Reducing losses in transformers - reducing operating costs

Introduction

 Until the 60s with transformer design strive to meet the specification requirements with a minimum transformer cost... For large transformers, the main thing was to limit the mass and overall dimensions to values determined by transport restrictions. At the same time, they tried to increase the magnetic flux density in the core, requiring manufacturers of electrical steel to manufacture steel that allows the transformer to operate at high induction with a minimum increase in losses and noise level.

At the end of the 60s, consumers of electrical energy realized the importance of full transformer cost and began to include the capitalized cost of losses in the tenders of transformer manufacturers' bids. However, the cost of losses was relatively low, and therefore there was no strict requirement to manufacture equipment with very low losses.

During the 70s, the cost of oil increased by about an order of magnitude, which led to an increase in the cost of other fuels and energy. The capitalized cost of losses increased accordingly.
Since then, the cost of energy, as well as the capitalized cost of waste, has continued to rise. There is no reason to believe that their value may significantly decrease in the future.
Therefore, it is objectively required when designing to achieve the lowest losses.

    Order a transformer now!

Reduction of losses

Losses in transformers

During the operation of the transformer, losses occur, which consist of no-load losses arising from the magnetization reversal of the amorphous steel of the core, and load losses, which are the sum of losses in the copper of the windings and additional losses in the walls of the tank and other metal parts caused by the leakage flux.

 The rise in energy costs has stimulated a reduction in both no-load losses and load losses, the latter especially in generator and other transformers with a high load factor. Over the past 30 years, transformer losses have been reduced by an average of 50%.

No-load losses

 In the 1950s, the previously used hot rolled steel was replaced by cold rolled steel having an oriented grain (domain) structure. Cold rolled steel has a high magnetic permeability and low magnetic flux losses in the longitudinal direction, i.e. in the direction of rolling. Over the past 30 years, there has been a significant improvement in performance and cold rolled steel, which has been driven by an increase in the capitalized cost of losses.

 With an increase in the capitalized value of the load losses, it is advisable to increase the induction in order to reduce the number of winding turns and thus the load losses. Therefore, it became necessary to create a steel capable of operating in transformers at relatively high values of induction at low specific losses.

The reduction in idle losses was due to three factors:

- the use of improved steel grades;

- improving the technology of manufacturing the magnetic system and, especially, cutting steel;

- improving the design of the core, and, above all, the joints of the sheets sleeps.

 Since the introduction of grain oriented transformer steel on the market, its quality has continuously improved and achieved impressive results.

Improvement of steel characteristics was due to:

- improving the orientation of domains;

- reducing the thickness of the sheets;

- purification of domains using laser treatment of the surface of the sheets.
                                                                                                                                                       потери в трансформаторной стали
Steel is now available with thicknesses of 0.27 and 0.23 mm for industrial applications.
Small quantities of 0.18 and 0.15 mm steel were made for trial use.
Improving the orientation and clearing the domains do not affect the transformer manufacturing technology, while a decrease in the thickness of steel sheets leads to an increase in the number of magnetic core sheets and to an increase in the mechanical sensitivity of the material. It is obvious that a decrease in losses in steel is associated with an increase in the labor intensity of the assembly and an increase in the cost of the material.
As for the noise level, its reduction caused by the use of improved materials is negligible compared to the reduction in losses. The use of the Hi-B grade with a sheet thickness of 0.27 mm, laser treated, reduced losses according to some data by 30%, while the results of measuring the noise level varied from a decrease in ZBB to an increase of 5 dB.
Today, there is steel on the market with a specific loss of 1.05 W / kg at a thickness of 0.3 mm, 1.00 W / kg at a thickness of 0.27 mm and an induction of 1.7 T.
About 50% losses in steel are eddy current losses, and 50% are hysteresis losses. Therefore, steelmakers strive to reduce sheet thickness. It can be expected that steel with a thickness of 0.15 mm can have specific losses of the order of 0.7 W / kg for the same induction.
Steel manufacturers offer a wide range of steels with different characteristics, and transformer manufacturer can choose steel depending on the design of the transformer and its required characteristics.

In fig. 4.1 shows the comparative characteristics of some grades of steel.

      

Amorphous steel

There is a certain rivalry between the two paths of development:

 a) The use of conventional carbon steel with improved orientation and controlled grain size and with reduced sheet thickness;

 b) Using amorphous steel tape.

 The use of amorphous steel requires new ideas in design and technology in order to fully exploit its benefits.
An amorphous material is obtained by rapid cooling in the form of a very thin ribbon with a thickness of no more than 0.02-0.03 mm. Despite significantly reduced losses, amorphous steel is unlikely to replace the ubiquitous carbon steel in transformers... The main disadvantages of amorphous steel are the low saturation induction, low utilization, and relatively high magnetostriction. In addition, fragility, the need for annealing in a magnetic field, mechanical sensitivity and high cost will also hinder its widespread use, at least in laminated magnetic circuits. However, there is a possibility of using amorphous steel in single-phase  distribution transformers with wound magnetic circuits. This may be appropriate for large values of capitalization losses.

