Page 216 - Sustainability in the process industry
P. 216

S o f t wa r e  T o o l s  193


                     material and operating unit appears exactly once. In the composition
                     phase, the nodes are linked stepwise, layer by layer, starting from the
                     shallowest end (i.e., the final-product end) of the remaining input
                     structure. The algorithm proceeds by assessing whether any of the
                     linked nodes violates any of the axioms described in Chapter 7.
                        The structure generation is performed transparently, and the
                     maximum structure that results is the input for the Solution Structures
                     Generation (SSG) algorithm. If a material has to be produced, then
                     the SSG algorithm generates all possible ways for its production. For
                     example, if M1 can be produced by O2 or by O3 then the possibilities
                     include production by O2 alone or by O3 alone or by using both.
                     Once an operating unit is included, its input materials must be
                     produced as well, and so forth. Materials are selected in a specific
                     order. The parent–child relation between steps ensures that the
                     materials are selected by the process according to this order.
                        The SSG algorithm yields all the combinatorially feasible solution
                     structures of a given problem. Unfortunately, the number of feasible
                     structures at this stage is  often too large to be enumerated explicitly.
                     Therefore, the Accelerated Branch-and-Bound (ABB) algorithm is
                     used to determine the optimal structure without generating all the
                     possible solutions. Input to this algorithm includes not only the
                     structural relationships between materials and operating units but
                     also such additional information as the costs of each raw material,
                     the fixed and proportional costs of operating units, and the constraints
                     (if any) on the quantity of materials and the capacity of operating
                     units.
                     9.2.2 S-Graph Studio
                     S-Graph Studio is a software package that enables the user to design
                     batch processes and to optimize them via various optimization
                     methods (S-Graph, 2009). The program also allows scheduling
                     problems to be defined using graphical tools. It has a modular
                     architecture, so different solvers can be used with the program.
                     S-Graph Studio uses the industry’s standard file format: BatchML (as
                     defined by the ISA-88 standard of the World Batch Forum), which is
                     used to exchange information between industry sites and plants as
                     well as for other purposes. The Excel file format can also be used for
                     both input and output. The software includes a solver that utilizes
                     the S-graph methodology developed at the University of Pannonia
                     (S-Graph, 2009).
                        One of the main goals of batch process optimization is to
                     minimize makespan—that is, finding the shortest time in which a
                     process can be completed using available resources. S-Graph Studio
                     can be used to define batch processes in terms of the tasks to be
                     performed, the available equipment units, and task completion times
                     (as a function of the equipment units used). This information is
                     necessary and sufficient for minimizing the makespan of a process.
   211   212   213   214   215   216   217   218   219   220   221