The creation of microgrids, the introduction of multiple renewable energy sources, and Internet of things (IoT) devices have led to the increased vulnerability of Smart Grids. The existing Intrusion Detection and Protection Systems (IDPS) designed for Information and Communications Technology (ICT) are generally regarded as inadequate for the protection of Smart Grids mainly because these systems do not include aspects of the power systems and the underlying communication technologies and protocols. Thus, they are not suitable for Smart Grids. This dissertation addresses the above need by introducing and implementing an IDPS for a Smart Grid that would incorporate some aspects of communication protocols that apply to the power system.First, we have comprehensively surveyed publications that address IDPS in Smart Grids. We discovered that even though some IDPS for Smart Grid ideas have been suggested, the industry has not implemented them. Second, we propose a cooperative distributed IDPS that can detect and isolate attacked microgrids from the rest of the Grid. Furthermore, we develop an algorithm to detect intrusions that can be applied in our IDPS. The algorithm helps isolate compromised microgrids from the rest of the Grid and readmit cured microgrids back to the Grid. Third, we introduce a performance measure for IDPS, "Detectability, "and implement a novel risk analysis of IDPS for Smart Grids using Ladder Logic and Fault Tree Analysis (FTA). Method analysis and simulation results show that this method can be effective in evaluating specific failure contributors in IDPS.