Power received by a unit surface area in a wavelength dierential d, the units are
W/m
2
m
2.2. Important Terms in PV Panel 11
Irradiance
The integral of the spectral irradiance extended to all wavelengths of interest and
the units are W/m
2
Radiation/Insolation
The time integral of the irradiance extended over a given period of time, therefore
radiation units are units of energy is given in kWh/m
2
day, or simply electrical units
per day.
2.2.2 Temperature
PV output will change with respect to temperature[8], because band gap, carrier concentra
tion are dependent on temperature. For maximum power output of the module, it is needed
to lower operating temperatures but increase the irradiance. The typical temperature coef
cient of power is 0.5%/
o
C for mono or polycrystalline silicon cells. The cell temperature
of open rack modules (e.g. aSi), however, is governed by several external factors such as
ambient temperature, irradiance level, wind speed, wind direction, and tilt angle of the mod
ule in the array. Temperature considered is actually the PV cell temperature and not its
temperature of the rack or atmospheric temperature. As per standard PV cell outputs, are
given in cell temperature of 25
o
.
2.2.3 NOCT
Nominal operating cell temperature (NOCT) is the cell temperature, when open circuited
panel subjected to a irradiance of 800W/m
2
, ambient temperature of 20
o
C and wind speed
of 1m/s at a module tilt angle 45
o
. This will give the idea of cell temperature rise in the PV
panel. Nominal value of NOCT will be around 43
o
50
o
C. this information is available in
the manufacturing data sheet.
NOCT temperature given in data sheet will be useful to nd approximate PV panel
temperature using known ambient temperature [2]
T
c
=
(T
NOCT
T
a
) E
800
(2.1)
12 Chapter 2. PV Panel Modelling
2.2.4 Standard Test Conditions(STC)
The STC (also known as SRC Standard Reporting Conditions) is dened with nominal cell
temperature 25
o
C, nominal irradiance level 1000W/m
2
at spectral distribution of Air Mass
1.5 solar spectral content. Many of the important parameter given in the data sheet is based
on this condition.
2.2.5 Open Circuit Voltage V
oc
Voltage across the PV panel, when it is open circuited (I
PV
= 0). This voltage will change
with respect to irradiance and temperature. V
oc
will increase with respect to irradiance,
decrease with respect to temperature(ve coecient
v
). This is the important parameter of
PV panel, which is used to calculate maximum PV voltage and important for system design.
Both V
oc
and
v
are given in the manufacturer data sheet at STC. All points lie in the xaxis
as shown in Fig.2.2 are open circuit voltages with respect to dierent irradiance.
2.2.6 Short Circuit Current I
sc
Current supplied by PV Panel when its terminals are shorted. This current will change
with respect to irradiance and temperature. I
sc
will proportionally increase with respect to
irradiance, logarithmically increase with respect to temperature(+ve coecient
i
). This is
the important parameter of PV panel and is used to calculate maximum fault current, it is
also important for system design. Both I
sc
and
i
are given in the manufacturer data sheet
at STC.All points lie in the yaxis as shown in Fig.2.2 are short circuit currents with respect
to dierent irradiance.
2.2.7 Maximum Power Point MPP
As shown in the Fig.2.2 PV voltage and currents are non linearly related, so at a particular
voltage, PV panel supplies maximum power, that inturn changes with respect to climatic
conditions (e.g. irradiance, Temperature, . . . etc.). The point(voltage, current) in Fig.2.2at
which PV panel supplies maximum power is called maximum power point (MPP). So to oper
ate PV at MPP requires separate controller called MPPT (maximum power point tracker).
MPPT will play an important role in PV system dynamics. Maximum power at STC is
considered as a power rating of the PV panel.
2.3. Manufacturer Data Sheet 13
2.2.8 Maximum Power Voltage and Current (V
m
&I
m
)
Voltage and current of the PV panel, when it is operating in maximum power point are
called maximum power voltage and current. Values of V
m
&I
m
are given in the data sheet at
STC.
2.2.9 Fill Factor
Every cell has a life expectancy. As time progresses, the quality of cell goes down. Hence, it is
essential to check the quality[9], periodically so that it can be discarded once the quality falls
below certain level. The quality of the cell called Fill Factor (FF) which can be calculated
as
FillFactor =
I
m
V
m
V
oc
I
sc
(2.2)
Ideally, the Fill Factor should be 1 or 100%. However, the actual value of FF is about 0.8
or 80%. A good panel has ll factor in the range of 0.7 to 0.8. For a bad panel it may be as
low as 0.4.
2.2.10 Temperature Coecient
v
&
i
Temperature coecient (
v
V/
o
C &
i
A/
o
C) given in the PV panel data sheet is used to
quantify change in the PV panel voltage and current with respect to temperature. This is
very important to model a PV panel.
2.3 Manufacturer Data Sheet
To model the practical PV panel, its required to use only the data given in the manufacturer
data sheet. Refer Fig.2.5 for the sample manufacturer data sheet. Clear understanding of
the data given in the data sheet is must for modelling of the PV panel. All the quantities
are already explained above.
2.4 PV Modelling
PV panel modelling is very important for dynamic analysis of PV system. Many models are
proposed in the literature.
14 Chapter 2. PV Panel Modelling
SANYO North America
Energy System Solutions Division
550 S. Winchester Blvd., Suite 510
San Jose, CA 95128, U.S.A.
www.sanyo.com/solar
solar@sec.sanyo.com
SANYO North America. All Rights Reserved. 1/6/2010
(OHFWULFDO6SHFLFDWLRQV
Model HIT Power 215A or HITN215A01
Rated Power (Pmax)
1
215 W
Maximum Power Voltage (Vpm) 42.0 V
Maximum Power Current (Ipm) 5.13 A
Open Circuit Voltage (Voc) 51.6 V
Short Circuit Current (Isc) 5.61 A
7HPSHUDWXUH&RHIFLHQW3PD[ 0.336%/ C
7HPSHUDWXUH&RHIFLHQW9RF 0.143 V/ C
7HPSHUDWXUH&RHIFLHQW,VF 1.96 mA/ C
NOCT 114.8F (46C)
CEC PTC Rating 199.6 W
&HOO(IFLHQF\ 19.3%
0RGXOH(IFLHQF\ 17.1%
Watts per Ft.
2
15.85 W
0D[LPXP6\VWHP9ROWDJH 600 V
Series Fuse Rating 15 A
Warranted Tolerance (/+) 0% / +10%
0HFKDQLFDO6SHFLFDWLRQV
,QWHUQDO%\SDVV'LRGHV %\SDVV'LRGHV
Module Area 13.56 Ft
2
(1.26m
2
)
Weight 35.3 Lbs. (16kg)
'LPHQVLRQV/[:[+ 62.2x31.4x1.8 in. (1580x798x46 mm)
Cable Length +Male/Female 46.45/40.55 in. (1180/1030 mm)
&DEOH6L]H&RQQHFWRU7\SH No. 12 AWG / MC4
TM
Locking Connectors
Static Wind / Snow Load 60PSF (2880Pa) / 39PSF (1867Pa)
3DOOHW'LPHQVLRQV/[:[+ 63.2x32x72.8 in. (1607x815x1850 mm)
4XDQWLW\SHU3DOOHW3DOOHW:HLJKW 34 pcs./1234.5 Lbs (560 kg)
4XDQWLW\SHU7UDLOHU 952 pcs.
2SHUDWLQJ&RQGLWLRQV6DIHW\5DWLQJV
Ambient Operating Temperature 4F to 115F (20C to 46C)
2
+DLO6DIHW\,PSDFW9HORFLW\ 1 hailstone (25mm) at 52 mph (23m/s)
)LUH6DIHW\&ODVVLFDWLRQ Class C
6DIHW\5DWLQJ&HUWLFDWLRQV UL 1703, cUL, CEC
/LPLWHG:DUUDQW\ 5 Years Workmanship, 20 Years Power Output
1
STC: Cell temp. 25C, AM1.5, 1000W/m
2
2
0RQWKO\DYHUDJHORZDQGKLJKRIWKHLQVWDOODWLRQVLWH
1RWH6SHFLFDWLRQVDQGLQIRUPDWLRQDERYHPD\FKDQJHZLWKRXWQRWLFH
$OOPRGXOHVFRQQHFWHGLQWKHVRODUDUUD\VKRXOGEHRIWKHVDPHPRGHOQXPEHU
'HSHQGHQFHRQ7HPSHUDWXUH
'HSHQGHQFHRQ,UUDGLDQFH
&
X
U
U
H
Q
W
9ROWDJH9
&
X
U
U
H
Q
W
9ROWDJH9
0.00
1.00
2.00
3.00
4.00
5.00
6.00
0 10 20 30 40 50 60
0
25
50
75
0.00
1.00
2.00
3.00
4.00
5.00
6.00
0 10 20 30 40 50 60
1000W/m
2
800W/m
2
600W/m
2
400W/m
2
200W/m
2
'LPHQVLRQV
Unit: inches (mm)
HIT Power 215A
31.4 (798)
Front Side Back
6
2
.
2
(
1
5
8
0
)
1.8 (46)
1.5 (37)
1
3
(
3
3
0
)
15.47 (393)
Ground (4 places)
0.165 (4.2) 0.205 (5.2)
0.165 (4.2) 0.205 (5.2)

+
4
0
.
5
5
(
1
0
3
0
)
4
6
.
4
5
(
1
1
8
0
)
4x mounting holes
0.276 (7)
Section AA
1
.
8
(
4
6
)
1.5 (37)
HIT
V +I
pv
R
S
V
T
1
_
_
V +I
pv
R
S
R
Sh
_
(2.3)
V
T
=
nKT
q
by neglecting R
s
and R
sh
I
pv
= I
SC
I
O
_
e
V
V
T
1
_
(2.4)
When I
pv
= 0 then V
oc
= V
T
ln
_
1 +
I
SC
I
O
_
(2.5)
When V
pv
= 0 then I
sc
= I
ph
From the Eq.2.5, open circuit voltage is logarithmically related with irradiance (PV panel
photon current)
2.4.1.1 Series Resistance R
s
The series resistive losses are present in practical solar cells. In fact, the current generated in
the solar cell bulk travels towards the contacts through resistive semiconductor material, both
in the base region( normally P typenot heavily doped) and in the narrow emitter region(N
type), which are normally heavily doped. Besides these two components, the resistance of the
16 Chapter 2. PV Panel Modelling
metal grid, contacts and current collecting bus also contribute to the total series resistive
losses. It is common practice to assume that these series losses can be represented by a
lumped resistor R
s
which is dependent on temperature is called the series resistance of the
solar cell.
2.4.1.2 Shunt Resistance R
sh
A number of shunt resistive losses are identied, such as localized shorts at the emitter layer
or perimeter shunts along cell borders are among the most common. This is represented
generally by a lumped resistor, R
sh
, in parallel with the intrinsic device.
2.4.2 Double Diode Model
In the above model, eect of non ohmic losses, due to recombination in the space charge
region of the solar cell is not considered. This is relevant at low voltage bias and can be
represented in an equivalent circuit by a second diode term with a saturation current I
02
,
which is dierent from the saturation current (I
0
) of the ideal solar cell diode, and given
ideality diode factor dierent from 1, normally diode with ideality factor 2 is used for this
purpose.
