Electric Power Systems Research 196 (2021) 107196

Contents lists available at ScienceDirect

Electric Power Systems Research

journal homepage: www.elsevier.com/locate/epsr

The impact of intermittent renewable energy generators on Lesotho

national electricity grid
Sebota Mokeke , Leboli Z. Thamae *
Energy Research Centre, Department of Physics and Electronics, National University of Lesotho, P.O. Roma 180, Lesotho

A R T I C L E I N F O A B S T R A C T

Keywords: Utility-scale integration of solar photovoltaic (PV) and wind farms has gained momentum as countries pursue
Critical clearing time sustainable power systems. Increased penetration of solar PV and wind alters the operation of power grids as they
Intermittent renewable energy generators have different electrical properties from conventional power plants. The paper assesses the dynamic and steady
Power system stability
state impacts of intermittent (variable) renewable energy generators (IREGs) on the Lesotho power grid.
Solar PV grid integration
Wind grid integration
Maximum allowable penetration of both solar PV and wind generators were determined for Ha-Ramarothole and
Letseng sub-stations, respectively. Dynamic impact studies were performed by applying the short circuit fault at
the bus bar with the least critical clearing time (CCT) and observing voltage, frequency and rotor angle as the
renewable energy capacity was varied. Steady state voltage analysis was performed based on the hourly load of
2018 and generation of both IREGs and Muela Hydropower. Voltage, frequency and rotor angle stability were
evaluated against the Grid Code of Lesotho. The impact studies were done using DigSILENT PowerFactory
software. Maximum allowable penetrations for solar PV was about 19% at Ramarothole substation while for
wind it was found to be 27% at Letseng substation. Simulations revealed that increased penetration of the IREGs
led to grid instability. The solar farm penetration of 36 MW connected at 132 kV resulted in grid instability
mainly from the rotor angle instability while the wind farm penetration of 52 MW connected at 88 kV resulted
from an overvoltage of 1.051 p.u. at the neighboring 33 kV substation of Tlokoeng.

1. Introduction
Advocacy for sustainable economic development coupled with the
need for environmental conservation has resulted in accelerated inte­
gration of renewable energy generators to the power grid. Amongst
renewable energy generators, solar and wind generation technologies
had taken the lead with worldwide capacity of 398 GW and 546 GW,
respectively, as of 2017 [1]. The EU had pledged to integrate renewable
energy penetration of 20% by 2020 and 27% by 2030, while Denmark
promised 50% penetration of renewable energy by 2030 and 100% by
2050 [2–4]. Furthermore, USA has set the renewable energy penetration
of 27% by 2030 while states like New Jersey and California have set
penetration levels of 50% by 2030 [5]. The power generated by inter­
mittent renewables such as wind and solar is variable and
non-deterministic, hence their integration to the grid results in opera­
tional challenges. Successful integration of these intermittent renewable
energy sources must be done such that the power system stability and
reliability are observed.
Increased penetration of intermittent renewable energy generators
(IREGs) affect power grid dynamics differently from conventional power
plants when subjected to disturbances as they exhibit different electrical
characteristics. IREGs (solar and wind) have low short circuit ratio and
contribute minimally to the power system inertia [6]. Power systems
with low short circuit ratio are more susceptible to voltage and fre­
quency instability when subjected to electrical faults. In addition, power
systems with high penetration of IREGs are not flexible as the generation
is not load following [7]. This variability of IREGs may result in
load-generation imbalance rendering the power system unstable. The
load-generation imbalance may be mitigated by fast spinning and
balancing reserves. Another challenge of IREGs is the dispatch strategy
that affect the power market structure. With the increased penetrations
of IREGs, there is need for optimal scheduling of renewable energy
sources and accurate forecasting. Limited reactive power compensation
in IREGs (especially PV) compromises the voltage stability of the power
system [8]. These challenges mandate that grid impact studies be carried
out before integration of IREGs to the utility grid.
Lesotho ranks amongst the top 15 countries with the highest wind
generation potential in Africa [9]. The wind resource assessments done

* Corresponding author.
E-mail address: zl.thamae@nul.ls (L.Z. Thamae).

