory, C.D. 2005. Converting herbaceous energy crops to bioethanol: a review with emphasis on pretreatment processes. Pages 1-11 in: Handbook of Industrial Biocatalysis. Hou, C.T., ed. Taylor & Francis, Boca Raton, FL.

Farrell, A.E., Plevin, R.J., Turner, B.T., Jones, A.D., O’Hare, M. and Kammen, D.M. 2006. Ethanol can contribute to energy and environmental goals. Science 311:506-509.

Himmel M.E., Adney W.S., Baker J.O., Nieves R.A. and Thomas S.R. 1993. Cellulases: structure, function and applications. Pages 144-161 in: Handbook on Bioethanol. Wyman C.E., ed. Taylor & Francis, Washington, DC.

Jeffries, T.W. and Jin, Y.S. 2004. Metabolic engineering for improved fermentation of pentoses by yeasts. Appl. Microbiol. Biotechnol. 63:495-509.

95

van Maris, A.J., Abbott, D.A., Bellissimi, E., van den Brink, J., Kuyper, M., Luttik, M.A., Wisselink, H.W., Scheffers W.A., van Dijken J.P. and Pronk J.T. 2006. Alcoholic fermentation of carbon sources in biomass hydrolysates by Saccharomyces cerevisiae: current status. Antonie Van Leeuwenhoek. Nov. 90(4):391-418.

Wyman, C.E., Dale, B.E., Elander, R.T., Holtzapple, M., Ladisch, M.R. and Lee, Y.Y. 2005. Coordinated development of leading biomass pretreatment technologies. Bioresource Technol. 96:1959-2032.

The mention of firm names or trade products does not imply they are endorsed or recommended by the US Department of Agriculture over other firms or similar products not mentioned.

96

ENERGY SAVINGS THROUGH BETTER USE OF SEPARATORS

Tristan Merediz*

Westfalia Separator, Inc., 100 Fairway Court, Northvale, NJ (901-751-0396) merediz.tristan@wsus.com

INTRODUCTION Disc nozzle separators have been used in the corn wet milling industry for almost 70 years. They are used as clarifier/concentrators for millstream thickening, gluten thickening and middlings concentration, and as classifier/concentrator/washer for primary separation. Their basic design and incorporation to the wet milling process streams modernized the industry, but the current design and mode of use needs to be reviewed to obtain optimum performance and energy savings. BASICS OF DISC NOZZLE SEPARATION A disc nozzle separator consists of a rotating bowl that encloses a disc stack and a series of equally spaced, peripheral nozzles for continuous discharge of the concentrated discharge (Figure 1). Continuous discharge separators discharge solids in a constant process. The solids are discharged via externally installed nozzles on the bowl periphery. The aim of the design process is to obtain solids in a form which is as dry as possible and also in a closed system.

Concentrate Discharge

(underflow)

Overflow Input Stream

Rotating Disc Stack

Nozzles

Figure 1. Disc nozzle centrifuge. The large number of conical discs increases the equivalent clarification area of the separator according to a modified form of Stokes Law. Stokes Law applied to a disc centrifuge is:

( )44444 344444 2143421

Ts

io rrzg

v

gdQ

Σ

−∗∗∗∗∗∗∗Δ∗

= 3322

tan32

18ϕωπ

ηρ

97

Vs describes the particle settling velocity in the gravitational field and is a function of fluid and particle characteristics, where

Q throughput capacity d particle diameter Δρ density difference between solid and liquid η dynamic viscosity g acceleration due to gravity

ΣT describes the equivalent clarification area of the disc stack and is solely a function of centrifuge geometry, where

ω angular velocity (rotating speed of the bowl) Z number of discs φ disc angle ro outer radius of disc ri inner radius of disc

Considering ΣT, disc type separators achieve clarification areas equivalent to around 80 football fields. The discs are assembled to form a disc stack. Small round or long rectangular spacers are positioned between the discs to produce separation channels to allow flow of fluid through the stack. The thickness of the spacers is determined so the height of the separation channel meets requirements of the specific product. The large number of parallel, interdisc spaces results in individual separating chambers. These separating chambers reduce the product volume to be separated into a large number of thin layers. This also reduces the sedimentation path to be covered by the particles to be separated, thereby accelerating the settling process. The nozzle diameter is selected to match the flow of solids into the bowl. In an ideal world, where the input stream conditions are constant, nozzle size would be selected to allow the release of all of the solids and an amount of liquid sufficient to achieve the desired concentration. If the nozzle size is too small, incoming solids cannot exit through the nozzle and solids will be lost through the overflow. Conversely, if the nozzle size is too large, too much liquid is released with the solids and the desired concentration is not reached. Unfortunately, process conditions are not constant and other control means must be used. HISTORICAL PERSPECTIVE When disc nozzle separators were introduced to the corn wet milling industry, they were introduced to achieve primary separation of starch and gluten. The other wet mill operations were still rather crude and many separations were achieved using flotation tanks. Consequently, the amount of foreign material coming to the primary separator could be large and the chance of plugging the nozzles was affected directly by the amount of foreign material.

98

The designers of the first separators for corn wet milling operations compensated for the amount of foreign material in the input stream by operating the separators with oversized nozzles to avoid plugging (Figure 2). Since this released more liquid than was desired to achieve the desired concentration, a portion of the nozzle discharge stream was recycled to the bowl to increase the concentration of solids.

Concentrate Discharge

(underflow)

Overflow

Recycle

Input Stream

Figure 2. Disc nozzle centrifuge with recycle to achieve desired concentration. This system of operation has been very successful and continues to be operated throughout the industry. It does have its drawbacks, mainly:

• Equipment requires minimum recycle. • Oversized nozzles are required to provide recycle. • If input solids decrease, recycle must increase to maintain solids discharge

concentration. • Power consumption remains constant or may increase. • Speed is constant.

Present day wet milling conditions are much better than they were 70 years ago when the separators were introduced and separations achieved are clean and precise. Consequently, the use of disc nozzle separators and the options for these separators need to be reevaluated. PRESENT OPERATIONS Process conditions in the wet mill are such that separations at each step of the process remove the desired material and allow the appropriate streams to move forward with little or no unwanted or undesired foreign matter. Consequently, the use of large, oversized nozzles in a disc nozzle separator is not necessary. Nozzles can be sized more closely to meet the input stream conditions, rate and solids loading, with consideration only for process changes that might be encountered, not for plugging. The small amount of recycle that may be required can be achieved with a small, external recycle pump (Figure 3). Additionally, since the recycle is external, the speed of the machine can be adjusted to meet the input stream conditions.

99

Figure 4.

External Recycle

Concentrate Discharge

(underflow)

Overflow

Input Stream

Figure 3. Centrifuge with external recycle. Power required to drive the bowl is a function of the amount of material being processed and the energy required to accelerate and decelerate the various streams.

In general, The power input can be described by (Figure 4).

Hp1 –accelerate input stream to V1 at R1 Hp2 –accelerate Q1 from V1 at R1 to V2 at R2 Hp3 –accelerate wash plus recycle to V2 at R2 Hp5 –accelerate overflow to V4 at R4

There is a certain amount of power recovered, described by

Hp2' – power from jet action Hp4' – power from overflow discharge

R2

R1 R4

R3 The largest power consumer, in spite of the power recovered, is the power of the nozzle discharge. The power consumed by the nozzles can be described as follows:

Factors contributing to the power consumed are given by: where Q is the nozzle discharge, given by

cd – discharge coefficient of the nozzle d – diameter of the nozzle orifice

2V**QHP ρ∝

( ) 5.02 ** PdcQ d∝

100

P is the fluid pressure at the nozzle entrance (the bowl periphery), given by

r - density of the slurry at the nozzle ω - angular velocity (rotating speed of the bowl) DR – radius difference between nozzle discharge and overflow discharge

V is the velocity of the jet discharge This reduces the power consumed by the nozzles to The power consumed by the nozzles is mainly a factor of the nozzle size, the nozzle discharge coefficient and the speed of the separator. Unfortunately, the larger the nozzle, the more inefficient it is, so that power recovery is worse for larger nozzles. From the above relations, operating with large nozzles on internal recycle machines consumes more power than required for smaller nozzles, and are more closely matched to the process conditions. Alternatively, the amount of recycle that is not required if the nozzles are matched to the process conditions would allow for more feed to the separator. Since horsepower is a direct function of nozzle size (nozzle discharge) and speed of the separator, it is advantageous to choose a nozzle size that is closest to the desired maximum process conditions to operate with the least possible recycle. Also, it is advantageous to adjust the speed of the separator to match input rate and nozzle discharge. SEPARATOR OPERATION Considering the above relations and the influence of speed and nozzle size on power consumption, consideration can be made of various operation possibilities. A conventional primary centrifuge that relies on internal recycle must operate with oversized nozzles at a fixed speed to maintain recycle. If the input rate changes, the recycle must be adjusted, but machine speed remains constant, so that at low feed rates the recycle must be high; there may be an actual increase in overall power consumption as a result of the increased recycle. A typical separator of this type would operate at 2,450 or 2,600 rpm. High recycle rates also have detrimental effects of added shear and heat to the product. For some applications and products, the reduction in particle size may be significant and may hinder separation. The temperature increase from input stream to underflow in separators operating with internal recycle varies from 2 to 5°C. A large number of separators operating with internal recycle can add a significant amount of heat to the process and additional energy costs in cooling. A primary separator operating with external recycle would be equipped with

22 ** RP Δ∝ ωρ

2222 ***** v RdcH dP Δ∝ ωρ

101

smaller nozzles which would reduce the nozzle contribution to power consumption by 25% at an operating speed of 2,600 rpm. The best configuration for a primary centrifuge, however, to allow a wide range of operation and to provide for optimum separation and product quality is to operate at higher speeds, with small recycle at high feed rates and then reduce the speed for lower flow rates. For this type of operation, a primary centrifuge might be set up to operate at 3,300 rpm for high input rates and then reduced to 2,600 rpm for low input rates. This speed change retains an equivalent, low amount of recycle, optimum Sigma Factor for the feed rate and reduces the nozzle power consumption by approximately 50%. Additionally, since there is no internal recycle, temperature increase through the separator is less than 1°C. From actual operating data, it is estimated that electrical power savings achieved with a newer separator with appropriately sized nozzles, low recycle and variable speed drive is $50,000/yr when compared to a conventional separator with oversized nozzles, internal recycle and constant speed drive. CONCLUSIONS Separator technology and operation has changed over the years. Plant conditions and economic conditions today require more attention to improvements and energy savings. The separators designed 70 years ago, although adequate for wet mill operation, do not provide the optimum performance and energy use. Technology exists to provide wet millers with significant energy savings and improved process performance.

102

ULTRASOUND PRETREATMENT OF CORN SLURRY TO ENHANCE SUGAR RELEASE

Samir Kumar Khanal*, David A. Grewell and J. (Hans) van Leeuwen

Iowa State University, Ames, IA

(515-294-7089) samirk@iastate.edu

ABSTRACT The effect of ultrasound pretreatment on sugar release from corn slurry was evaluated. Corn slurry samples were subjected to ultrasound pretreatment for 20 or 40 sec at 64 to 107 μmpp amplitudes of vibration. The resulting samples were enzyme treated to convert cornstarch into glucose. Comparing scanning electron micrographs of raw (control) and sonicated corn samples, sonication resulted in severe disruption of corn particles. Particle size declined 20 fold for sonicated samples at high power settings. Glucose release rate for sonicated samples was 3 times higher than controls. Enzymatic activity was enhanced when the corn slurry was sonicated with simultaneous addition of enzymes. Therefore, ultrasonic energy did not degrade or denature the enzymes. Thus, ultrasound has potential to enhance ethanol yield from cornstarch and may reduce production cost in commercial corn dry grind ethanol plants. INTRODUCTION Ethanol is a renewable clean fuel and is produced mainly from cornstarch in the US. Ethanol production in the US is expected to grow to 15 billion gallons by 2015 from the current 4.9 billion gallons (NCGA 2007). The majority of this growth is likely to come from the dry grind process. In a dry grind operation, corn is hammer milled, mashed, cooked and treated with enzymes to convert starch into glucose which is fermented to ethanol and recovered by distillation and molecular sieves.

