Transformer Design

1
Introduction

Constituents of transformer:
i. Introduction
ii. Constructional Details
iii. Transformer Core
iv. Transformer Winding
v. Output Equation
vi. Optimum Design

2
Constructional Details

3
Constructional Details

4
Constructional Details

5
Constructional Details
• The requirements of magnetic material
are,
 High permeability
 Low reluctance
 High saturation flux density
 Smaller area under B-H
curve
• For small transformers, the
laminations are in the form of E,I, C and
O as shown in figure.
• The percentage of silicon in the steel is
about 3.5. Above this value the steel
becomes very brittle and also very
6

hard to cut.
Transformer Core

7
Transformer Core
Transformer Core
Core Type Transformer Shell type transformer
The core has only one magnetic It has two magnetic circuits.
circuit.
Core has two limbs. Core has three limbs.
It has less mechanical protection It has better mechanical
to coil. protection to coil.
It has better cooling since more Cooling is not very effective.
surface is exposed to
atmosphere.
This transformer is easy to This transformer is not easy to
repair. repair.
The winding is surrounded Core is surrounded considerable
considerable part of core. part of winding of transformer.
Transformer Core

10
Transformer Core

11
Transformer Core

12
Transformer Core

13
Transformer Core
Stepped Core Construction

14
Transformer Core
Stepped Core Construction

• For large size transformer it is wasteful to use

rectangular cylindrical coils so circular cylindrical coils are
preferred. 15

• A common improvement on square core is to employ

cruciform core which demand at least two size of core
stripes.
Transformer Core
Stepped Core Construction

16
Transformer Core

Hw
Ww

17
Transformer Winding

 Types of transformer windings are,

Concentric
Sandwich
The winding consist of the current-
carrying conductors wound around
the sections of the core, and these
must be properly insulated,
supported and cooled to withstand
operational and test conditions. 18
Transformer Winding

Concentric winding
Transformer Winding

Sandwich Winding
Transformer Winding

Hw

Ww

21
Classification on Service
Details Distribution transformer Power Transformer

Capacity Upto 500kVA Above 500kVA

Voltage rating 11,22,33kV/440V 400/33kV;220/11kV…etc.,

Connection Δ/Y, 3φ, 4 Wire Δ/Δ; Δ/Y, 3φ, 4 Wire

Flux Density Upto 1.5 T Upto 1.7 T

Current Density Upto 2.6 A/mm2 Upto 3.3 A/mm2

Load 100% for few Hrs, Part load for Nearly on Full load
some time, No-load for few Hrs

Ratio of Iron Loss to Cu loss 1:3 1:1

Regulation 4 to 9% 6 to 10%

Cooling Self oil cooled Forced Oil Cooled 22

Output Equation of Transformer
• It relates the rated kVA output to the area of core & window
• The output kVA of a transformer depends on,
Flux Density (B) – related to Core area
Ampere Turns (AT) – related to Window area
• Window – Space inside the core – to accommodate primary &
secondary winding
Let,
N- No. of turns in transformer winding
f – Frequency of supply
23

Induced EMF/Turn , Et=E/N=4.44 f m

Output Equation of Transformer
• Window in a 1-phase transformer contains one primary & one
secondary winding.
Conductor area in Window
Window Space factor,K w 
Total area of Window
Ac
Window Space factor,K w 
Aw
 Conductor area in window, A c  K w A w  (2)
Current Density () is same in both the windings
Ip I
   s  (3)
ap a s
24
Output Equation of Transformer
Ip
Is
ap  ; as 
 
If we neglect magnetizing MMF, then (AN)primary = (AN)secondary
AN=IpNp=IsNs  (4)
Total Cu. Area in window, Ac=Cu.area of pry wdg + Cu.area of sec wdg
 No. of pry turns X area   No. of sec turns X area 
 
  
of X - section of pry conductor of X - section of sec conductor
 N p a p  N s as
 N p a p  N s as
Ip Is
 Np  Ns
 