In practice, only few devices exhibit a totally ideal I(V) characteristic with ideality coef
cient equal to unity, so it is common practice to also add a parameter n to account for non
idealities in the dark current diode and the single diode model can be modied to take this
eect into account.
ph
I
01
R
sh
R
s
I
pv
V
pv
I
02
Figure 2.7: Double Diode Model
I
pv
= I
ph
I
01
_
e
V
pv
+I
pv
R
s
nV
T
1
_
I
02
_
e
V
pv
+I
pv
R
s
2V
T
1
_
_
V
pv
+I
pv
R
s
R
sh
_
(2.6)
2.4. PV Modelling 17
In Fig.2.7, rst diode(I
0
)with ideality factor n accounts for dark current (independent of
irradiance) and second diode with a ideality factor of 2 accounts for non ohmic shunt losses
in the solar cell.
The above models have some disadvantages,
Many parameters (R
s
, R
sh
, I
01
, I
02
, I
ph
, V
T
) are not available, because it is depended on
property of the material. Also there are not given by manufacturer.
Semiconductor equations are derived from ve carrier transport dierential equations,
which assume uniform doping and crystalline material. But this is not true in the case
with practical module and thin lm PV (which have pin junction). So there is a
need to model PV Cell that uses only the information given by the manufacturer data
sheet. Also it must take care of change due to climatic condition (normally irradiance,
Temperature and wind speed). Translation equations are proposed to translate voltage
and current from one condition to another condition.
forth coming models will use only the data given in the data sheet as shown in the
Fig.2.5.
2.4.3 Sandstrom Equation
In the past, there is a need for translating PV current and voltage from one environmen
tal condition to another environmental condition[10]. First accepted translation equation is
proposed by an IEEE paper by J. D. Sandstrom of the Jet Propulsion Laboratory (JPL)
published in 1967. The paper presents some very good correlative results between experi
mental measurements and the resultant analytical predictions over a cell temperature range
from 20
o
C to 130
o
C, and over an irradiance range of 500W/m to 3000W/m. The equations
were widely used and standardized as IEC 891 model[11] correction procedure 1.
I
sc
= I
sc1
_
E
2
E
1
1
_
+
i
(T
2
T
1
) (2.7)
I
2
= I
1
+ I
sc
(2.8)
V
2
= V
1
v
(T
2
T
1
) I
sc
R
s
k(T
2
T
1
)I
2
(2.9)
P
2
= V
2
I
2
the above equations Eq. 2.15 & Eq. 2.16 are used to translate PV current and voltage from
one (I
1
& V
1
) irradiance, tempertaure (E
1
& T
1
normally in STC) to an another (I
2
& V
2
)
18 Chapter 2. PV Panel Modelling
irradance and temperature (E
2
& T
2
)
2.4.4 A.J. Anderson Translation equation
The above equations Eq.2.15 and Eq.2.16 suggest that the temperature eect on current
is independent of irradiance, if
i
assumed as a constant temperature coecient. The
irradiance eect on voltage (V
oc
) as in Eq. 2.16 is accounted for by the term I
sc
R
s
,
but by experiment, it is found (by Anderson[10]) that variation in voltage due to change
in irradiance found very little eect on PV voltage. So for dierent R
S
values Sandstrom
voltage equation found incorrect. Corrected Sandstrom equations proposed by Anderson are
given as below;
I
SC2
=
I
SC1
[1 + (T
1
T
2
)] [E
1
/E
2
]
(2.10)
V
OC2
=
V
OC1
[1 + (T
1
T
2
)]
_
1 + ln
E
1
E
2
_ (2.11)
I
2
= I
1
_
I
SC2
I
SC1
_
(2.12)
V
2
= V
1
_
V
OC2
V
OC1
_
(2.13)
P
2
= I
2
V
2
Where & are neutered temperature coecients for avoiding the error due to scaling
=
I
sc2
I
sc1
I
sc2
(T
2
T
1
)
& =
V
oc2
V
oc1
V
oc2
(T
2
T
1
)
the above equations would result in I
scl
translating to I
sc2
along the line R = 0 ohms,and
V
oc1
would translate to V
oc2
along the line R = . The IV point pairs in between would
similarly translate along lines of constant resistance with values of R =
V
1
I
1
and value of
ca be found by using experiment data, otherwise =
kT
q
. The above equations can be used
to nd all PV parameters by using manufacturer data in STC, but it is not providing the
relation between PV voltage and current.
2.4.5 Bleasser Equations
It is also a corrected form of Sandstrom equation, In current equation[12] eect of irradiance
during temperature change is included. Also, instead of transforming short circuit current
2.4. PV Modelling 19
I
sc
as in Eq.2.14, actual current (PV current for the present load)is transformed directly.
Voltage equation is mostly same as Sandstrom equation. This equation is standardized by
IEC 891[11] as a correction procedure 2.
I
sc
= I
sc1
_
E
2
E
1
1
_
+
i
(T
2
T
1
) (2.14)
I
2
= I
1
+ I
sc
(2.15)
V
2
= V
1
v
(T
2
T
1
) I
sc
R
s
k(T
2
T
1
)I
2
(2.16)
P
2
= V
2
I
2
The above equations also does not providing the relation between PV voltage and current. So
there is a need for an appropriate model which uses only the data given in the manufacturer
data sheet, and it has to provide the VI characteristic equation.
2.4.6 Michigan University Model
This model[13] is based on single diode model where R
sh
is neglected, and material property
parameters are assumed as variable and its value found by using known parameter, which is
given in the data sheet.
I
pv
(V
pv
) = I
max
i
I
max
i
e
V
pv
b(+1)(V
max
+
v
)
1
b
(2.17)
=
E
2
E
1N
I
max
=
I
SC
1 exp
(1/b)
= 1
V
min
V
max
+
v
i
= 1 +
TCi
100
(T
2
T
1
N)
v
= TCV (T
2
T
1N
)
PowerP
pv
= V
pv
I
pv
(2.18)
V
min
is the opencircuit voltage rating of the solar panel array for an eective intensity of
light less than 20% over the solar panels, T
1N
refers to a STC temperature T
1N
= 20
o
C,
E
1N
refers to a STC temperature, T
1N
= 1000W/m
2
, V
max
is open circuit voltage at STC.
TCi & TCV refers to the temperature coecient of panel given in the data sheet. Fit
variable b can be found by using available information in the data sheet.
20 Chapter 2. PV Panel Modelling
2.4.6.1 Calculation of b (t variable)
Maximum PV power can be calculated[14] by Eq.2.18, V
m
& I
m
is the PV current and
voltage at maximum power point, this information is given in the data sheet for STC. At
the maximum Power
dP
pv
dV
pv
= 0 differentiate Eq.2.18
0 = 1
_
V
m
b ( + 1 ) (V
max
+
v
)
_
exp
(
V
m
b(+1)(V
max
+
v
)
1
b
)
(2.19)
Using a guess value b = 0.09 calculate V
m
(1000W/m
2
, 25
o
C) by solving Eq.2.19. Equation
Eq.2.17 and Eq.2.18 are used to calculate I
m
(1000W/m
2
, 25
o
C) and P
m
(1000W/m
2
, 25
o
C).
These calculated values have to be compared with values from a module data sheet. Now the
guess value of t parameter b can adjusted iteratively to minimise the error of P
m
(1000W/m
2
, 25
o
C)
at STC. Fit variable b for SANYO 215A panel is calculated as b=0.079.
2.5 PV Model Implementation
MATLAB/Simulink is used to implement PV model. Both Anderson equation and Michigan
university models are implemented in the single block as shown in the Fig2.9. Here all the
required data sheet parameters, environmental (E & T), design datas are fed into to the
model as shown in Fig.2.8, by using given parameters. PV panel model calculates current
I
pv
based on the PV panel voltage V
pv
as per the Eq.2.17. Implemented model is simulated
Figure 2.8: PV Panel Model implementation
2.6. PV System Design 21
for dierent environmental conditions and results are matching with the given manufacturer
data sheet curve(refer Fig.2.2 for data sheet curve and Fig.2.5 for simulated results).
Figure 2.9: PV Panel Model in MATLAB/SIMULINK
2.6 PV System Design
Two PV plant with the capacity of 1MW (2 500kW) is modelled for the analysis as shown
in the Fig.2.10. Here SANYO 215A (215W) PV panel as in Fig.2.5 is used for simulation.
Many panels connected in series and parallel as shown in the Fig.2.11 for achieving high
voltage and power.
22 Chapter 2. PV Panel Modelling
2.6.1 Selection of PV Arrays
Number of panels required are selected based on the calculation given below; From the
PV
module
C
Inverter L
f
Filter
PV System
C
f
415V
PV
module
C
Inverter L
f
Filter
PV System
C
f
415V
415V
Grid
33kV
3ph
3ph
3ph
500kW
500kW
Figure 2.10: PV System Single line diagram
Fig.2.10 the expression for inverter output voltage is given as
v
i
= L
f
di
s
dt
+v
grid
(2.20)
If v
gird
= V
gm
sin(t)
For UPF i
s
= I
sm
sin(t)
So v
i
= L
f
I
sm
cos(t) +V
m
sin(t) (2.21)
v
i
= jX
L
i
s
+v
grid
(2.22)
Required inverter voltage is decided by grid voltage and voltage across inductor as in Eq.2.22.
Also inverter voltage is depends on DC bus voltage (PV Voltage) and method of switching.
maximum inverter voltage possible for Sine Triangle PWM is V
i
= V
pv
/2. Maximum inverter
output voltage required is calculated by assuming inductor voltage of 0.1pu, when rated
current ow through the inductor, and 10% increase in grid voltage is considered
Required max V
imax
= 0.1V
grid
+ 1.1 V
grid
(2.23)
V
pvmin
= 2V
imax
(2.24)
It is required to connect many PV panels in series to achieve the voltage requirement of
the inverter (as referred in Eq.2.23). To supply maximum possible power to the grid at all
2.6. PV System Design 23
environmental condition, minimum PV voltage (less irradiance and high temperature) must
be equal to two times of maximum required inverter voltage(as given in Eq.2.24).
For PCC line voltage of 415V, minimum of 20 SANYO 215A panels (N
s
= 20) have to
be connected in series as shown in the Fig.2.11, So maximum possible voltage is equal to
N
s
V
oc
= 51.3 20 = 1026V and minimum PV voltage is equal to Ns V
min
= 820V .
PV
module
PV
module
PV
module
PV
module
PV
module
PV
module
PV
module
PV
module
PV
module
C
Ns
2
1
1 2
Np
PV
module
C
500kW
Figure 2.11: Structure of PV Array
To achieve the required power rating (normally represented in STC), many PV string
must be connected in parallel as given in Eq.2.25, which are called as PV arrays.
N
p
=
P
pv
N
s
V
m
V
m
is at STC (2.25)
For 500kW PV system N
p
= 117 PV string must be connected in parallel. So 500kW PV
system need 2340 SANYO 215A panels. Actual rating of PV system is equal to 504.18kW.
P
pv
= (N
s
N
p
) P
panel
2.6.2 Inverter
Since, main focus is on studying transient response for large signal disturbances, so harmonics
generated by the inverter is not of much interest. So average model of inverter is sucient
for the analysis, but simulation carried out for both systems with and without inverter shows
no signicant dierence found.
Inverter Design Details
Two level three phase inverter with the rating of 500kW
24 Chapter 2. PV Panel Modelling
Sine triangle PWM with a switching frequency equal to 5kHz
LC lter is used
Inverter is modelled as a rst order system with the time constant equal to the half of
switching time (bandwidth is equal to twice that of switching frequency)
Constant current mode control is used to inject UPF power to the grid
Voltage controller is used to regulate the DC bus voltage.
2.6.3 Maximum Power Point Tracking (MPPT)
As observed fro the earlier studies we seen earlier due to the non linear property of PV panel,
at a particular operating point (V
pv
, I
pv
) PV generate maximum power is called maximum
power point, Hence it required to have an reliable technique to track the maximum power
point irrespective of the environmental conditions. A MPPT system is therefore always
implemented to optimize the eciency of the photovoltaic energy conversion. This tracking
system adjusts the inverter voltage reference signal as shown in the Fig.2.12 and hence, the
dc voltage at the output of the solar array.
