[Downloaded from www.aece.ro on Monday, December 11, 2017 at 17:09:42 (UTC) by 154.16.54.220. Redistribution subject to AECE license or copyright.
Advances in Electrical and Computer Engineering
Determination of Operation Characteristics of a
Synchronous Generator by Static Experimental
Tests
Ion-Daniel ILINA, Tiberiu TUDORACHE
Politehnica University of Bucharest, 060042, Romania
daniel.ilina@upb.ro
1
Abstract—This paper deals with the determination of the
operation characteristics of a Synchronous Generator (SG)
using static experimental tests (current and voltage decay
tests). Using these tests the magnetization characteristic, the
machine parameters (synchronous magnetization inductances
and differential magnetization inductances), the no-load
characteristic and the external characteristic were determined.
The magnetization characteristic, the parameters and the
operational characteristics provide important information
about the performance of synchronous generator, machine
which is currently used in most power plants. Compared to
classical experimental methods with moving rotor and
numerical methods of field computation that require the
knowledge of geometric dimensions and material properties,
static experimental methods are distinguished by several
advantages: simple implementation for any type of SG, quick
results, low power consumption and no effect on the drive
system where the machine is integrated. The mathematical
model of SG uses dedicated Matlab-Simulink programs and
the results obtained by static methods are compared with those
obtained by classical methods. Also to better approach the
phenomenon of magnetic saturation and a more accurate
estimate of the parameters, the magnetization characteristic,
synchronous magnetization inductances and differential
magnetization inductances are determined versus the total
magnetizing current.
Index Terms—electric machines, generators, parameter
estimation, performance evaluation, circuit testing.
I. INTRODUCTION
The synchronous machine is intensively used in different
industrial applications such as: electrical power plants,
electric drive systems, automation systems, robotics,
automotive. A lot of theoretical and experimental studies are
dedicated to this electrical machine. The operation
characteristics of synchronous machine offer important
information about the performance of the machine in
various working conditions. The methods for evaluating the
operation characteristics of synchronous machine are
numerous [1-4]. Obviously the most used of them are based
on experimental tests with moving rotor and these ones can
be considered classical [3, 4]. In case of large power
machines the classical tests are however difficult to be put in
practice and they involve important electric energy
consumption. In such cases mathematical estimations or
other experimental methods easy to implement are preferred
[5-16].
1 total magnetizing current and new parameters (differential
The work has been funded by the Sectoral Operational Programme
Human Resources Development 2007-2013 of the Ministry of European
Funds through the Financial Agreement POSDRU/159/1.5/S/ ID 134398.
Digital Object Identifier 10.4316/AECE.2016.02013
1582-7445 © 2016 AECE
Volume 16, Number 2, 2016
Mathematical estimations involve the use of dedicated
machine models that typically include the voltage equations
of different circuits of the machine and they can be
described by specific equivalent electric diagrams. The
mathematical models include the electrical parameters of the
machine that can be determined by electromagnetic field
computations or by other experimental methods. In the last
decades the numerical methods have been increasingly used
for electromagnetic field analysis of electrical machines of
being characterized by good versatility and accuracy [17-
21]. However these methods require a proper knowledge of
the electric and magnetic properties of involved materials as
well as the geometrical dimensions of the machine, data
which are not available in many cases.
Among the mathematical models used for the analysis of
electrical machines we can mention variable parameters
models, concentrated parameters models, models in phase
coordinates or in orthogonal coordinates.
In case of synchronous machine, due to its constructive
particularities and due to the ease of computations the d-q
mathematic model with concentrated parameters has been
the most used so far.
As mentioned before, to determine the operation
characteristics of synchronous machine we can use other
experimental methods that can be easily implemented
without important energy consumption. In this category we
can mention the static experimental methods (static rotor)
which offer some real advantages: easy implementation,
reduced consumption of energy, no impact on the electric
drive of the machine [22-28].
Usually in the case of low power synchronous generators
the operation characteristics are determined by classical
methods and in the case of high power synchronous
generators these characteristics are determined by static
methods. Most often these results obtained by the two
methods are not compared. In the same time the
magnetization characteristic is determined typically
depending on the field current considering the synchronous
magnetization inductances constant.
In this paper for the experimental tests was used a low
power synchronous generator and the operation
characteristics obtained by the static methods are compared
with those obtained by the classical methods. Also the
magnetization characteristic and the synchronous
magnetization inductances are determined depending on the
magnetization inductances) are introduced. Thus a better
93
 [Downloaded from www.aece.ro on Monday, December 11, 2017 at 17:09:42 (UTC) by 154.16.54.220. Redistribution subject to AECE license or copyright.]

