Improving the Photovoltaic Model

in PowerFactory

Farhan Mahmood

Degree project in
Electric Power Systems
Second Level,
Stockholm, Sweden 2012

XR-EE-ES 2012:017

Abstract
This master thesis project is carried out to improve the grid connected PV models in DigSilent
(PowerFactory).
A generic model in PowerFactory is already available in the form of a PV template in the library.
This model gives the basic understanding of the operation of PV system and has some basic
control systems in it. But there are many deficiencies of this model that need to be addressed.
This model has PV panel model with many assumptions and approximations. Also it does not
have any MPPT control to ensure PV system always operate on maximum power. There is only
one type of reactive power control is available based upon voltage deviation. And there is no AC
voltage and active power regulation is available. Under these circumstances, there arises a need
of an extensive model with all the essential control system to have detailed grid studies.
In this project a new model named KTH model, in PowerFactory is developed which is equipped
with many controls systems. Like generic model, this model also uses static generator to
represent the PV generator. The model has nominal rated power of 0.5 MVA and designed
power factor is 0.9. This model uses the same DC bus bar and capacitor model as of generic
model.
A detailed literature study is carried out to have information about the recent research in this
area. A new PV panel model is developed which demonstrated better output results as compared
to generic model. The main difference with the generic model is that this KTH model has a more
realistic PV panel model. Due to difference in modeling of PV panels, the output current and
power are different for two models.
Maximum Power Point Tracking (MPPT) control is developed with two different types of
methods Incremental Conductance and Perturb and Observe. The main objective of this control
is to calculate the DC reference voltage for the controller. Hence, by doing so, MPPT control
enables PV system to always operate on maximum power point. Comparison of these methods
shows that Incremental Conductance gives better results; therefore it is used in this project.
Setup A presents a detailed comparison between KTH model and generic model keeping the
same reactive power control in both models. This comparison indicates the validity of this
developed model. Then in Setup B, various additional control systems are implemented in
developed KTH model. These controls include active power control, four different types of
reactive power control and MPPT control. In the end extensive case studies are performed with
developed KTH model with all the essential controls implemented. Three different disturbances
are applied Irradiation change, External grid voltage change and three phase short circuit for the
both setups. Results of these case studies are compiled in the form of simulation plots and are
briefly explained. In the end, results are concluded and some future research tasks are suggested.
i

List of Abbreviations
PV

Photovoltaic

R&D

Research and Development

EPIA

European Photovoltaic Industry Association

EU

European Union

GW

Gigawatt

MW

Megawatt

MPPT

Maximum Power Point Tracking

IncCond

Incremental Conductance

P&O

Perturb and Observe

MPP

Maximum Power Point

STC

Standard Test Conditions

VSI

Voltage Source Inverter

CSI

Current Source Inverter

THD

Total Harmonic Distortion

LVRT

Low Voltage Ride Through

DSL

DIgSILENT Simulation Language

VSC

Voltage Souce Converter

PLL

Phase Locked Loop

RMS

Root Mean Square

EMT

Electromegnectic Transients

Impp

Current at maximum power point

Vmpp

Voltage at maximum power point

Isc

Short ciruit current of module

Voc

Open circuit voltage of module

AM

Air Mass

Euro Efficeincy

ii

Acknowledgments
All my acknowledgments go to ALLAH Almighty for bestowing me with health, knowledge and
patience to complete this work.
First of all, I would like to give appreciation to all the peoples who directly or indirectly assisted
me in completing my interesting research work for my master thesis and providing me an
opportunity to set important benchmark for my future professional career. It really has been a
memorable working experience added to my profile.
I acknowledge, with gratitude to my supervisor Robert Eriksson for all his supervision in this
project. He always supported me throughout my project work and was very kind to me at many
occasions. Especially, he helped me a lot in improving my thesis report.
Thereafter, I am greatly thankful and indebted to Afshin Samadi for his valuable time and
guidance throughout this project. He helped me a lot in model development process particularly
in PV modeling and MPPT control development. It was very kind of him that he always helped
me whenever I asked him anything regarding my project.
Special thanks should be paid to Prof. Mehrdad Ghandari who guided me during the last couple
of months of my projects. I believe that those months were the most important part of my thesis
as I learned a lot during these months. He specially, directed me in conducting various
comparisons and analysis of my developed PV model.
There are some special names that needs to be mentioned here who stood by me through thick
and thin of my stay here in Sweden , Shoaib Almas, Zeeshan Ali Khuram, Usman Malik, Amit
kumar Jha and Hamad Ali. Also I would be very delighted to thank Della Jose, who was my
project partner. We had many important and valuable technical discussions regarding our project
to keep it on track.
In the end I would like to thank my parents and family, who always supported me and boosted
my courage to complete this outstanding journey. Finally, I would like to state that all my
achievements in this life including this thesis is whole heartedly dedicated to my Father and
Mother as they always been my biggest motivation.

Farhan Mahmood
November 07, 2012
Stockholm, Sweden

iii

Table of Contents
1.Introduction......................................................................................................4
1.1Projectobjective..................................................................................................................................6
1.2Projectoutline.....................................................................................................................................7
1.3Projectcontribution.............................................................................................................................8

2.IntroductiontoPhotovoltaicSystem.....................................................................................9
2.1ElectricalcharacteristicsofPVsystem..................................................................................................9
2.2TypesofPVsystems...........................................................................................................................11
2.2.1 Standalonesystems............................................................................................................................11
2.2.2Gridconnectedsystems......................................................................................................................12
GridconnectedPVSystemComponents......................................................................................................12
2.3Inverters............................................................................................................................................13
2.3.1FunctionsofInverters.....................................................................................................................13
1. MaximumPowerPointTracking(MPPT).............................................................................................13
2. GridInterface.......................................................................................................................................13
3. PowerDecouplingbetweenACandDCside.......................................................................................13
4. GalvanicIsolationbetweeninputandoutput.....................................................................................14
2.4PVSystemconfigurations..................................................................................................................15
2.4.1Withrespecttoconnectionwithsystem............................................................................................15
2.4.2Withrespecttoconfigurationofinverter...........................................................................................16
2.5ImportantfactorstobeconsideredforPVinverter............................................................................18
2.6GridCodes.........................................................................................................................................19
2.6.1 Dynamicgridsupport......................................................................................................................20
2.6.2 Activepoweroutput.......................................................................................................................22
2.6.3 Reactivepowersupply....................................................................................................................23

3.PVModelingandMaximumPowerPointTracking........................................................25
3.1PVModule.........................................................................................................................................25
3.2MaximumPowerPointTracking(MPPT)............................................................................................28
3.2.1Background..........................................................................................................................................28
3.2.2MPPTScheme......................................................................................................................................29
3.3TypesofMPPTControl......................................................................................................................29
3.3.1PerturbandObserve(P&O)...................................................................................................................29
3.3.2IncrementalConductance(IncCond).......................................................................................................33
3.4ComparisonbetweenP&OandIncCond...........................................................................................36

4.PVModelsinPowerFactory....................................................................................................38
4.1GeneralPVmodelinginPowerFactory...............................................................................................38
4.2Staticgenerator.................................................................................................................................39

4.2.1Basicdata............................................................................................................................................40
4.2.2LoadFlowAnalysis...............................................................................................................................40
4.3GenericPVmodelinPowerfactory....................................................................................................42
4.3.1FrameofGenericPVModel.............................................................................................................43
4.4KTHModelinPowerFactory...............................................................................................................46
4.4.1Background.....................................................................................................................................46
4.4.2FeaturesofKTHModelinPowerFactory...........................................................................................46
4.4.3FrameofKTHModel........................................................................................................................47
SolarIrradiation(Slot1)................................................................................................................................47
Temperature(Slot2)....................................................................................................................................48
PVModule(Slot3).......................................................................................................................................48
MPPT(Slot4)................................................................................................................................................50
PLL(Slot5)....................................................................................................................................................50
DCBusbarandCapacitor(Slot10)..............................................................................................................51
Basicstructureofdaxisandqaxiscurrentcontrol.....................................................................................51
4.5ControlssystemsinKTHModel..........................................................................................................53
4.5.1Activepowercontrol(slot8)...............................................................................................................54
4.5.2MainController(slot11).....................................................................................................................57
4.5.3Qrefblock(slot5)................................................................................................................................60
4.6Comparisonoftwomodelsintermsofcontrolsystems......................................................................62

5.ResultsAnalysisandCaseStudies.........................................................................................62
5.1Singlelinediagramofthesystem.......................................................................................................63
5.2ComparisonandAnalysisofPowerFactoryModels(SetupA).............................................................64
5.2.1Case1:IrradiancechangeinPVsystem............................................................................................64
5.2.2Case2:Decreaseinexternalgridvoltageby1%................................................................................68
5.2.3Case3:Threephaseshortcircuit......................................................................................................72
5.3ResultsAnalysisofdevelopedKTHmodel(SetupB)...........................................................................76
5.3.1Case1:Irradiationchange.................................................................................................................76
a. Unitypowerfactor(QControl1).........................................................................................................79
b. DynamicPFoperationQ(P)(QControl2)............................................................................................80
c. DroopbasedreactivepowerQ(U)(QControl3).................................................................................83
d. ACvoltagecontrol(QControl4)..........................................................................................................84
5.3.2

Case2:Decreaseinexternalgridvoltageby10%......................................................................86

5.3.3

Case3:ShortcircuitfaultatMV(A)busbar..............................................................................91

5.3.4

ActivepowerCurtailment........................................................................................................94

5.3.5

Activepowercontrolwithsystemfrequency............................................................................98

5.3.6

ValidationofMPPTcontrol......................................................................................................99

5.4ComparisonbetweenKTHmodelandPWMconvertermodel(SetupC)............................................102

6.ConclusionsandFutureWork...............................................................................................107
Futurework..........................................................................................................................................108

7.References...................................................................................................................................109
8.Appendix......................................................................................................................................112
8.1ParametersusedforthenewPVmodel...........................................................................................112
8.2DSLcodesinsidethePVsystemblocks.............................................................................................116
8.3MATLABcodetodeterminethecontrollerparameters.....................................................................117
8.4DerivationforRsandVt...................................................................................................................119
8.5StabilityModels...............................................................................................................................120
8.6Buitincurrentcontrolle...................................................................................................................120

1. Introduction
Green Energy, the term is nowadays heard, seen and discussed in almost all the energy related
concepts. The world is running out of non-renewable energy, thus the green energy sources
gains more importance. Wind energy, solar energy, hydro-power, bio-gas energy etc. are the
most popular renewable energy sources. R&D in each of these areas is being carried out and
is continuing in all parts of the world. Out of which the solar energy is one of the cleanest and
the less expensive one. A few years ago the penetration of solar energy into the electricity
market was considerably negligible. But the recent statistics shows a drastic change in this
situation. As per EPIA forecast PV potential of the Sunbelt countries could range from 60 to
250 GW by 2020 and from 260 to 1,100 GW in 2030 [1]. Solar energy is highly promoted,
even the governments provides subsidies for the installation of Photovoltaic (PV) Systems.
According to the REN21 Renewables 2012 Global Status Report, there was 74% increase in
the installation of the PV plants in 2011 thereby increasing the total installed capacity to
70 GW worldwide. Large scale ground mounted systems are continued to be installed in
greater numbers which increases the market share of the solar energy to notable numbers.
During 2011 in EU, the solar PV capacity additions were higher than any other energy source
installation. Germany and Italy continued to lead the EU solar market when compared with
other countries. At the same time 2011 also saw the emergence of China as a strong player in
the solar energy sector [2]. A country based table on the installed PV capacity is shown in Fig.
1.1.

Fig. 1.1 Table on Worldwide PV system installation [2]

Power (GW)

PV market reports published by various organizations clearly show one major trend, the
installation of large scale grid connected PV plants in greater numbers. Also people are
getting more and more attracted towards solar energy which will result in an increase in the
small scale PV installations as well. Such an increase in PV installation can result in high
penetration of large amounts of PV energy into the electricity grid. According to the European
Photovoltaic Energy Association Report about 29.7 GW of PV systems is connected to the
grid in 2011 globally which was 16.8 GW the previous year. Out of 29.7 GW, 21.9 GW is in
Europe alone, of which 9.3 GW was in Italy and 7.5 GW in Germany during 2011. Below
Fig. 1.2 shows the increase in the amount of grid connected PV systems in Europe till 2011.

Year

Fig. 1.2 Evolution of new grid connected PV systems in Europe (MW) [2]

In Europe the on-grid capacity is far higher than the off-grid PV capacity of just 1% of the
PV installed capacity. But in countries such as USA, Australia and Korea the off-grid
capacities comes to several megawatts and so considered to play a significant role in the
penetration of the solar energy into the total energy market [2].
During 2011 considering the market scenario and energy demand, the contribution of the
installed PV capacity to the total electricity demand in Europe is 2%. A clearer picture can be
drawn from the Fig. 1.3 given below.

Contribution of PV towards Electricity demand (%)

Fig. 1.3 Actual vs. potential PV contribution to electricity demand in 2011(%) [2]

Fig. 1.3 shows that Italy and Germany leads the PV race, whereas Belgium, Czech and Spain
also show quite impressive numbers. The possible amount of PV contribution towards the
electricity demand is higher than the current contribution and has not achieved by most of the
countries except Czech Republic. Also Spain is very close to achieve their potential PV
contribution. For Czech Republic the possible % of contribution from PV towards the
electricity demand, based upon the installed capacity is 3.15 % whereas 3.14% is right now
their actual contribution. In case of Sweden, the numbers are not so good and even described
as an underestimated PV potential by European Photovoltaic Energy Association. The
cumulative installed capacity of PV systems in Sweden by the end of 2011 is 15MW and for
the year 2011, the annual installed PV power is 3MW.
Although by the end of 2011, the contribution of PV towards the global electricity demand is
0.5% and towards the peak power demand is 1%, but future looks quite promising [3]. Due to
the decrease in the PV cell manufacturing costs as compared with the previous years, more
and more positive efforts are coming to promote solar energy. So a better R&D and more
support schemes by the governing bodies can help this renewable energy source to become a
fair player in the world energy market.

1.1 Project objective

The high penetration of PV power into the existing electricity grid, demands more study and
analysis to enable safe and secure operation. There are several issues pertaining to this. One of
the major concerns is the impact of the PV system in the form of overvoltage that can result in
the voltage fluctuations. Another issue is the interaction of one PV system to another PV
6

system as well as its impact on the grid operations. Since the production of solar energy
depends upon weather, location and vary over time, PV system impact on the peak demand
and energy consumption can also be an area of interest. In order to reduce the impact of PV
system on the grid operations several rules termed as grid codes are formulated and issued in
countries such as Germany, Spain, Italy etc. that are necessary to be followed during grid
connection. Mainly these grid codes are issued for those PV systems which are connected at
least to the medium voltage power grid. A more detailed description on grid codes is given in
the coming Chapter 2.
The project aims to address and analyze some of the above discussed issues such as voltage
fluctuation problems, impact on grid operations etc. For such an analysis and study, a very
accurate PV system model is required that can give more realistic results. In PowerFactory
(power industry software), a PV system model is already available. The PowerFactory model
complies with the German Grid code and is simple in terms of design. Another model is also
available in PSCAD which is a very detailed model and the simulation using the same is a
time consuming process. So a much more detailed model complying with the standards in
terms of design in PowerFactory is the expected outcome of the project. The developed model
is used for utility grid connection studies. The results of this model are then compared with
the results of existing model available in PowerFactory.
When the term PV is mentioned, it consists of two main conversion systems. One is the
conversion of solar energy into DC power and the other is the DC to AC power conversion. It
is required to design PV panel model and Maximum Power Point Tracking (MPPT) control
for DC side design. It is also the objective of this project, to implement control systems of
converter such as DC voltage regulation, AC voltage regulation, Reactive power control and
Active power control etc.

1.2 Project outline

Chapter 2 is used to give a brief overview about the general Photo-voltaic system as
well as electrical characteristics of PV cells. Then an overlook on the Inverter
technology employed within the PV system is done and finally concluded with the
explanation of existing German grid codes.

Chapter 3 is dedicated to explain the modeling details of PV module inside the Power
Factory and also the details of modeling of Maximum Power Point Tracking (MPPT)
control. First, the PV panel model and its modeling methodology are explained in the
form of equivalent circuit, solar cell equations and modeling assumptions etc. Second
part of this chapter discusses about the MPPT control and two different methods used
in this project. In this respect, starting from introduction, background and types of
MPPT algorithms commonly used are presented.

Finally two MPPT algorithms used in this project, Perturb and Observe (P & O) and
Incremental Conductance (IncCond) are explained and results are compared in the
end.

Chapter 4 gives the brief inside into the modeling details of generic PV model
available in PowerFactory and newly developed PV model in PowerFactory. The main
idea of the project is to create a better PV system model than the existing generic
model, so it is necessary to be familiarized with the both models.
Chapter 5 is assigned for the analysis and results comparison between the existing
generic PV model in PowerFactory and the newly created PV model in Power Factory.
First, a comparison between two models is presented and after that some case studies
are presented for newly developed model in PowerFactory. In the end a comparison
between KTH model and PWM converter based model is presented.

Chapter 6 summarizes the conclusions drawn from this research project and it also
outlines the possible future work to refine the newly developed PV model in Power
Factory.

1.3 Project contribution

A new PV model developed in PowerFactory that can be used for PV system
simulations and grid studies.
A more realistic PV panel model with one diode model.
Two types of Maximum Power Point Tracking (MPPT) control developed Perturb
and Observation (P&O) and Incremental Conductance (IncCond).
A detailed comparison is performed between the Generic PowerFactory model and
the developed KTH model in PowerFactory.
Several case studies are performed with DC voltage control, AC voltage control,
Reactive power control, Active power control and MPPT control in developed
KTH model.
A comparison between KTH model having static generator and PWM converter
based PV model.

