Volume 53, Number 1, 2012 3

Field Test Process after Commissioning a

New Exciter System
Lajos KISS, Dániel DIVÉNYI, Dávid RAISZ and József ZERÉNYI
Abstract: Realization of a new exciter system generally consists of the following main steps:
1. The manufacturer issues a questionnaire on generator, exciter and network data. In case of installation of a new unit only
the Point of Common Coupling (PCC) parameters and data are discussed.
2. The manufacturer determines the optimal settings of time constants and gain factors of the exciter-controller-PSS system
by modeling its steady state and transient operation.
3. The commissioning procedure is executed by the manufacturer in frame of field tests taking into account the technical
specifications of the Grid Code, repeating the tests of main cases modeled in paragraph 1. (Except e.g. of short circuits case).
A new exciter system in general replaces an old one or is in cooperation with old existing systems. From the point of view of TSO
the new system is expected to damp the amplitudes of oscillations of active- and reactive power, voltage and frequency, both in
steady-state and during transients on all frequencies. In the last ten years we elaborated and successfully employed a method to
evaluate the performance of exciter systems – including PSS – without disturbing the steady-state operation of the units. The main
results of the last field test are documented in this paper. The field tests executed by the manufacturer (paragraph 2) cannot be
replaced by our tests, however our results give important additional information for the power plants and TSO.
Keywords: Inter-area oscillations, Large electric systems cooperation, Damping of active power- voltage- and frequency
oscillations, Constructing and tuning exciter system and PSS, Time-frequency analysis of state variables, Power system dynamics,
DFT analysis

1. INTRODUCTION 12 Fourier spectra gives a stable resultant one showing

the main characteristics of the measured variable (see
In case of performance evaluation of an exciter e.g. Fig. 5. a. and Fig. 6. a).
system the following four frequencies are the center of
interest: 2. THE CIRCUMSTANCES OF FIELD TEST
o Global-mode oscillations (0.015-0.04 Hz) formed
by the whole interconnected system. In this field test we compared an existing exciter
o Inter-area mode oscillations (0.13-0.35 Hz) which system performance with the new one, both cooperating
are observed between major groups of generation with the same type of generator. The tests were
plants. executed simultaneously on the two exciter systems.
o Electro-mechanical mode oscillations (1-2 Hz): The measured time functions were the following:
between a unit and the rest of the whole power ∆f: the frequency deviation time function from
system and the other units of the generating station. 50 Hz of the terminal voltage of unit, mHz,
o Torsional-mode ≈ 15 Hz. This is the first torsional ∆Ur, ∆Us, ∆Ut: the terminal voltage deviation of the
self-frequency of the turbine-generator shaft. unit from the nominal voltage, V,
∆P: the active power deviation of the unit, MW,
The method developed is explained in detail in [1].
∆Q: the reactive power deviation of the unit, MVAr,
There is no explicit disturbance signal; in fact the
perpetual low variations of state variables coming from The time functions above were registered in the
the network and cooperating units are considered as following five operating cases:
disturbance signals. Therefore the two units 1. MANUAL operation of the exciter system,
investigated are disturbed by the same signal 2. AVR (Automatic Voltage Regulation) operating
consequently the results are comparable. The whole test case,
of an operating case lasts 1200 sec divided into 12 3. {AVR+PSS} (AVR + Power System Stabilizer)
intervals of 100 s each, and a Fourier spectrum is active,
determined from every 100 sec interval. The average of 4. {AVR+U/Q} (AVR+ Central Voltage/Reactive
Power Controller) active,
 5. {AVR+PSS+U/Q} active.
Manuscript received October 3, 2011. This work is connected to the
scientific program of the "Development of quality-oriented and
harmonized R+D+I strategy and functional model at BME" project, During a simulation procedure we can assume,
supported by the New Hungary Development Plan (Project ID: that in MANUAL operating case the field voltage is
TÁMOP-4.2.1/B-09/1/KMR-2010-0002)

© 2012 – Mediamira Science Publisher. All rights reserved.