Advances in Electrical and Computer Engineering Volume 16, Number 2, 2016

approach of the magnetic saturation phenomenon is obtained As far as the magnetic fluxes are concerned, if we split
and the SG parameters are estimated more accurately. the linkage and leakage fluxes and neglect the mutual
The SG is analyzed using the d-q mathematical model and leakage inductances between the adjacent windings of the
the static experimental methods suppose current and voltage model, we obtain the relations [3, 22]:
decay tests in the stator and rotor circuits. In order to apply d  L s i d  dm , q  L s i q  qm , (6), (7)
the method mentioned in this paper to high power machines,
 D  L D i D  dm , Q  LQ iQ  qm , (8), (9)
the stationary rotor requires the possibility of manual or
mechanical positioning with respect to the stator.  E  L E i E  dm , (10)
where: Ψdm and Ψqm represent the components along d and q
II. MATHEMATICAL MODEL OF SYNCHRONOUS MACHINE axes of the linkage magnetic flux; Lsσ is the leakage
A synchronous machine is usually equipped with a three- inductivity of stator winding; LEσ, LDσ, LQσ are the leakage
phase symmetric winding system on the stator and with a inductances of the field circuit and of the D and Q circuits
field winding and a damping cage on the rotor (Fig. 1). The corresponding to the damping cage.
magnetic field produced by the field winding is oriented In order to take into account the magnetic saturation we
along the longitudinal axis (d) and that produced by the define the magnetization characteristics (for d and q axes) as
damping cage (situated in the main poles shoes) has two a dependence between linkage magnetic fluxes and total
components one along the longitudinal axis (d) and the other magnetizing current of the machine [3, 22]:
along the transversal axis (q). The d-q mathematical model  dm (i m ),  qm (i m ), (11), (12)
supposes the replacement of the three-phase winding system
2 2
of the synchronous machine with two windings situated on i m  i dm  i qm , (13)
the orthogonal axes d and q. The orthogonal model is based where:
on the following assumptions: anisotropic and symmetric i dm  i d  i E  i D , i qm  i q  iQ . (14), (15)
machine; the stator windings are three-phase, symmetrical
and sinusoidally distributed, without space harmonics; the The two magnetization curves are identical in case of
field coil is placed in the longitudinal axis; supplementary cylindrical rotor SGs and they are different in case of salient
short-circuited coils are placed in the d and q axes poles SGs. Taking into account the magnetization
respectively; the hysteresis phenomenon is neglected [3, 22]. characteristics defined before we determine the components
of the linkage magnetic fluxes by introducing two new
inductances [3, 22]:
 (i )  qm (i m )
dm  dm m  i dm , qm   i qm , (16), (17)
im im
 dm (i m )  qm (i m )
Ldm (i m )  , Lqm (i m )  , (18), (19)
im im
where: Ldm(im), Lqm(im) are magnetization synchronous
inductances depending on the total magnetizing current of
Figure 1. Synchronous machine: a) real model; b) orthogonal model
the machine.
In the same time, to take into account as precise as
The equivalence relations between the quantities of the possible the electromagnetic phenomena in transient
real synchronous machine and those of the d-q mathematical regimes we can take into account also the differential
model are obtained by Park’s transformation [3, 22]. Park’s magnetization inductances [3, 22]:
matrix is invertible so as: [ P ] 1  [ P ]T .  dm (i m )  qm (i m )
L dm dif (i m )  , Lqm dif (i m )  . (20), (21)
Taking the rotor as reference we can write the equations i m i m
of the d-q mathematical model of synchronous machine
operating as generator: III. DETERMINATION OF MAGNETIZATION
d d CHARACTERISTICS
id Rs  u d     r q , (1)
dt The magnetization characteristics of SG can be
d q determined by current decay tests carried out both along the
iq Rs  u q     r d , (2) longitudinal axis (d) and the transversal axis (q). These
dt
experimental tests suppose to supply the stator windings and
dE the field winding in DC, to short-circuit these circuits and to
iE RE  u E   , (3)
dt register the current and voltage decay [3, 22-26]. Neither the
d D dQ temperature variation of stator winding resistance nor the
iD RD   , i Q RQ   , (4), (5) residual short-circuit current due to the small short-circuit
dt dt
resistance are taken into account.
where: id, iq, iE, iD, iQ, ud, uq, uE, Ψd, Ψq, ΨE, ΨD, ΨQ
First the rotor should be positioned with the d axis along
represent the currents, the voltages and the magnetic fluxes
the U phase. Then the current decay tests can be done along
corresponding to the d - q mathematical model of the SG; Rs
the d axis and q axis.
- stator winding resistance per phase; RE - field circuit
resistance; ωr - angular speed of the rotor.