https://doi.org/10.1016/j.epsr.2021.107196
Received 23 November 2020; Received in revised form 21 January 2021; Accepted 23 March 2021
Available online 1 April 2021
0378-7796/© 2021 Elsevier B.V. All rights reserved.
S. Mokeke and L.Z. Thamae
Fig. 2.1. Detailed 3-phase grid connected PV generator.
at Letseng-la-Terae, Masitise, and Sani suggest the mean wind speeds of
5.97 m/s, 4.93 m/s at 10 m above ground and 5.5 m/s at 9 m above
ground respectively [10, 11]. The wind speeds at Letseng are particu­
larly good for utility-scale wind farms [11]. Furthermore, Taele et al.
suggest that Lesotho has an average global solar radiation of 5.5 – 7.2
kWh/m2 which is good for the solar PV power plants implementation
[12]. Lesotho generation master plan of 2010 suggests that Lesotho has a
wind potential of 758 MW and 361 MW of hydro potential [13]. Though
Lesotho has vast renewable energy sources, electricity access is only
around 40% [14]. The vast majority of Basotho mostly in rural areas lack
electricity access. The peak load for Lesotho has been increasing from
153 MW to 166.91 MW in 2016 to 2018 respectively [15]. The
increasing peak load is met by 74.46 MW (72 MW from Muela and 2.46
MW of small hydropower) generated locally while more than 50% of
electricity is imported from South Africa (Eskom) and Mozambique
(EdM), through a wheeling agreement with Eskom [15, 16]. This acts as
the clear indicator of energy insecurity. In order for Lesotho to reduce its
reliance on imported energy, the Government of Lesotho (GoL) aims to
install some off-grid and grid-connected renewable energy generators
from solar and wind. According to the Lesotho Country Action Agenda of
2017–2030, Lesotho aspires to install 375 MW capacity from the re­
newables with estimated demand of 300 MW by 2030 where excess
energy will be traded in Southern African Power Pool (SAPP) [17, 18].
This capacity will be constituted from the installation of micro-grids in
rural areas and utility scale solar PV and wind farms [19]. In response to
the energy security challenge, the GoL had launched the request for
expressions of interests to investigate the impact of IREGs to the national
grid [20].
The integration of utility-scale intermittent power plants remains a
research area of interest as integration of these generators increase
worldwide. In Canada, Tamimi et al. investigated the impact of solar PV
plants at transmission level and distribution level for Ontario power grid
[21]. From the study, it was concluded that utility-scale solar farms
compromise voltage stability of electrical network than distributed PV
Fig. 2.2. Control levels of 3-phase grid connected PV generator.
2
Electric Power Systems Research 196 (2021) 107196
generators. The research work conducted by Vittal et al. on the Western
US interconnection indicated that utility-scale solar PV farms may either
improve or deteriorate the transient stability of electrical network
depending on penetration level and location in the power grid [22].
Voltage stability studies conducted by Tan and Kirschen suggest that
displacement of the synchronous generators by PV power plants nega­
tively impact voltage stability [23]. The frequency response of South
Australia’s grid revealed that a trip from solar PV farm may result in
frequency oscillations which can cause power system instability [24].
The impact of the wind farms on the Danish power grid were performed
by Liu et al. where it was concluded that increasing penetrations of the
wind farms result in reduced critical clearing time (CCT), compromising
the transient stability of the power grid [25]. Further studies conducted
in China and UK suggest that depending on technology and system
configuration, both wind and solar power can destabilize voltage and
frequency of the power grid [26, 27]. However, Vittal et al. showed that
integration of wind farms improved voltage stability, while Sajadi et al.
found that it reduces line loading, hence improving power system sta­
bility [28, 29].
In Lesotho, the impact of utility-scale intermittent renewable energy
generators has not been conducted. This paper analyses the impact of
both solar PV and wind farms on Lesotho national electricity grid. The
emphasis is on the dynamic voltage, frequency and rotor angle responses
of the power grid after the fault is applied to the bus bar with the lowest
CCT. In addition, the maximum allowable penetration of both solar PV
and wind are determined at locations with good solar and wind re­
sources. To further investigate operational performance of the power
system, the voltage steady state is performed with the actual hourly load
for 2018. The study is performed using the DigSILENT PowerFactory
software. The simulations are evaluated against the Lesotho grid code to
determine the allowable penetration levels. Section 2 of this paper dis­
cusses solar PV model for utility-scale integration, wind turbine model
as it forms integral part in wind farms, fundamentals of power system
stability and underlying equations. Detailed methodology is outlined in
S. Mokeke and L.Z. Thamae
Fig. 2.3. Typical power curve of the wind turbine.
Section 3 while results and discussions are presented in Section 4. Sec­
tion 5 draws conclusions from results.
2. Theory
2.1. Solar photovoltaic power plants
The PV generators are constructed from solar cells which are made
from the P-N junction. The P-N junction produces direct current (DC)
from photovoltaic effect. The solar cells are connected in parallel and in
series to produce solar modules which are connected together to form
PV array. According to Conti and Raiti, output power from the PV array
can be estimated as [30]:
Ppv = ηAc Iβ (1)
where Ac : array surface area [m2]; η: efficiency of the PV system; Iβ :
incident radiation [W/m2] based on the angle of inclination (β).
Fig. 2.1 depicts utility-scale PV generator with its functional blocks.
The output voltage from PV array is fed to DC-DC converter, then to the
inverter and stepped up to grid compatible power. The DC-DC converter
(PV-Side Converter (PVSC)) is implemented by the DC-DC topologies
such as buck, boost and full-scale DC-DC converters. The PVSC main­
tains the input voltage to the inverter as the PV output varies due to
weather conditions. The DC-AC converter (Grid-Side Converter (GSC))
converts the DC power to AC power through the voltage source inverter
(VSI). The GSC uses the sinusoidal pulse width modulation to produce
AC power. The GSC employs additional controls to ensure that the power
and voltage delivered to the grid are maintained at the point of inter­
connect (POI).
According to Molina and Espejo, the control for grid-connected PV
generator consists of three hierarchical controls namely external, middle
and internal controls as depicted in Fig. 2.2 [31]. The active and reactive
power injected to the grid is determined by the external control. The
external control is implemented through the voltage and active power
control loops. The voltage loop consists of voltage-droop which modu­
lates the reactive power of the voltage source inverter in maintaining the
voltage at the POI [32]. The active power control loop matches the
instant power from the PV array to that injected to the grid. This is
achieved through the maximum power point tracking (MPPT) system.
The middle level control dynamically tracks the reference values set by
the external level in generating expected active and reactive power ex­
change between the grid and the VSI [32]. The reference values are
tracked through the use of proportional integral (PI) controllers and
implementation of the current control of the inverter in the
synchronous-rotating dq frame. The switching signals that drive the VSI
are generated by the internal level control using sinusoidal pulse width
modulation (SPWM). The phase locked loop (PLL) synchronizes the
output of the VSI to the utility grid power.
2.2. Wind power plants
The wind power plants are built from the aggregated wind turbines.
3
Electric Power Systems Research 196 (2021) 107196
Fig. 2.4. Block diagram for the wind turbine model.
The wind turbine operates based on two conversion principles. The rotor
blades capture kinetic energy of wind and turn it into rotational me­
chanical energy, then the generator converts rotational mechanical en­
ergy into electrical energy. The mechanical power (Pw) extracted from
blowing wind by the turbine is given by Eq. (2):
1
Pw = Cp (λ, β)ρAVw3 (2)
2
such that ρ represents air density (kg/m3); Vw represents wind speed (m/
s); A represents rotor area (m2) and Cp is power performance coefficient
(unitless) which is a function of tip-speed ratio and blade pitch angle.
Fig. 2.3 depicts power curve showing how the turbine electrical
power varies with wind speed. Below cut-in speed, the turbine does not
generate power. Beyond cut-in speed, the turbine starts to generate up to
rated speed where output power levels and drops to zero after cut-off
speed. Fig. 2.4 shows the standard wind turbine model showing how
the main block diagrams are connected. The wind speed model is
derived from historical measured wind speed. The wind speed model can
be represented by Eq. (3) where Vw (t) is instantaneous wind speed, and
Vwa (t), Vwr (t), Vwg (t),Vwt (t) are average, ramp, gust, and turbulence wind
speeds respectively [33].
Vw (t) = Vwa (t) + Vwr (t) + Vwg (t) + Vwt (t) (3)
The rotor model is represented by rotor blades which captures wind
kinetic energy and converts it into mechanical energy. The pitch angle
controls the output mechanical power of the wind turbine. The drive
drain consists of the rotating masses and the rotating shaft. The drive
train is normally modeled as the two mass system with the turbine and
the generator inertia connected by the shaft [34]. The generator model is
built from the induction generator for Type I – III and the synchronous
generator for the Type IV and the associated power electronics drives.
Furthermore, wind turbine employs the protection scheme based on
measured quantities such as rotor speed, frequency and voltage to
ensure safe operation.
2.3. Power system stability
The power system stability defines the ability of electric power sys­
tem for given operating condition, to gain equilibrium point after being
subjected to physical disturbance with the system variables being
bounded such that the entire system remains intact [35]. The power
system instability results from faults such as the tripping of generators,
transmission line faults and shorts circuit. The power system stability is
classified into rotor angle stability, frequency stability and voltage sta­
bility as depicted in Fig. 2.5.
The rotor angle stability defines the ability of the synchronous
generator within the inter-connected network to remain synchronized
after being subjected to a fault [36]. The rotor angle stability depends on
the balance between electromechanical torque and mechanical torque of
the synchronous generators within the power system [37]. The elec­
tromechanical change of the synchronous generator due to the fault can
be resolved into synchronous torque and damping torque as shown in
Eq. (4):
ΔCe = Cs Δδ + CD Δω (4)
such that Cs Δδ defines the synchronizing torque while CD Δω represents
damping torque where Cs and CD are the synchronizing torque and
S. Mokeke and L.Z. Thamae Electric Power Systems Research 196 (2021) 107196