There has been much debate on the net energy gain from the conversion of cornstarch to ethanol (Patzek 2004, Pimental and Patzek 2005). Farrell et al (2006) claimed the net negative energy gain reported in the literature was due to the omission of coproducts and use of obsolete data in the calculation. There are, however, possibilities of improving the economics of dry grind plants through process improvements, such as shortening of liquefaction and fermentation times, lowering enzyme dosages, improving overall starch hydrolysis and elimination of some unit processes. The use of ultrasonic technology could provide a practical solution to improve ethanol yield and reduce the production cost by addressing all of these. Ultrasound is a sound wave at a frequency above the normal hearing range of humans (> 15 to 20 kHz). When the ultrasound wave propagates in an aqueous medium such as corn slurry, it generates a repeating pattern of compressions and rarefactions in the medium. Microbubbles are formed in the rarefaction region due to excessively large negative pressure. As the wave fronts propagate, microbubbles oscillate under the influence of positive pressure, thereby growing to an unstable size before violently collapsing. The sudden and violent collapse

103

of huge numbers of microbubbles generates powerful hydromechanical shear forces in the bulk slurry surrounding the bubbles, thereby disintegrating corn particles (Kuttruff 1991). Ultrasound has been applied widely in various biological and chemical processes. The use of high power ultrasound enhanced starch-protein separation in a wet milling operation (Zhang et al 2005). Ebringerová et al (1998) employed ultrasound to aid extraction of active xylan and heteroxylan from corn cobs and corn hulls, respectively. Wood et al (1997) studied the effects of ultrasonic treatment on ethanol fermentation from mixed office paper. However, use of high power ultrasound in corn dry grind operations has not been investigated. The use of high power ultrasound has potential to reduce corn particle size and to free lipid bound starch. Thus, the integration of an ultrasonic unit prior to liquefaction and saccharification could enhance overall sugar release from corn for subsequent fermentation to ethanol. Based on these premises, our overall goal was to investigate the effectiveness of ultrasound pretreatment to release fermentable sugar from raw and cooked corn mashes. MATERIALS AND METHODS Corn Samples and Enzymes Preparation Corn slurry samples, both raw and cooked, were obtained weekly from Midwest Grain Processors (MGP), Lakota, IA, in a chilled container. Ground corn was obtained from Lincolnway Energy (LE), Nevada, IA, and was mixed with double distilled water to prepare mash before sonication. The corn slurry sample from MGP contained a partial amount of alpha-amylase and was stored at 4°C prior to use to reduce enzyme activity. Two types of enzymes were studied, namely STARGENTM 001 obtained from Genencor International (Palo Alto, CA) and amyloglucosidase from Aspergillus niger (Sigma-Aldrich, St. Louis, MO). STARGENTM 001 enzyme was used in raw corn slurry samples from MGP and corn mash samples from LE; glucoamylase was used only in cooked corn slurry samples obtained from MGP. Ultrasonic Pretreatment and Incubation Raw and cooked corn slurry samples were sonicated using a bench scale ultrasonic unit (Branson 2000 Series) for 20 or 40 sec. The system had a maximum power output of 2.2 kW and operated at 20 kHz. Tests were carried out at three amplitudes of vibration (64, 86 and 102 μmpp) corresponding to power levels of low, medium and high, respectively. Initial sonication tests were conducted using 10 ml corn slurry sample mixed with 25 ml acetate buffer (pH 4.3) and 0.05 % (v/v) enzyme in a 50 ml polypropylene centrifuge tube. Enzymes were added to corn slurry samples after sonication (SA) or prior to sonication (SD). Control samples were not subjected to sonication. Experimental conditions are summarized in Table 1. After sonication, samples were incubated for 3 hr in a rotary shaker at 100 rpm and 32°C. All tests were conducted in triplicate with different batches of corn slurry samples.

104

Table 1. Experimental conditions.

Glass Sonication Chamber The use of polypropylene centrifuge tubes resulted in an increase of corn slurry temperature during sonication. Thus, additional experiments were conducted using glass sonication chambers to examine the effect of temperature on glucose yield. To calculate the temperature increase (Δθ) during sonication, corn slurry temperature was measured before and after ultrasound treatment. Analytical Procedures After sonication, 10% (v/v) 4M HCl Tris buffer (pH 7.0) was added to terminate enzyme activity. Samples were centrifuged at 10,000 x g for 20 min and sieved through a 200 mesh (US standard screen). Supernatant was analyzed for glucose concentration using a modified dinitrosalicylic acid (DNS) method (Miller 1954). Absorbance of the sample was measured at a wavelength of 570 nm using a spectrophotometer (ThermoSpectronic Genesys 2, model W1APP11, Rochester, NY). Glucose concentrations were calculated from the calibration curve obtained using absorbance data for standard solutions of D-glucose reacted with DNS reagent.

105

The particle size distribution of the corn slurry before and after ultrasonic treatment at different power levels was determined using a Malvern particle size analyzer (Mastersizer 2000, Malvern, Worcestershire, UK). In addition, scanning electron microscopy (SEM) examination of corn slurry samples was carried out. Prior to SEM examination, samples were fixed and imaged using a JEOL 5800LV SEM (Japan Electron Optics Laboratory, Peabody, MA) at 10 kV with a SIS ADDA II for digital image capture (Soft Imaging Systems, Lakewood, CO). Energy Calculations The ultrasonic dose was estimated using the following relationship:

Edensity = Qavg x t [Eq 1]

Where, Pavg = average power (W); Qavg = average power density (W/ml); V= volume of sonicated sample (ml); Edensity = energy density (J/ml); and t = sonication time (sec).

The total energy dissipated into each sample (Ein) was calculated based on average power and sonication time as indicated below:

[Eq 2] Where, t0 and tf are the initial and final times of sonication. The total energy delivered during sonication (Eout) was calculated based on the chemical energy of the additional glucose produced compared to the control. The change of glucose mass yield compared to the control was calculated, and the energy of the glucose was estimated by assuming a conservative energy density of 15,740 kJ/kg for glucose if fully oxidized (Patzek 2004). The overall efficiency (Eff) of sonication was calculated using the following equation:

[Eq 3]

RESULTS AND DISCUSSION Scanning Electron Microscopy (SEM) Examination SEM of raw and cooked corn slurry samples before and after sonication at high power level are shown in Figure 1. Figures 1A and 1C show cells that appear almost fully intact; there are starch granules confined within the cells. With ultrasonic treatment for 40 sec, nearly complete disintegration of cells was observed with large numbers of fragmented cell materials (Figures 1B and 1D). Several micropores were visible within the disintegrated corn particles.

VPQ avg

avg =

tPEPdtE avgin

t

tinf == ∫ ~

0

%100×−

=in

inout

EEEEff

106

There is a 5 fold magnification difference between the Figures 1C and 1D. At the same magnification, the treated sample would appear as indistinguishable particles. Thus, using SEM images we observed changes in structures of corn particles following ultrasound pretreatment.

C

100m 20mm

200m 200m

A B

D

Particle Size Distribution The effect of ultrasonic treatment on corn particles was examined by sonicating both cooked and raw corn slurry samples and the resulting particle size was compared with nonsonicated samples (controls). The peak of the particle size distribution curve shifted from 800 to 80 μm following sonication at high power levels for cooked corn slurry samples (Figure 2C). In addition, particle size reduction was related directly to power level and sonication time. Particle size reduction at the higher power level and longer sonication time was in agreement with glucose yield under similar conditions. Similar results were obtained for raw corn slurry (data not shown).

Figure 1. SEM images of corn slurry: (A) raw corn (control); (B) raw corn sonicated (40 sec); (c) cooked corn (control); (D) cooked corn sonicated (40 sec).

107

0.00

1.00

2.00

3.00

4.00

5.00

6.00

0.0 0.1 1.0 10.0 100.0 1000.0 10000.0

Particle Size (um)

Volu

me

(%)

ControlCEU 20CEU 40CU 20CU 40

SD20

SA20

SA20

SD40

Low power

A

0.00

1.00

2.00

3.00

4.00

5.00

6.00

0.0 0.1 1.0 10.0 100.0 1000.0 10000.0

Particle Size (um)

Volu

me

(%)

ControlCEU 20CEU 40CU 20CU 40

SD20

SD40

SA20

SA20

B

0.00

1.00

2.00

3.00

4.00

5.00

6.00

0.0 0.1 1.0 10.0 100.0 1000.0 10000.0

Particle Size (um)

Volu

me

(%)

ControlCEU 20CEU 40CU 20CU 40SA40

SA20

SD40

SD20

High power

C

Figure 2. Particle size distributions of cooked corn slurry for (A) low power, (B) medium power, (C) high power. SD=enzymes added prior to sonication. SA=enzymes added after sonication. 20, 40= length of sonication interval (sec).

108

Glucose Release Concentrations of glucose released from raw corn slurry samples are presented in Figure 3. The highest glucose release increases of 32 and 27%, relative to the controls, were obtained for SD40 at low and medium power inputs, respectively. Glucose release decreased 22% at the high power setting for SD40. A similar trend was observed for SD20, in which the glucose release decreased by 11%. These tests were conducted without temperature control and the final temperature of the ultrasound treated samples increased with the power level and treatment time. Therefore, these decreases in glucose concentration could be attributed to the excessive ultrasonic treatment, which may have resulted in degradation or denaturing of the enzymes at high power settings. This finding is in close agreement with glucose release data without enzymes (SA20 and SA40), which showed improvement in glucose release following sonication irrespective of power levels. The higher power level did not cause the gelatinization of starch due to sonication. Additional sugar yield from ultrasound treated samples could also be due to the release of starch that was bound to lipids and did not have access to the hydrolyzing enzyme.

-

10

20

30

40

50

60

70

80

Control SA20 SA40 SD20 SD40

Glu

cose

(g/L

)

Low PowerMedium PowerHigh Power

Control SA20 SA40 SD20 SD40

Figure 3. Glucose release of raw corn slurry at varying power input levels. SD=enzymes added prior to sonication. SA=enzymes added after sonication. 20, 40= length of sonication interval (sec).

Concentrations of glucose released for cooked corn slurry samples are depicted in Figure 4. Because alpha-amylase had already been added to the corn slurry during cooking, only glucoamylase was added during the sonication test. The highest increase in glucose concentration (30%) with respect to the control group was obtained for SD40 at low and medium power inputs. The findings were in agreement with those for raw corn samples (Figure 3). The glucose release decreased 60% at high power levels. The SD20 group at high power setting had a 23% increase in glucose concentration, which was not observed in the raw corn slurry experiments.

109

0

20

40

60

80

100

120

140

160

180

Control SA20 SA40 SD20 SD40

Glu

cose

(g/L

)Low PowerMedium PowerHigh Power

Control SA20 SA40 SD20 SD40

Figure 4. Glucose release of cooked corn slurry at varying power input levels. SD=enzymes added prior to sonication. SA=enzymes added after sonication. 20, 40= length of sonication interval (sec).

To characterize the sonication effects on the rate of glucose release, corn slurry samples were treated at medium power level (amplitude of 80 μmpp) for 40 sec. In these experiments, a ground corn sample obtained from LE was used. Data were fitted with standard reaction rate kinetics of Arrhenius form:

Where G(t) is the glucose concentration as a function of time (t), G∞ is the glucose concentration at time infinity, k is the reaction rate coefficient and R is the universal gas constant. The reaction rate (k) for the ultrasonic treated sample was 3 fold higher than the control (Figure 5). This enhancement in glucose release was due to the fact that no enzymes were introduced into the test until the final saccharification step. In previous experiments, diffusion of enzymes into the corn particles was not eliminated completely.

⎟⎟⎟⎟

⎠

⎞

⎜⎜⎜⎜

⎝

⎛ −

−= ∞kR

t

eGtG 1)([Eq 4]

110

0

10

20

30

40

50

60

70

0 1 2 3 4 5 6 7 8 9

Time (h)

Glu

cose

(g/L

)

Ultrasonic treatmentControl

G(t)=31(1-e-t/4.76)

G(t)=145(1-e-t/15.15)

Figure 5. Glucose release of raw corn slurry at different saccharification times. Temperature Controlled Sonication Experiments Relative glucose releases at various power settings for the cooked corn slurry with the glass sonication chamber are shown in Figure 6. The relative glucose release was as high as 27% with temperature increase (Δθ) of less than 5°C during 40 sec sonication at lower power input. Thus, increased glucose release of sonicated samples was not attributed to thermal effect but was related to particle size reduction, better mixing due to streaming effects and release of additional lipid bound starch. The relative temperature increase was in direct proportion to power input and sonication time. The chamber geometry as well as the mechanical impedance of the base material may have contributed to various levels of attenuation. Sugar release did not improve for corn slurry samples with prior enzyme addition at the higher amplitude. There was a possibility of enzyme denaturation at higher power input.

