Systems Informatics and Analysis - Engineering and Science of Biomass Feedstock Production and Provision - page 217

biorefi neries, and the satellite storage and preprocessing locations, has been of
particular interest.
• Applications at the on-farm production level have been very limited. As pointed
out before, this is possibly due to the lack of commercial farming of energy
crops. This leads to lack of data to support such models.
• Many models have used Microsoft Excel ® to store and retrieve data as well as to
provide user interfaces for scenario development.
• The application of GIS-based approaches has recently increased, either to esti-
mate the availability of biomass at a local or regional level [ 123 - 125 ] or as part
of a decision-making model [ 90 , 126 - 128 ]. Use of GIS provides accurate infor-
mation, which means that the model predictions can be more realistic and readily
implementable. A challenge is to make the information provided by a GIS sys-
tem compatible with the decision-making model, which often requires work on
software and informatics. Moreover, the computations become more challeng-
ing. However, with the availability of better computing facilities as well as
greater accessibility to GIS data, the application of such approaches is expected
to increase in the future.

8.5 Future Challenges and Recommendations

The review has also identifi ed some research gaps that must be addressed in the
future. These are summarized below:

• Reliable input data that are experimentally validated are needed for model
simulations. Currently, there is a substantial lack of data related to actual yield of
the crop, fi eld losses, machinery performance, and storage losses. While most
models use values reported in the literature, these data points are extremely lim-
ited. Recently, Shastri et al. [ 129 ] showed the value of incorporating experimental
results in a modeling framework. Such approaches should be adapted more often.
• The models should account for the inherent uncertainties in the system, such as
weather, yield, maturity schedule, and equipment breakdown.
• The input data, model constraints, and assumptions must be standardized. The
life-cycle impact assessment studies have shown that differences in assumptions
and system boundaries can vastly impact the results.
• Storage of biomass has often been ignored in many early models. However, sea-
sonal availability will defi nitely necessitate storage. As pointed out in Chap. 7 ,
quality degradation and total biomass loss can severely impact the feedstock
supply. Therefore, storage costs and design of storage facilities must be a part of
the models.
• Model validation is important to build trust among the users. There has been a
considerable amount of work on validating the crop growth models with fi eld
studies. However, such efforts for other levels of models described here have
been limited. Lack of a commercial-scale operational system for second-
generation biofuels makes validation challenging. As an alternative, the models