Study On Optimal Operation Of Hydropower Stations Under The Multi-Scale Load Demands Of Power Grid

With the upgarading of power system to adapt to large-scale and high proportion of new energy,the requirement on flexible regulation increases in both power supply side and consumption side.Due to the advantages of abundant resources,fast response and flexible start-stop,effectively and reliably providing regulation capacity has become a significant task for hydropower in the process of power system transformation.However,the optimization of hydropower stations has not been able to completely fulfill the need of practical usage.In the medium and long-term scale,the optimal scheduling mostly operates referring to the maximum benefit,rather than the actual demand of the power grid.In the short-term scale,most of the peak shaving plans are developed with deterministic runoff,which impose a great difference between the actual effect and the expected effect.In the real-time scale,it is common that researchers attach importance to the economic and safety demand of the single response to load regulation but fail to consider the problem of reservoir level deviation caused by frequent regulation demand under the rising new energy penetration rate.Therefore,how to satisfiy the actual regulation needs of the power grid while meeting the operation requirements is an urgent problem for hydropower stations to be solved.In this paper,supported by optimization theory and decision-making method under uncertatinty,the optimal operation models and methods of hydropower stations in response to load demands under the time-scales of "medium and long-term,short-term,and real-time" are deeply studied.The main contents and innovations are as follows:(1)In the medium and long-term scale,the peaking scheduling perspective is used to improve the response ability of hydropower station to the load demands of power grid.Firstly,a Segmentation principle based on load characteristics for whole dispatching period is firstly established to obtain the set of sub-dispatch periods.Then,a calculation method of water level at the end of sub-dispatching period considering the peak shaving preference of dispatchers is proposed.On this basis,an adaptive Segmentation cutting load method including the Segmentation principle,heuristic cutting load method,and full-cycle reconfiguration is constructed.Finally,the proposed method is applied to the Three Gorges Reservoir in China for constructing a generation scheduling in falling stage.The operational results indicates that under the generation scheduling obtained by the proposed method,the hydropower station is able to fully and flexibly respond to the load demand,and reduce the peak and stabilize the load fluctuation,providing strong support for the optimal operation of hydropower stations to meet load demands of power grid.(2)In the short-term scale,the key point is enabling the hydropower station to provide more effective and reliable peak shaving capacity to the power grid.For this purpose,a theoretical model framework of information gap decision-making theory is introduced to quantify the impact of uncertatinty,which also supported the establishment of the shortterm peak shaving dispatching model of hydropower station considering the uncertainty of inflow.Then,the adaptive Segmentation cutting load method and the dichotomy iterative algorithm are used to solve the equivalent converted single-layer model.Further analysis is conducted on the sensitivity of peak shaving effect to the fluctuation range of inflow and the risk attitude of decision-makers.The case study demonstrates that the proposed model can provide important scheduling information such as satisfactory peak shaving scheme and key peak shaving index in the uncertain environment to decision-makers with different risk attitudes.Moreover,the unit robustness peak shaving cost and unit opportuneness peak shaving value curve are formed to provide more reliable and objective decision-making suggestions,which further verified the effectiveness of the proposed model in uncertain scheduling and peak shaving decision-making.(3)In the short-term scale,in order to make cascade hydropower stations provide more peak shaving capacity to the power grid through a safe and stable way,a two-layer nested model for daily generation scheduling based on information gap robustness and inner-plant economic operation is formulated to improve the guidance for the daily generation scheduling.Then,the adaptive Segmentation successive cutting load method is proposed to determine load distribution schemes for each station.At the same time,the multi-plan solution of the dynamic programming method considering the adjustable flexibility is introduced so that the original inner model is modified into a multi-plan model of in-station economic operation.The operational results indicate that the daily generation scheduling considering the adjustable range of unit can reserve the adjustment margin of the unit before the implementation of the plan,and provide suggestions for the unit operation scheme under different water inflow scenarios in the next period,enhance the stability of the unit and the applicability of the daily generation plan under inflow uncertainty.(4)Focusing on the short-term scheduling destruction caused by frequent response to regulation demand of cascade hydropower stations,the quantitative criterions of hydropower flexibility in real-time scale are established.Then,a real-time dispatching strategy that combining real-time and long-term perspectives for hydropower stations considering frequent load fluctuations is proposed.Meanwhile,the adjustment order of each hydropower station is determined by the long-term flexibility,and the output adjustment of each hydropower station is determined by the real-time flexibility.The two are integrated as the real-time decision-making scheme.The results show that the proposed real-time dispatching strategy can not only meet the real-time regulation requirements of power grid,but also effectively reduce the level deviation.Compared with the traditional real-time dispatching method,it is more suitable for the current operation environment,and can provide operation suggestions to effectively coordinate real-time,short-term and medium and long-term generation plans,so as to maximize the response ability of hydropower to the load demands of power grid.