-5

0

5

10

15

20

25

30

35

40

45

50

SA SA SD SD

Group

Rel

ativ

e ch

ange

in g

luco

se (%

)

GI for Low Power (at 80% amplitude)

GI for Medium Power (at % 60 amplitude)

GI for High Power (at 100% amplitude)

Δθ=

±8.2

°C

Δθ=

±7.1

2°C

Δθ=

±5.0

5°C

Δθ=

±5.1

°C

Δθ=

±4.9

5°C

Δθ=

±3.2

°C

Δθ=

±7.1

5°C

Δθ=

±7.8

5C

Δθ=

±6.3

7°C

Δθ=

±5.9

7°C

Δθ=

±5.6

2°C

Δθ=

±3.9

7°C

20 2040 40

Figure 6. Relative change in glucose concentration at high power setting with glass sonication chamber (cooked corn slurry). Δθ=temperature increase during sonication. GI-glucose release relative to control. SA=enzymes added after sonication. 20, 40= length of sonication interval (sec).

111

Energy Balance The overall efficiency of the glucose release from raw corn slurry ranged from 70 to 125%, depending on treatment condition. Efficiency greater than 100% indicated additional chemical energy from the release of extra sugar following sonication. At shorter treatment times, the efficiency was better, however, compared to higher power settings and longer treatment times. CONCLUSIONS Ultrasonic pretreatment of corn slurry resulted in a 20 fold reduction in corn particle size. Enzyme addition during sonication yielded higher glucose release than enzyme addition after ultrasound treatment. Glucose release of sonicated samples increased 30% with respect to the controls. The increase in glucose release from ultrasound treated samples was attributed to reduction in particle size, better mixing due to streaming effects and release of additional lipid bound starch. Ultrasound treatment at high power for longer sonication time resulted in 60% reduction in glucose release due to denaturing of enzymes. The rate of glucose release for sonicated corn samples was 3 fold higher than the nonsonicated samples. Studies are underway to optimize ultrasonic conditions (power input, sonication time, solids level) for continuous treatment and to examine final ethanol yield of sonicated samples. ACKNOWLEDGEMENTS Funded in part by the Grow Iowa Value Fund. Special thanks to Branson Ultrasonics for supplying the high power ultrasonic system, Melissa Montalbo and Gowrishankar Srinivasan for their help in the experiements, and to Dawn Leegard of Midwest Grain Processors (MGP), Lakota, IA,and Larson Dunn, Lincolnway Energy (LE), Nevada, IA, for supplying corn samples.

LITERATURE CITED Ebringerová, Z., Hromádková, J.H., Alföldi and Ibalová, V. 1998. The immunologically active

xylan from ultrasound-treated corn cobs: Extractability, structure and properties. Carbohydr. Polym. 37:231-239.

Farrell, A.E., Plevin, R.J., Turner, B.R., Jones, A.D., O’Hare, M. and Kammen, D.M. 2006. Ethanol can contribute to energy and environmental goals. Science 311:505-508.

Kuttruff, H. 1991. Ultrasonics Fundamentals and Applications. Elsevier, Essex, England. Miller, G.L. 1954. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal.

Chem. 31:426-428. NCGA. 2007. www.ncga.com/news/presentations/PDF/022007ProducingFoodFuel.pdf. Accessed

March 15. National Corn Growers Association, Washington, DC. Patzek, T.W. 2004. Thermodynamics of the corn-ethanol biofuel cycle. Crit. Rev. Plant Sci.

23:517-567. Pimental, D. and Patzek, T.W. 2005. Ethanol production using corn, switchgrass, and wood;

biodiesel production using soybean and sunflower. Nat. Resource Res. 14:65-76. Wood, B.E., Aldrich, H.C. and Ingam, L.O. 1997. Ultrasound stimulates ethanol production during

the simultaneous saccharification and fermentation of mixed waste office paper. Biotechnol. Prog. 13:323-327.

Zhang, Z., Niu, Y., Eckhoff, S.R. and Feng, H. 2005. Sonication enhanced cornstarch separation. Starch 57:240-245.

112

ENERGY AND PROTEIN – A GLOBAL PERSPECTIVE

David A. Cook*

Cargill, Inc., 10383 165th Ave. NW, Elk River, MN 55330 (763-241-3362) dave_a_cook@cargill.com

With a projected world population approaching 9 billion people by mid century, combined with continued growth in caloric intake per capita, it should not be surprising that food production must double to feed the world’s people (FAO 2004). And as affluence grows within developing countries, consumption of animal protein increases. Thus, livestock production must also grow at a dramatic rate to meet the ensuing growth in demand for food. The growth in biofuel production, however, creates a scenario which challenges societal ability to meet growing demands for food. Livestock producers have long enjoyed a near limitless supply of high quality cereal grains for animal production. In periods of price pressure, the industry has grown in its acceptance of alternative ingredients to reduce production costs. In the era of biofuels, however, changes in feeding practices will be forced upon the industry merely as a means to remain profitable. The question to ponder is what will be available to the industry to feed livestock? ETHANOL PRODUCTION Current bioethanol production in the US only meets 2% of the transportation fuel mix while the US Department of Energy is targeting 30% replacement of petroleum products by 2025 (Ragauskas et al 2006). Similar targets have been set in the EU and other parts of the world for total biofuel production. Corn as a feedstock, dominates bioethanol production, with the US leading the way in ethanol production from corn. Brazil, who has led in ethanol production until recently, produces nearly all of their ethanol from sugar cane. Other feedstocks include wheat, barley, rye, milo (sorghum) and rice. The predominant coproducts produced from ethanol production from grain sources are distillers dried grains with solubles (DDGS) and wet distillers grains. In the US alone, more than 8 million metric tons of DDGS were produced in 2005 with projections estimated at near 25 million metric tons by 2015. BIODIESEL PRODUCTION Biodiesel, while doubling in production during the last couple of years, pales in comparison to bioethanol production in the US. This is in stark contrast to the EU where biodiesel is the predominant biofuel. According to the European Biodiesel Board, production in 2005 increased 65% compared to 2004. Germany is, by far, the leader in biodiesel production in Europe and represents more than 50% of the production. Feedstocks for biodiesel are largely vegetable oils, but do include some animal fats and waste grease. The primary vegetable oil source in Europe is rapeseed while in the US soybean oil serves as the primary source. Other countries with significant biodiesel production include China, Brazil and India (Emerging Markets Online 2006). A key point to recognize is that, while soybean and rapeseed oil

113

represent the major oils produced in the US and Europe, yield of biodiesel per acre varies widely and may impact oil choices in the future (Brown 2006). Oil from palm and coconut far exceed production from soybean and rapeseed (Table 1). While the production regions for these crops are limited to tropical climates, it is reasonable to assume that availability of palm oil for biodiesel production will increase. Table 1. Biodiesel yield per acre of selected crops (Brown 2006).

The implication of vegetable oil choice for biodiesel production has broad reaching implications. First of all, rapeseed as the oil of choice in Europe is anticipated to reduce domestic production of soybean meal from imported soybeans. Moreover, availability of rapeseed meal will increase. In the US, it is questionable whether soybean meal production will increase as a result of demand for soybean oil for biodiesel. A short term rise in soybean meal supply may occur as carry over of beans declines. However, a key will be planting commitments for corn vs soybeans. Some experts are suggesting that demand for corn from bioethanol will not increase corn acreage in the long run due to the need for crop rotations to fix nitrogen and break pest cycles (Anonymous 2007). IMPLICATIONS TOWARDS ANIMAL NUTRITION With the growth in demand for starch for bioethanol and fats and oils for biodiesel, a critical question to consider is where will the energy come from to feed livestock. Recently, the ratio of soybean meal price to corn (or wheat) fell to historical lows. In China, local price for soybean meal decreased to 120% of corn (Cargill Internal Information). Similar price relationships were experienced in other countries, including the US. These historic price relationships created unique dynamics in formulating diets for livestock. Soybean meal became valued more as an energy source than a protein source. A closer look at the coproducts from bioethanol and biodiesel sheds additional light on the growing challenges of feeding livestock in the new millennium. DISTILLERS GRAINS Wet and dried distillers grains traditionally have been fed to ruminant animals (eg, dairy and beef cattle). The high fiber content can be fermented readily in the rumen. Nonruminant species (eg, pigs and poultry), on the other hand, traditionally have not used (or used to a limited degree) DDGS. A recent publication from Feedinfo News Service (Anonymous 2007) focused on the impact of the ethanol boom on feeding nonruminants. Principle barriers to use were identified as: 1) variability in nutrient content, consistency between grades and digestibility,

Crop Gal/Acre Oil Palm 508 Coconut 230 Rapeseed 102 Peanut 90 Sunflower 82 Soybean 56

114

2) perceived poor handling characteristics, small particle size, 3) lack of fast, accurate and inexpensive methods of estimating amino acid digestibility and 4) understanding and managing effects of corn oil in DDGS on pork and poultry fat quality. Spiehs et al (2002) surveyed 10 ethanol plants in Minnesota and South Dakota. They indicated variations among some DDGS producers were as great as variations across all plants surveyed. Source of corn will contribute to this variation but consider the graph in Figure 1. These data arise from a single crop year of samples moving through the Cargill Animal Nutrition analytical lab in North America. The noted variation in net energy of gain for cattle suggests additional sources of variation in addition to corn source. These may include the fermentation process itself, but more importantly, the level of solubles (syrup) added to the DDGS.

69 70 71 72 73 74 75 76 77 78 79 80 81 82

Corn Distillers Grain

69 70 71 72 73 74 75 76 77 78 79 80 81 82

Corn Distillers GrainDDGS

Figure 1. Variation in net energy of gain for corn compared to dried distillers grains. Handling characteristics are largely an issue with transportation. New railcars have been designed to address unloading issues associated with DDGS. Ability to address variation in amino acid digestibility, however, looms as a key concern. Stein et al (2006) determined amino acid digestibility of 10 DDGS samples taken from 10 plants of similar construction with all plants marketing the same branded DDGS. Of particular importance to nonruminants is the key amino acid, lysine. Digestibility of lysine varied by 20 percentage points (43.9 to 63.0% lysine digestibility). While drying conditions contribute to the resulting amino acid digestibility, variation in corn source and addition of solubles (syrup) also plays a role. The high level of polyunsaturated fat in DDGS limits their use in diets for finishing swine. Polyunsaturated fat or oil, when fed to pigs, can be deposited directly into fat stores and result in what is known in the industry as “soft” fat. In most countries of the world, the market discounts “soft” or oily fat. As a result, some pork producers will limit DDGS use, especially in late finishing diets. In total, DDGS can serve as a good source of energy and amino acids for most livestock species. Monitoring suppliers for consistency in DDGS composition is the first step in

115

expanding confidence and DDGS use. Developing a rapid procedure for measuring amino acid digestibility will be required to optimally utilize this ingredient. PROTEIN INGREDIENTS FROM BIODIESEL The implications of biodiesel on availability of primary protein ingredients are less clear. Change in supply of soybean meal is unclear. Growth in biodiesel production in Brazil and China should support growth in supply in those regions. Impact on supply in the US is less apparent. It appears that demand from bioethanol for corn in the US will be the key driver. High corn prices combined with a low spread between soybean meal and corn price will call for higher soybean meal usage. How much of this increase in soybean meal usage can be off set by increase use of DDGS is unclear. The situation in Europe appears a bit clearer. Rapeseed oil will dominate biodiesel production, thus increasing supply of rapeseed meal and decreasing domestic production of soybean meal. Rapseed meal (or canola meal in North America) has a lower concentration of protein and amino acids than soybean meal. Traditionally, rapeseed meal has been viewed as a lower quality source of amino acids than soybean meal. Rapeseed meal faces a similar challenge as DDGS in that over processing has been a historical norm, resulting in a lack of confidence in the ingredient. Newkirk at al (2002) evaluated 31 canola meal samples and found a range in lysine digestibility of more than 20% (65.5 to 85.7%). While in a similar range as DDGS, rapeseed meal faces an additional challenge with the presence of two antinutritional factors, glucosinolates and urusic acid. Generally this is not considered an issue with use of new varieties of rapeseed known as double zero (00) rapeseed. The exception, however, is in developing regions of the world, such as the former eastern block countries and parts of Asia. In these regions, the prevalence of the double zero variety is not known. The unknown is growth in palm oil production. A palm plantation requires 7 years to achieve adequate production for the plantation to generate positive cash flow. Palm kernel meal has developed a reputation of quality as processing methodologies vary widely. The ingredient is high in fiber, contains a moderate level of protein (16 to 18%), and contains 7 to 9% fat. Limited attention has been placed on this ingredient globally as it traditionally has been used at low levels and production is confined to tropical climates. Palm oil is an ingredient to pay attention to, given its potential as a biodiesel source. FEEDING ANIMALS IN THE FUTURE Many questions are yet to be answered. Davis et al (2000) demonstrated pigs can extract at least 75% of their resting energy needs from amino acids. This helps us to conclude that protein levels much higher than those fed traditionally can be tolerated without loss of performance. In fact, Kidd et al (2005) demonstrated that a 10% increase in amino acid content of diets for modern broilers resulted in improved animal performance and economic return. Thus, use of higher protein diets does not appear to be a concern for animal performance although geographies with strict waste management controls may not be able to benefit from increasing protein levels in the diet.

