Wind power in thermal power systems
Licentiate thesis, 2008
Wind power is a key technology in the effort to transform the power system in order to reduce its climate impact. However, the ability of wind power to reduce CO2-emissions depends on the properties of the power system. As large-scale wind power is integrated in a power system (typically 20% wind power grid penetration and beyond), the intermittent nature of wind power will result in an increase in variations in load on the other units in the system (e.g. thermal power plants). There are three main alternatives for the power system to respond to this increase in variations; by an increase in power plant cycling (i.e. starting/stopping), by an increase in part load operation hours or by wind power curtailment. In a system dominated by thermal base load units, the integration of large-scale wind power might thus result in an increase in costs and emissions related to thermal operation and/or an increase in wind power curtailment.
This work investigates consequences of large-scale wind power integration in a thermal power system. Operation of units in a regional thermal power system have been investigated and quantified by means of modelling. Furthermore, the possibility to reduce the influence from wind power variations by means of introducing a variation moderator or demand side management has been evaluated. A variation moderator is a unit with the ability to reallocate power in time. Pumped hydro power stations, compressed air energy storages, batteries and transmission lines providing power exchange are examples of variation moderators which are all investigated in this work. Demand side management will allow for some flexibility in the allocation of parts of the load and in this work evaluates the ability of plug-in hybrid electric vehicles (PHEV:s) to provide such flexibility. The simulations are carried out with a simplified version of the western Denmark power system as a starting point. This, since this system already has a high share of wind power and wind power data is available.
For the system studied, simulations show that for 24% wind power grid penetration, variations in wind power generation results in start-up emissions corresponding to 5% of the total CO2-emissions of the power generation system. It is also shown that the inclusion of start-up and minimum load level aspects has an impact on the dispatch of units in the system. By integration of a moderator in the wind-thermal system, emissions are reduced with 7.2% mainly due to a decrease in power plant cycling. At wind power grid penetration levels of around 20%, a daily balanced moderator is sufficient, whereas more extensive storage capacity is required at higher wind power penetration levels such as for the 40% penetration level investigated in this work.
The ability of PHEV:s to manage wind power variations depends strongly on the choice of PHEV integration strategy. With a strategy where the power system is free to allocate the load in time, emissions from the power system (excluding vehicles) are up to 6% lower than a corresponding system without PHEV:s. A passive approach to PHEV integration results in increased emissions from the system. For the system studied, the PHEV share of the electricity consumption should be around 12% to obtain maximum reduction in emission. This, since this share of PHEV capacity approximately corresponds to the difference in average load between night and day (evens out variations in load profile).