Many MPPT techniques are proposed[15] in the literature, most popular being incremen
tal conductance method and P&O (Perturbation & Observation method). P&O method is
used due to its simplicity and easy implementation. the P&O method works by perturbing
V
pv
and observing the impact of this change on the output power of the PV array as shown
in the Fig2.12.
At each cycle, V
pv
and I
pv
are measured to calculate P(k) . This value of P(k) is
compared to the value P(k 1) calculated at the previous cycle. If the output power has
increased, V
pv
is adjusted further in the same direction as in the previous cycle. If the
output power has decreased, V
pv
is perturbed in the opposite direction as in the previous
cycle. is thus perturbed at every MPPT cycle. When the maximum power point is reached,
oscillates around the optimal value . This causes a power loss that increases with the step
size of the perturbation. If this step width is large, the MPPT algorithm responds quickly
to sudden changes in operating conditions. On the other hand, if the step size is small
the losses under stable or slowly changing conditions will be lower but the system will not
respond quickly to rapid changes in temperature or irradiance. There is always trade o
2.6. PV System Design 25
Figure 2.12: Flowchart of P&O MPPT Algorithm
between speed and steady state loss. Here 0.1V perturbation with the time 2ms is used
to accelerate the simulation speed. MPPT tracking range is set between 700 to 1026 V.
Experiments demonstrated that the maximum power point tracking (MPPT) dictates the
dynamic behavior of the PV generator
Chapter 3
PV System Control
Synchronizing the PV system to the grid needs information about certain grid parameters,
these informations are obtained from PLL. Here synchronous reference frame (d&q axis)
based control, where three phase quantity is translated into a rotating space vector, which
is used to control the power ow. In this study, grid voltage vector is aligned towards d
axis. So aligning of current vector to the grid d axis ensures a unity power factor operation.
Separate controllers are used to control active (I
d
) , reactive (I
q
) power, and separate voltage
controller is used to control the DC bus voltage. PLL is used to track the gird frequency
and voltage.
3.1 Synchronous Reference Frame Strategy
Synchronous reference frame (SRF) strategy is a popular method used in the generation of
the current reference and constant current control interface uses this SRF based. The syn
chronous reference frame strategy uses coordinate transformations to generate the current
reference. It employs the well known Clarkes Transformation and Parks Transformation for
this purpose. uses the primitive machine model, where balanced current ow in the in
duction machine (RYB) static three phase winding (RBY) result into rotating ux space
vector. Synchronous reference is given as reference to the controller and the controller output
is obtained, reverse transformations are employed to transform the quantities back to the
actual threephase system.
26
3.1. Synchronous Reference Frame Strategy 27
3.1.1 Synchronous Frame Transformation for Three Phase Bal
anced System
v
a
= V
m
sin(t ) (3.1)
v
b
= V
m
sin(t 120) (3.2)
v
c
= V
m
sin(t + 120) (3.3)
Three phase balanced system can be easily represented by two variables(phase quantities).
Since third being found by using (v
a
+v
b
+v
c
= 0), it is possible to transform the three phase
balanced system (a, b, c) to a stationary two phase system ( ) as shown in Fig.3.1 By
a
b
c
d
q
V
s
abc to
 
to dq
Figure 3.1: Frame Transformation
using orthogonal decomposition of vector,
_
v
_
=
_
1
1
2
1
2
0
3
2
3
2
_
_
_
v
a
v
b
v
c
_
_
(3.4)
for balanced system, where v
a
+v
b
+v
c
= 0
v
=
3
2
v
a
(3.5)
v
3
2
(v
b
v
c
) (3.6)
v
and v
vary sinusoidally with respect to time, this stationary reference frame can be
translated[16] to synchronously rotating reference frame by using Parks Transformation.
Space vector
V = v
+ jv
_
(3.7)
In general transforming(mapping) the data from one Cartesian coordinate to another co
ordinate system, with angle between the coordinate is , then;
V
d
+ jV
q
= (v
+jv
)e
j
(3.8)
By aligning the daxis on the space vector,
V
d
=
3
2
V
m
(3.9)
V
q
= 0
3.1.2 Synchronous Frame Transformation for Three Phase Unbal
anced System
Since frequent faults in the systems are unbalanced in nature (e.g. LG), it is required
to understand the eect of unbalance in the (v
d
&v
q
). In general unbalanced three phase
quantities can be represented as shown in the Fig.3.2, by three balanced vectors namely
positive, negative and zero sequence component.
2
0
1
V
a
V
b
V
c
V
ao
, V
bo
, V
co
V
a1
V
b1
V
c2
V
a2
V
b2
V
c1
Figure 3.2: Sequence Component of Unbalanced System
v
a
= V
0m
sin(t +
0
) + V
1m
sin(t +
1
) + V
2m
sin(t +
2
) (3.10)
v
b
= V
0m
sin(t +
0
) + V
1m
sin(t +
1
120) + V
2m
sin(t +
2
+ 120) (3.11)
v
c
= V
0m
sin(t +
0
) + V
1m
sin(t +
1
+ 120) + V
2m
sin(t +
2
120) (3.12)
3.1. Synchronous Reference Frame Strategy 29
After substituting Eq.3.10, 3.11 & 3.12 in Eq.3.4, zero sequence component will cancel out
each other and only positive and negative sequence component will be present.
v
=
3
2
(V
1m
sin(t +
1
) + V
2m
sin(t +
2
)) (3.13)
v
3
2
(V
1m
(sin(t +
1
120) sin(t +
1
+ 120))) +
3
2
(V
2m
(sin(t +
2
+ 120) sin(t +
2
120))) (3.14)
By substituting Eq.3.13, 3.14 in Eq.3.7, assuming d axis is aligned towards positive sequence
voltage.
= t
2
+
1
(3.15)
V
d
=
3
2
(V
1m
V
2m
cos(2t +
1
+
2
)) (3.16)
V
q
=
3
2
(V
2m
sin(2t +
1
+
2
)) (3.17)
From the above equations, due to the unbalance in the three phase quantity, d axis component
of the three phase quantity will have eect of second harmonics negative sequence component.
This is because negative sequence space vector will move in opposite direction with respect
to positive sequence vector as shown in the Fig.3.3. For bolted LG fault on Cphase,
(V
c
= 0 V
a
= 239V (rms phase voltage), and V
b
= 239V (rms phase voltage)). The
sequence components are shown in the Fig.3.3, equivalent d & q axis components are shown
in the Fig.3.4.
250
500
30
210
60
240
90
270
120
300
150
330
180 0
100
200
30
210
60
240
90
270
120
300
150
330
180 0
25
50
30
210
60
240
90
270
120
300
150
330
180 0
Unbalanced System Sequence Components d q 0 Components
V
a
V
b
V
c
=0
V
a1
V
a0
V
a2
V
0
V
d
V
q
Figure 3.3: Sequence Component for LG Fault at Phase C
30 Chapter 3. PV System Control
0 0.01 0.02 0.03 0.04 0.05 0.06
200
100
0
100
200
300
400
500
600
time s
V
o
l
t
a
g
e
V
V
d
V
q
Figure 3.4: Sequence component for LG fault at phase C
3.2 Phase Locked Loops (PLL)
To supply unity power factor current and to maintain synchronism with the grid, PV sys
tem needs grid angle,voltage and frequency, Synchronous reference frame PLL with moving
average lter[17] is implemented as shown in the Fig.3.5. Moving average lter is used
to eliminate 2
nd
harmonic negative sequence component as in Eq.3.16. Dynamics of PLL
have signicant role in transient behaviour of PV system. Because important parameters
(V
sd
, V
sq
, sinj(), cos(), ) are obtained from PLL, which have signicant inuence in
the modulation index. The parameter of the controller chosen using symmetrical optimum
Figure 3.5: Simplied structure of SRF PLL with lter
method[18]. value of k
p
= 0.5 and T = 0.08. Performance of PLL shown in the Fig3.6. in
this case LG fault is created at time t=0.25s.
3.3. Current Reference Generation 31
Figure 3.6: Performance of PLL during fault at t=0.5s
3.3 Current Reference Generation
All the power generated by the PV must be inject to the grid, at the same DC bus voltage
have to maintained constant. Power injected by the inverter to the gird in dq domain is
given as,
S =
3
2
V
m
I
m
=
2
3
(V
d
+jV
q
)(I
d
jI
q
)
=
2
3
(V
d
I
d
+ V
q
I
q
+j(V
q
I
d
V
d
I
q
))
So active power is P =
2
3
(V
d
I
d
+ V
q
I
q
)
V
q
= 0 space vector aligned towards daxis
P =
2
3
(V
d
I
d
) (3.18)
Power Generated by PV P
pv
= V
pv
I
pv
(3.19)
Equating Eq.3.18 and Eq.3.19, current reference set by the PV system is
I
d
=
3V
pv
I
pv
2V
d
(3.20)
I
dref
= I
d
I
dcap
(3.21)
I
dref
is the current reference to the current controller, I
dcap
 Current required by the capacitor
to maintain constant voltage. To protect and increase the life time of the inverter, current
32 Chapter 3. PV System Control
reference is limited to its rated current or 1.5 to 2 times of the rated current. In this study
current reference is limited to 1.5 times of the rated current.
3.4 Filter Design
LC lter is used to lter out the harmonics as specied by standard IEEE 1527, for grid
connected application. For LC lter, value of the inductor is selected as same like L lter.
value of C must be chosen[19], such that resonant frequency of the lter is higher than the
bandwidth of the closed current loop. Normally maximum inductor voltage is limited to
0.1pu of the rated voltage. Maximum possible inductor can selected as shown below,
V
L
f
= 0.1 V
sph
V
sph
=
415
3
= 239.6V
L
f
I
s
= 239.6V where I
srms
= 701A for 500kW PV Plant
L
f
= 0.1088mH
C
f
= 470F
Resonant Frequency F
s
=
1
2
LC
F
s
= 545.57Hz (3.22)
For ease of simulation, lter capacitance is considered in distribution system side (33kV)
along with shunt capacitance of the system, so transformation of capacitance from one
voltage level to another voltage done by using energy balance.
1
2
C
f1
V
s1
2
=
1
2
C
f2
V
s2
2
HereV
s1
= 415V ;
V
(
s2) = 33kV
C
1f
= 470F
So C
2f
= 470
_
415
33000
_
2
C
2f
= 0.0743F (3.23)
Inductor is designed using area product method and its resistance R
f
= 1.02m
3.5. Current Controller 33
3.5 Current Controller
Controls the current injected to the grid and convert the current reference I
dref
(as in
Eq.3.21), I
qref
(equal to zero for UPF) into required inverter voltage by controlling the mod
ulation index. Decoupled Id & Iq are used to simplify the MIMO[16] system into two SISO
system. From the Fig.3.7,
R
f
L
f
V
sa V
ia
i
sa
R
f
L
f
V
sb V
ib
i
sb
Inverter
side
R
f
L
f
V
sc V
ic
PCC
(Grid)
i
sc
Figure 3.7: Inverter  Filter  Grid
v
ia
= L
f
di
sa
dt
+ i
sa
R
f
+v
sa
(3.24)
v
ib
= L
f
di
sb
dt
+i
sb
R
f
+v
sb
(3.25)
v
ic
= L
f
di
sc
dt
+i
sc
R
f
+ v
sc
(3.26)
if v
sa
= V
sm
sin(t)
for UPF i
sa
= I
sm
sin(t)
So v
ia
= L
f
I
sm
cos(t) +V
sm
sin(t) (3.27)
Current injected to the grid (at PCC) i
sa
can be controlled by controlling the inverter voltage
v
ia
as given in the Eq.3.27. Inverter voltage can be controlled by switching the inverter in a
dierent modulating signal.
