TY - JOUR
T1 - A reactive-iterative optimization algorithm for scheduling of air separation units under uncertainty in electricity prices
AU - Basán, Natalia P.
AU - Cóccola, Mariana E.
AU - Dondo, Rodolfo G.
AU - Guarnaschelli, Armando
AU - Schweickardt, Gustavo A.
AU - Méndez, Carlos A.
N1 - Publisher Copyright:
© 2020 Elsevier Ltd
PY - 2020/11/2
Y1 - 2020/11/2
N2 - The high energy demand in power-intensive processes and the possibility of reducing the energy bills by an optimal scheduling are the motivation for incorporating energy consideration in the production scheduling of air separation plants. Optimization opportunities exist at different time scales for day-ahead scheduling decisions and real-time decisions regarding all fluctuations in electricity prices. Consequently, this paper presents a reactive-iterative optimization approach, integrating the rolling horizon (RH) concept into an iterative solution algorithm, for optimizing production decisions when an industry participates in both the day-ahead electricity market and the spot electricity market. A novel discrete-time MILP formulation is used as a basis of the proposal, which allows adjusting production rates to electricity prices varying hourly or faster. Several scenarios from a real-life air separation industrial plant are solved to show interesting trade-offs between the predictive approach and the reactive-iterative strategy.
AB - The high energy demand in power-intensive processes and the possibility of reducing the energy bills by an optimal scheduling are the motivation for incorporating energy consideration in the production scheduling of air separation plants. Optimization opportunities exist at different time scales for day-ahead scheduling decisions and real-time decisions regarding all fluctuations in electricity prices. Consequently, this paper presents a reactive-iterative optimization approach, integrating the rolling horizon (RH) concept into an iterative solution algorithm, for optimizing production decisions when an industry participates in both the day-ahead electricity market and the spot electricity market. A novel discrete-time MILP formulation is used as a basis of the proposal, which allows adjusting production rates to electricity prices varying hourly or faster. Several scenarios from a real-life air separation industrial plant are solved to show interesting trade-offs between the predictive approach and the reactive-iterative strategy.
KW - Air separation plant
KW - Continuous power-intensive processes
KW - Discrete-time MILP scheduling model
KW - Electricity price uncertainty
KW - Reactive optimization approach
UR - http://www.scopus.com/inward/record.url?scp=85089441219&partnerID=8YFLogxK
U2 - 10.1016/j.compchemeng.2020.107050
DO - 10.1016/j.compchemeng.2020.107050
M3 - Article
AN - SCOPUS:85089441219
SN - 0098-1354
VL - 142
JO - Computers and Chemical Engineering
JF - Computers and Chemical Engineering
M1 - 107050
ER -