A Variant of Dynamic Programming Algorithm for Unit Commitment Optimization in Combined Heat and Power Production Planning

Combined heat and power (CHP) production is an increasingly important power production technology for efficient utilization of primary energy resources and for reducing CO2 emissions. In CHP plants, generation of heat and power follows joint characteristics, which means that production planning must be done in coordination. A CHP system may also comprise separate heat and power components to make the system more flexible and reliable under the variable heat and power demands. The possibility to buy and sell electric power on the spot market is also considered. In this paper, we address the unit commitment in multi-period CHP production in a general sense. We present a variant of the dynamic programming algorithm for solving the problem based on linear relaxation of the ON/OFF states of the plants and sequential commitment of units one by one. The time complexity of the algorithm is O(|U|×T)O(ED), where |U| is the number of plants, T is the number of hours in the planning horizon and O(ED) is the time complexity of solving hourly economic dispatch (ED) problem. The numerical results with realistic power plants show that the algorithm is not only slightly more accurate (0.08-0.5%) than the unit decommitment (UD) algorithm but much more efficient (3-5 times faster on the average).