94
 [Downloaded from www.aece.ro on Monday, December 11, 2017 at 17:09:42 (UTC) by 154.16.54.220. Redistribution subject to AECE license or copyright.]

Advances in Electrical and Computer Engineering Volume 16, Number 2, 2016

A. Longitudinal magnetization characteristic (d axis) From (27) and (30) we obtain the expression of d axis
In order to carry out the d axis current decay tests, the U magnetization characteristic:
phase winding should be series connected with the parallel 3 t  3 2 t 
connected V and W phase windings as shown in Fig. 2. Both  dm  R s  iU dt L s IU 0   uUVW dt. (31)
2 0 2 3 0
the field circuit and stator circuit are DC supplied (currents
Using (18), (20), (28) and (31) we can determine the
IE0, IU0). Then the stator circuit is short-circuited while the
synchronous magnetization inductance ( L dm ) and
field circuit remains DC supplied the stator current decay
being registered. differential magnetization inductance ( L dm dif ).

B. Transversal magnetization characteristic (q axis)

The rotor position should be kept with d axis along phase
U, but in this case only V and W phase windings should be
series connected (Fig. 3).

Figure 2. Windings connections for d axis current decay test

Using Park’s transformation in static state (   0 ) and

taking into account the type of connection of stator windings
for which iV  iW  iU / 2 , we obtain the expressions [3,
22]: Figure 3. Windings connections for q axis current decay test

3 2 We supply the stator circuit and field circuit with DC (IV0,

id   iU , u d  uUVW  , (22), (23)
2 3 IE0) and then the stator circuit is short-circuited and we
i q  0, u q  0. (24), (25) register the current decay. Using Park’s transformation in
static state (   0 ) and taking into account the connection
From (1) for static state (  r  0 ) we short-circuit the
type of stator windings for which iU  0, iV  iW we obtain
stator windings ( uUVW  0 ) and we obtain the voltage
the expressions [3, 22]:
equation of d axis stator circuit:
u
dd i q  2  iV , U q  VW , (32), (33)
id Rs   . (26) 2
dt
By integrating (26) and using (6), (11), (13), (14), (16), i d  0 , U d  0. (34), (35)
(22) and (23) we obtain a mathematical expression for the d If we short circuit the stator circuit ( uVW  0 ) with static
axis magnetization characteristics: state (  r  0 ) we obtain the corresponding equation of q
3 t  3 axis:
 dm  R s  iU dt L s I U 0  ( dm ) f , (27)
2 0 2 dq
iq Rs   . (36)
im  3 / 2  I U 0  I E 0 . (28) dt
We can notice that the quantities in (27), (28) are known If we integrate the previous expression and use (7), (12),
or can be determined. Thus the stator windings resistance (13), (15), (17), (32) and (33) we obtain the mathematical
(Rs) can be determined by measurement in DC and the expression of q axis magnetization characteristic:
leakage inductance Lsσ can be approximated with the  t  
 2  Rs   iV dt  Ls  2  IV 0   I E2 0  2  IV2 0
homopolar inductance. This inductance can be emphasized  
 qm   0  , (37)
when a homopolar currents system flowing through the 2  IV 0
stator windings determines only leakage magnetic fields [3,
4]. i m  I E2 0  2  I V2 0 . (38)
The integral from (27) is determined from current decay The quantities appearing in (37) and (38) are known or
characteristic and  dm  f represents the total magnetic flux they can be determined. The integral from (37) can be
produced by the field current. This magnetic flux can be determined by registering the stator current decay. From
determined by a field current decay test with open stator (19), (21), (37) and (38) we can determine the magnetization
circuit ( i d  0 ) registering the voltage at the stator synchronous inductance ( L qm ) and the differential
terminals. In this case the equation of stator circuit along d magnetization inductance ( Lqm dif ).
axis becomes:
dd IV. DETERMINING THE NO-LOAD AND EXTERNAL
ud  
. (29)
dt CHARACTERISTIC OF SG
By integrating (29) and taking into account (6), (23) The no-load and external characteristics are defined as:
results: U l  f ( I E ) n ct . , U l  f ( I ) n ct. , (39)
2 t  I 0 I E  ct .
( dm ) f   uUVW dt. (30) cos  2  ct .
3 0
where Ul is the line voltage across the generator terminals, IE

95
 [Downloaded from www.aece.ro on Monday, December 11, 2017 at 17:09:42 (UTC) by 154.16.54.220. Redistribution subject to AECE license or copyright.]

