PV Generator Modelling in EMTP

Andreas Theocharis Marjan Popov

Department of Engineering and Physics Department of Electrical Sustainable Energy
University of Karlstad, KAU Delft University of Technology, TUDelft
Karlstad, Sweden Delft, The Netherlands
[email protected] [email protected]

Abstract— In this paper, the development of a photovoltaic condition changes. The proposed model is verified by
generator model in EMTP is presented. The model is based on measurements and it is suitable for transients and dynamic
the linearization of the current-voltage nonlinear equations of a power systems studies.
PV generator. The model is realized by using standard electrical
components and MODELS language in ATP-EMTP. The
operational environmental conditions are taken into account and II. MODEL DESCRITION AND IMPLEMENTATION
the proposed model is suitable for transients and dynamic power In this section, a compendious description of the linearized
systems studies. PV model based on [10] as well as the model implementation
into EMTP by using ATPDraw are presented.
Keywords— ATP-EMTP, photovoltaic systems, power systems
dynamics, power systems transients component A. Model Description
The PV model is developed by linearization of any voltage-
I. INTRODUCTION current equation. Linearization is based on Taylor series
The high penetration of hybrid systems of various energy approximation in the space [vpv(t0), vpv(t)] and evaluating the
sources are very popular and bring significant technical and expression at vpv(t0) where it is assumed that at time t0 the
research challenges such as its transient and dynamic voltage and the current of the PV generator have values vpv(t0)
behaviour, smart protection and self-healing techniques, power and ipv(t0) respectively, at time instant t=t0+Δt the operation
quality and energy management, energy storage and load point has values vpv(t) and ipv(t) respectively and Δt is the time
demand coverage, reliability and controllability or other. step. By keeping only the first two terms of the series function,
one arrives at the following expression
Literature progress on research and the applications of
hybrid systems reference the necessity for the development of
grid-components and systems models for transient and i pv ( t ) = I0 − v pv ( t ) rpv0 (1)
dynamic response studies [1-4] as well as protection
requirements against variety of faults [5-8]. A direct
combination to those studies concerns the development of where rpv0 is defined as the incremental resistance at the point
reliable units’ models as power systems components which [vpv(t0), ipv(t0)] and the parameter I0 is reflects a dependent
take into account the variability of the operational system current source because its value depends on the values of the
conditions and the internal non-linear attributes of system's radiation G, the absolute temperature Ta and the position of the
components. This indicates the demand of effective electric operation point at the previous time instant t0. The incremental
power management (power sharing among generation units, resistance and the depended current source I0 are determined
exploitation of excess power production) by using storage respectively using
devices in combination to advanced control and management
methods [9-16]. Towards that cutting-edge research topic, the −1
development of grid component computational models for  
inclusion into the most authoritative software platforms, like (
rpv0 =  − di pv dv pv )  v pv ( t0 ),i pv ( t0 ) 
 (2)
for example electromagnetic transient simulators and real time   
digital simulators, is important.
In this paper, the inclusion of photovoltaic (PV) generator
I 0 = i pv ( t0 ) + v pv ( t0 ) rpv0 (3)
model into Electromagnetic Transients Program (EMTP) will
be presented. The PV generator model is realized by using
standard components in combination to MODELS language in The equivalent electrical circuit that represents (1) is given
EMTP software. The advantage of the proposed model, in in Fig. 1 and defines the position of the operation point of the
comparison to other PV models developed in EMTP, comprise PV generator on the ipv–vpv plane at the time instant t=t0+Δt
its capability to reproduce the voltage-current non-linear using values from the previous time instant t0 through the
characteristic curve and to take into account operational update of the parameters I0 and rpv0.

978-1-5090-4479-5/17/$31.00 ©2017 IEEE 400

 TABLE I. MODELS STRUCTURE

Fig. 1. Equivalent electrical circuit of the PV generator.

