Conclusions and Directions for Further Research

To summarize, we have explained the importance of energy throughout the world. We clarified that energy plays a vital role in today's human lives, so planning and scheduling the transportation of energy in an almost optimal situation is inevitable. Therefore, modeling the energy freight-transportation network requires modeling real problems and then solving them with methods such as operation research or fuzzy logic.

However, there are many components involved in a transportation network that affect network modeling. Modes of transportation alter the planning and modeling of a network. For instance, maritime transportation differs greatly from other modes of transporting energy—that is, trucks and trains. It includes specific characteristics and requires decision support models which would be appropriate to solve its problems. More research in maritime transportation problems has been done recently than ever before, but the field still needs much attention compared to other modes. To model and solve more realistic problems in maritime transporta­tion, there has to be development of optimization algorithms and computing power.

While trying to design a model that is able to minimize the cost, travel destina­tion, and time, it is necessary that the risk of such models when becomes applicable should be at the lowest possible level. Some presented models in previous sections were specific to energy transportation, and others were general in transporting haz­ardous materials of which energy is a part. Some models were explained, whereas for other problems we confined ourselves to a review of research about the trans­portation of energy that has been categorized in different fields from modes of transportation to single- or multiple-objective models. Readers were referred to sources that deal more extensively with the problems.

Erkut et al. [26] suggest that researchers emphasize global routing problems on stochastic time-varying networks because it has received almost no attention. The problem is so close to reality and most of maritime transportations are global and goes through international waters.

Furthermore, risk models and the probabilities of risk during freight transporta­tion still need to be studied. The field would be rather difficult to survey, because there is no agreement on general accident probabilities and conflicting numbers are reported by different researchers. Lack of essential data limits improvements in such fields, and perhaps more attention should be paid to quantifying and modeling perceived risks. In general, risk and its relevant topic in energy freight transporta­tion is of high importance, but unfortunately it has attracted little attention.

During the previous decade, attacks on energy freights increased as the price of energy rose. Energy freight can be a significant target to terrorists around the world. This fact raises the interest in the security of such freight. The US federal government, for instance, now requires hazmat truckers to submit to fingerprinting and criminal background checks [71 J. However, security as an important factor in freight transportation has not yet received much attention from operations research­ers. Obviously, the problem is complex, and many parameters should be considered while modeling. Erkut et al. [26] propose three dimensions for operation research­ers to focus on security issue: rerouting around major cities, changes in the model­ing of incidence risks, and route-planning methodology.

In addition, as fleets become larger, network planning problems become harder. So the need arises for a new generation of researchers and planners who have less practical but more academic backgrounds. As computer technology advances, new software and optimization-based decision-support systems are introduced for the varieties of applications in energy freight transportation. These advances make it easier to model all of the important problem components. This new generation of planners is more adapted to computers and software and therefore is capable of modeling realistic issues and finding good solutions to hard problems in a reason­able amount of time.