Download MODELS OF EXCITATION SYSTEMS

MODELS OF EXCITATION SYSTEMS
Milan Ivanović*, Sanja Ivković, Dragan Dabić, Nikola Georgijević, Dragan Popović
University of Belgrade, Electrical Engineering Institute Nikola Tesla,
Koste Glavinića 8a, Belgrade*
Abstract: In this paper, an overview of the preliminary results, obtained within Digsilent
PowerFactory is given. Digsilent PowerFactory have been used as reference program for validating
models used for the study "System parameters of excitation and turbine control for power plants in
the EPS". Objective of the study is to capture, analyze and adjust the relevant parameters of the
excitation and turbine control systems for all plants within EPS. The final objective is to improve
the quality of regulation, to provide reserves for primary, secondary and tertiary frequency control
and to provide resources for generation and absorption of reactive power for the primary voltage
control. For the study, model of southeast Europe power system has been created in the DINST
software, developed in the Electrical Institute Nikola Tesla. As a referent program, Digsilent
PowerFactory has been used. In the paper, comparative results of measurements and simulations of
the response voltage regulator on a stepping change of the reference voltage.
Key words: statism, generator voltage, excitation control, voltage - reactive states, transmission
network.
MODELI SISTEMA ZA REGULACIJU POBUDE
Milan Ivanović*, Sanja Ivković, Dragan Dabić, Nikola Georgijević, Dragan Popović
Univerzitet u Beogradu, Elektrotehnički institut Nikola Tesla, Koste Glavinića 8a, Beograd*
Apstract: U radu je dat pregled dela preliminarnih rezultata dobijenih simulacijama u programskom paketu Digsilent PowerFactory, koji je korišćen kao referentni program za proveru modela
korišćenih u studiji "Sistemski parametri regulacije pobude i turbinske regulacije u elektranama
EPS". Njen osnovni cilj je snimanje, analiza i podešavanje relevantnih parametara sistema
regulacije pobude i sistema turbinske regulacije u svim elektranama EPS. Krajnji cilj je poboljšanje
kvaliteta regulacije, obezbeđenje rezerve za primarnu, sekundarnu i tercijernu regulaciju učestanosti
i obezbeđenje kapaciteta za proizvodnju i apsorpciju reaktivne snage na generatorskim jedinicama
za primarnu regulaciju napona. Za potrebe studije, formiran je model EES jugoistočne Evrope u
programskom paketu DINST, razvijenom u elektrotehničkom institutu Nikola Tesla. Kao uporedni
program, korišćen je Digsilent PowerFactory. U radu su prikazani uporedni rezultati merenja i
simulacije odziva regulatora pobude na odskočnu promenu referentnog napona.
Ključne reči: statizam, napon generatora, regulacija pobude, naponsko-reaktivna stanja, prenosna
mreža.
1. INTRODUCTION
Basic object of the study "System parameters of excitation and turbine control for power plants in
the EPS" is measurement, analysis and setting of relevant parameters of primary, secondary and
tertiary frequency control as well as excitation control for all power plants within EPS (Public
Enterprise Electric Power Industry of Serbia), in order to provide reserves for primary, secondary
and tertiary frequency control, to provide resources for generation and absorption of reactive power
for the primary voltage control and to improve their quality. The intention was to perform
simulations on the real model of Serbian Power System, in its wide environment. In that way, socalled system aspects, which are only relevant for the study objective, would be considered. The
final objective of the study is to improve the quality of primary, secondary and tertiary frequency
control and primary voltage control, to provide resources for generation and absorption of reactive
power for the primary voltage control, and their harmonization with the existing regulations.
For step response simulations, software Digsilent PowerFactory 14.1 (Educational license),
have been used as a referent program. In this paper, some of preliminary results of simulations are
presented, as well as automatic voltage regulator structure, comparison of step response simulation
and measured response.
2. STATIC EXCITER SYSTEMS
2.1. Controller structure
One of tested exciter systems is static exciter system, which is developed at the Electrical
Engineering Institute Nikola Tesla. Exciter system is equipped with both an automatic voltage
regulator - AVR (exciter voltage control) and a manual regulator (exciter current control). It has two
identical, independent excitation channels. Only one of these channels operates at the time and
covers all operating regimes of the generator, including field forcing; the other is a backup channel.
Each channel contains digital controller of the exciter and the thyristor bridge.
Equivalent models of the generator voltage controller and the exciter current controller are
given in Figure 1 and Figure 2.
