Unlike the federal government, several states have altered their biom­ass incentives to enhance the positive impact on rural communities. The Minnesota experience, often called the Minnesota Model, is instructive.

In the early 1980s, Minnesota’s state ethanol incentive mirrored that of the federal incentive—a partial exemption from the gasoline tax. That incentive succeeded in making the price of ethanol competitive with other gasoline additives. The demand for ethanol-blended gasoline soared. But the demand was met entirely by ethanol imported into the state from out — of-state, large manufacturing facilities owned by one multinational cor­poration. Minnesota farmers and Minnesota’s farming communities did not benefit from the expanded consumption of ethanol inside the state.

To remedy this problem, in the mid-1980s, Minnesota converted its state ethanol incentive from a consumer-oriented excise tax exemption to a producer-oriented direct payment. Instead of reducing state gasoline taxes by a couple of pennies for a 10% ethanol blend, the state paid 20^/gal for ethanol produced within the state. To encourage the construction of many plants in different parts of the state, the incentive, which ran for 10 yr, applied only to the first 15 million gal produced.

Some argued that by encouraging many small biorefineries, the gov­ernment was encouraging higher-cost and more inefficient biorefineries because of the engineering economies of scale involved. Indeed, an internal study by the Institute for Local Self-Reliance concluded that a 150 million gal/yr ethanol facility had unit costs about 15-20^/gal less than those of a 15 gal/yr facility.

Thus, the 20^ incentive made up for the difference between small and large biorefineries. The result was that rather than one or two 100 million gal/yr plants, by 2002 Minnesota was home to 15 ethanol plants, the aver­age capacity of which was 15 million gal/yr. The scale of the plants also encouraged farmer ownership. In 2002, 12 of the 15 plants were owned by more than 9000 grain farmers. These plants provided almost 10% of the transportation fuel sold in the state.

The proliferation of small plants led to an unanticipated technological dynamic. Because of the large number of plants built, several engineering firms competed with each other to design and build the least expensive and most efficient facility. Yields of ethanol in dry mills quickly rose from 2.5 to >2.8 gal/bushel. The large number of plants, coupled with equal num­bers of plants being built in surrounding states, accelerated the engineering and operational learning curves. The result was to rapidly reduce the cost of ethanol produced from small dry mills.

A 1998 study by the USDA, a follow-up to a 1987 ethanol plant survey, examined the comparative economics of small — and medium-sized dry mills and large wet mills. In 1987, small — and mid-sized dry mills had cash oper­ating costs of 50^/gal. Large wet mills had a cash operating cost of 47^/gal. By 1998, dry mills had dropped their operating costs to 41.7tf, whereas wet mills had dropped theirs only marginally, from 47.2 to 46 c The 1998 report concluded, "Wet mill variable costs appear to have remained very stable at about 46 cents per gallon. Improved energy cost management was offset by several factors, including waste management and overhead… In contrast, dry mills have experienced a l5-percent reduction in operating costs, due to the effects of reduced energy, labor and maintenance expenditures and possibly economy of scale" (3).

The experimental data presented herein are the result of exploratory research aimed at bracketing the necessary moisture and inoculum loads for effective pilot-scale distributed upgrading of wheat straw stems for produc­tion of straw-thermoplastic composites (4,15). An exploratory approach was chosen for these tests because full-scale outdoor systems having few envi­ronmental controls would be difficult if not impossible to closely control. Both temperature and moisture levels vary owing to variations in heat,

humidity, and wind. Precise real-time control of temperature and moisture in such a system would be very expensive, and thus counter to the goal of using distributed systems. We therefore decided to choose conditions of moisture and inoculum that provided average fungal degradation of the stems and rapid competitiveness of the inoculated fungus at a single tem­perature. We then conducted statistical analyses of the data in order to predict the expected variability of final composition in response to varia­tions in initial moisture and inoculum.

