Impact of PV and Wind Penetration into a

Distribution Network Using Etap

Rashed R. Waqfi Mutasim Nour
School of Engineering and Physical Science School of Engineering and Physical Science
Heriot-Watt University Dubai Campus Heriot-Watt University Dubai Campus
Dubai, UAE Dubai, UAE
Rw122@hw.ac.uk Mutasim.Nour@hw.ac.uk

Abstract — According to IRENA the number of countries customers, even when unexpected equipment failures or other
adopted a target to ratify the renewable energy on their factors reduce the amount of available electricity” [4].
electricity grid went from 43 countries to 164 over a course of 10 However, high level of DG penetration is expected to give a
years only. These targets which will increase renewable energy negative impact to the grid which has been investigated in
share on the electricity grid, implies the use of Distributed previous research articles, DG can cause a voltage dip,
Generation (DG) like roof-top PV panels or the construction of flickers, and variability in generation, frequency instability
small scale dispersed renewable-based power plant like wind or and raise the short circuit currents [5-8].
PV farms. The aim of this paper is to study the impacts of high
PV and wind penetration into the grid. To address these impacts, Karimi et al. [9] have addressed the issues which might
a comprehensive study performed on a test system provided by arise with high penetration of PV into the distribution
IEEE to model the integration of DG plants into a 33kV network. Voltage variation across distribution feeders was the
distribution network using Etap software. Three levels of first issue to point at, because of the intermittency of PV
integration were modeled to reflect the impact of DG plants power due to irradiance fluctuation caused by clouds,
penetration on gradual basis, 6 scenarios have been assumed and furthermore, it was mentioned that high level of penetration
two types of tests have been carried out to investigate their between (30-50%) can cause harmonic distortion for current
impact on the grid; fault level and frequency transient stability and voltage at distribution feeders. Several mitigation methods
were performed individually at each level of penetration. have been illustrated by authors to minimize the impact of PV
Additionally, in order to examine the grid performance during penetration like installing fixed or switched capacitor bank
extreme weather conditions, several events have been assumed to
and voltage regulator at distribution buses. Kinneth A. et al.
model the intermittent properties of renewable energy
[10] have investigated the negative impact of high penetration
generation. The obtained results showed very good performance
of the grid for penetration less than 30% while higher level of PV
of DG PV using a 16-bus radial system of Cape Town city
and wind penetration requires further attention and modification using Digsilent Powerfactory Software and the study showed
on the grid side. how high DG penetration can cause a reverse power flow
because of high current injection at distribution feeders that
Keywords— distributed generation DG; DFIG; ETAP; 30-Bus increase the loading of the distribution transformers also.
system; transient stability; inertial response; PV penetration; wind
Other researchers have studied the impact of renewable-
energy.
based DG on the transit stability of the grid. Viet N. et al [11]
modeled 2 distribution networks with high level of PV
I. INTRODUCTION penetration using MATLAB, to study the Low Voltage Ride
The term DG has been classified into different schemes Through (LVRT) characteristic on the network stability. The
depending on several features that have been addressed like, presented results showed that up to 25% of penetration the DG
DG capacity, voltage, technology and dispatchability. While improves the grid transient stability, while higher level of DG
numerous researchers have given a specified capacity of DG penetration results in less grid stability during disturbances. A
plants and voltage range, others have linked it to the sudden loss of PV generation during faults does not give
technology or dispatchability of the plants [1-3]. However, synchronous generators enough time to timely supply the
following a comprehensive literature review, DG can be missed generation.
defined as “Any power generation located at distribution In addition, PV and wind as DG technologies lack the
network or customer’s site of the meter, which contributes inertial support which is notable in the conventional
with the centralized power plants to meet the demand synchronous generator, the lack of inertia results in frequency
regardless of the capacity, dispatchability, technology of DG excursion at the time of incidents like generation or load loss
and the voltage where the system is placed”. (supply and demand mismatch). The frequency deviation is
Power quality is a term linked to the grid reliability determined by equation (1) [12]:
according to what North American Electric Reliability
Corporation (NERC) has declared. Thus, the reliability can be d 1 
Pg − Pl =  J system .ω el  (1)
defined as “the ability to meet the electricity needs of end-use dt  2 