4 ACTA ELECTROTEHNICA

constant. In this situation the amplitude of oscillations the manufacturer and the power plant the most
of state variables will be at minimum level. But in important run is {AVR+PSS+U/Q} since units are
practice such conditions are not realized. In MANUAL generally in this operating mode. In the remaining cases
operation of the exciter system generally the field we analyzed the effect of parameters on the studied
current is kept constant by an appropriate simplified phenomena step by step to improve our MATLAB
exciter structure. SIMULINK model (Section IV). To facilitate the
The rated apparent power of the units was comparison of the field test results, the figures of time
259 MVA, and the rated voltage 15.75 kV. The type of functions are located side by side on the same pages.
the new static exciter system is IEEE Type ST5B The quick estimation of the frequency of state variables
cooperating with an IEEE Type 2B PSS. The type of is fostered by changing the timescale in appropriate
the old static exciter system is SG-413. The old exciter figures to a 15 second interval.
system does not comply with the IEEE standard while Our time domain investigations (Fig. 1-4.) are
its PSS has the conventional structure consisting of a completed with frequency domain investigations
high pass filter, a low pass filter and two lead-lags. (Fig. 5-10.). With Fourier spectra we can point out
For a general evaluation of oscillation of a state numerically the effect of the exciter system, PSS and
variable we adopt an index number, SF{Δx}, as defined U/Q controller on the amplitude of oscillations of state
below, where x is the state variable in question. variables. For an expert it is very useful to inspect the
N
Fourier spectra because these clearly show 1) the most
SF{x}   xi 1
2 (1) important frequencies of disturbance signals originating
from the network in the place of the power plant and
from the cooperating units and 2) the effect of the
N: total number of spectrum lines; in our case =
exciter system on the state variables.
200.
Δx: magnitude of one spectrum line. 3.1. General review of field test
During the field tests only the power controlling The parts a) of Fig. 1-4. prove that tests were
circuit of the turbine governor was active and the speed performed on two units simultaneously indeed in the
controller circuit was out of operation. same operating cases.
Furthermore by comparing the Fourier spectra of
3. ASSESSMENT OF THE FIELD TEST frequency one can observe that the damping effect of
RESULTS exciter systems with and without PSS on the frequency
of the terminal voltage of the units is practically the
Because of the limitations on the extent of the
same.
paper only two of five operating cases are presented
here; namely {AVR+U/Q} and {AVR+PSS+U/Q}. For
Δf, mHz

Δf, mHz

40 40

0 0
0 20 40 60 80 100 120 140 160 180 200 0 20 40 60 80 100 120 140 160 180 200
-40 t, sec -40
t, sec

a) frequency deviation
ΔUr, V

ΔUr, V

40 40

0 0
0 20 40 60 80 100 120 140 160 180 200 0 20 40 60 80 100 120 140 160 180 200
-40 t, sec -40
t, sec
b) terminal voltage deviation
2 2
ΔP,MW

ΔP,MW

0 0
0 20 40 60 80 100 120 50 55 60 65 0 20 40 60 80 100 120 50 55 60 65
-2 -2
t, sec t, sec
c) active power deviation
ΔQ, MVAr

ΔQ, MVAr

6 6

0 0
0 20 40 60 80 100 120 140 160 180 200 0 20 40 60 80 100 120 140 160 180 200
-6 -6
t, sec t, sec
d) reactive power deviation
Fig. 1. Time functions of field test results of a 259 MVA, 15,75 kV Fig. 2. Time functions of field test results of a 259 MVA, 15,75 kV
TVV-221 type unit in steady state operation (IEEE ST5B in TVV-221 type unit in steady state operation (SG-413 in {AVR
{AVR+U/Q}). +U/Q}).
 Volume 53, Number 1, 2012 5

Δf, mHz

Δf, mHz
40 40

0 0
200 220 240 260 280 300 320 340 360 380 400 200 250 300 350 400
-40 -40
t, sec t, sec
a) frequency deviation
ΔUr, V

ΔUr, V
40 40

0 0
200 220 240 260 280 300 320 340 360 380 400 200 220 240 260 280 300 320 200 205 210 215
-40 -40
t, sec t, sec
b) terminal voltage deviation
ΔP,MW

ΔP,MW
2 2

0 0
200 220 240 260 280 300 320 200 205 210 215 200 220 240 260 280 300 320 200 205 210 215
-2 -2
t, sec t, sec
c) active power deviation
ΔQ, MVAr

ΔQ,MVAr
6 6

0 0
200 220 240 260 280 300 320 340 360 380 400 200 220 240 260 280 300 320 200 205 210 215
-6 -6
t, sec t, sec
d) reactive power deviation
Fig. 3. Time functions of field test results of a 259 MVA, 15,75 kV Fig. 4. Time functions of field test results of a 259 MVA, 15,75 kV
TVV-221 type unit in steady state operation (IEEE ST5B + IEEE Type TVV-221 type unit in steady state operation (SG-413 + PSS in
2B PSS in {AVR+PSS+U/Q}). {AVR+PSS+U/Q}).