116

Added fat to offset losses of starch can be a challenge with certain species. As mentioned previously, high levels of polyunsaturated fat in diets for ruminants can reduce fiber fermentation in the rumen and diets high in polyunsaturated fats for pigs can reduce the firmness of fat and lead to “soft” fat. Can the livestock industry survive without starch? The simple answer is yes. Animals do not have a requirement for starch. They do, however, have a requirement for glucose. So the question must transform into how will glucose be supplied? Glucose is the universal fuel of life. The brain is an obligatory glucose user. Glucose is required to produce lactose for milk production. Starch is the key source of glucose for nonruminants while ruminants create glucose from volatile fatty acids, primarily, proprionic acid. Thus, ruminants can derive a significant portion of their glucose from fiber fermentation, while nonruminants are largely limited to dietary sources of glucose. Especially in the case of nonruminants, a source of glucose in the form of starch is required to achieve optimal performance. Even in this age of biofuels, livestock will compete for starch. Question yet to be answered is at what cost? Food prices are increasing and will continue to increase. At some point, animal performance will have to be sacrificed to remain economically viable. Our challenge of doubling food production by mid century should provide motivation to identify alternative feedstocks for biofuel production.

LITERATURE CITED Anonymous. 2007. The ethanol boom: what impact on synthetic methionine usage? Feed

Info News Service. March 5, 2007. Brown, L.R. 2006. Plan B 2.0: Rescuing A Planet Under Stress and A Civilization In

Trouble. Norton, New York, NY. Davis, J.A., Greer, F.R. and Benevenga, N.J. 2000. Urea production is increased in neonatal

piglets infused with alanine at 25, 50, and 75% of resting energy needs. J. Nutr. 130:1971-1977.

FAO. 2004. Summary of World Food and Agricultural Statistics. Kidd, M.T., Corzo, A., Hoehler, D., Miller, E.R. and Dozier, W.A. 2005. Broiler

responsiveness (Ross x 708) to diets varying in amino acid density. Poultry Sci. 84:1389-1396.

Ragauskas, A.J., Williams, C.K., Davison, B.H., Britovsek, G., Cairney, J., Eckert, C.A., Fredrick Jr., W.J., Hallet, J.P., Leak, D.J., Liotta, C.L., Mielenz, J.R., Murphy, R., Templer, R. and Tschaplinski, T. 2006. The path forward for biofuels and biomaterials. Science 311:484-489.

Spiehs, M.J., Whitney, M.H. and Shurson, G.C. 2002. Nutrient database for distillers dried grains with solubles produced from new ethanol plants in Minnesota and South Dakota. J. Anim. Sci. 80:2639-2645.

Stein, H.H., Gibson, M.L. and Boersma, M.G. 2006. Amino acid and energy digestibility in ten samples of distillers dried grains with solubles fed to growing pigs. J. Anim. Sci. 84:853-860.

117

THE USE OF BIOSOLIDS TO GENERATE STEAM AT DRY GRIND ETHANOL PRODUCTION FACILITIES

Gregory Coil*

M. A. Mortenson Company, 700 Meadow Lane North, Minneapolis, MN 55422

(763-287-5365) greg.coil@mortenson.com ABSTRACT Contained herein is information on a completed biofuel project and on studies for projects at other facilities. We demonstrated technical and economic viability of a fluidized bed combustor in burning dry grind ethanol plant high moisture coproducts (dewatered distillers solubles or syrup) and using waste heat from the combustion to generate steam for process heat loads at the facility. The studies involved use of ethanol plant coproducts and fractionated grain products as fuel for digesters and solid fuel boilers in ethanol production facility models. Historically, the energy ratio for such facilities has been 1:1.34 (DOE 2006). We increased that ratio to 3.2:1. In addition to the thermal performance of the demonstration project, the project acted as a pollution control device through thermal oxidation of volatile organic compounds in grain dryer exhaust gases, truck load out fumes as well as control of sulfur, nitrogen and particulate emissions. In further studies for separate ethanol production facilities, we have shown technical and economic feasibility for use of thin stillage as an anaerobic digester feedstock, as well as technical and economic feasibility for use of fractionated grain products (eg, bran and germ cake) as fuel for solid fuel boilers to generate steam and electricity for the facilities. In the models generated in these studies, the production facility becomes self sufficient for heat and electrical needs in steady state operations. Keywords: fluidized bed, dewatered distillers solubles, syrup, Copeland reactor, energy ratio, thin stillage, anaerobic digesters, solid fuel boilers THE SYRUP BURNER INTRODUCTION In mid 2001, management personnel at Corn Plus LLLP, anticipating a decrease in dried distillers grain with solubles (DDGS) prices, looked into determining if a cogeneration opportunity existed for use of DDGS, generated as a coproduct of the ethanol production process. The basic question being asked was: is DDGS worth more in terms of energy or as a coproduct? A feasibility study was funded for use of the coproduct as an energy source, with relatively unfavorable financial results (eg, paybacks greater than 7 yr) for use as a cogeneration fuel due to the low regional cost of electricity (Waffenschmidt 2001). They did indicate, however, that it might be economically viable to generate steam using some or all of the coproduct as a fuel, without generating electricity.

118

Further, they concluded that use of a biomass fuel for steam generation only was economically feasible, due, in part, to the relatively low electrical energy costs at that time. The use of dewatered distillers solubles (syrup) was a preferred fuel, due to a number of factors. The primary economic driver for such a facility investment was the facility natural gas usage and increasing market costs for this nonrenewable fuel. Natural gas consumption at the ethanol production facility was approximately 16 million therms/yr. With costs of natural gas near $10 per dekatherm, this consumption would represent a cost of $16 million/yr. Preliminary design information projected potential fuel displacement of 50 to 75% of natural gas consumption. With capital investment costs between $15 and 20 million, the payback was better. In addition to the natural gas cost issues, Corn Plus was being requested to correct an air emission problem identified by the Environmental Protection Agency (EPA) and the Minnesota Pollution Control Authority (MPCA). The problem involved potential emissions in excess of limits for both volatile organic compounds (VOCs) and particulate emissions primarily contributed by the DDGS drying process. A Consent Decree was issued from the EPA in February, 2003, to correct this issue. RESULTS Two solutions were proposed to correct the emissions problem. The first was the use of a regenerative thermal oxidizer (RTO). Thermal oxidation is a lower capital cost, more traditional technology for mitigation of VOC emissions. It would have required an investment of $4 million, with no favorable economic benefit for Corn Plus beyond avoidance of EPA and MPCA fines for additional emissions. In addition, a thermal oxidizer probably would increase consumption of natural gas at the facility, due to the gas required to run the oxidizer. The other proposed solution to the emissions issue was to combine the combustion of high moisture content coproducts in a single pass fluidized bed thermal reactor (Copeland reactor), with injection of dryer vapors into the same reactor, effectively creating a biofuel heated RTO. Heat from the incineration process would be recovered in a waste heat boiler. The proposed alternative solution was studied in depth through the use of a pilot incinerator study conducted by Hazen Research in Golden, CO (Mudgett 2002). The fuel was analyzed chemically to determine the energy and environmental characteristics of combustion products. Parameters analyzed included BTU, moisture, sulfur, nitrogen, oxygen, ash, chlorine and carbon. In addition, nonorganic contributions of sulfur and chlorine were determined, for tabulation of contributions to combustion gases. Laboratory testing was based on standard ASTM tests for coal fuels. In addition to laboratory analyses, prototype testing was required to generate empirical data regarding combustion of the fuel stream. Since a fluidized bed is a dynamic system, a prototype test helps to establish such things as appropriate calcium feed rate to prevent ash fusion and appropriate excess air setting for complete combustion without a decrease in thermal efficiency. Such testing also allowed characterization of system parameters such as excess fuel rate required for sustaining combustion, ash removal rates, ash characteristics after mixing with fluidizing materials and fluidizing bed material replacement rates. Data were used in generation of basic design criteria and prediction of environmental parameters in a full scale fluidized bed

119

thermal reactor. In addition to prototype and laboratory testing, plant testing was performed to obtain plant operational data in the condition modeled, and to provide some assurance of consistent fuel energy content. The proposed alternative solution involved combustion of the syrup coproduct and dryer exhaust gases in a fluidized bed reactor to generate heat necessary to produce the steam demand for the facility. Syrup is essentially the fines and soluble components of corn mash left over from the distillation process. Appropriate controls were used to ensure environmental parameters were maintained within EPA specified parameters. This solution is unique and beneficial for the following reasons:

• Eliminates the VOC and particulate air emissions problem. • Displaces use of a fossil fuel (natural gas) through the use of a portion of syrup as a biomass

fuel. • Reduces the amount of distillers grains to be dried, further decreasing the natural gas

consumption rate at the facility. • Generates more economic benefit to the agricultural business sector as energy than the sale

of syrup applied to DDGS and dried to be sold as food. • Removes some minor restrictions inherent in ethanol plant operations, resulting in an

increase in ethanol production from the facility with no other modifications. • Increases the energy ratio for ethanol production from 1.34 Btu ethanol output for every

nonsolar Btu input to 2.9 Btu ethanol output for every nonsolar Btu input. Since fuel costs are the second or third largest cost for an ethanol facility (behind corn and, possibly, debt service), this last benefit makes production of ethanol more competitive with production of oxidants (eg, gasoline) from petroleum.

The project also had effects on the distillers dried grain coproducts. These effects were increasing the protein content of the coproducts and lowering the energy (fat) content. In addition, the amount of distillers grains produced decreased by 19% due to reduced syrup being processed with the distillers grains. While the mass balance of the distillers grains would lead one to believe the decrease in distillers grains produced would be greater, it was determined the dryers actually were incinerating a significant portion of the syrup (eg, 33 to 50%) in the drying process. Removing the syrup from the coproduct being dried did not have as much of an impact as original mass balances indicated. FRACTIONATED GRAINS AND THIN STILLAGE INTRODUCTION Research was conducted for facilities investigating closed loop concepts, where all energy production (heat and electricity) is on site from nonfossil fuel sources. The reasoning behind the research was that, by fractionating incoming grain, more overall value was available from the

120

Flu

id B

ed

Rea

ctor

D

DG

Dry

er

Was

te H

eat

Boi

ler

Bag

hous

e &

Em

issi

ons

Stac

k F

luid

Bed

Blo

wer

Figu

re 1

. C

once

ptua

l sch

emat

ic.

121

grain, both in the form of energy and in the form of high protein coproducts. With a few key process changes, such a closed loop is possible (Harris Group 2006). The incoming grain was fractionated in this process, yielding bran, germ and grits (endosperm). Bran is a low value coproduct, with much more energy value than food value. Oil in the germ is a high value coproduct, while germ cake resulting from oil removal is of somewhat lesser value as a coproduct, but not as low as bran. Grits, composed of starch and protein, were fed into a fermentation and distillation process to make ethanol, with a high value, high protein distillers grain coproduct resulting. Thin stillage from the ethanol process was separated and sent to a set of anaerobic digesters to convert the solubles to biogas. Thin stillage has a high COD (eg, 100,000 to 150,000 mg/L), which makes it an ideal candidate for digestion. Water was recovered from the digestion process for reuse in the ethanol process. RESULTS Biogas provided sufficient energy to run a combined cycle combustion turbine which, in concert with steam from a solid fuel boiler system fueled by bran and some germ cake to run a steam turbine and provide extraction steam, provides all facility heat and electrical needs (Figure 2). More energy per pound of grain is available in this scenario when compared to the syrup burner, due to the lack of water (requiring additional Btus to flash off) in the fuel feedstock. In addition to the energy available, CO2, oil recovered from the germ, high protein distillers grains and pelletized ash from burning of bran and germ cake are salable coproducts (in addition to ethanol sales). Due to the confidential nature of the studies, actual payback or return on investment data is not available, but can be stated as adequate to have attracted investment capital for a project to commence this fall.

Figure 2. Diagram of ethanol plant equipped to provide facility steam and electricity.

122

LITERATURE CITED

DOE. 2006. Net energy balance for bioethanol production and use. US DOE, Energy Efficiency and Renewable Energy Biomass Program. February 8. www1.eere.energy.gov/biomass/printable_versions/net_energy_balance.html

Waffenschmidt, D. 2001. Michaels engineering DDGS biofuels feasibility study. (Not publicly released).