978-1-5090-5454-1/17/$31.00 ©2017 IEEE

 Where Pg is the generated power, Pl is the load, ωel is the WT cut-in speed is 3 m/s and the rated wind speed is 9 m/s,
angular frequency and Jsystem is system’s inertia. Hence, when which gives the advantage for power generation most of the
the grid is penetrated with high level of PV and wind the time. In terms of the generator equipped in the wind turbine,
overall system inertia will decrease which will result in higher the selected generator is Doubly Fed Induction Generator
frequency deviation. The current frequency control methods (DFIG) as variable speed control WTG is more preferable to
like inherent inertial support, Automatic Generation Control maximize the power captured from wind energy over the fixed
(AGC) and primary frequency controller will need to be speed WTG. Hence, in this paper the DFIG was selected since
modified soon in order to save system inertia, as the non- it is considered to be a mature technology with lower cost of
synchronous technologies like PV and wind are being manufacturing than other variable speed WTG [17-19].
promoted widely [13].
Following the WT selection and in order to reach to the
The aim of this paper is to investigate the impact of PV intended level of DG penetration, each wind farm is designed
and Wind penetration into the distribution grid. The to have a 12 MW installed capacity. This will be obtained by
penetration level of the PV and Wind generations is studied installing 6 numbers of 2 MW Vestas WT. All WTs are
for different percentages to mimic the different scenarios of connected to 0.69KV bus, which in turn connected to a 33KV
renewable integration to the existing grid. bus by using 15 MVA step up transformer. Fig. 2 shows
typical arrangement of the wind farm.
II. METHODOLOGY
A. System Modeling
The method used to study DG impact on the grid is to
model the electrical grid with different level of DG penetration
that reflects the gradual increase of renewables on the grid.
Fig. 1 shows the generation mix in the model from
conventional synchronous generator and renewable-based DG.
PV and wind penetration started at 15%, then 30% and finally
50%.

Fig. 2. Typical Arrangment of wind farm

Fig. 1. Block diagram of system modeling C. PV Farm Specifications

The IEEE 30-Bus Test System was utilized for system The PV farm designed to be of 5 MW installed capacity,
modeling. It is selected due to the fact that it is mainly a 33kV each farm consists of five PV arrays, while each array consist
system, and according to the American National Standard of 4550 panels. Subsequently, the inverters are connecting the
Institute (ANSI) the distribution voltage level starts from 69 PV arrays to 0.48KV bus which is connected to a 33KV bus
KV up to 480 V [14], which is the last stage before power by using a 6 MVA step-up transformer. Fig. 3 shows block
feeding to the customers. The IEEE 30-Bus network is a diagram for PV farm to grid connection via DC-AC inverter.
portion of American Electric Power System in Midwestern US The selected inverter supports Maximum Power Point Tracker
[15]. The system total connected load is 300MW. The power (MPPT) which allows for maximum solar energy extracting.
generated by synchronous generators will be reduced as DG
plants will compensate this reduction. Thus, the required level
of penetration will be achieved as shown in table I.
TABLE I. POWER GENERATED BY DG PLANTS

