A Development of Demand Response Model and Algorithm Based on State-Task Network for Industrial Facilities in Smart Grid

Abstract

Smart grid integrates information communication technologies (ICT) with electric grid and enables two-way communication between utility suppliers and electricity consumers. In smart grid, demand response (DR) energy management is a key technology which allows electricity consumers to actively participate in the management of electric grid. DR improves the reliability and reduces electricity cost of consumers by reducing the wholesale peak demand of the electric grid. On the demand side, industrial facilities consume huge amounts of electricity, highlighting the urgent need to implement DR energy management. A general model of DR energy management systems for industrial facilities was firstly introduced in this study. The model consists of model elements, model architecture, and approaches to industrial DR. Based on state-task network (STN), the model divides the processing tasks in industrial facilities into non-schedulable tasks (NSTs) and schedulable tasks (STs) to implement DR. The proposed model provides a straightforward means of designing and analyzing DR systems in industrial facilities and assists in developing standards for such systems. Besides, an example of this DR model was presented with a steel manufacturing facility. Based on the model, a DR energy management scheme for industrial facilities was introduced utilizing the state-task network (STN) and mixed integer linear programming (MILP). Considering day-ahead hourly electricity prices, the DR problem in industrial facilities was formed into a series of linear constraints and one objective function. Through minimizing the energy costs of overall industrial facilities, the scheme determines the scheduling of STs and distributed energy resources (DERs). In this way, the electricity demand of overall industrial facilities is shifted from peak periods (with high electricity prices) to off-peak periods (with low electricity prices), which not only improves the reliability of the electric power system, but also reduces energy costs for industrial facilities. Finally, the performance of the DR energy management scheme was verified based on oxygen generation facilities. |스마트그리드는 전력망과정보통신기술(ICT)를 통합하고 전력공급자와 전력이용자간의 양방향통신을 가능하게 한다. 스마트그리드에서 수요반응(DR) 에너지 관리는 전력이용자가 능동적으로 참여하여 전력망을 관리하도록 하는 핵심기술이다. 수요반응을 통하여 전력 이용자의 신뢰성을 높이고 전력망에서 대량의peak수요를 줄임으로써 전력요금을 줄일 수 있다. 수요측면에서 산업설비는 많은 양의 전력을 소비하기 때문에 수요반응 에너지 관리의 도입이 시급히 요구되고 있다. 본 연구의 도입부에서는 산업설비용 수요반응 에너지 관리 시스템의 일반적인 모델에 대하여 소개하고 있다. 수요반응 에너지 관리 시스템 모델은 모델요소, 모델구조로 구성되며 산업용 수요반응에 적용된다. State-Task Network(STN) 기반의 모델은 수요반응을 적용하기 위해 공정에 따라 산업설비를Non-Schedulable Tasks(NSTs)와 Schedulable Tasks(STs)로 나눈다. 이렇게 제안된 모델은 산업설비의 수요반응 시스템을 설계하고 분석하는 간단한 방법을 제공하며, 시스템의 표준을 제정하는데 도움을 준다. 그리고 본 연구의 수요반응 모델은 제철공장의 설비를 예로 들어 설계하였다. 모델을기반으로, State-Task Network(STN)과 Mixed Integer Linear Programming (MILP)를 이용한 산업설비용 수요반응 에너지 관리 시스템을 설계하였다. 하루 전의 시간대별 전력요금을 고려할때, 산업시설에서의 수요반응문제는 1차 방정식과 최대 혹은 최소값으로 나타낸다. 전반적인 산업설비의 에너지 비용 최소화를 통해 운영계획은 Schedulable Task와 Distributed Energy Resources (DERs)의 스케줄을 결정한다. 이러한 방식은 전력시스템의 신뢰성을 높이고 산업설비의 전력요금을 줄이기 위해 전반적인 산업설비의 전력수요는 Peak 구간(전력가격이높은구간)에서 Off-peak구간(전력가격이낮은구간)으로 옮겨진다. 마지막으로 수요반응 에너지 관리 방식의 성능은 산소 발전설비에 기초하여 확인하였다.; Smart grid integrates information communication technologies (ICT) with electric grid and enables two-way communication between utility suppliers and electricity consumers. In smart grid, demand response (DR) energy management is a key technology which allows electricity consumers to actively participate in the management of electric grid. DR improves the reliability and reduces electricity cost of consumers by reducing the wholesale peak demand of the electric grid. On the demand side, industrial facilities consume huge amounts of electricity, highlighting the urgent need to implement DR energy management. A general model of DR energy management systems for industrial facilities was firstly introduced in this study. The model consists of model elements, model architecture, and approaches to industrial DR. Based on state-task network (STN), the model divides the processing tasks in industrial facilities into non-schedulable tasks (NSTs) and schedulable tasks (STs) to implement DR. The proposed model provides a straightforward means of designing and analyzing DR systems in industrial facilities and assists in developing standards for such systems. Besides, an example of this DR model was presented with a steel manufacturing facility. Based on the model, a DR energy management scheme for industrial facilities was introduced utilizing the state-task network (STN) and mixed integer linear programming (MILP). Considering day-ahead hourly electricity prices, the DR problem in industrial facilities was formed into a series of linear constraints and one objective function. Through minimizing the energy costs of overall industrial facilities, the scheme determines the scheduling of STs and distributed energy resources (DERs). In this way, the electricity demand of overall industrial facilities is shifted from peak periods (with high electricity prices) to off-peak periods (with low electricity prices), which not only improves the reliability of the electric power system, but also reduces energy costs for industrial facilities. Finally, the performance of the DR energy management scheme was verified based on oxygen generation facilities.