Putaohua sand reservoir with medium-low porosity and medium-low permeability is heterogeneity in nature at D404 block of Xinzhan oilfield, Songliao Basin. By integrating core test with petrophysics logs data to classify flow units and to built the accurate permeability prediction model for identifying oil and water formation and interpreting the degree of water flooded layers. Putaohua reservoir would be divided into four distinct sub flow units on the principles of the flow unit classification, therefore the relationship between porosity and permeability is significantly improved, and different flow units are identified based on core data and wire-logging. The multiple-statistic equation of the Flow Zone Indicator (FZI) is built up by screening compensation density log, deep induction resistivity log, and middle induction resistivity log which are able to best characterize the accumulation property of reservoir, and an identification model of flow unit is obtained. The porosity permeability models of each flow unit are established by using core analysis data, which obtain effective result in Putaohua reservoir of D404 Block, and permeability prediction accuracy is improved. The workflow methodology could be used in other sandstone formation.