Detection of faults in the transformer LTS devices. Enhanced DRM method
According to the data of ORGRES, a leading engineering company in the power industry, 13.5% of 35-750kV transformer failures are due to the LTC devices. According to data of MRSK Sibir Branch of Krasnoyarskenergo, more than 23% of 35kV transformer failures are due to the LTC devices. Causes of LTC faults include failures in the contactors and switches, burns on the contactor devices contacts, seizure of contactor mechanisms, deterioration of mechanical strength of steel details and of a bakelized-paper shaft.
For revealing the causes of LTC failures and for automation of the diagnosis the SKB EP Company entered the instrumentation market with the PKR-2 and PKR-2M instruments for the transformer LTS testing. Their main function is recording the oscillograms of contators operation and the circular diagram across all the phases at a time The results are output in graph and tabular forms. The instruments are furnished with large color displays of high brightness and contacts that facilitate graphs analysis.
The PKR-2 is intended for recording the circular diagram and for LTC oscillography. Convenient external clamps facilitate the instrument connection to contactor contacts without oil drainage. A circular diagram and an oscillogram across the leads are recorded in one switchover within the time limit of 10 minutes. After each measurement the instrument automatically computes and displays the results in three forms.
- A developed circular diagram (Fig. 2);
- A circular diagram built based on switchover input data graphs;
- Table of reduced values of a circular diagram (Fig. 3).
The Table gives computed values of the switch contacts condition at control shaft rotation.
- К1/К2 opening is an angle of the output drive shaft rotation to the moment of the first vibration opening of the first and second contacts of a contactor;
- К1/К2 closing is an angle of the output drive shaft rotation to the moment of the first vibration closing of the first and second contacts of a contactor;
- I1/I2 retreat is an angle of the output drive shaft rotation to the moment of the first vibration disconnection of the first and second contacts of a selector from the transformer taps-off;
- I1/I2 stop is an angle of the output drive shaft rotation to the moment of the first vibration closing of the selector I1 with a transformer tap-off by a contactor.
These measured characteristics give an idea on the LTC condition. A circular diagram after measurements as in our case has been built incorrectly, and results obtained given in the table were not calculated for A phase (Fig. 2). The result was affected by excess of permissible values during calculations, e.g., contact chatter excess, which is obvious on the initial graphs (Fig. 3). It is also observed on the phase A graph (Fig. 3) where considerable chatter occurred and impacted the computation of a circular diagram. In this case the subsequent taps-off need multiple measurements. If a circular diagram is incorrect during repeated and even additional measurements, then it is necessary to check the connection circuit of the instrument to LTC and all its connections.
When reviewing the input data graphs in the PKR-2 software format, both taps-off (U1 and U2) will be overlapped (Fig. 5).
Due to possible faults in the tested switches (incorrect connection of LTC reactor taps-off) the tests should be performed after each mounting and adjustment works. If such a fault is detected, the circular diagrams cannot be built using the PKR-2, and for revealing the cause, the following input data graphs shall be analyzed:
Considered is the example of LTC switch-over of a volt-additional autotransformer (a booster) with incorrectly connected taps-off. Connection of "General" tap-off is mixed up with one of the taps-off of a control winding. It is obvious that initial graphs for this type of fault differ from a reference graph given above. Further review of graphs allows us identify which taps-off were mixed up during LTC mounting.
The PKR-2 performs oscillography of a contactor of LTC resistor devices. Mind that oscillography shall be performed on the switch disconnected from the working voltage, at low test currents and voltages; the nature of contacts vibration can be different and it is not standardized, that is, it is not a cause for rejection. Break of a current circuit evidences incorrect alternation of contactor contacts operation and excessive deviation of switching intervals from the norm, which evidences their wear or poor adjustment. In this very case the contactor is rejected, it is subjected to inspection and repair (Fig. 7).
Fig. 8 gives an oscillogram of LTC contactor of RS type with a circuit break detected at one of the operating sections of the arch extinction contacts that was caused by coal formed.
The main peculiarity of a break on this oscillogram is that it shows the contacts wear, rather than break of a current-limiting coil, which is due to the following facts:
- in direct switch-over the break occurs on the left, in the reverse switch-over it occurs on the right;
- "bridge" position is observed, which is possible only under coil circuit integrity;
- break is accompanied by chatter, jumps achieve normal level.
One of the PKR-2 benefits is an option of non-destructive control and diagnosis of LTC devices using the DRM-test. You do not need to remove the contactor tank cover. Analysis of graphs for winding resistance measurements at taps-off switchover allows rejection against operable/failed criteria, and very often indicates on the defect nature, which allows elimination of unnecessary openings and repair of properly operating LTC devices. As the data base on the graphs of the known defects for particular LTC devices increases, a more detailed diagnosis can be made.
- Express-diagnostics of transformer LTC at any weather conditions;
- Construction of a diagram for contactor operation without opening the LTC tank;
- Analysis of graphs for the measured item directly on the instrument;
- Localization of LTC defects, e.g., detection of a break of current limiting resistors, bad contact of a selector, etc.
Graphs in Fig. 9 present oscillograms of contactor switchover. The figure on the left shows primary results of resistance measurement during switch-over, and the figure on the right shows the results of measurement after mathematical analysis. As is seen from the figures, the DRM-method in the PKR-2M is much better implemented, and owing to mathematical analysis the obtained oscillogram came close to the results that could be obtained at direct connection to the contactor contacts. Insofar as the non-destructive testing of the LTC contactor state is not labor-intensive, it should be undertaken at any scheduled and off-schedule control of power transformers.
The use of PKR-2 and PKR-2M instruments considerably reduces expenses and labor costs of a company, raises transformer diagnosis quality and allows elimination of unnecessary repair.