PROTECTION SCHEMES FOR STATOR INTERNAL FAULTS IN SYNCHRONOUS GENERATOR

Abstract

A synchronous generator, being a large rotating machine and the heart of a power system network, is one of the most critical and complex equipment in the system. It is subjected to hazardous operating conditions compared to other power system equipment. As the capacity of generators has increased massively in recent times, the loss of a large generator interrupts the power supply, overloading the rest of the system. Nuisance tripping of a large generator has an equivalent catastrophic outcome to a substantial number of concurrent faults, potentially leading to widespread instability in the power grid and enormous economic loss. Further, the time duration and cost for maintenance/replacement of the parts of the synchronous generator are too high and may be in the order of several weeks. Though the circumstances causing generator failure are unpredictable, several studies have identified stator winding faults as the generator's most common type of fault. A stator fault in the generator may cause severe damage to the winding and core, as well as the connected power system. Due to the complex configuration of the generator in a substation and extensive connections with different elements, its protection is very critical as clearance of stator fault would require tripping all the breakers connected to each section of the generator in the generating substation. Some state-of-the-art numeric protection techniques based on differential, transverse differential, and harmonics have been established to detect phase, inter-turn, and ground faults in the SG. In addition, researchers have proposed several strategies to enhance the existing methods for detecting stator faults. However, sensitivity issues due to CT errors, depending on winding configuration and loading and source of harmonics, are the major limitations of these schemes. Nevertheless, it has always been challenging to address all internal stator faults with one protection algorithm. Moreover, sensitivity and selectivity during CT saturation phenomenon are still major challenges of the several existing techniques. The protection of the generator requires faster and more reliable operation of the relay for an internal stator fault. Therefore, it is crucial to develop a new combined generator protection scheme that not only yields enhanced sensitivity to all internal faults to prevent hazards to the system but, at the same time, also prevents unwanted tripping of the generator with better stability characteristics during heavy through faults. Hence, due to various issues, synchronous generators demand a fast and reliable protection scheme for hassle-free operations. The work presented in this thesis concentrates on developing improved protection algorithms using both numerical and statistical analysis based on advanced signal processing that significantly improves accuracy, reliability, and speed, which are the utmost requirements for the protection of synchronous generators. In the first approach, a new internal fault detection and classification technique for synchronous generators has been developed. The internal fault and abnormal/external-fault conditions are distinguished by utilizing the instantaneous phase angle between the Negative Sequence Component (NSC) of the terminal voltage and current. The suggested technique can accurately detect all internal stator defects, including inter-turn faults, and is not affected by external faults, even in the presence of CT saturation during heavy through-fault conditions. The effectiveness of the suggested approach has been validated using a model of a Phase Domain Synchronous Machine that was developed in the Real Time Digital Simulator (RTDS®). As it relies on the negative sequence component, its inability to identify internal balanced three-phase faults and faulty phase identification are the limiting factors. To avoid the limitations of the first approach, a new stator fault detection and classification scheme for SGs, based on an amplified discrete Teager-kaiser energy operator (ADTKEO) of a new differential component (DCSI), is introduced. The detection of an internal fault relies on the satisfaction of the ADTKEO (ξ) with respect to a predetermined threshold, as well as the examination of the second-harmonic content in the secondary current of the current transformer (CT). Subsequently, the classification and faulty phase identification for internal fault is carried out using the parameter ξ. The performance of the proposed scheme is tested on various fault cases obtained from a simulation network modeled in RTDS and validated on the developed laboratory prototype. The results indicate that the proposed scheme not only provides 100% winding coverage against phase-to-ground faults but is also capable of detecting other phase faults within a quarter cycle for low-resistance grounded SGs except interturn faults. To overcome the above limitations, a protection scheme based on the calculation of ‘d-q’ components of stator currents at neutral and terminal ends of the SG is presented in the next chapter. Detection of internal faults is achieved based on the rate of change of ‘q’ component of the stator current at the neutral end compared to a pre-defined threshold. Thereafter, discrimination between internal faults and external conditions is obtained by taking the difference of rate of change of ‘q’ component at the terminal and neutral ends. Subsequently, internal fault classification is carried out based on the rate of change of zero component of the neutral end current. The proposed scheme is tested and validated using various fault cases obtained from a simulation network developed in RTDS and from an experimental setup of the SG, respectively. It can detect all types of internal faults, including inter-turn faults, regardless of the generator winding configuration. It is also capable of distinguishing internal faults from external conditions within a quarter cycle, independent of current transformer errors and saturation, as well as variable loading conditions and SG rating. The presented research is expected to provide a substantial contribution to the detection and differentiation of stator faults with enhanced sensitivity and without compromising stability. The fast fault clearing time has also ensured along with reliability. At the same time, the protection scheme can classify the internal faults. Moreover, this work also emphasized better relay stability during external faults, particularly during heavy through faults, and accurate discrimination between internal and external faults, considering the CT saturation phenomenon and mismatches. Furthermore, the possible on-field implementation of the proposed schemes has also been discussed.