Effect of Variable PV Power on the Grid Power

Factor under Different Load Conditions

Deema Al-Baik Vinod Khadkikar

Electrical Power Engineering Electrical Power Engineering
Masdar Institute of Science and Technology Masdar Institute of Science and Technology
Abu-Dhabi, United Arab Emirates Abu-Dhabi, United Arab Emirates
[email protected] [email protected]

Abstract—The interconnection of renewable sources with the grid change on a system with high bus voltage is negligible,
carries various issues. At different time of the day, there exist however, the low voltage bus experiences a larger voltage
variations in the output power of the PV. In this paper, the effect change. The possible effect of PV on the power system due to
of different amount of photovoltaic (PV) real power injection load power changes is highlighted in [6]. It is explained that the
under different linear load conditions, such as, balanced RL and load power follows a predictable change in demand and that
unbalanced RL loads, has been studied. As the power generation the voltage at the load should remain within certain limits.
from the PV changes, depending on the loads connected to the There are several ways to regulate the voltage, one of which is
system the power factor of the grid might change from leading to the use of switched capacitors. However, if a PV is installed in
lagging or vice-versa. Multiple simulation cases are carried out
the system, it might interfere with the control scheme being
with different loads to observe the changes in the power factor of
implemented and this may lead to additional problems [6]. In
the grid as the real power from the PV varies. Furthermore, it is
observed that under unbalanced load condition, the balance [7], the effect of PV on a network with multiple loads, such as,
active power injection from PV system may increase the current resistive load, RL load, and a combination of RL and RC load,
unbalance factor on the upstream of the point of common has been discussed. The paper also investigates the effect of
coupling. This paper suggests that a thorough study should be PV penetration on harmonics in low voltage systems under the
conducted while connecting a PV solar plan to a feeder where the different loads.
load could be highly inductive or unbalanced. In this paper, the effect of different power injection from
PV system under RL load condition is addressed. Furthermore,
Keywords- Grid-connected PV system, power factor changes, the influence of unbalanced load on the upstream grid current
unbalanced load. unbalance, which so far has not been studied, is also
thoroughly discussed.
I. INTRODUCTION
Recently the renewable energy sources based power II. SYSTEM OVERVIEW
generation systems are becoming attractive solution to produce In this study, the photovoltaic arrays are modeled by a DC
the electric power. The scarcity of non-renewable fuels and the source. Similar consideration is done in [4,5]. The PV
increasing environmental concerns are instrumental for the generated power is injected to the main grid through an
technologies advancement in the area of renewable energy inverter. The overall system diagram is shown in the Fig. 1. For
sources. One of the widely used renewable energy source is the this system the voltage level is considered as 381V (line to
solar energy. A lot of research has showed interest in line). The transformer turn ratio therefore is 1:1. Tabel-1 gives
photovoltaic technology, and maximum power extraction [1-2]. the system parameter for the study system. The switching
There is extensive research in developing new algorithms to operation of the PV inverter generally produces harmonics in
achieve maximum power point tracking [3]. the output PV currents. An output filter, therefore, should be
However, the grid interconnection of renewable energy used to minimize the current distortion level in the PV output
sources has brought new challenges to the electric power as well the grid current. In this paper, an LC filter is used. The
system. Some of the issues are harmonics injected by inverter- inductor size and cost are greater than the capacitor, and
based system, voltage rise, voltage fluctuations, voltage usually, the inductor is designed to be as small as possible. In
stability and so on. In [5], the effect of different power factor of addition, the big inductor causes larger voltage drop and
the PV output power injection on the distribution network reduces the system dynamics. Its value should be less than 0.1
voltage is discussed. It is discussed that as the power factor of pu as this limits the voltage drop across it [6]. The inductor
the PV goes from negative to positive, in other words, the PV value can be found from the equation given below [7, 9, 11].
changes from absorbing reactive power to delivering the
reactive power, the feeder voltage increases. The paper
suggests a solution for regulating the power factor of the PV by VDC
implementing reactive power compensation devices. L= (1)
8!imax fs
Furthermore, it also shows that the effect of PV power factor
Where, Δimax is the maximum ripple allowed, which can be In this system, the PV supplies only active power to the grid,
so the d-axis current is set to the PV rated current, and the q-
10-25% of rated current [7,10]. VDC is the maximum DC link
voltage and fS is the inverter PWM switching frequency. axis component is set to zero.

III. SIMULATION RESULTS

In this section a detailed MATLAB/Simulink based
simulation study is discussed. The PV power is varied in steps
of 10, from 0 up to its rated capacity to study the effect on the
grid. Fig. 3 shows the effect of changing PV power on the grid
side power factor. In this case a balanced RL load is
considered. As noticed from the Fig. 3, the input grid side
power factor becomes lower (more inductive in this case) as
the active power from the PV increases.

