STABILITY ENHANCEMENT OF POWER SYSTEM UNDER CYBER ATTACK
This study addresses the reliability of power systems by examining their own importance and the effect of cyber-attack on power systems during sudden load shifts in frequency disturbing aspects. Using the continuous power flow approach with static var compensator (SVC) and static synchronous compensator (STATCOM), the maximum load capacity of the load buses is calculated. The result shows that the FACTS devices' responsive power support depends on the FACTS devices being properly placed in the network. The stable limit of speed control for load frequency control (LFC) and integral controller gain for automatic generation control (AGC) is extracted from their characteristic equations to evaluate the effect of cyber-attack on power systems. Simulations are conducted to demonstrate the frequency variations and oscillations of the power system, depending on the nature of the cyber-attack (positive biased or negative biased attack). Finally, to eliminate these oscillations, a feedback LFC block with a three-input shift is proposed.
The alternator frequency varies with a power station's load switch. The frequency sensor detects the system frequency and the LFC sets the primary mover speed to compensate for the system frequency according to the signal from the frequency sensor. In this study, the vulnerable quantity is known as speed regulation. Because of cyber-attack, malfunctioning of the governor speed regulator makes the prime mover's speed out of reach and results for unstable device rate.
Appropriate selection of integral controller gain (KI) is essential for the proper functioning of AGC. Improper choice of the KI value results in a governor failure in setting the correct point to restore the frequency of the device. This research describes KI as the other weak cyber-attack quantity. Any KI change due to unauthorized access to the AGC loop may cause system frequency oscillation that disturbs the stability of the system. Cyber-attack is divided into two forms in this work, one is positively biased, and the other is negatively biased, depending on the affected value of speed regulation (R) and integral controller gain (KI). First, the control system is believed not to be targeted by unauthorized individuals. In this condition, if a sudden change under load (increase in load) occurs, the device frequency will drop for a brief instant below the nominal frequency. The frequency sensor detects the device frequency drop and proper speed control sends signals to the LFC. The LFC sets the governor speed control to compensate for the prime mover speed and regulate the system frequency according to the signals from the frequency sensor. It will only be necessary for the governor's speed limit to be properly set. Due to cyber-attack, LFC may not be able to set the regulation properly and if this occurs then the system frequency will oscillate and make the system unstable. Using the Routh-Hurwitz series, the stable speed limit (R) can be obtained from the characteristic equation of the LFC loop.
This study considers an insulated power plant with the following parameters at a nominal frequency of 50Hz with 250MW turbine output power and a sudden change in a load of 50MW . From the above formula, the stability limit of speed control is obtained as R > 0.0133 using the power system parameters of the Table 5.1and Routh-Hurwitz set. From Fig. 1(a), If R=0.05, the frequency variance of the device is stable. Therefore, the fixed value of R should be 0.05 for the balanced operation of this power station. Any deviation from this fixed value results in the unstable variance of the device frequency. The uncertainty of system frequency deviation renders the governor unable to account for the difference in frequency i.e. In the case of a sudden change in load, the device frequency will not be restored. The deviation from the defined value of speed regulation is due to the cyber attack on LFC (Unauthorized access to LFC control).
Based on the governor's speed control, a shift in the system load can result in a steady-state frequency variance with the primary LFC loop. The primary LFC loop takes a considerable amount of time to restore unallowable device frequency. A modification is needed for LFC to reduce the frequency deviation to zero. By adding an integral controller to operate on the load preference setting to adjust the speed setting level, the adjustment can be accomplished. The integral controller increases the type of system through 1 which forces the final deviation of frequency to zero. The LFC system is the Automatic Generation Control (AGC) with the addition of a secondary circuit (integral controller). It is important to change the integral controller to gain KI for a satisfactory transient response. Adding an integral controller in parallel with LFC enables the governor to set acceptable points to increase turbine speed in the event of a sudden increase in load and to restore the system frequency faster than the primary LFC circuit.
Appropriate KI value is determined using the equation of AGC loop characteristics from which the function of the closed-loop transfer is obtained. Unauthorized access to AGC control can result in deviation of the system frequency and sometimes render the deviation in nature oscillating. If the system frequency deviation is unpredictable, the system frequency will not be restored by the president. Moreover, the attack on AGC may cause unwanted delay to restore system frequency. The frequency difference oscillation contributes to system frequency instability. Because of this oscillation, the governor will not set the right point to restore the frequency of the device. The oscillation deviation in the green curve is higher, suggesting a serious attack on AGC and the governor will no longer be able to restore the frequency of the system. The more the positive biased attack occurs, the greater the effect on the intensity of the program would appear. The frequency of the AGC attack depends on the type of attack. A negative biased attack means that the value of KI decreases from the set value due to unauthorized access to the AGC system. This attack may be counter to the objectives of using the integral controller with the primary LFC loop to obtain an AGC system to reduce to zero the frequency deviation. To increase the frequency deviation to zero, the primary LFC control loop is adjusted to form AGC. It is clear that frequency is deviated from nominal frequency due to a negative biased attack, which leads to unwanted delays in restoring the frequency of the system. A negative biased attack on AGC is not as severe as it is positive because it does not oscillate in essence the frequency deviation .
Figure-1. Frequency deviation step response for different values of governor speed regulation
Figure-2. Frequency deviation step response for LFC under positively biased attack condition
- M.K. Jalboub, A.M. Ihbal, H.S. Rajamtani, and R. A. Abd-Alhameed; Determination of static voltage stability margin of the power system prior to voltage collapse. 8th International Multi-Conference on Systems, Signals and Devices (SSD) 2011, 1, 1-6, 10.1109/ssd.2011.5981420.
- Hassan, M. (2016). Damping Performance Enhancement of a Power System by STATCOM and its Frequency Stability Consideration (Doctoral dissertation, Khulna University of Engineering & Technology (KUET), Khulna, Bangladesh.).
- P. Kundur, N. J. Balu, and M. G. Lauby. Power System Stability and Control; McGrawhill: New York, 1994; pp. 1-200.
- K. Prasertwong, N. Mithulananthan, and D. Thakur; Understanding low-frequency oscillation in power systems. International Journal of Electrical Engineering Education 2010, 47, 248-262, https://doi.org/10.7227/IJEEE.47.3.2.