2. Photovoltaic System
Introduction to Photovoltaic System
Photovoltaic cells are made of semiconductor materials which have four valence electrons in
the outer shell and the most widely used semiconductor for making the PV cells is Silicon. In
case of semiconductors the conduction band is empty but the band gap between the
conduction band and the valence band is very low. Because of low band gap the easy lifting
of electrons from the valence band to the conduction when hit by the charged photons is
possible. This phenomenon which is termed as the photovoltaic effect is employed for
generating the current in the PV cells and is depicted in Fig. 2.1.

Fig. 2.1 Processes in an irradiated solar cell [3]

2.1 Electrical characteristics of PV system

Equivalent circuit with a single diode is the most popular PV model to describe the
characteristics of the photovoltaic system and is shown in Fig. 2.2. A non-irradiated solar cell
will behave very much similar to that of a diode. The diode based model with a controlled
current source can be used to predict the behavior of the PV system under various levels of
irradiation, temperature and load conditions. The series resistance Rs represents the voltage
drop during the transfer of charge carriers from the semiconductor junction to the external
contacts. The parallel resistance Rsh represents the leakage currents at the cell edges. With the
variation of irradiation and cell temperature, there will be considerable change in current and
voltage which result in the increase or decrease of PV cell power output. It is very much
essential to set the PV system to function at MPP (Maximum Power Point) to deliver the
maximum possible power and the below model can be used to design, the MPPT (Maximum
Power Point Tracking), inverter sizing and its control.

Fig. 2.2 Single diode equivalent circuit of a photovoltaic cell [4]

While describing the V-I characteristics of the PV system there are three main measurement
points as well as the obtained values which are important to be mentioned. 1) MPP 2) Shortcircuit measurement 3) Open circuit measurement.
MPP is the Maximum Power Point at which the photovoltaic system delivers the maximum
power for a particular irradiance and temperature, from which the voltage at MPP, Vmpp and
the current at MPP, Impp can be obtained. Short circuit measurement with a zero voltage can
give the short circuit current, Isc and the open circuit measurement with a disconnected load
can provide the open circuit voltage, Voc , all these measurement points can be identified in
Fig. 2.3.
The typical I-V and P-V characteristic of the solar cell at Standard Test Conditions (E=1000
W/m2, =25 degrees, AM=1.5) is shown in Fig. 2.3.

Fig. 2.3 IV and VI characteristic of a PV cell with MPP [3]

It can be noticed from the figure that the value of Vmpp is lower than Voc and also the Impp
value is lower than the short circuit current Isc . The cell efficiency will be higher if the
operating voltage and current are Vmpp and Impp for a particular irradiance and temperature
other than any other operating points.

10

The voltage and current dependence on the change of irradiation and temperature for general
PV cell is shown in Fig. 2.4 and Fig. 2.5. It depicts that the cell voltage has higher
temperature dependence whereas the current has got higher irradiance dependence which will
also result in the change of MPP.

Fig.2.4 Temperature Dependence of Solar Cells [4]

Fig.2.5 Irradiation dependence of Solar cells [4]

2.2 Types of PV systems

There are two major ways in which the photovoltaic system can be implemented for an
application
2.2.1 Standalone systems
This system can be for a domestic or non-domestic application. Mostly for domestic
application there will be battery storage which can act as a hybrid system in order to ensure
continuous power supply in case of no irradiance. Off grid domestic application usually are of
11

1 kW in size and can provide electricity for lighting, refrigeration and other low power loads
[3]. Non-domestic application first comes in the form of terrestrial PV systems and later
implemented for telecommunication, water pumping, navigational aids etc. In these cases
even for small power generation there is a considerable value and is implemented as a better
choice than any other power generation method. Stand-alone system can be turned into a more
reliable system when combined with wind turbines, diesel generators, battery storage etc. to
form the hybrid systems and is implemented in several applications [3].
2.2.2 Grid connected systems
As the name indicates these are PV systems connected to the utility electricity network and
can be distributed or centralized in nature. The distributed PV systems mostly are of roof
mounted, can supply power to the grid connected customer or directly to the grid and will be
connected to the low voltage transmission network. To be grid connected, a roof-mounted
1 MW PV system is considered to be large enough according to the PV standards [3].
Centralized systems usually are of higher power rating usually more than 1 MW and normally
connect with the medium or low voltage transmission network depending upon its rating.
These systems are mostly ground mounted and are equipped to supply the maximum power to
the grid satisfying the grid codes. Since the project discuss the grid connected PV system in
particular, a brief description about the basic components of such a system is given in the
coming section.
Grid connected PV System Components
A grid connected system can be divided into two major sections which is the solar power
conversion unit and the interfacing unit. The power conversion unit comes with the solar
panels, mounting equipment, DC-DC converters (if necessary) and DC cabling which helps to
convert the solar energy into useful DC power. DC Power produced by PV Array goes to the
DC Bus connected to it. After this a DC-AC inverter, converts the generated DC power to AC
power and falls in to the interfacing area of the PV system.

Fig.2.6 Grid connected PV system components

12

Then with the AC cabling the PV system can be connected to either low or medium voltage
transmission grid. Normally, PV system is connected to medium voltage AC grid through a
step up transformer as shown in Fig 2.6.

2.3 Inverters
Inverters are used in a photovoltaic system to convert generated DC power into the useable
AC power and to connect PV system to the AC grid. These kinds of inverters are usually fully
synchronized in voltage and frequency with the grid. The inverters ensure that PV system
should always operate at such operating point so that it gives maximum power. For this
purpose a special MPPT (Maximum Power Point Tracker) control is used with the PV
inverters. This MPPT control will be explained in details in chapter 4.
2.3.1 Functions of Inverters
When the focus is on the power electronics function of the PV inverter and all the additional
constrains like safety, efficiency etc. are ignored then all PV Inverters can be evaluated with
these following basic functions as specified [9].
1. Maximum Power Point Tracking (MPPT)
The inverter controls the DC voltage in order to ensure that the PV system always operates on
the point where it gives the maximum possible power point. This MPPT (Maximum Power
Point Tracker) function is very important as it is the main factor of efficiency of the PV
system. PV system should adapt with the environmental conditions like solar irradiance,
environmental temperature and shading conditions, and must shift the operating point to
maximum power as any of the above mentioned parameters changes in the system.
2. Grid Interface
This is the most important part of the grid connected PV inverters. Usually, Voltage Source
Inverters (VSI) is used for the connection of PV generator to the AC grid. This type of
inverter has the buck characteristics (output voltage is always smaller than input voltage).
This is done by using the transformer (inductor). Although there are some transformers less
topologies existing, those will be explained later in this chapter.
3. Power Decoupling between AC and DC side
The power fluctuations between AC and DC side have to be adjusted by some energy storage
device that can bear these changes and provide smoothness in the system. In this prospective,
electrolytic capacitors are used to provide this kind of decoupling. These electrolytic
capacitors form the DC link, and the capacity of this link is very critical with respect to the
life time of the inverters.
13

4. Galvanic Isolation between input and output

There are two different types of topologies i.e. with or without galvanic isolation that can be
implemented in PV systems. In the first method a proper galvanic isolation is provided, but
there is no isolation in the later method.
a) Conventional Transformer with Galvanic Isolation

Most commonly used method for the galvanic isolation is the conventional transformer
operating on grid frequency. This is tried and tested method and is being used right from the
start of this PV technology. But this has some disadvantages like high weight, high cost,
additional losses and non-unity power factor. In this topology, MPPT is performed by the
main inverter as shown in Fig. 2.7. By controlling AC current the power that is fed into the
grid can be controlled.

Fig. 2.7 Conventional Transformer Isolation Method [9]

When the power fed into the grid is changed, keeping the DC power constant, DC link
capacitor is charged or discharged thus changing the voltage at the terminal of the PV
generator.
b) High frequency DCDC converters without galvanic isolation

Due to various drawbacks of conventional transformer, there arises a need to search for
alternative method for this galvanic isolation. As a result high frequency transformer topology
arises as a new method that does not have any isolation. This method uses high frequency
DC-DC converter and does not have any transformer in the system. MPPT and change of
voltage magnitude is performed by DC-DC converter. PV system with DC-DC converters are
shown in Fig. 2.8.

14

Fig. 2.8 DCDC Converter Topology without transformer [9]

Transformer less topology as shown in Fig. 2.8 is an upcoming technology, and in

development stage. There would be no transformer in the system and switching is done by
IGBTs used in the form of a bridge. Neutral conductor of AC side connected to inverter is
grounded. This technology is developing as it has less overall losses, lighter in weight and
cheaper than conventional grid frequency transformer topologies. However there would be
some switching losses in these methods. In addition, topologies without transformer increase
the control over the system voltage and power because transformer limits the control of the
grid current. When there is no special need of galvanic isolation then this topology is simple,
efficient and cheap to use.
There are still many challenges involved in this type of method due to grounding current
problems and safety hazards issues generated due to leakage problem. This demands the need
of a special measurement facility for the currents. If safety and EMC aspects are considered,
current flowing through the earth has to be limited and can be considered as the biggest
challenge in this technology. Special Residual Current Devices are used in the inverters to
have monitoring ground leakage current in this respect [9].

2.4 PV System configurations

Generally types of inverters are classified here in two categories as follows
I.
With respect to connection with system
II. With respect to configuration of inverter
2.4.1 With respect to connection with system
According to type of connection to the system, inverters are categorized into the two
Types:
a) Single phase inverter

Single phase inverters are used for applications like a roof top of a house or office. These
types of inverters are usually available up to 5 kW of rating.

15

b) Three phase inverter

Three phase inverters are usually related to larger systems, and are only used for grid
connected PV systems. Three phase converters are practically implemented in a PV system by
having three single phase converters connected to each load terminals. This is because for a
three wire topology relatively higher DC voltage values (around 600 V for a
400 V three
phase grid) is required and is limited to 1000 V due to safety reasons in installation
procedures [8].
2.4.2 With respect to configuration of inverter
Depending upon the way in which the inverter and PV array is configured, inverters can be
categorized into four general types. They are central inverters, string inverters, multi string
inverters and AC module inverters and are shown in Fig. 2.9.

=
AC

AC

a) Central Inverter

b) String Inverter

=
=
=

=
AC

=
=
AC

c) Multi-string Inverter

c) AC module Inverter

Fig. 2.9 Different configuration of PVSystem with inverters [9]

a) Central Inverters:

PV modules are connected in series and parallel to get the higher power level and finally
connected to a single converter in the end. Series connection of modules is called a string.
16

This kind of inverter has enough voltage at its DC side i.e. from 150 V to 1000 V that there is
no need to use an intermediate DC-DC converter to boost the DC voltage up to a reasonable
level. It is available in several kW to 1 MW range of power and can be used for almost all
levels of voltage applications.
Central inverter has got the advantage of high inverter efficiency at a low cost per watt. As
efficiency is one of the major concerns in the PV systems, makes central inverter a better
economical choice. Due to a single converter this configuration has low total harmonic
distortion (THD) losses in the system. Therefore it is still first choice of medium and large
scale PV systems. Central inverters are mainly built with three phase full bridges having
IGBTs and low frequency system [9]. The structure of the central inverter can be seen in
Fig. 2.8 (a).
This kind of inverter has the disadvantages especially for the small roof top application that it
results in mismatching losses between modules of string and missing individual Maximum
Power Point Tracking MPPT [9]. Also it has high losses in the DC cables. It has the
availability issue as if inverter trips, whole generation is out of the system. Shading effect due
to variable irradiance in the system can make it a bad choice as an inverter.
b) String Inverters

For the applications where different panel modules cannot be operated on the same orientation
and if the system is subjected to different shading conditions, string inverters are the best
choice [9]. As shown in Fig. 2.8 (b), it does not have any parallel connection. Each inverter is
responsible for each string having its own MPPT control. When the system has many strings
and each string is different from others in configuration then this type of inverter is ideal to
use. This arrangement facilitates to use it under the constraints like different orientation of the
parts of the roofs, different shading conditions and types or number of modules in each string.
It is available from 0.4 kW up to 2 kW of power [9].
String inverter has higher price per kW as compared to Central inverters because of its low
power per unit. Mostly string inverters are used in low power application with single phase
full bridge topology with low frequency transformer on the AC side for isolation.
c) Multi String Inverter

This type of inverter is also a type of string inverter with additional DC-DC converter for each
string. Basically it is a string inverter with two or more inputs. Each string having a DC-DC
converter is used as the input to main DC-AC inverter as shown in Fig 2.8 (c). With this kind
of arrangement one can have more power output of the inverter without having to sacrifice the
advantage of string technology [9].
Multi String Inverter has the disadvantages that due to two power conversion level it has more
power losses and less efficiency as compared to string inverters. It has the advantage that the
user can have a big freedom with respect to the input voltage range of the converter because
17

of an additional DC-DC converter. Also by having a separate MPPT control of each string it
is more efficient in that kind of operation. Multi string inverter has the power range of 1 kW
to 6 kW [9].
d) Ac Module Inverter

The AC module inverter has the arrangement that each module has its own inverter and
MPPT control. Several module inverters are used to compensate the high power level as this
inverter has lower power handling capability. In this case each inverter can be directly
connected to AC grid that is why it is called an AC module Inverter. This type of inverter is
shown in Fig 2.8 (d).
The main advantage of this inverter is that no DC wiring is necessary and risk of electric arc
and firing is eliminated in this case. It has low acceptance due to several disadvantages.
Firstly it has low power per unit that leads to low efficiency and high costs. Secondly, this
type of inverter technology does not reach the life time of the PV modules [10].

2.5 Important factors to be considered for PV inverter

There are some inverter constraints that should be kept in mind while selecting an appropriate
inverter for a PV system. First of all efficiency of PV inverter is most important, as inverter is
the interfacing part of the PV system. Inverters are prone to AC as well as DC side
disturbances. Any reduction in the inverter efficiency would straight away cause reduction in
the overall efficiency of PV system. Also PV array operate at the rated power for only a few
hours in a year, because the changing solar irradiance so in this condition inverter
predominately operate under part load. Generally, increasing the power rating will increase
the efficiency of the inverter.
Euro efficiency or European efficiency of PV inverters can be used as an authentic method to
compare the efficiencies of different inverters, and can be expressed by the following
equation [11].
0.03

0.06

0.13

0.1

0.48

0.2

The 0.03 % factor means that inverter is operating at 5% efficiency for a period of 0.03 time
period out of the total operating time period. Above equation consider the amount of time in
percentage that PV inverter is expected to work at partial loads or at different levels of
irradiation.

18

Fig. 2.10 Efficiency over a Range of Relative Photovoltaic Generator Powers [13]

Another aspect of PV inverter is anti-islanding protection. When PV inverter does not

disconnect to utility even when it has been shut down, then it causes problems as it continues
to supply power to grid even if it is not required. This problem arises as, in grid there are
some circuits that may resonate with grid frequency, and in this case this issue can be
dangerous for the grid side. So, inverter should be disconnected from the grid after it has been
shut down to avoid any kind of accident.
Harmonics in the output power provided by the inverter to the utility should be minimized as
much as possible, because these harmonics causes the distortion in the grid voltage and
current. It can be said that power quality can be improved by minimizing the THD (Total
Harmonic Distortion) from the grid voltage. Finally the PV inverter should be compatible
with the PV array such that the inverter can perform the MPPT operation efficiently. It should
also be able to adjust with the dynamics of the MPPT operation during irradiance changing
conditions.

2.6 Grid Codes

Grid connected photovoltaic systems should comply with the grid or utility requirements for a
safe and steady operation. Nowadays it is getting very difficult for the manufacturers and the
developers to comply with all these various requirements put forward by various bodies.
There are different types of documents like national standards, grid codes, company
regulations or rules etc. which will at the end confuse with regard to its realization and
compliance. There are several initiatives taken by various organizations to get a unified grid
code at least in the European Union and the most notable are as follows [7]:

19

European Committee for Electro-technical Standardization(CENELEC)

Network of Excellence of DER laboratories and Pre-standardization(DER-Lab)
International white book on the Grid Integration of Static Converters
Address project: Active Distribution network with full integration of demand and
distributed energy resources
EU-DEEP: The birth of a European Distributed Energy partnership

Even though the photovoltaic system has not reached a level where it can compete with other
conventional energy sources, the systems which have almost reached its advanced stage of
installation and manufacturing exits in Germany, Italy and Spain and all these countries
follow their own grid codes. As mentioned most of the countries follows their own grid codes
and is not so uniform. In order to avoid further complications, all the European country
specifications are recommended to comply with the European standards [7]:

EN 50160 Voltage characteristics of electricity supplied by public distribution systems

EN 50438 Requirements for the connection of micro-generators in parallel with public
low-voltage distribution systems

German grid code is found to be used as a reference code for various studies and is the most
updated one. So in this section the various requirements that the grid connected PV systems
should follow during its installation is presented mainly based upon the German grid code.
Existing renewable plants connected to the low and medium grids mostly does not contribute
much to the grid stability. But according to the current guidelines PV plants or renewable
plants connected to the grid should support the grid to maintain its stability and may also
remain connected to the grid during fault. Some of the major and should be mentioned
requirements are described in the following sections.
2.6.1 Dynamic grid support
Fault ride through which can also be termed as Low Voltage Ride through (LVRT) is the
capability of any electric system to remain connected to the grid in case of temporary voltage
drops or load change. So with the dynamic grid support the main aim is to obtain the LVRT
capability for the renewable system, which is the PV system.
During LVRT, the system must be capable to carry out any of the following options:
Remain connected to the grid
Improve the voltage stability by providing reactive power
Disconnect during fault and reconnect immediately after the fault clearance
The limiting curves of voltage for the renewable system is shown in Fig. 2.11

20

 Fig.2.11 Limiting curves of voltage in the event of a network fault [6]

For any voltage drop of 150 ms, the PV system should not disconnect from the system.
During voltage dips above the borderline 1, the voltage stability will not get affected
considerably and need not require a disconnection. When the voltage drop is between the
borderlines 1 and 2 then the system must be capable of fault ride through. Once the voltage
dip crosses the borderline 2 it is always allowed to have short disconnection and if necessary,
longer disconnections as well as agreed by the system operator. There is no requirement for
the system to remain connected to the grid if the voltage drop is below the blue line [5].
Next during the event of a network fault, how the voltage support of the renewable system
connected to a medium voltage grid should happen is shown in Fig. 2.12. If the system
voltage variation is within the dead band of 10%, no reactive current support is required or
enabled. This dead band is to prevent unwanted injection of reactive current. For a voltage
drop of more than 10% of the network voltage, the system should support the grid by
supplying reactive current, and is recommended to act within a time period of 20ms of the
fault. The reactive current support can be even of 100% of the rated current during necessary
conditions [5].