Successful work on bonding several layers of steel tape with a compound up to a thickness of 0.15 mm may open up the possibility of using amorphous steel in laminated magneto-cylinders. Since the losses in this steel are almost independent of the direction of magnetization, connections (joints) can be made very simple without increasing losses. The general consensus is that in the near future the use of amorphous steel will be limited to distribution transformers, provided that its price is below $ 2.5 per kg and the cost of losses is above $ 2.5 thousand per kW.

Load losses

Unlike no-load losses, the reduction in load losses was not accompanied by significant material improvements. Load losses consist of basic losses PR in the winding wire, additional losses in the wire due to eddy currents and skin effect, and additional losses in the tank walls and metal parts of the structure.

Reduced wire loss

 The main method for reducing load losses was to reduce the current density in the wire by increasing its cross section. However, this had two negative consequences. The first is an increase in the space occupied by the windings, which increased the size of the core, and, consequently, its mass and no-load losses. Secondly, an increase in the cross-section of the wire led to an increase in additional losses in the wire, i.e., losses caused by eddy currents and the surface effect. The use of a compact wire, consisting of a large number of insulated and transposed conductors with common insulation, partially eliminated the first drawback and, to a large extent, the second.
Currently, transposed wire is used in large transformers, in which the number of elementary conductors can reach 80. The wire can be insulated with epoxy resin, which, after polymerization during the drying process, imparts greater rigidity to the wire, which increases the strength of the windings when exposed to short-circuit currents.

Reducing added losses

 The additional losses in the metal parts external to the windings are caused by the leakage flux created by the windings, which depends on the ampere-turns and the configuration of the windings and does not depend on the current density. As the losses in the windings decrease, the share of additional losses outside the windings in the load losses increases, especially in transformers with a large value of the short-circuit resistance.
Previously, the control of the stray field was carried out primarily in order to avoid unacceptable heating at certain points of the tank walls and other metal parts, especially in transformers of high power or with a high value of short-circuit resistance. Today, such a control of the stray field is also carried out to reduce additional losses. Measures to reduce additional losses are to use conductive shields to deflect the magnetic flux from the protected surface, or electromagnetic shunts that collect and direct part of the magnetic flux in the desired direction. Non-magnetic, electrically conductive shields prevent leakage flux from penetrating into magnetic material, in which high losses can be induced.
The advantage of such screens is their simplicity and the ability to give them the required shape to protect surfaces of complex configuration. Their disadvantage is that losses occur in the screen itself, which must be estimated, and the screens themselves must be cooled. In this case, there should be a control of the leakage flux deflected by the screen, which can induce losses in other parts made of magnetic material.
Electromagnetic shunts direct the flow in paths where there can be only small losses, preferably in paths outside the tank walls and other metal parts. The advantage of magnetic shunts, recruited from electrical steel, is better control of the leakage flux and losses created by this flux. The disadvantage is the difficulty of giving the shunts the required shape to protect parts of a complex configuration.
In addition to electromagnetic shields with high electrical conductivity and electromagnetic shunts, it is sometimes practiced to replace individual metal parts of a structure with parts made of insulating materials with high mechanical strength.
In addition, some structural details located in the region of a strong field, for example, bushing adapters, can be made of non-magnetic materials with a relative magnetic permeability ð from 1.1 to 1.8 and a high conductivity of the order of 0.8-1.0 Ohm • mm2 / m.
Shielding can slightly change the value of the short-circuit resistance (by tenths of a percent).
In three-rod transformers that do not have delta-connected windings, the zero sequence resistance can almost double as a result of shielding the tank walls.
Experimental data confirm the effectiveness of the use of shields and shunts to reduce the additional losses and temperature of local heating of metal parts. According to some reports, the best result is obtained by shielding the tank walls with electromagnetic shunts, and metal parts near the outlets of high currents - by electromagnetic conductive shields.
Electromagnetic shunts are sometimes used to protect the yoke beams. Such shielding can reduce the added loss in the metal parts to be protected by more than 50%.
However, any shielding must be accompanied by a zero scattering control. if the screens are installed incorrectly, the additional losses may not only not decrease, but also increase.
Currently, the additional losses can range from 10 to 40% load losses. It can be assumed that the reduction in load losses, as well as no-load losses, achieved over the past decades, was largely stimulated by the high specific capitalized cost of losses.