Fig. 2.5. Classification of power system stability.

V2 = E − jXI (5)

EV2
P2 = sin(φ) (6)
X

V22 EV2
Q2 = + cos(φ) (7)
X X
Eliminating φ from Eqs. (6) and (7) results in Eq. (8) as:
( ) ( )
V24 + 2Q2 X − E2 V22 + X 2 P22 + Q22 = 0 (8)

Fig. 2.6. SMIB equivalent circuit.
The plot of V2 against P2 and Q2 in Fig. 2.7 indicates the static voltage
stability of the power system. The plot gives the maximum permissible
active and reactive power that can be sustained by the electrical network
without losing voltage stability.
Frequency stability defines ability of the power system to maintain
nominal frequency after power system is subjected to disturbance [34].
Frequency stability is caused by sudden increase of the load or generator
loss within power system leading to load generation imbalance. The
frequency response of synchronous generator due to load and generation
imbalance can be characterized by:
2HT dfCOI
= pm − pe = Δp (9)
f0 dt

where pm is total mechanical power; pe is total electrical power; f0 is

Fig. 2.7. P, Q and V plot for SMIB power system.
damping torque coefficient, respectively. Insufficient synchronizing
torque results in aperiodic or non-oscillatory instability whereas lack of
damping torque results in oscillatory instability. In the case of the
asynchronous generators (Type III, and Type IV wind turbines), the rotor
and the blades are decoupled from the power system through power
electronics hence do not contribute to either synchronizing torque or
damping torque after some electrical faults.
Voltage stability defines ability of the power system to maintain bus
voltages within prescribed operational limits after being subjected to a
disturbance and in normal operation [34]. The voltage stability problem
can be characterized by single machine infinite bus (SMIB) system as
depicted in Fig. 2.6 under the assumption that line reactance (X) is much
greater than line resistance (R).
From Fig. 2.6 the voltage, active power and reactive power at the
load terminal are respectively given as:
nominal frequency; fCOI is center of inertia frequency and HT is the total
system inertia.
After the frequency disturbance, the system rate of frequency varies
according to the total system inertia such that for large system inertia,
the frequency rate is small and for small system inertia, there is fast rate
of frequency change.
3. Methodology
The study considered both dynamic and steady state performances of
the Lesotho Electricity Company (LEC) network. The dynamic analysis
was performed by applying 3-phase short circuit faults to suitable bus
bar (most stressed) of the electrical network such that the chosen loca­
tion results in the maximum system instability. The stressed point was
chosen based on critical clearing time (CCT) criterion. The CCT defines
longest time the generator remains synchronized after being disturbed
[34]. After the fault was applied, time-domain responses of voltage,
frequency and rotor angle were evaluated against the grid code of
Lesotho to ensure that grid code is observed before and after the fault
conditions. The steady state performance was evaluated by analyzing

4
S. Mokeke and L.Z. Thamae
Fig. 3.1. Block diagram of the PV generator in DigSILENT.
Fig. 3.2. Block diagram of the DFIG in DigSILENT.
hourly voltage performance against the grid code of Lesotho. The steady
state analysis was based on the hourly load from 2018. The load flow
and stability simulations were performed using DigSILENT software.
DigSILENT is the computer-aided software suitable for analyzing elec­
trical transmission, distribution and industrial networks. In addition,
DigSILENT can model and analyze electrical networks with variable
renewable energy generators.
3.1. System modeling and configuration
The study was carried out based on the computer model of the na­
tional electricity network, which was supplied by the utility (LEC) and
then modeled using the DigSILENT software. LEC electrical network
consists of the Muela hydropower generating unit with capacity of 72
MW, 2 MW of the mini-hydro at Mantsonyane and the external grid from
the neighboring South-African power grid. The electrical transmission
5
Electric Power Systems Research 196 (2021) 107196
network is built from interconnected substations with voltage levels of
132 kV, 88 kV, 66 kV and 33 kV. The distribution network is built with
22 kV and 11 kV voltage levels. The demand voltage levels are consti­
tuted of 380 V and 240 V. Solar PV farm was modeled using DigSILENT
aggregated model with an array capacity of 0.5 MW and the power
factor of 0.95. The 0.5 MW solar PV arrays were connected in parallel to
model the required solar PV farm capacity under consideration. Fig. 3.1
depicts the main block diagram showing how the PV array model is
constructed.
The model consists of four main blocks namely the photovoltaic, DC-
link, controller and the inverter blocks. The inverter is modeled as the
static generator and interfaces with the grid. The PV model represents
the number of PV modules connected in parallel and in series to form the
solar PV array. The DC-link implements the MPPT algorithm. The
controller maintains the control for the PV system ensuring that the
voltage and power at the POI is maintained.
S. Mokeke and L.Z. Thamae Electric Power Systems Research 196 (2021) 107196

