Regulation of a Transformer:

1. When a transformer is loaded with a constant primary voltage, the secondary voltage
decreases because of its internal resistance and leakage reactance.
Let 0V2 = secondary terminal voltage at no-load.
= E2 = E1K= KV1 because at no-load the impedance drop is negligible.
V2 = secondary terminal voltage on full-load.
The change in secondary terminal voltage from no-load to full-load is = 0V2 − V2. This
change divided by 0V2 is known as regulation ‘down’. If this change is divided by V2, i.e.,
full-load secondary terminal voltage, then it is called regulation ‘up’.
oV 2  V 2 oV 2  V 2
% regn ‘down’ =  100and % regn ‘up’ =  100
oV 2 V2
In further treatment, unless stated otherwise, regulation is to be taken as regulation
‘down’.
The change in secondary terminal voltage from no load to full-load, expressed as a
percentage of no-load secondary voltage is,
= νr cos φ ± νx sin φ (approximately)
Or more accurately
1
=  r cos    x sin    ( x cos  r sin  )2
200
% regn = νr cos φ ± νx sin φ ...approximately.
The lesser this value, the better the transformer, because a good transformer should keep
its secondary terminal voltage as constant as possible under all conditions of load.
2. The regulation may also be explained in terms of primary values.
In Fig. 1 (a) the approximate equivalent circuit of a transformer is shown and in Fig. 1 (b),
(c) and (d) the vector diagrams corresponding to different power factors are shown.
The secondary no-load terminal voltage as referred to primary is E′2 = E2/K = E1 = V1
and if the secondary full-load voltage as referred to primary is V ′ 2 (= V2/K) then

V 1  V 2/
%regn   100
V1
 Fig 1
From the vector diagram, it is clear that if angle between V1 and V ′2 is neglected, then the
value of numerical difference V1 − V ′ 2 is given by (I1 R01 cos φ + I1 X01 sin φ) for
lagging p.f.
I 1Ro1cos   I 1Xo1sin 
%regn   100 =νr cos φ + νx sin φ
V1
I 1Ro1  100 I 1Xo1  100
Where  vr and  vx
V1 V1
As before, if angle between V1 and V2′ is not negligible, then
1
%regn  (vr cos   vx sin  )  (vx cos  vr sin  )2
200
3. In the above definitions of regulation, primary voltage was supposed to be kept constant
and the changes in secondary terminal voltage were considered.
As the transformer is loaded, the secondary terminal voltage falls (for a lagging p.f.). Hence,
to keep the output voltage constant, the primary voltage must be increased. The rise in
primary voltage required to maintain rated output voltage from no-load to full-load at a
given power factor expressed as percentage of rated primary voltage gives the regulation of
the transformer.
Suppose primary voltage has to be raised from its rated value V1 to V1′, then

V 1/  V 1
%regn   100
V1
Example 1. A-100 kVA transformer has 400 turns on the primary and 80 turns on the
secondary. The primary and secondary resistances are 0.3 Ω and 0.01 Ω respectively and
the corresponding leakage reactance are 1.1 and 0.035 Ω respectively. The supply voltage is
2200 V. Calculate (i) equivalent impedance referred to primary and (ii) the voltage
regulation and the secondary terminal voltage for full load having a power factor of 0.8
leading.

Solution.

K 2 = 0.3 + 0.01/ 1 / 5  = 0.55 Ω

2
K = 80/400 = 1/5, R1 = 0.3 Ω, R01 = R1 + R2/

K 2 = 1.1 + 0.035/ 1 / 5  = 1.975 Ω

2
X01 = X1 + X2/

(i) Z01 = 0.55 + j 1.975 = 2.05 ∠74.44º

K 2 Z01 = 1 / 5  (0.55 + j 1.975) = (0.022 + j 0.079)

2
(ii) Z02 =

3
No-load secondary voltage = KV1 = (1/5) × 2200 = 440 V, I2 = 10 × 10 /440 = 227.3 A
Full-load voltage drop as referred to secondary
= I2 (R02 cos φ − X02 sin φ)
= 227.3 (0.022 × 0.8 − 0.079 × 0.6) = − 6.77 V
% regn. = − 6.77 × 100/440 = − 1.54
Secondary terminal voltage on load = 440 − (− 6.77) = 446.77 V

Losses in a Transformer:
In a static transformer, there are no friction or wind age losses. Hence, the only losses
occurring are:
(i) Core or Iron Loss: It includes both hysteresis loss and eddy current loss. Because the
core flux in a transformer remains practically constant for all loads (its variation being 1 to
3% from no-load to full-load).
The core loss is practically the same at all loads.