21

Fig. 2.12 Principle of voltage support at the event of network fault [5]

2.6.2 Active power output

The PV system is recommended to be capable of decreasing its active power output at any
operating conditions as and when required by the system operator. These target values can be
pre-defined by the system operator at the common coupling point. The active power output
should change with at least 10% of the network connection capacity per minute without the
disconnection of the plant from the grid [5].
In case of system frequency increase beyond 50.2 Hz, then the active power output should
reduce with a gradient of 40% of the generators instantaneously available capacity per Hertz.
Once the frequency reach back to 50.05 Hz then the active power output can be increased
such that the system frequency will not exceed 50.2 Hz [5]. Fig. 2.13 clearly shows how the
active power output control according to the system frequency can be designed and realized
for a renewable system and in this case for the PV system.

22

Fig. 2.13 Active power reduction in case of over frequency [5]

2.6.3 Reactive power supply

Current PV systems are designed to provide only active power. But even during normal
operation there are possibilities of very slow or low voltage fluctuations in the system which
is necessary to be kept within the acceptable limits. In order to ensure the same, the grid codes
recommend the exchange of reactive power between the grid and PV system to a very limited
extent. In case of PV system normally the inverters are oversized to meet the reactive power
requirements [6].
A PV system can carry out the reactive power exchange based upon the three possible
operations listed below [5]:
Power factor based operation- can be constant or dynamic
Fixed reactive power value(Q in MVAR)
Variable reactive power depending upon the voltage, Q(U)
In case of constant power factor, the PV system will supply reactive power irrespective of the
AC bus voltage. The dynamic power factor operation which is shown in Fig. 2.14 is such that
the value of power factor and thus the reactive power varies with the change in active power
produced, independent of the AC bus voltage. For the PV plants the recommended variable
power factor operation is between 0.9 under-excited and 0.9 over-exited states [5].

23

Fig. 2.14 Dynamic power factor operation characteristic [6]

24

3. PV Modeling & Maximum Power

Point Tracking
In this chapter firstly the Photovoltaic system modeling is discussed. In this respect some
derivations and solar cell basic model will be presented. Secondly, assumptions and
approximation, on which the model is built, are presented. The other part of this chapter
contains the Maximum Power Point Tracking (MPPT). In this respect, different methods of
MPPT will be first discussed and then modeling procedure on which it is modeled is
described. In the end comparison between two MPPT methods is presented.

3.1 PV Module
Inside the PV module, basic solar cell equations are used to calculate current on the basis of
input voltage at any instant. Some derivations of basic solar cell equations are carried out to
calculate the unknown parameters using the method as indicated in [4]. The equivalent circuit
of solar cell with a series resistance is used for derivation of the PV equations as shown in Fig
3.1. This series resistance Rs refers to losses due to poor conductivity in the solar cell.
In this model one diode model is used for modeling of PV module as shown in Fig 3.1. One
diode model is used for the sake of avoiding complexity in the system.

Fig 3.1 The equivalent circuit of a solar cell with a series resistance

From Fig 3.1, the output current of the PV module can be calculated using (3.1) and (3.2).
The total output current of this model is the difference between photo electric current and
as shown in (3.1). This diode current is itself equal to a quantity having
diode current
multiplication of (dark current) and an exponential term as shown in (3.2).

25

(3.1)
(3.2)
In case of short circuit conditions, using (3.2), the photo electric current is equal to the short
circuit current of the PV module as shown in (3.3).
(3.3)
In an ideal solar cell the saturation current (dark current) can be calculated by using the open
circuit conditions in the method used in [4]. It can be worth noticing that this current only
depends upon the temperature of the module and is independent from the irradiance
variations. The expression for the dark current is given by
(3.4)

Where Voct is the open circuit voltage with just temperature dependence and Isct is the short
circuit current with just temperature dependence.
The unknown parameters in (3.2) i.e. Rs and Vt are derived using the method explained in
Appendix 8.4.
PV Array does not operate on Standard Test Conditions (STC) at all the time. Whereas all the
parameters inside the PV module data sheet, are given on Standard Test Conditions (STC), so
corrections should be made to convert all the voltages and currents in the calculation
accordingly. Also any changes in temperature and irradiance of the system should be
incorporated in the calculations, so that PV system could modify its voltage and current based
upon any change in the irradiance and temperature. So these issues are included in these
calculations using (3.5) - (3.8) as described in [4] and [14].
Short circuit current, which is almost linear function of irradiance and temperature, so
accordingly expression for short circuit current is approximated as follows.

(3.5)
(3.6)

26

Where Isct is the short circuit current with just temperature dependence,
is the temperature
correction factor for current and is defined in the data sheet of the PV module. Isc is the short
circuit current with both temperature and irradiance.
Open circuit voltage can be expressed as linear function of temperature and irradiance.
(3.7)
(3.8)

Where Voct is the open circuit voltage with just temperature dependence and
is the
temperature correction factor for the voltage which is defined in the data sheet of PV Module.
Also Voc is the open circuit voltage with both temperature and irradiance dependence.
Where
Tref = Temperature at the STC considered as 25 C.
Isc1= Short circuit current at STC (defined in the data sheet of the module)
Voc1= Open circuit voltage at STC (defined in the data sheet the module)
E = Irradiance at the output
T =Temperature at the output
ESTC = Irradiance at STC

These above derived (3.1) - (3.8) are used in the DSL code of the PV module inside the PV
array model for calculations. But (3.2) cannot be solved numerically, as current is the function
of voltage and current i.e. I=f (I,V). So Newton-Raphson method is employed in order to
calculate the current.
The general expression for Newton-Raphson is described by (3.9) in which current for the
next iteration can be estimated.

(3.9)

Where
I1=Calculated current for the next iteration
Iini=Initial value of current
F(I0)=Function to be solved
F(I0)= Derivative of the function

27

Equation (3.2) is taken as a function to be solved as given by

exp

(3.10)

The derivative of this equation is calculated and is shown below

exp

(3.11)

In the first iteration, the current value for the next iteration is calculated from (3.9). In this
respect 4 iterations are used to estimate the value of the current precisely. In MATLAB for
loop is used to run 4 iterations, but in PowerFactory loop cannot be applied so (3.9)-(3.11) are
used as it four times to estimate the final value of the current.

3.2 Maximum Power Point Tracking (MPPT)

3.2.1 Background
The output of a photovoltaic system depends very much upon the weather conditions i.e. solar
irradiation, temperature and shading effects due to clouds. These parameters in a PV system
never remain constant, instead kept on changing at each instant. Also mostly solar cells in the
market can achieve a maximum of 15-20% energy conversion. But after a lot human effort
and research, Atla electronics (USA) reached to an efficiency of 23.5 % as verified by
National Renewable Energy Laboratory (NREL) in 2012[15] [16]. So it the power electronic
converters that provide the controllability over this energy conversion issue from sun light to
electricity. Under these circumstances it is very important to for power electronic converter to
have such a control that can ensure to have a maximum possible output power from PV array
with changing weather conditions.
MPPT function is to regulate the DC output voltage or current in such a way that the
maximum possible power can be obtained, with respect to any changes in weather conditions.
The maximum power point (MPP) is the point at which system has the highest possible
efficiency. For one weather condition there can only be one operating point in the system that
can give the optimal maximum efficiency. Therefore to track this point in the system is very
important in order to increase the system efficiency. Particularly, the system having any kind
of converter needs Maximum Power Point Tracking (MPPT) in order to make sure that it
deliver maximum power to other side.

28

PV array characteristics are nonlinear and greatly depend upon weather condition as shown in
Fig 3.2. Due to this weather dependence there is one single operation point in each
characteristics that will give us the maximum point out of array. Particularly with grid
connected PV system, it is very important for a converter to track the maximum power point
with every change in the solar irradiation and temperature. Therefore MPPT control in a PV
system became very essential for an efficient system.

Fig 3.2 Irradiation and Temperature Impact on powervoltage characteristics of PV array

[14]

3.2.2 MPPT Scheme

The power generated by a PV system is nonlinear, but static function of PV generator
terminal voltage i.e. Vdc. For each irradiation and temperature value it exhibits a global peak
in power corresponding to an optimum voltage. Maximum power can be obtained from the
PV system, if its dc voltage is regulated at an optimum value. Therefore function of MPPT
control in a PV system is to find the value of optimum dc voltage and issue the dc voltage
reference command i.e. Vdcref. This reference dc voltage is then further given to the main
controller that will regulate dc voltage accordingly.

3.3 Types of MPPT Control

There are many MPPT techniques proposed and implemented in PV system e.g. Perturb and
Observe (P & O)/ Hill Climbing, Incremental Conductance (IncCond), Fractional Open
Circuit Voltage / Constant Voltage (CV) and Fractional Short Circuit Current as explained in
the literature [14], [17] and [18]. The methods vary in complexity, sensors required,
convergence speed, cost, range of effectiveness, implementation hardware, popularity, and in
other respects.
After many comparative studies and surveys, P&O and IncCond techniques can be said as two
most widely used MPPT method as indicated in [14], [17] and [19]. These two methods are
implemented in this project and are explained below in details.
3.3.1 Perturb and Observe (P & O)
P & O method is one of the most widely used MPPT methods because of its simplicity and
ease to implement. It works by creating a perturbation in terminal dc voltage of PV array and
observes its consequences on output power of PV array. If power increases with incremental
29

perturbation, it continues to make perturbation in the same direction otherwise it is reversed.

In case of power increase with incremental perturbation, the operating point would be on the
left hand side of the maximum power point.
If incremental perturbation causes power to decrease, it indicate that MPP has crossed and
operating point is somewhere on the right hand side of MPP. In this case, it reverses its
perturbation and start making detrimental perturbation in the voltage of PV array in order to
track the maximum power point. Once the MPP is achieved, the operating point would be at
maximum power point as shown in Fig 3.3.

Fig 3.3 Perturb and Observe (P & O) [20]

It has three conditions

I.
dP/dV>0 (Left hand side of MPP)
II. dP/dV<0 (Right hand side of MPP)
III. dP/dV=0 (At MPP)

(3.12)

It can also be seen from Fig 3.3 that voltage corresponding to MPP called Vmp and current I
called Imp. These values at STC are available in the data sheet of every PV module.
Fig 3.4 illustrates a flowchart of the algorithm used to implement this MPPT technique in this
project as explained in [17]. Like most of the other techniques, P &O also needs measurement
of Vpv and Ipv as input to its algorithm.

30

Fig 3.4 Flowchart of P & O Algorithm [7]

The algorithm calculates the power Ppv by multiplying Vpv and Ipv. After that it checks the
value of dP/dV, according to (3.12). If dP/dV is greater than zero, it set the value of k=1
making an incremental perturbation in the voltage. The quantity V is actually a voltage
increment step and is a design parameter. This parameter depends upon the tracking speed and
accuracy of algorithm [17] and is calculated by

(3.13)

Where
Vmp0= Voltage of PV array at maximum power point at STC
Tdelay= Time delay value between two measurements in seconds
freq= Frequency of MPPT algorithm in Hz (which is inverse of Tdelay)
step= Design Parameter
The parameter V is very important in deciding the speed of the algorithm as well as
accuracy. While designing the MPPT control it should be kept in mind that V should be
carefully selected, as bad choice of this parameters can lead to wrong operation of MPPT.
Generally, too large value of V would cause ripples in the output voltage, whereas too small
value of V can result in a slow response of MPPT at the output. It is a tradeoff between fast
tracking speed and accuracy of the output. Typical MPPT frequency used in the MPPT is 1020 Hz, but higher frequencies can be used in order to reduce the ripples in the output of PV
array.

31

Fig 3.5 shows the model of P&O MPPT technique implemented in PowerFactory for the PV
system. This model is based upon the algorithm shown in Fig 3.4. In this model Varray and
Iarray are two inputs taken from the output of the PV array. In the start, delay blocks are used in
front of voltage and current inputs, in order to take measurement values of Varray and Iarray
after specified interval. Due to first main voltage delay block, a delayed value of voltage
obtained i.e. V123, which is further passed through another voltage delay block to get the
value of Udelay. Hence Change in voltage is calculated as follows
123

(3.14)

Fig 3.5 Model of P & O MPPT technique implemented in Power Factory

To get delayed value of power, first delayed values of voltage and current i.e. V123 and I123
are multiplied to get delayed power. Similar to voltage, this power then passed through a
power delay block to get the value of change in power described as
123

(3.15)

By analyzing the DeltaP/DeltaV according to conditions presented in (3.12), the voltage value
for the next cycle can be calculated as

(3.16)
1
V

32

At the right hand side of the above equation, Vpv[n] is the voltage value of PV array for
current cycle whereas quantity kV is actually deciding factor in voltage increment or
decrement for the next cycle. More specifically, V is the constant value once calculated, but
magnitude and sign of k decides the increment or decrement in the value of voltage for the
next cycle.
P & O MPPT technique has one major drawback that once it tracks down MPP, it starts
oscillating around that maximum point. Exact MPP can never be reached with this method.
Also, these oscillations cause a power loss which depends upon the size of single perturbation
step. A solution to this conflicting situation is to have a variable perturbation size that gets
smaller towards the MPP as indicated in [14], [18] and [21]. This phenomenon of oscillation
around MPP creates severe problem in case of fast changes in environmental conditions.
Under fast changing of environmental conditions, P & O may fail to track the correct
maximum power point and can be diverged from actual point. Some techniques like three
point weight comparison or optimizing sampling rate have been introduced to overcome this
drawback [22].
3.3.2 Incremental Conductance (IncCond)
In order to avoid the drawbacks caused by P&O method, Incremental Conductance (IncCond)
MPPT techniques is developed. This method provides better results in tracking the MPP
without having many oscillations around maximum power point. This method based on the
fact that the slope of power curve is zero at MPP. Calculating the slope of power curve as
shown in Fig 3.6 as follows

0
0

(3.17)
(3.18)

Using (3.17) and (3.18), IncCond algorithm works. Particularly (3.18) is evaluated such that if
this equation holds it means MPP is achieved. In (3.18), instantaneous conductance (I/V) and
incremental conductance (dI/dV) are compared; when both of the conductances are equal
MPP is met.
From (3.17) and (3.18) it can be concluded that
I.
dI/dV>-I/V (Left of MPP)
II. dI/dV<-I/V (Right of MPP)
III. dI/dV=-I/V (At MPP)

33

(3.19)

Fig 3.6 Incremental Conductance (IncCond) [20]

From Fig 3.6, it is clear that if dI/dV>-I/V holds, operating point is at the left of the MPP and
voltage needs to be increased to track towards MPP. Similarly if dI/dV<-I/V holds then
voltage should be decrease in order to find MPP. But if dI/dV=-I/V holds, there is no need to
change the terminal dc voltage as in this case it is already at MPP.
This can be achieved by applying an intelligent algorithm as described by the flowchart of Fig
3.7. It starts by taking the measurement values of voltage vpv and current ipv. If there is no
change in vpv and ipv as compared to last cycle, it goes back and takes new measurements. As
soon as it observes some changes in vpv or ipv, it starts operating. From Fig 3.7 it can be
noticed that if (dipv/dvpv+ipv/vpv) is positive and larger than an small value, , then operating
point is at the left hand side of the MPP, and voltage is increased by one incremental step; if
(dipv/dvpv+ipv/vpv) is negative and larger than small value , then voltage is decreased by one
step. Once (dipv/dvpv+ipv/vpv) becomes an absolute value smaller than , MPP considered to
have been reached and voltage is kept unchanged, as soon as a new change is observed in ipv.
The magnitude of voltage incremental step V is calculated from (3.13). As already
explained, this step value determines the speed of the algorithm, so it should be carefully
selected.

34

Fig 3.7 Flowchart of IncCond Algorithm [7]

In IncCond method, small value , is very important in perturbation step calculation, as

according to (3.16) the voltage value for the next step is dependent on the value of k, which is
indirectly dependent on epsilon. This small value , is selected by considering a tradeoff
between the problem of not operating exactly at MPP and possibility of oscillating around it
[13]. If too high value is selected, it would cause operating point of PV system to move away
from MPP. On the other hand, if too low value is selected, it can cause more oscillations
around MPP. Therefore by hit and trail method a suitable value for each case should be
selected.
Fig 3.8 shows a model of IncCond which is implemented in PowerFactory for this project.
This model is based upon the algorithm shown in Fig 3.7. Similar to P & O method, in this
model here Varray and Iarray are two inputs taken from the output of the PV array. In the start,
delay blocks are used in front of voltage and current inputs, in order to take measurement
values of Varray and Iarray after specified interval. From main voltage and current delay blocks,
the delayed values of voltage and current are obtained i.e. V123 and I123. Then, in order to
get the values of change in voltage dv and change in current di for each cycle, these
values are passed through voltage delay and current delay blocks.

There are four inputs to the input selector i.e. delayed voltage, delayed currents, di and dv.
According to the algorithm shown in Fig. 3.7, input selector decides the value of k. After
calculating the value of k, using (3.13) it calculates the value of V in the next block named as
deltaU. This block also takes value of k as input, multiplies it with the V value and finally
gives the value kV. In the end, the value for the Vdcref i.e. reference dc voltage, is
calculated using (3.16).
35

Fig 3.8 Model of IncCond MPPT technique implemented in Power Factory

3.4 Comparison between P & O and IncCond

Incremental Conductance (IncCond) shows better performance in fast changing
environmental conditions. First advantage of IncCond is that it does not oscillate around MPP
as it was the case for P & O method. This is due to the check of condition di=0, which made it
possible to maintain a constant voltage across dc link once MPP is found. Furthermore, using
the conditions (3.20) and (3.21) helps in determining the relative location of MPP [23].
0

(3.20)

(3.21)

It can be observed from Fig 3.9 and Fig 3.10 that in case of P&O MPPT method, continous
oscillations around MPPT can be observed. But in case of IncCond these oscillations are
removed, by using some additional conditons i.e. di=0.This condition made it possible to
maintain a constant voltage across dc link once MPP is achieved.
These oscillations are also present in P&O method even when there is no disturbance in the
system as shown in Fig 3.10. But IncCond exhibits better results than P&O method as there

36

are no such oscillations present in this method. These oscillations causes power losses in the
system and are not desirable to have efficient power system.