3.2. Dynamic stability studies

The stability studies were conducted by performing the load flow

analysis under different penetration levels of the scenarios presented in
Table 1. For each case, the penetration levels were increased until grid
code was violated. The stability studies were performed on the electrical
network with the total load of 185.71 MW. Tables 2 and 3 show the
voltage and frequency limits for Lesotho Grid Code.

3.3. Steady state (quasi-dynamic) stability studies

The steady state was considered so that impact of hourly variations

of the IREGs was considered. To model steady state performance of the
LEC network, the hourly voltage performance of the network was
considered. The steady state analysis was performed according to
Fig. 3.6 based on the yearly load of 2018 and the hydro generation of
2018.
The hourly power from the wind turbine was obtained by assuming
the power curve of the Vestas V80 DFIG wind turbine. The hourly wind
Fig. 3.3. Simplified grid for PV integration.
speeds at Letseng were obtained from the measured wind speed at the
interval of 10 min at hub height of 50 m. The power from the wind
turbine (Pt) was found from Eq. (10) as follows:
⎧ ⎫
⎨ 0 : Vw 〈 Vcutin ; and Vw 〉 Vcutoff ⎬
Pt = Pout : Vcutin < Vw < Vrated (10)
⎩ ⎭
Prated : Vrated < Vw < Vcutoff

where Pout is the power of the turbine between cut-in wind speed (Vcutin)
and the rated wind speed (Vrated) of the wind turbine. Vw and Vcutoff
represent the wind speed at the turbine height (H) and the cutoff wind
speed of the turbine. The total power (Pwindfarm) from the wind farm was
obtained as:
∑N
Pwindfarm = n=1
Pt n (11)

where N represents total number of the wind turbines connected in

parallel.
The hourly output power for hourly PV output was calculated based
on solar radiation downloaded from PVGIS website [38]. The hourly
output power from the solar farm was calculated as:

Fig. 3.4. Simplified grid for wind integration.

ηpv Gt
Ppv = P (12)
ηSTC GSTC rated
The wind farm was modeled with a 2 MW doubly-fed induction
where ηpv and ηSTC are the solar PV array efficiency under operation
generator (DFIG) DigSILENT model with power factor of 0.9. The 2 MW
conditions and standard test conditions (STC) respectively. Incident
DFIG’s were connected in parallel to model the desired wind farm ca­
radiation on the solar array at operating conditions and at standard test
pacity. Fig. 3.2 depicts the wind model with its functional block dia­
conditions is represented as Gt and GSTC respectively. Capacity of the
gram. The wind turbine generator model comprises the wind model,
solar farm is represented as Prated . The array efficiency and the incident
mechanical model (rotor, shaft and gear box), the generator model and
radiation on are calculated based on the model developed by Hove et al.
the controls (PQ control, frequency control, etc.). The rotor blades
[39].
transform wind speed to the mechanical power that drives the shaft
which is connected to the generator via the gear box. The generator is
4. Results and discussions
modeled as a DFIG while the controls ensure that the wind turbine
operates within safety limits.
4.1. Critical clearing time (CCT) results
Figs. 3.3 and 3.4 present the simplified grid for the proposed sites,
where the solar PV farm and the wind farm will be integrated to the grid,
Table 4 shows the how CCT varies as the penetration levels of the
and the neighboring substations while Fig. 3.5 displays simplified LEC
wind and solar generators vary after the fault was applied to the
electrical network. The proposed sites for solar PV farm and wind farm
different buses. The clearing times considered were the least from
are Ramarothole and Letseng respectively. The neighboring substations
different bus categories. The comparative outlook of the results shows
serve as the important points to consider as they are close to the IREGs,
that the rotor angle stability was compromised as the penetration of the
hence affected more by the IREGs integration. The other point that was
solar PV farm generation increases while wind farm does not affect the
considered is the Muela hydropower substation as it is the main syn­
rotor angle stability, as the CCT remains constant as the wind generation
chronous generator for the LEC national electrical network.
increases.
Interestingly, the results indicate that the high voltage bus bars (132
kV) have the least critical clearing times. This suggests that these buses
destabilize the grid most when subjected to the fault. Based on this

6
S. Mokeke and L.Z. Thamae Electric Power Systems Research 196 (2021) 107196

Fig. 3.5. Simplified LEC network.

Table 1
Penetration levels for different scenarios.
Solar PV (MW) Wind (MW) Hybrid (Solar PV (MW) + Wind (MW))
Solar + Wind Solar +Wind