Percentage of DG Total Generation in MW

15 % 45 MW

30 % 90 MW
Fig. 3. Block diagram of grid connected PV farm.
50 % 150 MW
D. Scenarios
To model the penetration of the renewable-based DG, a The DG plants were gradually connected to the
number of wind and PV plants were connected gradually into Distribution network to reach to the required level of
the 33KV distribution buses to achieve the required penetration. Five scenarios have been considered which are
percentage of integration. listed in table II. Also the table shows the number of DG
plants at each percentage of penetration and at each scenario.
B. Wind Farm Specifications
The Wind Turbine (WT) used for system modeling is a
2MW rated power by Vestas model: (V110-2.0MW) [16]. The
 B. Transient stability calculation
TABLE II. LIST OF SCENRIOS (1-4) The transient stability analysis allows to thoroughly
Scenario Base 1 2 3 4 5 investigate the behavior of the grid during abrupt events or
faults. This will lead to understand to what extent is the grid
PV % 0 5% 9% 15 % 35 % 15 % rigid that it can withstand the contingencies. To achieve the
Wind % 0 10 % 21 % 35 % 15 % 35 % required analysis, a set of events have been considered which
Total DG % 0 15 % 30 % 50 % 50 % 50 % models incidents might face the grid to examine the frequency
No. of 12MW
0 3 6 9 4 9
excursion. The events for scenarios 1-4 are listed in table IV,
wind farm for scenario 5 the events are shown in table V.
No. of 5MW
0 2 5 10 21 10
PV farm
Actual DG in TABLE IV. SEQUENCE OF EVENTS FOR SCENARIOS (1 – 4)
0 46 97 158 158 158
MW
Time (s) Event
30 3-phase fault occurrence
From the data listed in table II, it can be noticed that for 70 3-phase fault clearance
scenarios 1 to 3, wind energy forms 70% of the total installed
DG capacity and PV forms 30%. While at scenario (4) wind 100 Loss of PV generation
forms 30% of the total installed DG capacity and PV forms 120 Loss of WT generation
70%. The level of penetrations in scenario 5 and 3 are exactly
the same however, the difference in scenarios will be in the TABLE V. SEQUENCE OF EVENTS FOR SCENARIO 5
assumed events in the transient stability test.
Time (s) Event
30 3-phase fault occurrence
III. RESULTS 70 3-phase fault clearance
Two types of tests were performed on the ETAP software; 100 Loss of WT generation
the fault current calculation and frequency transient stability. 120 Loss of PV generation
The significant results are represented in this section.
A. Fault current calculation
The loss of PV generation is based on a consideration that
In this research, 3 buses are selected to be faulted by 3- in some weather condition, a heavy clouds could pass over the
phase short circuit fault individually, in order to examine the city that prevent the sun light from reaching the PV panels to
impact of the DG on fault current and fault level. In all generate power. This formation of clouds can drop the power
electrical networks the fault level is a vital parameter that is generated by PV farms within few seconds to reach 10% of
known and pre-calculated at each point in the grid for the total installed capacity [20], in another word, the energy
protection purposes and sizing of the electrical installations. generated by PV will be assumed to be zero. The loss of WT
Table III illustrates the fault current at base scenario, where generation is based on a consideration that gust of wind is
0% of DG is integrated into the grid and then at scenario (4) established; the speed of wind exceeded the cut-off speed of
where 50% of DG integration. In addition, the percentage of the wind turbines which is 22 m/s. As a result, for protection
increment is calculated.
purposes the brakes of the rotor are applied to force them to
The selection of the 3 buses was based on the location of shut down and halt. Hence, the power generated will
the bus and its function. Bus-24 is a centralized bus and one dramatically decline; this situation will be modeled also as a
PV DG plants is coupled to it. Bus-30 is coming after long sudden loss of WT energy. The obtained results showed that
distribution feeder and is directly connected to load and 2 DG for penetration up to 30% no severe frequency deviation is
plants; wind and PV. Finally, Bus-19 is connecting several observed as shown in Fig. 4 which plots scenario 2. The
distribution buses together and no DG plants are connected to frequency dip is less than 0.5% with immediate recovery.
it, which explains the slight increment in fault current
comparing to the other buses.

TABLE III. FAULT CURRENT CALCULATION RESULTS

Bus ID Fault Current at Fault Current at Augmentation

0% DG (KA) 50% DG (KA) (%)
(Base Scenario ) (Scenario - 4)
Bus-24 9.7 11.7 21 %
Bus-30 3.1 4.7 52 %
Bus 19 8.2 8.8 7%

From table III it is noticeable that the fault current has

been increased in all buses, accordingly the fault level will rise
also. Thus, proactive measures should be taken to eliminate
the effect of such significant rise on the 3-phase fault current Fig. 4. Frequncy response at 30% (9% PV, 21% wind) of DG penetration.
and its impact on the grid apparatus and equipment.
 In scenario 3 as shown in Fig. 5, where the penetration of penetration of DG plants at load side resulted in significant
renewable-based DG is 50%, it can be noticed that no rise in fault level. This parameter under any circumstances
variation in frequency is observed during the fault, as the grid cannot be neglected. Essentially, the fault level is pre-
is getting more solid by the support of 19 DG plants. On the calculated and known for grid engineers in order to select the
other hand, the frequency distortion at second 100 at the time optimum size of the switchgear breakers and to determine the
of PV generation loss is very minor as PV forms only 30% of set point of the protection relays. The calculation has
the DG plants. However, the most significant result is at concluded that the augmentation in fault level for the buses
second 120 when the wind generation is curtailed. As the grid coupled to DG plants is between 20 – 50 % based on the
has lost all generation coming from DG plants which is equal location of the bus and the number of the connected DG
to 158MW, aaccordingly, the system become unstable, the plants. This increment is high enough to cause a malfunction
frequency started to decline sharply and the grid encountered a of the protection relays as well as to damage the switchgear.
power failure. As a result, a reconfiguration of the protection setting and
selection of the switchgear breaker is inevitably required [21].

Fig. 5. Frequncy response at scenario 3 .