Philippe Savoie,[3]1 Kevin J. Shinners,2
and Benjamin N. Binversie2

1Agribulture and Agri-Food Canada, 2560 Hochelaga Boulevard,
Sainte-Foy, Qubbec, Canada G1V 2J3, E-mail: psavoie@grr. ulaval. ca; and
2Biological Systems Engineering Department,

University of Wisconsin, 460 Henry Mall, Madison, WI 53706

Abstract

Mixing fresh silage in water resulted in partial segregation of grain from stover. Grain concentration was 75% in the sunk material when silage was relatively dry (64% moisture content [MC]) and 41% when silage was rela­tively wet (74% MC). Partial drying to remove 20 percentage units of mois­ture prior to water separation increased grain concentration to 92%, and complete drying increased grain concentration to 99%. Sieving without dry­ing followed by water separation resulted in a grain concentration of 79%. A byproduct of water separation is a large amount of soluble and deposited fine particles in the effluent: 18% of original dry matter after one separation, and between 21 and 26% after eight separations. In an industrial setting, hydrodynamic separation of silage with minimal pretreatment could pro­vide a feedstock with a high concentration of grain (75-80%). In a laboratory setting, hydrodynamic separation with prior oven drying could provide a method to separate grain from stover in corn silage by reaching a grain con­centration higher than 99%.

Index Entries: Corn; stover; grain; separation; silage.

Introduction

Various methods for separating corn grain from stover have been proposed. One approach is to shell the grain with a combine and subse­quently to harvest the residual stover with either a forage harvester or a baler (1). Another approach is to harvest, chop, and ensile the whole crop, and to separate components at removal from storage (2). Advantages of separation after storage include fast and efficient harvest in a single stream;

low-cost storage in high-capacity bunker silos without the need for grain drying; and separation throughout the year at relatively low work rates, with small size equipment, compared with the high rates handled during the short harvest season.

After ensiling, grain has been sorted from stover, at least partially, by mechanical sieving (3) or aerodynamic separation (4). Hydrodynamic sepa­ration has not previously been used for corn silage, but it is used industri­ally to separate heavier particles such as phosphatides from corn oil in the wet-milling process (5).

Hydrodynamic separation of grain from stover could be feasible if components exhibit significant differences in specific gravity or buoyancy. The specific gravity of corn grain has been observed to range from 1278 to 1380 kg/m3 (6). The specific gravity of corn stover components (stalk, cob, leaf, husk) is less well documented. Meanwhile, the specific gravity of for­age particles is in the order of 1500 kg/m3 (7). This would suggest poor hydrodynamic separation of grain from stover in water because both com­ponents would sink. However, empirical evidence shows that grain sinks more rapidly than stover, which tends to float because of buoyancy.

The objective of the present work was to evaluate the potential of hydrodynamic separation with water to sort grain from stover after ensiling. New data are presented on the specific gravity of corn grain and stover components after coarse chopping or grinding. Factors considered include harvest conditions (chop length and processing) and pretreatment of the silage (partial drying, sieving) prior to hydrodynamic separation.

In 2001, there were 88,000 acres of table grapes in California and the average production was 8.07 t/acre (1). The harvest season is from May through November, and table grapes can be sold as fresh fruit or for use in making juice, concentrate, or wine. Between 1991 and 2000, an average of 680,000 t of table grapes reached the fresh fruit market, each year. The average proportion of culled fruit in the fresh market is 25.5% (4). There­fore, we assume that 173,400 t of culled grapes would be available for use in ethanol production.

The average price for culled grapes diverted to other uses such as in the juice and concentrate markets during 1991 to 2000 was $161.70/t. We

use that price in our analysis because it represents the value of the next-best alternative to sale in the fresh fruit market. Table grapes produced in the San Joaquin Valley and sold in California must contain a minimum sugar content of 17% (5). That content implies an ethanol yield of 23.8 gal/t of grapes. This is determined by multiplying the sugar content by 100 and multiplying that result by a conversion coefficient of 1.4 gal of ethanol/t of grapes per unit of sugar (R. Murray, personal communication, 4/4/02). The estimated feedstock cost of culled table grapes becomes $6.79/gal of ethanol.

Raisins

The average annual production of raisins in California during 1991 through 2000 was 368,000 t. The Raisin Advisory Committee allocates a portion of the total production to "free tonnage" and the remainder to a "reserve pool." The reserve pool represents the potential feedstock source for ethanol production. Between 1979 and 2001, the average allocation to the reserve pool was 28% of total production. Using that proportion and the average production from 1991 through 2000 generates an expected annual reserve pool of 103,000 t of raisins. Although raisins are harvested from August through October, they can be stored for use as a feedstock in any month.

