(374ba) Biomass Transportation Model and Optimum Plant Size for the Production of Ethanol

Introduction

            The
increasing costs of fossil fuels and environmental concerns linked with
greenhouse gas emissions have sparked great interest in cellulosic-based second
generation biofuels. Despite considerable efforts over the last five years to
develop economically feasible process, some challenges still remain to commercialize
cellulosic-based biofuels. In the Midwest, corn stover is an abundant agricultural
residue and thus a potential feedstock for biofuels. To be used in biofuel
production, agricultural residues must be collected from many farms spread over
a large area and transported to a single production facility giving rise to
challenges associated with the logistics of harvesting, collection, storage,
and transportation. The economic success of biofuels from agricultural residues
and the reduction of greenhouse gas emissions during their life-cycle depend on
resolving the logistic complexity associated with biomass collection (Overend, 1982).

            The
low energy density of agricultural residues along with the logistic hurdles limit
the scale of bio-refineries; this issue does not affect petroleum refineries.
As the plant size increases, the reduction of depreciation (from capital
investment) due to economies of scale leads to a reduction in the production
cost; on the other hand, the increase in biomass needed to meet plant demand leads
to an increase in the transportation distances (due to a larger collection
area), thus increasing the transportation cost. As a result of these competing
factors, bio-refineries present an optimum plant size that minimizes the total production
cost. Historical data from the chemical industry indicates that the
scaling exponent is around 0.6, commonly known as the six-tenth rule of thumb (Peters and Timmerhaus, 1991); the rule has great influence from petroleum refineries and petrochemicals plants. Some ambiguity exists regarding if the rule applies when scaling capital investment for bio-refineries.

Model
Description

            A
detailed model is developed to quantify the collection of agricultural residues
and transportation costs as well as to assess the optimum plant size for
biorefineries. The model has two main elements; specifically, the production
and transportation cost components. A simplified case study for an ethanol
plant, using a dilute acid hydrolysis process, located in Macon county in
central Illinois is used for the production model. The economic analysis of the production process is
based on work performed at National
Renewable Energy Laboratory (NREL) and
reported by Aden et al. (2002); an Aspen
Plus model is used to assess the cost of the production process. A two-step
approach is used to develop the
transportation component. First, a detailed model referred to as the "Farm
Model", in which individual farms are modeled, is used to obtain a
non-dimensional transportation factor for a base case plant size. In a second
step, the dimensionless factor in a simplified transportation model that scales
the transportation cost with plant capacity. The simplified transportation
model coupled with the production cost model are used to perform the
optimization analysis; this model is referred to as the "Optimization
Model".

            The
model accounts for collection costs, transportation, storage of biomass, and cost
associated with the conversion of corn stover to ethanol. The model assumes the
farmland is uniformly distributed around the processing facility and that there
are no local variations in the characteristics of the crop (i.e., the
yield and the other factors do not vary within the collection area). The harvesting
cost is based on the collection of round bales on a second pass. The bailing
cost includes collecting, shredding, raking, and staging of bales at the edge
of field. Storage of the corn stover is considered in the model; corn stover is
only harvested a few months per year, leading to the need for storing the
feedstock to operate the bio-refinery year-round. Fixed cost and variable costs are accounted for in the
production cost model. The former includes capital depreciation, maintenance,
insurance and taxes, and labor. The later includes raw materials, utilities,
waste handling, and consumables.

Results

            The
Farm Model is used to assess three
transportation schemes. Scheme I, assumes the corn stover is transported from
the farm to the processing facility in a straight line. Scheme II, assumes the
corn stover is transported along the catheti (i.e., sides) of the right
triangle formed by the radius connecting the farm to the plant and the axes of
the Cartesian coordinate system. Finally, in Scheme III, the road distance is
calculated from the straight-line distance (from Scheme I) by means of a
winding factor. An average winding factor was calculated from the average
transportation distance from Scheme II (the average transportation distance
from Scheme I is also required) and used for Scheme III. A shift of the
distribution towards lower transportation distances is observed when using the
winding factor to calculate road distances compared to actual road distances
even though the average transportation distance is equal. In the transportation
industry, the hauling cost per distance decreases with increasing
transportation distance; higher unit cost per distance are charged for shorter
hauls (Glassner et al., 1998). The difference in the distributions of
transportation distances between actual road distances (Scheme II) and the ones
obtained through the winding factor (Scheme III) would lead to different total
transportation costs (i.e., for all the biomass transported to a given
plant). The transportation cost estimated from the use of the winding factor
would be greater than that paid from actual road distances due to the higher
number of shorter hauls.

            The
optimization problem requires that the transportation and production components
scale with capacity. One of the main components of the production cost is the
capital depreciation which is calculated from the capital investment. As
commonly done, the capital cost is expressed with an exponential
function to capture the effect of economy of scale. The investment for 2000
tonne/d and 907 tonne/d plants were obtained from the detailed model developed at the NREL (Aden et al., 2002). The capital investment scaling exponent was calculated.
The impact of plant size on the scaling exponent is also investigated.

            The
optimum capacity and minimum total production is greatly impacted by the farmer
participation. The rate of change of production cost (i.e., slope of the
curve) decreases as farmer participation increases; the rate of change
decreases significantly (i.e., the slope of the curve flatness) for
farmer participations greater than 0.4. The observed behavior is a result of
the relationship between the cost of delivered corn stover and farmer
participation. The cost of delivered corn stover is proportional to the square
root of the inverse of the farmer participation. As a consequence of the
proportionality, the reduction in cost for increasing farmer participation from
0.2 to 0.5 is greater than that of increasing it from 0.5 to 0.8. Biofuel
producers should target at least a farmers participation of 0.5; this point is in
the zone where the minimum production cost has a reduced sensitivity to the farmer
participation (greater than 0.4). An increase of farmer participation to 1.0
may require significantly more resources and the economic benefit, reduction of
the production cost, might not justify it. The optimum plant size increases
with increasing farmer participation.

Conclusion

            A
detailed model that includes cost of collection, transportation, storage, and
chemical transformation was developed to evaluate the optimum plant size of
bio-refineries. A simplified bio-ethanol refinery via dilute acid hydrolysis
from corn stover is presented as a case study. The conversion of straight-line,
farm-to-plant distances to road distance via a winding factor shifts the
distribution of transportation distances towards shorter hauls. The capital
investment scaling exponent was calculated. The cost of the delivered
corn stover is proportional to the square root of the inverse of the farmer
participation; as a consequence, biofuel producers using agricultural residues
as feedstock should target a farmer participation of fifty percent.

References

Aden, A., Ruth, M., Ibsen, K.,
Jechura, J., Neeves, K., Sheehan, J., Wallace, B., Montague, L., Slayton, A.,
Lukas, J., 2002. Lignocellulosic biomass to ethanol process design and
economics utilizing co-current dilute acid prehydrolysis and enzymatic
hydrolysis for corn stover. U.S. Department of Energy Laboratory, National
Reweable Energy Laboratory, Golden, CO.

Glassner, D., Hettenhaus, J.,
Schechinger, T., 1998. Corn stover collection project. Procedings of
BioEnergy'98?Expanding Bioenergy Partnerships, Madison, WI, pp. 1100-1110.

Overend, R.P., 1982. The average
haul distance and transportation work factors for biomass delivered to a
central plant. Biomass 2, 75-79.

Peters, M.S., Timmerhaus, K.D.,
1991. Plant design and economics for chemical engineers,  fourth ed.
McGraw-Hill, New York.