ARUSHA TECHNICAL COLLEGE

ELECTRICAL ENGINEERING DEPARTMENT

1.1 MOTOR SELECTION
When selecting motor for particular application, the criteria such as motor characteristics,
power supply, installation requirement and environmental conditions need to be considered.
Some common criteria to be considered when selecting a moto are as follows;
a) power rated
The rated power of a motor refers to the output power. When selecting a motor for a
particular application, the output power required should be computed based on the
torque and speed of the load and drive loses between the motor and the load.
b) Rated speed
The operating speed of a motor depends on the number of poles. The highest speed is
achieved with two pole motor. Therefore, for high speed application 2 pole motor s
should be selected. For low speeds motor with 4 or more pole can be used with
suitable transmission drive to achieve the required speed
c) Power supply Voltage and Frequency
The power supply to the motor can be single or three phase. The normal power supply
voltage in Tanzania is 220V for single phase and 380Vfor three phase, the supply
frequency is 50Hz.
d) Motor Characteristics
The most important motor characteristics that need to be considered are toque,
starting current and slip. Depending on the type of load, a motor with appropriate
characteristics has to be selected.
Most loads can be classified into one of three main type such as;
i. constant Torque
 Conveyors
 cranes and hoist
 extruders
ii. Variable Torque
 centrifugal pumps
 centrifugal fan
iii. Cylindrical loads
 presses
 reciprocating compressors

The following selection criteria are applied when sizing motors

i. operating torque for constant torque loads

ii. 100% load for variable torque loads
iii. Motor Breakdown for cylindrical loads
e) Service Factor
Service Factor (SF) is a measure of load above the rated capacity at which the motor can be
operated without overloaded or damaged for a short period of time. SF is 1 for standard
motor. If SF is 1.1 for particular motor, then that motor would be able to operate at 110% of

Electrical Machine selection by Melkior, U. F Page 1

its rated capacity occasionally when needed. If a motor with SF continuously operated above
100% load its life expectancy will be reduced.

f) Motor Duty Cycle

The duty cycle of a motor is the amount of time the motor can be operated out of every hour.
if the motor duty cycle is listed as continuous it means the motor can be operated 24 hours a
day and does not need to be turned off to cool down. if the duty cycle is rated 40minutes, it
means the motor can be safely operated for 40 minutes before it must be turned off to allow it
to cool.

Various duty cycle designations can be used to describe motor operating conditions. A typical
set of duty are listed in table

SI Continuous running duty The motor work at a constant load maintained for
sufficient time to allow the machine to reach thermal
equilibrium
S2 Short time duty The motor work at a constant load for a given time less
than that required to reach equilibrium, followed by a
time de-energized and rest for sufficient duration to re-
establish machine temperature with 2k of the coolant
temperature
S3 Intermittent periodic duty A sequence of identical duty cycle, each including a
time of operation at constant load and time de-
energized and at rest. in this duty the cycle is such that
the starting current does not significantly affect the
temperature rise
S4 Intermittent periodic duty Sequence of identical cycles, each cycle including
with starting significant starting time, time of operation at constant
load and time de-energized and at rest
S5 Intermittent periodic duty Sequence of identical cycles, each cycle consisting of a
with Electric Braking starting time, a time for operation at constant load, a
time of electric braking and a time de-energized and a
rest
S6 Continuous operation Sequence of identical cycles, each cycle consisting of a
periodic duty time of operation at constant load and a time of
operation at no load. There is no time de-energized and
rest.
S7 Continuous operation Sequence of identical cycles, each cycle consisting of a
periodic duty with electric starting time, a time for operation at constant load, a
braking time of electric braking. there is no time de-energized
and rest
S8 Continuous operation A sequence of identical duty cycle, each cycle
periodic duty with related consisting of a time of operation at constant load
load/speed changes corresponding to a predetermined speed of rotation,
followed by one or more times of operations at other
constant loads corresponding to different speed of
rotation (carried out for example by means of a
changes in number of poles in the case of induction
motor). There is no time de-energized and rest.

Electrical Machine selection by Melkior, U. F Page 2

S9 Duty with non-periodic A duty in which generally load and speed vary non-
load and speed variation periodically with the permissible operating range. this
duty includes frequently applied overload that may
greatly exceed the reference load
S10 Duty with Discrete A duty consisting of a specific number of discrete
constant loads and speed values of load (or equivalent loading) and if applicable
speed. each load/speed combination being maintained
for sufficient time to allow the machine to reach
thermal equilibrium. the minimum load within duty
cycle may have the value of zero (no- load or de-
energized and at rest)

g) Motor mounting configuration

Motors are normally designed for foot mounting or flange mounting to suit various
application

h) Method of Speed Control

Motor can be designed to operate at a constant speed, multiple speeds with multiple set of
windings at variable speed. The most common method of speed variation is using and
inverter to adjust the power supply frequency to the motor

i) Motor Insulation and Environmental Temperature

Motors are designed with different insulation classes so as they can be operated under
different environmental temperature. the maximum motor operating temperature for
commonly used insulation class are summarized as follows.

Insulation IEC 60085 Maximum Operating Allowable temperature Hot sport

Class Temperature allowed 0C rise at full load for allowance
service factor 1.0 (0C)
0
motors ( C)
A 105 105 60 5
B 130 130 80 10
F 155 155 105 10
H 180 180 125 15
R 220 220 150 30

The maximum operating temperature is determined by adding the rate ambient temperature
(normally 40 0C), the maximum allowable temperature rise and the hot spot allowance (hot
spot is a point at the centre of the motor ‘s windings where temperature is higher)

Maximum Operating Temperature (0C)=Rated Ambient Temperature + Allowable

Temperature Rise + Hot spot allowance

j) Motor Enclosure
The common motor enclosures are as follows

Electrical Machine selection by Melkior, U. F Page 3

 i. Totally enclosed fan cooled (TEFC) – designed to prevent exchange of air between
the inside and outside of the frame. a fan is attached to the shaft and blows air over
the frame during operation to remove heat.
ii. Open Drip Proof (ODP) –allows air to circulate through the winding for cooling but
prevent drops of liquid from falling into motor within 15 degree angle of vertical
typically used for indoor application in relatively clean, dry locations.
iii. Totally Enclosed Non-Ventilated (TENV) – similar to TEFC motor, but has no
cooling fan and relies on natural convection for cooling. suitable for use where motors
exposed to dirty or dampness
iv. Totally Enclosed Air Over (TEAO) – dust tight motors designed to be used with
shaft mounted fan or belt driven fan and the motor must be mounted within the
airflow of the fan.
v. Totally Enclosed Wash Down (TEWD)- designed to withstand high pressure
washing or other high humidity or wet environments. design for use in extremely
moist or chemical environment, but not for hazardous locations
vi. Explosion Proof (EXPL) –totally enclosed and is designed to withstand an explosion
of specific gas or vapor inside the motor casing and prevent the ignition outside the
motor by sparks, flashing or explosion.
vii. Hazardous Location (HAZ) –Motor designed to operate in a specific hazardous
locations

1.2 Motor Requirements for Special Applications

Some specific application requires motors with special design. a few such applications and
their requirement are listed;
a) crane and Hoist
 High starting torque
 high torque at low speed and low torque at high speeds
 acceleration and retardation have to be uniform
 steady braking of the motor against overhauling must be possible
 mechanical braking must be available for emergency conditions
 may require to work in dust environments
b) Mill Drive
 high starting torque
 capable of reverse rotation
 capable of four quadrant operation
 operate at wide range of speed
 duty cycle with frequent starts stops
 may require to work in dusty environments
 may need to withstand high ambient temperature

Electrical Machine selection by Melkior, U. F Page 4

 c) Machine Tools
 capable of high speed control
 capable of high speed operation
 duty cycle with frequent starts and stops
 able reverse direction
 low inertia for fast response
 able to provide precise positioning

1.3 Rating of Motor

The selection of motor for particular drive application based on the size of motor depends
upon the following two factors:
a) Maximum temperature rise for a given load.
b) Maximum torque required.

