AbstractTo monitor electrical activity throughout the power grid and mitigate outages, sensors known as phasor measurement units can installed. Due to implementation costs, it is desirable to minimize the number of sensors deployed while ensuring that the grid can be effectively monitored. This optimization problem motivates the graph theoretic power dominating set problem. In this paper, we propose a method for computing minimum power dominating sets via a set cover IP formulation and a novel constraint generation procedure. The set cover problem's constraints correspond to neighborhoods of zero forcing forts; we study their structural properties and show they can be separated with delayed row generation. In addition, we offer several computation enhancements which be be applied to our methodology as well as existing methods. The proposed and existing methods are evaluated in several computational experiments. In many of the larger test instances considered, the proposed method exhibits an order of magnitude runtime performance improvement.