3.5.1 Current Equations in dq Domain
Here synchronous reference frame based constant current control interface is used to connect
the PV system to the grid. All current equations must be transferred to dq domain. For
balanced system, current injected to the distribution system (at PCC) is given by,
i
sa
= I
sm
sin(t) (3.28)
i
sb
= I
sm
sin(t 120) (3.29)
i
sc
= I
sm
sin(t + 120) (3.30)
34 Chapter 3. PV System Control
Multiplying
3
2
Eq.3.24 as per Eq.3.5
v
i
= L
f
di
s
dt
+i
s
R
f
+v
s
(3.31)
Multiplying
3
2
(Eq.3.25  Eq.3.26) as per Eq.3.6
v
i
= L
f
di
s
dt
+i
s
R
f
+ v
s
(3.32)
So space vector
I
s
= i
s
+ji
s
and
V
s
= v
s
+ jv
s
from Eq.3.31 & Eq.3.32
V
i
= L
f
d
I
s
dt
+
I
s
R
f
+
V
s
(3.33)
As we know
+ j = (d + jq)e
where = t
2
(3.34)
(Eq.3.31 +j Eq.3.32)e
V
id
= L
f
dI
sd
dt
+I
sd
R
f
L
f
I
sq
+V
sd
(3.35)
V
iq
= L
f
dI
sq
dt
+ I
sq
R
f
+L
f
I
sd
+V
sq
(3.36)
Since space vector
V
s
is aligned with daxis, so
V
s
= V
sd
& V
sq
= 0. From the above equations
Eq.3.35 & Eq.3.36,daxis and qaxis components are coupled to each other, so its an MIMO
control, to convert this into SISO, feed forward term(as given in Eq.3.37), this will eliminate
the decoupling term as given in the equation.
V
id
= V
id
+L
f
I
sq
+
V
s
 (3.37)
V
id
= V
iq
L
f
I
sd
(3.38)
By substituting Eq.3.37 into Eq.3.35 and Eq.3.38 into Eq.3.36
L
f
_
dI
sd
dt
dI
sq
dt
_
=
_
R 0
0 R
__
I
d
I
q
_
+
_
V
sd
V
sq
_
(3.39)
From the Eq.3.39, it is possible to control I
sd
& I
sq
independently. Input to the controller is
current error (I
derror
& I
qerror
) and output of the controller is voltage (V
id
& V
iq
).
3.5. Current Controller 35
Figure 3.8: Current controller block diagram
3.5.2 Current Controller Design
PI controller is used for control, parameter of the controller is selected based on[20]. Here
lter inductor acts as a plant (where T
f
=
L
f
R
f
), and sampling time (T
s
/2) is taken as a
sensor time constant, is shown in Fig.3.8. Zero of the PI controller is used to remove the
pole of the plant where T
c
= T
f
.
T
c
=
L
R
= 0.1057s
After eliminating system pole, the closed loop transfer function[21](same for both d and q
current controller) is equal to
V
id
(s)
I
sd
(s)
=
KG
T
T
f
R
f
(1 + sT
s
)
s
2
+
s
T
+
GKK
s
R
f
T
T
f
T
= T
d
+ T
s
= T
switching
n
=
GKK
s
T
T
f
R
f
2
n
=
1
T
if = 0.707
So K =
T
f
R
f
2GT
K
s
(3.40)
Here T
f
= 0.1057s R
f
= 1.02m G = 1 T
= 0.0002, K
s
= 0.0001 and by substituting in
Eq.3.40
Calculated value Controller gain K = 0.272, Fig3.9 shows the response of the controller.
The Fig.3.10 shows the overall current controller with added decoupling terms.
36 Chapter 3. PV System Control
40
30
20
10
0
10
10
2
10
3
10
4
10
5
135
90
45
0
P
h
a
s
e
(
d
e
g
)
M
a
g
n
i
t
u
d
e
i
n
(
d
B
)
Bode Plot for Current Loop
Magnitude Plot
Phase Plot
Frequency (rad/s)
0 0.5 1 1.5 2 2.5
x 10
3
0
0.2
0.4
0.6
0.8
1
1.2
A
m
p
l
i
t
u
d
e
time (s)
Impulse Response of Current Loop
Figure 3.9: Current Controller Response
3.5.3 PV System Current Calculation
Grid voltage at PCC is sensed and as per Fig.3.10 current controller sets the inverter output
voltage, so PV system current is given by,
i
san
(s) =
(v
ian
v
san
)
sL
f
+R
f
n phase quantity
where v
ian
=
1
3
(v
ab
v
ca
) (3.41)
The above Eq.3.41 valid only if system is balanced, but unbalance is very common during
fault. Also transformer is Y ad there is no neutral terminal available to measure grid
phase voltages at PCC. So current must be derived using only line voltages, by applying
KVL in the below circuit.
3.5. Current Controller 37
Figure 3.10: Overall Current Controller Block Diagram
As shown in the Fig.3.11. Voltage across the lter (inductor with its resistance) can be
R
f
L
f
V
ica
i
sa
V
iab
R
f
L
f
V
sab
i
sb
V
ica
L
f
PCC
(Grid)
Inverter
side
R
f i
sc
Figure 3.11: Calculation of PV System Current
obtained by using the Eq.3.42
_
_
V
L
fa
V
L
fb
V
L
fc
_
_
=
_
_
0.6667 0.3333 0.3333
0.3333 0.3333 0.3333
0.3333 0.6667 0.3333
_
_
_
_
x
1
x
2
0
_
_
(3.42)
38 Chapter 3. PV System Control
Where
x
1
= v
iab
v
sab
x
2
= v
ibc
v
sbc
i
san
(s) =
V
L
fa
sL
f
+R
f
i
sbn
(s) =
V
L
fb
sL
f
+R
f
i
scn
(s) =
V
L
fc
sL
f
+R
f
3.6 DC bus Capacitor Voltage Controller
PV is directly connected across DC bus capacitor, So Voltage across capacitor must be
maintain constant to harvest maximum power from the PV, also to reduce output ripple
and to ensure balanced inverter output. Voltage is controlled by controlling the output
current as given in the Eq.3.43.
I
cap
= I
pv
S
a
I
sa
+S
b
I
sb
+ S
c
I
sc
(3.43)
From the above equation, by controlling the current injected to the gird, DC bus capacitor
voltage can be controlled. PI voltage controller is used to regulate the dc bus voltage. This
controller converts voltage error to current required by the capacitor as shown in the Fig.3.12
Voltage PI controller gives daxis current I
dcap
, which must be injected to the capacitor (to
Figure 3.12: Voltage Controller Block Diagram
regulate the voltage at the MPP level). This current is given as input to the current controller
3.6. DC bus Capacitor Voltage Controller 39
and converted into I
dc
(actual current through capacitor) and given to the plant.
I
dc
= K
2
I
dcap
where K
2
=
_
2
3
V
r
ms(L
L
)
V
dc
3.6.1 Voltage Controller Design
From the Fig.3.12, system have two poles at origin, also open loop must maintain 20dB
gain zero crossing for achieving better phase margin. Symmetrical optimum method is
used to design the controller, in this method, PI controller zero is selected such that, gain
zero crossing(at (
c
)) of the open loop transfer function placed symmetrically between PI
controller zero (1/T
v
) and open loop system dominant pole (1/T
). Current controller is
simplied as a rst order system as shown in the Fig.3.12.
Open loop transfer function is given by
V
dc
(s)
V
dcref
(s)
=
K
e
(1 + sT
v
)
S
2
T
v
(1 + sT
)
whereT
= 2T
+ T
s
K
e
= K
s
K
2
K
v
K
ct
open loop phase margin is equal to
Phase margin = tan
1
a tan
1
_
1
a
_
if a = 2
Phase margin = 36.87
o
c
=
1
aT
T
v
=
a
c
(3.44)
K
v
=
c
C
K
ct
K
2
K
s
(3.45)
Here K
s
= 1 ,
c
= 1000rad/s C = 10mF & K
ct
= 1
so K
v
=
10
K
2
T
v
= 0.002s
Voltage reference is generated by the MPPT, MPPT will always perturbate the capacitor
voltage reference (of the controller) to track the maximum power. in this current control
40 Chapter 3. PV System Control
interface, current loop act as a inner loop, voltage loop act as a outer loop and MPPT act
as a outer loop for voltage controller.
So Voltage controller must be slower than current controller and MPPT perturbation time
must be set slower than the voltage controller.
80
60
40
20
0
20
10
1
10
0
10
1
10
2
10
3
10
4
10
5
180
135
90
45
0
45
M
a
g
n
i
t
u
d
e
i
n
(
d
B
)
P
h
a
s
e
(
d
e
g
)
Frequency (rad/s)
Phase Plot
Magnitude Plot
Bode Plot for Current Loop
0 1 2 3 4 5 6 7
x 10
3
0
0.2
0.4
0.6
0.8
1
1.2
1.4
Impulse Response of Current Loop
A
m
p
l
i
t
u
d
e
time (s)
Figure 3.13: Voltage Controller Response
3.6.2 Dynamic Simulation of DC bus Capacitor
Since full simulation is done by mathematical equations using Simulink blocks, all the possible
voltage variation must be modelled. It is found that, capacitor voltage will vary due to
Change in PV maximum power point I
pv
& V
pv
Output Current (i
sa
, i
sb
, I
sc
)
Change in Maximum power point, taken into account by the voltage controller as shown in
the Fig.3.12 and later modelled either by injecting the capacitor current[22] or using power
balance in the voltage controller block diagram as shown in the g.3.14. Since average
3.7. Simulation of PV System 41
inverter model is used for simulation, so power balance is sucient to account for PV system
output current variations Here only active power balance is considered, since reactive power
Figure 3.14: Voltage Controller with Power Balance
wont aect capacitor voltage signicantly. Eect of active power balance on the capacitor
voltage is given by Eq.3.46
P
cap
= V
pv
I
pv
V
sd
I
sd
+ V
sq
I
sq
V
D
=
V
pv
I
pv
V
sd
I
sd
+V
sq
I
sq
V
pv
sC
(3.46)
by using the above equations, dynamics of the capacitor is successfully implemented, response
found satisfactory
3.7 Simulation of PV System
PV system is simulated by considering all the its dynamic elements, which already discussed
in the previous sections. PV system have nonlinear elements as shown below
C
dV
pv
dt
= I
pv
(S
a
i
sa
+ S
b
i
sb
+S
c
I
sc
) (3.47)
from the above Eq.3.47, I
pv
is non linear function of V
pv
. PV system have set of non linear
dynamics equations, solving those equations using small signal analysis is not valid for large
disturbances like faults. So all PV system elements are fully modelled by using its charac
teristic equations in MATLAB/SIMULINK as shown in the Fig.3.15.
Working of the PV System model is given below
PV panel supplies the current (I
pv
) based on the DC bus capacitor voltage (V
dc
= V
pv
)
42 Chapter 3. PV System Control
MPPT senses I
pv
&V
pv
, and set the voltage reference for the voltage controller. MPPT
always perturbate the DC bus voltage towards maximum power point.
Based on the power generated (V
p
v I
pv
) by the PV panel, synchronous current ref
erence (I
dpv
) is generated and fed to the current controller to inject all possible PV
power to the grid.
Voltage controller try to regulate the DC bus voltage as dictated by the MPPT and
generate the capacitor synchronous current reference (must be injected to the capaci
tor) to regulate the voltage. It fails during power imbalance (e.g. fault)
Capacitor current reference is subtracted (I
dpv
I
dcap
) from the PV current reference
and given to the current controller as the reference current (I
dref
).
