Table 5 reports the chemical composition of five components from corn. The level of starch in grain was not as high as expected (50-60% is typical), perhaps because of the crop’s late maturity. Fiber concentrations, either ADF or NDF, are good indicators to estimate the proportion of grain and stover in a whole-plant mix.

Table 6 shows the chemical composition of components after water separation. The sunk grain had a level of starch as expected but more fiber than expected. The sunk stover probably contained small fractions of grain that could not be separated manually. The suspended and floating materi­als also probably contained some small grain particles that had not sunk after the first water separation. The effluent DM containing soluble and very fine particles had on average 26% starch and 3% ADF, indicating a larger proportion of grain than stover components. Prior to water separa­tion, silages in experiment 2 had an average composition of 8.5% CP, 19.1% starch, 23.6% ADF, and 40.3% NDF.

Table 5 Chemical Composition on DM Basis of Mature Whole-Plant Corn Components* Component CP (%) P (%) Ca (%) K (%) Mg (%) Starch (%) ADF (%) NDF (%) Grain 4.9 b 0.25 a 0.03 d 0.31 d 0.11 d 42.5a 2.5 d 21.0 d Stalk 3.3 c 0.10 bc 0.16 b 1.04 a 0.17 b 0.4b 45.5 b 78.3 c Husk 3.9 c 0.07 cd 0.13 c 0.74 b 0.16 c 1.3b 42.2 c 84.4 b Cob 2.8 d 0.05 d 0.05 d 0.60 c 0.08 e 0.1b 44.5 b 90.6 a Leaf 5.7 a 0.13 b 0.44 a 0.23 d 0.25 a 1.3b 46.8 a 77.8 c SEM 0.3 0.02 0.01 0.05 0.01 2.3 0.8 2.7 LSD 0.7 0.03 0.02 0.12 0.01 5.2 1.8 6.1 * Average of three replications. Values with the same superscript letter in a given column indicate no significant difference (p < 0.05). SEM, standard error of means; LSD, least significant difference.

Table 6 Chemical Composition on DM Basis of Corn Silage Components After One or Eight Water Separations* Component CP (%) P (%) Ca (%) K (%) Mg (%) Starch (%) ADF (%) NDF (%) After one water separation (experiment 3) Sunk grain 1.9 d 0.06 a 0.02 a 0.15 a 0.03 d 55.0 a 5.3 c 16.9 a Sunk stover 4.0 c 0.10 b 0.07 b 0.22 b 0.07 c 16.7 b 29.7 b 50.9 b Suspended stover 5.1 b 0.11 b 0.18 a 0.23 b 0.09 b 6.8 d 39.0 a 63.4 a Floating material 6.0 a 0.16 a 0.18 a 0.38 a 0.14 a 9.5 c 36.5 a 61.5 a SEM 0.1 0.01 0.01 0.01 0.00 1.2 1.6 2.2 LSD 0.3 0.02 0.02 0.03 0.01 2.7 3.6 5.1 After eighth water separation (experiment 2) Effluent DM 15.5 0.64 0.95 3.13 0.73 25.9 3.2 5.0 ‘Average of three replications after one separation, average of 12 replications after eighth separation. Values with the same superscript letter in a given column indicate no significant difference (p < 0.05). SEM, standard error of means; LSD, least significant difference.

We assume that the ethanol production facility will be located to minimize the cost of transporting feedstock materials. Studies suggest that distances greater than 40-50 miles become unprofitable for utilizing agricultural waste as biomass feedstock (15). In our analysis, we assume that the cost of transportation to the ethanol facility is zero, because our goal is to compare production costs for alternative feedstock materials. The assumption regarding zero transportation costs may be realistic in cases in which small ethanol production facilities are located near a large source of a byproduct, such as an almond-processing plant. In our analy­sis, the price of each feedstock is based on the postharvest price at the initial point of sale, such as a packinghouse or a processing facility.

Lower transportation costs could ultimately make California-pro­duced ethanol competitive with imported supplies. Ethanol imported from the Midwest is splash-blended at fuel distribution centers. Ethanol plants using California feedstock materials might be located near the distribution centers, to minimize the cost of transporting ethanol before it is blended with gasoline.

Jim Hettenhaus1 and David Morris2

1cea Inc, Charlotte, NC; and
2Institute for Local Self Reliance, Minneapolis, MN

For large, economic, and sustainable harvest of biomass feedstock, major changes in cropping practice, collection, storage, and transportation are required. The challenges faced in supplying 0.7-1 million dry short tons (dt) for a single biorefinery are huge—five times larger than previous attempts.

