E LEKTROTEHNIKI VESTNIK 82(4): 183-190, 2015

O RIGINAL SCIENTIFIC PAPER

A complex Hydro-Power Plant Dynamic Model Integrated into

the Electrical Network

Klemen Nagode1, Botjan Murovec2

1
Savske elektrarne Ljubljana d.o.o., Gorenjska cesta 46, 1215 Medvode, Slovenija
2
Univerza v Ljubljani, Fakulteta za elektrotehniko, Traka 25, 1000 Ljubljana, Slovenija
E-mail: [email protected]

Abstract. The paper presents a complex modelling of a hydro-power plant synchronized with an electrical
network. A dynamic performance model of a synchronous generator, Kaplan turbine, turbine controller,
excitation system, automatic voltage regulator, transformer and simplified electrical network is designed in
accordance with the IEEE recommended practices and standards. The developed model which is validated with
the in situ measurements taken on the Mavie hydro-power plant and nearby 110 kV Slovenian electrical
network shows a great matching of the characteristics. An important property of the model is its ability to
simulate the various types of the power-system operation and the impact of the hydro-power plant control
systems on the electrical network, especially the hazard scenarios that cannot be tested on a real power system.

Keywords: hydro-power plant, simulation, excitation system, turbine controller, power system

Kompleksen dinamien model hidroelektrarne, vkljuene voltage regulator (AVR) with a reactive-power
v elektrino omreje controller in accordance with the IEEE
V lanku je predstavljen kompleksen model hidroelektrarne, standard [1].
sinhronizirane z elektrinim omrejem. Model obravnava 3. A turbine controller with the main active-
dinamine lastnosti sinhronskega generatorja, Kaplanove power Proportional-Integral-Derivative (PID)
turbine, turbinskega regulatorja, vzbujalnega sistema, control loop, guide-vane opening and runner-
avtomatskega regulatorja napetosti, energetskega blade opening sub-control loops.
transformatorja in poenostavljenega elektrinega omreja 4. Characteristics of the Kaplan water-turbine
skladno s standardi IEEE in prakso. Zasnovani dinamini dynamics with servomotors.
model je validiran z meritvami na hidroelektrarni Mavie in
5. A 119/10.5 kV block transformer.
blinjim 110 kV elektrinim omrejem ter izkazuje dobro
ujemanje dinaminih lastnosti z realnim sistemom. Prednostna 6. 110 kV power lines presented with three-phase
lastnost predstavljenega modela je v monosti simulacije PI-modelled sections.
razlinih obratovalnih stanj, prehodnih pojavov, vpliva 7. Three-phase voltage sources (determined with
regulacijskih sistemov na elektrino omreje ter prouevanje the short-circuit power) connected to the power
hazardnih scenarijev, katerih na realnem energetskem sistemu lines.
ni mogoe izvesti zaradi obratovalnih omejitev. 8. Three-phase parallel RLC loads connected to
the power lines.
1 INTRODUCTION
In our case, the platform for the simulation is Matlab
Nowadays, the hydro-power plants (HPP) present an with Simulink in conjunction with the
important worldwide source of the renewable energy. In SimPowerSystems analysis tool.
order to foster further the power-system development, In the literature, different approaches are used to
an appropriate modelling of the essential HPP study the power-system dynamic models. In [2]-[4], the
components is necessary in order to gain an advanced authors conduct a comprehensive survey of the load-
knowledge of their dynamic performance. This paper frequency control strategies and HPP models. The latter
focuses on a model, developed to cover the following are classified as linear (non-elastic) models and
elements of an underlying power system: nonlinear with an elastic water-column effect. The
1. A synchronous generator with a salient-pole classical control methodologies include the PID
rotor. controllers with the Bode diagram, Nyquist diagram and
2. A static excitation system based on the IEEE root-locus analysis. These approaches are of a great
ST6B type [1] that includes an automatic practical value for the implementation, however, they

