FI62430C - LINDNING FOER TRANSFORMATORER ELLER REAKTORER - Google Patents

Description

R5F »1 M« «» ““ 'UTUsjutKAisu 624 30

Jf »c ut ^.? Ä 'mm w 1 = n» ^ * ζ3γΓ · Patent rr. ? 11l "t ^ v ^ (51) K» .ik? /Int.a.3 H 01 P 27/28 SUOMI — FINLAND (21) T6oi »8it (M) H« k »wl« pMv »- AmBfcnlnwdf 25.02 .76 (23) AlkupUvt — GIM | h «t * d * t 25.02.76 (41) TutWt (vtklMkal - BIMt 07.12.76

Pttwitti · Jm r * Kist «Rih * l! Itut ............... ... . .

_. (44) Date of Nihttriksipanan It kuulfulkAlau. - λΐ.08.8?

Pltmth och r «gitt« rityralaan AmMcan utbfd oeh utl. * Krtft * n pubUcarad (32) (33) (31) Pyr> «« y «uoMceii * —Begird prtortMC 06.06.75

Norway-Norway (NO) 135 ^ 9 ^ (71) A / S National Industri, N-3000 Drammen, Norway-Norway (NO) (72) Petter I. Fergestad, Norway-Norway (NO) (71 *) Berggren Oy Ab (5 ^) Winding for Transformers or Reactors - LirJdning för transformatorer eller reaktorer The present invention relates to windings for transformers or reactors. The winding comprises a plurality of windings connected in series, in which the dimension of the conductor in the axial direction corresponds to the height of the winding, as is the case with windings made of metal sheets or strips. The winding is characterized by high windings with many turns in the middle and windings with decreasing height and speed towards the ends of the winding.

The winding is built symmetrically and, in the case of high-voltage operation, has an input connection in the middle winding.

!

A known problem in transformer construction is that it is possible to install the windings so that the surge voltage distribution becomes as linear as possible.

With regard to the overshoot calculation, it can be stated that the ratio of the surge voltage load between the disc windings and the corresponding linear voltage load can be expressed by the equation: 2 62430 Δβ surge ^ l / cj Δθ linear y Cs where Cj = winding capacitance with respect to ground Cs = winding series capacitance.

This equation holds between 2 <a <10-15.

A good surge voltage distribution or a non-oscillating state in the winding is achieved when the surge voltage in each winding is the same as the voltage with the operating cycle number, both calculated as a percentage of the effective voltage.

It is known that the high voltage winding of a transformer is divided into two parallel groups with the input circuit located in the middle of the winding.

In wafer windings, the aim is to make the voltage distribution linear by using layered windings. In this case, the series capacitance of the windings increases, and the voltage distribution thus becomes more advantageous. A completely linear voltage distribution cannot be achieved with this method. In practice, a doubling of the approximately linear voltage stress is the lowest that can be achieved. Such a winding comprising layered windings is laborious to manufacture. The manufacture of a layered wafer requires about twice as long as the manufacture of an uncoated wafer.

It is further known to use a combination of layered and non-layered disc windings to achieve the most linear surge distribution possible. The use of such a combined winding makes it possible to keep the filling factor at a reasonable value. It is desired to make the filling factor, namely the ratio of the conductor cross-section to the total cross-sectional area available for the conductor and the insulator, as large as possible.

A high filling factor is achieved by using tape or sheet winding. It is known to manufacture windings for small voltages from one or more strip or sheet windings of the same height connected in series. The use of this type of winding is limited to low powers, such as in distribution and power transformers, powers up to 10-15 MVA. This is due to the large additional losses of 3,62430 that occur at the ends of the windings. These losses increase as the scatter exam increases, which in turn increases as the power increases.

The new winding structure makes it possible to achieve a completely linear surge voltage distribution. In addition, the large additional losses that would otherwise occur at the ends of the winding due to radial scattering are avoided. This is possible by assembling the winding from sheet or strip windings whose speeds decrease from the input point of the phase voltage to the star point. Taking into account losses and short-circuit forces, the ampere speed per unit height is kept approximately constant. This means that the middle winding is made of a thin wide metal sheet. As the winding height is reduced, the thickness of the sheet or strip is increased. Thus, narrower windings are not normally wound using a sheet or tape but using a profile wire.