Ultimately, the farmer controls biomass sourcing for biorefineries. The availability of large quantities of residues, stover, and straw is greatly depen­dent on tillage practice. No tillage results in most of the residue available for removal, especially when cover crops are employed for erosion control. By contrast, no excess is available with conventional tillage. Since <20% of crop­land is no till and >60% is conventional till, a major shift in cropping practice is needed for sustainable removal of significant quantities.

Present collection costs are 1.5-2 times the delivered cost target—$35/ dt, including $20/acre or more net income for the farmer. Bulk collection is likely needed, because baling adds cost, $15/dt, and no value. One-pass harvest can lower the delivered feedstock cost to <$20/dt within a 15- to 20-mi radius. Prototypes for one-pass harvest of straw and stover are under development, adapting existing equipment. Many variations are possible, but until a better market definition is available, a new design is probably limited to paper studies.

One-pass harvest also reduces the risk of corn stover harvest if storage of wet harvested material is resolved. For the sugar platform, feedstock can be wet, above 65%, or dry, below 20%. Some are looking at adapting wet, bagasse-type storage, large 250,000-dt piles built via circulating liquor that conveys the feedstock from wagons or trailers directly from the field after it is washed and milled to a particle size that ensures good compaction and preservation in storage. Less area is required; fire is eliminated when stored above 65% moisture. The material processes easier because 80% of the solubles is removed in storage. Water management and other issues remain. Validation of this method is required for other crops such as stover and straw.

Because of the bulky nature, the cost of transportation is 20-40% of the cost within a 50-mi radius. Transportation from the field to a storage site following harvest needs to be kept short if truck requirements are to remain manageable. Collection within a 50-mi radius for one site requires about five times the trucks and wagons compared with a 15-mi radius. While bulk density can be increased, the cost of densification generally offsets any transport savings. Pipelines require huge initial investment. Short-line rail delivery from three or more collection sites to supply the plant appears most advantageous compared to trucking or pipelines.

In conclusion, potential processors want clean liquid, mostly ferment­able sugars, delivered to the processing plant. Thus, in addition to the above, using part of the storage time for value-added treatment offers more potential. Although this processing is probably under different regulatory requirements than storage, it may be segregated and controlled separately. Including preprocessing with harvesting, collection, and storage provides farmers ample opportunity to participate in the value chain, moving away from simply supplying a commodity.

Copyright © 2004 by Humana Press Inc.

All rights of any nature whatsoever reserved. 0273-2289/04/113/0005-0012/$25.00

Unsterile straw stems were used in all experiments because steriliza­tion of large quantities of straw for construction of large windrows would be uneconomical and impractical. For this strategy to be effective, it was necessary that the inoculated fungus be able to compete with the indig­enous organisms in the straw. Since white-rot fungi dominate in nature under conditions of nitrogen deprivation (9), experiments were previ­ously performed (4) to determine inoculum production conditions neces­sary to limit nitrogen addition during inoculation of the straw with the fungus. C/N ratios in the media tested ranged from about 29 to 89 and were adjusted by adding yeast extract to 20 g/L glucose solutions. The nitrogen-limited medium finally utilized for straw stem inoculum pro­duction contained 3.0 g/L of yeast extract, for a C/N of 32.6. Mycelial inocula for inoculation of straw stems were produced in this nitrogen — limited medium, using as inoculum the fungal pellets produced at Utah State University in YM broth (4).

Approximately 500 mL of wet fungal pellets of P. ostreatus grown at Utah State University in YM broth were transferred to a sterile blender and blended for 2 min, producing a slurry of finely chopped mycelial fragments. The optical density (OD) at 550 nm was determined for dilu­tions of this slurry using a standard UV/VIS spectrophotometer. The undiluted slurry was then inoculated to 1.0 OD into the fresh nitrogen — limited medium in sterile shake flasks and incubated for 5-7 d at 30°C, 135 rpm. The fungal pellets in the inoculum cultures were then transferred with the spent medium to a sterile blender and blended for 2 min. The OD at 550 nm was determined for dilutions of this slurry, and the concentra­tion of biomass was estimated from a previously measured calibration. The undiluted slurry was transferred to a sterile hand-pump garden sprayer for addition to the straw stems. No extraordinary measures were taken beyond this point to maintain sterility, except the use of initially sterile equipment.