At present, there is an oversupply of raisins in California. The reserve pool for the 2000-2001 season was 203,330 t, or about 66% of total produc­tion (6). Changing dynamics in international trade and market develop­ments in the wine industry will continue to have an impact on the size of the reserve pool.

Between 1993 and 2001, growers received an average of $329/t for raisins in the reserve pool. That price fell to $250/t in 2001 (6). We use the average annual price of $329/t in our analysis, while noting that raisins will be available at a lower price in the future, if the condition of excess supply continues. Hence, we consider also an alternative raisin price of $250/t. About 98 gallons of ethanol can be produced/t of raisins. Using the average reserve pool price of $32/t, the feedstock cost would be $3.36/gal of etha­nol. The feedstock cost is $2.55/gal when the 2001 reserve pool price of $250/t of raisins is considered.

Evening Primrose is a biannual wildflower named for its habit of opening yellow blossoms at dusk. The four-petaled flowers are strongly scented with a sweet perfume that attracts pollinating moths. Evening Primrose is easily cultivated; it prefers acid, neutral, or alkaline, well — drained soils and requires full sun. It does best when not having to compete for soil or space.

Because the Evening Primrose is a biannual, the stem shoots up only in the second year of growth. In the first year, a taproot is put down. In the second year a sprout shoots up. Producing a biomass material only every other year is a definite drawback to its use as a biofuel; however, the density of its xylem material is significantly greater than that of the other candidate plants, actually giving the evening primrose a competitive output.

Evening primrose is grown commercially for its seed oil (treats asthma, arthritis, headaches). Consequently, there is a great bed of knowledge con­cerning its seed production, cultivation, and even some insights into strat­egies for making it provide an annual shoot.

Stem construction consists of a thin, barklike epidermis. The xylem cylinder has an XPR of 1.3 and a typical stalk diameter of 1.5 cm. Xylem material has a density of about 536 kg/m3, which makes it almost compa­rable to hardwood. The pith is a solid spongy inner core that rots after about 2 mo of drydown.

Evening primrose has a dense "stem," probably indicating a high per­centage of lignin, and a correspondingly higher heating value than the other SSPs. Being a biannual is a mixed bag. The harvest is less frequent, which is an advantage, but the population density of harvestable plants is only half of what an annual or perennial would be.

We have described the economic feasibility of utilizing surplus and cull citrus, grapes, raisins, tree fruit, California-grown corn, and Midwest­ern corn as fuel ethanol feedstocks. The cost of production exceeded the price of ethanol for all of the feedstocks we considered. However, in the case of raisins, oranges, and grapes, surplus production could generate additional storage costs. Those costs provide an incentive for considering the sale of surplus or culled production to an ethanol producer at a price less than the market price of the surplus or culled product. For example, in the case of raisins, a fuel ethanol industry might provide higher net rev­enues to farmers who could sell raisins from the reserve pool for ethanol production, rather than storing them until they are released for sale in the fresh market or discarded, as required by the raisin marketing order.

We estimated that the ethanol producer would be willing to pay $66/t for raisins. That price will increase or decrease with changes in ethanol prices. At a price of $66/t, raisin growers would not choose to grow raisins for the ethanol market. However, raisin growers could benefit from having the ethanol market as an alternative diversion when reserve pool prices are low, or when the amount of raisins held in reserve is large. We showed that raisin producers in the San Joaquin Valley would have gained net revenue by selling raisins from storage for ethanol production in 6 of the 10 yr during 1992 and 2001.

One implication of our work is that an ethanol industry in the San Joaquin Valley might provide an economic benefit to raisin growers and producers of other agricultural products that need to be stored before they are sold. A second implication is that a broader range of potential market outlets may reduce uncertainty regarding net revenue, by alleviating some of the downside exposure to the high cost of maintaining reserve stocks.

Future work in this area might include extending our analysis to the grape concentrate and orange juice markets. That research would enhance our understanding of the potential economic impacts of developing an ethanol industry in the San Joaquin Valley. Given that the net returns from the production of ethanol are negative for all of the locally grown feed­stocks we considered, private firms will not choose to produce ethanol in California without a public subsidy. The amount of public funds required to support ethanol production can be reduced if alternative marketing opportunities are identified that allow producers to obtain feedstock mate­rials from farmers at prices that are lower than those in primary markets for surplus and culled products.