SF{Δf} = 1,198 mHz SF{Δf} = 1,098 mHz

F{Δf}, mHz

F{Δf}, mHz

0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
f, Hz f, Hz
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
a) frequency deviation
SF{avg(ΔUr, ΔUs, ΔUt)} = 1,426 V SF{avg(ΔUr, ΔUs, ΔUt)} = 1,687 V
avg(F{ΔU}), V

avg(F{ΔU}), V

0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
f, Hz f, Hz
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
b) average terminal voltage deviation
SF{ΔP} = 0,282 MW SF{ΔP} = 0,388 MW
F{ΔP}, MW

F{ΔP}, MW

0.06 0.06

0.04 0.04

0.02 0.02

f, Hz f, Hz
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
c) active power deviation
Fig. 5. Fourier transformed functions evaluated from the time functions Fig. 6. Fourier transformed functions evaluated from the time functions
of field test shown in Fig.1 (IEEE ST5B in operation {AVR+U/Q}). of field test shown in Fig.2 (SG-413 in operation {AVR+U/Q).
6 ACTA ELECTROTEHNICA

SF{Δf} = 1,165 mHz SF{Δf} = 1,211 mHz

F{Δf}, mHz

F{Δf}, mHz
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
f, Hz f, Hz
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
a) frequency deviation
SF{avg(ΔUr, ΔUs, ΔUt)} = 1,812 V SF{avg(ΔUr, ΔUs, ΔUt)} = 4,850 V
avg(F{ΔU}), V

avg(F{ΔU}), V
1.4 1.4
1.2 1.2
1 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
f, Hz f, Hz
0 0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
b) average terminal voltage deviation
SF{ΔP} = 0,234 MW SF{ΔP} = 0,268 MW
F{ΔP}, MW

0.06 F{ΔP}, MW
0.06

0.04 0.04

0.02 0.02

f, Hz f, Hz
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
c) active power deviation
Fig. 7. Fourier transformed functions evaluated from the time functions Fig. 8. Fourier transformed functions evaluated from the time functions
of field test shown in Fig.1 (IEEE ST5B in operation of field test shown in Fig.2 (SG-413 in operation {AVR+PSS+U/Q).
{AVR+PSS+U/Q}).

3.2. Investigations of different modes In {AVR+PSS+U/Q} and {AVR+U/Q} operating

cases these low frequency oscillations of state variables
Evaluating exciter systems in general the IEEE
can be observed in the case of IEEE ST5B exciter
ST5B provide much higher performance, its SF{ΔP}
system (Fig. 3.b) and d)). These are the effect of actions
and SF{ΔU} values are significantly lower (Table 1,
of its U/Q controller which were not set optimally.
Fig. 5-6. c)). Of course we expect from the newly
Since the field test the necessary steps to eliminate
installed exciter system to be better than the replaced
these oscillations have been determined and will be
one which is proven by this short comparison. However
done. As the U/Q controller of SG-413 exciter system
in this section the two PSS are analyzed from the point
did not generate such oscillations (Fig. 4.b) and d)) the
of view of damping different modes of frequency
damping effect on this frequency mode cannot be
oscillations.
compared.
Table 1: General evaluation of exciter system without PSS. To damp the oscillations in global mode frequency
general IEEE ST5B SG-413 range with the exciter system is far from center of
0,10-2,00 Hz interest recently although the amplitude of these
SF{Δf} 1,198 1,098
oscillations is significantly greater than that of the inter-
SF{ΔU} 1,426 1,687
SF{ΔP} 0,282 0,388 area modes. Nevertheless, based upon our model
investigation, with an appropriately constructed exciter
3.2.1. Global-mode oscillations system and tuned PSS it is possible to damp these low
Global mode of oscillations can be studied only by frequency oscillations.
analyzing the figures of time functions. (The time
3.2.2. Inter-area mode of oscillations
period of such oscillations is about 25-66 sec.) Red
color is used in figures for indicating the global mode. Effects of inter-area mode of oscillations can be
The global-mode of oscillations of state variables recognized both in figures of time functions and in
is determined by the difference of the generated and figures of Fourier spectra. (The time period is about
consumed active power of the whole system. Since the 2.8-7.7 sec.) Green color is used in figures for
active power of loads may depend on their terminal indicating the inter-area mode of oscillations.
voltage, the amplitude of global mode of oscillations – The amplitude of frequency deviation of inter-area
theoretically – may be influenced by the exciter system. mode is not influenced by these investigated exciter
systems significantly. However the SG-413 exciter
 Volume 53, Number 1, 2012 7