TABLE II. ENVIRONMENTAL CONDITIONS AND PARAMETERS

Fig. 2. The equivalent circuit of the PV generator using ATPDraw

The ATPDraw circuit is shown in Fig.2 where the TACS

B. Model Implementation source represents the depended current source I0, the TACS
The implementation of the PV model in EMTP is resistor represents rpv0 and MODELS block incorporates (3)
accomplished by reproducing the equivalent of Fig. 1 by using and (5) - (9). The Table I summarizes the MODELS structure
ATPDraw, TACS sources and MODELS. The parameter rpv0 is in accordance to the PV model.
determined by applying (2) on the nonlinear ipv(vpv) equation
III. RESULTS
  k vm  In this section, the PV model is verified by comparing
i pv = I sc 1 − k1  e 2 pv − 1  (4)
   simulations results with measured values. Moreover, PV
model's performance under severe loading and environmental
where the coefficients k1 up to k4 and m are conditions is examined.
k1 = 0.01175 (5)
A. Model Verification
m The model has been verified by constructing in ATPDraw
k2 = k4 Voc (6)
the simple circuit of the PV generator supplying a variable
k3 = ln ( I sc (1 + k1 ) − I mpp ) k1 I sc  (7) resistor. PV's voltage and current are taken as outputs by
varying the resistor. The process is repeated for different values
of temperature and radiation and the results are plotted using
k4 = ln (1 + k1 ) k1  (8)
MATLAB in order to easily insert the measured values for the
different cases. Table II summarizes the environmental
k   Vmpp  conditions and the Vmpp, Impp, Voc and Isc parameters values used
m = ln  3  ln   (9)
 k4   Voc  in each case [17].
The simulation results are shown in Fig. 3 for all three
and Vmpp is the maximum power point voltage, Impp is the
cases in two y-axis plot. By comparing the measured values to
maximum power point current, Voc is the open circuit voltage
the values resulted from the simulations one may conclude that
and Isc is the short circuit current. Finally, one derives that rpv0
the PV generator linearized implemented into EMTP model
is given by
can successfully reproduce the vpv vs ipv curves under different
environmental conditions.
m
m −1 k2 v pv ( t0 ) 
rpv0 = 1 ( I sc k1k2 m ) v pv ( t0 )  e   (10)

401
Fig. 3. Simulation results (lines) vs measured values (dots): in relation to
Table II, case 1, case 2 and case 3 correspond to green, orange and blue
colors respectively.

Fig. 4. The block diagram of the DC system used for PV model performance
investigation.

B. Model Performance Under Environmental and Load

Changes
The performance of the implemented model has been also
tested under step changes of the environmental conditions as
well as under step changes on the direct current (DC) load. To
this end, the simple system in Fig. 4 of the PV generator
supplying a DC load through a buck-boost converter (BBC)
and in the presence of a maximum power point tracker (MPPT)
of adaptive step size has been developed using ATPDraw.
The MPPT is designed by using a MODELS block for
which the output is the duty cycle at t time D(t). The inputs are
the vpv(t) and ipv(t) time t while the values from the previous
time t0 for vpv(t0), ipv(t0), D(t0) and the PV power Fig. 5. The P&O MPPT algorithm modified by adaptive step size control
ppv(t0)=vpv(t0)ipv(t0) are already stored in the block. In Fig. 5 block.
one reads the developed algorithm based on the perturb and
observe (P&O) method modified by a simplified adaptive step • Scenario 2: Constant environmental conditions and load
size block according to [18] for stability and fast response of change at time t=2s.
the maximum power point (MPP). In the adaptive step size
block the parameters a1, a2 and a3 are chosen according to ppv • Scenario 3: Environmental conditions change at times
vs vpv plot and the parameters d1, d2 and d3 for the step size ΔD t=2s and t=6s and constant load conditions.
are in the range of some tenths to some thousandths. The duty
The simulations results are presented in the following
cycle results as
subsections.
1) Scenario 1
D ( t ) = D ( t0 ) + k ΔD (11) In this scenario, the environmental conditions and
parameters of case 1 in Table II apply and the simulation
where k is the directional parameter that represents the results are shown in Fig. 6. By observation of Fig. 6, the
movement of the operational point on the ppv vs vpv plane. system approaches steady state after 0.2s. At t=0.3s the MPPT
is activated and leads the operational point at the maximum
System's performance has been tested under the following power conditions (MPC).
three scenarios where in all scenarios the system starts from
zero conditions and the MPPT is activated at time t=0.3s: 2) Scenario 2
• Scenario 1: Constant environmental conditions and
constant loading conditions.