1

N I  Ns Is
p p  25


1
AN  AN   2 AN  (5)
 
Output Equation of Transformer
Therefore, equating (2) & (5),
2 AN
KwAw 

1
AT  AN  K w A w  (6)
2
kVA rating of 1φ transformer is given by,
S  Q  VpIp  10-3  EpIp  10-3
Ep  E
 NpIp  10 -3
 from (1), Et  N 
Np
 Et  AT  103  (6)
1
 4.44 fm . K w A w  103 26
2
 2.22  fm  K w A w  103
Output Equation of Transformer
m
We know that, Bm  and m  Bm A i
Ai
 Q  2.22  f Bm A i A w K w   103 kVA
Three phase transformer:
• Each window has 2 primary & 2 Secondary
windings.
• Total Cu. Area in the window is given by, Hw

2N p I p 2Ns I s
Ac  
Ww

 
4 AT
Ac   (7)

4 AT
Compare (2) & (7),   K w Aw
27

K w Aw
AT 
4
Output Equation of Transformer
kVA rating of 3φ transformer,

Q  3 E p I p 103
 3Et NpIp  10-3
 3 Et  AT 103
1
 3  4.44  fm .  K w A w 103
4
 3.33 f Bm Ai A wK w 103 kVA

28
EMF per Turn

Ratio of Specific magnetic loading m

  r 
 to Electric loading  AT
Q  VpIp  103
 4.44 f m TpIp  103
 4.44 f m ( AT )  103
m
 4.44 f m  103
r
Q r Q  r  103
m 
2
3

4.44 f  10 4.44 f
Q  r  103 29
m 
4.44 f
EMF per Turn
w.k.t , E t  4.44 f m
Q  r  103 Q
 4.44 f  4.44 f  4.44 f  r  10 .
3

4.44 f 4.44 f
 4.44 f  r  103  Q
Et  K  Q

where, K  4.44 f  r  103

• K depends on the type, service condition & method of construction of

transformer.

30
EMF per Turn

Transformer Type Value of K

1φ Shell Type 1.0 to 1.2
1φ Core Type 0.75 to 0.85
3φ Shell Type 1.2 to 1.3
3φ Core Type Distribution Xmer 0.45 to 0.5
3φ Core Type Power Xmer 0.6 to 0.7

31
Optimum Design

• Transformer may be designed to make one of the following

quantitates as minimum.
i. Total Volume
ii. Total Weight
iii. Total Cost
iv. Total Losses
• In general, these requirements can be contradictory & it is
normally possible to satisfy only one of them.
m
• All these quantities vary with r 
AT 32
Optimum Design
Design for Minimum Cost

Let us consider a single phase transformer.

 Q  2.22  f Bm A i A w K w   103 kVA
Q  2.22  f Bm Ai A c  103 kVA  Ac  K w A w 
Assuming that  & B are constant, Ac.Ai – Constant
Let, A A  M2  (1)
c i
In optimum design, it aims to determining the minimum value of
total cost. r
m
AT
m  B m A i 33
1 Ac
AT  K w A w  
2 2
Optimum Design
Design for Minimum Cost
Bm A i 2 Bm A i
r 
Ac Ac
2
A i 2 Bm
r
Ac 
Ai 
 r    ( 2)
Ac 2 Bm

β is the function of r alone [δ & Bm – Constant]

From (1) & (2),
Ai  M  & Ac 
M  A A  M 
c i
2

 34
Optimum Design
Design for Minimum Cost
Let, Ct - Total cost of transformer active materials
Ci – Cost of iron
Cc – Cost of conductor
pi – Loss in iron/kg (W)
pc – Loss in Copper/kg (W)
li – Mean length of flux path in iron(m)
Lmt – Mean length of turn of transformer winding (m)
Gi – Weight of active iron (kg)
Gc – Weight of Copper (kg)
gi – Weight/m3 of iron
gc – Weight/m3 of Copper
C t  C i  C c  c iGi  c c G c
35
Optimum Design
Design for Minimum Cost
Ct  ci g ili Ai  ccg c Lmt Ac
where, ci & cc  specific cos ts of iron and copper respectively.
M
C t  c i giliM   c c gc Lmt