Air-dried straw stems (150 g dry wt at about 9-13 wt% moisture) were weighed onto a clean, dry, tared tray and spread in a thin (5-cm) layer. The homogenized mycelial inoculum slurry was then sprayed onto the stems, with frequent mixing of both the inoculum and the stems. Sufficient inocu­lum was added to reach the desired initial level of fungal inoculum in the stems. Periodically during addition of inoculum, a fan was used to blow nonsterile air across the tray of inoculated straw to evaporate excess water,

with frequent mixing of the straw. After the desired amount of inoculum was added, additional sterile distilled water was sprayed onto the straw as needed to reach the desired initial moisture content for the particular experiment. A separate sample of the well-mixed inoculum slurry was then added to a tared bottle and dried to constant weight at 105°C to determine the actual biomass concentration of the slurry. In addition, sev­eral small samples of the inoculated stems were transferred to tared bottles and dried to constant weight at 105°C to determine the actual moisture content of the inoculated stems.

Combinations of inoculum amounts and moisture contents tested in this exploratory study are shown in Fig. 1; controls lacking inoculum were also conducted at 0.4-0.77 g of H2O/g of stems and are not shown in Fig. 1. The first tests performed (4) are represented by the 12 points in the lower left-hand corner of Fig. 1. When these tests indicated that higher inoculum was needed for better selectivity and that higher moisture was needed for faster degradation (4), parameter testing moved to the combi­nations plotted in the upper middle and right-hand corner of Fig. 1.

The inoculated straw was added to clean, initially sterile columns fabricated from glass process pipe as previously described (4). The col­umns were prepared in triplicate with approx 50 g dry wt of inoculated stems in each column. The loaded columns were supplied with humidified oil-free instrument air at 193 kPa and a flow rate sufficient to turn over the air in the system once per day (about 10 mL/min). Approximately 2.5 g (dry wt) of straw was sampled from the top and bottom of each column initially and approx every 3 to 4 wk thereafter for 12 wk. The samples were com­bined, dried to constant weight overnight at 105°C, and ground to 60 mesh in a Wiley mill for compositional analyses.

Samples were analyzed for glucose by high-performance liquid chro­matography using an Aminex HPX-87H column at 65°C. The mobile phase was 5 mM H2SO4 at a flow rate of 0.5 mL/min. In addition, glucose concen­trations in some samples were determined by an enzymatic glucose test (Boehringer Mannheim GmbH, Mannheim, Germany).

Enzyme Assay

Filter paper activity, which describes the overall cellulolytic activity of an enzyme preparation, was determined by the method of Mandels et al. (16). A 1 x 6 cm strip of Whatman no.1 filter paper (Hillsboro, OR), which equals 50 mg of cellulose, served as the substrate and was added to the sample solution containing 0.5 mL of appropriate diluted enzyme (super­natant of culture broth) and 1.0 mL of 0.05 M citrate buffer (pH 4.8). After 60 min of incubation at 50°C, the hydrolysis was terminated by the addition of 3 mL of DNS solution, and the mixture was further assayed for reducing sugar content by the DNS method. One international filter paper unit (FPU) was defined as the amount of enzyme that releases 1 gmol of glucose/min under the assay conditions. Activities were reported as FPU/milliliter.

Amit Kumar, Jay B. Cameron, and Peter C. Flynn[2]

Department oaMechanical Enginee2ing,
Unive2bity oaAlberta, Edmonton,
Alberta, T6G 2G8, Canada,
E-mail: peter. alynn@ualberta. ca

Abstract

The cost of transporting wood chips by truck and by pipeline as a water slurry was determined. In a practical application of field delivery by truck of biomass to a pipeline inlet, the pipeline will only be economical at large capacity (>0.5 million dry t/yr for a one-way pipeline, and >1.25 million dry t/yr for a two-way pipeline that returns the carrier fluid to the pipeline inlet), and at medium to long distances (>75 km [one-way] and >470 km [two-way] at a capacity of 2 million dry t/yr). Mixed hardwood and softwood chips in western Canada rise in moisture level from about 50% to 67% when trans­ported in water; the loss in lower heating value (LHV) would preclude the use of water slurry pipelines for direct combustion applications. The same chips, when transported in a heavy gas oil, take up as much as 50% oil by weight and result in a fuel that is >30% oil on mass basis and is about two — thirds oil on a thermal basis. Uptake of water by straw during slurry trans­port is so extreme that it has effectively no LHV. Pipeline-delivered biomass could be used in processes that do not produce contained water as a vapor, such as supercritical water gasification.

Index Entries: Wood chips; pipeline; biomass; lower heating value; straw.