Advances in Electrical and Computer Engineering Volume 16, Number 2, 2016

is the field current, I is the load current [3, 4]. the expression of terminal voltage of SG per phase:
By the classical experimental methods these
characteristics are determined from no-load and load tests of
U E 0  X d I d 2  X q I q 2  Rs I . (46)
the machine [3, 4]. In some cases, especially for large SG,
these tests are difficult to be done and they are
uneconomical since they entail to decouple the SG from its
drive system and to use high power loads. By using static
experimental methods these drawbacks are removed since
they are easy to implement but also and economic.
A. No-load characteristic of SG
At no-load the voltage across the stator terminals of the
SG is influenced only by the excitation magnetic field. Figure 4. Phasor diagram for SG operating under resistive load
By considering U phase of the SG, the voltage at no-load
in steady state is given by the expression: At no-load the e.m.f. is equal with the terminal voltage,
U U    U . (40) this value being determined in the previous section of the
paper:
Taking into account the equivalence relations between the
quantities of the real machine and of the d-q mathematical 1 2 t  
E0       uUVW dt . (47)
model and the fact that the field circuit is positioned along d 3 
2 0 
axis results: Therefore, for a given value of the field current a current
2 decay test should be done and the decay of the terminal
U  d . (41) voltage is registered. The voltage drops across the
3
The magnetic flux along d axis, produced exclusively by synchronous reactances of the machine are determined from
the field current was determined before by current decay the reaction magnetic field corresponding to the operation of
tests with stator circuit open and by registering the voltage the machine under load. Thus:
across the terminal voltage, so:  1 2 r
X d I d    , (48)
2 t   2 3 d
d  ( dm ) f   
3 0
 uUVW dt. (42)
 1 2  r
From (40-42) results: X q I q    q . (49)
 2 3 

1 2 t  
UU       uUVW dt . (43) The reaction magnetic fluxes ( dr , qr ), produced by the
3 
2 0 
load currents, can be obtained from d and q axis current
1
The term is justified by the fact that the current decay tests (described in the previous chapter) carried out
2
2 1
decay tests are carried out by supplying the circuits in DC without supplying the field circuit. The constant is
and the voltage obtained at the stator terminals is determined 2 3
as the effective value of a sinusoidal quantity. If the stator obtained from Park’s transformation and from the
winding of the machine is star connected: equivalence of quantities corresponding to the current decay
tests and the effective values of the phasor quantities.
U l  3U U . (44)
Considering the d and q axes current decay tests without
The field current decay test is carried out for different supplying the field circuit and the relations (22), (26), (32)
values of this current in increasing order will result a and (36) results:
dependence of terminal voltage of the generator as a
3 t  t 
function of the field current (no-load characteristic). dr  R s  iU dt , qr  2 R s  iV dt. (50), (51)
2 0 0
B. External characteristic
From (48-51) we obtain:
The external characteristic of SG was determined for t 

resistive, inductive and capacitive loads. To determine the X d Id   R s  iU dt , (52)
terminal voltage under load operation we take into account 2 0
the voltage equation per phase of the salient pole SG [3]: 2 t 

U  E 0  R s I  jX d I d  jX q I q , (45)
where: E0 is the no-load e.m.f. phasor; Rs - is the stator
winding resistance per phase; Xd, Xq - are the synchronous
longitudinal and transversal reactances.
To solve this equation we use the phasor diagrams for the
three load types.
1) External characteristics – resistive load
If the SG operates under resistive load the phasor diagram
corresponding to (45) is shown in Fig. 4.
By solving geometrically the phasor diagram we obtain
X q I q  
 R s  iV dt. (53)
3 0
Since the stator current decay tests are done in DC, when
calculating the voltage drop on the stator resistance ( Rs I )
from (46), relation expressed in phasor quantities, we can
consider the decaying current divided by 2 . Using (46),
(47), (52) and (53) we can determine the line voltage
( U l  3U for star connection) across the SG terminals
corresponding to a given stator load current and to a given

96
 [Downloaded from www.aece.ro on Monday, December 11, 2017 at 17:09:42 (UTC) by 154.16.54.220. Redistribution subject to AECE license or copyright.]
Advances in Electrical and Computer Engineering
field current. The determination is done for increasing
values of the stator current resulting thus the dependence
between the terminal voltage and the load current (i.e. the
external characteristic).
2) External characteristics – inductive load
In case of an inductive load the corresponding phasor
diagram is shown in Fig. 5.
Figure 5. Phasor diagram for SG operating under inductive load
The expression of the phase voltage of SG obtained from
the phasor diagram is given below:
U E0  X d I d 2  X q I q 2  Rs I 2 .
The no-load e.m.f. can be computed using (47) as in the
case of resistive load. Sine the load of the SG is inductive
the voltage leads the load current with 90°. This phase-shift
angle should be introduced also in the current decay tests
carried out in DC. Thus, from the current decay tests for d
axis we can determine the quantities corresponding to the q
axis and from the current decay tests for q axis we can
determine the quantities corresponding to the d axis. Thus:
t 
2 along the phase U. This positioning of the rotor can be done
X d I d    R s  iV dt ,
3 0
t 

