Download IMPLEMENTING FREQUENCY REGULATION CAPABILITY IN SOLAR PHOTOVOLTAIC POWER PLANTS

Proceedings of the 4th Annual ISC Research Symposium
ISCRS 2010
April 21, 2010, Rolla, Missouri
IMPLEMENTING FREQUENCY REGULATION CAPABILITY IN SOLAR
PHOTOVOLTAIC POWER PLANTS
Venkata Ajay Kumar Pappu
[email protected]
Badrul H. Chowdhury
[email protected]
Jonathan W. Kimball
[email protected]
Electrical & Computer Engineering
Missouri S & T
ABSTRACT
Photovoltaic power plants pose some challenges when
integrated with the power grid. The PV plants always focus on
extracting the maximum power from the arrays. This makes the
PV system unavailable for helping in regulating the grid
frequency as compared to conventional synchronous
generators. A new technique for tracking a pseudo-maximum
power point for creating power reserve is presented. This will
help provide fast acting power to respond to rapid load changes
in a manner analogous to the inertia of a rotating machine. The
system is implemented in a two stage power conversion model
and has been developed in Simulink/Matlab. A GE-PV 200
solar array has been used for this purpose. The new technique
is discussed in detail and simulation results are provided.
1. INTRODUCTION
A microgrid is typically built around a low-voltage (LV)
distribution systems with distributed energy resources (DERs)
such as micro-turbines, fuel-cells, photovoltaic (PV) arrays, etc.
[1-3]. Photovoltaic systems form an important part of a
renewable energy generation portfolio since they are pollution
free when operating, are modular in nature which makes
construction easier, and they have relatively long life. In a
typical grid-tied operating mode, PV systems are made to
operate at the maximum power point by actively tracking the
array voltage and current at all times. In an energy-limited
system, such as a microgrid, this mode of operation is not
helpful as the PV system is generally not able to follow the load
demand or participate in frequency regulation. In addition, in
the special case when the microgrid is comprised of many other
DERs that are renewable in nature, load following will become
a critical function needed to maintain frequency and voltage of
the microgrid [4]. An example of this type of microgrid is a
forward operating military base. Since system frequency and
voltage are generally decoupled, mainly deviating from their
nominal values because of imbalances in the active power and
reactive power respectively, the design of the control and
optimization methodology becomes quite complex and
challenging.
A two-stage power conversion consisting of a dc-dc boost
converter and a single phase inverter with feedback control is
needed for the implementation of frequency regulation control
[5]. The dc-dc converter is driven by a modified maximum
power point tracker for the PV array. The power reserve
calculation is built into the MPPT algorithm.
The purpose of an MPPT algorithm is to extract the
maximum available power from the array, but in this paper, an
alternate approach of tracking a pseudo-MPP has been
implemented. This will give control over the amount of active
power that can be injected into the grid. Although any type of
PV cell will work, for the purpose at hand, a particular model of
the polycrystalline solar cell is adopted. The grid connection of
the solar modules has to guarantee maintenance of stability
when the microgrid operates in an islanded condition. The
reserve power from the array can be used to ride through the
intermittency of the resource, and help in system frequency
regulation.
For the analysis of the system the model has been simulated
using MatLab/Simulink version R2008b and PLECS version
3.0.2. The complete model is described in the following
sections followed by test results.
2. DESCRIPTION OF THE MODEL
2.1. Overview
The system being modeled is shown in Fig 1. It consists of
a two stage power conversion topology with the raw power
generated from the solar array being tracked for maximum
output using two different algorithms integrated into the MPP
controller. The Online search algorithm/Perturb and Observe
for the true MPP and the modified fractional Voc method for the
pseudo MPP. This controls the duty cycle of the boost converter
to maintain a constant voltage at the dc bus link.
Fig. 1. Topology of the two stage power conversion
1
For the above system a boost converter is used for stepping
up the PV array voltage to integrate it with a single phase
inverter and the grid. The inverter is controlled by a feedback
loop which injects the power into the grid at unity power factor.
The two stage conversion topology provides effective control
over the power transfer stages and provides flexibility to
implement storage devices at different stages.