Hysteresis loss Wh = η B1.6 max f V watt; eddy current loss We = P B 2 max f 2 t 2 watt
These losses are minimized by using steel of high silicon content for the core and by using
very thin laminations. Iron or core loss is found from the O.C. test. The input of the
transformer when on no load measures the core loss.
(ii) Copper loss. This loss is due to the ohmic resistance of the transformer windings. Total
Cu loss= I12 R1 + I22 R2 = I12 R01 + I22 R02. It is clear that Cu loss is proportional to
 current 
2 2
or kVA . In other words, Cu loss at half the full-load is one-fourth of that at full-
load.
The value of Cu loss is found from the short-circuit test

Efficiency of a Transformer:
As is the case with other types of electrical machines, the efficiency of a transformer at a
particular load and power factor is defined as the output divided by the input–the two being
measured in the same units (either watts or kilowatts).
Efficiency =Output/Input
But a transformer being a highly efficient piece of equipment, has very small loss, hence it is
impractical to try to measure transformer, efficiency by measuring input and output. These
quantities are nearly of the same size. A better method is to determine the losses and then to
calculate the efficiency from ;
Efficiency =Output /(Output + losses)= Output/( Output + Cu loss + iron loss)
Or η =( Input- Losses)/ Input =1-(losses/ Input)
It may be noted here that efficiency is based on power output in watts and not in volt-
amperes, although losses are proportional to VA. Hence, at any volt-ampere load, the
efficiency depends on power factor, being maximum at a power factor of unity.
Efficiency can be computed by determining core loss from no-load or open-circuit test and
Cu loss from the short-circuit test.

Condition for Maximum Efficiency:

Cu loss = I12 R01 or I22 R02 = Wcu
Iron loss = Hysteresis loss + Eddy current loss = Wh + We = Wi
Considering primary side,
Primary input = V1I1 cos φ1

V 1I 1cos 1  losses V 1I 1cos 1  I 12 Ro1  Wi

η= 
V 1I 1cos 1 V 1I 1cos 1
 I 1R01 Wi
 1 
V 1cos 1 V 1I 1cos 1
Differentiating both sides with respect to I1, we get

dη Ro1 Wi
 0 
dI 1 V 1cos 1 V 1I 12 cos 1
For η to be maximum,
dη
 0 Hence, the above equation becomes
dI 1

Ro1 Wi
 or Wi = I12 R01 or I22 R02
V 1cos 1 V 1I 12 cos 1
Or
Cu loss = Iron loss

The output current corresponding to maximum efficiency is I2 = Wi / Ro2

It is this value of the output current which will make the Cu loss equal to the iron loss. By
proper design, it is possible to make the maximum efficiency occur at any desired load.
Note. (i) If we are given iron loss and full load Cu loss, then the load at which two losses
would be equal (i.e. corresponding to maximum efficiency) is given by

ironloss
= full load x ( )
F .L.culoss

(ii) The efficiency at any load is given by

x  full  loadKVA  p. f .
η=  100
( x  full  loadKVA  p. f .)  Wcu  Wi
Where x = ratio of actual to full-load kVA
Wi = iron loss in kW ; Wcu = Cu loss in kW.
Example 2. In a 25-kVA, 2000/200 V, single-phase transformer, the iron and full-load copper
losses are 350 and 400 W respectively. Calculate the efficiency at unity power factor on (i)
full load (ii) half full-load.
Solution.
(i) Full-load Unity p.f.
Total loss = 350 + 400 = 750 W
F.L. output at u.p.f. = 25 × 1 = 25 kW ; η = 25/25.75 = 0.97 or 97%
(ii) Half F.L. Unity p.f.

