The growing adoption of electric vehicles has recently attracted increasing attention of scientists and brought about pioneering research studies from various fields. Safety concerns particularly those regarding ground faults detection and protection, have extensively been addressed. Ground faults occur quite frequently in electric vehicles and they may be due to severe operation conditions, such as vibrations, twists or even crashes. Generally, the first ground fault is not dangerous, since the powertrain systems, namely, the DC bus where the batteries are connected, the power inverter and one or more AC machines, are generally ungrounded. The second ground fault, however, can produce malfunction in some systems, power loss or even serious damages. Locating the fault has often proved hard and time-consuming. For this reason, the present study focuses on developing a ground-fault detection method for electric vehicles capable of determining on which side, the DC or the AC, the ground fault is located. The method is based on the analysis of the voltage in a grounding resistor connected between the midpoint of the battery pack and ground. Based on the polarity and harmonics, it is possible to locate the ground fault. This method has been verified excellent results have been achieved using computer simulations and experimental tests in a 140-kW electronic power inverter fed by a 480 Vdc battery.

The use of variable speed drives made essential the presence of power electronic in electrical systems. The ac to dc and dc to ac conversions, thanks to rectifiers and inverters cause difficulties in the ground fault detection. In this article, a new method to detect ground faults in variable speed drives is presented. It is based on the analysis of the terminal voltage signal of a grounding resistor connected between the neutral at the secondary of the main power transformer and ground. Attending to the waveform, the ground fault can be detected on the ac grid side, dc positive or negative terminals, or ac inverter side. To verify the method, numerous simulations and experimental tests in a 140-kW power converter were carried out achieving satisfactory results.

Condition monitoring of large power transformers is of paramount importance in power systems. One of the most common indicators of transformer aging is moisture content in oil. In this article, a capacitive sensor able to online measuring moisture content is presented. The output of this sensor is recorded using a Frequency Response Analysis (FRA) device. Comparing the FRA results of a degraded oil with a reference FRA test, moisture content can be determined. A prototype sensor has been designed, built and tested. The performance of this sensor was simulated and laboratory tested with different moisture content oil samples. The results confirm that the proposed sensor can determine moisture content in oil. Therefore, it allows estimating oil degradation.

Frequency Response Analysis (FRA) is an extended technique for transformer condition assessment. In recent times, its proven sensitivity for detecting different type of faults have generated interest for rotating machine applications. However, at present the interpretation of FRA results is mostly made by visual inspection of curves and is a hard task for non-expert maintenance personnel. This paper presents a mathematical model to simplify diagnostic tasks from FRA results. The developed model is called Fault Diagram and it is conceptualized to be applied to any coil, so it could be extensive to any type of machine. The fault diagram is based on a paradigm which states that a healthy machine is a machine with infinite impedance fault. The model uses a first order frequency response for the purpose of fitting a linear function R3. The input of the model is the FRA curve of the machine, and the output is a point in a 3D space. This approach can give an accurate diagnosis without ambiguous interpretations about if the machine is healthy or not and can provide an estimation of the fault location in case of damage. For validating the method, it was applied in a pole of a 40 MVA synchronous generator.

Synchronous generators are key pieces in any power system. Therefore, it is essential to have effective protection and diagnosis techniques for these machines. Ground faults are the most common issue in rotor windings and several damages can be originated if a second fault occurs, consequently, the detection of the first one is crucial. After detecting the existence of a ground fault, its location would be helpful to reduce the time needed to repair the machine. This paper presents a new ground fault location method in field windings of synchronous machines based on the Frequency Response Analysis (FRA). The method provides an estimation, not only of the location of the faulty turn but also of the fault resistance. The method is based on the utilization of the resonance frequency as a distinctive parameter. The construction and fitting of some curves, where resonance frequency against fault resistance for each turn are collected, allows the fault location. This technique has been developed and successfully corroborated carrying out numerous experimental tests in a pole of a large hydro synchronous generator. Applying six different fitting techniques to reduce errors from fault curves, PCHIP and Modified Akima methods provide better results than the others tested.

High speed trains supply is, nowadays, performed by 2 × 25 kV power systems. If a ground fault occurs, its location and correction are crucial to minimize downtime. One of the methods for locating ground faults in power systems is the impedance method. This was a good solution for 1 × 25 kV railways power systems, as the ratio between reactance and fault distance is linear. However, the use of autotransformers between the catenary, feeder and rails makes this ratio non-linear in 2 × 25 kV configuration. This work presents a novel ground fault location method for 2 × 25 kV power systems. The new method is based on the ratio of the currents of the autotransformers and the substation transformer. The main contribution, compared to previous methods developed, is that this new method is insensitive to the fault current variations produced by the network voltage fluctuations or by the main transformer tap changer position, among others for 2 × 25 kV power systems. The novel approach has been corroborated by numerous experimental tests in the laboratory and computer simulations, with satisfactory results.