The United Kingdom is powered by a huge grid of pylons, conducting electricy across hundreds of miles. Thanks to this network of transmitters, we can switch on the lights, boil a kettle and visit the Specac website to read this article.
Like any piece of equipment, pylons require maintenance. As high voltage electricy (up to 750,000 volts) surges through the pylons, transformers are used to contain the current and prevent dangerous power surges. From time-to-time, the insulation in these transformers is inevitably worn down and requires replacement. But removing the material prematurely, when it is still effective, can be an expensive overhead easily avoided by predictive testing.
A Ph.D. student at Glasgow Caledonian University, Graeme Smith, spoke to Select Science about his studies into the testing of dielectric materials and how his technique could save signaficant money for the energy industry.
Saving money and time
Graeme told Select Science,
What I am trying to do is a quick, chemometric analysis into how the insulators are performing inside the equipment, to enhance predictive maintenance of this equipment... Currently, energy companies do not employ predictive maintenance on the electrical transformers, the mineral oil is replaced and then sent for dissolved gas analysis. The mineral oil could come back absolutely fine, but it doesn’t matter — they’ve already changed it.
To take out 100 L of oil and replace it, can be anything upwards from £1,000... But if you don’t need to do that, and you can do a test on a small sample of the oil to say it’s fine, this can save energy companies a lot of money.
Graeme uses FTIR to observe any changes in the spectra of a kind of oil used for insulation, using a principle component analysis to detect chemical breakdown markers in the sample. But ease-of-use and high throughput are required to keep up with the samples change over time and avoid background drift.
The main problem I had was with the traditional sodium chloride cell. I was having an issue with sample thickness. As the sample degrades it becomes thicker, more viscous and harder to sample. The more I put it in the traditional cell, even with the same sample, I was getting a huge background drift.
By using the Pearl FTIR Transmission Liquid Accessory, this issue was averted.
When you put the sample on the cell and close the Pearl, any excess is pushed out over the side, creating a uniform film thickness, removing any drift that I was seeing,” explains Smith, which enables him to effectively observe the growth of carbonyl peaks in the mineral oil under thermal stress.
The repeatability has sped up the results because I am now able to confidently get those results.
Simple dielectric oil analysis using the Pearl FTIR
Thanks to the Specac Pearl's speed-and-ease-of-use, Graeme Smith could to utilize a principal component analysis and compare similarities and differences between the FTIR Spectra of different combinations of the mineral oil with the innards of an electrical transformer (consisting of copper and Kraft paper) at both ambient and heated temperatures.
Each result shows a different response in the principal component analysis, which you would not be able to determine spectrally. For the first time, a statistical technique for the assessment of the dielectric oil's degradation has been shown. Smith's hopes are to correlate the results he has gathered with the tests commonly carried out by in-field service engineers, such as neutralisation number and interfacial tension, as a prediction model.