2.2. Photovoltaic Array
The solar array selected for simulation of the module for
experimental purposes is the GE-PV 200 W. This delivers a
peak power of 200W at the MPP. The general PV cell circuit is
shown in Fig. 2.
Rs
I ph
+
I pv
D
Ro
Rp
V pv
-
Fig. 2. PV cell circuit topology
The governing equation [6] for the current from the array is
Fig. 3. On-Line search MPP algorithm
(1)
Ipv and Vpv correspond to the current and voltage of the PV
array respectively. The photo current Iph is dependent on the
insolation and temperature of the solar array and the model
simulated is a temperature dependent model of the panel.
In Fig. 3, the error or the slope of the P-V curve of the
array is checked at every operating point and tracks the optimal
value for power and voltage.
2.3. MPPT Algorithm
The aim of the system design is to generate a reserve
power in a PV plant. To achieve this, a combination of two
different MPPT algorithms has been used to make the
controller track a false power point from the array. The online
search algorithm tracks the true MPP while the modified
fractional Voc method tracks the false MPP.
Fig. 3 shows the on-line search algorithm which checks for
the zero slope of the P-V curve and tracks the maximum power
from the array [7]. Fig. 4 is the modified fractional Voc
algorithm in which the tracking voltage is set at a fraction of
the open circuit voltage. The normal voltage at MPP is between
0.71 and 0.78 [8] of the open circuit voltage for the array. To
get a reserve power, the fraction is modified to a value between
0.8 and 0.9 which will enable the controller to maintain the
operating point at a lower power value and get a reserve power.
The PV plant controller will change the algorithms based
on the frequency input or load demand from the grid. In Fig. 4,
The term (K*Voc ) is the control variable.
Fig. 4. Modified fractional open circuit voltage algorithm
2
Fig. 5 shows how the two algorithms are implemented at
two different operating points to achieve control over the active
power from the PV plant. P1-P2 provides the required reserve
power similar to inertia of conventional generators for the
system.
Fig. 7. dc-ac power conversion system
A transformer is used to step up the voltage to the single
phase grid voltage level at 170 V peak. A filter on the primary
side of the transformer reduces the harmonics on the output
side. The dc bus voltage reference error is fed through a PI
block and fed to the controller which will maintain the dc link
capacitor voltage to stay constant at the required value.
Fig. 5. Tracking of the new MPP algorithm for two different
operating points on the P-V curve.
The simulink block which controls the change of the two
different algorithms is shown in Fig. 6. It samples the values of
voltage and current to calculate the change in the value of
reference voltage.
3. SYSTEM SIMULATION AND RESULTS
The system is studied under various scenarios as explained
below.
3.1. PV Array
For the purpose at hand, a GE-PV 200W solar panel is
selected. Its characteristics and maximum power at different
insolation levels is shown in Table.1.
Table 1. Normal power output from the PV array
INSOLATION
(W/m2)
1000
900
800
600
500
Fig. 6. MPPT controller block for both the algorithms with
embedded Matlab code.
2.4. Two Stage Converter System
A boost converter and a single phase inverter provide
effective control over the transfer of power from the PV array
to the grid. The dc-dc boost converter has the MPPT which
controls the duty ratio to maintain a constant voltage of 48V at
the dc-ac link.
The inverter has a closed loop control which takes in the
grid voltage, current and the dc bus reference voltage [9] as
inputs and controls the gate signals to transfer the power at
unity power factor. The design of this system in PLECS is
shown in Fig. 7.
MAXIMUM
POWER (W)
200
180
160
120
100
Ipv (A)
Vpv (V)
7.6
6.6
6.15
4.55
3.8
26.4
26.3
26
26.3
26
The output shown in Table 1 is under normal operation
conditions of the PV array at a temperature of 25 C. But when
operated under reserve power, the output is brought down and
helps in the participation of large scale PV system in frequency
regulation as explained in Section 3.2.
3.2. Case 1 - Reserve Power
The system is checked for the following specifications:
Insolation:
500 W/m2
Panel rating
200W
Temperature
25o C
Power generation
100 W
Value of K in fractional Voc is 0.9 with Voc = 32.9V
Grid voltage
170 V peak
The following graphs depict how the system values change
with change in the operation mode of the PV plant. For the
particular case, initially the system is acting with reserve power.