20 30 35 + 30 20 + 50
30 40 35 + 40 30 + 50
36 52 35 + 52 36 + 50

Table 2
Grid code voltage limits.
Normal Conditions Emergency Conditions
Normal Voltage Maximum Minimum Maximum Minimum
(kV) (kV) (kV) (kV) (kV) Fig. 3.6. Flow chart for the steady state analysis.
132 (100%) 138.6 125.4 (95%) 145.2 118.8 (90%)
(105%) (110%)
4.2. Solar PV simulation case
88 (100%) 92.6 (105%) 83.6 (95%) 96.8 (110%) 79.2 (90%)
66 (100%) 69.3 (105%) 62.7 (95%) 72.5 (110%) 59.4 (90%)
33 (100%) 34.7 (105%) 31.4 (95%) 36.3 (110%) 29.7 (90%) Fig. 4.1 shows the voltage response after the 0.1 s 3-phase short
circuit was applied at the bus with the lowest CCT (Mabote) at different
penetration levels. From Fig. 4.1, it is observed that at solar PV pene­
tration levels of 20 MW and 30 MW, the voltages were within the
Table 3 allowed limits and the network was able to recover from the fault as the
Grid code frequency limits.
voltages were damped to the stable limits. Comparing the steady state
Nominal frequency (Hz) 50 voltages before and after the fault, it is seen that the Muela voltage has
Upper limit (Hz) 50.5 reduced to 1.017 p.u. compared to that of 1.03 p.u. before the fault at
Lower limit (Hz) 49.5
penetration level of 20 MW. At penetration level of 30 MW, it is observed
that the voltage has dropped further to 0.98 p.u. This suggests that
observation, the Mabote bus bar was considered the most stressed point increased penetration of solar PV farm results in the worsening voltage
and the 3-phase short circuit was applied at it as the frequency and stability as the voltage at Muela is reduced. The rest of the nearby
voltage responses for other critical points were observed. However, the substation’s oscillations were within the required voltage limits. How­
CCT at Letseng was relatively long compared to other substations. The ever, at the penetration level of 36 MW, the voltage limit was violated,
long CCT for Letseng is caused by the fact that it resides at remote rural as the voltage at the critical buses did not return to the allowable steady
areas characterized by low load capacity. The other points of consider­ state limits. The most affected substation was Muela, which shows the
ation were point of common coupling (PCC), where the wind and solar worst deviation from the nominal voltage limits.
generators are connected to the grid, and the neighboring buses as they Fig. 4.2 indicates how frequency at the faulted bus bar (Mabote) and
are affected more due to their proximity to wind and solar farms. To critical bus bars varies as the penetration of the solar PV integration is
further investigate the impact of increased penetration of IREGs, time- increased. As the solar PV penetration increased, the oscillations for the
based simulations were performed. first swing were almost equal with the magnitude of 51.027 Hz and
49.486 Hz for the penetration of 20 MW, and 51.244 Hz and 49.445 Hz
for the 30 MW penetration at Muela. The other critical points had the

7
S. Mokeke and L.Z. Thamae Electric Power Systems Research 196 (2021) 107196

Table 4
Critical clearing times for solar PV and wind generation penetrations.
Substation No RE Solar generation CCT (sec) Wind generation CCT (sec)
20MW 30MW 36MW 30MW 40MW 50MW

Mabote 132kV 0.18 0.13 0.08 0.02 0.18 0.18 0.18

Muela 132kV 0.19 0.17 0.14 0.04 0.19 0.19 0.19
Mazenod Tx 132kV 0.19 0.14 0.09 0.03 0.19 0.19 0.19
Ramarothole 132kV 0.71 0.4 0.24 0.05 0.71 0.71 0.71
Khukhune 88kV 6.9 2.41 1.27 0.24 6.9 6.9 6.9
Letseng 88kV >20 >20 >20 >20 >20 >20 >20
Katse Intake 66kV 3.82 1.9 1.05 0.2 3.82 3.82 3.82
Matsoku 66kV 4.81 2.56 1.38 0.26 4.81 4.81 4.81
Botshabelo 33kV 0.78 0.44 0.26 0.08 0.78 0.78 0.78
Mazenod Dx 33kV 0.49 0.28 0.16 0.04 0.49 0.49 0.49

Fig. 4.1. Voltage response at the critical bus bars after the fault is applied at Mabote with different penetration levels of solar PV.

Fig. 4.2. Frequency response at critical bus bars after the fault is applied at Mabote with different penetration levels of solar PV.

same response with the highest frequency of 50.749 Hz and 43.837 Hz at
20 MW and 50.416 Hz and 44.586 Hz at 30 MW. The lower frequency
oscillations were observed at the buses which are at close proximity to
the solar plant. The lower frequency swings are attributed to the fact that
the solar plant does not provide the damping torque. In both 20 MW and
30 MW penetrations the oscillations were quickly damped to the
allowable frequency limits indicating that increased penetration of solar
PV is not the main contributor to frequency instability. For both 20 MW
and 30 MW penetrations, frequency stability was observed. At the
penetration level of 36 MW, the frequency stability was violated, as the
frequency was initially almost constant but went out of permissible
limits after about 15 s with Muela exhibiting the highest oscillations.
Fig. 4.3 depicts how the rotor angle and the active power at the
synchronous generator varies with varying penetration levels for the
solar PV integration. At penetration levels of 20 MW and 30 MW, the
rotor angle stability was observed, as the output power was constant
after the fault even though rotor angle was oscillatory. Comparing the
oscillations at 20 MW and 30 MW penetrations, it is observed that the
oscillations were increasing as the solar PV integration was increased
from 20 MW to 30 MW. The oscillations were between − 4.650◦ to

8
S. Mokeke and L.Z. Thamae Electric Power Systems Research 196 (2021) 107196

Fig. 4.3. Rotor angle and real power response of synchronous generator at Muela due to the fault at Mabote with solar PV at different penetration levels.

