(394d) Multi-Objective Planning Approach for Supply Chains Based on Biomass Processing By a Sequential Methodology Via Geographic Information System and Mathematical Programming
AIChE Annual Meeting
2018
2018 AIChE Annual Meeting
Sustainable Engineering Forum
Design, Analysis, and Optimization of Sustainable Energy Systems and Supply Chains II
Tuesday, October 30, 2018 - 4:30pm to 5:00pm
Biomass is a renewable resource with attractive characteristics to produce energy and biofuels. One of main reasons to use biomass as renewable source for energy production is that biomass can capture a significant amount of CO2 emissions during growing. In this respect, diverse studies have stated that biomass used for biofuels and energy production can contribute partially to solve the energy demand problem and mitigate the increasing of CO2 emissions. Most works have focused on developing new technologies, processes and processing system based on biomass. In addition, several works have focused in determining locations, raw materials and processing technologies. Nevertheless, most of them have based the options to select potential locations on their own criteria, neglecting important issues associated to viability; such as distance to roads for transportation or distance to water bodies to avoid affecting rivers and lakes.
Therefore, models based on Geographic Information Systems (GIS) are useful tools for data management and processing in order to determine viable locations in this type of systems. GIS models have not been widely used to assess the economic performance of a production system. For that reason, GIS methodologies can be complemented using mathematical programming strategies. Therefore, this paper proposes a sequential and multi-objective approach for the optimal planning of a residual biomass processing system. The methodology considers the selection of potential locations through a multi-criteria methodology based on GIS; which considered the uncertainty associated to the amount of available biomass by diverse scenarios generation. Additionally, the approach uses the viable locations obtained by the GIS methodology and then selects locations larger than a minimum size in order to define where processing facilities can be installed. Also, chosen locations by the GIS methodology can receive raw material from specific harvesting sites; which should be into a radius of 161 km from the centroid of processing facilities. Subsequently, data of potential locations for processing facilities and harvesting sites are fed in a mathematical formulation to determine the optimal supply chain topology considering various points of view. The mathematical programming approach consists of mass balances to define the interconnections between the different supply chain nodes, as well as, constraints to model the considered technologies. Moreover, mathematical model involves two objective functions. Environmental objective is modeled via the eco-efficiency while the economic performance is associated to net annual profit; which produce a multi-objective programming problem. It should be noted that the proposed approach takes into account raw material production variations, since several probable values for available biomass amount were taken into account when scenarios were generated. The methodology was tested in a case study for Mexico considering 8 types of residual biomass (corn bagasse, sugar cane bagasse, wheat straw, barley straw, sorghum straw, rice straw, pecan nut shell and agave residue), 8 types of products (ammonia, ethanol, methanol, acetaldehyde and 4 types of Fisher Tropsch liquids), 9 processing routes, as well as different number of potential processing plants according to the raw material. The number of potential processing facilities were 257 for corn straw, 57 for sugar cane bagasse, 3 for wheat straw, 0 for barley straw, 54 for sorghum straw, 47 for rice straw, 8 for pecan nut shell and 6 for agave residue. The results present the trade-off between both objectives through a Pareto curve, where each of points represent a different production system based on biomass. Also, the results illustrate that the proposed methodology can be used to determine the supply chain topology considering diverse initial highly viable locations and several objectives. Furthermore, results show that scale and number of considered locations have a strong influence over the transportation, capital and operating costs.
Therefore, models based on Geographic Information Systems (GIS) are useful tools for data management and processing in order to determine viable locations in this type of systems. GIS models have not been widely used to assess the economic performance of a production system. For that reason, GIS methodologies can be complemented using mathematical programming strategies. Therefore, this paper proposes a sequential and multi-objective approach for the optimal planning of a residual biomass processing system. The methodology considers the selection of potential locations through a multi-criteria methodology based on GIS; which considered the uncertainty associated to the amount of available biomass by diverse scenarios generation. Additionally, the approach uses the viable locations obtained by the GIS methodology and then selects locations larger than a minimum size in order to define where processing facilities can be installed. Also, chosen locations by the GIS methodology can receive raw material from specific harvesting sites; which should be into a radius of 161 km from the centroid of processing facilities. Subsequently, data of potential locations for processing facilities and harvesting sites are fed in a mathematical formulation to determine the optimal supply chain topology considering various points of view. The mathematical programming approach consists of mass balances to define the interconnections between the different supply chain nodes, as well as, constraints to model the considered technologies. Moreover, mathematical model involves two objective functions. Environmental objective is modeled via the eco-efficiency while the economic performance is associated to net annual profit; which produce a multi-objective programming problem. It should be noted that the proposed approach takes into account raw material production variations, since several probable values for available biomass amount were taken into account when scenarios were generated. The methodology was tested in a case study for Mexico considering 8 types of residual biomass (corn bagasse, sugar cane bagasse, wheat straw, barley straw, sorghum straw, rice straw, pecan nut shell and agave residue), 8 types of products (ammonia, ethanol, methanol, acetaldehyde and 4 types of Fisher Tropsch liquids), 9 processing routes, as well as different number of potential processing plants according to the raw material. The number of potential processing facilities were 257 for corn straw, 57 for sugar cane bagasse, 3 for wheat straw, 0 for barley straw, 54 for sorghum straw, 47 for rice straw, 8 for pecan nut shell and 6 for agave residue. The results present the trade-off between both objectives through a Pareto curve, where each of points represent a different production system based on biomass. Also, the results illustrate that the proposed methodology can be used to determine the supply chain topology considering diverse initial highly viable locations and several objectives. Furthermore, results show that scale and number of considered locations have a strong influence over the transportation, capital and operating costs.