Figure 1: The overall system diagram

TABLE 1: SYSTEM PARAMETERS

System Parameters Values
Nominal (Bus) VL-L, frequency 381 V, 50 Hz
Line parameters/km 0.347 + j1e-6 Ω
Photovoltaic rating 10 kW

Figure 2: The PLL-based control of the PV system

The capacitor is designed based on the maximum reactive

power absorbed by the filter at rated power. This constraint
achieves a limit on the drop of real power at the PV rating. The
reactive power should remain less than 5% of the rated power
[6, 8, 9]. The capacitance is calculated from the equation Figure 3: (a) Grid voltage, (b) load current, (c) PV and grid currents for PPV =
below. 0.3pu, (d) PV and grid currents for PPV = 0.6pu, (e) PV and grid currents for
PPV = 0.9pu.
! Pr
C= (2)
3x2" f1VLN
2
A. Balanced RL load with changing power factor, S = 9kVA
Where, ! is percentage of reactive power from rated power, In this case, the load power factor is changed from 90˚ to
Pr is rated inverter real power capacity, f1 is fundamental unity power factor in the step of 10˚. The active power injected
frequency and VLN is the line to neutral voltage. In order to by the PV system during this condition is kept constant at PPV=
connect the PV to the grid, it is required to synchronize the 0.5 pu. The effect of different load power factor on the grid
output with the grid voltage. Fig. 2 depicts the overall side input power factor (pfgrid) is reported in table-2. Moreover,
controller for the PV system inverter. A phase-locked loop the total load apparent power (S) is kept constant, that is, the
(PLL) is used for grid synchronization. The control is based on load active and reactive powers are changed to achieve the
desired load phase angle (θL).
the dq transformation of the output PV phase currents and
synchronizing the d and q components with the grid voltage.
 As noticed from the table-2, when there is pure inductive
load on the system, the power factor seen from the grid appears
as leading. This is mainly due to the fact that in this case the
PV injected active power (as load active power demand is zero)
flows towards main grid. The combination of PV active power
and the load reactive power causes the grid power factor to
change, in this case, leading power factor. As the load power
factor is changed from pure inductive (90o) towards the unity
power factor, in step of 10o, the grid side power factor slowly
starts to change from leading to lagging and finally becomes
unity for pure resistive load.

TABLE 2: BALANCED LOAD WITH PPV =0.5 PU AND CONSTANT S=9 KVA
Figure 4: The grid power factor vs. load power factor for balanced RL loads
θL (deg) Pfload P (kW) Q (kVAR) pfgrid with changing power factors

90° (lag) 0 0 9.000 0.44 lead

Fig. 4 shows the relationship between load power factor
80° (lag) 0.17 lag 1.563 8.863 0.31 lead and input side grid power factor for different cases, namely,
70° (lag) 0.34 lag 3.078 8.457 0.17 lead
when PPV=0.5pu with load S=9kVA, PPV=1.0pu with load
60° (lag) 0.50 lag 4.500 7.794 0.02 lead
S=9kVA, PPV=0.5pu with load S=6kVA and PPV=1.0pu with
50° (lag) 0.64 lag 5.785 6.894 0.15 lag
load S=6kVA.
40° (lag) 0.77 lag 6.890 5.785 0.35 lag
30° (lag) 0.87 lag 7.794 4.500 0.55 lag
20° (lag) 0.94 lag 8.457 3.080 0.76 lag
10° (lag) 0.98 lag 8.863 1.563 0.93 lag B. Unbalanced RL load, S = 5kVA
0° 1 9.000 0 1 In this case study, the load on the system is considered
linear unbalanced. The presence of unbalance in currents is
measured as current unbalance factor (CUF) and is given as:
Table-3 gives the similar results when the active power
injected from the PV system is changed to 1 pu. Since in this
case the injected active power from PV system (PPV=10 kVA)
[
max I a − I avg , I b − I avg , I c − I avg ] x100% (3)
is always more than the considered load apparent power (S=9 I avg
kVA), the input power factor is noticed to be always leading The excessive current unbalance may lead to unbalance in
except for the pure resistive load. The grid power factor is the grid voltage. The effect of different PV active power
leading since the load requires small P when it is inductive, and injection in the presence of unbalanced load with load current
the power from the PV flows to the grid. As the load requires CUF=33% is given in Table-4. The unbalanced load current
more real power, power flowing to the grid decreases slightly, profile and the grid current profiles under different PV output
however, Q is also decreasing so the power factor changes power levels are shown in the Fig. 5.
slowly. When θL= 20°, Q decreases more than P, and the power
factor starts increasing because the change in Q is big. When
the load becomes purely resistive (Q=0), the grid power factor
TABLE 4: EFFECT OF PPV ON THE GRID SIDE CURRENT UNBALANCE
becomes unity. [CONSTANT LOAD OF 5 KVA]
PPV (pu) %CUFig PFA PFB PFC
TABLE 3: BALANCED LOAD WITH PPV =1.0 PU AND CONSTANT S=9 KVA
0 33 0.55 lag 0.90 lag 0.80 lag
0.1 34 0.35 lag 0.86 lag 0.70 lag
! L (deg) Pfload P (kW) Q (kVAR) pfgrid 0.2 34.5 0.10 lag 0.80 lag 0.40 lag
90° 0 0 9.000 0.70 lead 0.3 35 0.20 lead 0.70 lag 0
80° 0.17 lag 1.563 8.863 0.67 lead 0.4 28 0.40 lead 0.55 lag 0.40 lead
70° 0.34 lag 3.08 8.457 0.60 lead 0.5 32 0.57 lead 0.30 lag 0.65 lead
60° 0.50 lag 4.500 7.794 0.55 lead 0.6 34 0.70 lead 0 0.80 lead
50° 0.64 lag 5.785 6.890 0.50 lead 0.7 33 0.77 lead 0.33 lead 0.88 lead
40° 0.77 lag 6.894 5.785 0.45 lead 0.8 33 0.82 lead 0.55 lead 0.90 lead
30° 0.87 lag 7.794 4.500 0.43 lead 0.9 31.5 0.86 lead 0.70 lead 0.94 lead
20° 0.94 lag 8.457 3.078 0.45 lead 1 29 0.90 lead 0.80 lead 0.95 lead
10° 0.98 lag 8.863 1.563 0.60 lead
0° 1 9.000 0 1
 to the grid. Between 0.6 to 0.9 pu, the average current is
increasing, however, the CUF is still increasing. This is due to
the difference between phase-C current and the average current
that exceeds the rate of change of average current. At 0.9 pu,
the difference becomes small compared to the rate of increase
in average current, and thus CUF decreases. Fig. 6 below
depicts the results of this case, showing power factor of the
grid, and the CUF vs. the PV power.

