As power systems transition toward higher shares of renewable generation, the displacement of synchronous machines is reducing system controllability and increasing the operational complexity faced by system operators. In this context, maintaining secure and economically efficient operation requires advanced network management methodologies capable of exploiting flexibility-providing resources while explicitly accounting for the uncertainty associated with renewable generation and load variability.

This paper proposes a multi-period Chance-Constrained Security-Constrained Optimal Power Flow (CC-SC-OPF) formulation based on a full AC network model and solved in the nonlinear programming (NLP) domain. Uncertainty in renewable generation and load is represented through probabilistic chance constraints imposed on key security limits. The main contribution of this work is the adoption of differentiable rational surrogate functions to approximate chance-constraint violation indicators. In contrast to conventional mixed-integer formulations, which rely on binary variables to model constraint violations, the proposed reformulation eliminates discrete decision variables and enables the use of efficient NLP solvers while preserving the nonlinear AC power-flow representation. The proposed framework addresses a central challenge in contemporary power-system operation: integrating uncertainty-aware security assessment into decision-making processes with computational performance compatible with operational planning and, potentially, real-time applications. Deterministic OPF formulations may underestimate operational risk, whereas robust formulations may lead to overly conservative solutions. Chance-constrained optimization provides an attractive compromise between these two extremes, but its practical adoption depends critically on the availability of computationally tractable formulations.

The methodology is evaluated on a modified IEEE 5-bus transmission system considering multiple operating scenarios, time periods, and N-1 contingencies. Preliminary results show that the proposed NLP formulation achieves a 97.7% reduction in solution time relative to a benchmark Mixed-Integer Quadratic Programming (MIQP) formulation, while producing similar generation dispatch schedules and objective values. These findings indicate that smooth continuous reformulations based on rational surrogate functions constitute a practical and computationally efficient alternative to mixed-integer chance-constrained SC-OPF models, thereby enabling faster and more scalable uncertainty-aware security assessment in renewable-dominated power systems.