Introduction

Carbon-based power generation facilities do not typically rely on delivery of fuel by highway truck. Oil — and gas-fired plants rely on pipe­lines, and coal-based facilities typically either are located at the mine mouth or rely on rail or ship for fuel delivery. The reason for this is the high cost and high congestion that would be associated with delivery of large ton­nages of fuel to modern, large power plants.

Numerous biomass power plants are small and utilize truck delivery of fuel. However, in a previous work (1), we noted that optimum size for straw — and wood-based biomass power plants in a western Canadian set-

ting were 450 MW or greater for straw and wood from harvesting the whole forest, and that cost of power increased sharply at sizes below about 200 MW. For forest harvest residues (limbs and tops), which are more widely dispersed, the optimum size was 137 MW.

A 450-MW biomass power plant burning 2.1 million dry t/yr of wood chips would require 17 truck deliveries per hour at 20 t/truck (2). High­way transportation of fuel is a significant cost element, contributing at optimum power plant size 25, 14, and 38% of the total cost of power gen­eration from direct combustion of straw, wood from harvesting the whole forest, and forest harvest residues, respectively (1). In the present work, we evaluated pipeline delivery of biomass to a power generation plant, to avoid road congestion (and likely resistance by nearby residents), and to reduce overall fuel transportation cost.

Two carrier mediums are considered for biomass: water and oil. We review the inherent economics of truck vs pipeline transport, and then evaluate a case of field delivery of biomass by short-haul truck to a pipe­line terminal. We also evaluate the impact of water and oil absorption by the biomass fuel. Finally, we discuss the prospects for pipeline transport of biomass.

In this article, all costs are reported in year 2000 US dollars; Canadian dollars are converted to US at an exchange rate of 1.52 Cdn$/US$.

The larger-scale columns, while necessary to produce treated straw stems for use in the preparation of composite formulations for extrusion testing, were also a good test of the sensitivity of the system to scale up to larger columns and to changes in inoculum source and inoculation method. After 6 wk of treatment in the small columns used to inoculate the drums, 28-30% xylan was degraded while about 18.5% of glucan was degraded (Table 6). This gave ЛХ/AG ratios of 1.52-1.64. For comparison, at 40.0 mg of P. ostreatus/g of stems and 1.60 g of H2O/g of stems, the regression models predict 27.2% xylan degradation and 21.1% glucan deg­radation at 6 wk, for a AX/AG of 1.29. Clearly, P. ostreatus was dominant in the small columns used to prepare enrichment inoculum for the barrels. However, the AX/AG values were well outside the range predicted by the regression models. Apparently, the nitrogen-limited inoculum produced by Utah State University and shipped to INEEL for these columns was either more active or better acclimated to the nitrogen-limited conditions in the straw stems. This effect was repeatable, indicating that the history of the inoculum used may have a significant effect on AX/AG, although the actual glucan and xylan conversions were not far from the predicted values. The inoculum produced for the small columns used to inoculate the drums was produced differently than the inoculum used to inoculate the small columns in the moisture and inoculum tests—it was better accli­mated to nitrogen-limited conditions. This is because the mycelia were transferred directly into the nitrogen-limited medium (C/N of 32.6) with­out first being grown in the carbon-limited YM broth (C/N of 7.7). Since both enrichment steps in the production of the mycelial inoculum were carried out in the nitrogen-limited medium, this likely resulted in a myce­lial inoculum that was better acclimated to low-nitrogen conditions when it was added to the straw stems, which have a C/N of about 80 (17). Thus, system performance is sensitive to inoculum source and history.

The altered inoculation method also resulted in a different degrada­tion pattern than that observed in the small-column tests. After 6 and 12 wk of treatment in the drums, only 11.9 and 24.2% of the xylan was degraded, respectively (Table 6). Glucan degradation was similarly reduced, with only 13.4 and 26.8% of the glucan degraded at 6 and 12 wk, respectively. This equates to AX/AG values of 0.89 and 0.90 at 6 and 12 wk, respectively. Thus, selective degradation did not occur in the larger-scale columns, which indicates that P. ostreatus was not dominant. In fact, less degradation occurred in the barrels after 6 wk of degradation than was either observed or predicted in the small columns. It is likely that the low levels of degra­dation observed in the drums were owing to slower colonization of the fresh stems by the P. ostreatus growing in the solid enrichment inoculum. In the small-column tests, the indigenous microbes were shown to degrade about 15-20% of the polysaccharides in 12 wk in the absence of P. ostreatus (4,15), which likely represents the most easily accessible glucan and xylan fractions. If P. ostreatus were to colonize the straw more slowly from the solid enrichment inoculum, the primary effect on the degradation system would be to extend the treatment time necessary to reach the desired levels of xylan and/or glucan degradation in the final product. Thus, inoculating the fresh stems with partially degraded stems before introduction to the drums was an ineffective inoculation method when compared with inocu­lating by spraying homogenized mycelia onto the stem surfaces. The non­selective degradation pattern in the drums may or may not be a detriment to the physical properties of straw-thermoplastic composites produced from Degrade1 and Degrade2 straw stems, since selectively degraded stems have not yet been compared with nonselectively degraded stems.

