Design of Low-Voltage Distributed Photovoltaic Systems Oriented to Improve Fault Ride through Capability

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 04 Issue: 08 | Aug -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1297
Design of Low-Voltage Distributed Photovoltaic Systems oriented to
improve fault ride through capability
R. V. GOWTHAM1, Dr. P. SUJATHA2
1PG Student Dept. of EEE Engineering College, Anantapur, Andhra Pradesh, India
2Professor Dept. of EEE Engineering College, Anantapur, Andhra Pradesh, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Photovoltaic (PV) systems have gained
prominence as economically-viable, long-term alternatives to
conventional, non-renewable sources of energy. The optimum
design characteristics that Distributed Generation
photovoltaic units incorporated in order to meet the
requirements of Low-Voltage Ride Through Capability
(LVRTC) are evaluated in this paper. Without the optimal
design of DG-PV units that are connected to low voltage
distribution networks, certain limitations regarding
performance rises in the network transient phenomenon
condition. The analysis evaluates that, with appropriate
selection of equivalent interconnecting reactance XDG in the
oversized interfaced inverters of DG-PV in merging with high
penetration levels yields satisfaction of LVRTC demands
without violating the protection limits of the network.
Furthermore, along the distribution lines, uniform dispersion
is can be observed concerning the demand of LVRTC for
voltage selectivity. Finally, considering the derived outcomes,
optimum design of DG-PV about LVRTC withintheframework
of reasonable constraints, can be applied to any low-voltage
distribution network. In order to checkthe performanceofthe
entire network system, addition of a new LV network to the
existing network is done. The results are evaluated and
compared in MATLAB simulation software.
Key Words: Disrtibuted Generation photovoltaic units, low
voltage ride through capability, Photovoltaic system.
1. INTRODUCTION
Due to limited capacity of fossil fuels, the world is
focussing on the usage of renewable power generation in
terms of electricity production and fulfilling its growing
demand. Hence, it will not be a wrong to say that renewable
power generation will be a significantsourceof energyin the
near future. Furthermore,amongmanynotablegreen energy
sources, the world is mainly focussingontheusageofSOLAR
energy to meet their demand. Photovoltaic is the process of
converting sunlight directly into electricity using solar cells.
With increasing consumption of electric energy in the
context of limited conventional energy sources such as coal,
natural gas and crude oil it is necessarytoseek anduseother
energy sources. Hence the utilization of non-conventional
energy is increasing in day to day life. Mainly the world is
focussing on solar energy and wind energy. The majority of
the power can be evaluated using solar energy compared
with wind energy due to certain parameters. Due to some of
considerable advantages like flexible installation,minimum
upkeep etc., Solar energy or Photovoltaic(PV)energy power
plants are nowadays exhibiting globally a notable
development. The grid connected Distributed Generation
Photo Voltaic ( DG-PV) is the most essential assurance ofPV
systems because of some parameters like proven very
efficiency and cost consideration. In spite of that, DG-PV
penetration level (PL) is usually restricted (by distribution
companies) to the level of less than 20% of the substation
installed power due to some constraints. In this way, PV
production impact on the distribution grid operation,
especially in the event of electrical power systemunbalance,
is significantly reduced. However, due to such strict
constraints, the progress of the further development and
incorporation of renewable energy sources (RESs) into the
electrical power systems is limited to certain extent. To get
over this, all RES generators (wind parks, PV systems and so
on) have to acquire operating features akin to that of
conventional power plants (based on synchronous
generators), in order to contribute to a fault event too.
Therefore the standard of Low-Voltage Ride Through
Capability (LVRTC) has been adopted in all types of RES
attached to high- and medium-voltage grid.Inparticular, the
high power level of wind farms favoured the integration of
LVRTC concept in this type of RES and much improvement
has already been achieved.
With the growth in PV plants, especially that are
connected totransmissionnetwork,manycountriesinglobal
research markets started toimplementtheLVRTCschemein
PV units [1]. Regarding the LVRTC scheme,the global energy
markets vary. The Fig. 1 below describes that with the
implementation of LVRTC scheme in many countriessuchas
Germany, Taiwan, North America, Australia, Denmark etc.,
the desired behaviour of the interconnectedunitsincaseofa
voltage drop at their point of common coupling (PCC) is
observed [2]. Particularly the Fig. 1 evaluates that the time
interval that a DG unit has to stay connected to the grid
depends on the voltage dip level, mainly as the voltage
becomes lower, this time interval decreases.