Received 15 May 2015

Accepted 30 June 2015
184 NAGODE, MUROVEC

show a poor dynamic performance in the case of the 2 HPP MODELLING

parameter deviations or nonlinearities. The modern
control techniques use more complex algorithms like A typical interconnection diagram of HPP with a
the adaptive and variable-structure methods as well as synchronous machine is depicted in Fig. 1. The vital
intelligent and robust approaches. These are more subsystems included in the simulation are presented
adequate in the case of uncertainties, disturbances, load with solid lines, whereas the neglected functional parts
variations and errors in modelling. are marked with dashed lines.
In [5]-[6] the authors propose a load-frequency
control of a multi-area source with an optimal output
feedback controller and neuro-fuzzy hybrid intelligent
PI control. The drawback of their contribution is that the
models of the hydro, steam and gas turbines used in the
simulation are too much simplified.
A fractional-order PID controller for the AVR system
using a chaotic multi-objective optimization is studied
in [7]. The authors rely only on the first-order transfer
functions for the exciter and generator and neglect the
turbine and electrical network integration.
A simulation model of the Shkopeti HPP, Albania, is
introduced by Prillwitz et al. [8]. A comparison between
the HPP model and the measured signals is presented,
Figure 1. Typical interconnection system of a synchronous
however, the impact on the electrical network is not machine [13].
included.
The International Electrotechnical Commission (IEC)
and the Institute of Electrical and Electronics Engineers 2.1 Synchronous generator
(IEEE) standards [1], [9]-[13] describe the
A three-phase salient-pole synchronous generator with
recommended approaches to design and develop various
the nominal power of 25 MVA, voltage of 10.5 kV and
turbine, excitation-system, generator and load models.
power factor of cos=0.8 is included in the model. A
These standards also define the guidelines for the
dynamic model with the stator variables within the
functional verification and fulfilment of the tender
reference frame fixed in the rotor (SimPowerSystem
guarantees. Notwithstanding their important role, it is
synchronous machine per unit (p.u.) standard block) is
often very difficult to extract adequate
presented with the Parks voltage differential equations
recommendations in the development.
[17] in the following expanded form (1-7):
The impact of a specific load (charging of electrical
dqs
r
vehicles) on the power network stability is addressed in r
uqs Rs iqs
r
r ds
r
(1)
[14] where the authors predict the daily voltage dt
fluctuations in 2030 for the case that 10 % of the dds
r
uds Rs ids r qs
r r r (2)
vehicles would rely on the electrical power. In [15]- dt
[16], the authors optimize the reactive-power d0s (3)
u0s Rs i0s
compensation and active-power scheduling but with no dt
consideration of the features of the power-generation d'fdr
ufd Rfd ifd
'r ' 'r (4)
process. dt
In this paper, a complete dynamic model of the d'kdr
(5)
u kd Rkd ikd
'r ' 'r
Mavie HPP (on the Sava river, Slovenia) together dt
with its subsystems is proposed in Sec. 2. Compared d'kq r
1
u kq1 Rkq1 ikq1
'r ' 'r (6)
with the previously described researches, this model dt
integrates the functional components of the entire HPP d'kqr
2
2 Rkq2 ikq2
production and transmission system based on the real
'r
u kq ' 'r (7)
dt
power-system measurements. Moreover, the where urqs and urds denote the stator voltages in the q and
synchronous generator, excitation system and water d direction. Voltage u0s is related to the three-phase
turbine with a controller connected to the 110 kV stationary voltages independent of reference-frame
network include the non-linear dynamics that reflect the speed that is equal to rotor speed r at transformation
characteristics of a real power system. In Sec. 3, the to the rotor reference frame. Equ. (4-7) describes the
simulation results and validation of the model based on rotor excitation-field voltage u'rfd and damper winding
real measurements are described. The impact of the voltages: u'rkd in the d axis and u'rkq1 and u'rkq2 in the q
HPP operation on the 110 kV electrical network is axis. The indices referring to voltage u, resistance R,
presented in Sec. 4. Conclusions are summarized in current i, and magnetic flux linkage , indicate the q - q
Sec. 5.
A COMPLEX HYDRO-POWER PLANT DYNAMIC MODEL INTEGRATED INTO THE ELECTRICAL NETWORK 185

axis, d - d axis, s - stator of the generator, f - rotor-field Parameter Vref denotes the voltage-reference setpoint
winding and k - rotor-damper winding. and Vc denotes the low-pass filtered feedback voltage
The model of the synchronous generator includes the from the generator terminals.
manufacturers [18] real values for the synchronous, The output of the PI controller enters the inner loop
transient and subtransient reactances as well as short- which includes a field-voltage regulator with a
circuit time constants (Td, Td and Tq) for the d and q proportional control (gain Km) and pre-control loop
directions. (gain Kff). The output of the controller is further
The stator resistance per unit is calculated as: calculated:
R (8)
Rspu s
Zb u min (( I lr ) (
K ci I fd K lr ; u PI
(12)
with phase stator resistance Rs and base impedance Zb Kg
expressed in (9): Efd
Tg s 1
) K m u PI K ff ) V b