Acknowledgments

We appreciate support from the California Department of Food and Agriculture; the California Agricultural Technology Institute; and the Cen­ter for Agricultural Business at the California State University, Fresno. We appreciate also the helpful comments from two anonymous reviewers, and we retain responsibility for any oversights and omissions. This article is Contribution Number TP 03-10 of the California Water Institute.

References

1. California Department of Food and Agriculture. (2002), Agricultural Resource Direc­tory, Sacramento, CA.

2. California Department of Food and Agriculture. (1980), Ethanol Yields by Crop, Sacra­mento, CA. Unpublished.

3. Urbanchuk, J. M. and Kapell, J. (2002), Ethanol and the Local Community, Renewable Fuels Association, Washington DC.

4. California Department of Food and Agriculture. (2001), Agricultural Resource Direc­tory, Sacramento, CA.

5. California Department of Food and Agriculture. (2002), Synopsis of Standardization Provision and Procedures, Sacramento, CA.

6. Raisin Administrative Committee. (2002), Analysis Report, Fresno, CA.

7. United States Department of Agriculture (USDA). (2002), Market News Reports, Mar­ket News Service, AMS, Washington DC.

8. California Energy Commission. (2001), Costs and Benefits of a Biomass-to-Ethanol Pro­duction Industry in California, Sacramento, CA.

9. Yancey, M. (2003), Paper presented at the California Ethanol Workshop, April 14, Sacramento, CA (www. bbiothanol. com/doe/ca/Yancey_CA_DOE. pdf).

10. California Energy Commission. (1999), Evaluation of Bio-Mass to Ethanol Fuel Potential in California: Report to the Governor, Sacramento, CA.

11. Pacific Gas and Electric. (2003), Advice Letter No. 2471-G, issued by Karen A. Tomcala, Vice President, Regulatory Relations, San Francisco, CA.

12. Northwestern Corporation. (2002), Northwestern Energy Signs Gas Supply Contract to Serve New South Dakota Ethanol Plant, News Center News Display, Sioux Falls, SD.

13. Whims, J. (2002), Corn Based Ethanol Costs and Margins: Attachment 1. Agricultural Marketing Resource Center, Department of Agricultural Economics, Kansas State University, KA (www. agmrc. org/corn/info/ksuethl. pdf).

14. Shapouri, H., Gallagher, P., and Graboski, M. S. (2002), Agricultural Economic Report No. 808, US Department of Agriculture, Office of the Chief Economist.

15. Gabler, E., Kalter, R. J., Boisvert, R. N., Walker, L. P., Pellerin, R. A., Rao, A. M., Hang, Y. D. (1981), Cornell agricultural economics staff paper, no. 81-27.

16. Oxy Fuel News. (2002), Oxy Fuel News Price Report, 1995-2002, Hart Publications, Potomac, MD.

Can federal research efforts focus on technologies applicable for smaller facilities? What would such research look like? This is a key chal­lenge for researchers and policy makers. Engineering economies of scale do exist, as do management and marketing economies of scale. However, there are technologies that lend themselves more to modular expansion. Tech­nologies should as much as possible allow the farmer to capture the value added from storing, preprocessing, and perhaps even processing the crop on site.

In the late 1970s, the DOE launched its wind energy initiative. Wind energy has dramatic economies of scale. The power output varies by the square of the diameter of the turbine’s blades and by the cube of the increase in the wind speed. The DOE focused on building very large — diameter wind turbines. These megaturbines contributed relatively little to the technological advances in the wind energy field. Much more impor­tant for wind energy development was the design of buyback tariff struc­tures in California in the early 1980s and the wind energy mandate in Minnesota in the mid-1990s.

As the wind energy industry grew, advances in the electronics and construction design of wind turbines grew even more rapidly. As they did, wind turbines became larger in a more organic way, moving from the 200-kW machines of the early 1980s to the 750-kW machines of the late 1990s to the new 1.5-MW machines.

