Suppose a smart meter is connected to the mainline of the network. Would it be wise to say that the data captured through this meter can be used for fault location in the sub-lateral branches of the line in the same network?
Attached is the figure. The SM is connected to the mainline from where lateral branches go to loads and other sources etc. Suppose, at t = 4 ; fault 3 occurs while the rest of the sections are healthy, what could be the possible approach to locate fault 3 if we have only one meter connected at main bus (line).
Yes definitely, with the help of substantial lowering of voltage in the vicinity of fault with direction of reactive power flow increase and decrease in active power flow (due to shorting of load).
Subrata Mukhopadhyay: "substantial lowering of voltage in the vicinity of fault with direction of reactive power flow increase and decrease in active power flow" absolutely correct!
But, without additional instrumentation, would indications solely at the SM be able to identify the problem as Fault 3 rather than Fault 1 or Fault 2 or some combination?
It would still be helpful if the SM is connected uniquely to a subset of loads in isolating the problem just to the subset. This is a bit like identifying "bus islands", all the buses connected to a unique subset of generators in a network.
Additional circuit breakers could help isolate the fault in the identified subset of loads.
(and there's the perverse, but real world, case when the SM itself has a fault)
For low impedance, high current faults, the protection will trip and deenergize the feeder in a few 100ms. I don't know if smart meters can capture what happened during the fault. And as the power is off, unless the smart meter has a battery, it won't be able to communicate.
In some regions (usually rural), and if the fault is high impedance, the feeder can be operated with a sustained phase-earth fault, maybe for several hours. In this case, clearly smart meters can help by reporting the voltage profile.
There are papers on the subject - search "fault location using smart meters".
Today, this sort of high-speed fault data capture and signaling can be performed by modern distribution feeder RTUs, so utilities are busy installing them on their MV networks. Using smart FRTUs lead to the same sort of question - how many need to be installed in order to accurately locate faults.
The precision of any diagnosis will be limited by the amount and quality of data available {number of locations, detail/quality of what is measured, model of the network, loads, generation}.
Dear Subrata Mukhopadhyay thats a great answer, i agree. But that would depend on the number of meters to be installed at specific positions. My question here is, if i have a SM at substation, the feeder is carrying power for its -n number of lateral branches. The meter placement is for single ended transmission system. How would you identify that the fault is at n-1 location.
I agree with you John Wasson and thankyou for the suggestion. But the question here is to have a single meter for single-ended transmission measurement system (for example) and locating the fault location.
Tom Berry That's a great example you have shared. No doubt, RTUs or its variants are supposed to install, how does it respond to the question. Second, what is its topology in-terms of fault detection for distributed system.
Would it be practical to train a neural network (NN) to diagnose faults based on inputs from the sensor (S).
{ sensor performance inputs } -> NN -> output specific fault F1 or F2 or F3
The training data set would define (discover) unique sensor properties when each specific fault was purposely made.
Because there could be many combinations of faults, for example for F1, F2, F3 single faults
no faults
F1 fault
F2 fault
F3 fault
there are 4 possibilities. But the number of fault conditions increases rapidly for multiple faults with many devices, Fi, i = 1, n. So the AI approach may not be practical.
- Badges
- Science topic
- Similar topics
- Mathematics