U2 (9)
Zb b The first argument of the min function in Equ. (12)
Sb
denotes the field-current limiter with limitation
where Ub denotes the base phase voltage and phase parameter Ilr and gains Kci and Klr. A low-pass filter for
apparent power Sb. the field voltage with time constant Tg and gain Kg is
Inertia constant H [19] included in the turbine- used for the error-signal calculation.
generator rotor model is expressed in (10):
J n2 2.3 Water turbine with a digital turbine controller
H 5,48 10 9 (10)
Sn A dynamic model of a double-regulated (Kaplan) water
where J is the inertia of the rotor, n is the synchronous turbine with a turbine controller is implemented. The
speed (in revolutions per minute) and Sn is the generator controller inputs are: the reference of active power Pref,
nominal apparent power. active power Pe, speed of the generator w, speed
reference wref, net head dH, opening position of guide
2.2 Static excitation system vane Yv and opening position of the runner Yg. The
A dynamic model of the static excitation system with an output of the PID controller indicates setpoint opening
overexcitation limiter OverexcLim2, underexcitation for guide vane and runner that change water flow
limiter UnderexcLimIEEE2 and reactive-power through the turbine. The main PID controller error
controller VAR Type II is realized (Fig. 2) based on the signal is expressed as (13):
IEEE ST6B Type [1]. The error signal of the voltage
regulator (input in the PI controller) is calculated with ePID wref w R ( Pref Pe ) (13)
(11):
e max(Vuel;Vvar Vref Vc Voel ) Vs (11) where R denotes the steady state droop parameter.
Referring to the optimal operating efficiency without
where Vuel, Vvar, Voel and Vs denote the output of the cavitation, appropriate Yg/Yv ratio is considered that is a
underexcitation limiter, reactive-power controller, nonlinear function of the head:
overexcitation limiter and power-system stabilizer, Yg f (dH ,Yv ) (14)
respectively.
realized with a 2D lookup table.

Figure 2. Model of the static excitation system.

186 NAGODE, MUROVEC

U b2 2 (20)
Rfe
According to the real HPP measurement practice P0
[13], [20], a nonlinear relation between guide-vane 3
position Yv and active power Pe is modelled with a where Rfe presents the ohmic resistance of
nonlinear function (15): magnetization, Ub2 denotes the phase voltage on the
low-voltage side and P0 is the power in the open-circuit
Pe f (Yv ,dH ) (15) state.
Magnetization inductance Lmpu is expressed as (21):
Equ. (15) is the input to the transfer function of the U b2
water turbine (16): X I
Lmpu (21)
Pm ( s ) 1 s Aturb Tturb (16)
Z b2 Z b2

