This thesis considers the operation of corrective System Protection Schemes (SPS) in electricity networks. Such systems have the potential to greatly enhance the utilisation of the network, but malfunctions may expose the system to blackouts. Choosing the optimal operational strategy is therefore subject to a cost-benefit trade-off between the benefits of increased system utilisation and the risk of outages. This thesis proposes to formulate the problem in three different phases, namely steady-state, SPS action and response and impact assessment. This is used to formalise the objective of the network operator from a system perspective: minimising the operational costs considering unreliable SPS operation. This thesis also presents a generic modelling framework for SPS failure modes and related impact assessments. From an optimisation perspective, the problem presents two major difficulties. First, the computation of the load-shedding impact may require the analysis of a complex dynamical system. The second barrier is that the set of possible SPS outcomes is potentially large, and depends on the decision variables (generator dispatch and SPS configuration). The problem is firstly addressed in a simple system which serves to illustrate the salient properties of the problem at hand. It is demonstrated that the optimal SPS configurations are always part of a finite set of 'candidate solutions' that are associated with the provision of security against specific outcomes scenarios of the SPS (including its failure modes). The same network is used to investigate SPS operational benefits and risks in a year-round basis. Next, this thesis proposes a heuristic method to address the two complexities described above. The method is tested on the RTS 24-Bus IEEE System. Impact assessments are performed in a basic cascading simulator which provides a range of possible impacts from SPS malfunction scenarios. The method is shown to provide an optimised balance between benefits and risks.