Current controller sets the inverter pole voltages based on the current reference and
control the power ow to the grid, for maintaining UPF I
sq
= 0.
Based on the grid voltage (at PCC) space vector magnitude (V
sd
& V
sq
) with its angle
(), current controller sets the inverter voltage space vector (V
id
& V
iq
) with certain
angle to control the current. So based on the grid voltage, PV system supplies the
current.
PLL is used to track the grid voltage space vector and its angle, this is given as an
input to the current controllers, so all transformation is carried out based on the grid
space vector angle.
Forcing current space vector lie on the gird voltage vector assures UPF operation. PLL
gives voltage, phase and frequency information of the grid. So it is act as an important
role in synchronization.
Frame Transformation is used to convert all the synchronous quantity to a actual
three phase quantity (dq to abc). by using inverter pole voltages (set by the current
controller) and the grid voltages, PV output currents are computed and injected to
the grid at PCC.
PV system injects current to the grid based on the available PV power and grid voltage.
3.7. Simulation of PV System 43
Figure 3.15: Simulation of PV system
Chapter 4
Distribution System
Large PV plants( MW size) supplies power to either 33kV or 66kV, roof top PV plants are
design to supply in 415V secondary distribution system. Some medium range PV plants(of
few hundred kW) will supply power to 11kV primary circuits.
4.1 Important terms in Distribution System
As per IEEE 1547 standard (as shown in the Fig.4.1)
Electrical Power System (EPS)  Facilities that deliver electric power to a load
Area EPSAn EPS that serves Local EPSs
Local EPS  An EPS contained entirely within a single premises or group of premises
Distributed Resources (DR) Sources of electric power that are not directly connected
to a bulk power transmission system. DR includes both generators and energy storage
technologies
Distributed Generation  Electric generation facilities connected to an Area EPS through
a PCC, a subset of DR
Island  A condition in which a portion of an Area EPS is energized solely by one or
more Local EPSs through the associated PCCs while that portion of the Area EPS is
electrically separated from the rest of the Area EPS
Point of Common Coupling (PCC) The point where a Local EPS is connected to an
Area EPS
44
4.2. System Study 45
Figure 4.1: DG interconnection Terms IEEE std. 1547
4.2 System Study
Two photovoltaic system (2 500kW) with the aggregate capacity of 1MW is connected to
the grid(as shown in the Fig.4.2).
Distribution system under consideration contains[23]
33kV four bus distribution system is supplied by grid substation (Area EPS) and PV
system
Local load connected between PV system and the substation.
Three distribution lines, connecting PV to local load (Line 1), local load to substation
(Line 2) and substation to substation local load (Line 3)
Loads have its own reactive power compensation using static capacitors.
Only RL load is considered
Filter Capacitor of the PV system is transformed to 33kV (from 415V) and considered
as a shunt capacitor for Bus 1
Y n transformer, primary LV (415V ) on PV system side and grounded Y HV (33kV )
on grid side.
Substation model as an innite grid using its short circuit MVA
46 Chapter 4. Distribution System
R
g1
L
g1
R
1L
L
1L
R
g2
R
g3
L
g3
C
1L
Load1
C
2L
Grid
L
2L
R
2L
Load2
C
s
R
s
L
s
Transformer
PV
module
C
Inverter R
f L
f
Filter
PV System
Distribution System
415V
C
f
33 kV
Line1
Line2 Line3
Bus 1
Bus 2
Bus 3
Bus 4
3ph
1MW
V
sg
V
f
V
1L
V
2L
V
2L
L
g2
Figure 4.2: Single Line Schematic Diagram of the Photovoltaic System Interfaced with
Distribution Network
Substation along with load2 act as an Area EPS, PV system with its local Load1 act as a
Local EPS, and bus1 is considered as PCC.
4.3 Distribution System Parameters
Lumped parameters are used to model[24] the distribution system, simple radial 4 bus dis
tribution system is considered
4.3.1 Modelling of Load
Simple RL load with a capacitor reactive power compensation is modelled, as per the stan
dard for islanding simulation, capacitor must be chosen such that resonance will happen at
rated frequency (50Hz).
R
L
=
V
2
P
load
Quality factor Q
f
= 1.7
L
L
=
V
2
Q
f
P
l
oad
C
L
=
Q
f
P
load
V
2
= 2f where f = 50Hz
here Line V oltage V = 33kV
4.3. Distribution System Parameters 47
4.3.2 Innite Grid Model
Here substation is modelled as an innite grid by using its short circuit MVA (SMVA) at
Bus3
SMV A = 500MW
V = 33kV
if
R
s
Z
s
= 0.15
Z
s
=
V
2
SMV A
= 2.178
R
s
= 0.15 Z
s
= 0.3267
L
s
=
_
(Z
2
s
R
2
s
)
= 6.86mH
C
s
= 1F
4.3.3 Transformer
Transformer ( Y n) is modelled by using its leakage reactance and the resistance, all
parameters are translated with respect to high voltage secondary. Leakage reactance are
calculated using its short circuit test voltage (V
sc
), inverter grade transformer is designed for
relatively high leakage reactance, so V
sc
= 8%.
P
rated
= 1.25MW
%V
sc
= 8%
Efficiency = 99.8%
HV side Current I
rated
=
P
rated
V
rated
= 21.87A
V oltage drop V
L
=
%V
sc
V
rated
3
= 1.524kV
Z =
V
L
I
rated
= 69.7
R =
(1 )P
rated
3I
2
rated
= 1.74
L =
V
L
= 0.222H
So Transformer
X
R
= 40 (4.1)
48 Chapter 4. Distribution System
4.3.4 Distribution Line Model
The system under consideration have three distribution lines, conductor area of the transmis
sion line is chosen based on short circuit MVA and its continuous current carrying capability.
Resistivity () of the line is considered as 1.25 times of the given copper resistivity(at20
o
).
X
t
= 0.4/km (for 33kV Line)
L =
X
t
l
R =
1.25l
A
c
llength of the conductor, AcArea of the cross section of the conductor
Line No Length(km) Inductance L(mH) Resistance ()
Line 1 4 5.1 1.2
Line 2 5 6.4 1.6
Line 3 7 8.9 4.9
Table 4.1: Line Parameter
4.4 Distribution System Simulation
Distribution system as shown in the Fig.4.2 (excluding PV system) is have only linear ele
ments, so for the better simulation performance, full system is model as a state space equation
and state variable block of MATLAB/SIMULINK is used to implement the system. Loads
are modelled separately not included in the state space.
State variable matrix for the distribution system
x = Ax +Bu (4.2)
y = Cx +Du
4.4. Distribution System Simulation 49
Where
State V ariable x =
_
_
v
f
v
1L
v
s
v
2L
i
g1
i
g2
i
g3
i
s
_
_
State Space A =
_
_
0 0 0 0
1
C
f
0 0 0
0 0 0 0
1
C
1L
1
C
1L
0 0
0 0 0 0 0
1
C
s
1
C
s
1
C
s
0 0 0 0 0 0
1
C
2L
0
1
L
g1
1
L
g1
0 0
R
g1
L
g1
0 0 0
0
1
L
g2
1
L
g2
0 0
R
g2
L
g1
0 0
0 0
1
L
g3
1
L
g3
0 0
R
g3
L
g3
0
0 0
1
L
s
0 0 0 0
R
s
L
s
_
_
Input Matrix B =
_
_
1
C
f
0 0 0
0
1
C
1L
0 0
0 0 0 0
0 0
1
C
2L
0
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0
1
L
s
_
_
Input V ector u =
_
_
I
sa
I
1L
I
2L
v
sg
_
_
Required state variables can be taken as the output, so based on the analysis & state variable
of interest, C & D matrix can be formulated.
Chapter 5
System Simulation Method
PV system with distributed network is simulated in the MATLAB/SIMULINK environment,
since the model have many subsystem and dynamic elements, simulation takes much time to
simulate the system. PV system simulation shown in the literatures[25] are using PSCAD
and other packages with its inbuilt blocks, so always validity of the results are questioned.
So there is a need for reducing the simulation time.
5.1 Issues in System Simulation
Distribution voltage at PCC (bus 1) takes much time to reach steady state, because
during initial time, distribution system act as an voltage source connected to an RLC
circuit
Due to oscillation in the distribution system, PV system, which sense the PCC voltage
takes much time to reach steady state
PLL have to reach the steady state before PV system starts its operation, since MAT
LAB try to execute all the blocks simultaneously, so PV system oscillates for long
time
MPPT must start with the better voltage otherwise, due to MPPT dynamics, system
will take much time to track the maximum power point.
Less DC bus capacitor initial voltage results in an oscillations in the current & voltage
controller, this leads to an injection of high current to the system and slower the process
of reaching steady state.
50
5.2. Methods to Improve Simulation Performance 51
Both voltage & current controllers have saturation blocks to limit the maximum current
(rate of change in the current), since during initial condition many system parameters
are oscillating, limitation of the controller results in, inability of the controller to speed
up the settling process.
Since switching frequency set to 5kHz and sampling time to 0.1ms, xed point MAT
LAB simulation time must kept very low to record the dynamics, this leads to very
high simulation time, most often Computer/hardware failed to simulate the model.
Power balance added in the voltage loop leads to an initial DC bus voltage oscillation,
any change in the DC bus voltage leads to the change in the present MPPT voltage,
this makes system to stabilize at slower rate, since MPPT is relatively slower than the
other dynamic elements in the system
There is a need to reduce the initial settling time of the system, because many disturbances
must be simulated, while system is in the stable condition. Methods employed to reduce
settling time of the system must not aect the performance (dynamics) of the system.
5.2 Methods to Improve Simulation Performance
The above problems are overcome by using the below method
Making the MPPT voltage constant (near to MPP voltage) for few ms, this will avoid
the dynamics due to MPPT
Disabling the saturation block of the current and voltage controller for few ms will
speed up the controller action.
I
q
controller limiter block which decides the reactive power must be disabled during
the initial operation
Power balance feedback is disabled for rst few ms, to allow MPPT to track the
maximum power point.
All elements are modelled in mathematical equation, without using inbuilt MATLAB
power system blocks, this leads to an improved simulation time and performance.
52 Chapter 5. System Simulation Method
5.3 System Simulation
Simulation method must satisfy the below requirement
Simulation must take less time
It must be scalable to any distribution network
User friendly
Must adopt standard input format
Here System is divided into two parts
Non Linear PV system
Linear Distribution System
Distribution Line
Load
Nonlinear PV system is fully modelled as shown in the Fig.3.15 and linear distribution
system is modelled by using state space(Eq.4.2) equation, which explained in the last chapter.
From the Fig.5.1
PV system inject current based on the voltage at PCC, so output equation matrix
(Y = Cx+Du) C must formed such that bus 1 voltage (V
f
) must be one of the output
of state space and it gives the voltage information to the PV system.
PV system is injecting current to the distribution system, so PV system current
(i
sa
, i
sb
, i
sc
) must be added as the input to the state matrix (B & u) as shown in
the eq.4.2.
based on the load bus voltage (V
1L
& V
2L
), load will inject the current (I
1L
&I
2L
)) to
the distribution system, so load voltages V
1L
& V
2L
dened as output in the distribu
tion system state space, and load current (i
1L
& i
2L
must be dened as input to the
distribution system state space.
