Preserving prosumer privacy in a district level smart grid

This study presents the anonymization of consumer data in a district-level smart grid using the k-anonymity approach. The data utilized in this study covers the demographic information and associated energy consumption of consumers. The anonymization process is implemented at the prosumer level, considering their importance in sharing flexibility and distributed generation at the low voltage grid, and the fact that they need to interact with each other and the grid while keeping their data private. The proposed approach is tested under three anonymization scenarios: prosecutor, journalist and marketer. The smart grid data are investigated mostly under the prosecutor scenario with three risk levels: lowest, medium and highest. The results of the k-anonymity approach are compared to k-map and k-map + k-anonymity. No difference has been found between the three investigated approaches for the selected data set. Since, the aim of the k-anonymity is to not transform the information about any individual record among those k-1 individual, the recorded type and the number of attributes play a key role for the anonymization process. One of the risk is the using continuous attributes in the anonymization process which may cause the information lose in the anonymization process such as near real time energy consumptions. Hence we have focused on to anonymization of the consumers' demographic information, rather than their energy consumption.