In scenario 4, where the share of PV and wind is Fig. 7. Frequncy response at scenario 5.
exchanged, i.e. 70% PV and 30% wind of the total DG. The On the other hand, the transient stability test reveals the
system succeeded to recover after the loss of PV energy rest of the DG impact on the grid. At the initial level of DG
although it forms 70% of the total DG power. However, when penetration (scenario-1 and 2) the impact of the incidents
the system lost the generation coming from wind farms, the listed in table IV on frequency was very little as the power
system again failed to withstand the loss of DG plants coming from the conventional plants is the dominant.
generation and a power failure occurred as shown in Fig. 6. However, from the results it can be concluded that the critical
level of penetration can be determined to be 50% when a
blackout or catastrophic failure in grid developed. In scenario
(3), the loss of the 30% of DG in form of PV energy, which
equals 50MW, would not cause system fatal. However, at the
time grid lost the rest of DG plants power in form of wind, the
power system collapsed as the DG plants forms 50% of the
total generated power. This can be explained that synchronous
generators in the conventional power plant could not cope up
with sudden loss of the DG generation. To investigate why
this failure occurred; the operation of the synchronous
generator has to be understood. Basically, all synchronous
generators on the grid are paralleled together and running at
the same speed, to maintain the frequency at its desired value.
But to maintain the speed constant, the demand should match
the supply instantaneously and any deviation could result in
frequency excursion. When the load varies, the prime mover,
Fig. 6. Frequncy response at scenario 4. which is coupled to the synchronous generator to convert the
On the other hand, the result in scenario 5 is different as mechanical power into electrical power, responses to the load
shown in Fig. 7. The loss of 70% of DG power in form of variation immediately. If the load increased, the mechanical
wind was enough to let the system collapse, unlike the loss of power that drives the prime mover should increase too. In real
70% of DG in form of PV, where the system recovered world this happen by letting the governor of the prime mover
immediately. burns more fuel and produce more power [22]. Governors are
working in a time frame of millisecond and they are faster in
response than the grid operators, however, when the frequency
IV. DISCUSSION drops beyond a pre-specified value that the governor can cope
As the generation is in close proximity to the load side, the up, a complete frequency collapse will induce [23]. In this
fault current calculation has proofed that increasing the paper, the synchronous generators were supplying 150MW
before losing the PV power which is 50MW, afterwards, the REFERENCES
governors managed to compensate the power was being [1] Pepermans, G., Driesen, J., Haeseldonck, D., Belmans, R. &
generated by PV by burning more fuel. However, once the D’haeseleer, W. (2005) 'Distributed generation: definition, benefits and
generation of WT interrupted, the generators were supposed to issues', Energy Policy, 3(6), pp. 787–798.
cover the power was being produced by WT, which is [2] Dondi, P., Bayoumi, D., Haederli, C., Julian, D., and Suter, M. (2002)
approximately 105MW, but the governor failed to recover this 'Network Integration of Distributed Power Generation', Journal of
Power Sources, 106, pp. 1-9.
magnificent amount of power. As a result, the speed of the
generators started to fall down, accordingly the frequency [3] Breker, S., Claudi, A. & Sick, B. (2014) 'Capacity of Low-Voltage Grids
for Distributed Generation: Classification by Means of Stochastic
started to decline. On the other hand, in scenario (4) when the Simulations', IEEE, 30(2), pp. 689-700.
share of PV energy and wind energy has been swapped i.e. [4] Bhardwaj, A., Chanakya, M. & Sharma, R. K. (2013) 'Transmission
70% PV and 30% wind, the response of the grid during the System Reliability Planning', Enhanced Research Publication, 2(1), pp.
incidents has altered. The loss of 70% of DG in form of PV 4.
did not cause a failure in the power system, while at scenario [5] Barker, P. P. & Mello, R. W. (2000) 'Determining the Impact of
(5) when only the wind generation that forms 70% of the total Distributed Generation on Power Systems: 1. Radial Sistribution
DG curtailed, the system collapsed and could not ride out this Systems', IEEE, 3(0), pp. 1645-1650.
harsh condition. One explanation for this behavior is the [6] Dulăua, L. I., Abrudean, M. & Bică, D. (2014) 'Effect of Distributed
Generation on Electric Power Systems', Procedia Technology, 12(2014),
inertial feature of the wind turbines, although the inertia of the pp. 683-685.
WTGs is less than the conventional synchronous generator,
[7] Haque, M.M. and Wolfs, P. (2016) 'A review of high PV penetrations in