5.4. Simulation of Line faults 53
Figure 5.1: System Implementation in MATLAB/SIMULINK
In proposed method can be scalable to any distribution network, this can be done by chang
ing only the state space parameters (A, B, C&D), general program can be written in the
MATLAB, which provides the standard input to the simulation. Input le must provide all
important data/informations(e.g. controller, distribution system Panel....etc) to the simu
lation. Changing the parameter is quite easy in this kind of simulation. Since full system
is implemented in simple MATLAB/SIMULINK blocks, simulation is quite faster. So the
above simulation requirements are all achieved by using this proposed method as shown in
the Fig.5.1.
5.4 Simulation of Line faults
For creating faults at any part of distribution line during simulation, need modication in the
above state space, since MATLAB wont allow changes in the state space parameter during
simulation. Fault is simulated by adding a small ctitious capacitor in the network. From
54 Chapter 5. System Simulation Method
Figure 5.2: Simulation of Line Fault
the Fig.5.2, current state variable of the faulted line is split into two state variable based on
the fault point. If fault happen in the line2 then state variable i
g2
is split into i
g21
& i
g22
,
another state variable v
t
also included into the system, during fault v
t
made equal to zero
as shown in the g.5.2
5.5 Simulation of Loss of Grid (Islanding)
For simulating loss of grid, either Line1 or Line2 have to be open circuited, since islanding
studies have to be carried out for dierent local loads, so loss of line2 is simulated. Simulation
5.5. Simulation of Loss of Grid (Islanding) 55
of line outage need modication in the above state space, which is dierent from fault
simulation. This is done by taking the line (of interest)state variable (i
g2
) outside the state
space equation and implemented by using transfer function. Input to the Line2 transfer
function is the both end bus voltages (V
1L
&V
s
) (bus2 and bus 3). Based on this voltages
line transfer function injects current (i
g2
) to the distribution state space. So modied state
space equation will have one state variable less than steady state matrix and will have one
additional input (i
g2
) as shown in the Fig.5.3. Loss of line is simulated by forcing the current
i
g2
= 0.
Figure 5.3: Simulation of Loss of Grid(Islanding)
Chapter 6
Impact of DG on Area EPS
Standard IEEE 1547 deals with the performance requirements of the grid connected distribu
tion system (upto 10MW, 60Hz) interface. 29 important issues (both technical & managerial)
are discussed in this standard. impacts are fully depend on the amount of DG/DR added to
the Area EPS. Issues can be widely broadly classied into
Power Quality
DC current, Harmonics, Flicker etc
Management
Process, Agreement, Liability etc
Safety
Desensitization of Protection Device, Islanding
Damage to Area EPS & Customer Equipment
Overvoltage (Voltage Regulator), Shunt Capacitance switch, Circuit breaker, Fre
quency control, Network Protector, Reclosing etc
Reliability
Fuse, breaker Coordination, Time to repair etc
Stability
Fault, Voltage stability, Frequent Trip of DG etc
56
6.1. Tripping Parameters for Distribution System 57
6.1 Tripping Parameters for Distribution System
During abnormal conditions in the grid (e.g Voltage and Frequency variation at PCC) Pho
tovoltaic system must cease to energize the grid within a pre determined time as given in
Table.6.1 by the standard IEEE 1547. This information is very important for stability and
coordination study.
Voltage Range(% of base Voltage) Clearing Time(s)
V<50 0.16
50V<88 2
110<V<120 1
V 120 0.16
Table 6.1: System response to abnormal voltages
6.1.1 Thumb Rule for Islanding
Aggregate capacity of DG, which is less then one third of total minimum load of EPS
then ordinary islanding detection techniques is sucient
Simple protection (Voltage) sucient for DG with capacity is less than 50% of the
total load in local EPS minimum annual integrated electrical demand for a 15 minute
time period
Impact of DG is less, when it account only 15% of annual peak load(simple impact
study is sucient)
Less chance of interference (Reliability), if DG not contribute more than 10% of max
imum fault current
Circuit Breaker is safe, if fault current (DG +Area EPS) < 8590% short circuit rating
DG rating less than 30kW than simple study is sucient
Chapter 7
Simulation Results
PV system model is simulated for dierent grid disturbances and eect of each PV system
dynamic element on a particular disturbances are analysed. Important dynamic elements in
the PV systems are
MPPT (Maximum Power Point Tracker)
Voltage Controller (with DC bus capacitor)
Phase Locked Loops (PLL)
Current Controller (with ltet inductor)
7.1 Case 1: Normal Operation of Full PV System
PV System is operated at the irradiance of 1000W/m
2
, &T
call
= 25
o
, frequency of 50Hz and
grid Voltage of 1pu at PCC. Performance of the PV system model found satisfactory and
matches with the actual data.
58
7.1. Case 1: Normal Operation of Full PV System 59
Figure 7.1: PV System Response for MPPT perturbation =0.1, E = 1000W/m
2
, T = 25
o
C
Figure 7.2: Distribution system Line Currents i
g1
, i
g2
, i
g3
, & i
s1
for MPPT=0.1, E =
1000W/m
2
, T = 25
o
C
60 Chapter 7. Simulation Results
Figure 7.3: Grid Voltage at PCC, Phase a Current & Voltage at PCC & i
ga1
, i
ga2
for
MPPT=0.1, E = 1000W/m
2
, T = 25
o
C
7.2 Case 2: Change in Irradiance from 1000W/m
2
to
500W/m
2
Irradiance E is changed from 1000W/m
2
to 500W/m
2
at time t=0.25s. Since MPPT is the
important dynamic element for environmental change. Simulation is done for two dierent
MPPT perturbation (0.1 & 0.01). Study is carried out on PV system with the capacity of
500kW.
7.2.1 MPPT =0.1
The below gure shown the response of dierent PV elements for irradiance change
Concluding remarks
As per Fig.7.4, PLL dont have any eect on dynamics of the PV system during climatic
disturbances
MPPT, current and voltage controller is acts as an important role in system dynamics.
Here bandwidth of the current control loop is around 2500rad/s, and voltage controller
7.2. Case 2: Change in Irradiance from 1000W/m
2
to 500W/m
2
61
Figure 7.4: (a)Change in Irradance, (b) Three Phase Output Current of the PV System
MPPT=0.1
Figure 7.5: PV System Response for Irradiance Change (1000W/m
2
to 500W/m
2
)
MPPT=0.1
is around 1100rad/s. so system will need minimum of 0.4ms to settle.
There will be the slight dip in the capacitor voltage due to the pre disturbance(t=0.25s)
62 Chapter 7. Simulation Results
Figure 7.6: (a) Transient Portion of PV Current , Power & System Output Current
MPPT=0.1
current injected to the system (which is higher than the PV current), after that voltage
controller dynamics will be dictated by MPPT as shown in Fig.7.4.
Eect of this disturbance in the grid is very less, since grid is stronger and local load
also less than PV system, so no signicant change found in grid voltages.
To study the eect of MPPT , perturbation is changed from 0.1 to 0.01.
7.2.2 MPPT=0.01
In this case, it is found that maximum power is tracked after 2.45s as shown in the Fig.7.7,
so MPPT took much time to respond, since PV system rating is less (500kW) no much eect
is felt in the grid side, but high power PV will aects the performance of the weak grid.
7.3. Case 3: Grid Voltage Disturbance 63
Figure 7.7: (a) Transient Portion of PV Current ,Voltage, Power & System Output Current
MPPT=0.01
7.3 Case 3: Grid Voltage Disturbance
Response of photovoltaic plant for variation in the grid voltages at PCC is analysed by using
MPPT perturbation=0.1
7.3.1 Grid Voltage Variation from 1pu to 0.85pu at t=0.25s
Simulation is done for irradiance=1000W/m
2
and T
cell
= 25
o
Concluding Remarks
MPPT have very less inuence on the grid voltage disturbances as shown in the
Fig.7.11. This is only true when voltage controller is employed to control the DC
bus voltage
PLL act as an important role in system dynamics, since PLL is the element which
informs the PV system about the grid voltage vector. Based on the voltage vector PV
injects power to the grid. So PLL act as an important role in PV system dynamics for
grid voltage variation as shown in the Fig.7.8
Voltage and current controller also have inuence in the dynamics.
64 Chapter 7. Simulation Results
0 0.1 0.2 0.3 0.4 0.5
499
500
501
502
503
0 0.1 0.2 0.3 0.4 0.5
590
600
610
620
time (s)
P
V
C
u
r
r
e
n
t
I
p
v
(
A
)
time (s)
P
V
P
o
w
e
r
P
p
v
(
k
W
)
(a) PV Power during Voltage dip from 1pu to 0.85pu
(a) PV Current
P
V
V
o
l
t
a
g
e
V
p
v
(
V
)
0 0.1 0.2 0.3 0.4 0.5
810
820
830
840
time (s)
P
L
L

V
s
d
(
V
)
0 0.1 0.2 0.3 0.4 0.5
0
500
1000
time (s)
(c) PV Voltage
(c) PLL Output V
sd
Figure 7.8: Transient Response of the PV system for Grid Voltage Variation 1pu to 0.85pu
Figure 7.9: Grid Voltage (varying from 1pu to 0.85pu)at PCC and PV System Current
7.3. Case 3: Grid Voltage Disturbance 65
Figure 7.10: Transient Portion of PV system Elements for Grid Voltage Variation of 1pu to
0.85pu
Figure 7.11: PV Voltage Transient for Dierent MPPT Perturbation during Grid Voltage
Variation of 1pu to 0.85pu
66 Chapter 7. Simulation Results
To inject the all PV power to the grid, PV system output current will increases with
decrease in grid voltage at PCC
7.3.2 Grid Voltage Variation from 1pu to 1.1pu at t=0.25s
Simulation is done for irradiance=800W/m
2
and T
cell
= 25
o
to inject the all PV power to the grid, PV system output current will decrease with increase
Figure 7.12: Transient Response of the PV system for Grid Voltage Variation 1pu to 1.1pu
in grid voltage at PCC, other remarks are as same like previous one.
7.4 Case 3 : Change in Grid Frequency
Change in grid frequency from 50 to 45Hz at time t=0.25s is simulated under climatic
conditions E = 1000W/m
2
& T
cell
= 25
o
.
It observed that only PLL and current controllers have considerable eect on dynamics
behaviour of the system. PV system able to track frequency within a half cycle as shown
in the Fig.7.14
MPPT & voltage controller will have very Less impact
7.5. Case 4: Three Phase Fault 67
Figure 7.13: Grid Voltage (varying from 1pu to 1.1pu)at PCC and PV System Current
Figure 7.14: PLL Dynamics for Grid Frequency Variation from 50Hz to 45Hz
7.5 Case 4: Three Phase Fault
Three phase bolted fault is created in the dierent lines in the distribution network as shown
in the Fig.7.15 at time t=0.25s.
68 Chapter 7. Simulation Results
Simulation is carried out for dierent controllers and MPPT perturbation
R
g1
L
g1
R
1L
L
1L
R
g2
R
g3
L
g3
C
1L
Load1
C
2L
Grid
L
2L
R
2L
Load2
C
s
R
s
L
s
Transformer
Distribution System
C
f
33 kV
Line1 Line2 Line3
Bus 1
Bus 2
Bus 3
Bus 4
V
sg
V
f
V
1L
V
sg
V
2L
L
g2
i
sa
PV system
Current
injection
Line4
i
g1
i
g2
i
g3
i
sg
i
1L
i
2L
PCC
Figure 7.15: Single Line Schematic of Distribution System
7.5.1 Three Phase Fault in Line1
Three phase fault is created on the transformer high voltage secondary terminal, under the
climatic condition of E = 1000W/m
2
and T
cell
= 25
o
.