X qIq   R s  iU dt. (56)
2 0
In case of an inductive load the no-load e.m.f. is
determined from the field current decay test by registering
the voltage across the stator terminals. The voltage drops
across the synchronous reactances are determined also from
current decay tests along d and q axes without supplying the
field circuit. From these tests carried out for increasing
values of the stator current, we can determine using (38) the
terminal voltage obtaining thus the external characteristic of
the SG.
3) External characteristics – capacitive load
If the load is capacitive the phasor diagram of the
machine corresponding to (45) is the following:
Figure 6. Phasor diagram for SG operating under capacitive load
The expression of the terminal voltage corresponding to
the phasor diagram shown in Fig. 6 is the following:
U E 0  X d I d 2  X q I q 2  R s I 2 .
In a similar way with the case of inductive and resistive
loads the no-load e.m.f. can be determined using (47) by
field current decay tests. In case of capacitive loads the
terminal voltage lags the load current with an angle of 90°
Volume 16, Number 2, 2016
and this phase-shift angle should be considered during the
current decay tests carried out in DC. The voltage drops on
the synchronous reactances of the machine can be
determined using (55), (56) by the stator current decay tests
along the two axes. The experimental measurements are
carried out for increasing values of the stator currents,
resulting thus the dependence between the terminal voltage
and the load current which means the external characteristic
of the SG.
V. EXPERIMENTAL TESTS AND RESULTS
The experimental tests were carried out for a salient pole
SG characterized by the following main data:
Sn = 2500 [VA]; Un = 208/120 [V]; In = 6.95/12 [A];
nn = 1500 [rot/min]; Uen = 70 [V]; Ien = 2.93 [A];
cos φn = 0.8.
The phase resistance values was determined by
(54) measurements in DC its value being Rs = 1 [Ω]. The leakage
inductance can be approximated with the homopolar
inductance which was determined Lsσ = 0.0044 [H].
The acquisition of currents and voltages during the
experimental tests were done using an oscilloscope
(Tektronix TDS 2000B) and the mathematical calculations
were done using Matlab-Simulink dedicated programs. In
Fig. 7 is presented the experimental setup used for the
laboratory measurements. All the experimental static tests
supposed first to position the rotor of the SG with its d axis
(55)
using the electric diagram shown in Fig. 8.
Figure 7. Experimental setup
Figure 8. Electric diagram used for positioning the rotor with its d axis
along phase U
For that purpose the V and W phase windings are series
connected and supplied from an autotransformer with a
reduced AC voltage. The field circuit is short-circuited by
means of an ammeter. The rotor is rotated manually or using
a mechanical system until the ammeter indication in the
field circuit (AE) is zero.
(57)
A. Determination of longitudinal magnetization
characteristic (d axis)
To determine the longitudinal magnetization
characteristic by current decay tests we use the following
electric diagram.
97
 [Downloaded from www.aece.ro on Monday, December 11, 2017 at 17:09:42 (UTC) by 154.16.54.220. Redistribution subject to AECE license or copyright.]
Advances in Electrical and Computer Engineering
Figure 9. Electric diagram for determining the longitudinal magnetization
characteristic
We can notice that the supply of stator circuit and of field
circuit should be carried out from DC voltage sources that
allow short circuits at the terminals. For solving this
problem we used as DC voltage source a shunt DC
generator. By following the procedures described in section
III for increasing values of stator and field currents we
obtain the longitudinal magnetization characteristic, the
synchronous magnetization inductances and the differential
magnetization inductances of the SG. In the next figures are
shown the registered results during the current and voltage
decay tests for the stator circuit.
Figure 10. Stator current decay for IU0 = 7 [A] and IE0 = 3 [A]
Figure 11. Terminal voltage decay for a field current IE0 = 3 [A]
In Fig. 12 is presented the longitudinal magnetization
characteristic of the SG as a function of the total
magnetizing current.
Figure 12. Longitudinal magnetization characteristic (d axis) of SG
98
Volume 16, Number 2, 2016
Figure 13. Variation of synchronous magnetization inductance (continuous
line) and of differential magnetization inductance (interrupt line)