Calculated determination of the leakage fluxпотери в трансформаторе

 Currently, sophisticated computational methods are used to determine the magnetic leakage flux. Such calculations, for example, using the finite element method, can be performed for a two-dimensional zero, and in more complex cases, for a three-dimensional zero. Computer programs based on these methods make it possible to determine the most advantageous position of protective devices (screens or shunts), the value of the losses created by the leakage flux and the temperature at the place of the greatest losses. In fig. 4.2. shows the distribution of losses in the tank wall caused by the stray field in the absence and presence of protective elements. The curves are obtained by calculation on a computer using the finite element method.

Loss measurement

Much attention should be paid to the measurement of losses. Loss measurement accuracy is important for transformer manufacturer, since it allows you to correctly assess the changes made by materials and construction. For the consumer, measurement accuracy is important to correctly estimate the full capitalized cost and compare losses.

Measurement of no-load losses

No-load losses depend on the voltage value, its frequency and shape. No-load losses have two components - hysteresis losses and eddy current losses.
The hysteresis loss is a function of the maximum induction value and depends on the average value of the applied voltage. Eddy current losses are a function of frequency and are therefore sensitive to the harmonic composition of the voltage.
     A higher power factor value for no-load loss measurements results in higher accuracy than load loss measurements.
 However, there are other issues to consider.:
- Instrument transformers and wattmeters must have appropriate frequency characteristics;
- The resistance of the test voltage source should be small enough for all harmonics to have minimal voltage waveform distortion caused by the non-sinusoidal excitation current of the transformer;
- The formula for converting the measured loss to sinusoidal form assumes 50% hysteresis loss and 50% eddy current loss. This assumption is not true enough for all modern steel grades;
- The core temperature affects the value of no-load losses caused by eddy currents. Deviations in no-load losses due to temperature changes can be significant. So, when measuring losses at 21 ° C and at 50 ° C on a 50 MB • A, 110 / 10.5 kV transformer, a decrease in losses with an increase in temperature was noted. At a nominal induction of 1.77 T, the decrease was 1.2%, and at an induction of 1.6 T, it was 3.3%.
There was no noticeable change in idle losses during operation.
 It should also be borne in mind that no-load losses may increase after impulse testing. The difference can be less than 4% on average. The reason for this may be insulation breakdowns at the ends of the sheets due to the presence of burrs. There are cases when the controlled absence of burrs made it possible to avoid an increase in losses after

Measurement of load losses

At a low value of the power factor, errors in the measuring circuit, especially in instrument transformers and wattmeters, lead to a significant error in the measured losses. The lower the power factor, the larger the error can be. If the power factor is 0.01 when measuring the loss, an error in phase angle of one minute (290 microradians) will cause an error in the measured power of 2.9%.
The accuracy of measuring the load loss at a power factor of not less than 0.01, equal to 3 % is considered acceptable. Further improvement of accuracy requires a very large investment.
However, some firms report 1 % accuracy at 0.01 power factor and 0.5% accuracy when measuring no-load losses on transformers up to 300 MB-A.

Capitalization of losses

Total cost of the transformer and its optimization

The cost of the transformer, taking into account the cost of operation for the entire service life, consists of the following components:

- the price of the transformer;
- the cost of on-site installation;
- the cost of preventive work and maintenance;
- the cost of losses.

Reducing the cost of a transformer can be achieved by reducing the investment of active materials (electrical steel and copper). But at the same time, losses will increase. On the contrary, to reduce losses, additional investment of active materials is required, the use of more expensive materials, for example, steel with reduced specific losses.
The buyer of the transformer together with the manufacturer can choose the best option for the technical and economic characteristics of the transformer within the parameters regulated by the standards, such as maximum heating temperatures, etc. Usually, when comparing options for the same transformer, the main indicator is losses.
Although the efficiency of modern transformers exceeds 99 percent, the cost of losses for the entire service life, given at the time of installation of the transformer, can exceed its price. Based on the load graph of the transformer and the cost of electrical energy, the annual cost of no-load and load losses can be determined. The cost of losses in each year of the entire life of the transformer can be covered by the annual income earned from the bank's compound interest rate when the transformer is installed. This amount, sufficient to cover the cost of loss in each year of the transformer's life, is the capitalized cost of the loss.

The total capitalized cost is the sum of the cost to install the transformer (including its cost) and the capitalized cost of loss. There is an inverse relationship between these values. Therefore, there is an optimum of the total cost when the technical and economic characteristics of the transformer change.
It is possible to perform calculations for each year, taking into account changes in various parameters over time: energy costs, losses and the amount of bank interest. It is difficult to predict the changes in these parameters over the entire 25-year service life. Therefore, constant values of the parameters are taken and the calculations are reduced to the determination of two components: no-load losses and load losses.
But that's a topic for another article.

To the list of articles

Total comments: 1

  • юрий
    By yuri Added on October 25, 2018 at 11:45 pm

    There is a laboratory-tested technology for reducing the losses of the XX power transformer by a technical method to 1% from the nominal level without changing the parameters of the generated power, which is currently being patented.

    To answer

leave a comment

Your email will not be published.

en_GBEN