| Abstract: |
Ethanol, the common name for ethyl alcohol, is fuel grade alcohol that is
predominately produced through the fermentation of simple carbohydrates by
yeasts. In the United States, the carbohydrate feedstock most commonly used in
the commercial production of ethanol is yellow dent corn (YDC). The use of
ethanol in combustion engines emits less greenhouse gasses than its petroleum
equivalent, and it is widely hoped that the increased substitution of
petroleum by ethanol will reduce US dependence on imported oil and decrease
greenhouse gas emissions. Production of ethanol within the United States is
expected to double, from 3.4 billion gallons in 2004, to about seven billion
gallons in the next five years. Two processes currently being utilized to
produce ethanol from YDC are dry milling and wet milling. The wet mill process
is more versatile than the dry mill process in that it produces a greater
variety of products; starch, corn syrup, ethanol, Splenda, etc., which allows
for the wet mill to better react to market conditions. However, the costs of
construction and operation of a wet mill are much greater than those of a dry
mill. If ethanol is the target product, then it can be produced at a lower
cost and more efficiently in a dry mill plant than in a wet mill plant, under
current economic conditions. Of the more than 70 US ethanol plants currently
in production, only a few are of the wet mill variety. A descriptive
engineering spreadsheet model (DM model) was developed to model the dry mill
ethanol production process. This model was created to better understand the
economics of the ethanol dry mill production process and how the profitability
of dry mill plants is affected under different conditions. It was also
developed to determine the economic and environmental costs and benefits of
utilizing new and different technologies in the dry mill process.
Specifically, this model was constructed to conduct an economic analysis for
novel processes of obtaining greater alcohol yields in the dry mill process by
conducting a secondary fermentation of sugars converted from lignocellulosics
found in the dry mill co-product, distiller’s grains. This research is being
conducted at Purdue University, Michigan State, Iowa State, USDA, and NCAUR
under a grant from the US Department of Energy. The DM model is more
technically precise, and more transparent, than other models of the dry mill
process that have been constructed for similar purposes. The Tiffany and
Eidman model (TE model) uses broad generalities of the dry mill process, based
on the current state of production, to approximate the sensitivities of the
process to changes in variables. The TE model parameters were well researched,
but the model suffers from several drawbacks. The main limitations of this
model are that the results are very sensitive to the input values chosen by
the user. Unlike the DM model, complex manipulations, such as determining the
effect of new technologies would require accurate parameter estimates using
the TE model. The McAloon model [11].uses highly technical engineering
software (ASPEN) that acts essentially as a “black box” in the dry mill
production process. This very exact model does not allow for a more general
examination of the dry mill process. Changes in the production process would
necessitate precise engineering plans. Similar to the TE and McAloon models,
the DM model is a spreadsheet model, but unlike the McAloon model it is
completely self-contained. The DM model is a feed backward model, input
requirements (corn, enzymes, chemicals, utilities, etc) are calculated based
on the user entered values for annual production and process parameters. The
mass flow rates, in pounds per hour were then calculated and used in
estimating the size, in dimension or power, of each major piece of equipment.
The cost associated with each piece of major equipment was then calculated as
an exponential function of its corresponding size. Total capital costs
associated with a dry mill plant were then estimated using the percentage of
equipment costs method [13]. It was found that the DM model estimates of the
total capital costs associated with medium to large dry mill plants (those
with the capacity to produce between 10 and 100 million gallons of ethanol a
year) were within 5% of total fixed costs estimated by BBI [2]. Operating
costs were compared with the 2002 USDA survey results and also found to be
very close [15]. A companion document, “Economic and Technical Analysis of Dry
Milling: Model User’s Manual,” staff paper no 06-05, explains how the model is
used to conduct analysis of dry milling alternatives. |