Cu loss = 400 × 1 / 2 = 100 W. Iron loss remains constant at 350 W, Total loss = 100 + 350
2

= 450 W.
Half-load output at u.p.f. = 12.5 kW
η= 12.5/(12.5 + 0.45) = 96.52%

Example 3. A 11000/230 V, 150-kVA, 1-phase, 50-Hz transformer has core loss of 1.6 kW
and F.L. Cu loss of 1.4 kW. Determine
(i) The kVA load for max. Efficiency and value of max. Efficiency at unity p.f.
(ii) The efficiency at half F.L. 0.8 p.f. leading.
Solution. (i) Load kVA corresponding to maximum efficiency is

ironloss 1.6
= F.L. kVA ×  150   160 KVA
F .L.culoss 1.4
Since Cu loss equals iron loss at maximum efficiency, total loss = 1.4 + 1.4 = 2.8 kW ;
output = 160 × 1 = 160 Kw
ηmax = 160/162.8 = 0.982 or 98.2%

(ii) Cu loss at half full-load = 1.6 × 1 / 2 = 0.4 kW ; Total loss = 1.4 + 0.4 = 1.8 Kw
2

Half F.L. output at 0.8 p.f.= (150/2) × 0.8 = 60 Kw

Efficiency = 60/(60 + 1.8) = 0.97 or 97%
All-day Efficiency:
The ordinary or commercial efficiency of a transformer is given by the ratio
Output in watts/Input in watts
But there are certain types of transformers whose performance cannot be judged by this
efficiency. Transformers used for supplying lighting and general network i.e., distribution
transformers have their primaries energized all the twenty-four hours, although their second
Aries supply little or no-load much of the time during the day except during the house
lighting period. It means that whereas core loss occurs throughout the day, the Cu loss occurs
only when the transformers are loaded. Hence, it is considered a good practice to design such
transformers so that core losses are very low. The Cu losses are relatively less important,
because they depend on the load. The performance of such is compared on the basis of energy
consumed during a certain time period, usually a day of 24 hours.
ηall-day = Output in kWh/Input in kWh (For 24 hours)
This efficiency is always less than the commercial efficiency of a transformer.
To find this all-day efficiency or (as it is also called) energy efficiency, we have to know the
load cycle on the transformer i.e., how much and how long the transformer is loaded during
24 hours. Practical calculations are facilitated by making use of a load factor.
Parallel Operation of Single-phase Transformers:
For supplying a load in excess of the rating of an existing transformer, a second transformer
may be connected in parallel with it as shown in Fig. 4. It is seen that primary windings are
connected to the supply bus bars and secondary windings are connected to the load bus-bars.
In connecting two or more than two transformers in parallel, it is essential that their
terminals of similar polarities are joined to the same bus-bars as in Fig. 4. If this is not done,
the two e.m.fs. induced in the secondaries which are paralleled with incorrect polarities, will
act together in the local secondary circuit even when supplying no load and will hence
produce the equivalent of a dead short-circuit as shown in Fig. 5.

Fig 4
There are certain definite conditions which must be satisfied in order to avoid any local
circulating currents and to ensure that the transformers share the common load in
proportion to their kVA ratings. The conditions are :
1. Primary windings of the transformers should be suitable for the supply system voltage
and frequency.
2. The transformers should be properly connected with regard to polarity.
3. The voltage ratings of both primaries and secondaries should be identical. In other
words, the transformers should have the same turn ratio i.e. transformation ratio.
4. The percentage impedances should be equal in magnitude and have the same X/R ratio
in order to avoid circulating currents and operation at different power factors.
5. With transformers having different kVA ratings, the equivalent impedances should be
inversely proportional to the individual kVA rating if circulating currents are to be avoided.
Of these conditions, (1) is easily comprehended; condition (2) is absolutely essential
(otherwise paralleling with incorrect polarities will result in dead short-circuit). There is
some latitude possible with conditions (3) and (4). If condition (3) is not exactly satisfied i.e.
the two transformers have slightly different transformation or voltage ratios, even then
parallel operation is possible. But due to inequality of induced e.m.fs. in secondary, there will
be even on no-load, some circulating current between them (and therefore between the
primary windings also) when secondary terminals are connected in parallel. When
secondary are loaded, this localized circulating current will tend to produce unequal loading
condition. Hence, it may be impossible to take full kVA output from the parallel connected
group without one of the transformers becoming over-heated.