Mudgett, H. 2002. Operation summary: evaluation of fludized-bed combustion of a corn byproduct derived form ethanol production. Hazen Research, August 23. (Not publicly released).

Harris Group. 2006. Feasibility analysis of closed loop biomass to energy ethanol plan. Summer. (Not publicly released).

123

BENCHMARKING INDUSTRIAL ENERGY PERFORMANCE: THE ENERGY STAR APPROACH

Gale A. Boyd1 and Walt Tunnessen2*

1Duke University, Durham, NC and

2Climate Protection Partnership Division, US EPA, Energy Star, MS 6202J, 1200 Pennsylvania Avenue NW, Washington, DC 20460

(202-343-9965) tunnessen.walt@epamail.epa.gov ABSTRACT Benchmarking energy performance is key for driving energy efficiency. While this is understood by most industrial energy managers, access to industry wide energy performance benchmarks usually are unavailable. Consequently, energy managers frequently do not know if their most efficient plants actually are efficient when compared to the entire industry. To overcome this issue and to encourage greater energy performance within an industrial sector, the US Environmental Protection Agency (EPA) Energy Star program works with selected industries to develop plant level energy performance benchmarking and rating tools called energy performance indicators (EPI) which enable energy managers to evaluate the relative performance of their plants as compared to the industry and provides the basis for recognition for superior achievement in the form of the Energy Star plant label. To address differences among plants, the Energy Star approach to benchmarking draws on statistical methods to develop an underlying model that is the foundation of the EPI tool. This approach requires the most important drivers for energy use within an industry be identified and normalized to provide a meaningful comparison. It has been used to develop a plant level energy performance benchmarking tool for corn wet mills. We will describe the collaborative approach with industry the Energy Star program uses through industry focuses to develop these empirical comparisons. We will provide examples from glass, pharmaceuticals, food, paper and other industrial focus initiatives and discuss the steps used to insure confidentiality of the plant/company data but allows a plant to compare itself against its peers. We will describe how the EPI benchmarking tool is used for awarding the Energy Star to qualifying industrial plants. INTRODUCTION The term “industrial competitiveness” describes a process for firms striving to attain higher levels of performance and profitability. They do so by producing and selling more products at lower costs and higher profits. Firms compare themselves to their peers using basic metrics in this race to achieve higher performance. The performance metrics of production and sales is apparent in the market place but cost comparisons are more elusive. Firms use a variety of benchmarking approaches to develop the cost comparisons. Since energy is one of many costs faced by industry, benchmarking energy performance is critical to driving energy efficiency in the race for competitiveness. While this is understood by most industrial energy managers, access to industry wide energy performance benchmarks usually is

124

unavailable. Consequently, energy managers frequently do not know if their most efficient plants actually are efficient when compared to the entire industry. By not having an industry wide energy performance benchmark, many energy managers also do not know what is achievable. As a result, some firms and energy managers may believe that additional cost savings from greater energy efficiency are not possible or even necessary. This lack of information of what is achievable in effect becomes a barrier to greater energy efficiency. To overcome this issue, the US Environmental Protection Agency (EPA) Energy Star program has developed tools to enable meaningful plant level benchmarking of energy performance within a variety of industries. The Energy Star energy performance indicators (EPI) enable the evaluation of relative performance and provides the basis for public recognition for superior achievement in the form of the Energy Star plant label. To develop the EPIs, the Energy Star program uses a collaborative process with companies in several industries. We will describe the collaborative approach with industry that the EPA uses through the industry focuses to develop these empirical benchmarking tools. We will discuss the steps used to insure confidentiality of the plant/company data but allows a plant to compare itself against its peers. We will provide examples from glass, pharmaceuticals, food, paper and other industrial focus initiatives and how the EPI benchmarking tool is used for awarding the Energy Star to qualifying industrial plants. THE INDUSTRY FOCUS PROCESS The detailed nature of energy use in manufacturing is industry specific but the basic interest in benchmarking is not. To address the industry specific nature of energy management issues and the specific needs for a benchmarking tool tailored for individual manufacturing sectors, EPA organizes Energy Star industrial focuses. The focus process involves direct interaction between the Energy Star team with energy managers from companies within a sector via a series of conference calls and an annual workshop. Asdescribed on the Energy Star web site, the process focuses build upon the energy management resources available through Energy Star, and within an industry: 1) engage energy managers in discussions of strategic, corporate energy management, 2) provide tools to enhance energy management and performance, 3) uncover energy saving opportunities and practices and 4) encourage sharing of energy management techniques (Energy Star 2007). The collaborative nature of the industrial focus is critical for creating an energy benchmarking tool that will be accepted and used by the industry. The process begins by introducing the basic concept of a plant level energy benchmark that takes into account major difference among plants that influence energy use. Energy managers are briefed on the data sets available to the Energy Star team through existing sources such as the Census Bureau. Industry energy managers are asked to help identify these major plant differences and ways to capture and measure such variables using existing available data. If existing data sets are considered insufficient, additional sources, including voluntarily supplied industry data, are considered.

125

To address differences among plants, the Energy Star approach to benchmarking uses a statistical analysis. The statistical approach used to develop the EPI is described in more detail for automobile assembly, cement and corn wet milling in Boyd 2005a, 2005b, 2006a, 2007b. This requires that the most important drivers for energy use within an industry be identified to provide a meaningful comparison. While the specifics vary across sectors, the approach considers four major effects: 1) product mix, 2) physical plant size or productive capacity, 3) process inputs and 4) external variables, such as weather and utilization rates. Each one of these broad categories may have a specific measurable impact on the energy use of a manufacturing plant. The analysis to determine the magnitude and significance of each of these impacts and the process of normalizing these effects to determine the range of industry performance is conducted by the Energy Star statistical team. Statistical methods are used to measure how far any plant in the data base is from a predicted level of energy use, based on the plant level data for the basic energy drivers. This gives a measure of performance for every plant from the industry model predicted norm, ie, a normalized energy efficiency. The range of normalized performance from the best plant to worst plant is summarized as a statistical distribution that assigns a percentile score of 1 to 100 to every plant. This process is used to develop the underling equations and statistical distributions that result in a set of multivariate relationships that is the EPI. The equations and distributions in the EPI is presented as an Excel spread sheet and is designed to be user friendly. When a draft version of the EPI is completed by the Energy Star team, the industry testing and review process begins. To test the EPI, industry participants are asked to put either representative or actual data from plants in their companies into the EPI spreadsheet and consider the results. Typical data inputs include production, capacity, utilization, products, materials and energy. The results of the EPI reflect the metrics the industry typically uses or is interested in seeing. Typically these performance metrics are based on a unit of production, such as MMBTU per vehicle in the auto assembly EPI. Additionally, all EPIs generate an Energy Star rating which is provided on a scale of 1 to 100, with a score of 100 representing the highest performance currently achievable. The rating provides the benchmark for gauging the energy performance of a given plant. The rating reflects what percentile of energy performance the plant is operating in, ie, a plant with and EPI rating of 80 is considered to be performing better than 80% of similar plants in the industry. Once energy managers run the numbers in the EPI spread sheet, they are asked to consider the following general questions to evaluate the EPI results. 1) Are the results intuitive? Why or why not? 2) Do you observe patterns, eg, plants producing similar products tend to be rated

as more (less) efficient? 3) Do the scores from the plant data used in testing conform to your expectations,

eg, does a plant viewed as a poor performer score low? 4) Does the ranking of multiple plants conform to your expectations? 5) Does the change over time in a specific plant conform to your expectations?

126

The development and testing is an iterative activity. Energy managers provide feedback and suggestions of additional measures to include in the model. Additional or different factors are considered and adjustments to the model are made. The statistical methods are used to test if hypothesized effects identified by energy managers are significant and if the estimated impact on energy use conforms to general expectations. If so, it is included in the next version of the model. When industry and Energy Star both are satisfied the model works and provides a fair gauge of energy performance, the EPI version is considered final and distributed via the Energy Star web site. The analysis may be updated. This allows for the benchmark to evolve as industry changes and newer data that reflect those changes become available. PROTECTION OF CONFIDENTIAL INFORMATION Plant data on energy and other costs typically are viewed by companies as confidential business information (CBI). Energy Star has worked to eliminate the need for CBI and provide protections to companies that participate in the focus. The first way Energy Star does this is to make use of existing government collected data that are protected by law from disclosure. The second way is to construct the EPI so that no individual plant information is embodied in the equations or spreadsheets and that companies can use the EPI without disclosing the data or results to anyone else. The EPI analysis uses confidential plant level data from the Center for Economic Studies (CES), US Census Bureau. Data from CES includes nonpublic, plant level data which are the bases of the government statistics on manufacturing. Title 13 of the US Code protects this data. CES allows researchers with special sworn status to access these confidential microdata at a Research Data Center (RDC). Census confidentiality rules and procedures prevent the disclosure of any information that would allow for the identification of a specific plant or firm’s activities. Duke University is an institutional partner with CES which provides access to this research project to this confidential data. Members of the Energy Star team are have special sworn status from CES, which has reviewed and approved the use of the data for this purpose. The advantage of using available data is there are no burdensome reporting requirements, the information is reported on a consistent basis for all plants and confidentiality is assured. The final form of the EPI results and corresponding spreadsheets do not contain any plant level data, only the statistical relationships described above. Company energy managers can use the spreadsheets in the privacy of their offices and do not need to reveal CBI to anyone else when using the model. Additional protections for CBI are discussed below regarding plant level recognition. EXAMPLES OF NORMALIZATION IN THE EPI A basic metric for normalization for benchmarking is a ratio of energy use to some common activity. This is the simplest example of normalization. This ratio captures the link between energy use and the most basic differences among or within plants over time. The denominator may be production output, physical size or amount of materials processed. The EPI uses a multivariate approach to normalization where multiple effects are considered

127

simultaneously. The next sections give industry specific examples of the four basic categories of effects that are considered. Product Mix Energy is the derived demand for energy services used in support of various manufacturing processes. The results of these processes are various intermediate and final products produced by the plant. It is common to evaluate energy efficiency use in terms of the intensity of energy input relative to desired energy services (eg, per lb compressed air), relative to a particular intermediate process (eg, per lb crushed limestone) or relative to the final product of the plants (eg, per ton finished cement shipped). Each of the energy services or intermediate processes contributes to the overall energy use, hence energy efficiency, of the entire manufacturing plant. However, not all plants produce exactly the same product. In fact, many plants produce multiple products. The diversities within and among plants give rise to a mix of derived demands for specific processes and energy services. To the extent the final product is the result of a series of energy using steps, the energy use of the plant will depend on the level and mix of products produced. Rather than specifying each process step individually, the approach used here is to identify those products that use significantly more (or less) energy and measure those energy requirements with a statistical comparison. An understanding of the production process is needed to identify what products to consider and how to specify the relationships we wish to estimate. This understanding comes from related Energy Star research, from group meetings and one on one discussions with focus participants from the respective industries. One approach to controlling for product mix is to segment the industry into natural product categories. This works best when there is no overlap among plants that produce the various basic products and there are sufficient numbers of plants to conduct the statistical comparisons among those resulting groups. This means each subgroup is treated as a separate industry for evaluation proposes. The glass industry is a good example, since flat, container and fiber glass are distinct products and each sector can be treated in a “stand alone” manner, although within group variation may exist. When such natural subsectors do not exist and multiple products are produced within a plant, additional approaches are needed. Even though the cement industry primarily produces a single well defined commodity, some plants produce smaller amounts of specialty cements, eg, for masonry or oil well applications. If specialty products require different energy use the statistical approach can estimate these differences. Corn wet millers have a common underlying process involving separation of gluten from starch. This separation and resulting preparation of the animal food coproducts for shipment results in similar basic energy demands for corn refineries. Differences arise from further treatment of the intermediate product, corn starch. It may be dried as a final product, further modified or used as a feedstock for sugar or ethanol production. It is the mix of these downstream products that can be used to identify differences in process energy service demand and the benchmark for a plant with a specific product mix. For some sectors the diversity of products is so large it is necessary to consider broad groups of products or activities. For example, fruit and vegetable canning and freezing involve a