The size of motor and its rating are mainly dependent upon the raise in temperature. The
temperature raise in turn depends upon the type of insulation used.

a) Temperature raise of motor

The various losses takes place in any motor will be converted into heat. The heat thus
produced will increase the temperature of various parts of the motor. The increase in
temperature is mainly dependent on the following two factors:
a) Amount of heat developed internally at uniform rate.
b) The amount of heat dissipated from the surface of the motor.

In fact, the continuous rating of a machine is that rating for which the final temperature raise
is equal to or just below the permissible value of the temperature raise for the insulating
material used in protection of motor windings. When the machine is overloaded for such a
long time that its final temperature raise exceeds the permissible limit, it is likely to be
damaged. Sometimes, it will results immediate breakdown of insulating material which will
cause a sudden short circuit in the motor, which may also lead to a fire. Since temperature
raise is one of the chief features in fixing the size of motor. The temperature raise will be
high in the beginning and will decrease gradually with the passage of time and finally the
temperature of the motor attains a steady-state value. At this point, the heat produced and
dissipated will be equal. The above circumstances make the heating calculations very
complex and practically impossible unless certain assumptions are made as:
a) Heat developed, i.e., losses remains constant during temperature raise.
b) The heat dissipation is directly proportional to the difference in the temperature of
motor and cooling medium, i.e., Newton's law of cooling hold's good.
c) The temperature of cooling medium remains unchanged.
d) The motor is assumed to be a homogeneous mass having the same and uniform
temperature in all parts. It implies high thermal conductivity.
e) For the determination of an expression for the temperature raise of an electrical
machine after time ‘t’ seconds from the instance of switching it on.

Electrical Machine selection by Melkior, U. F Page 5

Let P is the electrical power converted into heat (W or J/sec), M is the mass of active parts of
motor (kg), S is the specific heat of material (J/kg/°C), O is the temperature raise above the
cooling medium or ambient temperature (°C), A is the surface area of cooling, (m 2 ), θf is the
final temperature raise with constant load (°C), and λ is the coefficient of cooling or the rate
of heat dissipation (W/m2 /°C raise). Now, let us assume that the machine attains a
temperature raise of θ°C above ambient temperature after ‘t’ seconds of switching on the
machine and further raise of temperature by dθ in very small time ‘dt’ seconds.
𝑑𝜃
The rate at which the loss takes place or the heat is absorbed by the motor= 𝑀𝑆 𝑑𝑡 j/sec
The rate at which heat is dissipated = 𝐴𝜃λ J/sec
But, the rate at which the electrical power converted into heat = the rate at which the heat is
absorbed + the rate at which the heat dissipated by the motor.
𝑑𝜃
𝑃 = 𝑀𝑆 𝑑𝑡 + 𝐴𝜃λ 1
𝑑𝜃
𝑃 − 𝐴𝜃λ = 𝑀𝑆 𝑑𝑡 2
𝑀𝑆
𝑑𝑡 = 𝑃−𝐴𝜃λ 𝑑𝜃
𝑀𝑆
∫ 𝑑𝑡 = ∫ 𝑃−𝐴𝜃λ 𝑑𝜃
−1
𝑡 = 𝑀𝑆 log 𝑒 (𝑃 − 𝐴𝜃λ) × ( 𝐴λ) + 𝐾 3
Where K is the integration constant. Initially, at time t = 0 sec, temperature raise θ = 0°C.
By substituting t = 0 and θ = 0 in Equation (3), we get the integration constant (K):
−1
0 = 𝑀𝑆 log 𝑒 (𝑃 − 0) × ( 𝐴λ) + 𝐾
𝑀𝑆
𝐾= log 𝑒 (𝑃)
𝐴λ
Substituting the value of K in equation 3
−1 𝑀𝑆
𝑡 = 𝑀𝑆 log 𝑒 (𝑃 − 𝐴𝜃λ) × ( 𝐴λ) + log 𝑒 (𝑃) 4
𝐴λ
−𝑀𝑆
𝑡= [log 𝑒 (𝑃 − 𝐴𝜃λ) − log 𝑒 𝑃]
𝐴λ
−𝑀𝑆 (𝑃−𝐴𝜃λ)
𝑡= log 𝑒 [ ]
𝐴λ 𝑃
−𝐴λ𝑡 (𝑃−𝐴𝜃λ)
= log 𝑒 [ ]
𝑀𝑆 𝑃
By applying exponential on both side, we get:
−𝐴λ𝑡
𝐴𝜃λ
𝑒 ( 𝑀𝑆 ) = 1 − 𝑃
−𝐴λ𝑡
𝐴𝜃λ ( )
= 1−𝑒 𝑀𝑆
𝑃
−𝐴λ𝑡
𝑃
𝜃 = 𝐴λ [1 − 𝑒 ( 𝑀𝑆 ) ] 5
When ‘t’ is infinity, ‘θ’ approaches to its final steady-state temperature ‘θf’. So, by
substituting t = ∞ and θ = θf in Equation (5), we get:
𝑃
𝜃𝑓 = 𝐴λ [1 − 𝑒 ∞ ]
𝑃
𝜃𝑓 = 𝐴λ 6
Substituting in Equation 6 into 5

Electrical Machine selection by Melkior, U. F Page 6

 −𝐴λ𝑡
( )
𝜃 = 𝜃𝑓 [1 − 𝑒 𝑀𝑆 ]
𝑀𝑆
Let’s 𝑇ℎ = is known as heating time constant of motor.
𝐴λ
−𝑡
( )
= 𝜃𝑓 [1 − 𝑒 𝑇ℎ
] 7

is the equation of temperature rise with time

The temperature raise time curve or heating curve is exponential in nature as shown in Figure
1

figure 1: temperature raise time curve or heating curve

at t=Th,
𝜃 = 𝜃𝑓 [1 − 𝑒 −1 ]
𝜃 = 0.632𝜃𝑓
Thus, heating time constant can be defined as follows: The heating time constant is the time
taken by the machine to attain 63.2% of its final steady temperature raise (θf). The heating
time constant of the conventional electrical machines is usually within the range of 0.5–3 for
4 h.

b) Cooling of motor
Let us assume, if the supply to the motor is switched off, after attaining the final steady
temperature raise of ‘θf”, the motor starts cooling. When the machine is switched off, no heat
is produced, therefore:
Heat absorbed + heat dissipated = 0
𝑑𝜃
𝑀𝑆 𝑑𝑡 + 𝐴𝜃λ′ = 0 Where λ = heat dissipation during cooling of motor.
𝑀𝑆𝑑𝜃 + 𝐴𝜃λ′ dt = 0
𝑀𝑆 𝑑𝜃
dt = − λ′ 𝐴 Integrate both sides
𝜃

Electrical Machine selection by Melkior, U. F Page 7

 𝑀𝑆 𝑑𝜃
∫ dt = − λ′ 𝐴 ∫ 𝜃
8

𝑀𝑆
t = − λ′ 𝐴 log 𝑒 𝜃 + 𝐾′ 9
Where K’ is the integration constant. The value of K’ is obtained by using the initial
conditions, when t = 0 and θ = θf, we get:
𝑀𝑆
0 = − λ′ 𝐴 log 𝑒 𝜃𝑓 + 𝐾′
𝑀𝑆
K ′ = λ′ 𝐴 log 𝑒 𝜃𝑓 10
Substituting Equation (10) in Equation (9):
𝑀𝑆 𝑀𝑆
t = − λ′ 𝐴 log 𝑒 𝜃 + λ′ 𝐴 log 𝑒 𝜃𝑓
−λ′ 𝐴 𝜃
t = log 𝑒 [𝜃 ] 11
𝑀𝑆 𝑓
Applying exponentials on both side
−λ′ 𝐴 log𝑒 [
𝜃
]
[ t] 𝜃𝑓
𝑒 𝑀𝑆 =𝑒
−λ′ 𝐴
𝜃 [ t] 𝑀𝑆
=𝑒 𝑀𝑆 for 𝑇𝑐 = −λ′ 𝐴 is known as cooling time constant.
𝜃𝑓
−t
[ ]
𝜃 = 𝜃𝑓 𝑒 𝑇𝑐 12
The above relation is the equation of cooling of motor. The cooling curve is exponentially
decaying in nature as shown in Figure 2.