Pe ( s ) 1 s Bturb Tturb where X is the reactance of magnetization and I
denotes the inductive part of the magnetization current,
while the output of the water turbine is mechanical calculated as follows:
power Pm, that presents the input to the synchronous
generator. I I 0f 2 I fe 2 (22)
In accordance with the IEC recommendations for the with the common current in open-circuit I0f and resistive
hydro turbines [13], parameter Tturb is the water inertia part of current Ife at magnetization.
starting time, Aturb = -1 and Bturb = -0.5.
2.5 110 kV electrical network
2.4 Block transformer
The three-phase 110 kV transmission line sections with
A three-phase 119/10.5 kV block transformer in the lumped parameters in the vicinity of the Mavie HPP
Yd11 winding connection with no core saturation is are included in the dynamic model. Assuming that the
included in the model (Fig. 1). A function developed in three phases are balanced, R, L and C, the line
Matlab TR_calc_param calculates resistances R1pu, parameters are applied as the positive and zero sequence
R2pu, inductances L1pu, L2pu for both voltage windings, parameters, taking into account the inductive and
magnetization resistance Rmpu and magnetization capacitive couplings between the phases and the
inductance Lmpu. Besides the basic nominal transformer ground [22].
parameters (e.g. nominal power Sn, frequency fn, ) Table 1 presents the parameters of the 110 kV
obtained from the manufacturer [21], the calculated transmission line sections with length l, voltage sources
parameters present the input for the transformer block with three-phase short circuit power Sks and the RLC
(SimPowerSystems library). loads included in the electrical network. Besides the
The resistances p.u. for the high-voltage and low- Mavie HPP, four important connections to the
voltage windings are calculated with (17): switchgears are analysed in the model: the KL (Klee),
R R ME (Medvode), LAB (Labore) and OKR (Okroglo)
R1pu 1 , R2 pu 2 (17)
Z b1 Z b2 switchgears.
where R1, R2, Zb1 and Zb2 are the phase resistances of the Table 1: Parameters of the 110 kV electrical network [24]
high-voltage winding, phase resistance of the low- included in the model
voltage winding, base impedance of the high voltage- Functional part Sks [MVA] l (km) Pi [MW] Qi [MVAr]
winding and base impedance of the low-voltage
3f Us KL-ME 456 / -9.3 -2.7
winding, respectively.
Assuming that reactances X1pu and X2pu of both 3f PI line KL-ME / 8.310 / /
windings are equal and that the transverse reactance in 3f Us ME 29.5 / 9.05 2.02
the short circuit has a very small value, the leakage 3f RLC load ME / / 13.55 1.61
inductance p.u. is: 3f PI line ME-MA / 12.66 / /
L1pu X 1pu Z kpu2 ( R1pu R2pu ) 2
(18) HPP MA gen. 1 / / 3 2
HPP MA gen. 2 69.5 / 7 4.28
It is equal for both windings 1pu = 2pu . Parameter 3f PI line MA-LAB / 9.411 / /
Zkpu denotes the short-circuit impedance in p.u. 3f Us LAB-MA 655 / 0 0
Magnetization resistance Rmpu and inductance Lmpu 3f RLC load LAB / / 22.18 3.34
are derived from an open-circuit test. The calculation of
3f PI line LAB-OKR / 5.780 / /
resistance Rmpu is expressed with (19) and (20):
R 3f Us OKR-LAB 222 / 26.16 -1.36
Rmpu fe (19)
Z b2
The voltage sources and RLC loads are set to the
initial values of Pi and Qi, identical to the observed
measurements taken on the 21st January 2015. The
A COMPLEX HYDRO-POWER PLANT DYNAMIC MODEL INTEGRATED INTO THE ELECTRICAL NETWORK 187

results are recalculated with a load-flow tool in the results of model Pgm, Qm and fm are presented with a dashed
SimPowerSystems (Newton-Raphson iterative method line.
[23]) and show appropriate matching with the real
measurements. The same reference of the active power is gradually
changed on the SCADA touch screen of HPP and in
3 SIMULATION RESULTS AND VALIDATION OF simulation for a 0.12 step p.u. depicted in Fig. 3 (top
THE MODEL graph). Active power of the generator model Pgm
The proposed HPP dynamic model is implemented with (dashed line) and measured generator active power Pg
the real-system initial values of the Mavie HPP (solid line) follow the referential signal to minimize the
generator 1 with subsystems. The generator RMS (L1, L2 error from Equ. (13) presented in Fig. 3 (second graph).
and L3) currents are set to: Igm = 0.143 p.u., line RMS The reference for the reactive power Qref = 0.08 p.u. is
voltage Ugm = 1.009 p.u., active power Pgm = 0.12 p.u. kept constant during the simulation, and consequently
modelled reactive power Qm (Fig. 3, third graph, dashed
and reactive power Qm = 0.08 p.u.
line) and measured reactive power Q (solid line) remain
In the excitation system, the initial field voltage is set at 0.08 p.u. Modelled frequency of the generator fm,
to Ufm = 1.117 p.u., the voltage reference presented in Fig. 3 (bottom graph) with a dashed line as
Vref = 1.0105 p.u. and the reactive-power reference well as measured frequency f depicted with a solid line
Qref = 0.08 p.u. remain constant due to the stiffness of the 110 kV
The active-power reference of the turbine controller network.
presented in Equ. (13) is set to Pref = 0.12 p.u. (the same Fig. 4 shows matching of the results of modelled
as the feedback of generator active power Pgm), while generator current Igm, line voltage Ugm and 110 kV
the rotation-speed reference is wref = 1 p.u. (the transformer phase voltage UmL1 with the real system
generator is synchronized to the electrical network). The measurements Ig, Ug and UL1. The dynamic response of
guide-vane opening of the water turbine is the generator active power Pgm (Fig. 3) is proportional
Yv = 0.277 p.u., wheras the runner opening is positioned to generator current Igm (Fig. 4, top graph). Simulated
Yg
at Yg =0 p.u., with the reference to the ratio and head line voltage of the generator Ugm (Fig. 4, second graph)
Yv
dH from Equ. (14). remains constant due to unchanged generator reactive
The model simulation presents a stepwise increase power Qm. Simulated L1 phase voltage UmL1 on the
and decrease in the generator active power (Fig. 3) at a primary side of the 119/10.5 kV transformer is shown
constant reactive power and frequency in a 1000 s time on the bottom graph of Fig. 4.
frame (observed from 30 to 1030 s). The model of excitation field current Ifm and voltage
Ufm is shown in Fig. 5. Direct current Ifm that
magnetizes the rotor of the generator (Fig. 5, top graph)
is controlled with direct voltage Ufm (Fig. 5, bottom
graph) via a static thyristor excitation system. Compared
to the real excitation system (Uf and If, solid line), 0.04
p.u. lower field voltage Ufm (Fig. 5, second graph,
dashed line) is obtained at highest active power
Pg = 0.82 p.u. (at time t= 560 s).