Fig 3.9Comparison of P & O and IncCond MPPT

Comparison between P & O and IncCond MPPT in steady state
701
700.8

IncCond

700.6

P&O

700.4

Vdc (V)

700.2
700
699.8
699.6
699.4
699.2
699

10

time(s)

Fig 3.10 Comparison of P & O and IncCond MPPT in steady state

37

4. PV Models in PowerFactory
This chapter provides the basic overview of the PV models in PowerFactory. Modeling details
of generic PV model available in PowerFactory and newly developed KTH model in
PowerFactory are presented. The content of this chapter validate that newly developed KTH
model is much improved one as compared to the generic model available in PowerFactory.

4.1 General PV modeling in PowerFactory

A grid connected PV system without any battery storage mainly consists of a PV array and an
inverter with its DC link capacitor as shown in Fig. 4.1.

Fig. 4.1 Grid connected PV system model

So while modeling a PV system it is essential to model these basic components. The modeling
varies from software to software depending upon the availability of components and
requirements.

1. PV array

A PV array can be generally represented as a current source. At the same time in order to
study the behavior of a PV system it is necessary to include the electrical characteristics of a
PV module as discussed in Section 2.1. Such a PV array model can behave like a realistic PV
module, of which the output power varies with the change in system conditions such as
change in irradiation, temperature etc.





38

2. Interfacing converter

In PowerFactory the interfacing converter can be represented by either a static generator or a
PWM converter. In PowerFactory when PV system is represented by the static generator, then
there is no need to physically represent interfacing converter. In this case static generator or
includes the PV array and interfacing converter in it.
But in case of a PWM converter, PV array needs to be physically represented with a current
source connected to a DC bus bar along with a DC link capacitor. PWM converter is also
separately represented. The two different possible ways of PV system modeling in Power
Factory are shown in Fig. 4.2.

Fig. 4.2 Different PV system modeling in Power factory

In the PowerFactory there is a PV system component called the PV generator that is available
for PV studies which is modeled with the help of a static generator. In PSCAD a PWM
converter based PV system is developed and is currently available for grid studies. But in this
project static generator is used to model PV generator and will be explained in details in
coming section.

4.2 Static generator

Static generator is an easy to use model of any kind of static (no rotating) generator in Power
Factory. It can be used for following [24].
39

Photovoltaic Generators
Fuel Cells
Storage devices
HVDC Terminals
Reactive Power Compensators
Wind Generators

4.2.1 Basic data

As in this project, static generator is used as photovoltaic generator, so it is selected from
category list as Photovoltaic. Any number of parallel machines can be used according to
system rating requirements. In this respect, the total rating of generator would be equal to the
rating of a single generator multiplied by the number of parallel machines used. Also under
the ratings tab, it is required to specify the rating for one generator in MVA and the power
factor at which machine is designed as shown in Fig 4.3.

Fig. 4.3 Basic Data for Static Generator

4.2.2 Load Flow Analysis

A typical interface window of the static generator in load flow is shown in Fig. 4.4. All values
can be changed according to the requirement. The active power tab shows the power rating of
the PV system which is the active power output out of the static generator. For the load flow,
the variables need to be defined initially to calculate the load flow analysis; can be set using
40

static generator. In Fig 4.4 it can be seen that active power and reactive power are defined in
the start to calculate the load flow of the system. These variables can be changed according to
the requirement of the system.

Fig 4.4 Static generator power flow under steady state conditions

The reactive power limits of the PV system are based upon either capability curve of the static
generator shown in Fig. 4.5 or can be manually defined.
Capability curve of the static generator has active power on its Y-axis and reactive power is at
X-axis. All the values are in per unit. The inverter cannot operate on maximum active and
reactive power at the same time. Assume that the PV system is expected to work at 0.9 power
factor. In capability curve of the PV system shown in Fig. 4.5, the upper red line is the limit of
the active power that can be transferred at 0.9 pf. The vertical red lines denote the limits of the
41

reactive power transfer at one specified active power as shown as Qmax and Qmin in Fig 4.5.
According to the system operator, control system and the method used for reactive power
control, the reactive power that is to be transferred is decided. The blue line denotes the power
limit of the inverter. The inverter cannot operate outside its limit since it is limited by nominal
power of the generator.

Fig 4.5 Capability curve of the inverter

4.3 Generic PV model in PowerFactory

The basic Photovoltaic system in the generic model is developed using a static generator. This
PV system model is available in the newest version of PowerFactory tool in the form of a
template. This template consists of a PV generator along with basic controls and design
features. As the name suggests the model is generic and can be modified as per the
requirement such as ratings, type of control etc.

Fig 4.6 Photovoltaic template

42

4.3.1 Frame of Generic PV Model

The control frame of the PV System in generic model is shown in Fig 4.7. As mentioned
previously it is necessary to model the PV array, DC link capacitor dynamics by DSL
functions. The same can be observed in the control frame and the main blocks will be
explained in brief in the following sections.

Fig 4.7 Frame of Generic PV Model

Slot 1 & 2: Solar radiation and temperature

Solar radiation and temperature both are modeled as limited time integrators. The purpose of
these slots is to accumulate change of irradiance per second or change of temperature per
second and integrate them over a period of time. The output of these blocks is given to slot 3
which is the photovoltaic model.
Slot 3: Photovoltaic model

Photovoltaic model takes DC voltage, irradiance and temperature as the inputs. Based upon
PV equations written as DSL codes, the PV array is modeled and provides array current and
reference DC voltage as the outputs. It can be noticed that in this PV model there is no
maximum power point tracker (MPPT) control available. The output voltage of this model is
assumed to operate on maximum power point and is taken as equal to the reference dc voltage
43

vdcref for the controller. There are many unknown approximations used to calculate the
final module current. Inside the PV panel model variables and parameters used for calculation
are unknown and undefined. So it is very difficult to identify any scientific reference of these
equations.
Slot 4: DC bus bar and capacitor model

This slot represents the mode of DC bus bar and capacitor. With the help of this block the
dynamics of the DC side capacitor is included in the PV system. This slot has two inputs one
from the PV model in the form of array current and other from power measurement device in
the form of active power signal. The output of the block is the DC voltage across the capacitor
considered as the actual DC voltage, which is given as feedback to the PV module as well as
the input to the controller.
Slot 5: Controller block

Controller is the one most important part of the PV system and the control frame is shown in
Fig. 4.8. It is necessary to regulate the active and reactive power outputs of the static
generator according to the DC side output of the PV system consisting of the PV array and the
DC side capacitor. This purpose is realized with the help of the controller block. This block
has four inputs and two outputs. DC voltage from the capacitor model as actual DC voltage,
reference DC voltage from PV model, measured AC voltage and active power are the inputs.
The reference values of the id (for active power) and iq (for reactive power) are the outputs of
this controller. These currents values are given as input to the static generator, which is the
interfacing converter.
The upper part of the control frame is the DC voltage regulation which is implemented to
control the active power output of the PV system. The inputs are vdcref, vdcin and
dvdcref and the output is the d-axis component of the reference current, id_ref. The lower
part is the reactive power control. For the control the inputs are uac and uac0 and the output
is the q-axis reference current component, iq_ref. Reactive power control is implemented in
this model based upon a German grid recommendation shown in Fig. 2.12.
The equations represented in this block use a deadband of 10% of the nominal voltage and
determine the iq (q-axis component of the current). A deadband of 10% means, that it will not
provide or absorb the reactive power as long as voltage deviation would be within 10 % of
nominal voltage. This deadband value can be changed according to the requirement of the
system.
The q-axis component of the current is calculated as follows
K duac

(4.1)

44

Where
K= parameters named as droop
duac=Difference in actual voltage and steady state voltage

Fig 4.8 Control frame of Generic Model

In (4.1) the value of K is calculated on the basis of droop control and is denoted as droop in
the parameters. The definition of iq is written according to Transmission Code 2007 [5] and
System Service Ordinance SDLWindV.
In the end the values of id , iq and the voltage deviation duac enters the current limiter that gives
the final reference values of id and iq as id_ ref and iq_ref. Current limiter, limits the current in the
d axis in case of fault which is based upon the value of the current in q-axis. This limiter also
controls the value of id_ref, in case of short circuit fault, such that there would not be any active
power during the short circuit. Also current in q-axis is limited between maximum and
minimum absolute possible current. This upper and lower current limit values are used as
parameters and can be changed accordingly.
Slot 6: Static generator block

This block represents the static generator in the control frame. The additional inputs that are
to be given to the static generator can be given to this block which is internally connected to
45

the actual one. So the output from the controller block that is the d-axis and q-axis reference
current components as well as the cosref and sinref values from the phase measurement
unit are given to slot 6. Based upon which the output of the static generator is regulated.

4.4 KTH Model in PowerFactory

4.4.1 Background
It has already been discussed that a generic model of PV system is available in PowerFactory
as a template. This PV model gives the basic understanding about the working principle of a
PV system and is explained under section 4.3 in details. But this generic model has following
issues:

PV array model with lot of undefined assumptions and internal variables.

No MPPT control
Only one type of reactive power control available
No active power curtailment

So under these circumstances, there arises a need of a model which could explain the PV
system in more details and can exhibit PV system with more controls in it.
4.4.2 Features of KTH Model in PowerFactory
This KTH model is equipped with following additional features as compared to generic
model:
1. A more realistic PV Array Model
Based upon the equations with authentic scientific references.
Incorporates the environmental impacts (Irradiation & Temperature)
One diode model is used to model PV module
2. Two types of MPPT control:
Perturb and Observation (P & O)
Incremental Conductance (IncCond)
3. Four types of Q control
Unity Power Factor (Q Control 1)
Dynamic Power Factor (Q Control 2)
Droop based Control (Q Control 3)
AC Voltage Control (Q Control 4)
4. Active Power Control
Active power curtailment based on given Pref
Based upon frequency
46

4.4.3 Frame of KTH Model

General overview of the KTH model is shown in Fig 4.9. This is the complete frame of the
new KTH model in which solar irradiation model, temperature model, PV array model,
different control systems, measurement devices and static generator are shown. This will be
elaborated in details in coming section.

Fig 4.9 Frame of a KTH Model

Solar Irradiation (Slot1)

Solar irradiation is modeled in the form of an integrator that will sum all the potential change
of solar irradiation per second dE and integrate over time. The variable dE is given as
input to this integrator.
The value of irradiation E along with another input Ein is then given to a summation
point that performs summation and finally gives the value of Eout which is the final value
of solar irradiation given to the PV array model .The solar irradiation model is shown in Fig
4.10. This model can be used to see the effects of changing solar irradiation in the form of
step on the output voltage, current and power of the PV array.
47

Temperature (Slot 2)

Similar to solar irradiation, temperature slot is modeled in the form of an integrator that will
sum all the potential change of temperature per second dtheta and integrate over time. The
variable dtheta is given as input to this integrator and it gives the temperature at the output.
This model is shown in Fig 4.11.
This value of temperature theta along with another input thetain is then given to a
summation point that performs summation and finally gives the value of thetaout which is
the final value of temperature given to the PV array model. Temperature slot is modeled like
this in order to have a step change of temperature.

Fig 4.10 Solar Irradiation Slot of a KTH Model

PV Module (Slot 3)

PV module is indicated as slot 3 in the frame of KTH model in Fig 4.9. The detailed PV array
model is shown in Fig 4.12.
This new PV array model is modeled in a more accurate way using the equations with
authentic scientific references. It also incorporates the environmental impacts (irradiation and
temperature) variation and output current has dependency on these environmental variables.
48

Fig 4.11 Temperature Slot of a KTH Model

Fig 4.12 PV module slot of KTH Model

49

Photovoltaic system is modeled by considering the basic equations of solar cell. This PV array
model is shown in Fig 4.12.
This model consists of several blocks in it. Input to this PV array model is dc voltage, solar
irradiation and temperature. At the output, it gives PV array current and PV array voltage.
This PV model has one main block (Photovoltaic Module), which has red colored box around
it. All the main equations are placed inside this PV module box. On the basis of these
calculations it gives voltage and current of PV array at the output. Inside the PV module all
calculations are performed for one PV module, but this PV array consists of a number of
series and parallel modules that on a whole makes this PV array.
Input dc voltage udc is in per unit and is passed through a low pass filter to attenuate the
high frequency under abnormal conditions and in normal conditions this block is deactivated.
After this the voltage is multiplied with its base value to get the actual value of the voltage.
This filtered voltage is divided by the number of series modules (ns) to make it input dc
voltage for one module, and is given to PV module for the calculations.
The other two inputs to the PV module are the solar irradiation Eout and temperature
thetaout, which are the values coming out from a solar irradiation block and temperature
block.
MPPT (Slot 4)

The slot 4 in the frame of KTH PV model represents MPPT control and has already been
explained in details under the section 3.2-3.4. The output of this slot gives the reference dc
voltage i.e Vdcref, which can be given to controller (slot 11) in the frame of new model via an
active power control block (slot 8). The details of these blocks will be explained in coming
sections.
PLL (Slot 5)

This slot contains a Phase Locked Loop (PLL) device that generates an output signal whose
phase is related to the phase of an input reference signal. PLL contains an internal oscillator
that is synchronized by being phase locked with some grid power signal. For this case, it is
locked to voltage in the system. In new model, this PLL is connected to LV bus bar.
The basic structure of a PLL is shown in Fig. 4.13. Phase detector produces a signal
proportional to the phase difference between V (input reference signal) and V (output signal).
Loop Filter is a low pass filter to cut off higher frequency signals. Finally, voltage controlled
oscillator adjust its frequency such that phase of input signal and output signal are matched.
PLL slot in this KTH model is same as of generic model.

50

 Fig 4.13 The basic structure of a PLL

The explanation of Slot 6, 7, 8.9,12,13 and 14 will be presented under the section 4.5.
DC Bus bar and Capacitor (Slot 10)

This slot has two inputs i.e. one from the PV model in the form of array current and other
from power measurement device in the form of active power signal as shown in Fig 4.14. The
output of this block is the dc voltage across the capacitor, which is used as the feedback to the
PV module and as well as the input to the controller.
In this model, a current is first calculated by dividing the measured active power by actual dc
voltage. This current is compared with input array current, which gives the actual current in
the capacitor. Finally, integrating this current gives the actual dc voltage across the capacitor.
The dynamics of the capacitor is given by
(4.2)

(4.3)
Pac is measured from active power measurement block, iph is the current output of PV array
and udc is actual dc voltage across capacitor. The capacitance value C is used as a parameter
and can be changed accordingly.
The description about other blocks of KTH model frame will be explained under the section
4.5 as those slots are related to control of PV system.

51

Fig 4.14 Dc Bus bar and Capacitor Slot

Basic structure of daxis and qaxis current control

In the stability model of static generator shown in Appendix 8.5, the inputs: id_ref, iq_ref,
cosref, sinref are used to determine the modulation index in the dq-reference frame. An inbuilt
current controller is available within the PWM converter of static generator as shown in
Appendix 8.6. The use of the built-in current controller is optional.
From the output definition of the stability model, it is clear that the converter output current is
projected on the dq-reference frame which is iq and id. So it is necessary to obtain the
reference values also in the dq-frame for which an outside controller is designed to obtain the
values of id_ref and iq_ref. This outside controller will be explained in the section 4.5.2.
The input reference values are compared with the actual converter output current id and iq.
Finally using the built-in current controllers, the d-axis and q-axis components of the
modulation index are calculated that in turns provide gate signal to VSC valves. Furthermore,
to obtain the values of cosref and sinref, a phase measurement unit is employed along with the
PWM converter. Based upon these reference angles, the d-axis and q-axis components of the
modulation index are transformed back to the global (abc) reference frame.
Finally, the active power can be regulated by controlling id which is based on dc voltage
regulation of PV system. And reactive power can be regulated by controlling iq which is
either based upon AC voltage regulation or reactive power control strategies.

52

4.5 Control system in KTH Model

The external control scheme will enable to obtain the reference value of the output current in
the dq-reference frame so that the desired ac output can be obtained from the PV system. The
main idea of the additional control is to provide an extra control over the output components.
For the current PV system several control are implemented in PowerFactory for the regulation
of DC voltage, AC voltage, reactive power output and active power output.
As explained previously the current generic PowerFactory PV model is equipped with active
power control based on the DC voltage regulation and the reactive power control based upon
the AC-voltage regulation. Along with those controls, extra controls are implemented in the
KTH model with options to choose depending upon the requirement. Various grid codes and
transmission regulations demand various requirements to be met by the grid connected PV
systems, which can be realized using these control techniques.
In Fig 4.9, slots 6, 7, 8 and 13 belongs to the active power curtailment of the PV system.
Frequency measurement (Slot 6)

Slot 6 is the frequency measurement unit. It measures the frequency of the system and gives
its output to active power reduction block (slot 7).
Frequency measurement (Slot 7 and Slot 8)

Slot 7 is the active power reduction block. Slot 8 is active power curtailment block. Slot7 and
Slot 8 will be explained under the section 4.5.1.
AC voltage measurement (Slot 9)

Slot 9 is AC voltage measurement device, which measures the AC voltage in per unit and
gives it to controller for AC voltage regulation.
Main controller (Slot 11)

Slot 11 is the main controller. The output out of the controller (slot 11) is given to static
generator. Slot 11 will be explained under the section 4.5.2.
Static generator (Slot 12)

Slot 12 represents the static generator which is directly associated with VSC in the system. It
has already been explained under section 4.2.
Power measurement (Slot 13 & Slot 14)

Slot 13 and 14 are the power measurement unit which will provide the measured AC power
from the AC side of the PV system. It should be kept in mind that slot 13 measures the active
power and reactive power and gives the corresponding measured value to controller, Qref
block and active power control block. In contrast, slot 14 is also a measurement device, but it
only measures the active power in MW and serves as the input to the DC bus bar and
capacitor slot. Both active power as well as reactive power are measured and are given to
various slots for further controller design.