Figure 2: Cooling curve

From the cooling equation, at time t = Tc:

𝜃 = 𝜃𝑓 𝑒 −1
𝜃 = 0.368𝜃𝑓
Thus, we can define the cooling time constant as the time required for cooling the machine
down to 36.8% of the initial temperature raise above the ambient temperature. The heating
and cooling curves follows an exponential law. Heating time constant and cooling time
constant may be different for the same machine and also the cooling time constant of rotating
machine is larger than its heating time constant, due to poorer ventilation conditions when the
machine cools.
Figure 3 (a) and (b) shows the heating and cooling curves of a motor for short-time and
intermittent loads.

Electrical Machine selection by Melkior, U. F Page 8

Figure 3: (a) Short-time load motor (b) intermittent-time load motor

Example 1:
An induction motor has a final steady-state temperature raise of 50°C when running at its
rated output. Calculate its half-hour rating for the same temperature raise if the copper losses
at the rated output are 1.5 times its constant losses. The heating time constant is 60 min.
Solution:
Given data:
Final steady temperature (θf) = 50°C.
Time constant (τh) = 60 min.
1
Rating (t) = 2hour = 30 min.
The copper loss = 1.5 × constant loss
𝑊𝑐𝑢 = 1.5 × 𝑊𝑖
let ‘P’ be the rated output,

Total loss at full load = Wcu +Wi.

But, the temperature raise is proportional to the losses.

∴ θ ∝ Wloss.

Let, θf be the temperature raise at full load.

𝜃𝑓1 be the temperature raise with short-time rating
𝜃𝑓 𝑊 +𝑊
𝑐𝑢 𝑖
= 1.5𝑥
𝜃𝑓1 2 +𝑊
𝑖
2.5
= 1.5𝑥 2 +1
The temperature raise after 30 min of operation should not exceed θf = 50°C. Now, from the
equation of temperature raise of motor:
−30
𝜃𝑓 = 𝜃𝑓1 (1 − 𝑒 60 )
50 = 0.393𝜃𝑓1
𝜃𝑓1 = 128.07 0C

2.5
𝜃𝑓 = 𝜃𝑓1 (1.5𝑥 2 +1)
2.5
50 = 128.07 (1.5𝑥 2 +1)

Electrical Machine selection by Melkior, U. F Page 9

𝑥 = 2.058
Hence, the half-hour rating of machine is 2.058 times its continuous rating.

Tutorial Question
1. A 10-kW motor has a heating time constant and cooling time constant of 45 and 70
min, respectively. The final temperature attained is 60°C. Find the temperature of
motor after 45 min full-load run and then switched of for 30 min.
2. The heating time constant of a 80-kW motor is 60 min. The temperature raise is 65°C
when runs continuously on full load. Find the half-hour rating of motor for the same
temperature raise. Assume that the losses are proportional to the square of the load
and the motor cools to ambient temperature between each load cycle.
3. The heating time constant and final steady temperature of a motor on continuous
running is 60 min and 40°C. Find out the temperature
a) after 25 min at this load,
b) after 45 min at this load,
c) if the temperature raise at half-hour rating is 40°C, find the maximum steady
temperature,
d) what will be the time required to increase the temperature from 25°C to 40°C
at one-and-half-hour rating.
4. The heating time constant of a motor is 90 min with 1-hr rating as 200 W. The
maximum efficiency of motor occurs at 80% of full load. Determine the continuous
rating of the motor.

1.4 Methods of Rating Motors

When load fluctuates over a given cycle, as in rolling mills, etc., the raise of motor is
determined accurately by finding the heating and cooling curves of motor, when working on
given cycle. The various methods for determining the rating of motor for continuous duty and
variable load are:
a) Equivalent current method.
b) Equivalent torque method.
c) Equivalent power method.

1.4.1. Equivalent current method

In this method, the actual current may be replaced by an equivalent current method (Ieq),
which produces the same losses in the motor as the actual current.

𝐼12 𝑡1 + 𝐼22 𝑡2 + 𝐼32 𝑡3 + ⋯ 𝐼𝑛2 𝑡𝑛

𝐼𝑒𝑞 = √
𝑡1 + 𝑡2 + 𝑡3 + ⋯ 𝑡𝑛
where I1, I2, I3, …, In be the load currents within short intervals of t1, t2, …, tn over a period
of time ‘T ’ seconds

Electrical Machine selection by Melkior, U. F Page 10

Figure 4 Load cycle for equivalent current method

1.4.2. Equivalent power method

In this method, if the load cycle is given in HP or kW verses time, then the motor rating can
be directly found as follows
𝑃12 𝑡1 + 𝑃22 𝑡2 + 𝑃32 𝑡3 + ⋯ 𝑃𝑛2 𝑡𝑛
𝑀𝑜𝑡𝑜𝑟 𝑅𝑎𝑡𝑖𝑛𝑔 = √
𝑡1 + 𝑡2 + 𝑡3 + ⋯ 𝑡𝑛

Figure 5 Load cycle for equivalent power method

Load changes uniformly; load cycle varies

1 2 2 2 2 1 2
𝑃 𝑡 + 𝑃 𝑡 + 𝑃 𝑡 + 𝑃 𝑡 +
𝑀𝑜𝑡𝑜𝑟 𝑅𝑎𝑡𝑖𝑛𝑔 = √3
1 1 2 2 0 3 2 4 3 𝑃5 𝑡5
𝑡1 + 𝑡2 + 𝑡3 + 𝑡4 + 𝑡5

Figure 6Load cycle for uniform load variation

Electrical Machine selection by Melkior, U. F Page 11

Note: If the power, load, or torque changes uniformly, then ∫ P2dt has to be taken for that
period.
If the load curve consisting of negative power, i.e., power returned to the source

𝑃2 − 𝑃1 2 2 2 𝑡4 𝑃4
2
𝑛 [𝑃1 + 𝑡] 𝑑𝑡 + 𝑃1 𝑡2 + 𝑃3 𝑡3 + ∫ [ 𝑡] 𝑑𝑡
√ 𝑡1 0 𝑡4
𝑚𝑜𝑡𝑜𝑟 𝑟𝑎𝑡𝑖𝑛𝑔 = ∫
0 𝑡1 + 𝑡2 + 𝑡3 +𝑡4

Figure 7 Load cycle for negative power

1.4.3. Equivalent torque method

This method is used to compute the motor heating rating effect, for short time and
intermittent loads where the torque is varying

Figure 8 Load cycle for equivalent torque method

T1, T2, T3, T4, and T5 be the load torques develop during the periods t1, t2, t3, t4, and t5
seconds now the equivalent torque can be calculated by considering time for one complete
cycle and RMS value of load torques at different times.

Electrical Machine selection by Melkior, U. F Page 12

 𝑇12 𝑡1 + 𝑇22 𝑡2 + 𝑇32 𝑡3 + 𝑇42 𝑡4 + 𝑇52 𝑡5
𝐸𝑞𝑢𝑖𝑣𝑎𝑙𝑒𝑛𝑡 𝑇𝑜𝑟𝑞𝑢𝑒 (𝑇) = √
𝑡1 + 𝑡2 + 𝑡3 + 𝑡4 + 𝑡5

Example
A motor operates continuously on the following load cycle. 20 kW for 10 sec, 10 kW for 15
sec, 30 kW for 5 sec, 50 kW for 20 sec, 40 kW for 10 sec, and idle for 5 sec. Draw the load
diagram and find the size of the motor required.