Fig 5
If condition (4) is not exactly satisfied i.e. impedance triangles are not identical in shape and
size, parallel operation will still be possible, but the power factors at which the two
transformers operate will be different from the power factor of the common load. Therefore,
in this case, the two transformers will not share the load in proportion to their kVA ratings.
It should be noted that the impedances of two transformers may differ in magnitude and in
quality (i.e. ratio of equivalent resistance to reactance). It is worthwhile to distinguish
between the percentage and numerical value of an impedance. For example, consider two
transformers having ratings in the ratio 1 : 2. It is obvious that to carry double the current,
the latter must have half the impedance of the former for the same regulation. For parallel
operation, the regulation must be the same, this condition being enforced by the very fact of
their being connected in parallel. It means that the currents carried by the two transformers
are proportional to their ratings provided their numerical impedances are inversely
proportional to these ratings and their percentage impedances are identical.
If the quality of the two percentage impedances is different (i.e. ratio of percentage resistance
to reactance is different), then this will result in divergence of phase angle of the two currents,
with the result that one transformer will be operating with a higher and the other with a
lower power factor than that of the combined load.
(a) Case 1. Ideal Case
We will first consider the ideal case of two transformers having the same voltage ratio and
having impedance voltage triangles identical in size and shape.
Let E be the no-load secondary voltage of each transformer and V2 the terminal voltage ;
IA and IB the currents supplied by them and I-the total current, lagging behind V2 by an
angle φ (Fig. 6 (a).

Fig 6(a) Fig 6(b)

In Fig. 6 (b) a single triangle ABC represents the identical impedance voltage triangles of
both the transformers. The currents IA and IB of the individual transformers are in phase
with the load current I and are inversely proportional to the respective impedances.
Following relations are obvious.
I = IA + IB; V2 = E − IAZA = E − IBZB = E – IZAB
Also IAZA = IBZB or IA/IB = ZB/ZA
IA = IZB/(ZA + ZB) and IB = IZA /(ZA + ZB)
(b) Case 2. Equal Voltage Ratios:
Let us assume that no-load voltages of both secondary is the same i.e. EA = EB = E, and that
the two voltages are coincident i.e. there is no phase difference between EA and EB, which
would be true if the magnetizing currents of the two transformers are not much different
from each other. Under these conditions, both primaries and secondary of the two
transformers can be connected in parallel and there will circulate no current between them
on on-load.
However, if we neglect the magnetizing admittances, the two transformers can be
connected as shown by their equivalent circuits in Fig. 7. The vector diagram is shown in
Fig. 8.

Fig 7
From Fig. 7 (a) or (b) it is seen that it represents two impedances in parallel. Considering
all values consistently with reference to secondary, let
ZA, ZB = impedances of the transformers
IA, IB = their respective currents
V2 = common terminal voltage
I = combined current
It is seen that IAZA = IBZB = IZAB ...(i)
Where ZAB is the combined impedance of ZA and ZB in parallel.
1/ZAB = 1/ZA + 1/ZB
Hence ZAB = ZAZB/(ZA + ZB) ...(ii)
From equation (i), we get IA = IZAB/ ZA = IZB/(ZA + ZB) and IB = IZAB/ZB = IZA/ (ZA
+ ZB)
Multiplying both sides by common terminal voltage V2, we have
ZB ZA
V 2I A  V 2I Similarly V 2 I B  V 2 I
Z A  ZB Z A  ZB
3
Let V2I × 10 = S-the combined load kVA. Then, the kVA carried by each transformer is

ZB 1 ZA 1
SA  S S and SB  S S ...(iii)
ZA  Z B 1  Z A / ZB ZA  Z B 1  ZB / Z A

Hence, SA and SB are obtained in magnitude as well as in phase from the above vectorial
Expressions.

Fig 8
The above problem may be solved graphically, although somewhat more laboriously. As
shown in Fig.8, drawn IA and IB with an angular difference of (φA−φB) and magnitude
(according to some suitable scale) inversely proportional to the respective impedances.
Vector sum of IA and IB gives total combined current I. The phase angle and magnitude of
I will be known from the conditions of loading, so that angle φbetween V2 and I will be
known. Inserting this, the transformer currents IA and IB become known in magnitude and
phase with respect to V2.

