This is the third and final part of an exciting three-part Model Monday post series highlighting RSCAD’s phase domain synchronous machine model and how it can be used for the closed-loop testing of a generator protection relay. Click here to read part one, and here to read part two!
In the past two Model Monday posts, we discussed the phase domain synchronous machine (PDSM) model available in the RSCAD library, and provided an overview of the procedure for setting up a simulation case, assigning signals to the GTAO component, and scaling those signals for input to a generator protection relay under test. This third and final post in the series will describe the process for testing the relay for various faults, and will show results for those tests.
Communication between the user’s workstation computer and the SEL 300G Multifunction Generator Relay is established using the AcSELerator Quickset Software, available from the relay manufacturer. A physical connection must be established between the relay’s serial port and the computer’s serial port (or between Ethernet ports). In the Quickset software, the existing settings of the relay should be read and saved as a backup. Details for setting this particular relay are not included in this blog post, but RSCAD users can find them in the PDF document available in the Tutorial > SAMPLES > Generator Protection and Excitation folder in the File Manager library.
In this example case, we will turn off all possible elements at the outset of testing and activate elements one by one. As each element is activated, a relevant fault is applied, and trip signal and corresponding bits to that elements are observed in order to test the functionality of the relay for that element. Many of the synchronous generator protection elements have overlapping functions, so it may be necessary for the user to set time-delays of certain elements to be unusually small (relative to time-delay settings used in the practical operation of synchronous generators) in order to successfully observe the trip signal for that element.
The results can be viewed using plots gathered from the RSCAD RunTime interface, by observing the physical relay’s front panel, and by viewing data recorded by the relay and reported from within the relay’s Human Machine Interface (HMI), which shows measured instantaneous values and phasor diagrams, the status of relay bits, and a virtual version of the front panel.
Results for testing a stator-ground fault
Many tests can be performed with varying grounding impedances and the fault at varying positions on the winding. Here, we show the results for a ground fault applied in the middle of the phase A winding with the neutral grounded through a low impedance (0.5 Ohm).
With low impedance grounding, the phase differential element (87) is capable of detecting stator-ground faults. The following plot from the RSCAD RunTime interface shows the terminal voltages, machine currents, and neutral voltage and current. In addition to the large flow of current in the stator winding A2, which causes the phase differential element to trip, the neutral also experiences large overvoltage and overcurrent in this case. This causes other protection elements such as neutral overvoltage (64G1) and neutral overcurrent (50N) to be activated as well. This is shown in the photograph of the SEL 300G relay front panel below, after the simulated fault was applied. It is also reflected in the AcSELerator software’s recorded events from the SEL 300G relay, shown below.
Results for testing a loss-of-field excitation fault
Generators are normally operated so they are slightly overexcited – thus, normal stable operation exists in the first quadrant. However, when field excitation is lost, the generator must absorb reactive power and consequently operation exists in the fourth quadrant. This area of operation is unstable and should be avoided. If there is no excitation and the system can sustain the voltage and provide necessary reactive power, the machine will act as an induction generator – otherwise, loss of synchronism will occur. Overheating and potential machine damage can also occur with sufficiently low excitation.
This fault is simulated by switching the Excitation Source switch (mentioned in our first post) to manual – letting the machine receive its excitation voltage from an RSCAD RunTime slider rather than an exciter model. The relay element detects the fault using MHO characteristics and sends the trip signal to the load breaker. The figure below shows the MHO characterists of the relay. Overlaid on this is the trajectory of the imaginary part of the impedance versus the real part, seen from the machine terminals. It can be seen from the front-panel display as shown by the HMI of the SEL 300G relay that element 40 has been activated.
Results for testing a turn-to-turn fault
Although the machine model has only one point of fault on phase A, a turn-to-turn fault can be modeled by connecting the node AJ to the stator terminal of phase A, as shown in the RSCAD Draft module screenshot below. Element 46 of the relay uses the negative sequence to detect turn-to-turn faults. The front-panel display as shown by the HMI of the SEL 300G relay, showing activation of element 46, is shown below.
If you enjoyed this Model Monday, check out the Protection System Testing section of our available Technical Publications page. The list includes many interesting and comprehensive examples of generator and other protection system testing by RTDS Simulator users. If you would like further information on any of these papers, or have a question about this Model Monday post series on generator modelling and protection testing, please contact us at firstname.lastname@example.org.