corresponding to d axis
In Fig. 13 are shown the variations of the synchronous
magnetization inductance and of the differential
magnetization inductance as a function of the total
magnetizing current.
By analyzing the previous results we can notice the
saturation of longitudinal magnetization characteristic, the
smooth variation of the synchronous magnetization
inductance and the accentuated variation of the differential
magnetization inductance.
B. Determination of transversal magnetization
characteristic (q axis)
The transversal magnetization characteristic is determined
experimentally by a procedure similar to the one used for
the longitudinal magnetization characteristic.
The phases of the experimental tests are described in a
previous section and the electric diagram used for
measurements is shown in Fig. 14. The registered current
decay is shown in Fig. 15 and the transversal magnetization
characteristic is presented graphically in Fig. 16.
Figure 14. Electric diagram used for the determination of transversal
magnetization characteristic
Figure 15. Stator current decay for IV0 = 7 [A] and IE0 = 3 [A]
We can notice that this characteristic has almost no
saturation effect. This fact can be explained by the
particularities of the SG used for measurements. This
machine has four salient poles, the ration between the height
and width of the magnetic pole being of approximately 2.2
which means the inter-polar zone is very large.
 [Downloaded from www.aece.ro on Monday, December 11, 2017 at 17:09:42 (UTC) by 154.16.54.220. Redistribution subject to AECE license or copyright.]
Advances in Electrical and Computer Engineering
Figure 16. Transversal magnetization characteristic (q axis) of SG
In Fig. 17 is represented the variation of synchronous
magnetization inductance and of differential magnetization
inductance as functions of total magnetizing current.
In case of q axis we can notice the decreasing trend for
both the synchronous magnetization inductance and the
differential magnetization inductance (more accentuated for
the last one) with the increase of the total magnetizing
current.
Figure 17. Variation of synchronous magnetization inductance (continuous
line) and of differential magnetization inductance (interrupt line)
corresponding to q axis
C. Determination of no-load characteristic and of external
characteristic of SG
The classical methods (i.e. no-load test and test under
load) for determining the no-load and external
characteristics of SG suppose first to turn the rotor of the
machine at the rated speed. Then to determine the no-load
characteristic of the machine the field circuit should be DC
supplied and the stator circuit should be open. Then we
measure the stator terminal voltage and we represent
graphically the voltage versus the field current. In case of
the external characteristic of the SG the field circuit is DC
supplied (usually with rated current value) and we connect
at the stator terminals a three-phase load and we measure the
terminal voltage and represent graphically the voltage versus
the load current. The static methods used for determining
the no-load and external characteristics of the SG suppose to
position first the rotor with its d axis along the U phase and
to carry out current decay tests for the stator current and for
the field current as described previously.
1) Determination of no-load characteristic of SG
To determine this characteristic we use the electric
diagram shown in Fig. 9. Then we carry out the field current
decay test with the stator circuit open and we register the
voltage at the stator terminals. These tests are done for
different field current values between: 0  I en . Using (21)
and (22) we obtain the no-load characteristic (Fig. 18).
Volume 16, Number 2, 2016
Figure 18. No-load characteristics determined by classical methods
(continuous line) and by static methods (interrupt line)
We can notice that the difference between the two curves
in Fig. 18 is very small. If we consider as reference the
curve obtained by classical methods, the relative errors
obtained by using static methods are less than 6.12%.
2) Determination of external characteristic
The determination of external characteristic by static
methods supposes first the determination of no-load e.m.f.,
which is equal to the terminal voltage determined in the
previous section. The no-load e.m.f. is determined for the
rated value of the field current. Then the current decay tests
are done for the d and q stator currents at increasing current