The rating of the motor = RMS value of the load

Figure 10 Load cycle

Tutorial
1. The load cycle of a motor in driving some equipment is as follows. 0−3 min 40 kW,
3−7 min No-load 7−12 min 30 kW, 12−15 min 20 kW, 15−18 min 50 kW. The load
repeated indefinitely. Draw the load cycle and suggest suitable continuous rating of
the motor.
2. A motor has to perform the following load cycle: Load rising uniformly from 0 to 100
kW in 10 s, constant load 300 kW for 5 sec, constant load 200 kW for 15 sec.
Regenerative braking power returned falling uniform from 50 to 0 kW in 5 s. Decking
period 4 s, motor stationary. Draw the load cycle and suggest a suitable continuous
rated motor.
3. A motor has the following load cycle. Load raising uniformly from 100 to 200 kW in
5 s. Continuous load 50 kW for 10 s regenerative braking kW returned to the supply
50 kW to 0 kW for 3 s and idle for 2 s. Draw the load diagram neatly for one cycle.
Find the size of continuously rated motor for the above duty. The load cycle is
repeated indefinitely.

Electrical Machine selection by Melkior, U. F Page 13

1.5 LOAD EQUALIZATION
1.5.1 Introduction
The load fluctuations take place in many of the industrial drives such as rolling mills,
planning machines presses, and reciprocating pumps, where the load on the motor varies
widely within a span of few seconds. The sudden and peak load requires very large current
from the supply results high voltage drop in the system or alternately would require very
large size of cables. It is very essential to smooth out fluctuating load is known as ‘load
equalization’. The load equalization involves the storage of energy during the off-peak period
and gives out during the peak load period. Load equalization process is commonly achieved
by means of a flywheel. A flywheel is a big wheel that is mounted on the same shaft of
motor, if the speed of the motor is not to be reversed or a heavy rotating body that acts as a
reservoir for absorbing and redistributing stored energy is also known as flywheel.

1.5.2 Function of flywheel

To operate the flywheel efficiently, the driving motor should have drooping speed
characteristics. During the light load, the acceleration of the flywheel is increased and it
stores the kinetic energy and at the time of peak load, the flywheel slows down and the stored
kinetic energy is given out to the load; so that, the demand of the load from the motor or
supply is reduced.

Figure 11 Flywheel

It is necessary that the motor used for load equalization should have drooping characteristics.
The flywheel is not used with motors having constant speed for example synchronous motor.
The torque developed by the motor and the load torque required as well as the speed
variations with time are shown in Figure 12.

Electrical Machine selection by Melkior, U. F Page 14

Figure 12 Motor torque, load torque, and speed variations against time

1.5.3 Flywheel calculations

Let us consider a flywheel is attached to a variable speed motor to achieve load equalization.
Let 𝑇𝐿 be the load torque (assumed constant during particular interval) in N-m. 𝑇𝑀 is the
motor torque in N-M, 𝑇𝐹 is the flywheel torque in N-m, 𝑇0 is the no-load torque in N-m, 𝜔0
is the motor speed on no-load in rad/sec, ω is the motor speed at any instant in rad/sec, and J
is the moment of inertia of flywheel in kg-m2.

𝑠 = 𝜔0 − 𝜔 Motor slip

Let us consider that the load on the motor is increasing; during this period, the flywheel will
decelerate and impart its stored kinetic energy to the load. The torque required to be supplied
by the motor

𝑇𝑀 = 𝑇𝐿 − 𝑇𝐹 1

The kinetic energy given by the flywheel when its speed reduced from 𝜔0 to ω is:
1
𝐾𝐸 = 2 𝐽(𝜔02 − 𝜔2 )

1
𝐾𝐸 = 𝐽(𝜔0 + 𝜔)(𝜔0 − 𝜔) 2
2

1
𝐾𝐸 = 2 𝐽(𝜔0 + 𝜔)𝑠

1
(𝜔0 + 𝜔) = 𝜔 Mean speed
2

𝐾𝐸 = 𝐽𝜔𝑠

The power given out by the flywheel = the rate of change of the energy given up by the
flywheel.
𝑑
𝑃 = 𝑑 𝐽𝜔𝑠
𝑡

Electrical Machine selection by Melkior, U. F Page 15

 𝑑𝑠
𝑃 = 𝐽𝜔 𝑑
𝑡

The flywheel Torque

𝑝𝑜𝑤𝑒𝑟 𝑔𝑖𝑣𝑒𝑛 𝑜𝑢𝑡 𝑏𝑦 𝑓𝑙𝑦𝑤ℎ𝑒𝑒𝑙
𝑇𝐹 = 𝜔

𝑑𝑆
𝐽𝜔
𝑑𝑡
= 𝜔

𝑑𝑆
𝑇𝐹 = 𝐽 𝑑𝑡 3

put equation 3 into equation 1

𝑑𝑆
𝑇𝑀 = 𝑇𝐿 − 𝐽 𝑑𝑡 4

If the slip, i.e., drop in speed limited to 10%, then the slip is proportional to the motor torque:

𝑆 ∝ 𝑇𝑀

𝑆 = 𝐾𝑇𝑀 5

put equation 5 into 4

𝑑𝐾𝑇𝑀
𝑇𝑀 = 𝑇𝐿 − 𝐽 𝑑𝑡

Rearrange
𝑑𝑇𝑀 𝑑𝑡
= 𝐽𝐾
𝑇𝐿 −𝑇𝑀

Integrating both sides

𝑑𝑇𝑀 𝑑𝑡
∫𝑇 = ∫ 𝐽𝐾
𝐿 −𝑇𝑀

𝑡
− 𝑙𝑜𝑔𝑒 (𝑇𝐿 − 𝑇𝑀 ) = 𝐽𝐾 + 𝐶

where C is proportionality constant. At time t = 0, the motor torque will be equals to the no-
load torque:

i.e., at t = 0, TM = T0.

The value of ‘C’ can be determined by using the initial conditions.

𝑜
− 𝑙𝑜𝑔𝑒 (𝑇𝐿 − 𝑇0 ) = 𝐽𝐾 + 𝐶

𝐶 = − 𝑙𝑜𝑔𝑒 ( 𝑇𝐿 − 𝑇0 )

Substituting ‘C’

Electrical Machine selection by Melkior, U. F Page 16

 𝑡
− 𝑙𝑜𝑔𝑒 (𝑇𝐿 − 𝑇𝑀 ) = 𝐽𝐾 − 𝑙𝑜𝑔𝑒 ( 𝑇𝐿 − 𝑇0 )

Rearrange
(𝑇 −𝑇 ) −𝑡
log 𝑒 [ ( 𝑇𝐿 −𝑇𝑀) ] = 𝐽𝐾
𝐿 0

Applying exponentials on both sides

−𝑡
(𝑇𝐿 −𝑇𝑀 )
= 𝑒 𝐽𝐾
( 𝑇𝐿 −𝑇0 )

−𝑡
(𝑇𝐿 − 𝑇𝑀 ) = ( 𝑇𝐿 − 𝑇0 )𝑒 𝐽𝐾

Now consider that the load is totally removed or decreasing, the motor starts accelerating and
so the KE is stored by the flywheel.
𝑑𝑆
Hence, the flywheel regains its normal speed; therefore, the slip decreases, i.e. 𝑑𝑡 is negative.
Now, motor torque will be:

𝑇𝑀 = 𝑇𝑜 + 𝑇𝐹

But,
𝑑𝑆
𝑇𝐹 = −𝐽 𝑑𝑡

𝑑𝑆
𝑇𝑀 = 𝑇𝑜 + 𝐽 𝑑𝑡

𝑆 ∝ 𝑇𝑀

𝑆 = 𝐾𝑇𝑀
𝑑𝐾𝑇𝑀
𝑇𝑀 = 𝑇𝑜 − 𝐽 𝑑𝑡

Rearrange and integrate

𝑑𝑇𝑀 𝑑𝑡
∫𝑇 = − ∫ 𝐽𝐾
𝑀 −𝑇𝐿

−𝑡
log 𝑒 (𝑇𝑀 − 𝑇𝐿 ) = + 𝐶2
𝐽𝐾

where ‘C2’ is integration constant. The value of constant can be obtained by substituting the
initial conditions
𝑎𝑡 𝑡 = 0, 𝑇𝑀 = 𝑇𝑀1

motor torque when load is decreased

0
log 𝑒 (𝑇𝑀1 − 𝑇0 ) = 𝐽𝐾 + 𝐶2

Electrical Machine selection by Melkior, U. F Page 17

𝐶2 = log 𝑒 (𝑇𝑀1 − 𝑇0 )

By substituting ‘C2’
−𝑡
log 𝑒 (𝑇𝑀 − 𝑇0 ) = 𝐽𝐾 + log 𝑒 (𝑇𝑀1 − 𝑇0 )

Rearrange and take exponential both sides

(𝑇 −𝑇 ) −𝑡
log 𝑒 [ (𝑇𝑀1 −𝑇𝐿) ] = 𝐽𝐾
𝑀 0

(𝑇 −𝑇𝑜 )
log𝑒 [ 𝑀1 ] −𝑡
(𝑇𝑀 −𝑇0 )
𝑒 = 𝑒 𝐽𝐾
−𝑡
(𝑇𝑀 − 𝑇0 ) = (𝑇𝑀1 − 𝑇0 )𝑒 𝐽𝐾
−𝑡
𝑇𝑀 = 𝑇0 + (𝑇𝑀1 − 𝑇0 )𝑒 𝐽𝐾

Example

A 15-HP, three-phase, eight-pole, and 50-Hz induction motor provided with a flywheel has to
supply a load torque of 600 N-m for 10 s followed by a no-load during which the flywheel
regains the full speed. The full-load slip of the motor is 4% and the torque−speed curve may
be assumed linear over the working range. Find the moment of inertia of the flywheel if the
motor torque is not to exceed twice the full-load torque.

Given data:
P0= 15 HP = 15 × 735.5 = 11.03 kW.
Number of poles P = 8
f = 50 Hz
Sf = 0.04
t = 10 sec
TL = 600 N-m
TM = 2TFL
T0 = 0.

Solution

Now synchronous speed can be calculated as follows

120𝑓
𝑁𝑠 = 𝑝

120×50
𝑁𝑠 = = 750 𝑟𝑝𝑚
8

full load torque

Electrical Machine selection by Melkior, U. F Page 18

 60𝑃
𝑇𝐹𝐿 = 2𝜋𝑁0
𝑟

𝑁𝑟 = 𝑁𝑠 (1 − 𝑆𝐹 )

𝑁𝑟 = 750(1 − 0.04) = 720 𝑟𝑝𝑚

60×11030
𝑇𝐹𝐿 = = 146.39 𝑁𝑚
2𝜋×720

𝑇𝑀 = 2𝑇𝐹𝐿 = 2 × 146.39 = 292.78 𝑁𝑚

Slip speed is calculated as

𝑠𝑙𝑖𝑝 𝑠𝑝𝑒𝑒𝑑 = 𝑆𝐹 𝑁𝑠 = 0.04 × 750 = 30𝑟𝑝𝑚

2𝜋×30
== = 3.14 𝑟𝑎𝑑/𝑠
60

and
𝑆 3.14
𝐾=𝑇 = 146.39=0.0214
𝐹𝐿

−𝑡
𝑇𝑀 −𝑇𝐿 = −(𝑇𝐿 − 𝑇0 )𝑒 𝐽𝐾

take natural logarithm both side

−𝑡 (𝑇 −𝑇 )
= log 𝑒 [ (𝑇𝐿 −𝑇𝑀) ]
𝐽𝐾 𝐿 0

−𝑡 600−292.78
= log 𝑒 [ ] = 0.669
𝐽𝐾 600

𝑡
= 0.669,
𝐽𝐾

10
𝐽 = 0.669×0.0214 = 698.49 𝑘𝑔𝑚2

1.6 Harmonics in Electrical Machines

1.6.1 Production of Harmonics
In a power system, any item of equipment which draws current from the supply which is
proportional to the applied voltage is termed a “linear” load. Examples of linear loads include
resistance heaters and incandescent lamps. The current and voltage waveforms associated
with linear loads are shown in Figure 13.

Electrical Machine selection by Melkior, U. F Page 19

FIGURE 13 Voltage and Current Waveforms for Linear Load
The term “nonlinear” is used to describe loads which draw current from the supply that is
dissimilar in shape to the applied voltage. Examples include discharge lighting, computers
and variable speed drives.
The simple single-phase full wave rectifier supplying a load containing both inductance and
resistance with impedance of the AC supply represented by the inductance Lac, the DC load
current (Idc) and AC input current (Iac), respectively.

FIGURE 14a Single Phase Full Wave Rectifier

FIGURE 14b Load and AC Supply Currents

Simple single line diagram comprising source voltage (u) and source impedance (LN). The
harmonic current (iN) passing through the source impedance produces a voltage drop (UL)
according the following formula:
𝑑𝑖
𝑈𝐿 = 𝐿𝑁 𝑑𝑡𝑁 1

Electrical Machine selection by Melkior, U. F Page 20

FIGURE 15a Simple Single Line Diagram

FIGURE 15b Load Current and Volt Drop Waveforms

The voltage drop across the source impedance (UL) is subtracted from the induced voltage
(U), resulting in the distortion of the supply voltage waveform,

Electrical Machine selection by Melkior, U. F Page 21

FIGURE 16 How Voltage Distortion is Produced (Simplified)
The majority of nonlinear loads equipment is that utilize power semiconductor devices for
power conversion (e.g., rectifiers). They include, for example, computer switched mode
power systems (SMPS) for converting AC to DC. Typical current waveform of a computer
switched mode power supplier unit.

FIGURE 17 Typical Computer Nonlinear Load

In order to appreciate why the current waveform shown in figure 17 is of a “pulsed” nature, it
is beneficial to consider the design of switched mode power supplies.

FIGURE 18 Single-phase Switched Mode Power Supply

This type of power supply uses capacitors to smooth the rectified DC voltage and current
prior to it being supplied to other internal subsystems and components. The semiconductor
diode rectifiers are unidirectional devices (i.e., they conduct in one direction only). The
additional function of the capacitor is to store energy which is drawn by the load as

Electrical Machine selection by Melkior, U. F Page 22

necessary. When the input voltage (Vi) is higher in value than the capacitor voltage (Vc), the
appropriate diode will conduct and non-sinusoidal, “pulsed” current will be drawn from the
supply.

Harmonic voltages and currents are integer multiples of the fundamental frequency. On 60
Hz supplies, for instance, the 5th harmonic is 300 Hz; the 7th harmonic is 420 Hz, and so
forth. When all harmonic voltages and currents are added to the fundamental, a waveform
known as a “complex wave” is formed.

FIGURE 19 Construction of Complex Wave

figure 19 shows Symmetrical waveform in which the positive portion of the wave is identical
to the negative portion. It should be noted that symmetrical waveforms only contain odd
harmonics, whereas asymmetrical waveforms (where both positive and negative portions are
different) contain even and odd harmonics and possibly DC components. An example of an
asymmetrical waveform would be that produced by a half wave rectifier.

1.6.2 Characteristic Harmonic Currents

Power conversion using full wave rectifiers generates idealized characteristic harmonic
currents given by the formula:
ℎ = 𝑛𝑝 ± 1 2
Where
h = order of harmonics
n = an integer 1, 2, 3…
p = number of current pulses per cycle
It should be noted that in the “ideal” harmonic theory, the following hypotheses are assumed
for all rectifiers:
a) The impedance of the AC supply network is zero.
b) The DC component of the rectifier configuration is uniform.
c) An AC line or commutating reactor is not used in front of the rectifier.
d) The AC supply network is symmetrical (i.e., balanced).
e) The AC supply is sinusoidal, free from harmonics.
f) There are no “overlap” or delay angles for the devices.

Electrical Machine selection by Melkior, U. F Page 23

Note: Any divergence from any of the above hypotheses will introduce “non-characteristic”
harmonics, including possibly DC, into the harmonic series. In practical terms, it has to
be noted that supply networks or connected equipment are never “idealized” (i.e.,
based on the above hypotheses) and, therefore, any actual harmonic currents measured
will not be exactly as calculated using the above simplified formula.

In addition, in idealized harmonic theory, the magnitude of the harmonic current is stated as
the reciprocal of the harmonic number:
1
𝐼=ℎ 3

Therefore, in theory, the 5th harmonic current and 7th harmonic current, for example, should
represent 20% and 14% of the total rms current, respectively. However, again, this is never
transferred into practical reality as the magnitudes of the various harmonic currents are
determined by the per-phase inductance of the AC supply connected, the rectifier and the
impedance of the rectifier as “seen” by the AC supply.
In rectifiers without added inductance (e.g., AC line reactors) it is not uncommon to measure
5th harmonic current up to ~80% on single-phase rectifiers and 65% on three-phase rectifiers.
Figure 20 illustrates a typical single-phase computer switched mode power supply with full
wave bridge rectifier.

FIGURE 20 Computer Power Supply with Single-phase Full Wave Bridge Rectifier

1.6.3 Effect of Harmonic Currents on Impedance(s)

Figure 21shows a simplified form that when a nonlinear load draws distorted (non-sinusoidal)
current from the supply, that distorted current passes through all of the impedance between
the load and power source. The associated harmonic currents passing through the system
impedance cause voltage drops for each harmonic frequency based on Ohm’s Law (Vh = Ih ×
Zh).
The vector sum of all the individual voltage drops results in total voltage distortion, the
magnitude of which depends on the system impedance and the levels of harmonic currents at
each harmonic frequency.

Electrical Machine selection by Melkior, U. F Page 24

FIGURE 21 Distorted Currents Induce Voltage Distortion

Figure 22 shows in detail the effect individual harmonic currents have on the impedances
within the power system and the associated voltages drops for each. Note that the “total
harmonic voltage distortion”, Vthd (based on the vector sum of all individual harmonics), is
reduced as more impedance is introduced between the nonlinear load and the source.

FIGURE 22 How Individual Harmonic Voltage Drops Develop Across System Impedances

1.6.4 Calculation of Voltage Distortion

Any periodic (repetitive) complex waveform is composed of a sinusoidal component at the
fundamental frequency and a number of harmonic components which are integral multipliers
of the fundamental frequency. The instantaneous value of voltage for non-sinusoidal
waveform or complex wave can be expressed as:

𝑣 = 𝑉0 + 𝑉1 𝑠𝑖𝑛(𝜔𝑡 + 𝜑1 ) + 𝑉2 𝑠𝑖𝑛(2𝜔𝑡 + 𝜑2 ) + 𝑉3 𝑠𝑖𝑛(3𝜔𝑡 + 𝜑3 ) + ⋯ + 𝑉𝑛 𝑠𝑖𝑛(𝑛𝜔𝑡 +

𝜑𝑛 ) 4
Where
v = instantaneous value at any time t
V0 = direct (or mean) value (DC component)
V1 = rms value of the fundamental component
V2 = rms value of the second harmonic component
V3 = rms value of the third harmonic component
Vn = rms value of the nth harmonic component
φ = relative angular frequency
ω = 2πf

Electrical Machine selection by Melkior, U. F Page 25

f = frequency of fundamental component (1/f defining the time over which the complex wave
repeats itself)
It is usually more convenient, however, to interpret a complex wave by means of “Fourier
Series” and associated analysis methods. Ignoring any DC components in the above formula,
where V1 and I1 represent the fundamental voltage and current, respectively, the instantaneous
rms voltage, Vh, can be represented as a Fourier Series:
∞ ∞

𝑣𝑡 = ∑ 𝑣ℎ (𝑡) = ∑ √2𝑉ℎ 𝑠𝑖𝑛(ℎ𝜔0 𝑡 + 𝜑ℎ ) 5

ℎ=1 ℎ=1
The rms value of voltage can be expressed as:

∞
1 𝑇 2
𝑣𝑟𝑚𝑠 = √ ∫ 𝑣 (𝑡)𝑑𝑡 = √∑ 𝑉ℎ2 = √𝑉12 + 𝑉22 + 𝑉32 + ⋯ + 𝑉𝑛2 6
𝑇 0
ℎ=1

∞
1 𝑇2
𝐼𝑟𝑚𝑠 = √ ∫ 𝑖 (𝑡)𝑑𝑡 = √∑ 𝐼ℎ2 = √𝐼12 + 𝐼 + 𝐼32 + ⋯ + 𝐼𝑛2 7
𝑇 0
ℎ=1

The rms voltage or current “total harmonic distortion”, Vthd and Ithd, respectively can be
expressed as:
∑∞ 2
ℎ=2 𝑉ℎ × 100% √𝑉22 + 𝑉32 + 𝑉42 + ⋯ + 𝑉𝑛2
𝑉𝑡ℎ𝑑 = = × 100% 8
𝑉1 𝑉1
∑∞ 2
ℎ=2 𝐼ℎ × 100% √𝐼22 + 𝐼32 + 𝐼42 + ⋯ + 𝐼𝑛2
𝐼𝑡ℎ𝑑 = = × 100% 9
𝐼1 𝐼1

Other simple but practical harmonic formulae include:

Total rms current:
2 2
𝐼𝑟𝑚𝑠 = √𝐼𝐹𝑢𝑛𝑑 + 𝐼ℎ𝑎𝑟𝑚 10

𝐼𝑡ℎ𝑑 2
√
𝐼𝑟𝑚𝑠 = 𝐼𝑓𝑢𝑛𝑑 1 + ( ) 11
100
𝐼𝑟𝑚𝑠
𝐼𝑓𝑢𝑛𝑑 = 12
2
√1 + 𝐼𝑡ℎ𝑑
Total fundamental current distortion:
2
𝐼
𝐼𝑡ℎ𝑑(𝐹𝑢𝑑) = √(𝐼 𝑟𝑚𝑠 ) − 1 13
𝑓𝑢𝑛𝑑

Total Demand Distortion (TDD)

√∑∞ 2
ℎ=2 𝐼ℎ √𝐼22 + 𝐼32 + 𝐼42 + ⋯ + 𝐼𝑛2
𝐼𝑇𝐷𝐷 = = 14
𝐼𝑙𝑜𝑎𝑑 𝐼𝑙𝑜𝑎𝑑

Where
Iload = maximum demand load current (fundamental) at the PCC

Electrical Machine selection by Melkior, U. F Page 26

TDD = ‘total demand distortion’ of current (expressed as measured total harmonic current
distortion, per unit of load current. For example, a 30% total current distortion
measured against a 50% load would result in a TDD of 15%).

1.6.5 Effects of Harmonics

[Link] Generators
The effects of harmonic voltages and harmonic currents are significantly more pronounced on
generators due to their source impedance being typically three to four times that of utility
transformers. The major impact of voltage and current harmonics is the increase in machine
heating caused by increased iron losses, and copper losses, both frequency dependent.

a) Thermal Losses
The iron losses comprise two separate losses, “hysteresis losses” and “eddy current losses”.
The hysteresis loss is the power consumed due to nonlinearity of the generator’s flux
density/magnetizing force curve and the subsequent reversal in the generator’s core magnetic
field each time the current changes polarity (i.e., 120 times a second for 60 Hz supplies).
Higher hysteresis losses occur at harmonic frequencies due to the more rapid reversals
compared to those at fundamental frequency. Hysteresis losses are proportional to frequency
and the square of the magnetic flux.
Eddy currents circulate in the iron core, windings and other component parts of the generator
induced by the stray magnetic fields around the turns in the generator windings. Eddy
currents produce losses which increase in proportion to the square of the frequency. The
relationship of eddy current losses and harmonics is given by:
ℎ𝑚𝑎𝑥

𝑃𝐸𝐶 = 𝑃𝐸𝐹 ∑ 𝐼ℎ2 ℎ2

ℎ=1
Where
PEC = total eddy current losses
PEF = eddy current losses at full load at fundamental frequency
Ih = rms current (per unit) harmonic h
h = harmonic number
On linear loads, the eddy current losses are fairly minor but become more significant as the
harmonic loading increases.

Copper losses are dissipated in the generator windings when current is passed through the winding
resistance. For a given load, when harmonic currents are present, the total rms value of the current
passing through the windings will be increased, thereby increasing the losses according to:
2
𝑃𝐶𝑈 = 𝐼𝑟𝑚𝑠 𝑅
Where
PCU = total copper losses
Irms = total rms current
R = resistance of winding
The copper losses are also influenced by a phenomenon termed “skin effect”. Skin effect
refers to the tendency of current flow to be confined in a conductor to a layer close to its
outer surface. At fundamental frequency, the skin effect is negligible and the distribution of
current across the cable is uniform. However, at harmonic frequencies, skin effect is
substantially more pronounced, significantly reducing the effective cross sectional area of the
conductor and increasing its resistance. The higher the resistance, the higher the I2R losses.
The harmonic stator current drawn by the nonlinear load will result in air gap fluxes, similar

Electrical Machine selection by Melkior, U. F Page 27

in shape to the fundamental but rotating at harmonic frequencies, inducing currents in the
rotor iron and windings, adding to rotor losses and the associated temperature rise.

b) Effect of Sequence Components

Harmonic currents occur in pairs each having a negative or a positive sequence rotation. The
5th harmonic is negative sequence and induces in the rotor a negatively-rotating 6th harmonic;
the 7th harmonic is positive and similarly induces a positively-rotating 6th harmonic in the
rotor. The two contra-rotating 6th harmonic systems in the rotor result in “ “damper” winding
currents which are stationary with respect to the rotor, causing additional, localized losses
and subsequent heating. Within the rotor, this effect is similar to that caused by single-phase
or unbalanced-phase operation and overheating is unlikely provided the rotor pole faces are
laminated. Similarly, the 11th and 13th harmonics will induce both negatively and positively
rotating 12th harmonics in the rotor.
Harmonic pairs, in addition to causing additional heating, can create mechanical oscillations
on the generator shaft. For 5th and 7th harmonic frequency, for example, the torque pulsations
will be at six times for fundamental (300 Hz based on 50 Hz fundamental, relative to the
stator). These result from the interaction between harmonic and fundamental frequencies
exciting a specific mechanical resonant frequency. Shafts can be severely stressed due to
these oscillations.

c) Voltage Distortion
A generator is designed to produce sinusoidal voltage at its terminals, but when nonlinear
current is drawn, the harmonic currents interact with the system impedances to produce
voltage drops at each individual harmonic frequency, thereby causing voltage distortion. A
simplified equivalent circuit for one phase of a three-phase generator is shown in Figure 23.
Note that the resistances are ignored as they are relatively small compared to the reactance.
The iron and damper windings, Dd and Dq, respectively, prevent any rapid flux changes in the
rotor.

FIGURE 23 Equivalent Circuits for a Generator

To calculate the rms harmonic voltage due to the respective harmonic current, the following method
can be used:
𝑉ℎ = √3 × 𝐼ℎ × ℎ × 𝑋𝑔𝑒𝑛
Where
Vh = L-L rms voltage of the harmonic h
Ih = harmonic current at order h
Xgen = generator reactance, in ohms
h = harmonic number
To calculate the L-L rms harmonic voltage as a percentage of rms fundamental voltage:

Electrical Machine selection by Melkior, U. F Page 28

 𝑉ℎ × 100%
𝑉𝑟𝑚𝑠
Where
Vh = L-L rms voltage of the harmonic h
Vrms = L-L fundamental rms voltage

Example
Calculate the L-L rms 5th harmonic voltage for a 480 V generator with reactance X of 0.0346
ohms when the 5th harmonic current is 135 A. Also, express the harmonic voltage as a
percentage of the fundamental rms voltage.
𝑉ℎ = √3 × 𝐼ℎ × ℎ × 𝑋𝑔𝑒𝑛
= 1.73 × 135 × 5 × 0.0346
𝐕𝐡 = 𝟒𝟎. 𝟒𝟓 𝐕
5th harmonic voltage as a percentage of the fundamental rms voltage
𝑉ℎ × 100%
=
𝑉𝑟𝑚𝑠

40.45 × 100%
480
= 𝟖. 𝟒𝟑%

[Link] Induction Motors

a) Thermal Losses
Harmonics distortion raises the losses in AC induction motors in a way similar to that
apparent in transformers with increased heating, due to additional copper losses and iron
losses (eddy current and hysteresis losses) in the stator winding, rotor circuit and rotor
laminations. These losses are further compounded by skin effect, especially at frequencies
above 300 Hz.
Leakage magnetic fields caused by harmonic currents in the stator and rotor end windings
produce additional stray frequency eddy current dependent losses. Substantial iron losses can
also be produced in induction motors with skewed rotors due to high-frequency-induced
currents and rapid flux changes (i.e., due to hysteresis) in the stator and rotor. The magnitude
of the iron losses is dependent on the iron loss characteristic of the laminations and the angle
of skew. The formulae used to calculate copper losses and eddy current losses used for
transformers are also applicable for induction motors:
2
𝑃𝐶𝑈 = 𝐼𝑟𝑚𝑠 𝑅
Where
PCU = total copper losses
Irms = total rms current
R = resistance of winding

The eddy current losses can be calculated

ℎ𝑚𝑎𝑥

𝑃𝐸𝐶 = 𝑃𝐸𝐹 ∑ 𝐼ℎ2 ℎ2

ℎ=1
Where
PEC = total eddy current losses

Electrical Machine selection by Melkior, U. F Page 29

PEF = eddy current losses at full load at fundamental frequency
Ih = rms current (per unit) harmonic h
h = harmonic number

Motors with deep bar or double cage rotors are susceptible to additional losses, particularly on highly
polluted supplies containing high order harmonics. This can in extreme cases, lead to “hot rotors”
which, due to conduction along the shaft, can degrade the bearing lubrication and result in bearing
collapse. Harmonic currents also can result in bearing currents. This can be prevented through the use
of an insulated bearing, a practice common in AC variable frequency drive-fed AC motors.
Overheating imposes significant limits on the effective life of an induction motor. For every 10°C rise
in temperature (continuous) above rated temperature, the life of motor insulation may be reduced by
as much as 50%. Squirrel cage rotors can normally withstand higher temperature levels compared
to wound rotors. The motor’s windings, especially if insulation is class B or below, are also
susceptible to damage due high levels of dv/dt (i.e., rate of rise of voltage) such as those attributed to
line notching and associated ringing.

b) Effect of Harmonic Sequence Components

Harmonic sequence components also adversely affect induction motors. Positive sequence
components (i.e., 7th, 13th, 19th…) will assist torque production, whereas the negative
sequence components (5th, 13th, 17th…) will act against the direction of rotation resulting in
torque pulsations which are significant. Zero sequence harmonics (i.e., triplens) do not rotate
(i.e., they are stationary); any harmonic energy associated with them is dissipated as heat. The
magnitude of torque pulsations due to sequence components can be estimated as follows to
assess possible shaft torsional vibration problems based on a nominal voltage:
2 2 − 2𝐼 2 𝐼 2 cos(𝜃
𝑇3𝑘 = √(𝐼𝑛+ + 𝐼𝑛− 𝑛+ 𝑛− 𝑛+ − 𝜃𝑛_ ))

where
In+, In– = per unit values
n+ = represents the 1 + 3k harmonic orders
n– = represents the 1 – 3k harmonic orders

For example, if we consider a 50 Hz supply with 4% voltage distortion based on motor

harmonic currents of 0.03 and 0.02 per unit for the 5th and 7th harmonics, respectively.
Assuming no phase displacement between harmonics and full value of voltage, the torque
will have a varying component at 300 Hz with amplitude of torque fluctuation of 0.01 per
unit. If the harmonics have the worst case phase relationship, then the amplitude of the torque
fluctuation will be 0.05 per unit.

The “Harmonic Voltage Factor” (HVF) can be defined as:

𝑛+∞
𝑉𝑛2
𝐻𝑉𝐹 = √ ∑
𝑛
𝑛=5

Where
n = odd harmonic, not including triplens
Vn = per unit magnitude of the voltage at the n-th harmonic

Example
1. Calculate the HVF based on the per unit harmonic voltages of 0.09, 0.065, 0.042,
0.038 for the 5th, 7th, 11th and 13th harmonics, respectively.

Electrical Machine selection by Melkior, U. F Page 30

 𝑛+∞
𝑉𝑛2
𝐻𝑉𝐹 = √ ∑
𝑛
𝑛=5

0.092 0.0682 0.0422 0.0382

𝐻𝑉𝐹 = √ + + +
5 7 11 13
𝐻𝑉𝐹 = 0.0505 𝑜𝑟 5.05% 𝑉𝑜𝑙𝑡𝑎𝑔𝑒 𝑑𝑖𝑠𝑡𝑜𝑟𝑡𝑎𝑡𝑖𝑜𝑛

2. An 850 HP/630 kW 6-pulse AC PWM drive with full load input current of 962 A is to
be supplied from a 2000 kVA 6000 V/480 V transformer with 6% impedance. The
drive manufacturer has advised the characteristic harmonic currents to the 25th based
on a 2% DC bus reactor on this transformer source to be as below.

Harmonic spectrum to 25th is:

5th – 32.27%
7th – 11.21%
11th – 7.37%
13th – 3.91%
17th – 3.45%
19th – 2.35%
23rd – 1.82%
25th – 1.51%
Calculate the source short circuit capacity and supply reactance on secondary of transformer:
Transformer full load current (A)
𝑇𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚𝑒𝑟 𝐾𝑉𝐴
=
𝑠𝑒𝑐𝑜𝑛𝑑𝑎𝑟𝑦 𝑉𝑜𝑙𝑡𝑎𝑔𝑒 × √3
2000 × 103
= = 2406 A
√3 × 480
Short circuit current
𝑇𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚𝑒𝑟 𝑓𝑢𝑙𝑙 𝑙𝑜𝑎𝑑 𝑐𝑢𝑟𝑟𝑒𝑛𝑡
=
𝑡𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚𝑒𝑟 𝑖𝑚𝑝𝑒𝑑𝑎𝑛𝑐𝑒
2406
= = 𝟒𝟎, 𝟏𝟎𝟎 𝐀 = 𝟒𝟎. 𝟏 𝐀
0.06
Short circuit MVA
√3 × 𝑉 × 𝑠ℎ𝑜𝑟𝑡 𝑐𝑖𝑟𝑐𝑢𝑖𝑡 𝑐𝑢𝑟𝑟𝑒𝑛𝑡
√3 × 480 × 40100 = 𝟑𝟑. 𝟑𝟒 𝐌𝐕𝐀
𝑉2
Supply Reactance =
𝑆ℎ𝑜𝑟𝑡 𝑐𝑖𝑟𝑐𝑢𝑖𝑡 𝑀𝑉𝐴
4802
Supply Reactance = = 𝟎. 𝟎𝟎𝟔𝟗𝟐 
33.34 × 106

Once the above information has been established, the individual voltage per harmonic order can be
calculated using the following method:

Harmonic current per order = Irms × Ih%

Irms =Total rms input current
Ih% =percentage harmonic current at harmonic order (e.g. 32.27 at 5th )

Electrical Machine selection by Melkior, U. F Page 31

Voltage (L-L) per harmonic order =√3hXsupply Ih

Where
h =harmonic Order (i.e. 5th =5, 11th =11 ….)
Xsupply = Supply reactance in ohms
Ih= harmonic current (A) at harmonic order h

𝑉ℎ ×100%
Harmonic voltage as a percentage of L-L rms voltage =
𝑉𝑟𝑚𝑠
Where as
𝑉ℎ =Harmonic voltage at order h
𝑉𝑟𝑚𝑠 =System L-L rms voltage

for 5th Harmonic

5th harmonic current =Irms × Ih%
5th harmonic current =962 × 0.3227
=𝟑𝟏𝟎. 𝟒𝟒𝐀

5th harmonic Voltage (L-L)= √3 × h × Xsupply × Ih

= √3 × 5 × 0.00692 × 310.44

= 𝟏𝟖. 𝟔 𝐕

𝑉ℎ ×100%
Harmonic voltage as a percentage of L-L rms voltage =
𝑉𝑟𝑚𝑠
18.6 × 100%
480
= 3.88%

Similarly for 7th harmonic

7th harmonic current =Irms × Ih%
7th harmonic current =962 × 0.1121
=𝟏𝟎𝟕. 𝟖𝟒𝐀

5th harmonic Voltage (L-L)= √3 × h × Xsupply × Ih

= √3 × 7 × 0.00692 × 𝟏𝟎𝟕. 𝟖𝟒

= 𝟗. 𝟎𝟒 𝐕

𝑉ℎ ×100%
Harmonic voltage as a percentage of L-L rms voltage =
𝑉𝑟𝑚𝑠
9.04 × 100%
480
= 1.88%

Electrical Machine selection by Melkior, U. F Page 32

 Harmonics 𝐈𝐡% 𝐈𝐡 (𝐀) 𝐕𝐡 (𝐕) 𝐕𝐡%
5 32.27% 310.44 18.6 3.87
7 11.21% 107.84 9.0 1.88
11 7.37% 70.90 9.3 1.95
13 3.91% 37.61 5.9 1.22
17 3.45% 33.19 6.8 1.41
19 2.35% 22.61 5.1 1.07
23 1.82% 17.51 4.8 1.00
25 1.51% 14.53 4.3 0.91

once the table above has been constructed, the total harmonics Voltage distortioncan be
calculated as
ℎ=25
2
𝑉𝑡ℎ𝑑 = √ ∑ (𝑉ℎ% )
ℎ=5

2 2 2 2
𝑉𝑡ℎ𝑑 = √(𝑉5% ) + (𝑉7% ) + (𝑉11
% %
) + ⋯ + (𝑉25 )

𝑉𝑡ℎ𝑑 = √3.872 + 1.882 + 1.952 + 1.222 + 1.412 + 1.072 + 1.002 + 0.912

𝑉𝑡ℎ𝑑 = 5.4%

Tutorial
1. The same 850 HP/630 kW, 480 V, 962 A, 6-pulse AC PWM is to be supplied by a
generator of rating 2000 kVA and sub-transient reactance (Xd″) of 16%. Calculate the
voltage distortion on the generator attributed to drive. The percentage harmonic
current spectrum based on the generator connection is given below. However, this
reduced harmonic current can result in significantly higher voltage distortion, as can
be seen below.
Harmonic spectrum to 25th is:
5th – 25.63%
7th – 7.8%
11th – 5.27%
13th – 3.43%
17th – 1.73%
19th – 1.55%
23rd – 0.93%
25th – 0.78%
Calculate the source short circuit capacity and reactance of the generator:

2. An 850 HP/630 kW 6-pulse AC PWM drive with full load input current of 962 A is to
be supplied from a 2000 kVA 6000 V/480 V transformer with 6% impedance. The
drive manufacturer has advised the characteristic harmonic currents to the 25th based
on a 2% DC bus reactor on this transformer source to be as below.
Note: The harmonics up to the 50th should be used in all practical applications. In this
instance, only up to the 25th harmonic is used for simplicity in order to demonstrate
the principles. For higher order harmonics, the calculations are exactly the same).
Harmonic spectrum to 25th is:

Electrical Machine selection by Melkior, U. F Page 33

 5th – 32.27%
7th – 11.21%
11th – 7.37%
13th – 3.91%
17th – 3.45%
19th – 2.35%
23rd – 1.82%
25th – 1.51%

Electrical Machine selection by Melkior, U. F Page 34