Differeniating Ct with respect to β,
d 1 1 / 2 1
C t  c i g i li M()  c c g c Lmt M 3 / 2
d 2 2
d
For minimum cost, Ct  0
d
1 1
c i giliM() 1/ 2  c c gc Lmt M 3 / 2
2 2
c i gili  c c gc Lmt 1
Ac 36

c i gili  c c gc Lmt
Ai
Optimum Design
Design for Minimum Cost
c i gili A i  c c gc Lmt A c
ci G i  cc G c
Ci  C c
Hence for minimum cost, the cost of iron must be equal to the cost of
copper.
Similarly,
For minimum volume of transformer,
Gi Gc
Volume of iron = Volume of Copper  or G i  gi
gi gc Gc gc

For minimum weight of transformer,

Weight of iron = Weight of Conductor Gi  G c

For minimum loss, 37

Iron loss = I2R loss in conductor Pi  x Pc 2
Optimum Design
Design for Minimum Loss and Maximum Efficiency

Total losses at full load = Pi+Pc

At any fraction x of full load, total losses = Pi+x2Pc
If output at a fraction of x of full load is xQ.
xQ
Efficiency, x  xQ  P  x 2 P
i c d x
0
Condition for maximum efficiency is, dx
dx

 
xQ  Pi  x 2 Pc Q - Q  2xPc xQ
0
dx 
xQ  Pi  x Pc2
2

xQ  P  x P Q  Q  2xP xQ

i
2
c c

xQ  P  x P   Q  2xP x
i
2
c c
38
xQ  Pi  x Pc  xQ  x Pc  x Pc
2 2 2

Pi  x 2 Pc
Optimum Design
Design for Minimum Loss and Maximum Efficiency

Variable losses = Constant losses

Pi piG i

Pc pcG c
piG i Gi 2 pc
x 
2
or x for maximum efficiency
pcG c Gc pi

39
Design of Core

• Core type transformer : Rectangular/Square /Stepped cross

section
• Shell type transformer : Rectangular cross section

40
Design of Core
Rectangular Core

• For core type distribution transformer & small power

transformer for moderate & low voltages
Depth
 1.4 to 2
Width
Rectangular coils are used.
• For shell type,
Width of Central limb
 2 to 3
Depth of Core

41
Design of Core
Square & Stepped Core

• Used when circular coils are required for high voltage

distribution and power transformer.
• Circular coils are preferred for their better mechanical
strength.

• Circle representing the inner surface of the tubular form

carrying the windings (Circumscribing Circle) 42
Design of Core
Square & Stepped Core

• Dia of Circumscribing circle is larger in Square core than

Stepped core with the same area of cross section.
• Thus the length of mean turn(Lmt) is reduced in stepped core
and reduces the cost of copper and copper loss.
• However, with large number of steps, a large number of
different sizes of laminations are used.

43
Design of Core
Square Core

Let, d - diameter of circumscribing circle

a – side of square
Diameter, d  a 2  a 2  2a 2  2  a
d
a
2 2
 d 
Gross core area, Agi  Area of square  a   
2