128

diverse range of foodstuffs and final products. To make the benchmarking tractable it may be appropriate to add up apples and oranges, so long as they are both being made into frozen fruit juice. The statistical approach used by Energy Star is well suited to testing if a particular grouping of products is appropriate for benchmarking differences in energy. Other industries, like pharmaceuticals or autos, that also have a diverse range of products may be treated differently for benchmarking. For automobile assembly, the size of the vehicle turns out to be a good measure for product differences when evaluating energy use. This is because energy consumption in auto assembly is dominated by the painting process. For pharmaceuticals there is enormous product diversity but major differences in energy use arise among the three basic activities in pharmaceuticals, R&D laboratories, active ingredient preparation and fill and finish. While there are many products within those areas, these basic production activities are viewed as principle drivers of energy use in this sector. Physical Size To include size as a normalizing factor in the EPI, a meaningful measure of size or capacity is needed. It may be measured on an input basis (corn wet milling and cement), output basis (auto assembly) or physical size (pharmaceuticals). In some cases, there may be advantages to larger scales of production. If it is the case that a larger production capacity or larger physical plant size has less than proportionate requirements for energy consumption, there are economies of scale with respect to energy use. For several sectors, including auto assembly, corn wet milling and pharmaceuticals, the EPI development tested models that would capture the bigger is better phenomenon. This was not found to be the case for auto assemblies and corn wet milling; the results are not final for pharmaceuticals. For cement, it was found that larger kilns were an advantage but larger numbers of kilns were not. Regardless of the industry specific results obtained to date, size and economies of scale remains an important area for normalization in the benchmarking approach. Process Inputs Other process inputs can be helpful in developing a statistical benchmark of energy use. Process inputs such as materials, labor or production hours may be good proxy measures of overall production activity when measures of production output are not available or have specific shortcomings. If complex or highly variable pricing is used to compute the total value of shipments (TVS) of a plant, then TVS may not be a good measure of production to compare energy use (Freeman et al 1997). If a physical measure of output is not readily available and pricing makes the value of shipments a questionable measure of production, then physical inputs can be a useful proxy. For some industries, the basic material input is so ubiquitous that it makes sense to view energy use per unit of basic input rather than (diverse) outputs. Process inputs also may be useful in measuring utilization, either directly or indirectly. Corn wet milling is an example of a sector where the energy use per unit of material input, ie, corn processed, is appropriate. The energy use at the plant, expressed in terms of bushels of corn processed, will be influenced by the mix of final products but include the energy

129

use for processing coproducts common to all plants. In this manner, the level of corn processed captures a number of common energy components in a succinct fashion. The level of corn processed also is a good way to capture plant utilization, since the capacity of a plant commonly is expressed in terms of corn volume processing capability per day. Sometimes physical production data are lacking in some way but material flows can be used. For example, sand, lime, soda ash and cullet (scrap glass) are the primary inputs to glass manufacturing. Since the Census only collects data on the value of glass shipped, these basic materials provide a good control for the level of physical production at a glass plant. Moreover, if the materials mix to produce different types of glass directly impacts energy uses, the statistical model can apply different weights to the materials mix in the same manner it does with product mix. In other words, product/process level differences in energy use are inferred from the volumes and types of materials used in production. Differences in material quality also may be considered in the statistical normalization, if they are measured on a consistent plant level basis across the industry. For cement plants, the hardness and moisture content of the limestone may influence energy use but no consistent data are available for this, leaving it the subject of future analysis if data can be collected. When levels of materials or outputs are not measured in common units and value units may introduce other problems, production labor rates may control for differences in production activity between plants and differences in utilization rates within plants. While the link between energy and labor is not as direct as energy and production, the fact that it takes both labor and energy to manufacture a product allows an indirect link to be estimated. One advantage is that labor hours can provide a common denominator in terms of measurement. External Factors There are many things under the control of a plant or energy manager, but one they cannot control is the weather. In most manufacturing plants, heating, ventilation and cooling (HVAC) contributes to energy demand and weather determines how much is required to maintain comfort. Since the benchmarking approach used here is annual seasonal variation, it does not enter into the analysis but differences due to the location of a plant and annual variation from the locations norm will play a role. The approach that has been taken for all sectors under study is to include heating and cooling degree days (HDD and CDD) into the analysis to determine how much these location driven differences in weather impact energy use. In principle, all plants have some part of energy use that is HVAC related but when the HVAC component of energy use is small relative to total plant consumption the statistical approach may not be able to measure the effect accurately enough to meet tests for reliability. For sectors like automobile and pharmaceutical manufacturing, the approach finds significant impacts of HDD and CDD on energy use. For sectors such as cement, glass, food processing and corn wet milling, we have not been able to estimate any impact so these factors are treated as de-minis for the purposed of annual, plant level benchmarks. Other location dependent impacts can be included, or at least tested, using the statistical approach. As part of the focus review process, altitude was proposed as having an effect on

130

cement kiln energy use, because of combustion oxygen differences. Altitude was included in the analysis of the cement energy data, but no measurable effect was seen. This type of hypothesis testing allowed by the statistical approach provides for a dialogue between researchers and industry participants to understand the drivers of energy use in the various sectors. This last example underscores the role of the focus process in developing the EPI. The analysis is not conducted in a vacuum. This understanding comes from related Energy Star research (eg, the Energy Guides) and from group and one on one discussions with focus participants from the respective industries. Industry participants provide guidance on what factors they feel are contributors to energy use in their respective sectors. The Energy Star research team suggests ways these may be measured or proxied and represent those effects in a statistically testable form. If the effect can be measured reliably in a statistical sense, using the data available, the effect is included in the EPI. RECOGNIZING EFFICIENT PLANTS In addition to offering industrial energy managers a tool to gauge their plant’s energy performance to the industry, the EPI provides the basis for providing public recognition of superior energy performance. Rating and recognizing products, homes, buildings and industrial plants that are energy efficient has been the hallmark of the Energy Star program. By earning the Energy Star, plants can distinguish themselves within their industry and demonstrate effective energy management to their customers and the community. Corporate energy managers can use the Energy Star to reward efficient plants and motivate less efficient ones. Today, the Energy Star brand is one of the most widely recognized brands in the US with a brand recognition rate that exceeds 65%. As discussed earlier, the EPI provides an energy performance rating on a scale of 1 to 100, with 100 being the best energy performance possible. This rating reflects the plant’s energy performance when compared to the industry. For example, a plant that scores a 75 or higher would be in the top quartile of observed energy performance within the industry. If a plant’s rating in the EPI is a 75 or higher, the plant is considered by the EPA to be energy efficient and is eligible for being awarded the Energy Star. In August 2006, the EPA established procedures for companies whose plants have an EPI score in the 75th percentile or higher to apply for the Energy Star. To be awarded the Energy Star, the plant’s EPI score must be based on 12 mo of recent production and energy data, the data used to generate the EPI score must be verified as accurate by a professional engineer, the score must be EPI verified by the EPA or a designated reviewer (the designated third party reviewer provides protection of CBI to the company under a legal nondisclosure agreement.) and the plant must not be involved or have had any major violations of the federal Clean Air Act. On September 13, 2006, EPA officially announced the first 17 plants to receive Energy Star recognition for superior energy efficiency in their respective industry. At that time, 7 companies from 3 industries, automobile assembly, corn wet milling and cement manufacturing, received awards. Several more companies are either in the application process or stated their intent to apply in the near future. For just those 17 plants, differences between actual energy use

131

and energy use of a similar plant performing at the 50th percentile in terms of energy efficiency amounted to 3 billion lb/yr CO2 emissions that otherwise would have been produced. Since then, several other facilities have been awarded the Energy Star. Currently, there are more than 25 industrial plants in the US that have been awarded the Energy Star. As additional industry specific analyses are completed, Energy Star recognition will be extended to those industries as well. SUMMARY In this paper, we provide an introduction and overview to the process of developing manufacturing plant energy efficiency benchmarks. The process involves interaction with representatives from companies within a selected industry in an industry focus. This focus includes many activities, one of which is the development of the EPI. The analytic approach to the EPI is based on a statistical methodology but the focus process is a critical component. Energy Star does not develop the EPI in a vacuum. Industry testing and feedback is critical to the EPI development process. The result is a bird’s eye view of plant energy performance relative to the range of comparable plants in the industry. Since no plant is the same as another the statistical approach provides the normalization and distribution used to rank a plant from 1 to 100. This percentile score allows energy managers to assess individual plants on an industry wide basis. The percentile score also is the basis for EPA recognition through awarding a manufacturing plant Energy Star.

LITERATURE CITED Boyd, G.A. 2006. Development of a Performance-based Industrial Energy Efficiency

Indicator for Cement Manufacturing Plants. Argonne National Laboratory, ANL/DIS-06-3. July.

Boyd, G.A. 2006. Development of a Performance-based Industrial Energy Efficiency Indicator for Corn Refining Plants. Energystar.gov/ia/business/industry/ANL-DIS-06-04pdf. Argonne National Laboratory, ANL/DIS-06-4. July.

Boyd, G.A. 2005. Development of a Performance-based Industrial Energy Efficiency Indicator for Automobile Assembly Plants. Argonne National Laboratory, ANL/DIS-05-3. May.

Boyd, G.A. 2005. A method for measuring the efficiency gap between average and best practice energy use: the Energy Star™ industrial energy performance indicator. J. Industrial Ecol. 9:51-65.

Energy Star. 2007. Industries in Focus. Energystar.gov/index.cfm?c=in_focus.bus_industries_focus. Accessed March 13. US EPA and Department of Energy, Washington, EC.

Freeman, S.L., Niefer, M.J. and Room, J.M. 1997. Measuring industrial energy intensity: practical issues and problems. Energy Policy 25:703-714.

132

MEMBRANE PROCESSES - OPPORTUNITIES IN CORN PROCESSING

William J. Koros*

Georgia Institute of Technology, Atlanta, GA 30332-0100

(404-385-2845) wjk@chbe.gatech.edu

ABSTRACT

As corn is considered increasingly a commodity fuel as well as a food, processing efficiency becomes more important; membranes may offer opportunities in this regard. The term “disruptive technology” is sometimes used to describe approaches that are radically different from the incumbent technology in a field. Typically, such a technology initially performs less well than the incumbent leader but by improvement, becomes a dominant force in the field. In this context, membrane processes may be disruptive technologies. The energy required to achieve some large scale separation tasks, for instance water purification, can be as much as a full order of magnitude lower via the membrane approach vs thermal alternatives. Polymeric media, ceramic, carbon, zeolitic and metal membranes are appealing from the standpoints of separation precision and chemical inertness. Based on such materials, lower cost robust membrane modules are required to be practical. Viable ways will be considered to achieve this goal to enable such advanced materials to have an impact in various applications, including corn processing. INTRODUCTION

Membrane separation processes offer theoretical advantages in energy efficiency relative to conventional thermally driven counterparts. Significant energy challenges face the corn processing industry. These two facts present an opportunity; however, energy issues have not been pursued as aggressively in corn processing as in other industries. It is useful to consider possible hurdles that must be overcome to change this situation.

Considering cases outside the corn processing industry provides a broad perspective on

current and developing capabilities of membrane technology. As editor in chief of the Journal of Membrane Science, the author is a membrane expert, but certainly not a sophisticated practitioner of corn processing technology. As such, I will seek to improve dialog between technology developers and technology users to identify realistic opportunities and hurdles to realizing those opportunities in corn processing.

In this spirit, an overview of the nature of membranes and their fundamental advantages over standard thermally driven separations will be provided. Nevertheless, some realities regarding the challenges to replacing energy inefficient separation processes will be highlighted. Some of these challenges related to concentration polarization in micro-, ultra- and nanofiltration will be noted. Indeed, related issues are faced in bioprocessing as well as water treatment and recovery areas. While general concepts related to these challenges are understood readily, approaches to deal with them are specific to each application and can be complex. Specific solutions to niche problems, therefore, will not be the focus of this presentation.

133

Therefore, while processing variables are ultimately critical, I will focus instead on an approach to make additions to the membrane toolbox comprising economical new materials and structures. This latter approach was the one that ultimately enabled membranes to capture the market lead in large scale micromolecular separations of aqueous feeds via reverse osmosis. Indeed, despite their known theoretical advantages, membranes for desalination of brackish and sea water only recently have become accepted as the preferred technology in that area. Achieving a similar advance in core areas of corn processing would have major economic impact.

As is always the case in practical situations, reducing cost was a key factor in enabling

the transformation for the aqueous separation case. When the need to integrate four critical capabilities noted below was understood, improved materials, lower cost and higher reliability manufacturing processes were introduced. This enabled the revolution in aqueous reverse osmosis and created a large new industry with positive environmental and energy benefits.