C. Unbalanced RL load, S = 15kVA

In this case the unbalanced load total apparent power is
increased to 15 kVA. The load current has CUF of 18%. The
effect of different PV active power injection is given in Table-
5. In this case, the load power demand is always greater than
the PV injected active power. For this reason, based on the
individual per phase reactive power demands, there is gradual
increase in the grid side current CUF.

TABLE 5: EFFECT OF PPV ON THE GRID SIDE CURRENT UNBALANCE

[CONSTANT LOAD OF 15 KVA]
PPV (pu) %CUFig PFA PFB PFC
0 18 0.9 lag 0.85 lag 0.94 lag
Figure 5: (a) The grid voltage, (b) three-phase load currents, (c) three-phase 0.1 20 0.88 lag 0.82 lag 0.93 lag
grid currents at PPV = 0.3pu, (d) three-phase grid currents at PPV = 0.6pu, (e) 0.2 22 0.86 lag 0.80 lag 0.92 lag
three-phase grid currents at PPV = 0.9pu. 0.3 23 0.83 lag 0.76 lag 0.90 lag
0.4 25 0.80 lag 0.73 lag 0.90 lag
0.5 27 0.75 lag 0.69 lag 0.88 lag
It can be noticed from Table-4 and Fig. 5 that as power
0.6 28.5 0.68 lag 0.63 lag 0.86 lag
injected by PV increases the grid side current CUF also
0.7 30 0.60 lag 0.57 lag 0.84 lag
changes. In this case the total load power S=5 kVA (less than
0.8 31 0.47 lag 0.50 lag 0.81 lag
rated PV rating). There is slight change in the grid side current
0.9 31.5 0.32 lag 0.40 lag 0.77 lag
CUF. This is mainly due to reverse power flow the PV towards
1 30.5 0.15 lag 0.30 lag 0.73 lag
the main grid. The grid side current power factors, for all three
phases, are also given in the Table-4. From PPV = 0 to 0.2 pu,
the CUF increases quickly. This is due to the decrease in the
grid currents, as the PV supplies more power. At almost 0.25 IV. CONCLUSION
pu, the power factor of phase-A becomes leading, and some
In this paper, several cases were studied to understand the
current starts flowing to the grid. Therefore, the average current
impact of changing PV power on the grid side power factor. It
of phase-A starts increasing, however, IB and IC are decreasing
more than the increase in IA, so the average current decreases is concluded that the grid power factor changes as the PV
and thus the CUF increases. At 0.6 pu, phase-C power factor active power injection level changes, depending on the loads
becomes leading, implying the reverse direction of power flow connected to the system. With balanced RL load of variable
power factor present in the system, the power factor changes
were recorded and were significant in most of the cases.
Furthermore, when the load is unbalanced, as the power from
the PV varies, the grid experiences changes in the power
factor as well as current unbalance. The simulation results
prove the need of performing such studies for systems with
high PV penetration. Further investigation should be
conducted for the large-scale renewable energy based system
integration to the electric network having balanced and
unbalanced PQ loads.