53

Qref block (Slot 15)

Slot 15 is the Qref block which is designed to provide the Qref value to the controller (slot
11). The obtained Qref value out of the Qref slot is used inside the controller for reactive
power regulation.
The explanation about slots related to control system of KTH model frame is summarized in
Table1.

Table 1: Details of control system slots in KTH Model Frame

Slot Number

Slot Title

Frequency measurement

Active power reduction based upon frequency

Active power curtailment

AC Voltage measurement

11

Main controller

12

Static generator

13 &14

Power measurement

15

Power factor control

4.5.1 Active power curtailment (slot 7 & Slot 8)

Active power control as the name indicates will regulate the actual active power output out of
the inverter according to the reference value of the active power. The active power control is
mainly employed in order to decrease externally the power while the PV system is operating
around its maximum power output by varying the voltage. The below block diagram, Fig.
4.15 shows the active power control technique employed within the system.

54

Fig. 4.15 Active Power Curtailment

As seen in the figure, there are two options available to determine the value of Pr. First
option is to manually input the desired value which is given by Pref and will be enabled if the
parameter Act inside the active power controller is set to 1. The next option is to obtain the
active power based upon Pfreq value. Pfreq is the active power reference value obtained out
of the active power reduction block which is shown in Fig. 4.15.
In the grid code there is a recommendation to reduce the active power in case of over
frequency as explained in Fig. 2.13. The same is realized in the active power reduction block
where the actual system frequency Fmeas is compared with the limits. Based upon which
the Pfreq is calculated and is then given to the active power controller which will be used
further for the calculation of Vdcref. For any Act parameter value (here it is set to 0) other
than 1, Pfreq option is enabled and is the usual choice during the normal operating
condition.
Vdcref is the output of the active power control block which goes into the main controller.
MPPT connected system will have the value of Vdcref coming out of the MPPT function
block. In case if there is requirement for constant Vdcref then the value of Vdcref into the
controller can be set as Vmpp. When the requirement is to reduce the active power at
maximum power output state as per the operator wish, then the required Vdcref can be
obtained using the PI controller from Pref.

55

Fig. 4.16 Active power reduction block

During all the operating conditions other than Pref option, if there is any frequency increase
in the system then the active power will be reduced with the help of active power reduction
block. To choose from all these various options such as MPPT based Vdcref, constant Vdcref
and Pref or Pfreq based Vdcref, a selector switch with parameter Act is employed. Three
following options can be selected.

When the switch parameter Act=0, then the output Vdcref comes out of the MPPT block.

For Act=1, then the manual Pref option is enabled which allows to increase or decrease
the value of active power as per requirement.

When the switch parameter Act=2 then the value of Vdcref out of active power control
block is a constant which is normally Vmpp.

Further to obtain the value of Vdcref based upon Pref or Pfreq the PI controller employed
is having the equation:

Where

Kp is the gain of the active power controller

Tp is the time constant of the active power controller in seconds
56

(4.4)

4.5.2 Main Controller (slot 11)

The block diagram of the main controller implemented to obtain the values of iq_ref and
id_ref is shown in Fig. 4.17.

Fig. 4.17 Controller block diagram

The topmost part is the reactive power control with Q and Qref as input and iq as output
as represented in green color box in figure. Down to reactive power control is the AC voltage
regulation with uac and Vacref as inputs and iv as output as shown in blue color box.
Next is iqref which is a directly calculated iq_ref value and is an output of the power
factor (Qref block). In order to choose from the three different iq_ref calculation method
available, an input selector is provided with a parameter named Flag. Since the value of
iq_ref that is to be given to the VSC should be in pu, the chosen value out of the input
selector is converted to pu by dividing it by ibase.
The value of ibase is calculated using (4.5) as follows:
(4.5)
Where, Sbase is the inverter rating in MVA and Vsd is the peak value of the line-ground
inverter AC side voltage in kV.

57

The lower part of the control is for the DC voltage regulation with udc and Vdcref as
inputs and yi4 as the output as shown in blue box. The output yi4 is further converter to pu
using ibase to obtain the desired value of id_ref in pu.
In this controller block, reactive power control used in generic model is also used as it as an
additional option to control reactive power. This type of control exists in Reactive Power
Support block in Fig 4.17.
a) DC voltage regulation
DC voltage regulation is realized with the help of a PI controller which is the most widely
used method as shown in Fig. 4.18. In this case the inputs Vdcref and udc are used to
determine the error between them. Then the error is passed through the PI controller to obtain
the value of id_ref which is the input to the static generator to control the active power.
Then with the help of the in-built current controller the modulation index is calculated which
can be used to generate the switching pulses for the inverter. The PI controller uses following
equation.

(4.6)

Where

Kdc is the gain of the DC voltage regulator

Tdc is the time constant of the DC voltage regulator in sec.

Fig. 4.18 DC voltage regulation

For the PI controller, the values of Kdc and Tdc are determined by trial and error method to get
the desired output.
b) AC voltage regulation
The main idea behind the introduction of a separate control for the AC voltage is to keep the
voltage at the coupling point within the steady state voltage limits. If there is an increase or
decrease in grid voltage then the ac-voltage control will act to return back the system voltage
to the normal operating voltage or Vacref. This Vacref can be assigned with the desired AC
58

voltage, which is to be maintained at the regulation point after the PV system gets connected
to the grid. The regulation is achieved with the injection or absorption of reactive power by
the inverter according to its capability/limits. Nowadays for the medium voltage PV system
grid connections, grid codes allow or demands the PV system to operate within 0.95 lead or
lag power factor. So within these limits as well as depending upon the inverter ratings the AC
voltage control can be utilized for a better control over the output AC voltage.
As shown in Fig. 4.19 the error of the AC reference voltage, Vacref and the actual AC-voltage
from the inverter Vac is passed through a PI controller, Fac(s) using (4.6) to obtain the iq_ref.
This iq_ref is used as input to static generator to regulate reactive power and hence AC
voltage at the regulation bus.

Fig. 4.19 AC voltage regulation

The limits of iq are determined using (4.7). First Vsd is calculated as the peak value of the lineground inverter AC side voltage. Then Qs is already known as it is the maximum reactive
power of the PV system in MVaR (reactive power corresponding to 0.9 power factor in this
model). Since as all other quantities are known, iq can be easily calculated from below
equation.
(4.7)

This value of iq is used as maximum (positive value) and minimum (negative value) limits as
the parameters of the main controller.
c) Reactive power control
The actual reactive power termed here as Q in Fig. 4.20 is regulated to follow the Qref
value which is the desired reactive power. The Qref, variable to the main controller comes
from the Qref block which will be explained in the next section. Then the actual reactive
power output of the converter is compared with the obtained reactive power reference value to
get the error. The error is then passes through the controller, Fq(s) using (4.6) to obtain the
value of iq_ref which is the input to the static generator for reactive power control.

59

Fig. 4.20 Reactive power control

4.5.3 Qref block (slot 15)

The Qref input shown in Fig. 4.20 comes from the Qref block. The block diagram of the
Qref block employed along with the inverter control, in order to obtain the value of Qref
based upon different strategies is shown in the below Fig. 4.21.

Fig. 4.21 Qref block

60

The Qref block outputs the Qref value based upon the choice selected from three options.
1. unity power factor
2. dynamic power factor
3. Q(U) operation
In case of unity power factor operation, there will not be any reactive power support, ie. Qref
is equal to zero. This case of unity power factor operation is nowadays not advised rather
dynamic power factor operation is recommended.
Dynamic power factor operation will enable the PV system to take part in the grid-voltage
support in case of disturbances. Since the PV panel does not provide any reactive power
support the inverter is oversized to meet the reactive power grid requirements.
Fig. 2.14 demonstrates the dynamic power factor operation implemented for the new PV
system. According to the active power output, the power factor of the PV system is varied to
obtain the required reactive power support independent of the system voltage.
Droop based reactive power control is the third option such that the Qref is calculated based
on the variation in the system voltage. If the grid voltage change is within the entered deadband D there will not be any reactive power injection which will eliminate the unnecessary
reactive power absorption or injection. If the variation of system voltage is beyond this deadband, then there will be reactive power injection or absorption according to the control curve
shown in Fig. 4.22.

Fig. 4.22 Q(U) curve

The obtained Qref is given to the main controller block and is used to get the desired value of
iq_ref.

61

4.6 Comparison of two models in terms of control systems

In the end, the control systems present in both models are summarized in the form of a table
and is shown as Table 2.
It can be observed from Table 2 that there are some new control systems implemented in the
KTH model which were not present in generic model. Active power control, AC voltage
control, reactive power control (based on active power) and MPPT control are the newly
implemented control in KTH model. These additional controls make this new KTH model
more advanced and more reliable as compared to generic model.
Table 2: Comparison of two models in terms of control system

Control Technique

Generic Model

KTH Model

DC Voltage Regulation

MPPT (Maximum Power Point Tracking)

Unity Power Factor Control Strategy 1

Dynamic Power Factor Control Strategy

2

Droop Control based on Voltage

Active Power Curtailment

Frequency based Active Power Control

Control Strategy 3

AC Voltage Regulation
Control Strategy 4

62

5. Model Analysis and Case Studies

In this chapter, three setups will be discussed. In Setup A, a comparison between the results of
generic model and newly developed KTH model will be presented. Second part of this chapter
describes Setup B, in which several case studies are presented with different control systems
in new KTH model. In Setup C a comparison between KTH model based on static generator
and PWM converter based model is presented. The simulations will be presented and results
will be explained in details.

5.1 Single line diagram of the system

The single line diagram of the power system used in the analysis of both generic model and
new KTH model in PowerFactory is shown in Fig 5.1.

Fig 5.1 Single line diagram of the system

The system includes a PV model of 0.5 MVA and is connected to LV bus of 0.4 kV. The PV
system includes the inverter and gives the AC quantities at the output. The measurement point
of AC power and AC voltage is at LV bus. After the LV bus, a 2 winding transformer is
placed to step up the voltage up to 10kV at the MV (A) bus. This is followed by two parallel
short transmission lines each of 10 km. The system is lossless, so all the components in this
system have no resistance. After the transmission line, it is connected to external grid via
another MV (B) bus of 10 kV. The detailed system specifications of above system are
summarized in the table and are given in Appendix 8.1.

63

5.2 Comparison and Analysis of PowerFactory Models (Setup A)

This comparison is based on following facts:

Generic model and developed KTH model have the same reactive power control.
There is no MPPT control in both models at this time.
This comparison is without any additional controls developed in new model.

It is noticeable that PV array model in both models are different, which can be main
difference in two model at this stage. Reactive power control based upon voltage deviation is
used for both the models.
These models are compared by having different case studies. Three types of disturbances are
used to see its impact on the output of the models.

Case 1: Irradiance change in PV system

Case 2: Decrease in external grid voltage
Case 3: Three phase short circuit at MV (A) Bus

5.2.1 Case 1: Irradiance change in PV system

Irradiance in the system is decreased from 1000 W/m2 to 500 W/m2 at t=1 and brought back
to 1000 W/m2 at t=1.1. This is shown in Fig 5.2 below.
Irradiation change
1100

1000
Solar Irradiation

E(W/m2)

900

800

700

600

500

400
0.9

1.1

1.2
time(s)

1.3

Fig 5.2 Solar irradiation change

64

1.4

1.5

DC voltage of the both models with the irradiance change is shown in Fig 5.3. It can be seen
from the below Fig 5.3, that general response of DC voltage is same for both models. As in
this case a constant dc voltage reference is used for the DC voltage control so very quickly
after the irradiation change, voltage comes to its initial value. The change in the voltage due
to irradiation variation is less in case of KTH model as compared to generic model.
Irradiation change
1.2

1.15

Generic Model
KTH Model

X: 1.104
Y: 1.119

Udc(p.u)

1.1
X: 1.107
Y: 1.091

1.05

X: 1.318
Y: 1

0.95

0.9

0.85
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.3 DC voltage of PV system

Fig 5.4 shows the DC current of two models which is the main difference in the two models at
this time. It can be seen that current decrease is less in case of KTH model with the irradiation
decrease. This current difference is due to the different nature of the modeling method of the
PV array model. For KTH model, basic solar cell equations are used from authentic scientific
references and every variable and parameter is defined, whereas in case of generic model a lot
of assumptions are used in the modeling procedure. Also many undefined internal variable
and parameters are used.

65

Irradiation change
1.1
1
X: 1.25
Y: 1

0.9
Generic Model
KTH Model

Idc(p.u)

0.8
0.7
0.6
X: 1.086
Y: 0.4701

0.5
0.4

X: 1.084
Y: 0.4094

0.3
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.4 DC current of PV system

Also value of current obtained at E=500 W/m2 is very close to expected value in case of KTH
model, whereas in generic model value of current is somehow lower than expected. In other
words, with new KTH model we are getting more current and hence more power at a given
irradiation value.
Fig 5.5 depicts the active power of the PV system for both models. The green curve
representing expected result (curve from MATLAB equations). It is noticeable that red curve
representing the KTH model has power output which is closer to expected result as compared
to generic model.
AC voltage of the both model is shown in Fig 5.6 and is quite close to each other. Fig 5.7
represents the reactive power of the PV system in case of irradiation change. This reactive
power control is based on deadband, and in this case voltage variation is within deadband.
Therefore there is no reactive power regulation expected. Also in figure reactive power shown
is negligible. First of all reactive power is plotted in Mvar, secondly scale of y-axis is so
small, therefore difference between the reactive powers of two models is very minute. This
difference can be due to numerical calculation difference in two models.

66

Irradiation change
1.1
1
Generic Model
KTH Model
MATLAB

0.9

Pac(p.u)

0.8
0.7
0.6
X: 1.1
Y: 0.477

X: 1.078
Y: 0.4697

0.5
0.4

X: 1.091
Y: 0.4086

0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.5 Active Power of PV system

Irradiation change
1.0005
Generic Model
KTH Model

1.0004
1.0003

Uac(p.u)

1.0002
1.0001
1
0.9999
0.9998
X: 1.322
Y: 0.9997

0.9997
0.9996
0.9

1.1

1.2
time(s)

67

1.3

1.4

1.5

Fig 5.6 AC voltage at LV terminal

-3

10

Irradiation change

x 10

Generic Model
KTH Model

X: 1.107
Y: 0.008136

Qac(Mvar)

X: 1.34
Y: 5.6e-005

-5
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.7 Reactive Power of PV system

5.2.2 Case 2: Decrease in external grid voltage by 1%

This section illustrates the response of two models in case of external grid voltage decrease by
1 %. It should be noted that this disturbance is permanent i.e. at t=1 external grid voltage is
decreased by 1 % permanently. As both models have same reactive power control at this stage
so it is very interesting to see whether both models show same behavior or not in case of grid
voltage change.
The DC voltage increases a bit in both models with external grid voltage decrease at t=1 and
comes back to original value due to constant dc voltage reference setting. DC voltage is
shown in Fig 5.8.

68

External grid voltage decrease by 1%

Generic Model
KTH Model

1.0025

Udc(p.u)

1.002

X: 1.005
Y: 1.002
X: 1.005
Y: 1.002

1.0015

1.001

1.0005
X: 1.286
Y: 1

0.9995
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.8 DC voltage of PV system

Fig 5.9 and 5.10 shows the DC current and active power of the two models respectively. It is
clear from Fig 5.9 that response of the DC current of PV system is close to each other in case
of external grid voltage decrease. Similarly active power for both models as shown in Fig 5.10
exhibits the same response and small difference is due to the different modeling of PV panel.
The important thing to be mentioned here is in case of external grid voltage change, variation
in DC voltage and active power is very small. This is justified as in case of voltage decrease
of external grid, PV system should not change its active power instead reactive power and AC
voltage should regulate in this case.

69

External grid voltage decrease by 1%

1.0005

1
X: 1.217
Y: 1

Generic Model
KTH Model

Idc(p.u)

0.9995

0.999

0.9985

X: 1.005
Y: 0.9978

0.998

0.9975
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.9 DC current of PV system

External grid voltage decrease by 1%

1.004
Generic Model
KTH Model

1.002

Pac(p.u)

X: 1.25
Y: 1

0.998

0.996

0.994

0.992

0.99
0.9

X: 1.001
Y: 0.9917

X: 1.001
Y: 0.9908

1.1

1.2
time(s)

1.3

Fig 5.10 Active power of PV system

70

1.4

1.5

The response of AC voltage and reactive power in this case is shown in Fig 5.11 and 5.12
respectively. As both models have same reactive power control, the response of the AC
voltage is almost the same. AC voltage is decreased by same percentage as external grid
voltage is decreased. It should also be noticed that there is no AC voltage control applied,
therefore AC voltage at LV terminal is decreased.

External grid voltage decrease by 1%

1.002

Generic Model
KTH Model

X: 0.959
Y: 0.9997

Uac(p.u)

0.998

0.996

0.994

0.992
X: 1.216
Y: 0.9898

0.99

0.988
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.11 AC Voltage at LV terminal

Reactive power for both models shown in Fig 5.12 is also close to each other. The voltage
variation is within deadband so there would be no reactive power regulation expected in this
case. In figure reactive power is almost negligible. Also scale at y-axis is so small and
reactive power shown is in Mvar therefore difference between the reactive power values is
also very small. This difference can be due to numerical calculation difference in two models.