Note. (a) In equation (iii) above, it is not necessary to use the ohmic values of resistances
and reactances, because only impedance ratios are required.
(b) The two percentage impedances must be adjusted to the same kVA in the case of
transformers of different rating.
(c) From equation (iii) above, it is seen that if two transformers having the same rating and
the same transformation ratio are to share the load equally, then their impedances should be
equal i.e. equal resistances and reactances and not numerical equality of impedances. In
general, for transformers of different ratings but same transformation ratio, their equivalent
impedances must be inversely proportional to their ratings if each transformer is to assume
a load in proportion to its rating. For example, as said earlier, a transformer operating in
parallel with another of twice the rating, must have an impedance twice that of the large
transformer in order that the load may be properly shared between them.
Example 4. Two 1-phase transformers with equal turns have impedances of (0.5 + j3) ohm
and (0.6 + j10) ohm with respect to the secondary. If they operate in parallel, determine
how they will share a total load of 100 kW at p.f. 0.8 lagging ?
Solution. ZA = 0.5 + j 3 = 3.04 ∠80.6° ZB = 0.6 + j 10 = 10.02 ∠86.6°
ZA + ZB = 1.1 + j13 = 13.05∠85.2°
Now, a load of 100 kW at 0.8 p.f. means a kVA of 100/0.8 = 125. Hence,
S = 125 ∠ − 36.9°

ZB 125  36.9o  10.0286.6o

SA  S   96  35.5o
ZA  Z B 13.0585.2o

= a load of 96 × cos 35.5° = 78.2 kW

ZA 125  36.9o  3.0480.6o

SB  S   29.1  41.5o
ZA  Z B 13.0585.2o

= a load of 29.1 × cos 41.5° = 21.8 Kw

Example 5. Two 1-phase transformers A and B rated at 250 kVA each are operated in
parallel on both sides. Percentage impedances for A and B are (1 + j 6) and (1.2 + j 4.8)
respectively. Compute the load shared by each when the total load is 500 kVA at 0.8 p.f.
lagging.
Solution.

ZA 1  j6 ZB 1.2  j 4.8
  0.552.1o ,   0.45  2.5o
Z A  Z B 2.2  j10.8 Z A  Z B 2.2  j10.8

ZB
SA  S  500  36.9O  0.45  2.5O  225  39.4O
Z A  ZB

ZA
SB  S  500  36.9O  0.552.1O  275  34.8O
Z A  ZB

Obviously, transformer B is overloaded to the extent of (275 − 250) × 100/250 = 10%. It

carries (275/500) × 100 = 55% of the total load.
(c) Case 3. Unequal Voltage Ratios:
In this case, the voltage ratios (or transformation ratios) of the two transformers are
different. It means that their no-load secondary voltages are unequal. Such cases can be more
easily handled by phasor algebra than graphically.
Let EA, EB = no-load secondary e.m.f.s of the two transformers.
ZL = load impedance across the secondary.
The equivalent circuit and vector diagram are also shown in Fig. 9 and 10.

Fig 9
It is seen that even when secondary are on no-load, there will be some cross-current in
them because of inequality in their induced e.m.fs. This circulating current IC is given by
IC = (EA − EB)/(ZA + ZB) ...(i)
As the induced e.m.fs. of the two transformers are equal to the total drops in their
respective circuits.
EA = IA ZA + V2 ; EB = IBZB + V2
Now, V2 = IZL = (IA + IB) ZL
Where ZL = load impedance
EA = IA ZA + (IA + IB) ZL ...(ii)
EB = IBZB + (IA + IB) ZL ...(iii)
EA − EB = IA ZA − IB ZB ...(iv)
IA = [(EA − EB) + IB ZB]/ZA

Fig 10
Substituting this value of IA in equation (iii), we get
EB = IBZB + [{(EA − EB) + IBZB}/ZA +IB]/ ZL
IB = [EB ZA − (EA − EB)ZL ]/[ZAZB + ZL.(ZA + ZB)] ...(v)
From the symmetry of the expression, we get
IA = [EAZB + (EA − EB)ZL[ZAZB + ZL(ZA + ZB)] ...(vi)
E A Z B  EB Z A
Also, I = IA + IB =
Z AZ B  Z L ( Z A  Z B )

