Smart grid promises a reliable and secure electricity infrastructure to meet the future demand growth. However, the increase of data types and data amount from advanced smart grid introduce new privacy issues, which have to be resolved for customers. This paper presents a cooperative state estimation technique that protects the privacy of users' daily activities. By exploiting the kernel of an electric grid configuration matrix, we develop an error free state estimation technique that can hide the behavioral information of users effectively. The proposed scheme can obfuscate the privacy-prone data without compromising the performance of state estimation. We evaluate our obfuscation scheme using data from 1349 meters in 5 IEEE Electric Test Bus Systems. Our simulation results demonstrate high level of illegibility and resilience of our scheme with an affordable communication overhead. © 2011 IEEE.