Then through an external signal, when it detects a drop in
3
frequency, the power generation is increased to its true
maximum value. The voltage and current behavior of the array
are shown in Fig. 8.
Fig. 10. Boost converter output voltage
The boost converter output voltage was observed to have a
ripple of 5V. This does not have a significant effect on the grid
voltage at which the current is being injected.
Figs. 11 and 12 show the two different values of current
that are being given to the grid side under reserve power and
full power mode respectively.
Fig. 8. Solar panel Voltage and current
In Fig. 8, the voltage which is initially at 0.9*Voc at a
value of 29.6V changes to a value of 26V and the current also
increases from 2.5A to 3.8A when it gets an external signal that
there is a drop in system frequency.
Fig. 9 shows the change in power from 75W to 100W that
is being generated by the solar panel.
Fig. 11. Current injected into the grid under reserve power
mode
As shown in Fig. 11, the current has a peak value of 0.88 A
peak which is giving a power of 74.8 W at 170 V peak. The
difference in the power generated and given is due to the loss in
the resistor used in the filter on the primary side of the
transformer.
Fig 9. PV array output power for 500W/m2
With a change in the grid frequency, the power generation
from the solar panel is increased or decreased according to
whether the frequency has dropped or risen respectively. This is
necessary from all types of generators in the power grid to help
in maintaining near constant frequency [10].
A closed loop bus voltage control is used for maintaining a
constant voltage at the boost converter output terminals as
shown in Fig. 10.
Fig. 12. Current injected into the grid under high power mode
In Fig. 12, the current has a value of 1.17 A at 170V peak
that gives a power of 99.45 W. Again, the slight difference in
power is due to the resistance in the filter.
4
3.3 Case 2 – Output under uniformly varying
insolation
In this case the system is checked for varying input
insolation to the array. The results, shown in Fig. 13, suggest
that it follows the change in insolation with time.
power. The current injected at the grid end also varies and is
shown in Fig. 17.
Fig. 13. Output power under varying insolation for reserve
power mode with a peak at 160W for 1000 W/m2 insolation.
Fig. 16. System response under rapidly varyin insolation
for true power mode.
In Fig 14 the power curve is for high power mode for the
same variation of insolation.
Fig. 17. Power injected into the grid for the rapidly varying
insolation under true power mode.
Fig. 14. Output power for high power mode with peak at 200W
for an insolation of 1000 W/m2
3.4 Case 3 – Rapidly varying insolation
The system is checked if it is able to track the power under
rapidly varying insolation. The results are shown in Fig 15.
Table 2 gives the details of reserve power from the PV
plant for various insolation values when the system is tracking
0.9* Voc voltage in the reserve power mode. As can be observed
there is a significant amount of difference in power for the
panel. If this is applied to large scale PV systems, it will help in
a better frequency control participation.
Table 2. Reserve power values for various insolation levels.
INSOLATION
(W/m2)
Fig. 15. System response under rapidly varying insolation for
reserve power mode.
Fig. 16 depicts that, for the same rapid variation in
insolation values, the system will track the true maximum
1000
900
800
600
500
TRUE MAX
POWER
(W)
200
180
160
120
100
FALSE
MAX
POWER
160
140
120
90
75
RESERVE
POWER
(W)
40
40
40
30
25
4. CONCLUSIONS
The PV technology, although promising in many aspects,
suffers from several drawbacks such as relatively higher cost,
low efficiencies, inherent output intermittency and daily
5
availability that could be lower than 50%. Yet, it is one of the
more mature, easily available renewable energy technologies
that is currently being used for utility-scale generation. The
frequency regulation capability for the PV plant has been
proposed for possible use in load following schemes,
particularly in microgrid application. This method allows for
effective control of solar power for various purposes without
compromising the efficiency of conversion.
The system has been tested under constant and varying
insolation cases. The frequency regulation capability has been
simulated by using the control over the MPPT algorithm based
on the external frequency signal of the grid which conveys the
need for active power.