The conditions for the linear surge voltage distribution are represented by the following formulas, in which the notations are the same as in the circuit diagram of Figure 1:. · Ni Δθ1 'N (i + 1) * Δθ (i + 1) 1 n Σ Np

(Ιηή - N (Ϊ + 1) Pg + P ~ i * i r II

Csi Ni c (i + 1) Ni ji 11 where i is between 1 and n - 1.

The speeds of the windings can be expected to decrease in a geometric series, for example. In addition, the same radial filling factor must be maintained in all windings.

N1 + 1 = k · N. III

Then Equation II takes the form:

csi = k ° s (i * l) + Γ- · k · ° ji IV

Provided that the radial filling factor and the ampere speed per unit height remain constant, the equation is obtained: 4 62430

^ si "k Cs (i + 1). V

By placing in addition to Equation IV a practically suitable value

Cji = 0.25 Cs (i + 1) VI

the equation is obtained: 1 kn ”^ - 1

1 = k + k K-T_TL. . 0.25 VII

If n = 10 is placed in this equation and applied to the first 4-5 windings, i.e. i = 1, 2, 5, 4, 5, k = 0.73 is obtained, which is fairly accurate throughout this range.

That is, the condition of the linear surge voltage distribution is satisfied if the coil height and the coil speed decrease in a geometric series with a coefficient k of 0.73 · When the radial filling coefficient is constant, a constant ampere-rotation density in the height direction is obtained. This is quite accurate for the first 4-5 windings in our example. Farther from the intake point, the speed change may deviate from the geometric series and / or the radial fill factor may vary somewhat while the condition is still met. In practice, however, a deviation from the linear distribution away from the intake point can be accepted. The distal part of the winding can thus be made as a conventional disc winding. This fraction can be between 3-5 5? ... 60-70% of the total winding length depending on the production equipment and methods of each manufacturer. The rest of the winding is made up of leaf or strip windings.

A calculation example is presented to show that the condition of a linear surge-voltage distribution can be satisfied by realistic means, and the example is not intended to be anything other than an explanation. In practice, a deviation from the linear surge voltage distribution is taken into account in order to take into account economic factors such as the simplicity of the structure and the presence of as few winding heights as possible. In other words, a compromise is made between the most linear surge voltage distribution possible and the presence of as few different winding types as possible.

The new winding method with the feed point in the middle also provides 5,62430 advantages over conventional sheet metal winding in terms of the additional losses caused by the radial scattered fields at the ends of the winding. In a winding made either of one complete sheet or of a total of series-connected leaf windings of the same height, there are large additional losses at the ends of the winding. These losses are due to radial scattered fields that cause uneven current distribution. Attempts have been made to avoid these losses by protecting the ends of the winding. Such protections are neither economical nor practical in high power and high voltage transformers. In the new winding, the highest windings are located in a scattered field parallel to the sheet or strip. Near the ends of the winding, the windings are so narrow that the additional losses due to the radial scattered field remain considerably smaller. Thus, this type of winding has significantly lower additional losses than other types of winding because it has high turns with many turns, where the scattering field is parallel to the winding, and low windings with few turns at the ends, where the radial scattering field is the largest.

Figure 2 shows an embodiment of the winding, in which 1 denotes the take-up windings and 2, 3, 4, 5, 6 and 7 windings, the height of which decreases symmetrically on each side of the take-up winding.

Figure 3 shows an embodiment with a take-up coil 1 and d.

the tapered leaf windings 2, 3a and 4o, but at the ends of which the disc windings 5, 6, 7 and 8 are equal to each other. a * a * a a

Fig. 4 shows an embodiment with a take-up coil 1b, leaf coils 2 ^, 3 ^ and ^ of equal height symmetrically on each side of the take-up coil and, in addition, disc coils 5b, 6 ^, 7b and 8 ^ at the ends of the winding. This structure comprises the smallest number of different windings.

The new winding is very resistant to radial forces, especially if epoxy-coated prespan is used as the layer insulation. The part most susceptible to "bumping" is the center of the winding, in which case a very strong structure can then be provided. The winding can be made self-supporting and thus does not need to be supported in the heart. Besides, the winding is easier to manufacture with a precise tolerance in the height direction. In this case, the axial short-circuit forces were smaller.