Figure 6: The grid PF and CUF vs. PV power for unbalanced RL load (5kVA)
 REFERENCES [8] Erika Twining; Holmes, D.G.; , "Grid current regulation of a three-phase
voltage source inverter with an LCL input filter," IEEE Transactions on
Power Electron., vol.18, no.3, pp. 888- 895, May 2003.
[1] Agorreta, J.L.; Borrega, M.; López, J.; Marroyo, L.; , "Modeling and [9] Liserre, M.; Blaabjerg, F.; Hansen, S.; , "Design and control of an LCL-
Control of N-Paralleled Grid-Connected Inverters With LCL Filter filter-based three-phase active rectifier," IEEE Transactions on Ind.
Coupled Due to Grid Impedance in PV Plants," IEEE Transactions on Applic., vol.41, no.5, pp. 1281- 1291, Sept.-Oct. 2005.
Power Electron., vol.26, no.3, pp.770-785, March 2011.
[10] Wang, T.C.Y.; Zhihong Ye; Gautam Sinha; Xiaoming Yuan; , "Output
[2] Barbosa, P.G.; Braga, H.A.C.; Rodrigues, Md.C.B.; Teixeira, E.C.; , filter design for a grid-interconnected three-phase inverter," in Proc.
"Boost current multilevel inverter and its application on single-phase IEEE Power Electron. Spec. Conf. (PESC), 15-19 June 2003, pp. 779-
grid-connected photovoltaic systems," IEEE Transactions on Power 784.
Electron., vol.21, no.4, pp. 1116- 1124, July 2006.
[11] Yongqiang Lang; Dianguo Xu; Hadianamrei, S.R.; Hongfei Ma; , "A
[3] Kadri, R.; Gaubert, J.-P.; Champenois, G.; , "An Improved Maximum Novel Design Method of LCL Type Utility Interface for Three-Phase
Power Point Tracking for Photovoltaic Grid-Connected Inverter Based Voltage Source Rectifier," in Proc. IEEE Power Electron. Spec. Conf.,
on Voltage-Oriented Control," IEEE Transactions on Ind. Electron., 16 June 2005, pp.313-317.
vol.58, no.1, pp.66-75, Jan. 2011.
[12] Min-Young Park; Min-Hun Chi; Jong-Hyoung Park; Heung-Geun Kim;
[4] Castilla, M.; Miret, J.; Sosa, J.L.; Matas, J.; de Vicuña, L.G.; , "Grid- Tae-Won Chun; Em-Cheol Nho; , "LCL-filter design for grid-connected
Fault Control Scheme for Three-Phase Photovoltaic Inverters With PCS using total harmonic distortion and ripple attenuation factor," in
Adjustable Power Quality Characteristics," IEEE Transactions on Power Proc. IEEE Power Electron. Conf. (IPEC), 21-24 June 2010, pp.1688-
Electron., vol.25, no.12, pp.2930-2940, Dec. 2010. 1694.
[5] Xiaowei Wang; Jun Gao; Wenping Hu; Zhengang shi; Baofeng Tang; , [13] Xing Wei; Lan Xiao; Zhilei Yao; Chunying Gong; , "Design of LCL
"Research of effect on distribution network with penetration of filter for wind power inverter," in Proc. IEEE World Non-Grid-
photovoltaic system," in Proc. IEEE Universities Power Engineering Connected Wind Power and Energy Conf. (WNWEC), 5-7 Nov. 2010,
Conference (UPEC), 31 2010-Sept. 3 2010, pp.1-4. pp.1-6.
[6] Walling, R.A.; Saint, R.; Dugan, R.C.; Burke, J.; Kojovic, L.A.; , [14] Wang, T.C.Y.; Zhihong Ye; Gautam Sinha; Xiaoming Yuan; , "Output
"Summary of Distributed Resources Impact on Power Delivery filter design for a grid-interconnected three-phase inverter," in Proc.
Systems," IEEE Transactions on Power Del., vol.23, no.3, pp.1636- IEEE Power Electron. Spec. Conf. (PESC), 15-19 June 2003, pp. 779-
1644, July 2008. 784.
[7] Benhabib, M.C.; Myrzik, J.M.A.; Duarte, J.L.; , "Harmonic effects [15] Ahmed, K.H.; Finney, S.J.; Williams, B.W.; , "Passive Filter Design for
caused by large scale PV installations in LV network," in Proc. IEEE Three-Phase Inverter Interfacing in Distributed Generation," in Proc.
Electrical Power Quality and Utilisation (EPQU), 9-11 Oct. 2007, pp.1- IEEE Compatibility in Power Electron. (CPE), May 29-June 1 2007,
6. pp.1-9.