 2
Agi  0.5d 2
Let the stacking factor, Sf=0.9. 44

Net core area, Ai  0.9  0.5d 2  0.45d 2

Design of Core
Square Core
• Gross core area includes insulation area
• Net core area excludes insulation area
Area of Circumscribing circle is 
d2
4
Ratio of net core area to Area of Circumscribing circle is
2
0.45d
 0.58
 2
d
4
Ratio of gross core area to Area of Circumscribing
2
circle is
0.5d 45
 0.64

d2
4
Design of Core
Square Core

Useful ratio in design – Core area factor,

Net Core area

KC 
Squareof Circumscribing Circle
A i 0.45d 2
 2  2
 0.45
d d

46
Design of Core
Stepped Core or Cruciform Core

Let, a – Length of the rectangle

b – breadth of the rectangle
d – diameter of the circumscribing circle and diagonal
of the rectangle.
θ – Angle b/w the diagonal and length of the rectangle.
b • The max. core area for a given ‘d’ is
(a-b)/2
obtained by the max value of ‘θ’
• For max value of ‘θ’,
d b a dA gi
θ 0
d
(a-b)/2
• From the figure,
a 47
a cos     a  d cos 
d
b
sin    b  d sin 
d
Design of Core
Stepped Core or Cruciform Core
b

Two stepped core can be divided in to 3 rectangles.

(a-b)/2
Referring to the fig shown,
d
ab ab b a
Gross core area,Agi  ab   b   b θ
 2   2 
(a-b)/2
2(a  b)
 ab  b
2
a
 ab  ab  b 2  2ab  b 2
On substituting ‘a’ and ‘b’ in the above equations,
A gi  2(d cos )(d sin )  (d sin ) 2
A gi  2d 2 cos  sin   d 2 sin 2 
A gi  d 2 sin 2  d 2 sin 2 
For max value of ‘θ’, 48
dA gi
0
d
Design of Core
Stepped Core or Cruciform Core
dA gi
i.e.,  d 2 2 cos 2  d 2 2 sin  cos   0
d
d 2 2 cos 2  d 2 2 sin  cos 
2 cos 2  sin 2
sin 2
2
cos 2
tan 2  2
2  tan 1 2

  tan 1 2  31.720

1
2
Therefore, if the θ=31.720, the dimensions ‘a’ & ‘b’ will give
maximum area of core for a specified ‘d’.
a 49
cos     a  d cos   a  d cos(31.720 )  0.85d
d
b
sin    b  d sin   b  d sin(31.720 )  0.53d
d
Design of Core
Stepped Core or Cruciform Core
Gross core area,
A gi  2ab  b 2
A gi  2(0.85d )(0.53d )  (0.53d ) 2
A gi  0.618d 2
Let stacking factor,S f  0.9,
Net core area,Ai  Stacking factor X Gross Core area
Ai  0.9  0.618d 2  0.56d 2
The ratios,
Net core area 0.56d 2
  0.71
Area of Circumscribing circle 
d2
4 50
Gross core area 0.618d 2
  0.79
Area of Circumscribing circle  2
d
4
Design of Core
Stepped Core or Cruciform Core
Core area factor,
Net Core area
KC 
Squareof Circumscribing Circle
A i 0.56d 2
 2  2
 0.56
d d
Ratios of Multi-stepped Cores,
Ratio Square Core Cruciform Core 3-Stepped Core 4-Stepped Core

Net core area 0.64 0.79 0.84 0.87

Area of Circumscribing circle

Gross core area 0.58 0.71 0.75 0.78

Area of Circumscribing circle
51

Core area factor,KC 0.45 0.56 0.6 0.62

Design of Core
Choice of Flux Density in Core
 Flux density decides,
 Area of cross section of the core
 Core loss
 Choice of flux density depends on,
 Service condition (DT/PT)
 Material used
Cold rolled Silicon Steel-Work with higher flux density
 Upto 132 kV : 1.55 wb/m2
 132 to 275kV: 1.6 wb/m2
 275 to 400kV: 1.7 wb/m2
Hot rolled Silicon Steel – Work with lower flux density 52
 Distribution transformer : 1.1 to 1.4 wb/m2
 Power Transformer : 1.2 to 1.5 wb/m2
Overall Dimensions
Single phase Core Type

Hy

Hw H

Hy

a
W

Dy

d Ww d
D
53
Overall Dimensions
Single phase Shell Type
a

2a Hw

Ww

Depth
b Over
winding

54
Overall Dimensions
Thee phase Core Type

Hy Hy
a

Hw H Hw H

Hy Hy

a
W

Dy Dy

d Ww d Ww d

D D 55
Design of Winding
 Transformer windings: HV winding & LV winding
 Winding Design involves:
 Determination of no. of turns: based on kVA rating & EMF per turn
 Area of cross section of conductor used: Based on rated current and Current density
 No. of turns of LV winding is estimated first using given data.
 Then, no. of turns of HV winding is calculated to the voltage rating.