Fig. 4.4. Voltage response at critical bus bars after the fault is applied at Mabote with different penetration levels of wind power.

fault, the voltage quickly settles to the constant voltage levels without
Table 5
overshooting. In addition, only the nearby substations (Letseng, Tlo­
Voltage levels as penetration increases at Letseng.
koeng and Mahlasela) were affected by the fault while Khukhune was
Penetration (MW) Letseng p.u Tlokoeng p.u. Mahlasela p.u. less affected by the fault. Taking a closer look at Table 5, the voltage
30 1.02 0.99 1 levels of the affected substations keep on increasing as the penetration
40 1.045 1.021 1.02 level of the wind generators increase.
52 1.07 1.051 1.02
At the penetration level of 52 MW, steady state voltage at the Tlo­
koeng substation was greater than 1.05 p.u. which is stipulated as the
74.050◦ at 20 MW penetration and − 19.270◦ to 73.060◦ at 30 MW limit under the grid code of Lesotho. From this, it is concluded that
penetration. At 36 MW, the rotor angle stability was violated and the under the current operating conditions, the 52 MW penetration cannot
active power at Muela hydropower collapsed. be exceeded as it violates the grid code. Hence a slightly lower pene­
From Figs. 4.1 to 4.3, it could be concluded that the solar PV pene­ tration of 50 MW, which is about 27% of the total load, is chosen as the
tration level of 36 MW leads to voltage, frequency and rotor angle maximum allowable limit.
instability, hence a slightly lower penetration of 35 MW acts as the The frequency response due the fault is shown in Fig. 4.5 after the
maximum allowable limit for solar PV at Ramarothole, leading to 0.1 s fault is applied at Mabote. As the penetration levels were increased,
around 19% of the total load of 185.71 MW. Furthermore, the rotor frequency undershoot also increased, showing some small oscillations.
angle was varying most as the penetration levels increased and this The frequency undershoot was 49.644 Hz, 48.498 Hz and 48.358 Hz at
suggests the high penetrations of solar PV affect rotor angle more than the penetration levels of 30 MW, 40 MW and 52 MW, respectively, for
frequency and voltage stability. Letseng and Tlokoeng. The frequency perturbations were severe at the
nearby substations (Letseng, Tlokoeng and Mahlasela) while Khukhune
experienced very small frequency disturbance. This suggests that the
4.3. Wind simulation case increasing penetration of the wind generation does not result in severe
frequency instability as the oscillations were quickly damped to the
Fig. 4.4 depicts the voltage response, after the 0.1 s 3-phase short allowable frequency limits.
circuit is applied to the Mabote substation, under different wind power Fig. 4.6 displays the rotor angle response of the Muela synchronous
generation penetration levels. From Fig. 4.4, it is seen that after the

9
S. Mokeke and L.Z. Thamae Electric Power Systems Research 196 (2021) 107196

Fig. 4.5. Frequency response at critical bus bars after the fault is applied at Mabote with different penetration levels of wind power.

Fig. 4.6. Rotor angle and real power response of synchronous generator at Muela due to the fault at Mabote with wind farm at different penetration levels.

generator before and after the fault was applied at Mabote substation.
From all different penetration’s levels of wind farm, the rotor angle
stability was observed as the output power from the generators
remained constant. The oscillations had the highest overshoot of
84.690◦ and the lowest undershoot of 9.040◦ . Contrary to the solar farm
generation where the increased penetration results in worsened rotor
angle stability, the wind farm generation does not lead to increased rotor
angle instability as the penetrations increase.
4.4. Hybrid simulation case
Hybrid simulation case considered maximum solar penetration of 35
MW and varying the wind generator capacity. It also assumed the
maximum wind capacity of 50 MW as the solar PV generator capacity

Fig. 4.7. Voltage variations at PCC for Letseng and Ramarothole and at Eskom infeed and Muela Hydro Power.

10
S. Mokeke and L.Z. Thamae Electric Power Systems Research 196 (2021) 107196

Fig. 4.8. Voltage variations at the neighboring substations to the solar and wind farms with active power injected from solar and wind farms.

Table 6
Maximum and minimum voltage variations at various substations displays.
Bus No RE Solar case (50 MW) Wind Case (36 MW) Wind 50 MW+ Solar 36 MW
Min (p.u) Max (p.u) Min (p.u) Max (p.u) Min (p.u) Max (p.u) Min (p.u) Max (p.u)

Tlokoeng 0.88 0.95 0.88 0.95 0.88 1.08 0.88 1.08

Khukhune 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
Mahlasela 0.96 0.98 0.96 0.98 0.96 1.04 0.96 1.04
ThabaneMorena 1.00 1.03 1.00 1.03 1.00 1.03 1.00 1.03
Matelile 1.01 1.03 1.01 1.03 1.01 1.03 1.01 1.03
Mafeteng 1.00 1.03 1.00 1.03 1.00 1.03 1.00 1.03
Letseng 0.91 0.96 0.91 0.96 0.91 1.09 0.91 1.09
Ramarothole 1.02 1.03 1.02 1.04 1.02 1.04 1.02 1.04
Muela 1.02 1.03 1.02 1.03 1.02 1.03 1.02 1.03
Mabote 1.03 1.03 1.03 1.03 1.03 1.03 1.03 1.03
Mazenod Tx 1.02 1.03 1.02 1.03 1.02 1.03 1.02 1.03

was varied. For both cases, the voltage, frequency and rotor angle
exhibited the same responses due the fault at Mabote as in solar PV only
and wind only cases. The penetration level of 50 MW was found to be the
maximum penetration as the wind generation was varied and solar PV
kept constant at 35 MW. When the wind capacity was held constant at
50 MW and solar generation varied, it was found that the 36 MW
resulted in system instability hence 35 MW was chosen as the maximum
penetration for the hybrid. It was observed that hybridizing solar PV and
wind at Ramarothole and Letseng does not result in increased penetra­
tion of either solar PV or wind capacity.
voltage variations at various substationsdisplays the minimum and the
maximum voltage limits for the period of 2018. Most of the voltage
limits at the substations were observed except Tlokoeng which experi­
enced the over voltage (1.08 p.u.) and the under voltage (0.88 p.u.) at
times. Also, at Letseng, the under-voltage limits were experienced with
the minimum voltage of 0.91 p.u., but this does not result from wind
penetrations as that voltage level was maintained when there are no
renewables.
5. Conclusions

