A supply chain is defined as a network of facilities that performs the functions of procurement of materials, transformation of these materials into intermediate and finished products, and distribution of these products to customers. A typical supply chain comprises suppliers (production plants), storage facilities (Warehouses) and customers. It involves two basic processes tightly integrated with each other: (i) the production planning and inventory control process, which deals with manufacturing, storage, and their interfaces, and (ii) the distribution and logistics process, which determines how products are retrieved and transported from the warehouse to retailers. A hydrogen infrastructure can be defined as the supply chain required to manufacture, store and deliver hydrogen to the consumer. Like any supply chain it consists of several distinct components. Production processes are required to convert primary energy resources into hydrogen. Storage units and terminals are needed to compensate for fluctuations in demand. Distribution systems are essential for transporting hydrogen from the production facilities to the point of sale, i.e., demand. Finally, dispensing/refuelling technologies allow transfer of hydrogen to users at forecourt retail stations.
In traditional SCM (supply chain management), minimizing costs or maximizing profit as a single objective is often the optimization focus. Moreover, safety of hydrogen transportation turns out to be one of the most basic functions of the hydrogen infrastructure system. Thus, safety level of hydrogen should also be taken into consideration when formulating a SC model even if it is difficult to quantify it as a monetary amount in the objective function. It can be expected that there is a conflict between these objectives, i.e. the most profitable infrastructure is not necessarily also the best safety level. Because of this trade-off there is not a single solution to this class of problem. Therefore, it is proposed to set up a multiobjective design problem whose solution will be a set of Pareto optimal possible design alternatives representing the trade-off among the different objectives. In this paper, hydrogen infrastructure system is formulated as a multiobjective stochastic MILP model, which is solved by using the standard E-constraint method. This formulation takes into account both SC profit and safety level, which is defined as the probability. The uncertainty associated to the safety level is represented by a set of scenarios with given probability of occurrence. Such scenarios together with their associated probabilities must be provided as input data into the model. And the amount of products to be produced and stored in the nodes of the SC, the flows of materials transported among the entities of the network and the product sales are considered as decision variables. At the end of the design horizon, a different value of cost and safety level is obtained for each particular realization of safety uncertainty.
To illustrate the features of the model, the results of an industrial case study conducted are used. It consists of a geographical region where 6 production sites have been identified for the potential installation of central production technologies. Demand for hydrogen by FCV drivers is expected at 10 major cities, acting as the markets in the formulation. Of the 6 central production sites, some are existing refineries, chemical complexes and natural gas compression stations. On the other hand, all of the 10 major cities can the produce hydrogen of its own demand to reduce excessive transportation cost. But local production has more production cost than the production in supply site. Of the two objective functions in this case study, the first is Profit, which is calculated for each scenario and time interval as the difference between the revenues and the total costs including production cost and transportation cost. The second objective function is safety level related to Frequency, which is the amount of products to be transported between the nodes of the SC. It means that as the amount of products to be transported is increased, the safety level decreases. The safety level is also associated with Consequence, which presents the uncertainty of this study.
To use hydrogen successfully as the fuel in the future, it is required to create an entirely new fuelling infrastructure, from production and storage to distribution. Any investment strategy for building up a hydrogen supply chain needs to be supported by rigorous quantitative analysis. But determining the optimal SC configuration is a difficult problem since a lot of factors and objectives must be taken into account when designing the network. To solving the high complexity of this problem, this paper represents multiobjective stochastic model. The developed mathematical model provides the efficient solutions, which represent an alternative SC configuration and corresponding investment strategy, each achieving a unique combination of safe and economic performance. This way of generating different possible configurations will help the decision-maker determine the best design according to the selected objectives.
KEYWORDS Hydrogen infrastructure, Multi-objective optimization, Mixed integer linear programming, Safety, Stochastic model