A novel multi-period mixed-integer linear optimization model for optimal distribution of byproduct gases, steam and power in an iron and steel plant

Abstract

Byproduct gases, steam and electricity play an important role in providing energy for production units in the iron and steel industry. Optimal distribution of byproduct gases, steam and electricity in an iron and steel plant can significantly decrease energy cost and reduce CO 2 emissions. However, such optimal distribution is not trivial because it involves many production units, steam turbines, combined heat and power units, and waste heat and energy recovery units, and several realistic operational features such as byproduct gas mixing, byproduct gas level control in dedicated gasholders, different levels of steam requirement, minimum heating and energy requirement, and maximum allowable burner switches, resulting in a large complex combinatorial problem. In this paper, we develop a novel multi-period mixed-integer linear programming model for optimal distribution of byproduct gases, steam, and power in an iron and steel plant. The consuming rates of byproduct gases are variable. Different byproduct gases are allowed to be mixed to satisfy minimum heating and energy requirement of production units. The steam is specifically classified as high, medium and low pressure. New binary variables are introduced to determine electricity purchase or sale decision with each having different price. The burner switching operation is correctly modeled with fewer binary variables allowing turning on and off at any time. Several important practical features such as fuel selection, gasholder level control, ramp rate variation, piecewise constant generation rates of byproduct gases, and piecewise constant demand profiles of byproduct gases, steam and electricity are also incorporated. The computational results demonstrate that the optimal operating cost is obtained within 2 CPU seconds for an industrial example using the proposed model, which is reduced by 6% compared to that from actual operation.

The business advantage of identifying and solving pseudo-continuous-integer periodical linear problems

“…Notice that the model in ( 2) is mathematically equivalent to (3) through (5) but the latter happens to have integer solutions for all its variables. Since both problems share the same A, c and T > 0 they share the same optimal basis (basic columns of matrix A) and the value for the dual variables.…”

A Review of Mathematical Programming in Integrated Iron‐ and Steelmaking

“…The intuitive capability of higher gasholder capacities having an impact on CO 2 emissions, because flaring can be reduced, is confirmed in this study. Zeng et al proposed an MILP model for by‐product gas distribution, which considered parallel use of fuels in burners. The model applies state‐of‐the‐art penalties and maintenance costs and minimizes operational costs as the objective.…”

Multi-Objective Optimization of Steam Power System Under Demand Uncertainty

“…A mixed-integer nonlinear programming (MINLP) model was developed to improve the utilization of surplus by-product gas from SPS by Zhang et al [2]. For optimal distribution of by-product gases, steam and electricity, a multi-period MINLP model of energy system in iron and steel plant was proposed by Wei et al [3] and Zeng et al [4]. Zhao, Rong and Feng [5,6] proposed an MINLP model for refinery consisting of production and utility systems.…”