What measurements are performed by the industry, for performing fault analysis on an electric power grid?
- Bus Voltages (Bus Nodes?)
- Line Currents (On transmission Lines connecting Buses?)
- Active & Reactive Power (At the Generating Units? & Loads?)
- Power Factor (At the Generators?)
- Electric Frequency (Not quite sure at the point of measurement, my guess would be at the generating units and at the load end?)
Your contribution to this thread, will be extremely helpful. Thank you
Main are,type of fault and reacetances ( X+,X- and Xzero) offered to fault, with the five point you said, could refer my paper "Microprocessor based fault type detection and load balance control using symmetrical components"......
Depends what you mean by fault analysis.
If you mean protection tripping, then see the manufacturers documenttion e.g. https://www.schneider-electric.com/en/product-range-presentation/935-easergy-sepam-series-80/
The raw measurements are voltage and current.
Frequency, and power are calculated from V and I by the relays or FPIs.
If you mean fault location, then e.g. see Article An Overview to Fault Location Methods in Distribution System...
Thank you so much Bhupendra Desai for your response. I do follow you on the type of faults as mentioned in your paper (LG,LLG,LLLG, LLL) however what do we mean by +ve and -ve sequence reactances?
Tom Berry Thank you so much for your detailed response, both the documentation and the article have been extremely helpful.
What I mean by fault analysis is the type and location of the fault for a transmission level system rated at 500kV.
One follow up question, for raw measurements of voltage & current are they measured as AC waveforms or RMS values of the wave-forms for the analysis? Also please specify why is this so?
The protection relays, usually receive inputs from CT and VTs. They use voltage and current phasors, meaning the amplitude and phases of fault current and voltage for overcurrent, directional overcurrent, distance, differential, REF,... protection and fault analysis. In some cases like the frequency relays, other parameters like frequency is also used.
Mani Ashouri what about the concept of voltage sag for the case of fault detection scenario? Wouldn`t for this case we will measure the node/bus voltage and see if it goes below a certain threshold, e.g if a fault is introduced on a bus the voltage at that point will go down to 0 for the duration of the fault.
And for such a scenario use the RMS value for the analysis?
If you mean the event recorders in numeric relays, i believe they have a predifined setting input for the rated voltage and for recording events in case of transient they record the instant voltage with the time step of 1 ms (areve/alstom Micom relays)
Re: voltage & current measurements. Digital relays sample the waveforms in order to get the fastest response possible. As fault currents are distorted, the samples are filtered and transformed to avoid unnecessary tripping on transient spikes. Typically relays use frequency tracking and Fourier transforms to determine the fundamental components of V,I per phase every half-cycle. These fundamental components are then used in the protection algorithms. In some cases, special algorithms will detect a fault within 4-5ms.
Re: sampling rate: modern digital protection relays should follow IEC 61869-9. This specifies 4800 Hz as the preferred rate for general measuring and protective accuracy classes, regardless of the power system frequency.
Samples of full window width would be needed to discriminate between fault and inrush. Refer my paper,” Microprocessor based one cycle differential relay for transformers ”....
By the time the frequency measurement has become a significant issue, the grid has failed. At 50Hz or 60Hz it doesn't matter where the measurement is taken because the frequency cannot be significantly different between ends of a grid.
Carl Van Wormer isn`t it the grid operators job to ensure that the demand and supplied energy balance is maintained to ensure that the frequency remains within the specified bounds, despite how small those bounds may be?
For many years the 24-hour average of the US grid was "exactly on" and electrical clocks that were synchronized to the AC line maintained the correct time, while crystal controlled clocks drifted because of their oscillator errors. The Wiki article https://en.wikipedia.org/wiki/Utility_frequency has a section on stability which references another Wiki at https://en.wikipedia.org/wiki/Electric_clock with a section on accuracy that discusses the possible removal of that mandate. I think the average frequency is still very close, but that may not be the case in the future. I have a device on my desk (made by a friend) that shows the phase shift between 60Hz and the current line frequency. I can watch the phase shift in the morning advancing to catch up with the losses in the previous evening.
Re: Allocating responsibility for frequency control is a political choice, and can be done with markets. In reality, the frequency obey sthe laws of physics and is subject to weather events which are sometimes extreme. Making the power system resilient comes at a price.
E.g. see https://www.aemo.com.au/Media-Centre/AEMO-publishes-final-report-into-the-South-Australian-state-wide-power-outage
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity