Three Phase Synchronous Generator — Detailed

Notes
Unit 2

Contents
2.1 Working Principle, Necessity of Constant Speed Operation, Speed Governor 1
2.2 EMF Equation, Distribution Factor, Pitch Factor . . . . . . . . . . . . . 4
2.3 Operation at No-Load and Load Conditions . . . . . . . . . . . . . . . . 6
2.4 Armature Reaction and Its Effects — Resistive, Inductive, Capacitive
Loads; Phasor Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.5 Concept of Synchronous Reactance and Impedance . . . . . . . . . . . 9
2.6 Voltage Regulation — Synchronous Impedance Method, Ampere-Turn
(MMF) Method, Zero Power Factor Method . . . . . . . . . . . . . . . 10
2.7 Parallel Operation of Two Synchronous Generators and Load Sharing with
Speed-Droop Characteristics . . . . . . . . . . . . . . . . . . . . . . . 13
2.8 Transient Condition, Transient and Sub-transient Reactance . . . . . . . 17
2.9 Synchronous Generator Connected to an Infinite Bus . . . . . . . . . . . 19
2.10 Limits of Active and Reactive Power Generation — Capability Curve . . 21
2.11 Losses and Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.1 Working Principle, Necessity of Constant Speed Operation, Speed

Governor
Working Principle

A synchronous generator (alternator) works on the principle of Faraday’s Law of

Electromagnetic Induction: whenever a conductor cuts magnetic flux lines, an
EMF is induced in it.
Construction basics:
• Stator (Armature): Stationary part carrying a 3-phase winding displaced 120°
electrical apart in space. The output power is drawn from here.
• Rotor (Field): Rotating part carrying the DC-excited field winding, which pro-
duces the main magnetic flux. Two types:
– Salient pole rotor — projecting poles, used for low-speed, large-diameter
machines (hydro generators).
– Cylindrical (non-salient) rotor — smooth cylindrical rotor, used for high-
speed machines (turbo-alternators driven by steam/gas turbines).

1
Operation: The rotor is driven by a prime mover (turbine, diesel engine) at a con-
stant speed. DC current fed to the rotor field winding (through slip rings/brushes or
brushless excitation) creates a rotating magnetic field. This rotating flux cuts the sta-
tionary armature conductors and induces a sinusoidal EMF in each phase, displaced
120° in time — because the windings are 120° apart in space.

Relation Between Speed, Frequency and Poles

𝑃 𝑁𝑠
𝑓=
120
where 𝑃 = number of poles, 𝑁𝑠 = synchronous speed in rpm, 𝑓 = frequency in Hz.

Necessity of Constant Speed Operation

1. Frequency stability: Since 𝑓 ∝ 𝑁𝑠 , any variation in rotor speed directly

changes the output frequency. Power systems and connected loads (motors,
electronics, clocks, other generators) require frequency to remain fixed (50 Hz
or 60 Hz).
2. Parallel operation: When multiple alternators operate in parallel on a grid,
all must run at speeds corresponding to the same system frequency; otherwise
they fall out of synchronism, causing large circulating currents and mechanical
stress.
3. Voltage stability: Induced EMF, 𝐸 = 4.44𝑓Φ𝑇 𝐾𝑤 , is also frequency dependent
— speed fluctuation causes voltage fluctuation.
4. Equipment protection: Motors and transformers are designed for rated fre-
quency; off-frequency operation causes overheating, mistiming, and inefficient
operation.

Speed Governor

A governor is a device fitted to the prime mover (turbine/engine) that automatically

controls the input (steam, water, or fuel) to maintain constant speed despite changes
in load.
Working:
• When load increases, the generator draws more power from the prime mover,
tending to slow it down.
• The governor senses this speed drop (via a flyball/centrifugal mechanism or elec-
tronic sensor) and increases the fuel/steam/water input to restore speed.
• When load decreases, the reverse happens — input is reduced to prevent over-
speeding.
Types of governors: Mechanical (centrifugal flyball), hydraulic, and elec-
tronic/digital governors (used in modern plants for faster, more precise response).
Droop characteristic: Practical governors are set with a small speed droop (typi-
cally 3–5%) so that speed falls slightly as load increases. This droop is essential for

2
stable load sharing when multiple generators run in parallel (covered in Section
2.7).

3
2.2 EMF Equation, Distribution Factor, Pitch Factor
EMF Equation (per phase)

For a sinusoidally distributed flux Φ (Wb per pole), frequency 𝑓 (Hz), and 𝑇 turns per
phase:
For a full-pitch, concentrated winding:

𝐸 = 4.44 𝑓 Φ 𝑇 (RMS value)

For a practical winding (distributed + short-pitched):

𝐸𝑝ℎ = 4.44 𝑓 Φ 𝑇 𝐾𝑑 𝐾𝑝

where:
• 𝐾𝑑 = distribution factor
• 𝐾𝑝 = pitch (coil-span) factor
• 𝐾𝑤 = 𝐾𝑑 × 𝐾𝑝 = winding factor
Derivation outline: Average EMF per conductor = 𝑑Φ/𝑑𝑡 over a pole pitch = 2Φ𝑓 .
RMS EMF per conductor = 1.11 × 2Φ𝑓 = 2.22 Φ𝑓 (using form factor 1.11 for sinu-
soidal wave). For 𝑇 turns (2T conductors) per phase: 𝐸 = 4.44𝑓Φ𝑇 , then corrected
by 𝐾𝑑 and 𝐾𝑝 for real windings.

Distribution Factor (𝐾𝑑 )

In practice, windings are distributed in several slots per pole per phase (not concen-
trated in one slot), to produce a more sinusoidal EMF waveform and better utilize the
stator periphery.
Because the coils in different slots are displaced by a slot angle 𝛽 , their individual
EMFs are not in time phase — they must be added phasorially, not arithmetically.

Phasor sum of EMFs of distributed coils sin(𝑚𝛽/2)

𝐾𝑑 = =
Arithmetic sum of EMFs of distributed coils 𝑚 sin(𝛽/2)

where 𝑚 = slots per pole per phase, 𝛽 = slot angle (electrical degrees) =
180°/(slots per pole).
• 𝐾𝑑 is always less than 1 (typically 0.9–0.96).
• Distribution reduces the total EMF slightly but greatly improves waveform shape
by suppressing harmonics.

4
Pitch Factor / Coil-Span Factor (𝐾𝑝 )

Windings are often short-pitched (chorded) — the coil span is made slightly less than
one full pole pitch (180° electrical) — instead of full-pitch.

𝛼
𝐾𝑝 = cos ( )
2
where 𝛼 = angle (electrical degrees) by which the coil is short-pitched (chording
angle).
Advantages of short-pitching:
1. Saves copper (shorter end connections).
2. Reduces/eliminates specific harmonics (e.g., 5th, 7th) from the output waveform,
giving a more sinusoidal EMF.
3. Reduces the fundamental EMF slightly (𝐾𝑝 < 1), which is an acceptable trade-
off.
Winding factor: 𝐾𝑤 = 𝐾𝑑 × 𝐾𝑝 (typically around 0.85–0.95), used directly in the
EMF equation.

5
2.3 Operation at No-Load and Load Conditions
No-Load Operation

• The alternator is driven at synchronous speed with the field winding excited by
DC current, but the armature (stator) circuit is open (no current flows in stator
windings).
• Only the rotor’s DC field current 𝐼𝑓 produces flux, which induces EMF 𝐸0 (open-
circuit EMF) in the armature as per the EMF equation.
• Since there is no armature current, there is no armature reaction and no volt-
age drop in the winding — terminal voltage equals the induced EMF: 𝑉 = 𝐸0 .
• The Open Circuit Characteristic (OCC) is obtained by plotting 𝐸0 vs field
current 𝐼𝑓 . It resembles the B-H curve of the magnetic circuit — linear initially,
then saturates at higher 𝐼𝑓 .

Load Operation

• When a load is connected, armature current 𝐼𝑎 flows in the stator windings.

• This current produces its own rotating magnetic field (armature reaction) that
interacts with the main field flux, altering the net air-gap flux and hence the
terminal voltage.
• The armature current also causes voltage drops across:
– Armature resistance (𝑅𝑎 ) — 𝐼𝑎 𝑅𝑎 drop (resistive, in phase with 𝐼𝑎 )
– Armature leakage reactance (𝑋𝑙 ) — 𝐼𝑎 𝑋𝑙 drop
– Armature reaction effect — represented as an additional reactance drop,
𝑋𝑎𝑟
• Combining leakage reactance and armature-reaction reactance gives the syn-
chronous reactance 𝑋𝑠 (Section 2.5).
• Terminal voltage under load:

𝑉 = 𝐸0 − 𝐼𝑎 𝑅𝑎 − 𝑗𝐼𝑎 𝑋𝑠 (phasor equation)

• The terminal voltage under load is generally less than the no-load EMF for resis-
tive and inductive loads (due to demagnetizing armature reaction and impedance
drops), but can be greater for leading (capacitive) loads (magnetizing armature
reaction).

6
2.4 Armature Reaction and Its Effects — Resistive, Inductive, Capaci-
tive Loads; Phasor Diagrams
Armature reaction = the effect of the magnetic field produced by armature (stator)
current on the main field flux produced by the rotor.
The nature of this effect depends on the phase angle between armature current
and induced EMF, i.e., on the power factor of the load.

(a) Unity Power Factor (Purely Resistive Load)

• Armature current 𝐼𝑎 is in phase with induced EMF 𝐸0 .

• The armature reaction flux is at 90° to the main flux — it is purely cross-
magnetizing.
• Effect: distorts the flux wave (strengthens flux under one pole tip, weakens un-
der the other) but does not significantly change its average magnitude.
• Terminal voltage drops slightly due to 𝑅𝑎 and 𝑋𝑠 drops only.
Phasor diagram (resistive load): 𝐸0 leads 𝑉 by the internal impedance drop (𝐼𝑎 𝑅𝑎
parallel to 𝐼𝑎 , 𝐼𝑎 𝑋𝑠 perpendicular to 𝐼𝑎 ); 𝐼𝑎 is in phase with 𝑉 .
E0
|\
| \ IaXs (perp. to Ia)
| \
| \___________ V (approx. along Ia)
| /
| IaRa (parallel to Ia)
|_/________________-> Ia (reference, in phase with V)

(b) Zero Power Factor Lagging (Purely Inductive Load)

• Armature current 𝐼𝑎 lags 𝐸0 by 90°.

• The armature reaction flux is now exactly opposite (in phase opposition) to the
main field flux.
• Effect: Purely demagnetizing — armature reaction directly reduces the net
air-gap flux, reducing induced EMF and terminal voltage significantly.
• This gives the largest voltage drop among the three cases; regulation is poor
(large positive % regulation).
Phasor diagram (pure inductive/lagging pf): 𝐸0 , 𝐼𝑎 𝑋𝑠 , 𝐼𝑎 𝑅𝑎 , and 𝑉 all lie
roughly along the same vertical line since 𝐼𝑎 is at 90°.
E0 (up)
|
| IaXs
|
V
|
| IaRa

7
 v
Ia -> (Ia lags V/E0 by 90 deg, drawn horizontally)

(c) Zero Power Factor Leading (Purely Capacitive Load)

• Armature current 𝐼𝑎 leads 𝐸0 by 90°.

• The armature reaction flux is now in phase with (aiding) the main field flux.
• Effect: Purely magnetizing — armature reaction increases the net flux, so in-
duced EMF and terminal voltage increase above 𝐸0 .
• This gives a negative voltage regulation (terminal voltage on load is higher
than no-load EMF).

General (Mixed) Load — Lagging or Leading Power Factor

For a general load with power factor cos 𝜙, armature current 𝐼𝑎 lags (or leads) 𝑉 by
angle 𝜙. Armature reaction has both a cross-magnetizing component and a demagne-
tizing (or magnetizing) component.
General phasor diagram (lagging pf, most common case):
E0
/|
/ IaXs
/ |
V ---+
\ |
\ IaRa
\ |
Ia (Ia lags V by phi)
Relation (phasor form):

𝐸0⃗ = 𝑉 ⃗ + 𝐼𝑎 𝑅𝑎 ∠0° + 𝐼𝑎 𝑋𝑠 ∠90°

(all added phasorially, with 𝐼𝑎 at angle −𝜙 relative to 𝑉 for lagging pf)
Summary Table:

Angle of 𝐼𝑎 w.r.t. Armature Reaction Terminal Voltage

Load Type 𝐸0 Effect vs 𝐸0
Resistive (UPF) 0° Cross-magnetizing Slightly less
Inductive (lagging, 90° lag Demagnetizing Much less (poor
ZPF) regulation)
Capacitive 90° lead Magnetizing Greater (𝑉 > 𝐸0 )
(leading, ZPF)
General lagging pf 0°–90° lag Partly Less
demagnetizing
General leading pf 0°–90° lead Partly magnetizing Less drop / may
rise

8
2.5 Concept of Synchronous Reactance and Impedance
When the armature carries current, three effects occur simultaneously:
1. Armature resistance drop (𝐼𝑎 𝑅𝑎 ) — real, due to winding resistance.
2. Armature leakage reactance drop (𝐼𝑎 𝑋𝑙 ) — due to leakage flux that links the
armature winding but not the main air gap (slot leakage, end-winding leakage,
tooth-tip leakage).
3. Armature reaction — effect of armature MMF on the main flux, treated as an
equivalent reactance drop 𝐼𝑎 𝑋𝑎𝑟 (since armature reaction is proportional to 𝐼𝑎
and lags it by 90° in its magnetic effect, it behaves exactly like a reactance).
Since both 𝑋𝑙 and 𝑋𝑎𝑟 cause voltage drops that are 90° out of phase with 𝐼𝑎 , they
are combined into a single fictitious reactance called the synchronous reactance:

𝑋𝑠 = 𝑋𝑙 + 𝑋𝑎𝑟

Combined with armature resistance, we get the synchronous impedance:

𝑍𝑠 = 𝑅𝑎 + 𝑗𝑋𝑠 , |𝑍𝑠 | = √𝑅𝑎2 + 𝑋𝑠2

This allows the entire complex behavior of armature reaction + leakage flux to be
represented by one simple equivalent circuit: an EMF source 𝐸0 in series with 𝑍𝑠 ,
supplying the terminal voltage 𝑉 and load.
Equivalent circuit (per phase):
E0 --[Ra]--[jXs]--*-- V (terminal) ---- Load
Voltage equation:

𝐸0⃗ = 𝑉 ⃗ + 𝐼𝑎⃗ (𝑅𝑎 + 𝑗𝑋𝑠 )

Note: 𝑋𝑠 is not constant — it varies with the degree of magnetic saturation, which
is why methods like the Zero Power Factor method (Section 2.6) are used to get a
more accurate, saturated value for regulation calculations at rated conditions.

9
2.6 Voltage Regulation — Synchronous Impedance Method, Ampere-
Turn (MMF) Method, Zero Power Factor Method
Definition of Voltage Regulation

Voltage regulation is the change in terminal voltage when full load (at a given power
factor) is thrown off, keeping field excitation and speed constant, expressed as a per-
centage of rated terminal voltage.

𝐸0 − 𝑉
%𝑉𝑅 = × 100
𝑉
where 𝐸0 = no-load induced EMF (with same excitation), 𝑉 = rated terminal voltage
on load.
• Lagging pf loads → positive regulation (𝐸0 > 𝑉 )
• Leading pf loads → often negative regulation (𝐸0 <𝑉)
Since directly loading a large alternator to determine regulation is impractical (needs
an actual load bank of that rating), indirect methods are used.

(1) Synchronous Impedance Method (EMF Method)

Tests required:
• Open Circuit Test (OCC): Alternator run at synchronous speed, armature open-
circuited; field current is varied and open-circuit voltage 𝐸0 is recorded → gives
OCC curve (𝐸0 vs 𝐼𝑓 ).
• Short Circuit Test (SCC): Armature terminals short-circuited through amme-
ters; field current gradually increased and short-circuit armature current 𝐼𝑠𝑐
recorded → gives SCC curve (𝐼𝑠𝑐 vs 𝐼𝑓 ), which is linear (no saturation since flux
is very low).
Procedure:
1. For a given field current 𝐼𝑓 , read 𝐸0 (open circuit) from OCC and 𝐼𝑠𝑐 (short
circuit) from SCC.
2. Since on short circuit, 𝑉 = 0, the entire 𝐸0 drops across 𝑍𝑠 :

𝐸0
𝑍𝑠 = (per phase, at that 𝐼𝑓 )
𝐼𝑠𝑐
3. Knowing 𝑅𝑎 (measured by DC test on armature, corrected for AC/skin effect),
calculate:

𝑋𝑠 = √𝑍𝑠2 − 𝑅𝑎2
4. Use 𝐸02 = (𝑉 cos 𝜙 + 𝐼𝑎 𝑅𝑎 )2 + (𝑉 sin 𝜙 ± 𝐼𝑎 𝑋𝑠 )2 (+ for lagging, − for leading)
to find 𝐸0 at rated 𝑉 , 𝐼𝑎 , and pf, then compute % regulation.

10
Limitation: This method over-estimates 𝑋𝑠 (since 𝐼𝑓 for OCC is taken in the satu-
rated region, but SCC is unsaturated), giving pessimistic (higher than actual) regula-
tion values. Hence called the “pessimistic method.”

(2) Ampere-Turn (MMF) Method

This method assumes armature resistance drop is negligible (or treated separately)
and works directly in terms of MMF (ampere-turns) rather than reactance.
Procedure:
1. From OCC, find field current 𝐼𝑓1 required to produce rated voltage 𝑉 (plus 𝐼𝑎 𝑅𝑎
drop) on open circuit.
2. From SCC, find field current 𝐼𝑓2 required to circulate rated armature current
𝐼𝑎 on short circuit — this 𝐼𝑓2 represents the field MMF needed to overcome the
armature reaction (demagnetizing) MMF and leakage reactance drop entirely.
3. Combine 𝐼𝑓1 and 𝐼𝑓2 phasorially (not arithmetically) — the angle between them
depends on load power factor:
• For lagging pf 2 + 𝐼 2 + 2𝐼 𝐼 sin 𝜙
𝜙: resultant 𝐼𝑓 = √𝐼𝑓1 𝑓2 𝑓1 𝑓2

• For leading pf: 𝐼𝑓 2 + 𝐼 2 − 2𝐼 𝐼 sin 𝜙

= √𝐼𝑓1 𝑓2 𝑓1 𝑓2

• For unity pf: 𝐼𝑓 = √𝐼𝑓1 2 + 𝐼2

𝑓2
4. Using this resultant 𝐼𝑓 , read the corresponding 𝐸0 from the OCC (extended if
necessary).
5. Calculate % regulation using 𝐸0 and 𝑉 .
Limitation: This method under-estimates regulation (optimistic method), since it
assumes all the reactance drop occurs on the saturated part of the OCC.

(3) Zero Power Factor (ZPF) / Potier Method

This is the most accurate practical method, correcting for the shortcomings of the
above two.
Tests required:
• OCC (as before).
• ZPF test: Alternator loaded with a purely inductive load (or a reac-
tor/synchronous condenser under-excited) adjusted to draw rated armature
current at zero lagging power factor, for at least two points — usually rated
current at rated voltage and at short-circuit (𝑉 = 0, giving the SCC point).
Procedure (Potier construction):
1. Plot the OCC.
2. Plot the ZPF characteristic (𝑉 vs 𝐼𝑓 at rated 𝐼𝑎 , zero pf) — this curve is iden-
tical in shape to OCC but shifted, since at ZPF the armature reaction is purely
demagnetizing.
3. From the ZPF full-load point, draw a horizontal line back to a length equal to the
short-circuit field current (this gives the Potier triangle), whose horizontal side

11
 = armature reaction MMF (equivalent field current), and vertical side = leakage
reactance drop 𝐼𝑎 𝑋𝑙 (Potier reactance).
4. Slide this triangle along under the OCC curve, keeping its horizontal side hor-
izontal and vertical side vertical, until the top vertex touches the OCC and the
lower vertex touches the ZPF curve — this locates the true leakage reactance
drop separately from the armature-reaction demagnetizing MMF.
5. From this construction, determine the actual field MMF needed to generate the
required 𝐸0 at the given load and pf; read 𝐸0 from OCC.
6. Compute % regulation as before.
Advantage: Separates leakage reactance (a fixed, saturation-independent quantity)
from armature-reaction MMF (which depends on saturation), giving results close to
actual measured regulation — considered the most reliable of the three classical meth-
ods.

12
2.7 Parallel Operation of Two Synchronous Generators and Load Shar-
ing with Speed-Droop Characteristics
Conditions for Paralleling (Synchronizing)

Before connecting an incoming alternator to a running bus/another alternator, these

conditions must be satisfied (checked using synchroscope or synchronizing lamps):
1. Equal terminal voltage (magnitude) as the bus.
2. Same phase sequence.
3. Same frequency as the bus.
4. Voltages in phase (zero phase angle difference) at the instant of closing the
breaker.

Why Parallel Operation is Needed

• To supply loads exceeding one machine’s capacity.

• To allow maintenance/shutdown of one machine without interrupting supply.
• To improve reliability and allow flexible loading according to demand (running
machines near their efficient operating point).

Load Sharing — Governor Droop Characteristic

Once synchronized, how the total load divides between two parallel generators is gov-
erned by their governor droop (speed-load) characteristics, not by their voltage
regulation.
Droop characteristic: Each generator’s governor is set so that its speed (and hence
frequency, since 𝑓 ∝ 𝑁 ) falls slightly as its electrical load (kW output) increases —
a drooping straight-line relationship between frequency and load:

𝑁𝑛𝑜-𝑙𝑜𝑎𝑑 − 𝑁𝑓𝑢𝑙𝑙-𝑙𝑜𝑎𝑑
Droop% = × 100 (typically 2–5%)
𝑁𝑟𝑎𝑡𝑒𝑑
Load sharing principle:
• All parallel-connected generators must operate at the same system frequency
(common bus frequency), since they are electrically locked together.
• Each machine’s real power (kW) output is determined by where its individual
droop line intersects the common operating frequency.
• If a machine’s governor setpoint (no-load frequency) is raised (by increas-
ing fuel/steam input reference), its droop line shifts upward — at the same sys-
tem frequency, this machine now takes a larger share of the load, while the
other machine’s share decreases correspondingly (since total load is fixed and
frequency is common).
Graphical explanation: Two drooping straight lines (frequency vs. kW load) inter-
sect the common operating frequency at different load points 𝑃1 and 𝑃2 , with $P_1 +
P_2 = $ total load. By adjusting the governor reference (parallel shift of droop line) of

13
either machine, load can be transferred from one machine to the other while system
frequency stays essentially at the set value.
f
|\ \
| \ \ <- Machine 2 droop line
| \ _________\____
| \ / operating f
| \ /
| \/ <- intersection = common freq
| /\
| / \ <- Machine 1 droop line
|___/____\_________________ kW load
P1 P2 (P1 + P2 = total load)
Reactive power (kVAR) sharing: Similarly controlled — determined by each ma-
chine’s field excitation. Increasing excitation of one machine increases its reac-
tive power output (and terminal voltage support) without significantly affecting real
power sharing, following a similar “droop” logic on the excitation–VAR characteristic.
Key point: Real power sharing is controlled by prime-mover governor settings (speed
droop); reactive power sharing is controlled by field excitation — the two controls are
largely independent (decoupled).

Lamp Methods of Synchronization

Before closing the breaker to parallel an incoming alternator with a running bus, ordi-
nary lamps can be used as a simple visual indicator that all synchronizing conditions
(equal voltage, equal frequency, matching phase sequence and phase angle) are met.
Circuit setup (typical arrangement):
• G1 = running alternator, already supplying the load through CB-1.
• G2 = incoming alternator, to be paralleled with G1 through CB-2 (kept open
initially).
• Voltmeter V1 measures the main bus bar voltage; V2 measures G2’s output volt-
age.
• A phase sequence indicator (small induction-motor-type device) confirms both
machines have the same phase rotation before any lamp test is attempted.

(a) Dark Lamp Method (all lamps connected straight)

• All three lamps connected across corresponding phases (R1–R2, Y1–Y2, B1–B2).
• Lamp brightness depends on the voltage difference between the incoming ma-
chine and the bus for each phase.
• As G2’s speed is brought close to synchronous speed, all three lamps brighten
and dim together, in unison.
• CB-2 is closed at the instant all three lamps go completely dark.
• Drawback: Since brightness ∝ (voltage difference)2 , lamps appear dark over a
range of angle near zero — making the exact synchronizing instant hard to judge

14
 precisely. It also cannot by itself reveal a wrong phase sequence.

(b) Two-Bright-One-Dark Method (crossed lamp connection) This is a more

informative variant, using a crossed lamp arrangement:
• L1 connected straight — across R1 and R2 (same phase).
• L2 connected crossed — across Y1 and B2.
• L3 connected crossed — across B1 and Y2.
• Lamps L1, L2, L3 are physically arranged in a triangle (hence also called the
“three-lamp” or “rotating lamp” method).
Procedure:
1. G2 is run up to approximately synchronous speed by its prime mover, with CB-2
open.
2. G2’s excitation is adjusted until V2 matches V1 in magnitude.
3. The lamps are observed to confirm phase coincidence and equal frequency be-
fore closing CB-2.
Phasor interpretation:
• The phase voltages of G1 (R1, Y1, B1) and G2 (R2, Y2, B2) can be represented
as two sets of phasors superimposed on one diagram, each rotating at its own
machine’s frequency.
• If both frequencies are exactly equal, both phasor sets rotate at the same
speed, so the angle between corresponding phasors (R1–R2, Y1–B2, B1–Y2) re-
mains constant.
• At the specific instant this constant angle brings R1 and R2 into coincidence:
L1 (straight) goes dark, while L2 and L3 (crossed) glow with equal brightness.
This is the correct instant to close CB-2.
If frequencies are unequal (say G2 runs faster than G1):
• The R2-Y2-B2 phasor set rotates faster than R1-Y1-B1, so the relative angle
between them keeps changing with time.
• Voltage across L1 keeps increasing, across L2 keeps decreasing, across L3
keeps increasing.
• At some point B2 and Y1 coincide → L2 goes dark momentarily, while L1 and
L3 remain lit.
• As the phasors continue to slip relative to each other, the point of darkness shifts
progressively from lamp to lamp — the lamps appear to go dark one after an-
other in a rotating sequence (e.g., anti-clockwise around the triangle).
• This rotating dark pattern indicates the sign of the frequency difference
— its direction tells the operator whether the incoming machine is running faster
or slower than the bus, and its speed of rotation indicates how large the differ-
ence is.
• The operator adjusts G2’s speed (via its governor) until the rotation slows and
stops, reaching the steady condition: L1 dark, L2 = L3 equally bright — at which
point CB-2 is closed and G2 is successfully paralleled with G1.
Key advantage of the two-lamp-crossed method over the simple dark lamp

15
method: it doesn’t just show whether the machines are synchronized — the direction
of lamp rotation tells the operator which way to correct the incoming machine’s speed,
and the closing instant (L1 dark, L2 = L3 equal) is much sharper and easier to judge
than “all lamps dark.”
Both lamp methods have largely been replaced in modern practice by synchroscopes
and automatic synchronizing relays, but remain instructive and are still used as a low-
cost backup method, especially for small generating sets.

16
2.8 Transient Condition, Transient and Sub-transient Reactance
Transient Condition

When a sudden disturbance occurs — such as a sudden short circuit at the terminals
of a synchronous generator, or sudden large load change — the armature current does
not jump instantly to its final steady-state value. Instead, it passes through distinct
phases because of the effect of induced currents in the rotor circuits (field winding
and damper/amortisseur windings) opposing the sudden change in flux (Lenz’s law).
At the very instant of a 3-phase short circuit, the flux linkage with rotor circuits cannot
change instantaneously, so the effective reactance the armature “sees” is much lower
initially (giving very high initial current), and it gradually rises to the final steady-state
synchronous reactance value as the induced rotor currents decay.

Three Distinct Periods

1. Sub-transient period (0 to ~2 cycles):

• Very high initial current.
• Caused by induced currents in both the damper winding and the field winding,
which strongly oppose the change in flux.
• Characterized by sub-transient reactance 𝑋𝑑″ (smallest reactance, largest cur-
rent) — the initial effective reactance.
2. Transient period (~2 cycles to a few cycles/several cycles, up to ~0.5–2
sec):
• Damper winding currents have decayed away (damper winding has low time
constant), but field winding currents are still decaying.
• Characterized by transient reactance 𝑋𝑑′ (medium value):

𝑋𝑑″ < 𝑋𝑑′ < 𝑋𝑑

3. Steady-state period:
• All induced transient currents have died out; only the steady DC field excitation
remains.
• Characterized by the normal synchronous reactance 𝑋𝑑 (largest value, small-
est current) — same as discussed in Section 2.5.

Short-Circuit Current Envelope

The symmetrical AC component of short-circuit current can be approximated as:

√ 1 1 ″ 1 1 ′ 1
𝑖𝑎𝑐 (𝑡) = 2 𝐸 [( ″ − ′ ) 𝑒−𝑡/𝑇𝑑 + ( ′ − ) 𝑒−𝑡/𝑇𝑑 + ] sin(𝜔𝑡 + 𝛼)
𝑋𝑑 𝑋𝑑 𝑋𝑑 𝑋𝑑 𝑋𝑑

17
where 𝑇𝑑″ = sub-transient time constant (very short, ~0.01–0.05 s), 𝑇𝑑′ = transient
time constant (longer, ~0.5–2 s).
A DC offset component (decaying with armature time constant 𝑇𝑎 ) is also present,
depending on the point-on-wave at which the fault occurs, making the actual fault
current asymmetrical initially.

Significance

• These reactances (𝑋𝑑″ , 𝑋𝑑′ , 𝑋𝑑 ) are essential for circuit breaker rating selec-
tion (must interrupt the higher transient/sub-transient fault currents), protec-
tive relay setting/coordination, and stability studies of power systems.
• 𝑋𝑑″ and 𝑋𝑑′ are much smaller than 𝑋𝑑 , meaning fault currents in the first few
cycles can be many times the rated current.

18
2.9 Synchronous Generator Connected to an Infinite Bus
Concept of Infinite Bus

An infinite bus is an idealized large power system (or grid) whose voltage magni-
tude and frequency remain constant, independent of the power drawn from or
delivered to it by any single generator connected to it — because the bus capacity
is very large compared to the individual machine.
When a synchronous generator is connected to an infinite bus:
• Its terminal voltage 𝑉 and frequency 𝑓 are fixed by the bus (cannot be
changed by adjusting the individual machine’s governor or excitation).
• The only quantities the operator can control on that single machine are:
1. Field excitation (𝐼𝑓 ) → controls reactive power (Q) output/absorption, and
hence power factor.
2. Prime-mover input (governor setting) → controls real power (P) output
— increasing steam/fuel input increases the power angle 𝛿 and hence real
power delivered, without changing speed/frequency (which is locked to the
bus).

Power Angle Equation

For a cylindrical-rotor machine connected to an infinite bus (neglecting 𝑅𝑎 ):

𝐸𝑉
𝑃 = sin 𝛿
𝑋𝑠

where 𝐸 = internal EMF, 𝑉 = bus (terminal) voltage, 𝑋𝑠 = synchronous reactance, 𝛿

= power (torque) angle between 𝐸 and 𝑉 .
• Increasing prime-mover input increases 𝛿 , increasing P delivered (up to the max-
imum at 𝛿 = 90°, the steady-state stability limit).
• Reactive power:

𝐸𝑉 cos 𝛿 − 𝑉 2
𝑄=
𝑋𝑠
• If 𝐸 cos 𝛿 > 𝑉 → machine delivers (supplies) reactive power to bus (overexcited,
lagging pf as seen from the system).
• If 𝐸 cos 𝛿 < 𝑉 → machine absorbs reactive power from bus (underexcited, lead-
ing pf / acts partly like an inductor).

Key Characteristics

• Since 𝑓 is fixed by the bus, the generator’s speed is rigidly locked to synchronous
speed — it cannot “run away” or slow down; any excess mechanical input beyond
what can be transferred as electrical power (at 𝛿 = 90°) causes loss of synchro-
nism (pole slipping).

19
• This is the standard model used for analyzing a single generator’s behavior
within a large interconnected grid.

20
2.10 Limits of Active and Reactive Power Generation — Capability
Curve
The capability curve (or capability chart) of a synchronous generator defines the
safe operating region in the P–Q (real power vs reactive power) plane, bounded by
several physical limits.

Axes

• Horizontal axis: Reactive power Q (MVAR) — leading (absorbing) to the left,

lagging (delivering) to the right.
• Vertical axis: Active power P (MW).
• The apparent power 𝑆 = √𝑃 2 + 𝑄2 is represented as radial distance from the
origin.

Limiting Boundaries

1. Armature (Stator) Current Limit:

• Determined by the maximum current the stator winding can carry continuously
without overheating
√ (thermal limit of armature copper/insulation).
• Since 𝑆 = 3𝑉 𝐼𝑎 , this limit appears as a circular arc of radius = rated MVA
(𝑆𝑟𝑎𝑡𝑒𝑑 ), centered at the origin.
2. Field (Rotor) Current Limit:
• Determined by maximum permissible field current, limited by rotor winding heat-
ing.
• Since 𝐸 ∝ 𝐼𝑓 , and the relation between P, Q, E, V, 𝑋𝑠 gives:

2 2
2 𝑉2 𝐸𝑉
𝑃 + (𝑄 + ) =( )
𝑋𝑠 𝑋𝑠
• This is also a circular arc, but centered at (𝑄 = −𝑉 2 /𝑋𝑠 , 𝑃 = 0) on the Q-
axis — i.e., offset into the leading/underexcited region. This boundary limits the
overexcited (lagging, high-Q-delivery) operating region.
3. Prime Mover (Turbine) Limit:
• A horizontal straight line at 𝑃 = 𝑃𝑚𝑎𝑥 (rated MW of the turbine/engine) —
cannot be exceeded regardless of electrical limits, since it’s a mechanical input
constraint.
4. Stability Limit (Underexcited Region):
• In the underexcited (leading pf, absorbing VAR) region, if excitation is reduced
too far, the machine risks loss of synchronism (falls out of step) and also stator
end-core heating due to increased leakage flux at low excitation.
• This appears as a curved boundary limiting how far into the leading-Q / under-
excited region the machine can safely operate, usually with a margin below the
theoretical steady-state stability limit (𝛿 = 90°).

21
Composite Capability Curve

The overall safe operating region is the area bounded simultaneously by the turbine
limit (horizontal line at 𝑃𝑚𝑎𝑥 ), the armature current limit (circular arc centered at
the origin), the field current limit (circular arc offset along the Q-axis, restricting the
overexcited/lagging side), and the underexcited stability limit (curved boundary on
the leading side).
P (MW)
|
Pmax |________________________ <- Turbine limit
| \
| Field current \
| limit (arc) \
| ___ \
| / \ \ <- Armature current
| / \ \ limit (arc)
|/ \_________________\
--------+----------------------------> Q (MVAR)
Underexcited Overexcited
(leading, stability (lagging,
limit region) delivering VAR)
Practical use: Plant operators use the capability curve to determine the maximum
safe combination of real and reactive power the generator can deliver at any oper-
ating point, ensuring none of the thermal or stability limits are exceeded simultane-
ously.

22
2.11 Losses and Efficiency
Classification of Losses in a Synchronous Generator

1. Copper (Electrical) Losses (𝐼 2 𝑅 losses) — vary with load:

• Armature copper loss: = 3𝐼𝑎2 𝑅𝑎 (per-phase resistance 𝑅𝑎 , for 3 phases) —
varies with the square of load current.
• Field copper loss: = 𝐼𝑓2 𝑅𝑓 — power dissipated in the DC field winding re-
sistance (relatively constant if excitation is maintained roughly constant, but
changes if excitation is adjusted for pf control).
• Loss in exciter (if a separate DC exciter or brushless exciter is used to supply
field current) — sometimes included depending on whether efficiency is calcu-
lated including the exciter.
2. Core (Iron) Losses — approximately constant, depend on flux and fre-
quency, not load current:
• Hysteresis loss: Due to repeated magnetization reversal of the core material.

1.6
𝑃ℎ = 𝐾ℎ 𝑓𝐵𝑚𝑎𝑥
• Eddy current loss: Due to circulating currents induced in the laminated core
by the alternating flux.

𝑃𝑒 = 𝐾𝑒 𝑓 2 𝐵𝑚𝑎𝑥
2

• Minimized by using laminated silicon-steel cores.

• Since flux and frequency are held nearly constant (governed by fixed excitation
and synchronous speed), core losses are essentially constant regardless of load.
3. Mechanical Losses — constant, independent of electrical load:
• Friction loss: At bearings.
• Windage loss: Air resistance to rotor rotation (and fan/blower losses if forced
ventilation/cooling is used).
• These depend only on speed (which is constant), so they remain essentially fixed.
4. Stray Load Losses:
• Additional losses caused by leakage flux inducing eddy currents in structural
steel parts, end-windings, etc., under load — usually small, estimated as a per-
centage of output or measured separately.

Constant Losses vs Variable Losses

• Constant losses (𝑊𝑐 ): Core loss + mechanical (friction & windage) loss + field
copper loss (if excitation held fixed) — do not change significantly with the load
current.
• Variable losses (𝑊𝑣 ): Armature copper loss (3𝐼𝑎2 𝑅𝑎 ) + stray load loss — in-
crease with the square of load current.

23
Efficiency
Output Power Output
𝜂= × 100 = × 100
Input Power Output + Losses
√
3 𝑉 𝐼𝑎 cos 𝜙
𝜂=√ × 100
3 𝑉 𝐼𝑎 cos 𝜙 + 𝑊𝑐 + 3𝐼𝑎2 𝑅𝑎 + other losses
√
(3-phase output = 3𝑉𝐿 𝐼𝐿 cos 𝜙, or 3𝑉𝑝ℎ 𝐼𝑝ℎ cos 𝜙)

Condition for Maximum Efficiency

For a given power factor and terminal voltage, efficiency is maximum when variable
(copper) losses equal constant losses:

3𝐼𝑎2 𝑅𝑎 = 𝑊𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡

This gives the load current (and hence the load) at which the machine operates most
efficiently — an important design/operational consideration for sizing generators rel-
ative to typical expected load.

Typical Efficiency Values

Large synchronous generators (utility-scale, tens to hundreds of MW) typically have

efficiencies in the range of 95–99%, since 𝐼 2 𝑅 and core losses form a very small
fraction of the very large output power. Smaller alternators have comparatively lower
efficiencies (85–92%) due to a larger relative proportion of fixed losses.

24