402
Fig. 6. Results for scenario 1: (a) vpv (red) and load voltage (green), (b) ipv Fig. 7. Results for scenario 2: (a) vpv (red) and load voltage (green), (b) ipv
(red) and load current (green), (c) ppv (red) and load power (green) and (d) (red) and load current (green), (c) ppv (red) and load power (green) and (d)
duty cycle. duty cycle.

In this scenario, the conditions of case 1 in Table II apply again where the DC load voltage and current have changed in
and Fig. 7 present the corresponding simulation results. In Fig. respect to the appropriate duty cycle new value.
7, one can see that after t=0.2s, where the system has
approached steady state, the MPPT is activated at t=0.3s and 3) Scenario 3
leads the operational point at MPC. At time t=2s the load In this scenario, from t=0s till t=2s the conditions of case 1,
changes and the MPPT leads the operational point at MPC from t>2s till t=6s the conditions of case 2 and from t>6s till
t=10s the conditions of case 3 in Table II apply respectively.

403
 Fig. 9. For scenario 3 the trajectories of the operational point on the vpv vs ipv
and vpv vs ppv planes.

Point A and end at Points I. The operational points move on the

vpv vs ipv and vpv vs ppv characteristic curves for each
environmental conditions and jump from the one characteristic
curve to the other when the step changes happen.

IV. DISCUSSION
The simulation results show the applicability of the
implemented model to represent the PV generator under
different environmental conditions by supplying the
corresponding Vmpp, Impp, Voc and Isc parameters values used in
each case. The step changes for all scenarios observed because
the environmental and loading conditions follow step changes
also.
The combination of the BBC with the MPPT successfully
drives the system to the MPC at which somehow increased
current and voltage ripple is observed in some cases.
Appropriate re-sizing of the filters used in the system could
reduce these ripples.
There is no aim to compare or further develop several
MPPT techniques in this paper and hence, the P&O method is
Fig. 8. Results for scenario 3: (a) vpv (red) and load voltage (green), (b) ipv
(red) and load current (green), (c) ppv (red) and load power (green) and (d)
fundamentally used. However, the basic adaptive step size
duty cycle. control that is used could be refined in order to reach faster and
more stable the MPC.
The simulation results are presented in Fig. 8 where one can
see that for all environmental changes the MPPT leads the
operational point at the corresponding MPC. Consequently, V. CONCUSION
both PV and DC load voltages and currents have changed in In this paper, a PV generator model is developed in EMTP
respect to the appropriate duty cycle values in each case. by using standard components in combination to MODELS in
Moreover, Fig. 9 shows the trajectories of the operational point ATPDraw environment. The proposed model reproduces the
on the vpv vs ipv and vpv vs ppv planes. The trajectories start from vpv vs ipv and vpv vs ppv characteristic curves by taking into