5. ACKNOWLEDGMENTS
[9]
[10]
[11]
[12]
The authors wish to acknowledge the Intelligent Systems
Center at Missouri S & T for the research support which made
it possible to carry out the project.
The authors gratefully acknowledge the help of Luke D.
Watson of the ECE department for his design of the control
block for the inverter.
[13]
6. REFERENCES
[14]
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
K. D. Brabandere, K. Vanthournout, D. J. Deconinck, G.
R. Belmans, 2007, “Control of Microgrids,” IEEE Power
Engineering Society General Meeting, pp. 1 – 7.
F. Katiraei, R. Iravani, N. Hatziargyriou, A. Dimeas,
May/June 2008, "Microgrid management: control and
operation aspects of Microgrids," IEEE Power and Energy
Magazine, vol. 6, pp. 54-65.
B. Kroposki, R. Lasseter, T. Ise, S. Morozumi, S.
Papathanassiou, May/June 2008, N.Hatziargyriou,
"Making microgrids work," IEEE Power and Energy
Magazine, vol. 6, pp 41-53.
N. Jayawarna, X. Wu, Y. Zhang, N. Jenkins, M. Barnes,
2006, “Stability of a MicroGrid,” The 3rd IET
International Conference on Power Electronics, Machines
and Drives, pp. 316 – 320.
R. Gules, P. D. Pellegrin, J. Hey, H. L. J. Imhoff, 2008, “A
Maximum Power Point Tracking System With Parallel
Connection for PV Stand-Alone Applications,” IEEE
Transactions on Industrial Electronics, vol. 55, Issue 7,
pp. 2674 – 2683.
J. A. Jiang, T. L. Huang, Y. T. Hsiao, C. H. Chen, 2005,
“Maximum Power Tracking for Photovoltaic Power
Systems,” Tamkang Journal of Science and Engineering,
vol. 8, No 2, pp. 147-153.
K. A. EL-Serafi, A. E. Kalas, M. H. Elfar, 2003, “OnLine Maximum Power Tracking In a Stand Alone
Photovoltaic System,” Suez Canal University, Faculty
of Engineering.
T. Esram, P. L. Chapman, 2007, “Comparison of
Photovoltaic Array Maximum power Point Tracking
[15]
6
Techniques,”
IEEE
Transactions
on
Energy
Conversion, Vol 22, No 2, pp. 439 - 449.
N.A Ninad, L.A.C. Lopes, 2007, “Operation of Single
Phase Grid Connected Inverters with Large DC Bus
Voltage Ripple,” IEEE Canada Electrical Power
Conference, EPC, pp. 172 – 176.
S. Wijnbergen, S. W. H. de Haan, J. G. Slootweg, 2005,
“A System for Dispersed Generator Participation in
Voltage Control and Primary Frequency Control,” 36th
IEEE Power Electronics Specialists Conference, pp. 2918
– 2924.
I. H. Altas, A. M. Sharaf, 1996, “A Novel Online MPP
Search Algorithms for PV Arrays,” IEEE Transactions on
Energy Conversion, Vol 11. No 4, pp. 748-754.
I. Serban, C.P. Ion, C. Marinescu, M. Georgescu, 2007,
“Frequency control and Unbalances Compensation in
Autonomous Micro-Grids Supplied by RES,” IEEE
International Electric Machines & Drives Conference,
IEMDC '07, vol.1, pp. 459 – 464.
M.E. Ropp, S. Gonzalez, 2009, “Development of a
MATLAB/Simulink Model of a Single-Phase GridConnected Photovoltaic System,” IEEE transactions on
Energy conversion, vol. 24, Issue 1, pp. 195 – 202.
M. Ciobotaru, T. Kerekes, R. Teodorescu, A. Bouscayrol,
2006, “PV inverter simulation using MATLAB/Simulink
graphical environment and PLECS blockset,” 32nd
Annual IEEE Conference on Industrial Electronics,
IECON, pp. 5313 – 5318.
X. Jinhui, Y. Zhongdong, S. Qipeng, S.Renzhong, 2008,
“Analyze and research of the inverter for grid-connecting
photovoltaic system,” Third International Conference on
Electric Utility Deregulation and Restructuring and Power
Technologies, DRPT, pp. 2530 – 2535.