Generator voltage controller
Thyristor bridge
Figure 1: Equivalent model of the generator voltage controller in s domain
Exciter current controller
Thyristor bridge
Figure 2: Equivalent model of the exciter current controller in s domain
Input of the generator voltage controller, shown in Figure 1, is generator reference voltage
(Vs ref). Stator voltage (Vs) is brought back to form the negative feedback. The output of the static
exciter is excitation voltage (Vf).
First block of the generator voltage controller is lead - lag compensator, whose transfer
function in s domain is (s + aVs)/ (s + bVs). In practice, the values of parameters aVs and bVs are
equalized, which reduces compensator transfer function to unity gain. Therefore, this block is
omitted from the model of the static exciter systems. The second block of the generator voltage
controller is PI controller, whose parameters are proportional gain (KVs) and the reciprocal of the
integrator time constant (cVs) [1] - [3].
Output of the generator voltage controller represents thyristor firing angle (αm), which
regulates DC output voltage of the thyristor bridge. Fully controlled thyristor bridge is non-linear
element, so its transfer function is linearized around its nominal operating point: KTM / (sTTM + 1),
where KTM is gain of the thyristor bridge and TTM is thyristor delay. Output voltage limit depends on
supply voltage of the thyristor bridge. Effective limits of thyristor bridge output voltage (VRmax and
VRmin) are adjustable.
The (linearized) transfer function of described static exciter system, in automatic regime
(excitation voltage control) can be represented as:
Gns ( s ) = K ns
s + cns
1
⋅
s
TTM s + 1
(1)
Block for reactive power compensation is not shown in Figure 1. Its adjustable parameter is
voltage droop (σ). Output of the block for reactive power compensation affects generator reference
voltage: increase of reactive power from zero to its nominal value should increase reference voltage
(in p.u.) for the value equal to voltage droop (σ). Block for reactive power compensation does not
affect reference voltage for under-excited regimes [5], [6].
3. MODELLING OF EXCITATION CONTROLLER
3.1. Digsilent PowerFactory environment
Composite frame is basic element for modeling generator controllers in Digsilent PowerFactory.
Each composite frame contains slots and signal interconnections, as shown in Figure 3. Within
internal library, several composite frames are available, and they differ from each other on the
number of slots and available signals. Each slot has predefined purpose - for connecting generator
(SYM), excitation controller (AVR), turbine controller (GOV), power system stabilizer (PSS), etc.
All of these elements can be found in the library or they can be created by the user [4].
Figure 3: Composite frame SYM Frame no_droop 2
3.2. Model of static exciter system
None of predefined AVR models from the PowerFactory library match structure of previously
described static excitation system. Therefore, new model, shown in Figure 4, has been created. In
order to simplify modeling, original transfer function (1) has been modified:
Gns ( s) =
Vf
eVs
1
⎛ c ⎞
= K ns ⎜1 + ns ⎟ ⋅
s ⎠ TTM s + 1
⎝
(2)
Figure 4: Model of static exciter system: avr_StSam_INT
Model shown in Figure 4 uses following signals:
- usetp -
generator reference voltage [p.u.]
- Qg
-
generator reactive power [Mvar]
- upss
-
output of the power system stabilizer [p.u.]
- sgnn
-
nominal apparent power of the generator [MVA]
- cosn
-
nominal power factor of the generator
-u
-
generator voltage (negative feedback) [p.u.]
All necessary signals were available only in composite frame SYM Frame no_droop 2 (Figure
3). Signals Qg, sgnn i cosn have been used for calculation of reactive power generation in p.u.
system, where nominal reactive power of the generator has been used as a base power:
q=
Qg
Qgn
=
Qg
(3)
Sgnn ⋅ sin(arccos(cos n))
For the modeling of voltage and reactive power measurement filters, first order elements have
been used, with time constants Tmu and Tmq, respectively. Model parameters are shown in Table 1.
Table 1: Model parameters
Parametar
Delay of the voltage filter
Voltage droop
Delay of the reactive power filter
Reciprocal of the integrator time constant
Proportional gain
Tiristor bridge gain
Tiristor bridge delay
Minimum output limit
Maximum output limit
Symbol
Tmu
Stat
Tmq
Cns
Kns
Ktm
Ttm
Vrmin
Vrmax
Unit
[s]
[%]
[s]
[r/s]
[p.u.]
[p.u.]
[s]
[p.u.]
[p.u.]