Figure 7.16: Transient Response of the PV system During Three Phase Fault at Line 1
7.5. Case 4: Three Phase Fault 69
Figure 7.17: Distribution System Line Currents During Fault at Line1
Figure 7.18: PV System Fault Current & Grid Voltage at PCC During Fault at Line1
70 Chapter 7. Simulation Results
Figure 7.19: Transient Portion of PV System Current & DC Bus Voltage During fault at
Line1
Concluding Remarks
To protect power electronic components and system , PV system output current is
limited to 1.5pu as shown in Fig.7.18.
Fault requires reactive power, but PV generate active power. This active power must
be dumped inside the capacitor which inturn rise the capacitor voltage
Increased capacitor voltage leads to less PV current I
pv
and power P
pv
(refer the PV
Cell characteristic).
Grid Supplies high current to the fault as shown in the Fig.7.17
PV system supplies 1.5pu fault current, which is around 1% of total fault current
During fault, DC bus voltage is decided by power ow and MPPT has less control
over DC bus capacitor voltage. So MPPT wont aect dynamics of the system during
bolted fault condition.
At the instant of fault, response of the PV depends on controller design, PWM tech
nique and DC bus capacitor.
7.6. Case5: Loss of Grid  Islanding 71
7.5.2 Three Phase Fault at Line1 with Inverter
The above fault simulation is done by considering average inverter model. Simulation also
tried with ideal inverter (refer Fig.7.20). Both the cases results are matching. Fault is
created in the Line1 at time t=0.3s.
Figure 7.20: PV System Fault Current & Grid Voltage at PCC During Fault at Line1 with
Inverter
7.6 Case5: Loss of Grid  Islanding
Loss of grid is simulated by forcing current i
sg
= 0 at time t=0.25s under irradiance E =
1000W/m
2
) & T
cell
= 25
o
. When grid is lost, PV system operates in isolation which normally
called as islanding. Islandng conditions are not acceptable by standard due to various safety
and quality issues. Lots of islanding detection methods are proposed, which normally sense
voltage, frequency or rate of change in frequency(ROCOF) and rate of change in voltage
angle. Simulation is done for dierent Load1 (refer Fig.7.15) values, with a quality factor
of 1.7. Island must be detected and PV system must be disconnected within 2s.
72 Chapter 7. Simulation Results
7.6.1 Loss of Grid, Load1 = 750kW, Q
f
= 1.7 Load2 = 70kW at t =
0.25s
Grid loss simulated at time t=0.25s
Figure 7.21: Transient Response of the PV system During Loss of Grid
Figure 7.22: Distribution System Line Currents During Loss of Grid
7.6. Case5: Loss of Grid  Islanding 73
Figure 7.23: Grid Voltage at PCC & PV System Current During Loss of Grid
Figure 7.24: Grid Voltage Vector & Grid Frequency During Loss of Grid
Concluding remarks
After disconnection of the grid, Power generated by the PV is much higher than the
74 Chapter 7. Simulation Results
connected load. Since PV controllers try to ensure constant power injection. This
active power mismatch leads to increment in the PCC voltage, in this case voltage
increased to 1.02pu, so simple passive voltage detection antiislanding method failed
to detect the island.
Since some of the reactive power(VAR) is supplied by the gird, loss of grid results in
loss of reactive power source. This active power mismatch leads to decrement in the
frequency. In this case frequency is decreased from 1pu to 0.85pu. so simple passive
frequency & ROCOF detection antiislanding method can detect the island.
Frequency of the system settles in the resonant frequency of the load(including distri
bution line) to ensure UPF operation.
PLL, current and voltage controllers act as an important role in the transient response.
7.6.2 Loss of Grid, Load1 = 500kW, Q
f
= 1.7 Load2 = 70kW at t =
0.25s
Fig.7.25 & Fig.7.28 shows the performance of the system during Loss of grid with Load
1=500kW. Since active power mismatch is high, so even simple passive voltage based anti
islanding method detects the island(refer Table.7.1)
Figure 7.25: Grid Voltage at PCC & PV System Current During Loss of Grid for Load
1=500kW
7.6. Case5: Loss of Grid  Islanding 75
7.6.3 Loss of Grid, Load1 = 1300kW, Q
f
= 1.7 Load2 = 70kW at t =
0.25s
Fig.7.26 & Fig.7.28 shows the performance of the system during Loss of grid with Load
1=1300kW. Since connected load is higher than the PV capacity. This active power mismatch
leads to decrement in the PCC voltage. In this case voltage decreased to 0.82pu. Power
mismatch is high, so even simple passive voltage based antiislanding method detects the
island(refer Table.7.1).
Figure 7.26: Grid Voltage at PCC & PV System Current During Loss of Grid for Load
1=1300kW
7.6.4 Loss of Grid, Load1 = 900kW, Q
f
= 1.7 Load2 = 70kW at t =
0.25s
Fig.7.27 & Fig.7.28 shows the performance of the system during Loss of grid with Load
1=900kW. Since connected load is slightly dierent from the PV capacity. Power mismatch
is low, so simple passive voltage based antiislanding method failed to detects the island(refer
Table.7.1).
76 Chapter 7. Simulation Results
Figure 7.27: Grid Voltage at PCC & PV System Current During Loss of Grid for Load
1=900kW
7.6.5 Summary of Voltage & Current Data
For all the cases Load2 = 70kW, quality factor (Q
f
) = 1.7 , irradiance E = 1000W/m
2
and
T
cell
= 25
o
maintained constant. With grid connected to the PV system(before disturbance)
peak voltage (V
PCC
) = 340.2V , Peak PV system output current(for 500kW) I
sa
= 979A,
grid voltage vector V
sd
= 510V and grid frequency = 314rad/s. The quantities mention
in the table.7.1 is measured at time t=0.6s (after loss of the grid)..
Load1 (kW) Peak V
PCC
(V ) Peak I
sa
(A) Ang. Frequency ( rad/s) Voltage Vector V
sd
(V )
500 404 829 241 595
750 347 965 266 519
900 326 1033 275 486
1300 282 1183 291 423
Table 7.1: Voltage at PCC & PV System (500kW) Output Current
7.7. Case 6: Study of Nuisance Trip for Fault Ride Through Enabled PV System 77
Figure 7.28: Grid Voltage Vector & Grid Frequency During Loss of Grid with Dierent
Connected Load
7.7 Case 6: Study of Nuisance Trip for Fault Ride
Through Enabled PV System
In the distribution network as shown in the Fig.7.15, fault in Line3 cannot trip the PV
system. For the high penetrated PV system, loss of PV supply leads to power system
instability. Many governments agencies are insisting on mandatory fault ride through capable
PV system to save the overall power system. So stringent studies have to be carried out for
high penetrated PV plants.
From the Fig.7.31, post fault voltage at the PCC is 205Vpeak, which around 60% of the
rated voltage, but as per the IEEE Std.1547 as shown in the table.6.1. For voltages(at PCC)
between 50% to 88% , PV have to cease the energization within 2s. If fault rectied within
2s, PV system will regains its normal state. Otherwise PV system will get disconnected from
the network, this leads to loss of generation and power system instability. So while planning
the PV system, detailed analysis have to be done on fault ride through and its eect on
system stability.
78 Chapter 7. Simulation Results
Figure 7.29: Transient Response of the PV system During Fault at Middle of Line3
Figure 7.30: Distribution System Line Currents During Fault at Middle of Line3
7.7. Case 6: Study of Nuisance Trip for Fault Ride Through Enabled PV System 79
Figure 7.31: Grid Voltage at PCC & PV System Current During Fault at Middle of Line3
Chapter 8
Conclusions
This project was aimed to study the behaviour of the photovoltaic system during both
climatic and grid disturbances. This study is important, because of increasing penetration
of the grid connected PV generation. Since not much study were done on fault behaviour of
inverter based generation, particularly fault response of the PV system.
Manufacturer data sheet based PV panel is modelled. Constant current control based(UPF)
interface is used to connect the PV system to the grid. Distribution system which assumes
innite grid as a source were modelled using state space matrix. The overall PV system is
modelled such that it takes very less time for simulation. The given system is implemented
by using MATLAB/SIMULINK. System model also takes care of unbalanced conditions.
This model can be used for any number of distribution system.
System model is used to simulated the dierent disturbances. Eect of each PV system
element for dierent disturbances were analysed. Those are
System with voltage controller showing better performance. Voltage controller reducing
the eect of the MPPT during the grid voltage disturbances
MPPT act as an important role, During large DC voltage disturbances under nonfault
conditions.
PLL dynamics act as an very important role in non fault grid disturbances(voltage and
frequency disturbances)
PV system fault current is limited to 1.5 to 2pu, so PV panel contribution to the
system fault is very less(depends on the penetration level)
80
81
MPPT have less eect on dynamics of the system during three phase bolted fault
condition.
MPPT act as an important role in system dynamics during climatic disturbances.
MPPT have much eect during
Absence of voltage controller
Large climatic disturbance
System with high PV penetration
PLL have very less inuence on system response, during climatic disturbances
Current and voltage controller have eect in system dynamics during all possible grid
and climatic disturbances. Current controller bandwidth is depends on the lter in
ductance(which act as a plant) and PI controller.
Change in controller parameter aects the transient response of the system during
fault, faster controller have less fault current overshoot and slower controller have high
fault current overshoot.
Saturation block(Limit) in the PI controller also aect the transient response of the
PV system.
Transient response of the PV system is depends on the design parameters(like con
troller, lter design, DC bus Capacitor). So it is not possible to standardize the
transient response of the system.
MPPT is helping the system to reach stable condition.
PV system behaviour during islanded conditions are also studied, islanding detection
becomes very dicult, if load demand is equal to PV generation
Since PV system is operating in UPF. Particularly for PV system, passive frequency
detection and ROCOF detection based antiislanding method is better performing than
voltage detection based antiislanding method.
82 Chapter 8. Conclusions
Nuisance trip study for fault ride through capable PV system is very important for
maintaining the power system stability. This study will be very critical for high pen
etrated System. There is no standard method is available to carry out this study for
high penetrated PV systems. So there is need for standardization of PV system fault
analysis and coordination with the protection devices.
Appendix A
MATLAB Simulation
A.1 Over all system
Grid
powergui
Cont inuous
N
A
B
C
PV System1
V
s
a
V
s
b
V
s
c
f
a
u
l
t
I
a
I
b
I
c
PV System
V
s
a
V
s
b
V
s
c
f
a
u
l
t
I
a
I
b
I
c
Distribution Network
A
B
C
Visa
Visb
Visc
R
Y
B1
N
Add2
Add1
Add
Figure A.1: Over all System
83
84 Appendix A. MATLAB Simulation
A.2 Distribution Network
Visc
3
Visb
2
Visa
1
N
4
B1
3
Y
2
R
1
Voltage Sensor
Vsa
Vsb
Vsc
R
y
B
N
Nt2
K
Nt1
K
Nt
K
C Phase
I
Vs
Vinv
B Phase
I
Vs
Vinv
A Phase
Isa
Vs
Vinv
Isc
3
Isb
2
Isa
1
Figure A.2: Distribution Network Implementation
A.3. Distribution System  Phase a 85
A.3 Distribution System  Phase a
Vinv
1
To Workspace8
Is
To Workspace7
Vpcc
To Workspace6
Ig3
To Workspace5
Ig2
To Workspace3
V2L
To Workspace2
Vs
To Workspace1
V1L
StateSpace11
x' = Ax+Bu
y = Cx+Du
Load2
x' = Ax+Bu
y = Cx+Du
Load1
x' = Ax+Bu
y = Cx+Du
Fault
Ig11
Ig12
Vt
Vs
2
Isa
1
Figure A.3: Distribution System  Phase a implementation
86 Appendix A. MATLAB Simulation
A
.