Fig-1: LVRTC schemes of several energy markets
With the aspects stated above, the standard of LVRTC is
extended so as to include the DG-PV units thatareconnected
to the low-voltage network is evaluated with the addition of
a single BUS bar to the existing network. The necessary
inverter over-sizing is thoroughly monitored in order to
meet LVRTC requirements. Hence, the interfaced inverters
are obliged to be coupled with the low-voltage network and
feed the faulty part for time intervals that last more than a
few line cycles. Therefore they have to be dimensioned for a
significantly longer duration of faulty conditions, calling for
more effective thermal management and much higher over
current and overvoltage strength, combinable termedasthe
over-dimensioning (OD) of PV inverters [3]. The analysis of
LVRTC application on LVDG-PV systems focuses especially
on the Penetration Level (PL). Studies about the PL limits of
PV units have been recently presented [4]. The inverter
design is taken into account and the evaluation of PL
parameter is extended in terms of meeting the current
LVRTC requirements satisfactorily. And also, in order to
fulfill the new operational demands and improving
distribution network during disturbances, the necessary
adjustments on the PV units must be incorporated.
In Section 2, through a detailedtheoretical analysisLVDG-PV
design guidelines are presented. In Section 3, the optimum
design of existing network is added with one low voltage
distribution network to check the optimum design is
applicable to any distribution network at higher PL values.
2 LVDG-PV design guidelines according to LVRTC
The main generalised theme of LVRTC scheme with
reference to the fig.1 is to restrict the unnecessary
simultaneous breakdown of multiple generation sources in
the network disturbance condition the generation sources
should be operating (or) connected to low-voltage
distribution network for some distributionofsomeduration
of time from the fault network and operation time is based
w.r.t electrical degrees function. Here the definition of the
term voltage selectivity rises because electrical distance in
above condition is expressed by voltage dip at Distributed
Generator’s Point of Common Coupling (PCC).
One of the important characteristic for the high
expansion of DG-PVs in focus of high penetration level is
generation loss is limited to the near ones that are close to
the faulty network part in case of short duration faults.
The distributed network have a radial structurei.e..,
if case of fault condition occurred in a network path, there is
no alterative distribution path. More over the units that are
closest to the fault networks are effected more. Therefore
according to the LVRTC scheme since there is no alternative
path the DG’s should fed the faulty part. Based on the above
aspects the design of interfaced inverter should be such that
the proper voltage selectivity is achieved so that on feeding
faulty part during the fault interval time, the interfaced
inverters should withstandthehighercurrentsthannominal
currents in order to stay connected with the distributed
network (or) in continuous operation condition. This ideas
of aspect evaluated several commercial inverters that are
supporting the above characteristics during over current
short term disturbances and so on LVRTC.
The performance and featuresoftheDGunitsduringa
disturbance should be nearly as the performance of
synchronous Generator. This can be done by adjusting the
operational characteristics of the synchronous Generator to
the terms of LVRTC so that in case of disturbance the DG
units must withstand higher currents than nominal rated
currents, since they shall be disconnected after atleast
0.15sec (referred to fig. 1) reaching steadyfaultyconditions.
Therefore from the above aspects, it is preferable for the
LVDG-units should behave as voltage sources instead of
constant source in disturbance condition so that they
provide higher than the nominal values of current. The
deviation between transient and steady-state currents can
be controlled with the LVDG-PV units in the networks [5].
There are certain aspects that should be incorporated in
design concept. They are :
1. All LVDG-PV units must behave as a voltage source
in series with a reactance in faulty conditions.
2. The algorithms related to interfaced inverters to
achieve above aspects must redefined. This is
because there should be a suitable response in case
of a high PL value so as to meet the desired voltage
selectivity value. In particular, in order to avoid
unnecessary generation divergences the inverter
signal must retain its pre-disturbance value in case
of short-term disturbances. In long-term
disturbances, the inverter controller is triggered
either to reduce current within safe- limits (or) trip
LVDG-PV units, if the time limit is exceeded.