6 62430

It should also be particularly emphasized that the type of winding described above has advantages in terms of manufacturing and insulation, which result in large economic savings compared to current winding methods. High voltage transformers today mostly use some type of layered wafer winding. Since it is known that the production of a layered wafer takes twice as long as the manufacture of an uncoated wafer, it is immediately apparent what savings can be achieved. In addition, the leaf winding is simpler to manufacture, and this still means a 30% saving. The new winding comprises fewer wafers than conventional. Thus, there will be less installation of spacers and height adjustments. In addition, the leaf winding is easier to insulate, and the insulation requires less space, so the filling factor can be increased. In this case, the need for active materials of the transformer can be reduced. The new winding structure can be used as a high-voltage winding or low-voltage winding in connection with any other type of winding. In a high voltage winding, the phase input is preferably in the middle and its ends are connected to ground or a star point, because otherwise large additional losses will occur. When a new winding is used as a low voltage winding, the impact resistance requirements are easier to meet and less prevalent.

Also in this case, the new winding end is capable of reducing additional losses.

The design of a low voltage winding depends on the voltage for which it is intended. The simplest possible low-voltage winding comprises a high-leaf coil in the middle, connected in series with the low windings at the ends, with radial stray fields occurring at the ends of the winding. The height of the middle winding can then suitably be at least 1/4 and preferably more than half of the height of the entire winding.

Figure 5 shows the simplest form of low voltage winding with a high middle winding 1 connected in series with the low windings 2, 3C, 4c and 5q, the low windings being located at the ends where radial stray fields occur.

Transformers in which both high-voltage and low-voltage windings are formed as described above can be constructed using a smaller distance between the windings. This: ·, <·! i will follow this in more detail with reference to Figures 6-8.

7 62430

In transformers whose input is in the middle of a high-voltage winding (A in Fig. 6), there may be a considerable voltage rise across the channel between the high-voltage and low-voltage windings under the influence of the surge voltage. This is because the voltage at the center B of the low voltage winding initially rises (capacitively) to a value having the same polarity as the effective voltage. It then oscillates to the opposite polarity as shown in Figure 7. Voltage Ae The channel between the high voltage and low voltage windings can then rise to a value of 130-140% of the effective voltage. This voltage becomes greatest when the oscillation at point B contains little overshoot, as is the case in Figure 7, which shows a distinct frequency. Due to the symmetrical structure in Figure 6, this happens regularly.

By reducing the voltage Ae, the insulation distance between the high voltage and low voltage windings can also be reduced and thus the material cost for both the core and the windings can be reduced.

The reduction of the voltage Ae is achieved by constructing a high voltage winding from the strip / sheet windings, as described above. In this case, the voltage in the middle of the high-voltage winding, with respect to the channel between the high-voltage and low-voltage windings, will decrease by a fraction of the voltage of the first winding (e.g. by $ 25). This causes point B to charge correspondingly less capacitively, and consequently Ae also decreases accordingly.

The voltage Ae can be further reduced by constructing a low voltage winding as shown in Fig. 8 by reference L. It has a central part made of one or more strip windings, while the outer parts are disc windings.

In this case, the part of the low-voltage winding which is located closest to the center of the high-voltage winding is made to be close to the potential of the second low-voltage output. This is approximately the value of O under the influence of the supply voltage. Thus, the symmetry is broken so that the voltage fluctuation in the low voltage winding receives several harmonic overvoltages and thus remains smaller.

It is also possible to connect the windings of a low-voltage winding in series so that the part of the winding which is closest to the high-voltage winding

Claims (6)

The center of the test is obtained at the maximum possible voltage under the influence of the induced test voltage. In this case, said part is located as close to the potential of the high-voltage winding as possible. Both of these measures are likely to reduce the requirement for an insulation distance between the high-voltage and low-voltage windings. Figure 8 shows a transformer in which both the high voltage winding H and the low voltage winding L are constructed according to the invention. Also in a three-phase transformer, all three windings are built according to the same principle. A device for use in windings of high-current transformers or reactors comprising windings whose conductors have an axial dimension corresponding to the height of the winding, characterized in that the height and speed of the windings connected in series decrease either to the winding following the winding or in groups from the center of the winding. Device according to Claim 1, in particular for use in high-voltage windings, characterized in that the halves of the winding are connected in parallel and their outer ends are connected to ground and / or to a star point. Device according to Claim 2, characterized in that the height of the winding and its speed decrease approximately in a geometric series. Device according to Claim 1, in particular for low-voltage windings, characterized in that all the windings of the winding are connected in series. Device according to Claim 4, characterized in that the central winding comprises at least a quarter, possibly more than half, of the height of the winding. Device for transformers according to claim 1, characterized by a combination in which the high-voltage winding is