4.5. Steady state analysis
Fig. 4.7 presents how the voltage at the point of common coupling
(Letseng and Ramarothole) changes as the load and the renewable en­
ergy sources (RES) vary. It also displays how the voltage at Muela and
Eskom infeed vary due to RES. As Eskom acts as the reference machine to
the LEC network, its voltage remains constant and it is not affected by
load and RES fluctuations. On the other hand, Muela voltage varied
within allowable limits as the RES and the load vary. The voltage at the
wind farm varied more while that at the solar farm varied slightly.
Fig. 4.8 displays the voltage at the neighboring substations and the
power output from the IREGs. The voltage variations at the substations
near the solar farm were within limits, and Mafeteng substation showed
the lowest voltage while Matelile showed the highest variations. The
voltage variations near the wind farm varied randomly as load and wind
power varied. Also, the voltage at the Tlokoeng substation sometimes
exceeded the required voltage limit of 1.05 p.u. to 1.08 p.u. during high
wind power generation. This suggests that the mitigation strategies such
as wind power curtailment have to be employed to ensure that the
voltage is within limits at all times. Table 6 – Maximum and minimum
The impact of both solar PV and wind power plants on the national
grid of Lesotho was investigated. The frequency, voltage and rotor angle
responses were observed after the fault was applied at the substation
with the lowest critical clearing time (CCT) under varying penetration
levels of solar PV and wind power generators. Furthermore, the steady
state voltage performance of the network was evaluated. For the solar
PV only scenario, the maximum allowable solar PV capacity was found
to be less than 36 MW, while for the wind only generation scenario, it
was found to be less than 52 MW. The 35 MW and 50 MW penetrations
were found to be the maximum allowable penetrations for solar PV farm
and wind farm, respectively. Under the hybrid case it was found that the
simultaneous integration of wind and solar PV does not result in either
the wind or solar PV penetration increase. From the observation, it is
concluded that 85 MW remains as the maximum allowable penetration
resulting from 35 MW of solar PV at Ramarothole and 50 MW from
Letseng, respectively, leading to about 19% and 27% of the peak load of
185.71 MW.
For all the scenarios considered, the results analysis indicated that
solar PV generation results in increased rotor angle instability relative to
the wind generation option. Furthermore, for both solar PV and wind