Fig-2: Single-phase equivalent circuit of LV distribution
network line with DG-PV units
limits (or) trip LVDG-PV units, if the time limit is
exceeded.
3. A Design guide line is presented on the basis of low
voltage network form. Only one distribution line is
considered for analysis. All connected LVDG-PV
units are considered to act as conventional AC
voltage sources based on the above analysis in fault
condition. A three-phase low-voltage network with
DG-PVs, operating under unity power factor in
steady state is considered. If a three-phase short
circuit occurs at the i-bus, then the single-phase
equivalent circuit of Fig. 2 stands.
A similar modeling of distribution network equipped
with DG units for disturbance studies has already been
presented in [6]. The three-phase short circuit is used for
dimensioning the network’s protection equipment, so the
worst case is included. The calculation of busvoltagesaftera
fault is carried out through the superpositionprocess,which
is absolutely equivalent to the classic method (Ybus) [7].
However, it offers a better sense of the radial distribution
network passive elements, which affect the bus voltages
after a three-phase short circuit. Considering a three-phase
short circuit at the i-bus in Fig. 2, the voltage at k-bus is
derived from the sum of contributions from all sources.
Particularly, for k < i (fig 2.1) it can be calculated as
Fig-2.1: Single-phase equivalent circuit of LV distribution
network line with fault at the bus i
Vk,SCi|K i), the following expression stands
Fig-2.2: Single-phase equivalent circuit of LV distribution
network line with fault at the bus i
Vk,SCi|K>i = Vk→k,SCi +Vlv→k,SCi + j→k,SCi (2)
+ j→k,SCi(2)
The calculation of the individual terms in (1) and (2)
arises from the superposition principle of sources.
Equation set for the analytical calculation of bus voltages
after the short circuit are as follows
For kk,SCi|kk,SCi|kk,SCi|j<kk,SCi|k<j<i= EDGj



j
kp SCiRp
LVpSCiLp
SCijDGj Z
ZZ
ZjX 1 ,,
,,
, )/(1
1
(6)
For k> i
Vk->k,SCi|k>i = EDGk
)/(1
1
, SCikDGk ZjX (7)
Vlv->k,SCi|k>i = 0 (8)
Vj->k,SCi|i<j<k = EDGj





1
1 ,,
1,,
, )/(1
1 k
p SCiRp
LVpSCiRp
SCijDGj Z
ZZ
ZjX
(9)
Vj->k,SCi|i<k<j = EDGj



j
kp SCiRp
LVpSCiLp
SCijDGj Z
ZZ
ZjX 1 ,,
,,
, )/(1
1
(10)
Zk,SCi = Zk,L,SCi//Zk,R,SCi//ZLK (11)

It is worth mentioning that if there is no generation or load
at k-bus, the above equations stand if EDGk becomes infinite,
for the specific k-bus. Considering that DG-PVs supply only
active power in steady state, the internal voltage of a DG
connected to the k-bus is modelled as follows
PDGk =
DGk
kkDGk
X
VE sin**||
(12)
Furthermore as the DGs operate under unity power factor,
the following equations stand
PDGk = Vk * IDGk (13)
tan k =
k
DGkDGk
V
IX *
(14)
Combining (14) and (15) the following expressions can be
given
tan k =
Vk
DGkDGk PX
2
*
(15)
Finally, from (12) to (15) the EDGk can be extracted as
EDGk =
)
*
())
*
(sin(*
*
2
1
2
1
tantan
VV k
DGkDGk
k
DGkDGk
k
DGkDGk
PXPX
V
XP


(16)
From equation 16 it is clear that the internal voltage of DG
units depends on bus voltage and DG active power which
steady state parameter at their PCC. From the load flow
analysis the pre fault bus voltages and the LV voltages are
extracted. It is obvious that these voltages changes if PDG
varies, above equations apply only at typical distribution
networks with radial structure
Euations (1)-(16) taking into account, the DG
contribution to a fault current in case of DG connected to the
k-bus is given by
IDGk,SCi =
DGk
SCikDGk
jX
VE ,
(17)
Example on how the above equation set can be used for the
determination of the minimum bus voltage. If the short
circuit takes place at its neighbouring bus minimum voltage
condition occurs at n-bus. Short circuit at n-1 bus is the
worst case because it leads n-bus to its smallest possible
voltage level(excluding the case of short circuit at n-bus
itself). The root mean suare(rms) voltage attheterminal bus
as follows
Vn,SCn-1 = Vn->n,SCn-1 = EDGn
)/(1
1
LVnDGn ZJX (18)
Considering that the internal rms-voltage of LVDG-PV unit
is roughly 1pu (in order to generate power under unity or
slightly leading power factor), we come up with the
marginal minimum rms-voltage at a ‘healthy’ bus
Vmin,rms(pu) =
dmin
1 , dmin = |
LVn
DGn
Z
JX
1 | (19)
Fig-3: Example of the proposed LVDG-PV design concept
for high PL values
Fig.3 depicts Vmin,rms as a function of dmin in the context of a
design example, considering a specific LVRTC scheme (also
presented in the same figure). Minimum voltage level canbe
achieved by setting dmin value in case of a three-phase sort
circuit at neighbouring bus. Specifically from fig.3 the
selection of dmin below 6.5 reassures that voltage at‘healthy’
buses shall always be > 0.15pu and so they have to stay
connected for atleast 0.625ms(more than 30 lines cycles in
50Hz systems). In this way, the philosophy of LVRTC is
served and so impermanent faulty conditions would have
limited impact on the available DG-PV power production.
3 Integration of new LV network to the existing
Distribution network
The optimum design of LVDG-PV systems which is
proposed in existing network(11) mainly depends on the
selection of XDG and PL values in the direction of high
possible values of buses voltages under faultycondition. The
above aspect mainly opposes the impact of the temporary
potential disturbance on the DG-PV power production.
Mainly the optimization process is came out for a network
with equal DG-PV units and it can be easily applied at any
distribution network with increased DG-PV penetration.
In order to prove this optimum design is applicable
to any distribution network here we are adding existing
network with one additional LV network is shown in fig4 to
check overall performances of the system at higher
penetration level.