4
P
V
S
y
s
t
e
m
Ic 4 Ib 3 Ia 2
1
V
o
lta
g
e
C
o
n
tro
lo
r
V
dcref
KV
fault
P
ow
er
Idvol
V
dc
T
ra
n
s
fe
r
F
c
n
4
1
0
.0
0
1
s
+
1
T
o
W
o
rk
s
p
a
c
e
9
fre
q
T
o
W
o
rk
s
p
a
c
e
8
v
m
s
u
T
o
W
o
rk
s
p
a
c
e
7
p
m
s
u
T
o
W
o
rk
s
p
a
c
e
6
te
m
p
T
o
W
o
rk
s
p
a
c
e
5
irra
d
T
o
W
o
rk
s
p
a
c
e
4
p
o
w
e
r
T
o
W
o
rk
s
p
a
c
e
3
tim
e
T
o
W
o
rk
s
p
a
c
e
2
p
llv
o
l
T
o
W
o
rk
s
p
a
c
e
1
Ip
v
c
t
S
w
itc
h
3
>
=
S
u
b
s
y
s
te
m
2
In1
O
ut1
S
u
b
s
y
s
te
m
Ipv
V
o
V
sd
IdIqV
sq
V
c
S
te
p
2
S
c
o
p
e
8
S
c
o
p
e
4
S
c
o
p
e
3
S
c
o
p
e
1
3
S
c
o
p
e
1
2
S
c
o
p
e
1
1
S
c
o
p
e
1
S
c
o
p
e
R
e
f
G
e
n
e
ra
to
r
V
m
pp
Im
pp
V
sd
Idref
P
V
P
a
n
e
l
T
c
V
N
s
N
p
T
c
I
Im
p
s
V
m
p
s
D
e
lta
V
ET
a
N
O
C
T
Is
c
V
o
c
V
m
in
bV
Im
p
0
V
m
p
0
T
_
C
e
ll
P
m
a
xIL
P
m
s
u
P
L
L
V
R
0
V
R
1
V
R
2
Freqw
A
ngle
V
sd
V
sq
sin1
cos1
V
sd2
V
sq2
V
a
V
b
M
P
P
T
Ipv
V
pv
V
m
pp
Im
pp
P
m
ax
In
v
e
rte
r
C
u
rre
n
t
IaS
Fa
IbS
Fb
IcS
Fc
Iinverter
In
d
u
c
to
r
&
C
u
r
r
e
n
t S
e
n
s
o
r
2
Id
v
Id
v
1
Id
v
2
Id
v
3
Id
v
4
Id
v
5
Ia
2
Ib
2
Ic
2
Ia
1
Ib
1
Ic
1
G
o
to
9
[Iq
]
G
o
to
8
[Id
]
G
o
to
7
[G
r]
G
o
to
6
[c
o
s
1
]
G
o
to
5
[s
in
1
]
G
o
to
4
[Id
re
f]
G
o
to
3
[Id
v
]
G
o
to
2
4
[p
o
w
e
r]
G
o
to
2
3
[V
s
d
2
]
G
o
to
2
2
[V
s
q
2
]
G
o
to
2
1
[Id
c
]
G
o
to
2
[V
s
q
]
G
o
to
1
9
[Iy
l]
G
o
to
1
8
[Irl]
G
o
to
1
7
[G
b
]
G
o
to
1
6
[G
y
]
G
o
to
1
5
[Ip
v
c
]
G
o
to
1
4
[V
o
fa
u
lt]
G
o
to
1
3
[K
]
G
o
to
1
2
[rc
]
G
o
to
1
1
[w
]
G
o
to
1
0
[V
s
d
]
G
o
to
1
[V
o
]
G
a
in
3
K

F
ro
m
9
[V
o
fa
u
lt]
F
ro
m
8
[Id
c
]
F
ro
m
6
[c
o
s
1
]
F
ro
m
5
[V
s
d
]
F
ro
m
4
[s
in
1
]
F
ro
m
3
9
[m
b
n
]
F
ro
m
3
8
[m
y
n
]
F
ro
m
3
7
[p
o
w
e
r]
F
ro
m
3
6
[m
rn
]
F
ro
m
3
5
[Iq
]
F
ro
m
3
4
[Id
]
F
ro
m
3
3
[V
o
]
F
ro
m
3
2
[Ip
v
c
]
F
ro
m
3
1
[V
s
q
]
F
ro
m
3
0
[V
s
q
]
F
ro
m
3
[K
]
F
ro
m
2
9
[V
s
d
]
F
ro
m
2
7
[Irl]
F
ro
m
2
6
[Iq
]
F
ro
m
2
5
[c
o
s
1
]
F
ro
m
2
4
[s
in
1
]
F
ro
m
2
3
[Id
re
f]
F
ro
m
2
2
[Id
v
]
F
ro
m
2
1
[G
y
]
F
ro
m
2
0
[V
o
]
F
ro
m
2
[V
o
]
F
ro
m
1
9
[V
s
q
]
F
ro
m
1
8
[Iy
l]
F
ro
m
1
7
[G
r]
F
ro
m
1
6
[G
b
]
F
ro
m
1
5
[Ib
l]
F
ro
m
1
4
[Ip
v
c
]
F
ro
m
1
3
[V
s
d
]
F
ro
m
1
2
[V
s
d
]
F
ro
m
1
1
[Id
]
F
ro
m
1
0
[w
]
F
ro
m
1
[V
o
]
F
ro
m
[V
o
]
D
iv
id
e
3
C
u
r
r
e
n
t
C
o
n
t
r
o
llo
r
Id
r
e
f
Id
v
IdwIqV
d
c
s
in
c
o
s
V
s
q
2
V
s
d
2
m
r
n
m
y
n
m
b
n
G
r
G
y
G
b
V
o
C
o
n
s
ta
n
t9
5
1
.6
C
o
n
s
ta
n
t8
4
8
C
o
n
s
ta
n
t7
7
.5
C
o
n
s
ta
n
t6
0
.2
5
C
o
n
s
ta
n
t5
4
2
C
o
n
s
ta
n
t4
5
.1
3
C
o
n
s
ta
n
t3
C

C
o
n
s
ta
n
t2
0
.1
4
3
C
o
n
s
ta
n
t1
3
0
.1
0
C
o
n
s
ta
n
t1
2
C

C
o
n
s
ta
n
t1
1
5
.6
1
C
o
n
s
ta
n
t1
0
4
6
C
o
n
s
ta
n
t1
1
1
7
C
o
n
s
ta
n
t
2
0
C
lo
c
k
2
C
lo
c
k
1
C
a
p
a
c
ito
r
V
o
lta
g
e
R
ip
p
le
Iinverter
Ipv
V
ripple
V
ripple1
A
B
C
to
D
Q
IaIbIcsin
cos
idiq
V
s
c
3
V
s
b
2
V
s
a
1
P
V
P
a
n
e
l
M
P
P
T
V
o
l
t
a
g
e
C
o
n
t
r
o
l
l
e
r
C
u
r
r
e
n
t
R
e
f
e
r
e
n
c
e
C
u
r
r
e
n
t
C
o
n
t
r
o
l
l
e
r
w
i
t
h
i
n
v
e
r
t
e
r
P
L
L
P
V
O
u
t
p
u
t
C
u
r
r
e
n
t
C
a
l
c
u
l
a
t
i
o
n
C
u
r
r
e
n
t
s
e
n
s
o
r
P
o
w
e
r
B
a
l
a
n
c
e
D
C
b
u
s
C
a
p
a
c
i
t
o
r
C
u
r
r
e
n
t
R
i
p
p
l
e
D
i
s
t
r
i
b
u
t
i
o
n
S
y
s
t
e
m
V
o
l
t
a
g
e
I
n
p
u
t
F
i
g
u
r
e
A
.
4
:
P
V
s
y
s
t
e
m
I
m
p
l
e
m
e
n
t
a
t
i
o
n
References
[1] www.nrdc.org/international/copenhagenaccords/,National Resource Defense Centre.
[2] Arunachalam. G, Talk on Indian electrical scenario, CSTEP IIScAA popular lecture
series.
[3] www.mnre.gov.in.
[4] www.wikipedia.org.
[5] Zeineldin. H.H, ElSaadany. Ehab F, Salama. M.M.A Impact of DG Interface Control
on Islanding Detection and Nondetection Zones,IEEE Transactions on Power Delivery,
vol. 21, No.3, July 2006.
[6] http://science.nasa.gov/sciencenews/, NASA Science News.
[7] Yun Tiam Tan, Impact on the Power System with a large penetration of PV Genera
tion,Phd. Thesis, Dept. of Electrical Engineering and Electronics UMSIT.
[8] Castaner. Luis, Silvertre. Santiago, Modelling of Photovoltaic System using PSpice,
John Wiley & sons Ltd.
[9] Umanand. L,Lecture notes on Solar Photovoltaic systems,CEDT, Indian Institute of
Science,Bangalore
[10] Anderson. A.J, Final Report for Task 2.0, NREL subcontract No.TAD4 1416601,
Oak Leaf Place (1995 ).
[11] Procedures for temperature and irradiance corrections to measured IV characteristics
of crystalline silicon photovoltaic devices,IEC 891 standard 1987
87
88 References
[12] Blaesser. G. PV Array Data translation procedure,13th ECPVSEC (1995)
[13] OrtizRivera. E.L, Peng. F.Z, Analytical Model for a photovoltaic module using the
electrical characteristics provided by the manufacturers data sheet,Proceedings of the
IEEE PESC 2005, pp.20872091.
[14] Bke Ulrich, A simple model of photovoltaic module electric characteristics,Philips
Research, Germany.
[15] Tiam Tan. Y, Kirschen. D.S, Jenkins. N, A Model of PV Generation Suitable for Sta
bility Analysis. IEEE Transactions on Energy Conversion.
[16] Arun Karuppasamy, Synchronous reference frame strategy based statcom for reactive
and harmonic current compensation,M.Tech Thesis, NITC,IIScBangalore
[17] Anirban Ghosal, Dr. Vinod John, A Method to Improve PLL Performance Under Abnor
mal Grid Conditions,NPEC, 2007. Chung, S.K PhaseLocked Loop for gridconnected
threephase power conversion systems, IEEE Proceedings on Electronic Power Applica
tions, vol. 147, no. 3, pp. 213219, 2000.
[18] Leonhard, W, Control of Electrical Drives, Third edition, New York: Springer, 2001
[19] Parikshith. BC, Integrated approach to lter design for grid connected power convert
ers, M.Sc Thesis, Department of Electrical Engineering, Indian Institute of Science,
July 2009.
[20] Sivaprasad, J.S,Narayanan, G. et. al, Vector control of threephase AC/DC frontend
converter, Sadhana, Vol. 33, Part 5, October 2008, pp. 591613.
[21] Ranganathan. VT,Course Material on Control of Electric Drives,Dept. of Electrical
Engineering, Indian Institute of Science,Bangalore
[22] Evans, P.D, HillCottingham, R.J, DC link current in PWM inverters, IEE Proceed
ings, Vol. 133, Pt. B, No. 4, JULY 1986.
[23] Amirnaser Yazdani, Prajna Paramita Dash, A Control Methodology and Character
ization of Dynamics for a Photovoltaic (PV) System Interfaced With a Distribution
Network, IEEE transaction on Power Delivery, Vol. 24. No 3, July 2009
References 89
[24] Cahier Techinque no.158, Calculation of short circuit currents, Schneider Electric
corporation.
[25] Sidhu T.S, Bejmert D, Shortcircuit current contribution from large scale pv power
plant in the context of Distribution power system protection , IET Conference on RPG,
Sept 2011
[26] IEEE Standard 15472003, IEEE Standard for Interconnecting Distributed Resources
With the Electric Power System