In 2001, the DOE launched a small wind turbine initiative. The objec­tive was to make wind energy economical in the many areas of the country that have lower wind speeds. This initiative is a 180° turn from the wind energy program of the late 1970s. It is too early to evaluate its results.

Regarding biomass, the DOE favors larger facilities, again because of their engineering economies of scale. The biomass program’s orientation to the scale of production systems mirrors that of the fossil fuel and nuclear programs. Yet biomass has characteristics that may lend itself to a different orientation. The cost of transporting biomass, e. g., is much higher than the cost of transporting fossil fuels or uranium. In addition, the farmer-ori­ented and rural economy objectives of the biomass program are not part of the fossil fuel or nuclear development program.

Both DOE and the Department of Agriculture are conducting research in biochemical production. How would a focus on smaller scale and modu­lar production units affect the research done under these programs?

The DOE and, to a lesser extent, the Department of Agriculture, have largely ignored questions of scale and ownership in their R&D efforts. The result could be that success in dramatically increasing the use of biomass for energy and industrial purposes may well not translate into higher farmer income or healthier rural communities. Yet these are formal objectives of both agencies’ missions. Taking these socioeconomic factors into account could well encourage a different R&D and commercialization strategy by the federal agencies in charge of the biomass program.

Notes

1. Catherine E. Woteki, Deputy Undersecretary, Research, Education and Economics, USDA, Testimony Before the House Agriculture Committee, Resource Conservation, Research and Forestry Subcom­mittee, May 14, 1996.

2. Achievements in Agricultural Utilization Research, ARS Committee on Research Achievements, ARS, Washington, DC, November 1966.

The same 25-yr period, 1941-1966 was examined in a PhD thesis at the University of Georgia by Harold B. Jones, Jr., of the USDA’s Economic Research Service. He found that by 1966, 9% of the projects undertaken by the regional laboratories had produced an economic return. That figure compared favorably with returns on food indus­try research. The cost-benefit ratio was 20 to 1 or better (cited in Always Something New: A Cavalcade of Scientific Discovery, USDA, Agricultural Research Service, Miscellaneous Publication 1507. November 1993).

3. Fred C. White, B. R. Eddleman, Joseph Purcell, et al, "Nature and Flow of Benefits from Ag-Food Research," in An Assessment of the United States Food and Agricultural Research System. Volume 2. Com­missioned Papers, IR-6 Information Report No. 5. Office of Technol­ogy Assessment, Washington, DC, December 1980, pp. vii-x.

4. Speech by Senator John F. Kennedy at the National Plowing Contest, Sioux Falls, SD, September 22, 1960.

5. David M. Russo and Edward McLaughlin, Farmers Can Get Bigger Share of Food Dollar. New York State College of Agriculture and Life Sciences, Cornell University, April 1991, estimates the retail price of corn flakes at $1.56 per 18-oz box. Corn represents about 10£ of the cost. The Economic Research Service, in Food Marketing and Price Spreads: Farm-to-Retail Price Spreads for Individual Food Items (2003) Washington, DC, estimated the year 2000 cost of an 18-oz box of corn flakes to be $2.14, with corn flakes representing 8£ of that cost.

6. Biomass RoadMap. Biomass Research and Development Technical Advisory Committee, Washington, DC, November 2002.

7. Personal communication from several ethanol plant managers in Minnesota. Also see Van Dyne, D. L., Employment and Economic Ben­efits of Ethanol Production in Missouri. Department of Agricultural Economics, University of Missouri-Columbia, February 2002.

Tests were designed to determine moisture levels and inoculum amounts that maximized hemicellulose degradation. In preliminary test­ing (4) it was shown that 10.9 mg of P. ostreatus/g of dry stems was suf­ficient to allow successful competition of the inoculated fungus with indigenous microbes. However, 10.9 mg of P. ostreatus/g of dry stems and a gravimetric moisture content of 0.77 g of H2O/g of straw were too low to effect significant degradation of the straw in 12 wk (15). In those tests a proxy variable, the ratio of cellulose and hemicellulose compositions (C/H), was used to indicate the relative change in composition occurring from indigenous microbes and by P. ostreatus. This ratio was used because P. ostreatus has been shown to be somewhat selective for hemicellulose and lignin degradation vs cellulose degradation (7,8), while the indig­enous microbes were shown in uninoculated controls to be nonselective for one polysaccharide or the other (4).