Fig-4: Layout of distribution network with DG-PV units
with sixth bus
The first step in optimization problems isthedefinitionof an
objective function with one or more variables. Through the
maximisation or minimisation(dependingonthecase)ofthe
objective function, the optimum values for these variables
are extracted. The parameters of the optimisation function
are the XDG reactance and the PL value. In this paper, the
definition of PL is based on the total load demand and is
given by the following equation
PL(%)= (20)
By making the commercial inverters’ supplyofover
current up to 2.8 times the IDG for several milliseconds,when
disturbances are occurring, the factor ODk at k-bus for
quality improvement on autonomous LV-PV systems[8] can
be defined as depicted below:
ODk(%)= (21)
In the optimization process for power quality
improvement or autonomous LV-PV equipment’s a
necessary (ODK) over-dimensioning percentage of DGK
converters in order to meet the LVRTC demands.
Here the existing distribution network isconnected
with one extra LV network to check overall performance of
the system at high penetration levels.
The objective function for the new network and its
constraints are as follows
Objective function
f(XDG, PL) = (XDG, PL)
Constraints ODk<ODlimit at k-bus
The main purpose of objective function is its
maximization, it means in order to satisfy the LVRTC
demands respecting the DG-PV units OD constraints ODlimits
ISC the bus voltages are to be maximize. The optimization
process has been conducted for different ODlimits values. The
actual PV inverter design and control are noted accordingto
the current standards [2],thefaultcurrentsmorethan120%
of the nominal value does not allowed. According to the
EN50160 standards the bus voltages are under steady-state
conditions when PL valuesvariesbetween20to120%[9].In
order to know how this factors affects the voltage selectivity
achievement under high PL values various ODlimits are
studied. In order to integrate new LV network tothe existing
network the system has to maintainreactanceXDG ateach DG
unit same as in the existing network.
Fig-5: XDG optimum value as a function of PL for different
ODlimit
From the fig-5 it is clear that the XDG at each DG unit
has obtained same as in the previous network(11) and
higher ODlimit makes possible the achievement of voltage
selectivity under a wider range of PL values. By taking one
example, if ODlimit is set equal to 100%, high-voltage
selectivity may be achieved under any PL value between the
acceptable ranges.
Fig-6: Percentage of buses with voltage higher than 0.1pu
as a function of PL, with ODlimit being a parameter, in case
of a short circuit at 1LV1 bus