Discussion

The specific gravity of particles changed with their size. Intact grain was much denser than coarsely chopped stover components. However, when material was finely ground, the density of all components increased and no further difference in density was observed. The results suggest that stalk and leaf have more micropores than husk and cob while grain had the least micropores. In principle, fine chopping and processing would con­tribute to an increase in the density of all components and the proportion of stover that sinks with grain in a water separation process.

In experiment 2, most of the corn grain sunk rapidly. Between the first and the eighth water separation, the amount of sunk grain increased by only 0.4% for the Prairie-du-Sac silage, 0.9% for the Binversie silage, 1.5% for the Ziegler silage, and 5.9% for the Arlington silage. Therefore, only one or two water separations would be needed to separate most grain.

When fresh silage was separated in water, the highest grain concen­tration achieved in the sunk material was 75%; this was observed with processed and relatively dry corn silage (64% MC). Short chopped mate­rial (8 mm) actually had a higher grain concentration (68%) than long material (17 mm), whose grain concentration was only 41% largely because of a high initial MC (74%). The moisture content had a greater impact than the physical form in the range that was observed (8- to 17-mm MPL, processed or not processed).

The effect of MC was even more apparent in experiment 3 when fresh material was compared with material partially dried (10 or 20 percentage units of moisture removal) or completely dried prior to water separation. Greater than 99% grain concentration was observed with bone-dry mate­rial. As corn silage becomes drier, stover pieces are likely to become more buoyant because of their large area compared to grain. Oven-drying is therefore a good pretreatment followed by water separation to concen­trate grain from corn silage. Because of the large amount of water in silage and the high cost to dry this material, the procedure could be used for small samples in the laboratory but would not likely be feasible in an industrial setting.

Without any pretreatment, hydrodynamic separation could allow the production of a concentrate of about 75% grain and 25% stover. It is difficult to achieve a higher grain concentration without having to par­tially dry or sieve the silage. Sieving increased the grain concentration in the sunk material to 79%. From an industrial point of view, water separa­tion would require recuperation of considerable amounts of soluble and deposited fine particles (18% of original DM after one separation, and between 21 and 26% after eight separations). Hydrodynamic separation could provide a feedstock with a high concentration in grain (75-80%), but it could not provide a stover-free feedstock without drying. Com­pared to pneumatic separation and sieving alone, hydrodynamic separa­tion is likely to require less energy when drying is not used. If pure corn

components are needed, the traditional combine thresher is more suitable for providing pure grain than any poststorage separation method applied to silage. Various harvest devices can be designed to collect the stover either simultaneously with threshing or afterward with another pass machine.

Acknowledgments

This research was partially supported by the USDA-ARS, UW Gradu­ate School, John Deere Technical Center, and Wisconsin Corn Promotion Board. We also acknowledge support from the Natural Science and Engi­neering Research Council of Canada and Agriculture and Agri-Food Canada.

References

1. Richey, C. B., Liljedhal, J. B., and Lechtenberg, V. L. (1982), Trans. ASAE 25(4), 834­839, 844.

2. Jenkins, B. M. and Sumner, H. R. (1986), Trans. ASAE 29(3), 824-836.

3. Ganesh, D. and Mowat, D. N. (1983), Can. J. Plant Sci. 63, 935-941.

4. Bilanski, W. K., Jones, D. K., and Mowat, D. N. (1986), Trans. ASAE 29(5), 1188-1192.

5. Corn Refiners Association. (1996), Corn Oil. 4th Edition. Corn Refiners Association, Washington, DC.

6. Gustafson, R. J. and Hall, G. E. (1972), Trans. ASAE 15(3), 523-525.

7. Pitt, R. E. (1983), Trans. ASAE 26(5), 1522-1527, 1532.

8. ASAE. (2002), Moisture Measurement—Forages. ASAE S358.2. Standards, 49th Edition. American Society of Agricultural Engineers, St. Joseph, MI.