Figure 4. Comparison between measured generator L1 RMS

Figure 3. Reference of active power Pref (measured and current Ig, line (L1-L2) RMS voltage Ug (solid lines) and the
simulated), measured generator active power Pg, reactive output of the generator model with Igm and Ugm (dashed lines).
power Q and frequency f depicted with a solid line, while the The 110 kV L1 RMS voltages of the transformer UL1
188 NAGODE, MUROVEC

(measurement solid line) and simulated UmL1 (dashed line) 17.25 m to 16.25 m) due to the lowered lake water level
are depicted in the bottom graph. and increased HPP tail water.

The response of the controlled water-turbine model is

presented in Fig. 6 (dashed line) with guide-vane Yvm 4 IMPACT OF THE HPP OPERATION ON THE
and runner Ygm opening. The measurements of both 110 KV ELECTRICAL NETWORK
openings (Yv and Yg) are presented with a solid line. The Referring to the proposed model of the 110 kV
displacement of the guide-vane opening Yvm (Fig. 6, top electrical network from Sec. 2.5 and initial state
graph) that participates in synchronization, emergency parameters (Table 1), the impact of the changed active
shut-down functions and load rejection are faster than power on the 110 kV network and frequency deviation
the runner change at opening. According to the increase in the range 0.025 p.u. with the same initial condition
in the Yvm up to 0.45 p.u., active power Pgm increases, is presented in Fig. 7.
whereas the runner (Fig. 6, second graph) stays in a In the first case (at a constant frequency), the
closed position. At higher referential active power Pref response of simulated active power Pm (grey dashed
(Fig. 3, top graph, from 0.48 p.u. up to 0.82 p.u.), runner line) compared to measurements P (black solid line) is
opening Ygm increases to provide more water through depicted in Fig. 7 on graphs (c), (d), and (e) for the
the turbine at consequently higher active power Pgm. In Klee-Medvode, Labore-Mavie and Okroglo-Labore
the simulation, real net head Hm (Fig. 6, bottom graph) 110 kV transmission lines, respectively. At the
is included to achieve the real cirumstances. beginning of the simulation, the direction of power flow
was from the Okroglo to the Klee switchgear (positive
Pm at Okroglo-Labore and negative Pm at Klee-
Medvode). The sink of the modelled active power on
the Klee-Medvode transmission line increases from
Pm = -9.3 MW to -14.5 MW, while active power Pm at
Labore-Mavie changes its direction (from source
Pm = 3.95 MW to sink Pm = -6.55 MW).

Figure 5. Comparison between measured excitation current If

and voltage Uf (solid line) and model of the excitation system
with Ifm and Ufm (dashed line).