Diverse types of materials and modules are needed to extend the advantages offered by

membranes beyond the first generation examples noted above. In fact, looking beyond liquid feeds is critical, since including gases and vapors opens new opportunities. Carbon dioxide recovery and purification is the most obvious gas separation application of interest to corn processors. Other examples such as gas management to control and optimize fermentation by respiration may be relevant. Factors such as oxygen tension, via use of inexpensive controlled atmospheres may be of interest. In any case, options that membranes offer will be considered to stimulate thoughts where this revolutionary technology may fit in the framework of innovative corn processing. FOUR ESSENTIAL ELEMENTS OF MEMBRANE TECHNOLOGY High efficiency modules with large amounts of membrane area per volume are a critical need to perform large scale separations. The numbers are impressive: for instance, hollow fiber modules can contain 10,000 m2/m3 of module, which is more than 100 times larger than early plate and frame units (Baker 2004). Advanced materials with tunable capabilities to separate diverse components have been a second key factor in the emergence of membranes as an applicable technology platform. Thanks to materials scientists, a rich array of choices now exists for the membrane technologist (Kesting 1985, Pinnau and Freeman 1999, Pixton and Paul 1994, Buxbaum 1993, Nair and Tsapatsis 2003, Akin and Lin 2002). While the materials themselves are important, modern membranes did not appear until the sophisticated capability was developed to control microscopic transport phenomena by tailoring morphology at multiple levels within a membrane cross section. For instance, in the thickness dimension of Figure 1, a submicrometer ultrathin selective skin region is supported atop low resistance transition and microporous substrate layers (Carruthers et al 2003). Engineering the properties of the ultra thin top layer enables providing this skin with micro-, ultra-, nanafiltration or micromolecular separation capabilities. In the case shown in Figure 1, the scale of these critical features is molecular in nature and too fine to be imaged even with the highest resolution microscopy. Recent examples of such morphology engineering involves hybrid structures comprising preassembled micro- or nanoscale functional entities dispersed in an engineered supporting matrix (Figoli et al 2002, Mahajan and Koros 2000). Such materials usually are referred to as

134

“mixed matrix” media, and they offer efficient “offline” engineering of the functional selective entities without compromising rapid economical production of large surface area modules. The last element needed to be able to develop a practical membrane module is the availability of an adequately high speed manufacturing method to rapidly link the above three elements into economical devices with minimal defects (Eykamp 1997).

0.1-0.2 μm selective skin region

transition region

microporoussubstrate

200 μm OD Fiber

0.1-0.2 μm selective skin region

transition region

microporoussubstrate

200 μm OD Fiber

The above four key elements needed to introduce any new type of membrane process, or even a new generation of the same type of membrane process are summarized in Figure 2. Also emphasized is a lesson learned in the introduction of early membranes for energy efficient separations: interlinkages among these elements also is crucial. Moreover, good connectivity at interlinkage boundaries between the four key elements is needed by viewing membrane technology holistically as an integrated subdiscipline within the chemical engineering paradigm (Koros 2004).

Figure 1. Typical asymmetric hollow fiber illustrating a thin selective layer and an open porous substrate. The selective layer may be dense, as shown here, or have tailored micro or nano pores that enable either molecular diffusion or viscous flow separation phenomena to occur. For high pressure operation, the fiber can be fed on the shell side, but for low pressure cases, bore side feed usually is preferred.

135

Advanced materials

High efficiency

modules

High speed

manufacturing

Micro-morphology control

Figure 2. Enabling elements needed for modern synthetic membrane technology.

Despite demanding requirements for selectivity and robustness in this next generation

of applications, membranes and modules must retain their attractive cost advantages. Realistically, therefore, any program to introduce truly new high performance membranes should incorporate hybrid materials within its enabling vision. A complete picture of membrane materials includes the spectrum ranging from purely inorganics and carbons to purely organic polymers. Current work has explored only the two extreme ends of this spectrum, plus a few hybrids containing 10 to 15% inorganic or carbon dispersed phases in a polymer continuous phase. For future demanding applications, the optimum position in the materials spectrum may be near the “midpoint” in hybrid composition. INTERSECTION OF MEMBRANE PROPERTIES AND PROCESSING CONDITIONS

Membrane processes can be deceptively simple, and therein lies a potential pitfall for the uninitiated practitioner. For instance, the ability to package huge surface area per unit volume was noted correctly earlier as an advantage of hollow fibers vs plate and frame units. While this advantage is attractive, in complex feeds with large membrane fouling tendencies, the ability to engineer high cross flow shear rates uniformly throughout the module may be of greater importance. In such a case, the tradeoff may require opting for lower volumetric module productivity to minimize cleaning requirements. In other cases, where particulate fouling is not the limiting issue, the low cost of hollow fiber membrane area per unit volume may allow operation with low transmembrane pressure drop. This option may enable running below the critical flux for serious concentration polarization, and thereby improve selectivity and actually suppress irreversible gel induced fouling in some cases.

136

CONCLUSIONS Whatever the situation and constraints involve, greater sophistication is now available to create advanced membrane structures, engineered modules and to predict their use with diverse input streams. This greater sophistication is valuable, is opening the door to broader use of this exciting technology and will enable major energy savings relative to competitive thermal options by introducing membrane processes for separations. Nevertheless, it clarified the need for a large scale integrated systematic approach to broaden the economical application of membranes to more challenging feed streams. This information highlights the need for modeling and analysis that starts at megascale plant systems and ranges down to the nanoscale where most separations ultimately occur. Materials science is a critical component; however, technologies to engineer supermolecular membrane morphologies and economical modules are equally critical to build such an expanded platform. Over the past decade or more, the global scientific community has taken action to promote the introduction of fuel cells, advanced batteries and solar cells; however, much less aggressive action is apparent to promote energy efficient separations. A concerted program focused on developing the membrane platform beyond its current state to enable rapid replacement of energy inefficient separation processes is needed badly.

LITERATURE CITED Akin, F.T. and Lin, Y.S. 2002. Oxidative coupling of methane in dense ceramic membrane

reactor with high yields. AIChE J. 48:2298-2306. Baker, R.W. 2004. Membrane Technology and Applications. 2nd ed. Wiley, West Sussex,

England. Buxbaum, R.E. 1993. Hydrogen transport through non-porous metal membranes of palladium-

coated niobium, tantalum and vanadium. J. Membrane Sci. 85:29. Carruthers, S.B., Ramos, G.L. and Koros, W.J. 2003. Morphology of integral-skin layers in

hollow-fiber gas-separation membranes. J. Appl. Poly. Sci. 90:399-411. Eykamp, W. 1997. Perry’s Chemical Engineers’ Handbook. Perry, R.H., Green D.W., eds.

Chap. 22, 7th Ed. McGraw Hill, New York, NY. Figoli, A., Sager, W. and Wessling, M. 2002. Synthesis of novel nanostructured mixed matrix

membranes. Desalination 148(1-3):401-405. Kesting, R.E. 1985. Synthetic polymeric membranes: a structural perspective. 2nd ed. Wiley

Interscience, New York, NY. Koros, W.J. 2004. Evolving beyond the thermal age of separation processes: membranes can

lead the way. AIChE J. 50:2326-2334. Mahajan, R. and Koros, W.J. 2000. Factors controlling successful formation of mixed-matrix

gas separation materials. Ind. Engr. Chem. Res. 39:2692-2696. Nair, S. and Tsapatsis, M. 2003. Handbook of Zeolite Science and Technology. Auerbach,

S.M., Carrado, K.A. and Dutta, P.K., eds., pp. 867-919, Marcel Dekker, New York, NY. Pinnau, I. and Freeman, B.D. 1999. Membrane formation and modification. Pinnau, I. and

Freeman, B.D., eds. ACS Symposium 744, Chap. 1. American Chemical Society, Washington, DC.

Pixton, M.R. and Paul, D.R. 1994. Polymeric Gas Separation Membranes. Paul, D.R. and Yampol’ski, Y.P., eds. Chap. 3, CRC Press, Boca Raton, FL.

137

POSTER ABSTRACTS

138

DEPOSITION CONTROL IN BIOPROCESS EQUIPMENT TO INCREASE PLANT EFFICIENCY

Carol B. Batton*

Nalco Company, 1601 West Diehl Road, Naperville, IL 60563-1198

(630-305-2148) cbatton@nalco.com

Grain can be processed in either wet or dry operations. Corn wet milling allows

separation and isolation of all corn components. Some components (germ meal, gluten feed,

gluten meal) are used in animal food. In dry grind, the primary operation is ethanol

production. Two coproducts are produced, ethanol and distillers dried grains with solubles

(DDGS).

Deposition and scaling impact several mill operations, whether the plant is wet milling

or dry grind. New problems are occurring due to many operating plants wanting to push

beyond the 120% nameplate rating. Traditionally, deposition and scaling were found in

evaporators and heat exchangers. In this operationally challenged environment, deposition is

occurring in areas of the plant that were not previously a problem. Deposition affects the

efficiency and economics of the plant and heat transfer is reduced. Heavy build up actually

can block flow, resulting in unplanned shutdowns. Reduced heat transfer, blocked flow and

shutdowns collectively result in significant production bottlenecks, increased operational

costs and decreased revenue.

Our approach to problem solving involves review, evaluation and recommendation of

mechanical, operational and chemical solutions. In an attempt to address proactively the

types of problems encountered in the grain industry, we have collected and identified a

number of deposits from various pieces of equipment. A two pronged approach to address

process deposition has been taken: 1) evaluation of products for the removal of existing

deposits and 2) chemical products to inhibit deposit formation. Laboratory findings and test

protocols used in evaluating the functional chemistries will be provided. Increased production

from improved deposition control and resulting energy cost savings will be detailed from a

case study.

139

140

OIL, CORN AND ETHANOL: NO MORE CHEAP FOOD

Ronald L. Belyea1*, Joe L. Horner2, Kent D. Rausch3 and M.E. Tumbleson3

1Animal Sciences and 2Social Science and Community Agriculture, University of Missouri,

Columbia, MO, 65211 (573-882-6354) belyear@missouri.edu and 3Agricultural and Biological Engineering, University of Illinois at Urbana-Champaign,

Urbana, IL 61801

The increase in ethanol production that has occurred during the past several years will continue with pervasive effects on the agriculture sector. Currently, ethanol production capacity is 5.4 billion gal/yr, equivalent to 1.9 billion bu corn (16% of the current corn crop). Plant capacity under construction will increase ethanol production to 11.2 billion gal/yr, using 4.1 billion bu corn. By 2009, production could reach 20 billion gal/yr and consume an amount of corn equivalent to half of the current corn crop. However, it is questionable if ethanol production will attain this level, because of various economic factors. These include petroleum prices, commodity supports and corn prices.

The world petroleum price, which has stabilized at $60/barrel, will decline to $50/

barrel by 2016 (FAPRI 2007). This could make ethanol less profitable. Government (US) supports are important drivers; these expire in 2008; without these incentives, ethanol production could be impacted negatively. Finally, corn prices and availability are major factors; corn is the single largest input cost in ethanol processing. Droughts and other climatic factors that reduce the corn crop could have large effects on corn prices and economics of ethanol production.

Increased ethanol production already is impacting corn prices. FAPRI (2007) projects

corn prices to average $3.17/bu in 2007 and $3.23/bu in 2008, well above the average $2.00/bu in 2005. They suggest corn prices will exceed $3.00/bu in each of the next 10 yr, while others suggest corn prices could be much higher. Increased corn prices will increase corn production by 85 million bu in 2007 and possibly by 90 million bu in 2008. Increased corn prices will have important impacts on the animal industries, reducing animal numbers and increasing production (food) costs substantially. These effects will result in increases in retail prices; the amount is unclear. Consumers probably will experience noticeably higher prices for most animal based food products; consumption of animal products could decline.

LITERATURE CITED

FAPRI. 2007. US Baseline Briefing Book. Rept. No. 02-07. Food and Agricultural Policy

Research Institute. University of Missouri, Columbia, MO.

141

142

COPRODUCTION OF FUEL ETHANOL AND NEW VALUE ADDED COPRODUCTS

David B. Johnston*

Eastern Regional Research Center, ARS, USDA, 600 East Mermaid Lane, Wyndmoor, PA 19038, USA

(215-836-3756) djohnston@arserrc.gov

Fuel ethanol production using the corn dry grind process produces distillers dried

grains with solubles (DDGS) and carbon dioxide as the only coproducts. DDGS are used

primarily as food for ruminant animals and have limited use in nonruminants due to high fiber

content. The market price of DDGS has decreased significantly as the number of dry grind

production facilities has increased. The extraction of new coproducts from corn processing

has been a long term focus of research in our laboratory. Recently, we began investigating

the potential and feasibility of “coproduction” of value added products with simultaneous

production of ethanol. Several possible strategies have been investigated for making products

that can enhance existing coproducts or produce ones that are unique and would represent

completely new animal food ingredients for the industry. A number of microbially derived

products can be produced in series (or parallel) with ethanol production while utilizing many

common unit operations. This strategy could increase the number of coproducts produced by

a corn processing facility and improve overall economics of fuel ethanol production. An

example of the integrated production of a value added product will be presented and an

integrated process model for production will be shown.

