Abstract
The short term scheduling of a national energy production for one week time horizon is considered. The problem is to optimize the use of various kinds of power plants such as nuclear power, conventional condensing power, counter pressure power and water power. Energy storages are also included in the optimization.
The problem has been decomposed into controllable hydro power and other energy sources. The discharges of each hydro power plant are optimized using modified LQP approach. Every river bed between seasonal reservoirs is optimized separately. The other energy sources and storages are modelled as an LP/MIP model, which determines the optimal on-off states and power output for each plant, assuming a predetermined level of hydro power production. The two models are solved in turns and they are coordinated by an algorithm based on marginal costs for different kinds of power plants and the power output of each plant.
As an example a system is solved with and without an energy storage system (ES). When an energy storage is utilized, the production costs for the system are about five per cent less than for a system excluding an energy storage. This is due to fewer start-ups needed and more efficient use of plants with small production costs.
| Original language | English |
|---|---|
| Pages (from-to) | 757-766 |
| Number of pages | 10 |
| Journal | Lecture Notes in Control and Information Sciences |
| Volume | 180 |
| Publication status | Published - 1992 |
| MoE publication type | A1 Journal article-refereed |