71

-4

3.5

External grid voltage decrease by 1%

x 10

Generic Model
KTH Model

X: 1.005
Y: 0.000337

Qac(Mvar)

2.5

1.5

1
X: 1.019
Y: 5.5e-005

0.5

0
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.12 Reactive power of PV system

5.2.3 Case 3: Three phase short circuit
In this section, a three phase short circuit is used as disturbance to check its impact on the
model output. As short circuit fault is the most severe disturbance in power system therefore
if model works well in this severe condition it will increase the validity of the model. Short
circuit event is created at t=1 and is removed at t=1.1 at MV (A) bus bar as shown in Fig 5.1.
Both models behave very similar to each other in case of short circuit fault in the system. DC
voltage of PV system is shown in Fig 5.13. At t=1 when short circuit happens, DC voltage for
the both models increases to open circuit voltage of PV system. When short circuit is removed
DC voltage comes back to its initial value.
Fig 5.14 and 5.15 shows the DC current and active power of the PV system during short
circuit, respectively. It can be noticed that during the short circuit fault duration i.e. 100ms
current and power of the PV system becomes zero and comes back to its initial value as soon
as the fault is removed. This response is very similar for the both generic and KTH model.
The small difference can be explained on the basis of the modeling difference of two PV
models.

72

Short circuit
1.3
X: 1.059
Y: 1.251

Generic Model
KTH Model

1.25

Udc(p.u)

1.2

1.15

1.1

1.05
X: 1.277
Y: 1

0.95
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.13 DC voltage of PV system

Short circuit
1.2

1
X: 1.265
Y: 1

Idc(p.u)

0.8
Generic Model
KTH Model

0.6

0.4

0.2
X: 1.055
Y: 5.989e-007

0
0.9

1.1

1.2
time(s)

1.3

Fig 5.14 DC current of PV system

73

1.4

1.5

Short circuit
1.2
Generic Model
KTH Model
1
X: 0.942
Y: 1

X: 1.267
Y: 1

Pac(p.u)

0.8

0.6

0.4

0.2
X: 1.052
Y: 0

0
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.15 Active power of PV system

Fig 5.16 and Fig 5.17 describes the AC voltage and reactive power of PV system for the both
models in case of short circuit fault. As expected, during short circuit i.e. from t=1 to t=1.1,
AC voltage at the LV terminal becomes zero for the both models. This AC voltage quickly
comes back to original value after the removal of this fault. As both models have the same
reactive power control so response of reactive power and AC voltage is expected to be quite
similar to each other.
This can also be confirmed from Fig 5.16 in which reactive power of PV system is shown in
case of short circuit in the system. The response of reactive power is very close for the both
models. As soon as the short circuit is removed, a reactive power is supplied by PV system in
order to maintain voltage at the LV terminal of PV system. It should be noted that in Fig 5.17,
the reactive power limits of the inverter are intentionally increased to see whether this model
is capable of regulating reactive power for voltage support. In this case reactive power limits
are increased up to 0.5 Mvar.

74

Short circuit
1.2
Generic Model
KTH Model
1
X: 0.966
Y: 0.9997

X: 1.225
Y: 0.9997

Uac(p.u)

0.8

0.6

0.4

0.2
X: 1.048
Y: 0

0
0.9

1.1

1.2
time(s)

1.3

1.4

1.5

Fig 5.16 AC Voltage at LV terminal

Short circuit
0.45
0.4

Generic Model
KTH Model

X: 1.101
Y: 0.4102

0.35

Qac(Mvar)

0.3
0.25
0.2
0.15
0.1
0.05

X: 1.008
Y: -0

0
-0.05
0.9

1.1

1.2
time(s)

1.3

Fig 5.17 Reactive power of PV system

75

1.4

1.5

Summary:
On the basis of above comparison results between the generic model and newly developed
KTH model, it can be concluded that overall response of the both models is quite similar. The
only difference is the different nature of modeling the PV array model. In this respect, PV
current and hence power is different. As both models are behaving in the same way with
similar reactive power control therefore this newly developed KTH model is validated by this
comparison.

5.3 Results Analysis of newly developed KTH model (Setup B)

After the validation of this developed model, detail results analysis of this model will be
presented. This KTH model is now equipped with following additional control systems in it.

MPPT control
Four types of reactive power controls
Active power control

The detailed features of this model have already been presented under the section 4.4.
Following simulations are based upon three different case studies:
Case 1: Decrease in irradiance of PV system from 1000 W/m2 to 500 W/m2 (Permanent
Disturbance)
Case 2: 10 % decrease in external grid voltage (Permanent Disturbance)
Case 3: Three phase short circuit at MV (A) Bus (for 100 ms)

These cases are analyzed and presented in the coming sections with different control
strategies. Note that now MPPT control is included in the system, so a time delay in the PV
array model is used in order to filter the unnecessary oscillations in the output. This delay
value is selected by considering a tradeoff between the oscillations in the output and to ensure
that model works for each case for this value. A delay value of T=0.05 second is selected for
this model and can be changed accordingly.

5.3.1 Case1: Irradiation change

For first case, solar irradiation is decreased from 1000 W/m2 to 500 W/m2. This decrease can
be realized as suddenly clouds appear and intensity of sunlight decreases. This disturbance is
made in order to check how this model behaves with sudden decrement in irradiation.

76

Fig 5.18 shows the change in irradiation in the form of a step decrease from 1000 W/m2 to
500 W/m2. The starting time of irradiation decrease is 15 sec and is a permanent disturbance.
Irradiation change
1000
X: 15
Y: 1000

E(W/m2)

900

800

700

600
X: 15
Y: 500

500
12

14

16

18

20
22
time(s)

24

26

28

30

Fig 5.18 Irradiance decrease from 1000W/m2 to 500 w/m2

DC voltage of the PV system responds to this irradiation change as shown in Fig 5.19. MPPT
is now in the system also a delay has been added to filter the high frequency oscillations.
Under these conditions, DC voltage decreases with decrease in irradiation and takes some
seconds to find its maximum power point and become steady. It can be seen from Fig 5.19
that DC voltage first give an overshoot in the start and quickly after that MPPT function start
working which causes the voltage to become steady value as determined by MPPT algorithm.
The value of DC voltage at 500 W/m2 irradiance is quite same as expected value, which
proves the validity of MPPT function.
Fig 5.20 shows the active power of the PV system with the sudden decrease in irradiation. It is
clear from Fig 5.20 that in steady state active power was 1 p.u. As disturbance applies in the
system at t=15, active power decrease and a small ripple is observed. Then very quickly
MPPT function works in order to find the point at which maximum possible power at 500
W/m2 irradiation can be obtained. This function track down MPP and a power of 0.477 p.u is
obtained which is the maximum power at this irradiation. Hence new model works well with
MPPT and regulate the DC voltage and active power with the change in irradiation of the PV
system.

77

Irradiation change
1

0.98

Udc(p.u)

0.96

X: 20.49
Y: 0.958

X: 15
Y: 0.9752

0.94

0.92

0.9

0.88
12

14

16

18

20
22
time(s)

24

26

28

30

28

30

Fig 5.19 DC voltage of PV system

Irradiation change
1.1

0.9

Pac(p.u)

0.8

0.7

0.6
X: 21.19
Y: 0.4771

0.5

0.4
12

14

16

18

20
22
time(s)

24

26

Fig 5.20 Active Power of PV system

In order to see how this new KTH model behaves with different reactive power controls,
following simulation are presented and results are discussed.

78

a. Unity power factor (Q Control 1)

In this control, PV system always operates at unity power factor. This works by initializing
reference value of reactive power Qref equal to zero. Based on this Qref, reactive power can
be controlled by using PI controller as already shown in Fig 4.20. Hence in this control
strategy, reactive power is controlled such that no reactive power is supplied to make it
possible to maintain unity power factor of the PV system.
Fig 5.21 shows that in case of irradiation decrease at t=15, reactive power is controlled to be
zero. As y-axis scale is very small and plot showing reactive power in Mvar, so reactive
power is almost zero in the below figure.
-5

Irradiation change

x 10

X: 15.11
Y: 3.5e-005

Qac(Mvar)

X: 22.85
Y: 0

0
-1
-2
-3
-4
X: 15.03
Y: -5.6e-005

-5
-6
12

14

16

18

20
22
time(s)

24

26

28

30

Fig 5.21 Reactive Power of PV system

In this type of reactive power control, AC voltage changes a bit which can be seen in Fig 5.22.
Therefore voltage is slightly changed with the change in irradiation of the system.

79

Irradiation change
1.05

X: 20.73
Y: 1.047

1.048

Vac(p.u)

1.046

1.044

1.042

1.04
X: 13.47
Y: 1.04

1.038
12

14

16

18

20
22
time(s)

24

26

28

30

Fig 5.22 AC voltage at LV terminal

b. Dynamic PF operation Q(P) (Q Control 2)

In dynamic PF operation, reactive power is regulated based on instantaneous value of active

power. Based on curve presented in Fig 2.14, reactive power varies depending upon active
power to keep the power factor within 0.9 inductive and capacitive.
Initially, according to PV system design it operates at 0.9 inductive power factor. Due to
decrease in active power, power factor from Fig 2.14 corresponding to this active power is
calculated i.e. 0.9954. Hence reactive power is regulated in order to maintain this power
factor in the PV system. Hence, reactive power is controlled on the basis of active power.
Fig 5.23 demonstrates the reactive power change on the basis of dynamic power factor
control. It can be seen in the figure that initially PV system was operating at 0.9 PF, it was
absorbing 0.2174 Mvar reactive power. When irradiation decreases, PV system reduces its
absorption of reactive power to 0.0205 Mvar in order to maintain a PF of 0.9954. Power
factor keep on changing with the change in the irradiation. That is why; this is called dynamic
power factor operation.

80

Irradiation change
0.1

X: 20.23
Y: 0.02057

0.05

Qac(Mvar)

-0.05

-0.1

-0.15

-0.2

-0.25
12

X: 13.8
Y: -0.2174

14

16

18

20
22
time(s)

24

26

28

30

Fig 5.23 Reactive Power of PV system

In this case, AC voltage changes a bit with change in irradiation as shown in Fig 5.24. Power
factor variation can be seen from Fig 5.25 which is evident that PF initially was 0.9 and
becomes 0.9954 as irradiation and active power decreases.

81

Irradiation change
1.08
X: 19.49
Y: 1.054

1.06

Vac(p.u)

1.04

1.02

0.98

0.96
12

X: 13.6
Y: 0.965

14

16

18

20
22
time(s)

24

26

28

30

28

30

Fig 5.24 AC voltage at LV terminal

Irradiation change
1.1
1.05

X: 20.98
Y: 0.9954

1
0.95

PF

0.9
0.85

X: 13.51
Y: 0.9

0.8
0.75
0.7
0.65
12

14

16

18

20
22
time(s)

24

Fig 5.25 Power Factor

82

26

c. Droop based reactive power Q(U) (QControl 3)

Droop based reactive power is designed to regulate reactive power based upon voltage
deviation at the regulating point.
Also this control is used to avoid unwanted injection of reactive power when the voltage
variation of the grid is within the dead band. This type of control is designed based on the
curve as shown in Fig 4.22. For the current system, dead band is defined as 0.05, which
means if the voltage variation would be within 5 % of the grid voltage, no reactive power will
be regulated. It is shown in Fig 5.26 that as irradiation changes, reactive power is almost
negligible and can be considered as zero.
-5

Irradiation change

x 10

X: 15.11
Y: 3.5e-005

Qac(Mvar)

X: 21.56
Y: 0

0
-1
-2
-3
-4
-5
-6
12

14

16

18

20
22
time(s)

24

26

28

30

Fig 5.26 Reactive Power of PV system

AC voltage variation can be seen in Fig 5.27, which proves that PV system should not inject
reactive power as the voltage variations in this figure is not more than dead band. This voltage
variation is not sufficient to initiate any reactive power regulation.

83

Irradiation change
1.05

X: 21.1
Y: 1.047

1.048

Vac(p.u)

1.046

1.044

1.042
X: 13.76
Y: 1.04

1.04

1.038
12

14

16

18

20
22
time(s)

24

26

28

30

Fig 5.27 AC voltage at LV terminal

d. AC voltage control (Q Control 4)

When reactive power control is based upon AC voltage regulation, the response of the AC
voltage and reactive power of the system is shown in Fig 5.28 and Fig 5.29 respectively. In
this case at steady state, system was operating at unity power factor and no reactive power
was supplied. As soon as irradiation in the system decreases AC voltage of the system starts
varying. Reactive power is regulated such that voltage of the system comes back to its initial
value.
Fig 5.28 shows AC voltage of the system with change in irradiation. It is clear from below
figure that as soon as irradiation decreases in the system, AC voltage control made it possible
to bring AC voltage back quickly to its initial value.
From Fig 5.29 it is clear that as the ac voltage tries to increase in the start with decrease in
irradiation, reactive power is absorbed from the grid in order to maintain a constant AC
voltage at LV terminal of PV system.

84

Irradiation change
1.0435
X: 15
Y: 1.043

1.043
1.0425

Uac(p.u)

1.042
1.0415
1.041
1.0405

X: 15.17
Y: 1.04

1.04
1.0395
1.039
14.9

14.95

15

15.05

15.1
time(s)

15.15

15.2

15.25

15.3

Fig 5.28 AC voltage at LV terminal

Irradiation change
0.005

Qac(Mvar)

-0.005

-0.01

-0.015

-0.02

X: 20.4
Y: -0.0233

-0.025

-0.03
12

14

16

18

20
22
time(s)

24

26

Fig 5.29Reactive power of PV system

85

28

30

In Fig 5.30 different reactive power control strategies with the change in irradiation are
compared. Reactive power for these different control strategies is plotted in the figure. The
Qcontrol 1 and Q control3 shows similar behavior as in both cases there is no regulation of
reactive power. The QControl1 is for unity power factor so there is no reactive power
supplied or absorbed. The Qcontrol 3 is based upon voltage variation at regulation point, as
voltage variation is within deadband so no reactive power is regulated. For Qcontrol 2,
absorption of reactive power is reduced on the basis of active power change in the system. In
Qcontrol 4, reactive power is absorbed from the grid in order to make AC voltage constant at
the regulation point.
Irradiation change
0.1

X: 18.33
Y: 0.02067

0.05

X: 23.43
Y: 0

Qac(Mvar)

0
X: 20.4
Y: -0.0233

-0.05

QCtrl1
QCtrl2
QCtrl3
QCtrl4

-0.1

-0.15

-0.2

-0.25
12

X: 13.68
Y: -0.2174

14

16

18

20
22
time(s)

24

26

28

30

Fig 5.30Reactive power with different Q Control Strategies

5.3.2 Case 2: Decrease in external grid voltage by 10%

Next case is to check how KTH model works in case of external grid voltage decrease by
10%. Obviously this is a big voltage drop in the system, so a lot of reactive power needs to be
regulated in order to maintain the voltage at LV terminal of PV system. This case is to check
the capability of newly developed KTH model in case of such a huge disturbance in the
system. Fig 5.31 shows the external grid voltage decrease by 10%, which is used a
disturbance in this case.

86

External grid voltage decrease by 10%

1
X: 13.68
Y: 1

Vac(p.u)

0.98

0.96

0.94

0.92

0.9
12

X: 20.44
Y: 0.9091

14

16

18

20
22
time(s)

24

26

28

30

Fig 5.31 External grid voltage decrease

Under normal conditions, reactive power operational limits are defined and PV system cannot
supply reactive power more than its limits. But for this case reactive power limits are
increased as much as to supply enough reactive power to compensate such a big voltage drop.
AC voltage regulation (Q Control 4) is used in this case to show, how this model act with this
disturbance. External grid voltage is decreased by 10% at t=15.
DC voltage of the PV system is shown in Fig 5.32, in which it can be observed that at t=15
second when grid voltage is decreased, its DC voltage increases in start but within a second it
is maintained at its nominal value.
Active power decreases to 0.94 p.u. when disturbance is applied in the system and very
quickly like DC voltage it comes back to its steady state value. Hence model works perfect
with respect to active power with change in grid voltage. Active power of PV system is shown
in Fig 5.33.

87

External grid voltage decrease by 10%

1.01
1.008

X: 15
Y: 1.009

1.006
1.004

Vdc(p.u)

1.002
X: 15.18
Y: 0.9994

1
0.998
0.996
0.994
0.992
0.99
14.95

15

15.05

15.1
time(s)

15.15

15.2

15.25

Fig 5.32 DC voltage of PV system

External grid voltage decrease by 10%
1.04

1.02
X: 15.12
Y: 1

Pac(p.u)

1
X: 14.98
Y: 1

0.98

0.96
X: 15
Y: 0.9334

0.94

0.92
14.95

15

15.05
time(s)

15.1

Fig 5.33 Active power of PV system

88

15.15

When external voltage decreases by 10% AC voltage at LV terminal of PV system also

decreases without any AC voltage control. But when AC voltage control (Q control4) is there
in the system, it helps to maintain a constant AC voltage at the terminal of PV system.
Fig 5.34 shows voltage decreases due to external grid voltage decrease, but due to AC voltage
control it quickly comes back to its original value in less than a second. It is very fast control
that helps to maintain a constant voltage at regulation point even with 10% grid voltage
decrease.
External grid voltage decrease by 10%
1.06

1.04
X: 14.97
Y: 1.04

X: 15.09
Y: 1.04

Vac(p.u)

1.02

0.98
X: 15
Y: 0.9561

0.96

0.94
14.95

15

15.05
time(s)

15.1

15.15

Fig 5.34 AC voltage at LV terminal

It is obvious from Fig 5.35 that PV system injects a lot of reactive power to external grid, in
order to bring the voltage at LV terminal back to desired value. It injects a reactive power of
about 0.3 Mvar to compensate decrease of external grid voltage. It shows that KTH model is
capable of providing reactive power in order to maintain a constant voltage at LV terminal of
PV system and to ensure it works in any disturbance of voltage.

