Research On Optimization And Cost Allocation Model Of Ancillary Services For Power Systems Integrated With High Share Of Renewable Energy

It is a consensus of the international community to vigorously develop renewable energy,mainly wind and solar energy,and promote green and low-carbon energy transition to achieve sustainable development of energy and economy.The integration of a high proportion of renewable energy into power grid will be one of the basic characteristics of future power systems.However,due to the inherently strong variability and uncertainty of wind and solar energy,the large-scale renewable energy integration increases the demand for peak regulation and reserve ancillary services,and also brings challenges to the operation and management of ancillary service markets.The traditional deterministic reserve scheduling methods are unable to meet the actual demand and it is essential to optimize the operating reserve schedule suitable for power systems with a high share of renewable energy.In addition,the increasing share of renewable energy in power system s increases the ancillary service costs,which should be properly allocated among the market entities that cause the need for these services.Under this background,this thesis conducts theoretical research on the reserve scheduling problem from the perspectives of ‘multiple uncertainty factors’ and‘multiple reserve resources’,and the cost allocation problem of reserve and peak regulation ancillary services from the perspective of ‘multiple market entities’.The main work of this thesis is summarized as follows:In terms of ‘multiple uncertainty factors’,a probabilistic reserve scheduling model is proposed considering multiple uncertainty factors,such as unscheduled generator outages,load forecast errors,and wind and photovoltaic power forecast errors.The scenario analysis method is adopted to calculate the expected energy not served(EENS)and expected energy curtailment(EEC)by combining the probability distributions of various uncertainty factors and the available reserve capacity of the system.Then,a stochastic security-constrained unit commitment model is established by incorporating the EENS and EEC as penalties.The formulations of EENS and EEC are further simplified by introducing relaxation variables,which improve the computational performance of the security-constrained unit commitment.Compared with the existing deterministic reserve scheduling methods,the proposed method can reduce the expected cost of the system and improve the economic efficiency.In terms of ‘multiple reserve resources’,a risk-averse reserve scheduling method considering wind farm reserve and demand-side response is proposed.Firstly,the reserves provided by conventional units,demand side resources,and wind farms are modeled,which are fully utilized to improve the operational flexibility of the power system.The uncertainty of wind power output is taken into account when modeling the reserve of wind farms.Then,the conditional value at risk is used as the risk measure,and a two-stage risk-averse reserve scheduling model is established.The conditional value at risk is adopted to manage the uncertainty of wind power output and ensure the safe operation of the system in scenarios with small probabilities and high risk.Finally,the influence of risk management and coordinated scheduling of multiple reserve resources on the reliability and economy of power system operation is analyzed.In terms of the uncertainty of ‘multiple market entities’,a reserve cost allocation model based on cooperative game is proposed.Some basic elements of cooperative game theory and the features of Shapley value are introduced.The influence of forecast errors of load,wind power,and photovoltaic power on reserve capacity costs is analyzed.It is pointed out that the ‘smoothing effect’ of forecast errors among load,wind power,and photovoltaic power provides a physical condition for the application of cooperative game theory in allocating reserve costs.Then,a chance-constrained reserve scheduling model is established,based on which a cost characteristic function that reflects the influence of the alliance’s uncertainty on the system operating cost is constructed.Finally,the Shapley value method is used to quantify the increased operating costs incurred by the uncertainties of the market entities,and the reserve capacity costs are allocated to the market entities based on the quantified operating costs.The effectiveness of the proposed reserve cost allocation method is verified by numerical examples,and the incentive effect of the method on reducing the forecast errors of load,wind power,and photovoltaic power is analyzed.In terms of the variability of ‘multiple market entities’,a peak regulation cost allocation model based on Shapley value framework is proposed.Firstly,the peak regulation mechanism of power systems is analyzed,and the market entities which cause the need for peak regulation service are identified.The alternative scenarios are constructed where load curves and the output curves of renewable energy are converted into flat lines.Then,an optimal scheduling model considering deep peak regulation and pumped storage is established to calculate the peak regulation costs in different scenarios.Finally,the peak regulation cost incurred by each market entity is calculated by the Shapley value.The paid peak regulation costs are allocated to the market entities in proportion to their respective contributions to the costs.Case studies based on the actual power system data of a province show that the proposed model can reflect the costs or values of peak regulation service contributed by different market entities and transmit the cost signal of peak regulation service to the market entities incurring the costs.