Preserving Privacy of AC Optimal Power Flow Models in Multi-Party Electric Grids

Abstract

The electric power grid is owned and operated by many competing providers who do not want to reveal their sensitive information to other providers. This competitive environment motivates the use of a third-party (cloud-computing) platform that combines information from each provider and jointly optimizes resources in the entire grid. Since the power system structure and component values are confidential, the third-party computing platform must be either absolutely secure or information needs to be masked before sharing with the third party. In this paper, we propose an approach to obfuscate the sensitive information and solve the multi-party AC optimal power flow problem in a shared computing platform. The initial obfuscation is separately performed by each party so that sensitive information is masked from the third party and all other participants.