Value
0.01
1
0.01
0.6
23.3
1
0.0011
-1.88
6.28
3.3. Gain calculation
In reports on regulator testing, proportional gain (K3) is given in absolute values, as a quotient of
excitation voltage change (∆Vf) and corresponding generator voltage change (∆Vg):
K3 =
∆U f
∆U g
=
472.7 V − 122.5 V
15900 V − 14270 V
⎡V ⎤
= 0.21 ⎢ ⎥
⎣V ⎦
(4)
Equivalent gain is calculated by using (4), according to responses of excitation voltage and
generator voltage for step change of generator voltage reference in automatic regime, as in Figure 5.
Figure 5: Step responses of excitation voltage and generator voltage
Calculated equivalent gain (K3) has to be converted to p.u. system. According to [7], which is
completely based on [8], for the base excitation voltage Uf0 should be used:
K ns =
∆U f / U f 0
∆U g / U gn
=
∆U f
⋅
U gn
∆U g U f 0
= K3 ⋅
U gn
Uf0
[r. j.]
(5)
In order to make possible comparison of measurements and simulation results, following base
values have been used:
- Ugn - nominal generator voltage
- Ufn - nominal field voltage
4. SIMULATION RESULTS
4.1. Step response
Comparison of measurements and simulation results, for the step change of reference voltage from
0.954 p.u. to 1.0031 p.u, is given in Figure 6.
1.1
1.0
0.9
0.8
Ug, meas [p.u.]
0.7
Uf, meas [p.u.]
0.6
Ug, sim [p.u.]
0.5
Uf, sim [p.u.]
0.4
0.3
0.2
0.1
0.0
19
20
21
22
t[s]
23
24
25
26
27
Figure 6: Comparison of measurements and simulation results
4.2. Voltage droop
For checking voltage droop of static exciter model, special test grid has been created. Initial voltage
value is set to 1 p.u. and referent voltage step value to 0. Initially, generator operates in no-load
regime. At zero point, generator reactive load becomes equal to generator nominal reactive power.
Voltage droop is set to 1%. Simulation results are given in Figures 7 and 8.
Figure 7: Simulated response of an AVR for step reactive loading of the generator
Figure 8: Detailed generator voltage response
As it can bee seen from Figure 8, new steady-state value of generator voltage is 1.01 p.u. The
result is the same as it would be if step change of voltage reference for 1% (σ = 1%), for generator
in no-load regime.
5. CONCLUSIONS
Achieved simulation results are almost identical to results of measurement which confirm structure
and parameters of created AVR model as well as generator parameters. It is very important that
incorporated block for reactive power compensation has also been verified. Adjustable parameter of
this block - voltage droop, is very important parameter in steady-state and dynamic analysis of
system security. The other important achievement is using measurements for parameter calculation
and for verification of created models. This approach will improve quality of analysis, their results,
and, which is the most important, performance of the power system.
This paper is based on the research results of the project "Increasing energy efficiency,
reliability and availability of the EPS power plants by determining generator capability and
application of the new methods of testing and remote monitoring", which is financed by the
Ministry of Education, Science and Technological Development.
REFERENCE
[1] "Matematički modeli objekata elektroenergetskog sistema", studija Instituta "Nikola Tesla",
Beograd, 1981.
[2] Joksimović D., Ćirić Z., Stojić Đ., Milojčić N., Arnautović D., Petrović D., "Parametri
pobudnih sistema", časopis "Elektroprivreda", br.3. 2011. str.198-206.
[3] Arnautović D., Ćirić Z., Stojić Đ., Milojčić N., Joksimović D., Milinković M., Veinović S.,
Bakić M., Palija V.,"Modernizacija, rekonstrukcija i razvoj sistema pobude sinhronih
generatora", Zbornik radova Instituta "Nikola Tesla", knjiga 21, 2011., str. 181-195.
[4] "User’s Manual, DIgSILENT Power Factory Version 14", DIgSILENT, GmbH, Gomaringen,
Germany, 2008.
[5] Popović D.P, Stojković M.Lj., "Naponsko-reaktivna stanja prenosnih mreža", monografija,
Institut "Nikola Tesla", Beograd, ISBN 978-86-83349-09-8, jun 2009. godine, str.295.
[6] Machowski J., Bialek W., Bubmy J., Power System Dynamics: Stability and Control, John
Wiley and Sons Inc., New York, 2008.
[7] "DIgSILENT Technical Documentation - Synchronous Generator", 2010.
[8] Kundur P., "Power System Stability and Control", McGraw-Hill, 1994.