they still possess the inertial response as WT stores significant LV distribution networks: Present status, impacts and mitigation
amount of kinetic energy, while PV energy lacks any amount measures', Renewable and Sustainable Energy Reviews, 62(2016), pp.
of inertia as no moving parts are involved in power generation 1195-1208.
[20]. [8] Ipinnimo O., Chowdhury, S., Chowdhurya S.P. & Mitra, J. (2013) 'A
review of voltage dip mitigation techniques with distributed generation
The inertial support is a very vital parameter for the grid, in electricity networks', Electric Power Systems Research, 103(10), pp.
as it strengthens it to resist any deviation in the frequency 28-30.
caused by abrupt load variation or faults. Also, the relation [9] Karimi, M., Mokhlis, H., Naidu, K., Uddin, S. and Bakar, A.H. (2016)
between the overall grid’s inertia and frequency variation has 'Photovoltaic penetration issues and impacts in distribution network – A
review', Renewable and Sustainable Energy Reviews, 53(364-0321), pp.
been clarified in equation (1) in introduction section. Thus, the 594–605.
methodologies used for frequency control have to be
[10] Kenneth, A.P. and Folly, K (2014) 'Voltage Rise Issue with High
developed in order to allow for higher level of wind and PV Penetration of Grid Connected PV', IFAC Proceedings Volumes, 47(3),
integration into the grid [12]. pp. 4959–4966.
Moreover, to ensure the system owns the required inertia, [11] Viet, N. H. and Yokoyama, A (2010) 'Impact of Fault Ride-Through
Characteristics of High-Penetration Photovoltaic Generation on
the Federal Energy Regulatory Commission (FERC) have Transient Stability', International Conference on Power System
released the order 661-A which requests wind farms operator Technology, (11700577), pp. 1-7.
to comply 3 requirements before network connection is [12] Tielens, P and Hertem, D. V. (2012) 'Grid Inertia and Frequency Control
granted. One of these requirements is to guarantee that Fault in Power Systems with High Penetration of Renewables', ELECTA
Ride Through (FRT) test is performed for the farm, which Research Group, pp. 1-5.
impose an unfailing connection of the farm to the grid during [13] Seneviratne, C. and Ozansoy, C. (2016) 'Frequency response due to a
the fault until its clearance. large generator loss with the increasing penetration of wind / PV
generation – A literature review',Renewable and Sustainable Energy
Reviews, 57(1364-0321), pp. 659–668.
V. RECOMMENDATION AND CONCLUSION [14] American National Standard Preferred Voltage Ratings for Electric
Power Systems and Equipment (60 Hz), ANSI C84.1-1989.
This research has concluded that the optimum level of DG
[15] University of Illinois (2016) IEEE 30-Bus System, Available
integration for the today’s grid is 30%, while higher level of at:http://icseg.iti.illinois.edu/ieee-30-bus-system/ (Accessed: 2nd Feb
integration requires a mitigation measure of the DG impact 2016).
illustrated in this paper, as well as, grid modification. The [16] Vestas (2015) 2MW Platform, Denmark: Vestas.
results obtained in this paper are in line with the [17] Kadam, D.P. & Kushare, B.E. (2012) 'Overview of Different Wind
recommendation published by IRENA in 2013 [24] when it Generator System and their Comparison ', IJESAT, 2(4), pp. 1076-1081.
stated that, if the level of DG penetration exceeded 30% the [18] Li, H. & Chen, Z. (2008) 'Overview of different wind generator systems
implementation of the Smart Grid become necessary. and their comparisons', IET Renewable Power Generation, 2(2), pp.
Additionally, the variability and uncertainty of the renewable- 123-138.
based DG can be managed by increasing the geographical area [19] Camm E.H., Behnke M., Bolado O., Bollen M., Bradt M., Brooks C. and
that the electrical network covers. Hence, the impact of the Dilling W. (2009) 'Characteristics of Wind Turbine Generators for Wind
Power Plants', IEEE,10903732(1932-5517), pp. 1-5.
weather condition would be limited to a certain area instead of
the whole grid. Also, implementing various renewable energy [20] Kassakian, J.G.& Schmalensee, R. (2011) The Future of the Electric
Grid, Massachusetts: Massachusetts Institute of Technology
technologies and not to rely on one technology makes the grid
[21] Hien, N. C., Mithulananthan, N. and Bansal, R. C. (2013) 'Location and
more flexible. Moreover, installing advanced weather Sizing of Distributed Generation Units for Loadabilty Enhancement in
forecasting devices at DG plants increases the resolution of Primary Feeder', IEEE,7(4), pp. 797 - 806.
weather predication and give early warning to the grid [22] Boldea, I. (2006) Synchronous Generator , 1st edn., 6000 Broken Sound
operators to keep some spinning reverse units ready to pick up Parkway, NW: CRC Press.
the load incase of contingencies. [23] NERC (2011) Balancing and Frequency Control, Princeton, NJ: NERC.
[24] Kempener R., Komor, P. & Hoke, A. (2013) Smart Grids and
Renewables, Colorado: IRENA.