89

External grid voltage decrease by 10%

0.4
0.35

X: 21.49
Y: 0.3075

0.3

Qac(Mvar)

0.25
0.2
0.15
0.1
0.05
0
-0.05
12

14

16

18

20
22
time(s)

24

26

Fig 5.35 Reactive power of PV system

90

28

30

5.3.3 Case 3: Short circuit fault at MV (A) bus bar

A three phase short circuit at MV (A) bus bar is created at t=15 and is removed at t=15.1. The
system response in this disturbance is very important to analyze. As MPPT control is also
activated, so its interaction with the severe disturbance can also be a point of interest in this
analysis. AC voltage regulation (Q Control 4) is used in this case to show, how this model act
with this disturbance.
Fig 5.36 shows the DC voltage response of the PV system with the short circuit at bus
MV(A). When short circuit is applied and quickly removed, DC voltage first increases to
open circuit voltage of the system and then starts decreasing. But in this case, due to MPPT
control a time delay in the system of T=0.05 was used to filter high frequency oscillations in
the output. This causes the system to take some time to reach the DC voltage, active power
and reactive power after the removal of short circuit fault. This can be seen in Fig 5.36 in case
of DC voltage of the PV system. It takes some time to attain a steady state value after the
disturbance is removed.
Short circuit fault at MV(A) bus
1.3

1.25
X: 15.18
Y: 1.251

Vdc(p.u)

1.2

1.15

1.1

1.05
X: 77.62
Y: 1

0.95
20

30

40

50

60
time(s)

70

80

90

100

Fig 5.36 DC voltage of PV system

Active power of the system as can be observed in Fig 5.37 reduces to zero in case of short
circuit fault duration. When fault is removed, it also takes some time to achieve the steady
state value of active power.

91

Short circuit fault at MV(A) bus

1.4
1.2

X: 15.1
Y: 1.368
X: 75.97
Y: 1

Pac(p.u)

0.8
0.6
0.4
0.2

X: 15
Y: 0

0
-0.2

20

30

40

50

60
time(s)

70

80

90

100

90

100

Fig 5.37 Active power of PV system

Short circuit fault at MV(A) bus

X: 15.1
Y: 2.308

Qac(Mvar)

1.5

0.5
X: 68.47
Y: -9.9e-005

0
20

30

40

50

60
time(s)

70

80

Fig 5.38 Reactive power of PV system

92

PV system injects the reactive power to the grid after the fault is removed, in order to bring
AC voltage at LV terminal back to its original steady state value. Reactive power is shown in
Fig 5.38 whereas AC voltage at LV terminal is presented in Fig 5.39.
AC voltage becomes zero during the short circuit fault, and as soon as fault is removed it
return back to its initial steady state value. Due to AC voltage regulation control, PV system
made it possible to very quickly regulate its AC voltage as shown in Fig 5.39.
Short circuit fault at MV(A) bus
1.6
1.4
X: 14.83
Y: 1.04

1.2

X: 15.28
Y: 1.04

Vac(p.u)

1
0.8
0.6
0.4
0.2

X: 15.06
Y: 0

0
-0.2
14.7

14.8

14.9

15

15.1
time(s)

15.2

15.3

Fig 5.39 AC voltage at LV terminal

93

15.4

15.5

5.3.4 Active power Curtailment

Active power control is another type of unique control which is new addition in the control
system of a KTH model. This control can be realized, when PV system is operating on its
maximum power point and there arises a need to reduce the active power coming out of the
PV system. This can be due to need of AC grid if for some reason; 100% active power is not
required. But active power produced by PV array remains constant at a particular weather
condition (irradiation, temperature and shading). Under these circumstances, active power
control is developed and is implemented in this new model.
This control is implemented by providing a reference active power Pref value as a parameter
and a PI controller is used to calculate the corresponding Vdcref for the main controller. For
this case parameter event Act at t=15 is created which set Act=1 and give the Pref value.
Hence at t=15 active power control starts operating.
Fig 5.40 shows the DC voltage decrease when active power of PV system is decreased by
setting the value of Pref= 85% P i.e. it mean value of Pref equals to 15 % decrease of nominal
value of active power.
Active power control
1.1

1.05
X: 13.92
Y: 1

Vdc(p.u)

0.95

0.9

0.85
X: 17.92
Y: 0.7847

0.8

0.75
12

14

16

18

20

22

24

time(s)

Fig 5.40 DC voltage of PV System with 15% Pref decrease

94

26

Active power control

1.04
1.02
1
X: 14.95
Y: 1

0.98

Pac(p.u)

0.96
0.94
0.92
0.9
0.88
X: 18.21
Y: 0.85

0.86
0.84
12

14

16

18

20

22

24

26

time(s)

Fig 5.41 Active power of PV System with 15% Pref decrease

Fig 5.41 demonstrates the active power of PV system when active power control is enabled. It
is clear that when Pref is reduced 15% of the nominal power, active power of PV system
starts following its reference value which results in a decrease of active power of the system
by 15 %.
In the end function of active power control with different percentage reduction of Pref value is
presented. Fig 5.42 to Fig 5.44 shows active power control operation with 5%, 10% and 15 %
decrease in Pref value respectively. PI controller used in this control is tuned so that active
power quickly follows the reference value of active power.

95

Active power control with 5% Pref decrease

1.02
Pref
Pac

1.01
X: 14.98
Y: 1

P(p.u)

0.99
0.98
0.97
0.96
X: 18.66
Y: 0.95

0.95
0.94
12

14

16

18

20

22

24

26

time(s)

Fig 5.42 Active power control with 5% Pref decrease

Active power control with 10% Pref decrease
1.02
Pref
Pac

1
X: 15
Y: 1.002

0.98

P(p.u)

0.96

0.94

0.92
X: 18.18
Y: 0.9

0.9

0.88
12

14

16

18

20

22

24

time(s)

Fig 5.43 Active power control with 10% Pref decrease

96

26

Active power control with 15% Pref decrease

1.02
1

Pref
Pac

X: 15
Y: 1.003

0.98
0.96

P(p.u)

0.94
0.92
0.9
0.88
X: 19.18
Y: 0.85

0.86
0.84
12

14

16

18

20

22

24

time(s)

Fig 5.44 Active power control with 15% Pref decrease

97

26

5.3.5 Active power control with system frequency

This active power control is implemented by manually changing the system frequency and
analyzing its impact on active power of the system. System frequency is manually increased
from 50 Hz to 51 Hz by creating a parameter event at t=15 and is changed back to its original
value at t=150. Frequency change of the system is shown in Fig 5.45.
Active power control based upon frequency
51.2

51
X: 45.51
Y: 51

Frequency (Hz)

50.8

50.6

50.4

50.2
X: 15
Y: 50

X: 223.5
Y: 50

50

49.8

50

100

150
time(s)

200

250

Fig 5.45 System frequency

According to the regulation presented in Fig 2.13, change in power with 1 Hz change in
frequency is calculated as 0.143 MW and can be seen in Fig 5.46. Active power decrease to
the value 0.143 MW and as soon as frequency comes back to 50 Hz, it starts increasing and
comes back to its steady state value at t=260.

98

Active power control based upon frequency

0.6
0.55
0.5

X: 7.629
Y: 0.4488

X: 260.8
Y: 0.4488

Active power (MW)

0.45
0.4
0.35
0.3
0.25
0.2

X: 150
Y: 0.1403

0.15
0.1

50

100

150
time(s)

200

250

Fig 5.46 System frequency

This type of control takes some time to reach to steady state value after the disturbances. This
can be explained on the fact that first of all MPPT is there in the system also due to active
power PI controller time constant it takes some time to reach steady state. This time constant
value has already been set after many tradeoffs and that is why same value is used.

5.3.6 Validation of MPPT control

First solar cell equations are written in MATLAB and a plot with changing irradiation is
plotted and is shown in Fig 5.47. As it can be seen in this figure that for each curve there is a
point at which power is maximum and is called MPP of the curve as indicated by data dips.
MPP gives approximately 700 V for 1000 W/m2 and 691V for 800 W/m2.

99

4.5

Irradiation decrease from 1000 to 200 in 5 steps

x 10

X: 700.8
Y: 4.488e+005

E=200
E=400
E=600
E=800
E=1000

4
3.5

X: 691.4
Y: 3.542e+005

Ppv (W)

X: 679.6
Y: 2.605e+005

2.5
X: 653.6
Y: 1.683e+005

2
1.5

X: 623.1
Y: 7.922e+004

1
0.5
0

100

200

300

400
500
Vpv (V)

600

700

800

900

Fig 5.47 PV characteristics

In order to validate the MPPT control, this plot from MATLAB based upon static equations is
compared with the dynamic behavior of MPPT control in PowerFactory. Fig 5.48(a) shows
how the PV characteristics changes when irradiation changes from 1000 W/m2 to 800 W/m2
in dynamics. In this figure a dynamic response from PowerFactory is plotted with these two
static curves. Here, green and blue are two static PV characteristics obtained from MATLAB
equations for 1000 W/m2 and 800 W/m2 irradiation respectively. The red curve shows how
PV characteristics changes in power factory when irradiation changes from 1000 W/m2 to 800
W/m2.

100

Irradiation decrease from 1000 to 800

x 10

4.5
4
3.5

P pv (W )

3
2.5
2
E=800
E=1000
PV graph (power factory)

1.5
1
0.5
0

100

200

300

400
500
Vpv (V)

600

700

800

900

Fig 5.48(a) PV characteristics with power factory

In order to analyze it in more details the Fig 5.48(b) shows the same response by having a
closer look of Fig 5.48(a).
5

Irradiation decrease from 1000 to 800

x 10
4.6

X: 700
Y: 4.488e+005

4.4

4.2

P pv (W)

4
E=800
E=1000
PV graph (power factory)

3.8
X: 691
Y: 3.536e+005

3.6

3.4

3.2
640

660

680

700
Vpv (V)

720

740

760

Fig 5.48(b) PV characteristics with power factory (closer look)

101

Fig 5.48 (b) shows that PV characteristics start from 1000 W/m2 and as MPPT is there in the
system, it takes a little while to track to its maximum power point for this irradiation (green
curve). This is why there are some oscillations in the start. As soon as irradiation decreases to
800 W/m2, active power decreases and now characteristics come to (blue curve). Now MPPT
again tracks for this curve and after some minor oscillations, it finds the MPP for 800 W/m2
irradiation.
The main purpose of showing this comparison between PowerFactory and MATLAB
equations is to show that MPPT control works fine. It is also clear that this control works
perfect even in case of dynamic behavior of MPPT, as red curve exactly starts with and ends
to the same point as was expected.
5.4 Comparison between KTH model (static generator) and PWM converter
model (Setup C)
In this section a comparison is presented between KTH model having the static generator and
a PWM converter based PV system model.
For this comparison following conditions are brought into consideration

There is no MPPT control in the system

Both models have same reactive power control (based on AC voltage-Q Control 4)

In this comparison only Case 1 is investigated:

Case 1: Irradiance change in PV system (Irradiation is decreased to 500W/m2 at t=0.3 and
increased back to 1000w/m2 at t=0.7)
In Fig 4.2 the basic structure of both models has already been shown. The brief description of
two models is presented as
1. KTH Model based on static generator having DC bus bar model
In this model the PV system is represented by a static generator. The DC bus bar model used
in this PV model calculates the DC voltage instead of directly measuring it.
2. Model based on PWM converter having no DC bus bar model
In this model a current source is used to represent the PV system with a DC bus bar
connecting to it as shown in Fig 4.2. The DC voltage is measured from the voltage measuring
device at the DC bus bar. This model is presented and explained in details in a master thesis
project named Comparison of a three phase single stage PV system in PSCAD and
PowerFactory.
102

The DC voltage of the two models is almost similar to each other in case of irradiation change
in the system. In PWM converter, there is a bit more oscillations as compared to static
generator. But overall response is quite similar. The DC voltage for the both models is shown
as follows
Irradiation change
800
PWM Converter
Static Generator

780
760
740

Udc(V)

720

X: 0.121
Y: 700

X: 0.5012
Y: 699.9

700
680
660
640

X: 0.3092
Y: 630.6

X: 0.3052
Y: 619

620
600

0.1

0.2

0.3

0.4

0.5
time(s)

0.6

0.7

0.8

0.9

Fig.5.49 DC voltage of PV system

The active power of the two models is shown in Fig.5.50. It is clear from the figure that both
models exhibit the similar response. There is only a minute difference in the form of a small
overshoot, but this can be ignored as overall behavior is almost the same.

103

Irradiation change
0.55
PWM converter
Static Generator

0.5
X: 0.123
Y: 0.4488

X: 0.9132
Y: 0.4485

0.45

Pac(MW)

0.4
0.35
0.3
0.25

X: 0.5222
Y: 0.2111

0.2

0.1

0.2

0.3

0.4

0.5
time(s)

0.6

0.7

0.8

0.9

Fig.5.50 Active power of PV system

The AC voltage of the two models is shown in Fig 5.51. It is clear from the data dip values in
the figure that AC voltage at the LV terminal in case of irradiation change is almost identical.
This response is based on Q control 4 in which AC voltage is controlled to a constant value. It
can be seen from Fig 5.51 that the irradiation decreases and then increases; AC voltage comes
back very fast to its controlled value for these two changes in both models. In case of PWM
converter there is bit more oscillations in the AC voltage as compared to static generator.
In Fig 5.52, a zoomed in view of Fig 5.51 is presented. In Fig 5.52-(a) a zoomed in view of
left peak in Fig 5.51 (peak due to first decrease in the irradiation of the system) is shown. It
can be seen that in case of static generator, voltage comes back very quickly to its controlled
value. Whereas in case of PWM converter it takes a bit longer time to reach to steady state
value. Similarly in the Fig 5.52(b) a zoomed in view of second peak of Fig 5.51 is presented.
This figure shows that AC voltage reaches its steady state value earlier for static generator
than PWM converter. There are also some oscillations in the AC voltage in case of PWM
converter, but static generator does not give much oscillations.

104

Irradiation change
1.045
PWM Converter
Static Generator

1.044
1.043

X: 0.3042
Y: 1.044

X: 0.3022
Y: 1.043

Vac(p.u)

1.042
1.041

X: 0.18
Y: 1.04

X: 0.5262
Y: 1.04

1.04
1.039
1.038
1.037
1.036
0.1

0.2

0.3

0.4

0.5
0.6
time(s)

0.7

0.8

0.9

Fig.5.51 AC voltage at LV terminal

Irradiation change

Irradiation change

1.045

1.042

1.044

1.041

1.043

1.04
Vac(p.u)

Vac(p.u)

PWM Converter
Static Generator

1.042

1.039

1.041

1.038

1.04

1.037

1.039

0.3

0.32 0.34
time(s)

0.36

1.036
0.65 0.7

PWM Converter
Static Generator

0.75 0.8
time(s)

0.85

Fig.5.52 (a) Zoomed in view of left peak of Fig 5.51 Fig.5.52 (b) Zoomed in view of right peak of Fig 5.51

105

Fig 5.53 shows the reactive power response of the two models with irradiation change in the
system. It should be noted that at t=0.3 when irradiation decreases, reactive power is absorbed
from the grid in order to stabilize the AC voltage to a constant value. In this case response of
the both models is quite similar. But at t=0.7 when the irradiation increases again both models
shows a bit different behavior. Reactive power from the PV system is released to the grid. In
case of static generator reactive power reaches to a steady state value without oscillations. On
the other hand, with PWM converter model, some oscillations can be observed in the reactive
power but finally it reaches to its steady state value.
Irradiation change
0.015
PWM Converter
Static Generator

0.01
0.005

X: 0.151
Y: 2.8e-005

Qac(Mvar)

0
-0.005
-0.01
-0.015
X: 0.4942
Y: -0.02297

-0.02
-0.025
-0.03

0.1

0.2

0.3

0.4

0.5
time(s)

0.6

0.7

0.8

Fig.5.53 Reactive power of the PV system

106

0.9

6. Conclusions and Future Work

This project focuses on the improvement of PV model in PowerFactory. In this respect, a new
model named KTH model in PowerFactory is developed which is equipped with many
controls system. The following conclusions are drawn from this project.
The developed KTH model in PowerFactory is compared in details with already available
generic model in PowerFactory using the same reactive power control for the both models.
For all types of disturbances, both models behave in a similar way, which validate the
developed KTH model in PowerFactory.
In generic model, PV panel is modeled with unknown approximations and parameters.
Therefore, in KTH model, a more realistic PV panel model is developed which is based on the
solar cell equations. This panel model gives power output which is more close to expected
one as compared to that of generic model.
As far as control systems are concerned, there were some basic controls available in generic
model, but still there is a need of many other controls. Therefore in KTH model, the following
new control systems are implemented as compared to generic model i.e. active power
curtailment, MPPT control and four different types of reactive power controls. These control
systems in a PV model are very essential in order to perform the effective grid studies,
making power system more reliable and efficient.
There was no MPPT control in the generic model; therefore in KTH model MPPT control is
implemented to increase the efficiency of the system. Because MPPT ensures the PV panel to
always operates on maximum power point in case of any change in irradiation. This MPPT
control causes some high frequency oscillations in the system. But at the same time it gives
more power at the output which increases the systems efficiency. A time delay in the PV
panel model is used to filter these high frequency oscillations. Two methods of MPPT control
are implemented in this project Perturb and Observe and Incremental Conductance. The
comparison of these two methods shows that Incremental Conductance provides better results
than Perturb and Observe. This is because Incremental Conductance does not give oscillations
around MPP.
Reactive power support is very important for a PV system according to the grid codes. In
generic model there is only one type of reactive power control present. But in KTH model
four different types of reactive power controls i.e. unity power factor, dynamic power factor,
droop control and AC voltage regulation are implemented. It provides flexibility in the KTH
model to use any one of the available reactive power control strategies, depending upon the
requirement of the system.