143

144

STARCH COMPONENTS AND PROPERTIES IN TRITICALE AND OTHER CEREALS

John Lu*, Byron Lee, Brian Beres, Andre Laroche, Denis Gaudet and Francois Eudes

Lethbridge Research Centre, Agriculture and Agri-Food Canada,

Lethbridge, Alberta, Canada T1J 4B1 (403-317-3302) luj@agr.gc.ca

Triticale (Triticosecale) is a man made cereal species originating from a cross between

wheat (Triticum) and rye (Secale). The high yield and low input requirement for production

make this species a potential source of industrial starches for western Canada. The

components and properties of grain starches were evaluated from triticale germplasm and

cultivars (predominantly hexaploid) and other cereal crops (wheat, barley and rye).

Variations in total starch contents and different starch components (amylose, amylopectin and

resistant starch) were observed among the milled grain flours of different genotypic samples.

AC Ultima, a spring triticale cultivar, exhibited potential for higher starch and amylose

contents. Pure starch granules were isolated from a single seed of each genotypic sample and

the granule sizes and shapes were observed and imaged under the optical and scanning

electron microscope. Evaluating micrographs, we demonstrated that barley and rye have

large starch granules (A type); whereas, wheat and triticale have large and small starch

granules. Triticale has a larger number of small starch granules (B type) but A type granules

represent the majority of starch mass in triticale grains.

145

146

BIOFUELS PRODUCTION IN THE PACIFIC NORTHWEST: OPPORTUNITIES AND CHALLENGES

Ganti S. Murthy*

Oregon State University, Corvallis OR 97330 (541-737-6210) murthy@engr.orst.edu

Ethanol production from renewable resources is one of the strategies to achieve US

energy security. Most ethanol in the US is produced from corn. Greater replacement of fossil

fuels with ethanol is possible by producing ethanol from other starch and cellulose rich

feedstocks. The Pacific Northwest region of the US has diverse agricultural and climatic

regions that limit large scale farming of a single feedstock. Feedstock availability in the

Pacific Northwest for ethanol production will be assessed.

Potatoes, sweet potatoes, sweet beets, rye and other small grains are high starch

feedstocks that are grown in limited quantities in different areas. Limited production poses a

major constraint for construction of large ethanol plants using locally produced feedstocks.

However, production of valuable coproducts such as color pigments and soluble dietary fiber

can make biofuels production economically viable.

Cellulose rich materials such as forest thinnings, paper pulp, wheat straw, grass straw

and municipal solid waste are feedstocks that have potential for ethanol production.

Transportation costs, dispersed feedstock sources and price pressure from animal food

markets are determining factors for economical production of ethanol.

Dispersed sources of limited feedstock necessitate development of novel technologies

that can process mixed feedstock. Determining the effects of feedstock variability on ethanol

production process is a critical need. Small modular units that process either starch or

cellulose based feedstocks for ethanol will lead to effective utilization of limited and seasonal

resources.

147

148

IMPROVEMENTS IN CORN GLUTEN DEWATERING

David W. Scheimann*

Nalco Company, 1601 W. Diehl Road, Naperville, IL 60563

(630-305-2389) dscheimann@nalco.com

In corn wet milling, the primary goal of the process is efficient separation of four primary components of the corn kernel: germ, fiber, starch and gluten. The gluten, or protein, fraction typically accounts for 5.7 to 9.7% of the initial weight of the corn kernel. The gluten fraction is an important component used in manufacture of a number of different grades of animal food. Gluten dewatering and drying operations are estimated to account for 26% of energy used in the wet milling operation. This makes gluten dewatering and drying the second largest consumer of energy with starch dewatering and drying being the largest at 32% (Galitsky et al 2003). Gluten is dewatered using rotary vacuum filters and may be viewed as a bottleneck or a rate limiting step within the wet mill. Seasonal variations in the corn crop or process related issues within the plant also impact gluten dewaterability. Gluten is a valuable coproduct. A primary criterion in establishing the value of the coproduct is the protein content. Gluten that cannot be processed on the filters as corn gluten meal may end up being diverted and further processed in other unit operations within the plant. When this occurs, the high value coproduct becomes a lower value coproduct, corn gluten feed, and may result in significant revenue loss for the processing plant. We will focus on the gluten dewatering process and the potential for improving performance and throughput by adding a processing aid. Significant improvements in gluten dewaterability can be achieved with conditioning of the heavy gluten.

LITERATURE CITED Galitsky, C., Worrell, E. and Ruth, M. 2003. Energy efficiency improvement and cost saving

opportunities for the corn wet milling industry: An ENERGY STAR Guide for Energy and Plant Managers. July 1, 2003. Lawrence Berkeley National Laboratory. Paper LBNL-52307. http://repositories.cdlib.org/lbnl/LBNL-52307.

149

150

SORGHUM AS A VIABLE RENEWABLE RESOURCE FOR BIOFUELS AND BIOBASED PRODUCTS

Xiao Wu1, Renyong Zhao1, Scott R. Bean2, Paul A. Seib1, James S. McLaren3,

Ronald L. Madl1, Mitch Tuinstra1, Mike C. Lenz4 and Donghai Wang*1

1Kansas State University, Manhattan, KS 66506, 2Grain Marketing and Production Research Center, ARS, USDA, Manhattan, KS 66502, 3StrathKirn Inc, Chesterfield, MO 63017 and

4Monsanto, Mt. Hope, KS 67108 (785-532-2919) dwang@ksu.edu The goal of this research is to improve bioconversion efficiency of sorghum for biofuels and biobased products. The main focus is to understand the relationship among genetics-structure-function-conversion and key factors impacting bioprocessing to selected products (eg, ethanol), as well as to develop an energy life cycle analysis model to quantify and prioritize potential savings from identified factors. Ethanol fermentation results from 70 sorghum genotypes and elite hybrids were used to determine the impact of compositional, structural and physical factors; key interactions impacting ethanol yield from sorghum grain and have been identified. Genetic lines with high and low ethanol fermentation efficiencies and some specific attributes that may be manipulated to improve the bioconversion rate of grain sorghum were identified. In general, ethanol yield increased as starch content increased. However, there was no linear relationship between starch content and fermentation efficiency. Key factors affecting the ethanol fermentation efficiency of grain sorghum included protein digestibility, protein and starch interaction, mash viscosity, amount of phenolic compounds, ratio of amylose to amylopectin and formation of amylose-lipid complex in the mass. Ethanol fermentation efficiency increased as amylose to amylopectin ratio decreased. Waxy sorghum had relatively higher fermentation efficiency. Ethanol yield and fermentation efficiency increased as protein digestibility and extractable protein increased. Mash viscosity, level of phenolic compounds and formation of amylose-lipid complexes had negative effects on ethanol fermentation. Analysis of the chemical composition of distillers dry grains with solubles (DDGS) from sorghum lines with high and low ethanol fermentation efficiencies showed residual starch in DDGS contained mostly resistant starch associated with amylose and in the forms of lipid-amylose complex. Commercial sorghum ethanol producers were contacted to gain actual practical information and data on processing parameters and energy utilization throughout the commercial process. The energy life cycle analysis model for sorghum has been developed to determine potential savings for selected factors. Acknowledgement Supported in part by grant number 2004-35504-14808 by the USDANational Research Initiative (NRI) Biobased Products and Bioenergy Program. Authors would like to thank Novozymes, Inc. for providing enzymes.

151

152

ENZYME PRODUCTION BY INDUSTRIALLY RELEVANT FUNGI CULTURED ON COPRODUCTS FROM CORN DRY GRIND ETHANOL PLANTS

Eduardo A. Ximenes1,2,3, Bruce S. Dien1*, Michael R. Ladisch2, Nathan S. Mosier2,

Michael A. Cotta1 and Xin-Liang Li1

1National Center for Agricultural Utilization Research, ARS, USDA,

1815 N. University Street, Peoria, IL 61604 (309-681-6270) dienb@ncaur.usda.gov,

2Purdue University, LORRE, 500 Central Drive, West Lafayette, IN and 3Present Address: Microbiology, University of Georgia, Athens, GA

Distillers dried grains with solubles (DDGS) is the major coproduct produced at a dry

grind facility. Currently, it is sold primarily as a ruminant animal food. DDGS is low cost

and relatively high in protein and fiber contents. DDGS was investigated as a carbon source

for extracellular hydrolytic enzyme production. Two filamentous fungi, Trichoderma reesei

Rut C-30 and Aspergillus niger NRRL 2001, noted for high cellulolytic and hemicellulolytic

enzyme titers, were grown on DDGS. DDGS was used either as delivered from the plant

(untreated) or after being pretreated with hot water (HW). Both microorganisms secreted a

broad range of enzymes when grown on DDGS. Higher xylanase titers were obtained when

cultured on HW-DDGS compared to growth on untreated DDGS. Maximum xylanase titers

were produced in 4 days for A. niger and 8 days for T. reesei in shake flask cultures. Larger

amounts of enzymes were produced in bioreactors (5 L) either equipped with rushton (for T.

reesei) or updraft marine impellers (A. niger). Initial production titers were lower for

bioreactor than for flask cultures, especially for T. reesei. Improvements of enzyme titers

were obtained using fed batch feeding schemes.

153

AUTHOR INDEX

Page

Bailly, Mathieu Ameridia 73

Bar, Daniel H. Ameridia 73

Batton, Carol B. Nalco 138

Bean, Scott R. Kansas State University 150

Belyea, Ronald L. University of Missouri 42, 140

Beres, Brian Agriculture Canada 144

Boyd, Gale A. Duke University 123

Coil, Gregory Mortenson 117

Cook, David A. Cargill 112

Cotta, Michael A. NCAUR/ARS/USDA 152

Coward-Kelly, Guillermo Novozymes 68

Dahlberg, Jeff Sorghum Producers 50

DeKam, Matthew University of Minnesota 76

Dien, Bruce S. NCAUR/ARS/USDA 87, 152

Enterline, William R. Andritz Sprout 18

Eudes, Francois Agriculture Canada 144

Fink, Rodney J. Western Illinois University 56

Fischer, James R. REE/USDA 2

Gaudet, Denis Agriculture Canada 144

Grewell, David A. Iowa State University 102

Hicks, Kevin B. ERRC/ARS/USDA 34

Horner, Joe L. University of Missouri 140

Johnston, David B. ERRC/ARS/USDA 34, 142

Jorgensen, Per Lina Novozymes 86

154

Page

Konieczny-Janda, Gerhard Genencor 34

Koros, William J. Georgia Institute of Technology 132

Khanal, Samir K. Iowa State University 102

Ladisch, Michael R. Purdue University 152

Laroche, Andre Agriculture Canada 144

Lee, Byron Agriculture Canada 144

Lenz, Mike C. Monsanto 150

Li, Mian Genencor 34

Li, Xin-Liang NCAUR/ARS/USDA 152

Lu, John Agriculture Canada 144

Lutin, Florence Ameridia 73

Madl, Ronald L. Kansas State University 150

McAloon, Andrew J. ERRC/ARS/USDA 30

McLaren, James S. StrathKirn 150

Merediz, Tristan Westfalia 96

Morey, R. Vance University of Minnesota 76

Mosier, Nathan S. Purdue University 152

Murthy, Ganti S. Oregon State University 42, 146

Nghiem, John ERRC/ARS/USDA 34

Pederson, Sven Novozymes 86

Rausch, Kent D. University of Illinois 42, 140

Robertson, John Alfa Laval 27

Scheimann, David W. Nalco 148

Seib, Paul A. Kansas State University 150

155

Page

Shetty, Jay K. Genencor 34

Soong, Chee-Leong Novozymes 68

Svonja, George Barr Rosin 21

Teunissen, Pauline Genencor 34

Tiffany, Douglas G. University of Minnesota 76

Tiller, Kelly J. University of Tennessee 13

Tuinstra, Mitch Kansas State University 150

Tumbleson, M. E. University of Illinois 42, 140

Tunnessen, Walt US EPA 123

Van Leeuwen, J. (Hans) Iowa State University 102

Vikso-Nielsen, Anders Novozymes 86

Wang, Donghai Kansas State University 150

Wang, Michael D. Argonne 69

Wenger, Kevin Novozymes 68

Wu, May M. Argonne 69

Wu, Xiao Kansas State University 150

Ximenes, Eduardo A. NCAUR/ARS/USDA 152

Zhao, Renyong Kansas State University 150