Figure 6. Measurement of Kaplan turbine guide vane Yv and

runner opening Yg (solid line) and model responses Yv and Yg Figure 7. Impact of the HPP Mavie operation on the Klee-
(dashed line) at the same water conditions (equal heads Medvode, Labore-Mavie and Okroglo-Labore 110 kV
H=Hm). transmission lines at constant frequency fm (grey dashed line)
and changed frequency ff (black dashed line). The
During the increase in generator active power Pm, net measurements of frequency f (graph (a)), generator power Pg
head Hm (Fig. 6, bottom graph) is decreased (from (graph (b)) and active power P on the transmission lines
(graphs (c), (d) and (e)) are depicted with a solid black line.
A COMPLEX HYDRO-POWER PLANT DYNAMIC MODEL INTEGRATED INTO THE ELECTRICAL NETWORK 189

On the Okroglo-Labore transmission line, the source of The presented dynamic model uses a useful software
active power Pm decreases (from Pm = 26.2 MW to equipment for further development of the HPP control
23.6 MW). The comparison presented in Fig. 7 shows a systems, study of the impact on electrical network and
correct matching between the simulated and actual allows to simulate the tests which are difficult to
mesurements. The largest difference between the model evaluate on a real power system.
and measurement results P= -3.6 MW is noticed on the
Labore-Mavie transmission line at time t= 470 s). REFERENCES
[1] The Institute of Electrical and Electronics Engineers (IEEE),
In the second case, the increase in frequency ff IEEE Std 421.5-2005 IEEE Recommended Practice for
(Fig. 7, graph (a), black dashed line) for +0.025 p.u in Excitation System Models for Power System Stability Studies,
the time range from 300 s to 500 s and decrease New York, USA, 2006.
of -0.025 p.u. from 700 to 900 s is simulated. At the [2] Shayeghi H., Shayanfar H. A., Jalili A. Load frequency
increase in frequency ff, simulated active power of control strategies: A state-of-the-art survey for the researcher,
Energy Conversion and Management, 50(2009), pp. 344-353,
generator Pgf decreases from 0.48 p.u. to -0.02 p.u. 2009.
(Fig. 7, graph (b), black dashed line) due to the steady- [3] Kishor Nandar, Saini R.P., Singh S.P., A review on
state droop R= 0.05 p.u. from Equ. (13). Consequently, hydropower plant models and control, Renewable &
the sink of the active power on the Klee-Medvode Sustainable Energy Reviews, 11(2007), pp. 776-796, 2007.
transmission line (Fig. 7, graph (c), black dashed line) [4] Naghizadeh R.A., Jazebi S., Vahidi B., Modelling hydro
power plants and tuning hydro governors as an educational
decreases for 3.6 MW, whereas the source of the active guideline, International review on Modelling and Simulation,
power on the Labore-Mavie transmission lines 5(4), pp. 1780-1790, 2012.
(Fig. 7, graph (d), black dashed line) and Okroglo- [5] Singh Parmar K.P., Majhi S., Kothari D.P., Load frequency
Labore (Fig. 7, graph (e), black dashed line), increases control of a realistic power system with multi-source power
generation, Electrical Power and Energy Systems, 42(2012),
for 7.1 MW and 1.8 MW, respectively. pp. 426-433, 2012.
At the time of 700 s, simulated frequency ff decreases [6] Prakash Surya, Sinha S.K., Simulation based neuro-fuzzy
for -0.025 p.u. (Fig. 7, graph (a)) and active power Pgf hybrid intelligent PI control approach in four-area load
(Fig. 7, graph (b)) increases to 0.98 p.u., however it is frequency control of interconnected power system, Applied
limited to 0.82 p.u. due to the turbine power range. The Soft Computing, 23(2014), pp.152-164,2014.
[7] Pan Indranil, Das Saptarshi, Frequency domain design of
transmission line active power (Fig. 7, graphs (c), (d) fractional order PID controller for AVR system using chaotic
and (e)) changes oppositely to the results obtained at the multi-objective optimization, Electrical Power and Energy
frequency increase. Systems, 51(2013), pp.106-118, 2013.
The total change in the transmission-line active [8] Prillwitz Fred, Al-Ali Salah Eddin, Haase Torsten, Weber
Harald, Saqe Lutfi, Simulation model of the hydro power plant
power (12.5 MW) is equal to the change in the Shkopeti, 6th EUROSIM Congress on Modelling and
generator active power (0.5 p.u.) due to the frequency Simulation, 9.-13. September, 2007, Ljubljana, Slovenia.
deviation. [9] International Electrotechnical Commission (IEC), IEC
60308:2005 Hydraulic TurbinesTesting of Control Systems,
5 CONCLUSIONS 2nd ed., Geneva, Switzerland, 2005.
[10] International Electrotechnical Commission (IEC), IEC 61362:
A complex dynamic model of a HPP integrated in a 2012 Guide to Specification of Hydraulic Turbine Control
110 kV electrical network is proposed. The basic Systems, 2nd ed.; Brussels, Belgium, 2012.
properties of the synchronous generator, static [11] International Electrotechnical Commission (IEC), CLC/TR
excitation system, water turbine, turbine controller, 60034-16-2:2004: Rotating electrical machines-Part 16-2:
power transformer and simplified 110 kV network are Excitation systems for synchronous machines Models for
power system studies, Brussels, Belgium, 2004.
realized in Matlab/Simulink with the SimPowerSystems [12] International Electrotechnical Commission (IEC), CLC/TR
library based on real-system parameters and IEEE 60034-16-3:2004: Rotating electrical machines Part 16-3:
recommended practice. Excitation systems for synchronous machines Dynamic
The results of using the developed dynamic model performance , Brussels, Belgium, 2004.
and its subsystems are validated with measurements [13] International Electrotechnical Commission (IEC), Draft IEC
61970: Energy Management System Application Program
taken at the Mavie HPP and show appropriate Interface (EMS-API) Part 457: Common Information Model
matching of the responses. The observed changes in the for Dynamics Profile, 1st ed., Brussels, Belgium, 2013.
active power from Pm = 0.12 p.u. to 0.82 p.u. strongly [14] Avdakovic Samir, Bosovic Adnan, Impact of charging a large
affect the power flow on the 110 kV transmission lines. number of electric vehicles on the power system voltage
stability, Elektrotehniki vestnik, 81(3), pp. 137-142, 2014.
The results obtained for the Labore-Mavie
[15] Volkanovski Andrija, epin Marko, Mavko Borut,
transmission line reveal a change in the active-power Optimization of reactive power compenzation in distribution
flow direction from Pm = 3.95 MW to Pm = -6.55 MW. networks, Elektrotehniki vestnik, 76(1-2), pp. 57-62, 2009.
When simulating the frequency deviation, the turbine [16] Gjorgiev Blae, epin Marko, Volkanovski Andrija, Kanev
Duko, Multi-objective power-generation scheduling:
controller with a steady-state droop parameter Slovenian power system case study, Elektrotehniki vestnik,
appropriately changes the generator active power 80(5), pp. 222-228, 2013.
without menacing the network stability. [17] Krause Paul C., Analysis of Electric Machinery, McGraw-Hill
Book Company, New York, 1986.
190 NAGODE, MUROVEC

