Electric power is one of the important infrastructures for our life. The electric power system is going to be changed dramatically by clean energy, which includes unstable power supplies such as photovoltaics and wind-power generation, and by introduction of distributed power generation. In order to maintain a balance between supply and demand, a mechanism called demand response, which manages user consumption of electricity depending on supply conditions at peak time, becomes the focus of attention. In this paper, we conducted research on the demand response. Since the number of electric-power users is immense, the management system will require large computation time to account for every user's request. In order to flexibly perform the demand response, we propose a method to break a large-scale problem into smaller-scaler problems to reduce computation time. We solve these small problems using a system with hierarchical structure for reducing computation time and for overall optimization. In this paper, we adopt and creatively use the reservation trade method and a brown-out optimization method for shift and reduction of demand, and we conduct experiments using computer simulations to confirm the effectiveness of our approach. We show that the proposed system can reduce computation time for a large-scale demand response.