Coordinated And Optimal Operation For Data Center Energy System Based On Distributed Renewable Energy Generation

With the accelerated development of digitization in recent years,China’s data center industry has been rapidly grown.It causes a significant growth of energy consumption and carbon emission.In addition,under the background of Carbon Peak and Carbon Neutrality,the government give directions for the operators to achieve high efficiency and low emission for the energy system in data centers.It is notable that considering the integration of renewable energy generation could reduce carbon emissions as well as the dependence on the power grid.Therefore,it is an inevitable trend to integrate distributed renewable energy generation to the energy system in data centers.However,the uncertainty and randomness of renewable energy would lead to negative impact on the operating stability of the data center.On this condition,considering the coordinated and optimal operation of the data center energy system,realizing the power demand response and improving the utilization of renewable energy,has important significance for the low-carbon and low-cost operation of the data centers.Therefore,taking the distributed renewable energy integrated data center energy system as the research object,this thesis established an effective demand response interaction mechanism between data center and power market.On this basis,single and multi-objective optimization methods are adopted to study the coordinated and optimal operation of electricity-thermal-load in the data center integrated microgrid considering waste heat recovery and renewable energy.The main research work of this thesis is listed as follows:(1)Firstly,the demand response interaction mechanism between data center and power market is established,and the optimal operation for the data center microgrid under the demand response market is studied.Specifically,the data center as well as other load aggregators submit the bidding information according to the designed mechanism.Then,with the bidding information from both demand and generation sides,the power market operator clears the market through the two-stage risk averse market clearing model,and issues the cleared time-of-use price.Finally,the data center conducts the optimal operation of the energy system based on the cleared time-of-use prices.(2)Secondly,under the background of demand response power market,the data center energy system is deeply modeled and optimized considering the system uncertainties.On the one hand,the delay-tolerable IT workload is characterized by the arrival-departure curve firstly.Then,the energy consumption management of the data center is optimized through scheduling the delay-tolerable IT workload.On the other hand,the thermal-electrical coupled data center integrated microgrid is established based on the waste heat recovery system.Finally,a two-stage stochastic programming model is established to realize the coordinated and optimal operation of the data center integrated microgrid.(3)Finally,considering the multiple evaluating targets of data center operation,a multicritera optimal operation framework is established for the data center integrated microgrid.Specifically,a multi-criteria optimization problem is established based on the operation cost,quality of service,carbon emission and stability of the data center firstly.Then,an improved augmented ε-constraint algorithm is proposed to solve the multi-criteria optimization problem and obtain the Pareto solution set.On this basis,the compromise relationship of the criteria are analyzed,Finally,a balanced decision-making method which can consider all criteria is proposed to obtain the final solution.It provides references for the decisionmaking of data center operators.