TRANSFORMER DESIGN

A Tried and Tested Product, But With Scope for Development

The transformer is a product that is more than one hundred years old. Many of the basic functions and solutions remain the same after all these years, in spite of the fact that transformers have undergone a comprehensive development process. One could therefore assume that most things that could be done have already been done; but still there is extensive ongoing R&D being conducted at a highly theoretical level. This research is looking at new materials for insulation, methods for condition monitoring and diagnostics, noise elimination, environmental impact and the possibilities of using superconductive materials.

Intensive search for New insulation Materials
The oil and cellulose material known as pressboard has been used in the insulation of transformers for one hundred years. Experience has shown that it works well, and over the years both the material and the methods have been refined.

Pressboard has, however, one obvious disadvantage: its good electrical properties drastically deteriorate if the material is damp. This means it must be dry. There must be a drying process during the production of the transformer, and during the erection work in the field -- the ingress of air must be limited to the absolute minimum, especially if the air is damp.

To avoid difficulties caused by sensitivity to moisture, we are researching other materials. To merely carry out tests with some plastic material and maintain that it works well in the lab, is not the way forward. The strength of pressboard is that we know that it can actually function over a long period of time -- 50 years is nothing unusual. To be sure that any new material will function for such a long time, there are two alternatives: either we must test the material over a very long period; or, amass a sufficient amount of theoretical knowledge to assure us that the new material really will function for the required number of years.

Naturally our primary aim is to increase our theoretical understanding of what happens inside a transformer. One route, for example, is to increase our understanding of the physical processes that occur in connection with a flashover, and how this is influenced by the materials present.

A new material should have the following properties:

• It must be resistant to electrical breakdown.
• Electrical flash-over must not slide along the materials surface (which it can do with pressboard).
• It must be homogeneous.
• The material must be mechanically stable.
• It must not be affected by or affect the oil.

Right now we are in the process of constructing a full-scale model with a new material. There remain, however, many questions that we have to find answers to. The knowledge gained from working with pressboard, e.g. when it comes to dimensioning, is to a large extent empirical and can therefore not be directly applied to other materials.

Condition Monitoring Helps the Customers Make Decisions
The material that could eventually be used to replace pressboard is unlikely to be any cheaper to buy. It could, on the other hand, contribute to lowering total costs if you can dispense with drying, and perhaps the dimensioning can be changed. In addition, better aging properties can hopefully prolong active life.

An important means for our customers to save money is to make the right investments at the right point in time -- and thereby of course also avoid making unnecessary investments.

In general, it is estimated that a transformer will last for 25 years. At the same time, we know that some last for 50 years and more. In this respect it is of interest to develop techniques for condition monitoring and diagnostics.

Privatization and deregulation of the power industry in many countries has changed the demands placed on the equipment. When everything was under the control of national power companies, the demands on the reliability of individual components were less. Then, problems at one location could be compensated for by taking power from somewhere else until the problem had been remedied. A smaller power producer who has undertaken to supply power to a major industry has no such options. Cuts in power supply cannot be compensated for and therefore must not occur.

Traditionally, the condition monitoring of power transformers takes place through keeping a check on temperature and oil levels, along with regular maintenance of the tap changer. In addition, oil samples are often taken once a year for gas analysis, to check that no problems are in the process of developing. Increased demands on non-stop supply also mean demands for early warning of incipient problems.

To meet these demands, we are now developing a gas detector for online use, which can detect several gases. We are also developing a detector that can trace partial discharge, which can be the first sign of a fault developing. To handle all the information provided by condition monitoring, a data collection unit with an accompanying communications system is also being developed, to enable information from different installations to be collected in a central control room.

When an indication comes in that something is not as it should be, it's a matter of being able to make the right diagnosis. This is going to demand considerably more advanced, and therefore more costly, equipment; it will not be possible to install this in the field, but it should be possible to take it out to the site if necessary.

A detailed gas analysis can give information about what is happening, but not where it is happening. Partial discharges emit sounds in a range of frequencies which can normally not be detected by the human ear. This is utilised so that, with the help of special microphones, placed in a certain way, the distance and direction of the sound can be located.

Over and above the information obtained through measurements, diagnostics is to a very large extent built on experience. In order to make all our collected experience available, we have started to incorporate it into knowledge-based diagnostic systems.

As a part of the general heightened environmental awareness, demand for quieter installations has increased. There are two ways of tackling the problem of unwanted noise. You can either muffle the existing noise or eliminate the cause.

If you want to deaden noise the simplest way is to case it in. A more advanced method of suppressing noise is to record and analyze the noise and then transmit the same noise, but offset in phase, so that the sound waves eliminate each other.

It is best however, in most cases, to start by tackling the problem at source, primarily through increasing our understanding of sound generation. In addition, the noise generated is often amplified by resonance phenomena. Such phenomena can be eliminated through construction, if dimensions that give resonance with the sounds that are generated are avoided.

So, that we can get a perspective on our products' environmental impact and prepare ourselves for possible future demands from the authorities and the world around, we have started to work with life cycle analysis. Such analyses indicate there is one factor in the transformer that exceeds all others by a very large margin. That is the energy loss during operation.

Superconductors and High-Voltage Direct Current in the Future
Suppose that one of our transformers has an efficiency of 99.7 per cent. If you work out how much energy this loss of 0.3 per cent corresponds to per year, and subsequently during the entire lifetime of the transformer, then we are talking about huge amounts of energy. The transformer's total environmental impact, then, is largely governed by the environmental impact of the electrical power producing installations.

The biggest contribution we can make to helping the environment is to improve the transformers' efficiency, which is a matter we are addressing.

If one wants to look further into the future, then there could be more dramatic innovations. One of them, of course, is the use of superconductive materials. Superconductors could give us smaller transformers and transformers able to effectively handle high frequencies at a low voltage.

This development has, however, not proceeded as rapidly as many hoped when the first discoveries were made. While it has been possible to discover materials that are superconductive at "high" temperatures i.e. temperatures over -70 degrees C, these materials are extremely expensive.

But there are still possible advantages to be gained with superconductors, so we are continuing to work with these materials. We have constructed one prototype and are about to start constructing a second.

There are signs that the transformers of the future will be of a totally different type than today. Some believe that alternating current will turn out to be a parenthesis in history, albeit a long one. If the energy of the future is to come from solar power stations in the Sahara, one might as well keep it as direct current. Already now it is cheaper to transport energy long distances as high-voltage direct current, and for most uses direct current is as good as alternating current. In those cases where alternating current is necessary, for example in motors, 50 Hz is in any case seldom ideal.

It will take a few decades before the time comes to scrap alternating current. Until then, we will continue to develop the management of it just as far as it is possible to go.