VLV AT
No. of turns in LV winding, N LV  (or)
Et I LV
where, VLV  Rated voltage of LV winding 56

I LV  Rated Current of LV winding

Design of Winding
VHV
No. of turns in HV winding, N HV  N LV 
VLV
where,VHV  Rated voltage of HV winding

57
Cooling of transformers
• Losses in transformer-Converted in heat energy.
• Heat developed is transmitted by,
• Conduction
• Convection
• Radiation
• The paths of heat flow are,
• From internal hot spot to the outer surface(in contact with oil)
• From outer surface to the oil
• From the oil to the tank
• From tank to the cooling medium-Air or water.
Cooling of transformers
Methods of cooling:
1. Air Natural (AN)-upto 1.5MVA
2. Air Blast (AB)
3. Oil natural (ON) – Upto 10 MVA
4. Oil Natural – Air Forced (ONAF)
5. Oil Forced– Air Natural (OFAN) – 30 MVA
6. Oil Forced– Air Forced (OFAF)
7. Oil Natural – Water Forced (ONWF) – Power plants
8. Oil Forced - Water Forced (OFWF) – Power plants
Cooling of transformers
Transformer Oil as Cooling Medium

 Specific heat dissipation due to convection is,

1
  4
conv  40.3  W / m 2 .0 C
H
where,  - Temperature differenceof the surface relative to oil,0C
H  Height of dissipating surface, m
 The average working temperature of oil is 50-600C.
 For   200C & H  0.5 to 1m,
conv  80 to 100 W/m2 .0 C.
 The value of the dissipation in air is 8 W/m2.0C. i.e, 10 times less
than oil.
Cooling of transformers
Temperature rise in plain walled tanks

 Transformer wall dissipates heat in radiation & convection.

 For a temperature rise of 400C above the ambient temperature of 200C, the heat
dissipations are as follows:
 Specific heat dissipation by radiation,rad=6 W/m2 .0C
 Specific heat dissipation by convection, conv=6.5 W/m2 .0C
 Total heat dissipation in plain wall 12.5 W/m2 .0C
 The temperature rise,
Total losses P  Pc
  i
Specific heat  Heat dissipating  12.5 St
   
Dissipation   surface of tank 

St – Heat dissipating surface

 Heat dissipating surface of tank : Total area of vertical sides+ One half area of top
cover(Air cooled) (Full area of top cover for oil cooled)
Design of tanks with cooling tubes
• Cooling tubes increases the heat dissipation
• Cooling tubes mounted on vertical sides of the transformer would not
proportional to increase in area. Because, the tubes prevents the
radiation from the tank in screened surfaces.
• But the cooling tubes increase circulation of oil and hence improve the
convection
Design of tanks with cooling tubes
Let, Dissipating surface of tank – St
Dissipating surface of tubes – XSt
Loss dissipated by surface of the tank by radiation and convection = 6  6.5St  12.5St
Loss dissipatedby  135
  6.5   XSt  8.8XSt
tubes by convection 100
Total loss dissipated
  12.5St  8.8XSt  12.5  8.8X St  (1)
by walls and tubes 
Actual total area of tank wall s and tubes  St  XSt  St (1  X )
Design of tanks with cooling tubes
Total lossesdissipated
Loss dissipated per m of dissipating surface 
2