404
account environmental and system’s loading conditions [8] Andreas Theocharis, Marjan Popov, Vladimir Terzija, “Computation
changes. The P&O technique for MPPT is used to approach the of internal voltage distribution in transformer windings by utilizing
a voltage distribution factor,” Electric Power Systems Research, vol.
MPC by combining a BBC controlled by an adaptive step size 138, pp. 11-17, September 2016.
block for the duty cycle determination. The simulation results [9] Kun Ding, XinGao Bian, HaiHao Liu, and Tao Peng, “A Matlab-
are compared to measured values and very good agreement is Simulink-based PV module model and its application under conditions
observed in the case of the PV generator as individual of nonuniform irradiance,” IEEE Transactions on Energy Conversion,
component as well as successful performance is recorded in the vol. 27, no. 4, December 2012.
case of the DC system operation. [10] Andreas D. Theocharis, Eleftheria C. Pyrgioti, “Development of a
linearized photovoltaic generator model for simulation studies with
On-going work is conducted in order to include non- electromagnetic transient programs,” ETEP International Transactions
uniform and partial shadowing conditions into the PV model. on Electrical Energy Systems, vol. 25, no. 3, pp. 454-470, March 2015.
Moreover, possible advantages by using the incremental [11] Vijay K. Sood and Prabhpreet Bhalla, “EMTP model of grid connected
conductance method should be investigated because the pv system,” International Conference on Power Systems Transients
(IPST2013) in Vancouver, Canada, July 18-20, 2013.
implemented model is based on the incremental resistance rpv.
[12] M. Amin Zamani, Amirnaser Yazdani, and Tarlochan S. Sidhu, “A
control strategy for enhanced operation of inverter-based microgrids
REFERENCES under transient disturbances and network faults,” IEEE Transactions on
Power Delivery, vol. 27, no. 4, October 2012.
[1] Sungwoo Bae, and Alexis Kwasinski, Member, “Dynamic modeling and
operation strategy for a microgrid with wind and photovoltaic [13] Kinattingal Sundareswaran, Peddapati Sankar, P. S. R. Nayak, Sishaj P.
resources,” IEEE Transactions on Smart Grid, vol. 3, no. 4, December Simon, and Sankaran Palani, “Enhanced energy output from a pv system
2012. under partial shaded conditions through artificial bee colony,” IEEE
Transactions on Sustainable Energy, vol. 6, no. 2, pp. 198-209, January
[2] Nimish Soni, Suryanarayana Doolla, and Mukul C. Chandorkar, 2015.
“Improvement of transient response in microgrids using virtual inertia,”
IEEE Transactions on Power Delivery, vol. 28, no. 3, July 2013. [14] Xiangjun Li, Dong Hui, and Xiaokang Lai, “Battery energy storage
station (bess)-based smoothing control of photovoltaic (pv) and wind
[3] Prajna Paramita Dash, and Mehrdad Kazerani, “Dynamic modeling and power generation fluctuations,” IEEE Transactions on Sustainable
performance analysis of a grid-connected current-source inverter-based Energy, vol. 4, no. 2, pp. 464-473, April 2013.
photovoltaic system,” IEEE Transactions on Sustainable Energy, vol. 2,
no. 4, October 2011. [15] Y.-J. Wang P.-C. Hsu, “Modelling of solar cells and modules using
piecewise linear parallel branches,” IET Renewable Power Generation,
[4] Andreas Theocharis, M. Popov, R. Seibold, S. Voss, “Analysis of vol. 5, no. 2, pp. 215-222, 2011.
switching effects of vacuum circuit breaker on dry type foil-winding
transformers validated by experiments,” IEEE Transactions on Power [16] Phatiphat Thounthong, Arkhom Luksanasakul, Poolsak Koseeyaporn,
Delivery, vol. 30, no. 1, pp. 351-359, February 2015. and Bernard Davat, “Intelligent model-based control of a standalone
photovoltaic/fuel cell power plant with supercapacitor energy storage,”
[5] Andreas Theocharis, M. Popov, “Modelling of foil-type transformer IEEE Transactions on Sustainable Energy, vol. 4, no. 1, pp. 240-249,
windings for computation of terminal impedance and internal voltage January 2013.
propagation,” IET Electric Power Applications, vol. 9, no. 2, pp. 128-
137, 2015. [17] S. Lalouni, D. Rekioua, T. Rekioua, and E. Matagneb, “Fuzzy logic
control of stand-alone photovoltaic system with battery storage,”
[6] Ye Zhao, Jean-François de Palma, Jerry Mosesian, Robert Lyons, Jr., Journal of Power Sources , vol. 193, no. 2, pp. 899-907, September
Member, IEEE, and Brad Lehman, “Line–line fault analysis and 2009.
protection challenges in solar photovoltaic arrays,” IEEE Transactions
on Industrial Electronics, vol. 60, no. 9, September 2013 [18] B. Azizian Isaloo, P. Amiri, “Improved variable step size incremental
conductance mppt method with high convergence speed for PV
[7] Yang Wang, Peng Zhang, Wenyuan Li, Weidong Xiao, and Ali systems,” Journal of Engineering Science and Technology (JESTEC),
Abdollahi, “Online overvoltage prevention control of photovoltaic vol. 11, no. 4, pp. 516-528, 2016.
generators in microgrids,” IEEE Transactions on Smart Grid, vol. 3, no.
4, December 2012.

405