system in {AVR+PSS+U/Q} operating case extremely 3.3. Final evaluation of the two exciter systems
increases the amplitude of terminal voltage deviations
Using the IEEE ST5B exciter and PSS (having p
and also increases to a less extent the active power
and ∆ω input channels) the active power oscillations in
deviations (Table 2, comparing Fig. 8.b-c) with Fig.
all important frequency ranges are damped
7.b-c), and Fig. 12. c) with d)).
appropriately. Although its PSS shows weaker
Table 2: Effect of PSS on inter-area mode performance than PSS of SG-413, regarding
inter-area IEEE ST5B SG-413 oscillations near the electro-mechanical mode. Despite
0,13- of better performance of PSS of SG-413 (having only p
PSS OFF PSS ON PSS OFF PSS ON
0,35 Hz
0,8857 0,9017
input channel) near the electromechanical self-
SF{Δf} 0,9149 0,8562 frequency, in consideration of performance in inter-area
(–3%) (+5%)
1,0645 4,0470 mode the new realizations of this type of exciter system
SF{ΔU} 0,8041 0,9477
(+32%) (+327%) are denied by our TSO.
0,0876 0,1834 With an executed and successful commissioning
SF{ΔP} 0,0955 0,1560
(–8%) (+17%)
and control field test the PSS tuning procedure is not
finished because the frequency of inter-area modes are
These terminal voltage- and active power
not constant as result of changing the structure of the
deviations, generated by SG-413 exciter system, are
whole cooperating network. After connected Turkey’s
exported to the network. These oscillations reacting to
power grid to the European Union Energy System
themselves and to other exciter systems produce
(2010-09-18) a new cc. 0.13 Hz inter-area mode of
positive feedbacks. This performance of exciter system
oscillations has appeared in frequency time function
is not acceptable by the Hungarian TSO in case of
(Fig. 9 and Fig. 10).
future realizations. The demonstration of these
phenomena is an important result of our field test 18
method. mHz
16
3.2.3. Electromechanical mode
14
Effects of electromechanical mode of oscillations
can be studied both in figures of time functions and in 12
figures of Fourier spectra. (The time period is about
10
0.5-1 sec.) Orange color is used for indicating these
oscillations. 8
Analyzing the Fourier transformed active power
6
functions of {AVR+PSS+U/Q} operating case the SG- 1140 1150 1160 1170 1180 1190 sec 1200
413 exciter system provides better performance than the (file fav.ADF; x-var t[s]) fav[Hz]

IEEE ST5B exciter system (Fig. 7.c) and Fig. 8.c)). The Fig. 9 Time function (∆fav) recorded by WAMS devices in Hungarian
amplitudes of power swings near the electromechanical system Trigger time: 2011-01-24-08:19:00.
mode (1.2 Hz) are more than two times bigger using
IEEE ST5B system. Similar conclusion can be deduced
from the active power time functions (Fig. 3.c), Fig.
4.c) and Fig. 11). 1.2 SF(Δf) = 2,1581 mHz
F(Δf) [mHz]

Considering the theoretic interval of 1

electromechanical modes (1-2 Hz) the PSS of SG-413
exciter system decreases the amplitudes of power 0.8
swings by 50% while PSS of IEEE ST5B system by 0.6
only 26% (Table 3). 0.4
Table 3: Effect of PSS on electromechanical mode 0.2
el-mech. IEEE ST5B SG-413
1,00- 0
PSS OFF PSS ON PSS OFF PSS ON 0 0.2 0.4 0.6 0.8 1
2,00 Hz
f [Hz]
0,2396 0,3186
SF{Δf} 0,3544 0,3045
(–32%) (+5%) Fig. 10 Fourier transformed function evaluated from the time function
0,7319 0,7914 (∆f) recorded by WAMS devices in Hungarian system Trigger time:
SF{ΔU} 0,6890 0,7387
(+6%) (+7%) 2011-01-24-08:00:00.
0,1603 0,0975
SF{ΔP} 0,2173 0,1935
(–26%) (–50%)