[18] Rade Konar, Ispitivanje sinhronog generatora HE Mavie,

Zagreb, 1988.
[19] Grigsby Leonard L., Power system stability and control, CRC
Press Taylor & Francis Group, 2nd ed., Boca Raton, 2007.
[20] Andino Hydropower Engineering d.o.o., HE Mavie, 2 Kaplan
agregata Turbinski regulator in ostala oprema 1.01.1134,
Ljubljana, 2011.
[21] Kolektor ETRA d.o.o., Tehnina dokumentacija za
transformator tip BT 25000-119/10,5 YNd5, Ljubljana, 2013.
[22] Hydro-Qubec, Transnergie Technologies, SimPowerSystems
for use with Simulink, The MathWorks, Version 3, Natick,
2003.
[23] Glover J. Duncan, Sarma Mulukutla S., Overbye Thomas J.,
Power system analysis and design, Cengage Learning, 4th ed.,
Stamford, 2010.
[24] Elektrointitut Milan Vidmar, Raziskave kratkostinih razmer v
slovenskem EES do leta 2004, Ljubljana, 1999.

Klemen Nagode received his Diploma Engineer degree in

Electrical Engineering from University of Ljubljana, Slovenia
in 2007. He is currently a postgraduate student at the Faculty
of Electrical Engineering and works in the Department of
Electrical Engineering of Savske elektrarne Ljubljana d.o.o.
His research interests include automation, measurements,
control systems, modelling and simulation of power systems.

Botjan Murovec is an associate professor at the Faculty of

Electrical Engineering of University of Ljubljana, Slovenia.
He gained his B.Sc, M.Sc and D.Sc. degrees in electrical
engineering in years 1996, 1999 and 2002, respectively. His
research interests include machine vision, bioinformatics,
combinatorial optimization and embedded systems.