9. ASTM. (2003), Standard Test Method for Specific Gravity of Soil Solids by Gas Pycnometer. Designation D5550-00. Volume04.08. American Society for Testing Materials, accessed at Website: www. astm. org.

10. Shinners, K. J., Jirovec, A. G., Shaver, R. D., and Bal, M. (2000), Appl. Eng. Agr. 16(4), 323-331.

11. ASAE. (2002), Method of Determining and Expressing Particle Size of Chopped Forage Material by Screening. ANSI/ASAE S424.1. Standards, 49th Edition. American Society of Agricultural Engineers, St. Joseph, MI.

12. Steel, R. G. D., Torrie, J. H., and Dickey, D. A. (1996), Principles and Procedures of Statistics: A Biometrical Approach. 3rd Edition. McGraw Hill, New York, NY.

Copyright © 2004 by Humana Press Inc.

All rights of any nature whatsoever reserved. 0273-2289/04/113/0055-0070/$25.00

Ethanol prices vary, over time, with changes in ethanol and corn pro­duction decisions, political statements from Washington regarding fuel policy, and expectations regarding the potential adoption of a national renewable fuels standard. These factors and others have caused moderate volatility in ethanol prices in recent years. Over the longer term, ethanol prices have been less volatile. For example, from January 1995 through May 2002, the average price of ethanol delivered to San Francisco ranged from $0.90 to $1.85/gal, with a mean value of $1.20/gal (16).

The net returns from producing ethanol are estimated by subtracting the adjusted production costs from the expected total revenue. We use a price of $1.20/gal of ethanol to represent a "base-case" scenario. Given that price, the estimated net returns range from a negative $5.92 ($1.20-$7.12)/ gal using California grapes to a negative $0.06 ($1.20-$1.26)/gal, using Midwestern corn.

David Morris[1]

Inntitute for Local Self-Reliance, Suite 303,
1313 5th Street SE, Minneapolin, MN 55415,
E-mail: dmorrin@ilnr. org

Abstract

This article addresses two questions: Has the effectiveness of the US government’s federal research and development (R&D) spending suffered from the post-1980 strategic change from freely shared and publicly owned to privately owned scientific advances? What criteria would a federal R&D program use to design a strategy that most effectively enhances the well­being of farmers and rural communities? Several studies found that the pre — 1980 US Department of Agriculture research strategy was very effective. No comparable studies have analyzed the comparative effectiveness of the post — 1980 strategy of restricting access to the results of public research. Recent experience and several analytical studies suggest that to significantly enhance the health of rural economies from an expanded use of plant matter as an industrial material, federal policy should channel scientific and engi­neering research into small- and medium-sized production and processing technologies and should encourage farmer-owned, value-added enterprises.

Index Entries: Ethanol; scale; ownership; research and development; effectiveness.

Introduction

One of the twentieth century’s greatest scientists and thinkers, Albert Einstein, observed, "Perfection of means and confusion of ends seem to characterize our society." Are the US government’s federal programs sup­porting industrial uses of biomass an example of such thinking?

The ends of the federal biomass programs are clear enough: enhanced national security, improved environmental protection, stronger rural economies. Research is a means to these ends. Are the federal biomass programs designed to most effectively achieve these ends?

Federal programs and policies channel scientific genius and entrepre­neurial energy and investment capital in specific directions. Regarding biomass, the overall impact of direct spending and tax incentives is not insubstantial.

The federal government spends more than $200 million annually directly on research and development (R&D) directed toward expanding industrial uses of biomass. Assuming a three — to one-average private-to — public investment match, this federal direct spending attracts an addi­tional $600 million in private spending. Federal biomass-related tax incentives presently "spend" more than $1 billion per year. This in turn attracts several billion dollars of private investment in these areas. The vast majority at present goes into expanding ethanol production.

This article examines some aspects of federal R&D efforts in the bio­mass area. It raises two areas of question. One is whether the post-1980 changes in the way the federal government conducts biomass research has made the biomass R&D effort more or less effective in achieving its stated goals. The observations are primarily directed toward the Department of Agriculture because the vast majority of its research is conducted in-house by permanent scientific staff.

The second area of question focuses on one of those goals, improving the well-being of America’s farmers and rural communities, and suggests that social criteria can and should inform and guide engineering research strategies. The observations are primarily directed toward the Department of Energy (DOE), although they are broadly applicable to agency heads and policy makers.