| The main goal of this project was to build a model of an off-gas distribution system in QIT Fer et Titane plant (Sorel, Quebec) and to apply a real-time optimization algorithm. During process, carbon monoxide is produced. This off-gas is reused as fuel by many different plant's units. Some of those units can also use alternative fuel during low CO gas availability.;Actually, QIT does not have equipment to stock CO gas. The only way to do a time-variant using of this gas is by variable use of products and by-products stock piles and equipments flows, and by equipment shutdowns. This way, the CO gas is used when it's available. Optimization take into account future perturbations happening during time horizon going from few days up to one week. For example, those perturbations could be a downturn in electrical power consumption or planned shutdown of equipments.;The model is a static and linear one and was built using mass and energy balances. Discrete variables were used to model decisions to operate or not equipments so the optimization problem is a mixed-integer linear programming problem.;The objective function to be minimized is a sum on all horizon time of costs caused by use of alternative fuel, equipments shutdowns and material handling from stock piles for each time step. Constraints are also defined for each time step: minimum and maximum of stock piles, flow limits for stock piles handling, equipments production limits, CO gas availability, number of equipments available to operation. GAMS software is used to simulation operation cases. (Abstract shortened by UMI.). |