values without supplying the field circuit. We use the
electric diagrams shown in Fig. 9 and Fig. 14 and depending
on the load type we use the relations described in section IV.
Thus, for resistive loads we use (24), (25) and (28), in
case of inductive loads we apply (25), (29) and (30) and for
capacitive loads the relations (25), (30) and (31) are used.
By applying the previously described procedure we
represent graphically the external characteristic of the SG
for resistive, inductive and capacitive loads, as shown in
Fig. 19. In the same figure is represented the external
characteristic obtained by classical experimental methods.
By analyzing comparatively the two characteristics we
can notice a good agreement between them, the relative
errors being smaller than 4.9%.
Figure 19. External characteristic (for Ien = 2.93 [A]) determined by
classical methods (continuous line) and by static methods (interrupt line)
VI. CONCLUSIONS
Based on the theoretical considerations, on the
experimental methods analyzed and on the results obtained
we can formulate some concluding remarks as follows.
The static experimental methods used for determining the
operation characteristics of SGs present several advantages
over the classical experimental methods: easy to implement
99
 [Downloaded from www.aece.ro on Monday, December 11, 2017 at 17:09:42 (UTC) by 154.16.54.220. Redistribution subject to AECE license or copyright.]
Advances in Electrical and Computer Engineering
and to process the results, reduced consumption of energy,
not impact on the electric drive of the machine.
The determination of magnetization characteristics as
function of total magnetizing current offers a better
approach of the magnetic saturation phenomenon.
The determination of synchronous magnetization
inductances as a function of total magnetizing current
(considered constant in many cases) and the introduction of
differential magnetization inductance lead to more precise
estimation of SG parameters.
The differences between the results (i.e. the operation
characteristics) obtained by classical methods and by static
methods can be explained by the inherent measurement
errors of meters used and of the acquisition system used for
registering the current and voltage decay curves.
This paper presents in detail specific calculation
procedures and the obtained results prove the usefulness and
accuracy of static methods used to determine the operation
characteristics of the SGs.
REFERENCES
[1] E. C. ShaRer, C. A. Gross, “Methods for Determining Linear
Synchronous Machine Parameters,” in Proc. 26th Southeastern
Symposium on System Theory, Athens, 1994, pp. 411-415.
doi:10.1109/SSST.1994.287842
[2] M. Kazerooni, N.C. Kar, “Methods for Determining the Parameters
and Characteristics of PMSM,” in Proc. IEEE Int. Conf. Electric
Machines & Drives Conference (IEMDC), Niagara Falls, 2011, pp.
955-960. doi:10.1109/IEMDC.2011.5994944
[3] I. Boldea, “The Electric Generators Handbook – Synchronous
Generators”, chapter 4, pp. 18-48, chapter 5, pp. 8-39, Taylor &
Francis Group CRC Press, 2006.
[4] P. Vas, “Parameter Estimation, Condition Monitoring, and Diagnosis
of Electrical Machines”, pp. 231-245, Oxford Science Publications,
1993.
[5] F.P. de Mello, L.N. Hannett, “Determination of Synchronous Machine
Electrical Characteristics by Test,” IEEE Transactions on Power
Apparatus and Systems, vol. PAS-102, no. 12, pp. 3810 – 3815, 1983.
doi:10.1109/TPAS.1983.317876
[6] A. Tumageanian, A. Keyhani, “Identification of Synchronous
Machine Linear Parameters for Standstill Step Voltage Input Data,”
IEEE Transactions on Energy Conversion, vol. 10, no. 2, pp. 232-240,
1995. doi:10.1109/60.391887
[7] J.J.R. Melgoza, G.T. Heydt, A. Keyhani, B.L. Agrawal, “Synchronous
machine parameter estimation using the Hartley series,” IEEE
Transactions on Energy Conversion, vol. 16 , no. 1, pp. 49-54, 2001.
doi:10.1109/60.911403
[8] B. Stumberger, G. Stumberger, D. Dolinar, A. Hamler, M. Trlep,
“Evaluation of Saturation and Cross-Magnetization Effects in Interior
Permanent-Magnet Synchronous Motor,” IEEE Transaction on
Industry Applications, vol. 39, no. 5, pp. 1264-1271, 2003.
doi:10.1109/TIA.2003.816538
[9] H.B. Karayaka, A. Keyhani, G.T. Heydt, B.L. Agrawal,
“Synchronous generator model identification and parameter
estimation from operating data,” IEEE Transactions on Energy
Conversion, vol. 18, no. 1, pp. 121-126, 2003.
doi:10.1109/10.1109/TEC.2002.808347
[10] R.D. Fard, M. Karrari, O.P. Malik, “Synchronous generator model
