Systems Informatics and Analysis - Engineering and Science of Biomass Feedstock Production and Provision - page 218

could be designed to validate the agricultural residue system such as corn stover,
which is more established.
• There is a disconnect between the assumptions and cost estimates among
different models. In particular, the disconnect between models developed at dif-
ferent scales (i.e., farm, regional, and national levels) needs to be addressed. For
example, national-level models often assume that biomass can be grown on
degraded lands that are often small in terms of area. Shastri et al. [ 1 ] have shown
that the per-unit cost of production for small farms (less than 100 ha) can be
substantially higher than the average cost. Such trends, though, are ignored in
national-level models.
• For realistic cost numbers, farms of all sizes typically observed in current agri-
culture must be considered. Costs are typically calculated assuming one farm
size, which is often quite large. This will underestimate the actual costs [ 1 ]. This
becomes even more important when we consider that farms may use only a frac-
tion of their land initially for growing energy crops.
• One option to address seasonal availability of feedstocks is to process multiple
feedstocks at different times in the biorefi nery. This would reduce storage
requirements substantially. Optimization of the BFPP system for such scenarios
has generated interest in the last few years [ 93 , 94 ]. Kenney et al. [ 130 ] have
proposed mixing of feedstocks to address signifi cant compositional variability in
feedstock. The supply chain logistics considering such modifi cations needs to be
further explored.
• Greater emphasis should be placed on incorporating the environmental and
social performance indicators explicitly in the modeling approaches. This may
require the solution of a multi-objective optimization problem to highlight the
trade-offs between different dimensions of sustainability.
• Efforts should be made to integrate models developed at different scales as well
as models addressing different aspects of BFPP (Fig. 8.6 ). There could be one
single model that covers all the scales. This would be extremely challenging
from a modeling and computational standpoint. Therefore, seamless integration
of multiple models addressing different questions should be targeted. In process
engineering, the CAPE-OPEN standard has been developed that enables the
seamless integration of process and equipment models of different scales ( http://
www.colan.org/ ) . Perhaps such an approach should be utilized. This opens up the
fi eld of multi-scale modeling for bioenergy systems.
• The role of informatics, including DSS, has been limited. This restricts the dis-
semination of decision-making tools that would be extremely valuable to a num-
ber of stakeholders. User-friendly DSS can enable even nonexperts to study
specifi c cases for decision making. Efforts should also be made to make these
DSSs web-based to further promote dissemination. Some model-based systems,
such as BPSys, IBSAL, and APSIM, have successfully shown the integration of
informatics with modeling and analysis. Jakku and Thorburn [ 131 ] emphasized
the value of social learning and have recommended a framework to develop
participatory DSS in agriculture.