By multiplying the numerator and denominator of this equation by 1/ZAZB and the result
by ZL we get
E A / Z A  EB Z B
V2 = IZL =
1 / Z A  1 / ZB  1 / ZL
The two equations (v) and (vi) then give the values of secondary currents. The primary
currents may be obtained by the division of transformation ratio i.e. K and by addition (if
not negligible) of the no-load current.
If ZA and ZB are small as compared to ZL i.e. when the transformers are not operated near
short circuit conditions, then equations for IA and IB can be put in a simpler and more easily
under-stand able form. Neglecting ZAZB in comparison with the expression ZL(ZA + ZB),
we have

E AZ B E  EB
IA   A ...(vii)
ZL (Z A  ZB ) Z A  ZB

EB Z A E  EB
IB   A ...(viii)
ZL (Z A  ZB ) Z A  ZB

The physical interpretation of the second term in equations (vii) and (viii) is that it represents
the cross-current between the secondaries. The first term shows how the actual load current
divides between the loads. The value of current circulating in transformer secondaries (even
when there is no load) is given by* IC = (EA − EB)/(ZA + ZB) assuming that EA > EB. It
lags behind EA by an angle α given by tan α = (XA + XB)/(RA + RB). If EA + EB the ratios
of the currents are inversely as the impedances (numerical values).
If in Eq. (iv) we substitute IB = I − IA and simplify, we get

IZ B E  EB
IA   A ...(ix)
(Z A  ZB ) Z A  ZB

Similarly, if we substitute IA = I − IB and simplify, then

IZ A E  EB
IB   A ...(x)
(Z A  ZB ) Z A  ZB

In a similar manner, value of terminal voltage V2 is given by

E AZ B  EB Z A  IZ AZ B
V2 = IZL = ...(xi)
Z A  ZB

These expressions give the values of transformer currents and terminal voltage in terms of
the load current. The value of V2 may also be found as under :
IA = (EA − V2)/ZA = (EA − V2)YA;IB = (EB − V2)YB
I = V2YL = IA + IB or V2YL = (EA − V2)YA + (EB − V2)YB
E AYA  EBYB
V2(YL + YA + YB) = EAYA + EBYB or V2 = ...(xii)
YL  YA  YB
Eq. (xi) gives V2 in terms of load current. But if only load kVA is given, the problem
becomes more complicated and involves the solution of a quadratic equation in V2.
Now, S = V2I. When we substitute this value of I in Eq. (xi), we get

E AZ B  EB Z A  SZ AZ B / V2
V2=
(Z A  Z B )

Or V2 2 (ZA + ZB) − V(EAZBEBZA) + SZAZB = 0

When V2 becomes known, then IA and IB may be directly found from

V2 = EA − IAZA and V2 = EB − IBZB.
Note. In the case considered above, it is found more convenient to work with numerical
values of impedances instead of % values.
Example 6. Two transformers A and B are joined in parallel to the same load. Determine the
current delivered by each transformer having given : open-circuit e.m.f. 6600 V for A and
6,400 V for B. Equivalent leakage impedance in terms of the secondary = 0.3 + j3 for A and
0.2 + j1 for B. The load impedance is 8 + j6.
Solution.

E A Z B  ( E A  EB ) Z L
IA 
Z AZ B  Z L ( Z A  Z B )

Here EA = 6,600 V ; EB = 6,400 V ; ZL = 8 + j 6; ZA = 0.3 + j 3 ; ZB = 0.2 + j 1

6600(0.2  j1)  (6600  6400)(8  j6)

IA 
(0.3  j3)(0.2  j1)  (8  j6)(0.3  j3  0.2  j1)
117 − j 156 = 195A in magnitude
Similarly,

EB Z A  ( E A  EB ) Z L
IB 
Z AZ B  Z L ( Z A  Z B )

6400(0.3  j3)  (6600  6400)(8  j6)

IB 
(0.3  j3)(0.2  j1)  (8  j6)  (0.5  j 4)
= 349 − j 231 = 421 A (in magnitude)