Owing in part to the low and thus more uncertain measurements of the nonxylan carbohydrate components of hemicellulose (galactan, arabinan, and mannan), we decided to change the proxy variable to one exhibiting less variability as the result of measurement uncertainty. Two additional proxy ratios were assessed: the ratio of glucan and xylan com­positions (G/X) and an estimate of the relative degradation of xylan vs glucan (AX/AG). This "degradation ratio" was calculated from the esti­mated conversions, which assumed little change in the sum of ash, lignin, and extractives. Implicit in this is an assumption of minimal mineralization of lignin to CO2, and thus losses of lignin are assumed to be from depoly­merization to extractives; extractives would increase, keeping the sum of lignin and extractives constant. This allowed a closed mass balance to be estimated and the amounts of xylan and glucan degraded to be calculated on an initial weight basis.

A comparison of these ratios for the preliminary testing at 0-10.9 mg of P. ostreatus/g of stems and 0.77 g of H2O/g of stems is shown in Table 3. The majority of the glucan/xylan-based ratios had lower standard deviations than the corresponding cellulose/hemicellulose-based ratios, which was expected because xylan represents a greater fraction of the straw stems than galactan, arabinan, and mannan (Table 1). The relative changes in the three proxy variables were consistent among the data. For example, when C/H and G/X did not change significantly, indicating nonselective degradation of glucan and xylan, the estimated degradation ratio AX/AG was about 1. Similarly, when C/H and G/X exhibited only a small increase, AX/AG was only slightly larger than 1, while larger increases in C/H and G/X corre­sponded with large increases in AX/AG. This may provide some support for the assumption used to estimate AX/AG. The proxy ratio chosen for use was the degradation ratio AX/AG; with this change, the con-clusions of the pre­liminary study were unchanged, that is, P. ostreatus was shown to out- compete the indigenous organisms by 56 d. In addition, from 0 to 22 d the AX/AG ratios did not change significantly from 1.0, suggesting that under

the conditions tested, in the initial 22 d of culture P. ostreatus did not domi­nate degradation of the stems.

Experiment 1: Specific Gravity of Corn Components

Several stalks of whole-plant corn were cut with a scythe at 10 cm from the ground at full maturity (early December 2002) near Madison, WI. Plants were separated manually into five components: grain, stalk, leaf, husk, and cob. The components were oven-dried at 103°C for 24 h (8) to estimate moisture content and the respective proportions of dry matter (DM). Specific gravity was estimated with a PMI Automated Gas Pyc­nometer (Porous Materials, Ithaca, NY), which measured the pressure change in helium gas as it surrounded the crop component in an enclosed volume (9). Three iterations were done for each sample in the pycnom­eter, and three replications were done for each component in two states: intact grain or coarsely chopped stover, and ground components. Chop­ping was done with a laboratory chopper set at 13-mm theoretical length of cut. Grinding was carried out with a Model 4 Thomas Wiley Mill using a 1-mm screen (Thomas Scientific). The specific gravity was corrected to a dry basis by mass balance:

pwm ph2o (l00- MC) 100 p h2o- p wm mc

in which pDM is the corrected specific gravity on a dry matter basis, pWM is the wet matter specific gravity as measured experimentally, pH2O is the specific gravity of water (1000 kg/m3), and MC is the moisture content on a wet basis (%).

We consider deciduous tree fruit production including peaches, plums, and nectarines. Between 1991 and 2000, California produced 664,000 t of tree fruit annually. The average annual cull rate is 25%, pro­viding 166,000 t of potential ethanol feedstock. This feedstock would be available seasonally from May through October.

The prices of culled tree fruit depend on the marketing options avail­able. Industry surveys completed in the spring of 2002 indicate that culled fruit prices range from $15 to $20/t. Hence, we use a price of $17/t in our analysis. Ethanol yields from culled fruit vary with the fruit selected. Nec­tarines have the highest yield 13 gal/t; peaches yield 12 gal/t; and plums generate 11 gal/t. We use the average of these estimates (12 gal/t) in our analysis. Hence, the estimated feedstock cost for culled fruit is $1.42/gal of ethanol.