Never mind the noise

Electrical noise is an ever-present issue when trying to make reliable and accurate measurements in high-voltage substation environments and it’s an issue that was proving especially problematic for one very large investor owned utility (IOU) in North America. In particular, the company was finding it impossible to test bushing CTs on transformers in 765 kV substations. Engineers from Megger, working on site in conjunction with their utility counterparts, have now developed a complete and effective solution to this challenging problem. The subject utility, one of the largest in the USA with over five million customers in 11 states, operates numerous 765 kV substations. A crucial part of the commissioning process for power transformers at these substations is the testing of CTs. Despite having tried instruments from several different manufacturers, this IOU had been unable to find equipment that would give reliable and repeatable results. Measurements falsely showing ratio errors of 100% and more were by no means unusual.

Having been made aware of the problem, Megger engineers visited a substation where tests needed to be performed urgently in rainy weather on a single-phase 765 kV autotransformer prior to energisation. The transformer is a 750 MVA 765/528 kV unit with a 13.8 kV tertiary winding. It is sited adjacent to 765 kV transmission lines that were in operation and energised during testing.

The transformer has 17 CTs – six on H1, four on X1, two on H0-X0, three on Y1 and two on Y2. The majority of CTs have ratios of 3000:5, but there are also some with ratios of 30,000:5. Tests were carried out with the latest Megger MRCT relay and current transformer tester , working in atrocious conditions – it was raining heavily and all work had to be carried out under a tent.

Standard practice was followed when setting up the test, with insulated wires from the top of the bushings brought down to ground level for easier connections. When testing was not underway, the bushings were grounded through these wires.

The first tests were carried out on the H1 bushing CTs. During the insulation resistance test, however, the test set gave a warning that live voltage was present on one of the bushings. Similar problems were encountered when attempts were made to test other CTs.

After carrying out further investigations, which were hampered at all stages by the constant rain, the engineers discovered that the root cause of the measurement problems was that the H1 bushing was acting as an antenna for electrical noise. To address this problem, a directly connected grounding wire was installed on the H1 bushing, even though this bushing is also grounded via the test set.

Table 1 shows the connections made to test H1 bushing CTs. It is to be noted that H2 test lead of the MRCT instrument was connected to X0 bushing instead of X1 for the return path. Since the height of X0 bushing is smaller than X1 bushing, the electric noise interference could be reduced by using X0 bushing for return path.

With this arrangement, it was made possible to make accurate measurements on all six CTs on the H1 bushing. The worst-case error was just 0.029%, and in most cases the error was well below 0.02%. Testing the CTs on the X1 bushing created an additional challenge, since the ground connection from H1 could not be removed because of the level of electrical noise. The standard connections for the test set, shown in Table 2, could not be used as this would have created a ground loop (since H1 test lead is grounded at the test set). A slightly modified connection arrangement, shown in Table 3, was adopted. This has the effect of giving test results that have incorrect polarity, with 180º phase angle, but it is easy to allow a fix for this.

Table 1 – Standard connections for H1 bushing CTs

Table 2 – Standard connections for X1 bushing CTs

Table 3 – Modified connections for X1 bushing CTs

The results for the X1 bushing CTs showed slightly higher errors – worst case 0.24% – than those for the H1 CTs. This was because the height of the X1 bushings made the measurements more susceptible to the effects of electrical noise induced by the nearby live transmission lines. Nevertheless, the accuracy was still very good, and the utility engineers were completely satisfied with the measurements.

A similar approach was used for measurements on the X0 bushing CTs, while for tests on the Y1 and Y2 bushing CTs all H1, X1 and X0 bushings were shorted and grounded together. In all cases, accurate and repeatable results were obtained easily and quickly.

The IOU engineers who witnessed the CT testing in the 765 kV transmission substation were very favourably impressed not only by the performance of Megger’s MRCT test set, but also by its speed of operation. Once the initial noise-related issues had been addressed, all the IEEE C57.13.1 recommended tests (saturation, ratio, polarity, winding resistance and CT demagnetization) were completed on all 17 CTs in less than three hours, despite the difficult working conditions. Based on their previous experience, these IOU engineers had expected these tests to take between one and two full days.

This experience has demonstrated conclusively that the Megger MRCT current transformer analyser is capable of delivering dependable and accurate results in high voltage substations where the high levels of electrical noise make it impossible for other instruments to operate correctly.