MATHEMATICAL MODELLING OF FINANCIAL MOVIVATED CYBER ATTACK BY FDI AND SECURITY CONSTRAINED OPTIMAL SCHEDULING OF VIRTUAL POWER PLANTS IN ELECTRICITY MARKET

Abstract

Power system is a wide interconnected network of electricity generation, transmission, and distribution systems. With the technical advancement, increase in day to-day demand from industrial, agricultural and residential consumers, and the disadvantages of monopoly system demanded the power sector for disintegration and deregulation. Power system deregulation and integrating communication devices are advantageous for better monitoring and decision making by the system operator. At the same time, it increases the risk of cyber intrusion. In 2003, the eastern United States and Canada had major power system blackout due to the failure of grid. Despite of the fact that the blackout was caused by factors that are other than cyber-attack, many academics believed that similar catastrophe may occur with targeted cyber intrusion. In 2007, Idaho National Lab researchers attempted to attack a synchronous generator. The attempt was successful, and the generator got self-destructed within minutes. This incident alarmed cyber-security decision-makers, prompting them to establish a critical infrastructure that is vulnerable to prevent cyber-attack. The existing bad data detection procedure in state estimation is incapable of detecting a certain sort of cyber-intrusion known as a stealth attack. Stealth attacks can be used to influence state estimate results for financial gain or to cause technical problems for the power system. Unbundling of transmission lines, ensuring social welfare among the power system utilities, promote investment in the electricity sector. The deregulated power system has brought up power market as an efficient tool and has created an enabling environment to accelerate the all-around development of power generation, transmission and distribution systems. The effective monitoring and decision making is achieved with the integration of communication lines and internet network. Cyber-Physical System technology is utilized for more safer and secure grid operations. In this dissertation, financially motivated false data attacks are investigated, by injecting manipulated data into day-ahead and real-time electricity markets operation. For determining the optimal attack vector, it is assumed that the attacker runs a bi-level optimization problem that comprises the attacker's profit maximisation objective and the market clearing problem. While manipulating measurement devices such as RTUs, the attacker needs to take care of being identified by the ISO’s bad data detection (BDD) mechanism. The proposed attacking model is implemented on the PJM-5 bus test system to demonstrates the potential impact of financially motivated cyber intrusions in the x power markets. During the attack, the relationship between market clearing power and LMPs is established. The simulation results are deduced to demonstrate the effect on locational Marginal pricing in achieving the attacker's goal of profit maximization. Secondly, Renewable energy generation has become more prominent in the power sector around the world. Large integration of RE sources into the electricity markets has brought further complexities in the markets. Distributed energy sources (DE) have limited participation in these markets. Considering uncertainties related to RE intermittence nature, and market prices small-scale REs such as wind power, solar PV power, ESSs, and utilities comprising CHPs, DG sets, flexible demands, etc., are aggregated in to single entity in the name of (VPP) and participate in the electricity markets. Hence, it is important to find out an optimal scheduling solution to these VPPs. In this dissertation two-stage stochastic programming approach for optimal scheduling of VPP in the electricity market is presented. The uncertainties are modelled using scenario bounds and are formulated using stochastic programming approach. Simulation results are carried out on 4-hour planning horizon. Since, electricity markets are competitive in nature, each and every market participant tries to maximize their profits through strategic bidding. Keeping in the view, the uncertainties related to RE generation, market prices and reserve deployment requests, VPP also tries to maximize its profit. It is necessary to take strategic decision to counter the other market participants. Therefore, a bi-level model is proposed for finding out optimal scheduling solution in the electricity markets. Uncertainties are modelled using scenario realization technique. VPP maximize its profit by making strategic decision on trading power in the DA and reserve markets. To exercise the power of VPP in altering market decision, the upper-level problem in the bi-level model address the VPP objective to maximize the profits, while lower level addresses the market clearing problem of both DA and reserve markets. The proposed model is then reformulated in to single level MILP problem using KKT optimality conditions and strong duality theorem. Finally, the proposed model is implemented on IEEE-24 reliability test bus system. The results are analysed based on the profit acquired by the VPP with and without flexible demands. The importance of the reserve market in balancing the system is demonstrated through appropriate scenarios, additionally, demand-side flexibility xi smoothens the load curve and connects the generating and demand side curves, allowing the VPPs to achieve the best profit. At the end, impact of strategic and non-strategic decision making on VPP’s profit is also analysed.