11
S. Mokeke and L.Z. Thamae
generation, the frequency stability was not severely affected as the
penetration levels increased. Increased solar PV integration did not
result in major voltage violations as the penetration increases. On the
other hand, increased wind integration resulted in increase of the steady
state voltage for the neighboring substations. At steady state, increasing
wind generation resulted in more voltage fluctuations than the solar PV
generation case. During high wind generation, the voltage at 33 kV
Tlokoeng substation exceeded the prescribed limit of 1.05 p.u. This
suggests that for the wind case there is the need to employ some voltage
control strategies to ensure that the voltage limits are always observed.
The current research did not consider how the penetrations are
affected for future years as load increases due to lack of comprehensive
future network planning expansions. It is highly recommended that
future research on impact of IREGs will consider network expansions
coupled with forecasted load, including the mitigation strategies that
need to be employed to reduce the negative impact of IREGs as their
penetration is increasing.
CRediT authorship contribution statement
Sebota Mokeke: Investigation, Methodology, Software, Validation,
Formal analysis, Writing – original draft. Leboli Z. Thamae: Concep­
tualization, Project administration, Supervision, Writing – review &
editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
References
[1] “RENEWABLES 2018 GLOBAL STATUS REPORT,” Renewable Energy Nolicy
Network Fot the 21st Century, 2018. [Online]. Available: http://www.ren21.net/st
atus-of-renewables/ren21-interactive-map/.
[2] Anonymous, “2020 climate & energy package,” Climate Action - European
Commission, 2016. https://ec.europa.eu/clima/policies/strategies/2020_en
(Accessed Sep. 19, 2019).
[3] Anonymous, “2030 climate & energy framework,” Climate Action - European
Commission, 2016. https://ec.europa.eu/clima/policies/strategies/2030_en
(Accessed Sep. 19, 2019).
[4] “The European Semester, ” European Commission - European Commission. https
://ec.europa.eu/info/business-economy-euro/economic-and-fiscal-policy-coordin
ation/eu-economic-governance-monitoring-prevention-correction/european-semes
ter_en (Accessed Sep. 19, 2019).
[5] “Updated renewable portfolio standards will lead to more renewable electricity
generation - Today in Energy - U.S. Energy Information Administration (EIA).” htt
ps://www.eia.gov/todayinenergy/detail.php?id=38492 (Accessed Sep. 19, 2019).
[6] P. Du, R. Baldick, A. Tuohy, Integration of Large-Scale Renewable Energy Into Bulk
Power Systems, Springer International Publishing, Cham, 2017.
[7] X. Liang, Emerging power quality challenges due to integration of renewable
energy sources, IEEE Trans. Ind. Appl. 53 (2) (2017) 855–866, https://doi.org/
10.1109/TIA.2016.2626253.
[8] R. Shah, N. Mithulananthan, R.C. Bansal, V.K. Ramachandaramurthy, A review of
key power system stability challenges for large-scale PV integration, Renew.
Sustain. Energy Rev. 41 (2015) 1423–1436, https://doi.org/10.1016/j.
rser.2014.09.027.
[9] J.A. Wisse, K. Stigter, Wind engineering in Africa, J. Wind Eng. Ind. Aerodyn. 95
(9–11) (2007) 908–927, https://doi.org/10.1016/j.jweia.2007.01.023.
[10] M. Mpholo, T. Mathaba, M. Letuma, Wind profile assessment at Masitise and Sani
in Lesotho for potential off-grid electricity generation, Energy Convers. Manag. 53
(1) (2012) 118–127, https://doi.org/10.1016/j.enconman.2011.07.015.
[11] T. Mathaba, M. Mpholo, M. Letuma, Velocity and power density analysis of the
wind at Letšeng-la-terae in Lesotho, Renew. Energy 46 (2012) 210–217, https://
doi.org/10.1016/j.renene.2012.04.003.
[12] B.M. Taele, L. Mokhutšoane, H. Narayan, Solar energy resources potential and
sustainable production of biomass energy in Lesotho, Unpublished (2010), https://
doi.org/10.13140/rg.2.1.2962.0882.
12
Electric Power Systems Research 196 (2021) 107196
[13] LEC, Lesotho Generation Master Plan (Pre-feasibility Study), Lesotho Electricity
Company, Maseru, 2010.
[14] M. Mpholo, et al., Rural household electrification in Lesotho, in: M. Mpholo,
D. Steuerwald, T. Kukeera (Eds.), Africa-EU Renewable Energy Research and
Innovation Symposium (RERIS), Springer International Publishing, Cham, 2018,
pp. 97–103.
[15] LEWA, “LEWA Annual Report 2017/18,” Lesotho Electricity and Water Authority,
Maseru, 2017. [Online]. Available: http://www.lewa.org.ls/library/annual_report_
17-18.pdf.
[16] W.J. Klunne, Small hydropower in Southern Africa – an overview of five countries
in the region, J. Energy South. Afr. 24 (3) (2013) 12.
[17] T. Hadjicostas, D. Padayachy, Lesotho Electrification Master Plan Grid
Development Plan, Department of Energy, 2018, 2017/387289/1.
[18] Lesotho Electricity & Water Authority, Electricity Supply Cost of Service Study –
LEWA Lesotho Final Report, LEWA, 2018.
[19] D. Fernandez, Lesotho Electrification Master Plan Off-Grid Master Plan,
Department of Energy, 2017, 2017/387289/1.
[20] LEC, “Lesotho Services for Renewable Energy Grid Integration Study.” 2019,
[Online]. Available: https://www.afdb.org/fileadmin/uploads/afdb/Docume
nts/Procurement/Project-related-Procurement/EOI_-_Lesotho_-_Services_for_Rene
wable_Energy_Grid_Integration_Study.pdf.
[21] B. Tamimi, C. Canizares, K. Bhattacharya, System stability impact of large-scale
and distributed solar photovoltaic generation: the Case of Ontario, Canada, IEEE
Trans. Sustain. Energy 4 (3) (2013) 680–688, https://doi.org/10.1109/
TSTE.2012.2235151.
[22] S. Eftekharnejad, V. Vittal, B.Keel Heydt, J. Loehr, Impact of increased penetration
of photovoltaic generation on power systems, IEEE Trans. Power Syst. 28 (2)
(2013) 893–901, https://doi.org/10.1109/TPWRS.2012.2216294.
[23] Y.T. Tan, D.S. Kirschen, Impact on the power system of a large penetration of
photovoltaic generation, in: Proceedings of the IEEE Power Engineering Society
General Meeting, Tampa, FL, USA, 2007, pp. 1–8, https://doi.org/10.1109/
PES.2007.385563.
[24] R. Yan, T.K. Saha, N. Modi, N.-.A. Masood, M. Mosadeghy, The combined effects of
high penetration of wind and PV on power system frequency response, Appl.
Energy 145 (2015) 320–330, https://doi.org/10.1016/j.apenergy.2015.02.044.
[25] C. Liu, Z. Chen, C.L. Bak, Z. Liu, P. Lund, P. Ronne-Hansen, Transient stability
assessment of power system with large amount of wind power penetration: the
Danish case study, in: Proceedings of the 10th International Power & Energy
Conference (IPEC), Ho Chi Minh City, 2012, pp. 461–467, https://doi.org/
10.1109/ASSCC.2012.6523312.
[26] M. Xu, et al., Influence of distributed generation on distribution system voltage
stability, in: Proceedings of the 5th International Conference on Electric Utility
Deregulation and Restructuring and Power Technologies (DRPT), Changsha, China,
2015, pp. 2352–2355, https://doi.org/10.1109/DRPT.2015.7432631.
[27] L. Holdsworth, Electrical stability of large, offshore wind farms, in: Proceedings of
the Seventh International Conference on AC and DC Transmission, London, UK,
2001, pp. 156–161, https://doi.org/10.1049/cp:20010535, vol. 2001.
[28] E. Vittal, M. O’Malley, A. Keane, A steady-state voltage stability analysis of power
systems with high penetrations of Wind, IEEE Trans. Power Syst. 25 (1) (Feb. 2010)
433–442, https://doi.org/10.1109/TPWRS.2009.2031491.
[29] A. Sajadi, R.M. Kolacinski, K.A. Loparo, Transient voltage stability of offshore wind
farms following faults on the collector system, in: Proceedings of the IEEE Power
and Energy Conference at Illinois (PECI), Urbana, IL, USA, 2016, pp. 1–5, https://
doi.org/10.1109/PECI.2016.7459227.
[30] S. Conti, S. Raiti, Probabilistic load flow using Monte Carlo techniques for
distribution networks with photovoltaic generators, Sol. Energy 81 (12) (2007)
1473–1481, https://doi.org/10.1016/j.solener.2007.02.007.
[31] M.G. Molina, E.J. Espejo, Modeling and simulation of grid-connected photovoltaic
energy conversion systems, Int. J. Hydrogen Energy 39 (16) (2014) 8702–8707,
https://doi.org/10.1016/j.ijhydene.2013.12.048.
[32] W. Cao and Y. Hu, Renewable energy: utilisation and system Integration. 2016.
[33] P.M. Anderson and A. Bose, “Stability simulation of wind turbine systems,” p. 5.
[34] M. Eremia, M. Shahidehpour (Eds.), Handbook of Electrical Power System
Dynamics: Modeling, Stability, and Control, IEEE Press, Piscataway, NJ, 2013.
[35] Definition and classification of power system stability IEEE/CIGRE joint task force
on stability terms and definitions, IEEE Trans. Power Syst. 19 (3) (2004)
1387–1401, https://doi.org/10.1109/TPWRS.2004.825981.
[36] K. Prabha, Power System Stability and Control, McGraw-Hill, New York, 1994.
[37] J. Machowski, J.W. Bialek, J.R. Bumby, Power System Dynamics, Jown Wiley
&Sons. Ltd, 2008.
[38] B. CATTANEO, Photovoltaic Geographical Information System (PVGIS), EU
Science Hub - European Commission, 2018. https://ec.europa.eu/jrc/en/pvgis.
Accessed Mar. 03, 2020.
[39] T. Hove, A method for predicting long-term average performance of photovoltaic
systems, Renew. Energy 21 (2) (2000) 207–229, https://doi.org/10.1016/S0960-
1481(99)00131-7.