In case of a short circuit at 1LV1 bus fig-6 presents the
percentage of buses with voltage higher than 0.1pu as a
function of PL, with ODlimit being a parameter.Shortcircuit at
1LV1 bus is the worst possible case regarding the voltage
levels. The results in fig-6 agrees with the ones in fig-5,
highlighting the fact that as ODlimit increasesmorebusesmay
preserve adequately high-voltage levels.. Assuming thatthis
disturbance is temporary, the selection of ODlimit and the
respective optimum XDG value can be set so as to preserve a
critical amount of DG-PVs in operation,especiallyforhighPL
values.
Fig-7: Impact of optimum XDG and PL values on the ratio
of bus short-circuit currents in line 1, in case of a three-
phase short circuit at 1LV5 bus (worst case)
Finally, an important outcome of this optimization
process is that the aforementioned results,regardingthe XDG
optimum value, do not remarkably violate the protective
limits of the distribution network. This is shown in Fig-7,
where the ratio of 1LV5 bus short-circuit current is
presented as a function of the PL value, with ODlimit being a
parameter. Particularly, the ISC varies between 1 and 1.35.
The ISC value comes below 1.25pu whenODlimit islowerthan
20%. Therefore we have to implement proposed design of
DG-PV with minimum reconfigurations in the network
protection scheme. As long as ODlimit is higher than 25% it
can be deduced that the restriction of ISC < 1.25pu still
permits the achievement of voltage selectivity under a wide
range of PL values.
From above optimization results it is clear that the ideal
characteristics fig5 and fig6 are achieved and after adding a
new LV network to the existingnetwork shortcircuitcurrent
ISC below 1.25pu still allow the system to maintain voltage
selectivity at higher penetration level has shown in fig-7.
4 CONCLUSION
In this paper the optimum design of low voltage
distribution network are connected with DG-PV units are
presented and applied withLVRTCschemeinordertosupply
the generated power from PV to distribution network at
higher penetration level. The simulationresultsexhibitedthe
according affects of the impedance of the DG units on the bus
voltages of the distributed network aftermath of the
happening of the fault. On topof the all aspects theconceptof
voltage selectivity been introduced. In order to control the
cost of the system over sizing of interfaced inverter are
limited. In order to integrate new LV network to the existing
network the system has to maintain reactance XDGat eachDG
unit same as in the existing network is achieved. The
optimum design with the introduction of a new LV network
to the existing networkshown betterperformancecompared
to existed five LV distribution network are represented
throughsimulationresultswhichshowssignificantlyincrease
in short circuit current where ISC < 1.25pu permits the
achievement of voltage selectivity under a wide range of PL
values. Obtained simulation result confirms the effective
development of DG-PV for respective increase in PL values.
REFERENCES
[1] Saeedul Islam, G.M., Al-Durra,A.,Muyeen,S.M.,Tamura,J.:
‘Low voltage ride through capability enhancement of grid
connected large scale photovoltaic system’. IECON 2011 –
37th Annual Conf. of the IEEE Industrial Electronics Society,
Melbourne, VIC, November 2011, pp. 884–889.
[2] Technical Guideline: BDEW (Bundesverband Deutsche
Elektrizitäts- und Wasserwirtschaft) : ‘Generating plants
connected to the medium-voltage network, guideline for
operating plant’s connection to and parallel operation with
the medium-voltage network’, June 2008.
[3] Demoulias, C.: ‘A new simple analytical method for
calculating the optimum inverter size in grid-connected PV
plants’, Electr. Power Syst. Res., 2010, 80, (10), pp. 1197–
1204.
[4] Sheng, W., Liu, K.-y., Cheng, S.: ‘Optimal power flow
algorithm and analysis in distribution system considering
distributed generation’, IET Gener. Trans. Distrib., 2014, 8,
(2), pp. 261–272.
[5] Miret, J., Castilla, M., Camacho, A., Vicuña, L.G.D., Matas,J.:
‘Control scheme for photovoltaic three-phase inverters to
minimize peak currents during unbalanced grid-voltage
sags’, IEEE Trans. Power Electron., 2012, 27, (10), pp. 4262–
4271.
[6] Renders, B., Vandevelde, L., Degroote, L., Stockman, K.,
Bollen, M.H.J.: ‘Distributed generationandthevoltageprofile
on distribution feeders during voltage dips’, Electr. Power
Syst. Res., 2010, 80, (12), pp. 1452–1458.

[7] Grainger, J.J., Stevenson, J.W.: ‘Power system analysis’
(McGraw-Hill, Inc., 1994).
[8] Christodoulou, C., Papanikolaou, N., Gonos, I.: ‘Power
quality assessment for autonomous residential PV systems’.
16th IEEE Int. Conf. on Harmonics and Quality of Power
(ICHQP 2014), Bucharest, Romania, 25–28 May 2014.
[10] EN 50160: ‘Voltage characteristics of electricity
supplied by public distribution system’, 2004.
[11] Ioannis I. Perpinias, Nick P. Papanikolaou, Emmanuel C.
Tatakis: ‘Optimum design of low-voltage distributed
photovoltaic systems oriented to enhanced fault ride
through capability’,2015.

Design of Low-Voltage Distributed Photovoltaic Systems Oriented to Improve Fault Ride through Capability

More Related Content

What's hot (20)

Similar to Design of Low-Voltage Distributed Photovoltaic Systems Oriented to Improve Fault Ride through Capability (20)

More from IRJET Journal (20)

Recently uploaded (20)

Design of Low-Voltage Distributed Photovoltaic Systems Oriented to Improve Fault Ride through Capability