identification for control application using volterra series”, IEEE
Transactions on Energy Conversion, vol. 20, no. 4, pp. 852-858, 2005.
doi:10.1109/TEC.2005.847997
[11] Y.N. Sarem, J. Poshtan, M. Ghomi, M. Poshtan, “Synchronous
generator parameters estimation,” in Proc. Intelligent and Advanced
Systems (ICIAS), Kuala Lumpur, 2007, pp. 870-875.
doi:10.1109/ICIAS.2007.4658511
100
Volume 16, Number 2, 2016
[12] R. Wamkeue, I. Kamwa, Fellow, “Hybrid-State-Model-Based Time-
Domain Identification of Synchronous Machine Parameters from
Saturated Load Rejection Test Records,” IEEE Transaction on Energy
Conversion, vol. 23, no. 1, pp. 68-77, 2008.
doi:10.1109/TEC.2007.914663
[13] T.L. Vandoorn, F.M. De Belie, T.J. Vyncke, J.A. Melkebeek,
“Generation of Multisinusoidal Test Signals for the Identification of
Synchronous-Machine Parameters by Using a Voltage-Source
Inverter,” IEEE Transactions on Industrial Electronics, vol. 57, no. 1,
pp.430-439, 2009. doi:10.1109/TIE.2009.2031135
[14] J. Huang, K. A. Corzine, M. Belkhayat, “Online Synchronous
Machine Parameter Extraction from Small-Signal Injection
Techniques,” IEEE Transactions on Energy Conversion, vol. 24, no.
1, pp. 43-51, 2009. doi:10.1109/TEC.2008.2008953
[15] G. Valverde, E. Kyriakides, G.T. Heydt, V. Terzija, “Nonlinear
Estimation of Synchronous Machine Parameters Using Operating
Data,” IEEE Transactions on Energy Conversion, vol. 26, no. 3, pp.
831-839, 2011. doi:10.1109/TEC.2011.2141136
[16] M. A. Arjona, M. Cisneros-González, C. Hernández, “Parameter
Estimation of a Synchronous Generator Using a Sine Cardinal
Perturbation and Mixed Stochastic–Deterministic Algorithms,” IEEE
Transactions on Industrial Electronics, vol. 58, no. 2, pp. 486-493,
2011. doi:10.1109/TIE.2010.2047833
[17] C. Ramirez, M. Tu Xuan, J.-J. Simond, “Synchronous machines
parameters determination using finite elements method,” in Proc.
International Conference on Electrical Machines (ICEM), Espoo,
2000, pp. 1130-1133.
[18] R. Escarela-Perez, E. Campero-Littlewood, T. Niewierowicz,
“Efficient finite-element computation of synchronous machine
transfer functions,” IEEE Transactions on Magnetics, vol. 38, no. 2,
pp. 1245-1248, 2002. doi:10.1109/20.996318
[19] T. Tudorache, M. Popescu, “Optimal Design Solutions for Permanent
Magnet Synchronous Machines,” Advances in Electrical and
Computer Engineering, vol.11, no.4, pp.77-82, 2011.
doi:10.4316/AECE.2011.04012
[20] T. Tudorache, M. Modreanu, “Design Solutions for Reducing the
Cogging Torque of PMSM,” Advances in Electrical and Computer
Engineering, vol.13, no.3, pp.59-64, 2013.
doi:10.4316/AECE.2013.03010
[21] L. Livadaru, A. Munteanu, A. Simion, B. Vîrlan, S. Benelghali,
“Study on the Fault-Tolerance Concept of the Five-Phase Permanent
Magnet Synchronous Generator,” Advances in Electrical and
Computer Engineering, vol.14, no.2, pp.77-84, 2014.
doi:10.4316/AECE.2014.02013
[22] I.D. Ilina, C. Ghiţă, “Determination of Magnetizing Characteristics of
Synchronous Machine by Static Methods,” in Proc. 6th International
Symposium On Advanced Topics in Electrical Engineering (ATEE),
Bucharest, 2008, pp.158-161.
[23] P.J. Turner, A.B.J. Reece, D.C. MacDonald, “The DC decay test for
determining synchronous machine parameters: measurement and
simulation,” IEEE Transactions on Energy Conversion, vol. 4, no. 4,
pp. 616-623, 1989. doi:10.1109/60.41720
[24] V.Z. Groza, “Experimental Determination of Synchronous Machine
Reactances from DC Decay at Standstill,” IEEE Transactions on
Instrumentation and Measurement, vol. 52, no. 1, pp. 158-164, 2003.
doi:10.1109/TIM.2003.809496
[25] F.S. Sellschopp, M.A. Arjona, “DC decay test for estimating d-axis
synchronous machine parameters: a two-transfer-function approach,”
IEEE Proceedings - Electric Power Applications, vol. 153, no. 1, pp.
123-128, 2006. doi:10.1049/ip-epa:20050248
[26] L. Vicol, I.A. Viorel, V. Iancu, “Identification of the Synchronous
Generator Parameters by Standstill Tests,” in Proc. 6-th International
Conference on Electromechanical and Power Systems, Chişinău,
2007, pp. 138-141.
[27] E.S. Boje, J.C. Balda, R.G. Harley, R.C. Beck, “Time-domain
identification of synchronous machine parameters from simple
standstill tests,” IEEE Transactions on Energy Conversion, vol. 5, no.
1, pp. 164-175, 1990. doi:10.1109/60.50828
[28] M. Hansi, O. Touhami, R. Ibtiouen, M. Fadel, S. Caux, “Estimation of
synchronous machine parameters by standstill test,” Mathematics and
Computers in Simulation, vol. 81, no. 2, pp. 277-289, 2010.