Abstract: In a distributed environment of the smart grid,data-intensive applications often involve complex transmissions between and within the data centers which may have to use large amounts of datasets.An application may need several datasets located in different data centers facing great challenges including the high cost of data movement between data centers and data dependency within the same center.Considering the efficient storage and management of large scale data in the smart grid,a two-stage strategy is proposed for the data placement.In the first stage,an initial classification is achieved by the K-means;while in the second,datasets are placed in different centers by a clustering scheme based on the data dependency.Simulations show that the algorithm can effectively reduce the cost of data movement while performing an even data distribution.