Abstarct: Abstract This thesis focuses on modeling the risk of deception attacks in cyber-physical power systems. The proposed attack is designed using a bi-level optimization model, where the attacker aims to deceive the operator and induce overloading beyond the capacity of critical network lines. The model is first transformed into a single-level linear optimization problem using KKT conditions and the Big-M method. Subsequently, risk assessment of the attack is conducted specifically on the standard IEEE 57-bus system. This simulation, employing the Monte Carlo method, is performed under 84 different uncertainty scenarios related to both the structural and physical parameters of the network, resulting in a total of 84,000 attack scenarios. Furthermore, the proposed attack model is comprehensively validated across five standard networks, including 5-bus, 14-bus, 30-bus, 57-bus, and 118-bus systems. All simulations and analyses are implemented in the MATLAB environment. The results of this study demonstrate that a deception attack, with minimal available information and minor changes to load measurements, can accurately target the network's vulnerable points. However, the level of information inaccuracy significantly impacts the success and precision of the attack. Specifically, an increase in information inaccuracy reduces the attack's accuracy while escalating the attacker's costs. The analysis identifies vulnerable buses and lines, such as buses 8, 1, 57, 37, and 9, and lines 8, 18, 61, 47, and 15, which require special protective and security measures. Beyond highlighting the critical vulnerabilities of the power network, this thesis demonstrates that the proposed attack model is capable of simulating complex cyber threats, providing valuable insights into the cybersecurity risks associated with cyber-physical power systems.