Double-ended travelling wave fault location using unsynchronized data

In Germany we say “Not macht erfinderisch” which stands for “Necessity is the mother of invention”.

Imagine a situation like shown in the figure above. A customer installed travelling wave recorder with GNSS receiver for precise double ended travelling wave fault location. Unfortunately, at one end the connection to the GNSS antenna is broken. Due to this problem, the time stamps t1 and t2 on both ends are not synchronized and the well-known formula for double ended travelling wave fault location is not working anymore.

What can we do if a fault occurs in this situation, but we want to get a precise fault location? Both devices trigger and save the fault records but the timestamps at both ends are not synchronized.

In this case we can apply single ended travelling wave fault location because for single ended travelling wave fault location the time synchronization of both ends is not necessary. For single ended travelling wave fault location, we need to measure the time difference between the initial travelling wave and it′s reflections from the fault:

With respect to the Bewley-Lattice diagram above the fault location from Terminal A can be calculated as follows:

In the same way the single ended travelling wave fault location can be applied to get the distance to fault from terminal B. In the case shown above, the first reflection from the fault to terminal B is related to the timestamp tB4.

Sometimes it is complicated to identify the right reflections for single ended travelling wave fault location. A good measure to justify the quality of the results from the single ended travelling wave fault location is to check, whether the fault location from terminal A and terminal B get the same result.

Another approach is to use the unsynchronized timestamps from both ends of the line for a double-ended travelling wave fault location according to the following formula:

A great advantage of the approach according to formula (4) is, that the propagation speed is not needed in this formula to calculate the distance to fault.

The quality of the result from formula (4) can be evaluated by checking the sum of the propagation times from the fault to the terminals. This sum should be equal to twice the propagation time over line:

Now let′s see an example:

First, we analyze the fault record from substation A. We see a fault in phase B. The time difference between the initial wave and the first reflection from the fault is 553,6 us.

The fault record from the remote end also shows a fault in phase B. Here the time difference between the initial wave and the first reflection from the fault is 1059,56 us.

The line length in this example is 238,4 km and the propagation speed is 295.020 km/s. With these values we can calculate the fault location.

The example proves that in case of problems with time synchronization the use of data from both ends of the line has several advantages. Checking the results of single ended travelling wave fault location from both ends increases the certainty of the result. The proposed new method of double ended travelling wave fault location using unsynchronized data has the additional benefit to be independent from the propagation velocity of the line.