107

Active power curtailment implemented in the KTH model provides a facility to reduce the
active power coming out of PV generator according to grid demand. This control was not
present in generic model.
In Setup B, various transient grid studies are performed in KTH model having all the
implemented controls. In this respect, it is concluded that KTH model in PowerFactory works
perfect in case of severe short circuit fault with any kind of control for reactive power. Hence
it provides a facility to perform short circuit analysis by creating fault anywhere in the system.
In case of external grid voltage change, KTH model is capable of providing reactive power to
the grid or absorbing reactive power from the grid. Hence, it increases the voltage stability in
the system. Also this model behaves well in case of any solar irradiation change in the system.
Especially with MPPT control we can have maximum possible power for any value of the
irradiation in system.
In Setup C, a comparison is carried out between KTH model with static generator and PWM
converter based PV model. It is concluded that both models behave in a similar way for the
change in irradiation in the system. In case of PWM converter based model, more oscillations
are observed as compared to KTH model based on static generator.
Future work
After the development of KTH model there can be many analyses which could be performed
using this model in PowerFactory. And still there can be many more additions or
improvement in this model possible.
Suggested possible future works in this respect can be as

This KTH PowerFactory model can be used to perform grid studies with two or more PV
system connected in parallel with each other. It would be interesting to analyze the
interaction between these parallel PV systems.

A more detailed solar irradiation model can be developed by taking into consideration the
following advanced aspects i.e. cloud modeling, different shading effects and angle of
incidence of sun light.

This PV model can be further investigated in order to comply with different grid codes
introduced by various countries.

Further grid studies can be performed from this developed model by having AC
generators and different loads in the system instead of Infinite bus.

108

7. References
[1]
Renewables
2012
Global
Status
Report
available
at
http://www.ren21.net/REN21Activities/Publications/GlobalStatusReport/GSR2012/tabid/792
18/Default.aspx as seen on 14-10-2012.
[2] European Photovoltaic Industry Association: Global Market Outlook for Photovoltaic
until 2016 available at:
http://www.epia.org/publications/epiapublications/globalmarketoutlookforphotovoltaicsuntil2
016.html as seen on 14-10-2012.
[3] Volker Quaschning,Understanding renewable energy systems, first published by Earth
scan in the USA and UK in 2005.
[4] Dezso Sera, Remus Teodorescu, Pedro Rodriguez, PV panel model based on data sheet
values, In Proc. IEEE International Symposium on Industrial Electronics ISIE 07, June 47,2007,pp.2392-2396.
[5] TransmissionCode 2007: Network and System Rules of the German Transmission System
Operators.
[6] E.Troester,New German grid codes for connecting PV systems to the medium voltage
power grids, 2nd International Workshop on Concentrating Photovoltaic Power Plants:
Optical Design, Production and Grid Connections.
[7] Recommendations for unified technical regulations for grid-connected PV systems 2009:
SUNRISE project available at:
http://www.pvsunrise.eu/fileadmin/PVS_docs/images/Recommendations_for_unified_technic
al_regulations_for_grid-connected_PV_systems__Public_.pdf as seen on 14-10-2012.
[8] Teodorescu R., Liserre M., Rodriguez P., Grid Converters for Photovoltaic and Wind
Power Systems, Wiley, United Kingdom, 2011.
[9] Fritz Schimpf, Lars E .Norum: Grid connected Converters for Photovoltaic, State of the
Art, Ideas for Improvement of Transformerless Inverter, NORPIE/2008, Nordic Workshop on
Power and Industrial Electronics, June 9-11, 2008.
[10] Nobert henze, Afred Engler, benjmen sahan performance of a novel three phase
photovoltaic inverter for module integration, 22nd Photovoltaic solar energy conference and
exhibition, 3-7 septempember, Milan Italy.

109

[11] Marion B., Validation of a Photovoltaic Module Energy Ratings Procedure at NREL,
NCPV Program Review Meeting 2000, Denver, Colorado, 16-19.04.2000.
[12] Andreas Wagner , Photovoltaik Engineering Handbuch fr Planung, Entwicklung
und Anwendung, Fachbereich Informations- und Elektrotechnik Sonnenstrae 96 44139
Dortmund.
[13] Standard Test Conditions (STC) in the Photovoltaic (PV) Industry IMT Solar: North
America 814 Wurlitzer Drive North Tonawanda, NY 14120 available at
http://www.imtsolar.com/public/files/IMT%20Solar_STC%20for%20PV%20APP%20NOTE.
pdf as seen on 21-08-2012.
[14] Vladimir Cuk, Paulo F.Ribeiro, joseph F.G.Cobben, Wil L.Kling, Fridirk R. Isleifsson,
henrik W.Bindner, Nis Martensen, Afshin Samadi, and Lennart Soder Considerations on the
Modeling of Photovoltaic Systems for Grid Impact Studies, 1st International Workshop on
Integration of Solar Power into Power systems, Aarhus, Denmark, 24 October, 2011.
[15] Santa Clara, ALTA DEVICES solar panel receives NREL verification of 23.5%
efficiency Press Release, February 6, 2012,available at
http://www.roeder-johnson.com/RJDocs/ALnrelverification0212.html as seen on 21-08-2012.

[16] Carl Franzen, Highest Solar Panel Efficiency Achieved, TPMIdeaLab, FEBRUARY 7,
2012, available at
http://idealab.talkingpointsmemo.com/2012/02/highest-solar-panelefficiency-achieved.php as seen on 15-09-2012.
[17]A.Yazdani, A.R.D.Fazio, H.Ghoddami, M. Russo, M. Kazerani, J.Jatskevich, K.Strunz, S
.Leva, J.A.Martinez, Modeling Guidelines and a Benchmark for Power System Simulation
Studies of Three-Phase Single-Stage Photovoltaic Systems,IEEE Transactions on Power
Delivery, Vol.26,No.2, April 2011.
[18] T. Esram and P.L. Chapman, "Comparison of photovoltaic array maximum power point
tracking techniques," IEEE Trans. Energy Conversion, Vol. 22, No. 2, pp. 439-449, Jun. 2007
[19] Wenkai Wu, N. Pongratananukul , Weihong. Qiu, k. Rustom, T. Kasparis, and I.
Bataresh, DSP based Multiple Peak Power Tracking for Expandable Power System, 2003
IEEE.
[20] Dezso Sera,Maximum Power Point Tracking (MPPT) for PV systems, Aalborg
University ,[email protected], www.et.aau.dk.
[21]W. Xiao and W. G. Dunford, A modified adaptive hill climbing MPPT method for
photovoltaic power systems, in Proc. 35th Annu. IEEE Power Electron. Spec. Conf., 2004,
pp. 19571963.

110

[22] M. Calavia1, J.M. Peri1, J.F. Sanz2 and J. Salln2 Comparison of MPPT strategies for
solar modules, International Conference on Renewable Energies and Power Quality
(ICREPQ10) Granada (Spain), 23th to 25th March, 2010.
[23] Nattorn Pongratananukul,Analysis and Simulation Tools for Solar Array Power
Systems, Doctoral dissertation, Department of Electrical and Computer Engineering at the
University of Central Florida Orlando, Florida.
[24] S.Weigel, Static Generator, DigSilent Technical Documentation, PF Version 14.0.516,
published by DigSilent GmbH Germany, 15-04- 2010.

111

8. Appendix
8.1 Parameters used for the new PV model
Table 8.1: System Parameters

PV system components

Value

Static Generator

0.5 MVA

DC side voltage of PV

0.7 kV

LV terminal AC voltage

0.4 kV

MV(A) terminal AC voltage

10 kV

MV(B )terminal AC voltage

10 kV
Transformer

Transformer, Nominal Power

0.63 MVA

Transformer, Voltage Ratio

0.4kV/10kV

Transformer, Vector Group

Dyn11

Transformer, Impedance

4%
Transmission Lines

Line 1, Total length

10km

Line 1, Resistance per km

Line 1, Reactance per km

Line 1, Impedance per km

Line 2, Total length

10km

Line 2, Resistance per km

Line 2, Reactance per km

Line 2, Impedance per km

112

Table 8.2: PV module Parameters

Parameters

Value

Time constant, T

0.05 sec

Short circuit current per module, Isc1

5 A

Open circuit voltage per module, Voc1

43.8 V

Current at MPP , Impp0

4.58 A

Voltage at MPP, Vmpp0

35 V

Tref

25 C

K I

0.0004

KV

0.0039

ESTC, Irradiance at STC

1000 W/m2

nSerialModules Number

20

nParallelModules Number

140

Table 8.3: DC Bus bar and Capacitor

DC Bus bar and Capacitor

parameters

Value

Capacitance

0.0172

Initial DC voltage

700 V

Nominal DC voltage

1 p.u

Rated Power

0.5 MVA

113

Table 8.4: MPPT Parameters

Track (MPPT module) Block

parameters

Value

Tdelay1

0.045 s

Vbase

Tdelay

0.05 s

Vmpp0_array

700 V

Step

125

Epsilon

0.01

Table 8.5: Active Power Reduction Parameters

Active Power Redn Block

parameters

Value

fUp

50.2Hz

fLow

50.05Hz

Gradient

0.05

Pbase

0.44884 MW

Tfilter

0.01 s

Table 8.7: Qref Block Parameters

Qref Block parameters

Value

PFlim

0.9

Ubase

0.4 kV

Pbase

0.44884 MW

PF_select:0=const_pf;1=dynamic pf;2=Q(U)

0/1/2

db

0.05

114

Table 8.8: Controller Parameters

Controller Block parameters

Value

Tfac

0.002

Tqfac1

0.001

ibase

1.02

ibase_iq

Tfqac

Kdc

0.005

Kqac

Tqac

0.041

Kac

0.5

Tac

0.0002131

Tfdc

Tdc

0.015

Flag 0=Reactivepower:1=Vac;2=Direct_iqref

0/1/2

id_min

iq_min

id_max

iq_max

115

Table 8.9: Active Power Control Parameters

Active power control Block

parameters

Value

Act 0=MPPT;1=InputPref;2=Constant VDCref

0/1/2

Pfreq_fun ; 0=off ; on=1

0/1

Tfdc

0.001 s

Flow

47.5 Hz

Fup

50.2 Hz

Vconst

700 V

Kp

Tp

0.0005

8.2 DSL codes inside the PV system blocks

Vt1=(2*Vmpp0-Voc1)*(Isc1-Impp0)/(Impp0+(Isc1-Impp0)*ln((Isc1-Impp0)/Isc1))
Rs1=(Vt1*ln((Isc1-Impp0)/Isc1)+Voc1-Vmpp0)/Impp0
Pmpp0=Impp0*Vmpp0
Isc2=Isc1*(1+(KI/100)*(thetaout-Tref))
Isc=Isc2*Eout/Estc
Iph=Isc;
Voc=Voc1+(KV*(thetaout-Tref))
! Voc=log(Iph/I0)*Vt1
Voc2=Voc*(log(Eout)/log(Estc))
I0=(Isc2/exp(Voc/Vt1))
! I=Iph-I0*exp(V/Vt1);
FI0=Istart-Iph+I0*(exp((V+Istart*Rs1)/Vt1)-1);
FpI0=1+I0*(exp((V+Istart*Rs1)/Vt1))*Rs1/Vt1;
I1=Istart-FI0/FpI0;
FI1=I1-Iph+I0*(exp((V+I1*Rs1)/Vt1)-1);
116

FpI1=1+I0*(exp((V+I1*Rs1)/Vt1))*Rs1/Vt1;
I2=I1-FI1/FpI1;
FI2=I2-Iph+I0*(exp((V+I2*Rs1)/Vt1)-1);
FpI2=1+I0*(exp((V+I2*Rs1)/Vt1))*Rs1/Vt1;
I3=I2-FI2/FpI2;
FI3=I3-Iph+I0*(exp((V+I3*Rs1)/Vt1)-1);
FpI3=1+I0*(exp((V+I3*Rs1)/Vt1))*Rs1/Vt1;
I4=I3-FI3/FpI3;
! I=I4;
! V1=V ;
I=lim(I4,0,5)
V1=lim(V,0,43.8)
P=V1*I
Parray=Varray*Iarray

8.3 MATLAB code to determine the controller parameters

clc;
clear all
close all
f=50;
phase=0
W=2*pi*f;
%system date
%%%%%%%based on the ... paper
Vmpp=0.700;
%************ Give Total Value
Impp=0.6412;
%************ Give Total Value
%PV output rating power
PpouPV=Vmpp*Impp%*1e6%MW
%############# PV output power
Ubase=0.4; %************ Give AC side rated line-line voltage
Vs=Ubase*sqrt(2/3)
pf=0.90;
%********** Give Power factor
ReactPower=PpouPV*tan(acos(pf))
Sbase=sqrt((PpouPV^2)+(ReactPower^2))
%%%%%%%%%%%%%%%%Inverter power
rating
Ppu=PpouPV/Sbase;
Zbase=(Ubase^2)/Sbase;
Udc=Vmpp;
%%%%%%%%%%%%%%% DC voltage
Zbasedc=(Udc^2/Sbase)
Ibase=Sbase/(sqrt(3)*Ubase);
Rbase=Udc/Impp;
Uphase=Ubase/sqrt(3);
%Vsd
Vspu=Vs/Uphase;
%Vs= 0.147%e3
L=0.1e-3;%*********
Inductance of the Series Recator connected with the
inverter
Xl=2*pi*50*L;
Lpu=Xl/Zbase;

117

R=3e-3;% was 1e-3

%******** Resistance of the Series Recator connected
with the inverter+Switching losses
Rpu=R/Zbase;
toi=.5e-3
toipu=.5e-3*W
%Dc Bus capacitor
C=17200e-6 %*********** DC side capacitor
Xc=(1/(W*C));
Cpu=Xc/Zbasedc;
%filter
Cf=10e-6;
%*********** Filter Capacitor if required
Xcf=-(1/(2*pi*50*Cf));
Cfpu=Xcf/Zbase;
%Refrtence
Iqref=0;
Idref=0;
Pref=PpouPV/2 %MW
Qref=0
%line
Lg=0.0424*(.4/10)^2
transformer rating
Xlg=2*pi*50*Lg
Lgpu=Xlg/Zbase;

%********* Line inductance note here the

Rg=0.06275*(.4/10)^2
%********** Line Reactance note here the
transformer rating
Rgpu=Rg/Zbase;
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%limits
DcBus_UpperLimit=Pref;
DcBus_LowerLimit=-Pref;
AcBus_UpperLimit=Pref;
AcBus_LowerLimit=-Pref;
% PI controller data
Kp=Lpu/toipu
%########### Gain to be given for the Inverter INBUILT controller
Ki=Rpu/toipu
%########### Timeconstant=1/Ki to be given for the
Inverter IN-BUILT controller

% how to calculate Vdc DC bus voltage

Vsd=Vs;
DeltaP=1*Pref/2;
Idref1=2/3*Pref/Vsd
DeltaQ=0;
Qs0=0;
Ps0=0;
Vtd= Vsd+(2*L*W/3/Vsd)*Qs0+(2*L/3/Vsd/toi)*DeltaP
Vtq= (2*L*W/3/Vsd)*Ps0-(2*L/3/Vsd/toi)*DeltaQ
Vt=sqrt(Vtd^2+Vtq^2)
Vdc=2*Vt
% Closed loof function of current controller loop
s=tf('s')
Gcl_current=1/(toi*s+1);
% figure, margin(Gcl_current)

118

%How to design controller for DC-bus Voltage regulation

WcV=25 %cross over frequancy of Dc_bus voltage loop
Gv=(-2/C)/s;
Gp=Gcl_current; %power controller dynamic
Kv=C/3/s;
Tdc=3/C
%#######Tdc#######initial controller transfer function(
Gol=-Kv*Gp*Gv ;% Open loop transfer function
K0=1/abs(freqresp(Gol, WcV));
Gol_uncomp=K0*Gol;
%figure, margin(Gol_uncomp)
[Gm,Pm,Wg,Wp] = margin(Gol_uncomp)
PM=60-Pm
Beta=(1-sin(PM*pi/180))/(1+sin(PM*pi/180))
tolead=1/WcV/sqrt(Beta)
Tlgg=Beta*tolead
Flead=(tolead*s+1)/(Beta*tolead*s+1);
%#######Tld=tolead,Tlg=Beta*Tolead###########
Kvlead=1/abs(freqresp(Flead*Gol_uncomp,WcV));
Gol_comp=Kvlead*Flead*Gol_uncomp;
GAINtot=K0*Kvlead

%#######Kdc##########

%figure, margin(Gol_comp)
%figure, step(feedback(Gol_comp,1))
zpk(Gol_comp);

8.4 Derivation for Rs and Vt

The unknown variables Vt and Rs are calculated by the method described as follows.
At MPPT dP/dV=0 gives the below expression

(8.1)

Solving for dV/dI in (8.1) gives

(8.2)
119

Also from (3.2), following expression can be derived for V.

ln

(8.3)

From (3.4), putting the value of I0 in (8.3) at MPP gives

ln

(8.4)

From (3.3) it is also known that Iph=Isc, putting above and solving for Rs yield the expression
as shown in (8.5).

(8.5)
From (8.3), taking derivate with respect to I gives
(8.6)

By comparing (8.2) and (8.6) the final derive expression for Vt is shown in (8.7).

(8.7)

Hence from (8.5) and (8.7) the values of Rs and Vt can be calculated respectively. It should
be noted these parameters are easy to calculate with these expression as all the entities
involved in these expressions are known from the data sheet of the PV module.

8.5 Stability Models

There are two possible types of models that are supported by static generator:
Current source model
Voltage source model
120

In this project, current source model for the RMS and EMT simulations are used. It takes the
id_ref ,iq_ref,cosref and sinref as input to the model. The output of the model provides
frequency and d-axis & q-axis components of current.

Fig 4.6 Input/output Definition of model for Stability Analysis [24]

The input and output definition of the parameters involved in current source model are shown
in Table 1 and Table 2 respectively.
Table 3: Input Definition of the Current Source Model

Parameters

Description

Unit

Id_ref

d_Axis Current Reference

p.u

Iq_ref

q_Axis Current Reference

p.u

cosref

Cos(dqReferenceAngle)

sinref

Sin(dqReferenceAngle)

Table 4: Output Definition of the Current Source Model

Parameters

Description

Unit

xspeed

Frequency

p.u

Id

Current, d_Axis

p.u

Iq

Current,q_Axis

p.u

The reader is referred to Technical reference for Static Generator available in the software
DigSilent (PowerFactory) [24] to have detailed overview of the stability models of static
generator.

121

8.6 Builtin Current controller

Fig 4.7 Builtin current controller in static generator

122