Total area
St (12.5  8.8 X ) (12.5  8.8 X )
Loss dissipated per m of dissipating surface 
2
  (2)
St (1  X ) (1  X )

Temperature rise in  Total loss

 
Transformer with cooling tubes  Loss Dissipated
Total losses,Ploss  Pi  Pc  (3)
Pi  Pc
From (1) and (3), we have,  
St (12.5  8.8 X )
Pi  Pc
(12.5  8.8 X ) 
 St
Pi  Pc 1  Pi  Pc 
8.8 X   12.5  X   12.5 
 St 8 .8   S t 
Design of tanks with cooling tubes
1  Pi  Pc  1  Pi  Pc 
Total area of cooling tubes  
 
 12.5  St    12.5St   (5)
8.8   St  8.8   
Let , lt  Length of tubes
d t  Diameter of tubes
 Surface area of tubes   d t lt
Total area of tubes
Total number of tubes, nt 
Area of each tube
1  Pi  Pc 
nt    12.5 S t   (6)
8.8 d t lt   
 Standard diameter of cooling tube is 50mm & length depends on the
height of the tank.
 Centre to centre spacing is 75mm.
C4

HT

D D
C3

WT

LT

Doc C1
C2
Design of tanks with cooling tubes
 Dimensions of the tank:
Let, C1 – Clearance b/w winding and tank along width
C2 - Clearance b/w winding and tank along length
C3 – Clearance b/w the transformer frame and tank at the bottom
C4 - Clearance b/w the transformer frame and tank at the top
Doc – Outer diameter of the coil.
Width of the tank, WT=2D+ Doc +2 C1 (For 3 Transformer)
= D+ Doc +2 C1 (For 1 Transformer)
Length of the tank, LT= Doc +2 C2
Height of the tank, HT=H+C3+ C4
Design of tanks with cooling tubes
 Clearance on the sides depends on the voltage &
power ratings.
 Clearance at the top depends on the oil height above
the assembled transformer & space for mounting the
terminals and tap changer.
 Clearance at the bottom depends on the space
required for mounting the frame.
Design of tanks with cooling tubes
Clearance in mm
Voltage kVA Rating
C1 C2 C3 C4

Up to 11kV <1000kVA 40 50 75 375

Upto 11 kV 1000-5000kVA 70 90 100 400

11kV – 33kV <1000kVA 75 100 75 450

11kV – 33kV 1000-5000kVA 85 125 100 475

Estimation of No-load Current
No-load Current of Transformer:
Magnetizing Component
Depends on MMF required to establish required flux
Loss Component
Depends on iron loss

70
Estimation of No-load Current
No-load current of Single phase Transformer
ly
Total Length of the core = 2lc
lC
Total Length of the yoke = 2ly lC

Here, lc=Hw=Height of Window

ly

ly= Ww=Width of Window

MMF for core=MMF per metre for max. flux density in core X Total length of
Core
= atc X 2lc= 2 atc lc
MMF for yoke=MMF per metre for max. flux density in yoke X Total length of
yoke
= aty X 2ly= 2 aty ly
Total Magnetizing MMF,AT0=MMF for Core+MMF for Yoke+MMF for joints 71

= 2 atc lc +2 aty ly +MMF for joints

The values of atc & aty are taken from B-H curve of transformer steel.
Estimation of No-load Current
No-load current of Single phase Transformer

Max. value of magnetizing current=AT0/Tp

If the magnetizing current is sinusoidal then,
RMS value of magnetising current, Im=AT0/√2Tp
If the magnetizing current is not sinusoidal,
RMS value of magnetising current, Im=AT0/KpkTp
The loss component of no-load current, Il=Pi/Vp
Where, Pi – Iron loss in Watts
Vp – Primary terminal voltage
Iron losses are calculated by finding the weight of cores and 72
yokes. Loss per kg is given by the manufacturer.
No-load current, I0  I m  Il
2 2
Welding Transformer

73
Welding Transformer

74
Welding Transformer

75