3.2.4. Torsional mode 4. MODEL INVESTIGATIONS

The first torsional self-frequency of the turbine
generator shaft is 14.5 Hz in this case. The amplitude of The results of the field tests were used to verify
active power deviation time function is negligible our MATLAB model. Because of the limited extent of
according to field test in harmony with our simulation the paper only the most important results are presented.
results.
8 ACTA ELECTROTEHNICA

IEEE-ST5B SG-413
4 4
PSS PSS
3 AVR 3 AVR

2 2
Active power [MW]

Active power [MW]

1 1

0 0

-1 -1

-2 -2

-3 -3

-4 -4
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Time [sec] Time [sec]

Fig. 11. The time function of active power deviation time function, if the disturbing signal is the +0.01 p.u. step change of the commanded value
of terminal voltage in t = 1 sec; in case of operation (IEEE ST5B + IEEE Type 2B PSS) and (SG-413 + conventional PSS).

4.1. Inter-area mode of oscillation by Fourier spectra of Fig. 7.b) where we can similarly
see that the amplitude of terminal voltage unit
The effect of inter-area mode can be followed by
cooperating with SG-413 is four times bigger around
analyzing the bell-functions (Fig. 12). The frequency of
0.2 Hz than on the another unit. Thus the modeling
the network was disturbed by the following signal and
procedure can be used to forecast the performance of
the amplitudes of the active power and terminal voltage
exciter system intended to build in.
deviation were registered:
f N  5  sin(2    F ) [mHz] (1) 4.2. Electromechanical mode
F: frequency of disturbing signal (0.01 .. 0.05 mHz). The time functions shown in Fig. 11 suggest that
the (SG-413+conventional PSS) has better performance
Results, given in Fig. 12.c) and d) are reinforced
than (IEEE ST5B+IEEE type 2B PSS). This result is
IEEE-ST5B SG-413
1.6 1.6
AVR AVR
PSS PSS
1.4 1.4
MAN MAN

1.2 1.2
Active power [MW]

Active power [MW]

1 1

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Frequency [Hz] Frequency [Hz]
a) b)

-3 -3
x 10 IEEE-ST5B x 10 SG-413
2 2
AVR AVR
1.8 PSS 1.8 PSS
MAN MAN
1.6 1.6

1.4 1.4
Terminal voltage [p.u.]

Terminal voltage [p.u.]

1.2 1.2

1 1

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Frequency [Hz] Frequency [Hz]
c) d)
Fig. 12. Sections of the “bell-function” of the amplitude of active power- and terminal voltage deviation time function, if the disturbing signal is
the network frequency: ΔfN=5·sin(2·π·F) [mHz]; where 0.1 < F < 0.5 Hz.
 Volume 53, Number 1, 2012 9

important for the power plant but does not answer the REFERENCES
questions regarding the network – power plant
1. L. Kiss and J. Zerényi, "A Simple Measuring Method to
cooperation. Our method does, so its results are Determine the Damping Efficiency of the Excitation Systems
important for TSOs. and Power System Stabilizers (PSS)," in Electric Power
Engineering, 1999. PowerTech Budapest 99. Int. Conf., pp. 104.
5. CONCLUSIONS
ACKNOWLEDGEMENT
1. The method demonstrated in this paper is a very
efficient one to numerically assess the performance This work is connected to the scientific program
of an existing and a new exciter+PSS system. of the "Development of quality-oriented and
2. The simulation of field test results gives an harmonized R+D+I strategy and functional model at
opportunity to improve our model parameters to BME" project, supported by the New Hungary
simulate cases which were executed during Development Plan (Project ID: TÁMOP-4.2.1/B-
commissioning and which we cannot accomplish in 09/1/KMR-2010-0002)
practice, e.g. transient phenomena following a 3F
short circuit in the network. Assoc. Prof. dr. eng. Lajos KISS
3. Our measuring method does not replace the field Eng. Jozsef ZERÉNYI
tests executed by manufacturer during Assoc. Prof. dr. eng. David RAISZ
commissioning, but can give important additional As. eng. PhD student Daniel DIVÉNYI
results. Department of Electric Power Engineering
Budapest University of Technology and Economics
Egry Jozsef u. 18., H-1111 Budapest
Tel. +36 1 4633021
E-mail: [email protected]