Simulation of Smart Functionalities of Photovoltaic

Inverters by Interfacing OpenDSS and Matlab

Meghasai, S. Monger, R. Vega and H. Krishnaswami

Department of Electrical and Computer Engineering and Texas Sustainable Energy Research Institute
University of Texas at San Antonio
San Antonio, Texas, USA

Abstract— With the increase in demand for clean and green operation issues and impacts on other PV systems connected to
power, several energy utilities are increasing renewable energy in the same LV circuit [3].
their generation portfolio. The high penetration of PV can create
problems of voltage fluctuation and frequency variation. The two Finding a software that can perform QSTS in all possible
major smart functionalities that address high PV penetration modes of operation is difficult. It can only be done, if the
include dynamic Volt/VAR control for voltage fluctuations and software offers a Common Object Model (COM) interface
active power control and curtailment to achieve frequency capability to run it through an external program and utilize the
regulation. Quasi-static time series simulation of the power grid capabilities of the external program to have an enhanced
that includes the two smart functionalities of the Photovoltaic evaluation [1]. The basic features required from a distributed
(PV) generator is needed. In this paper, the modeling and analysis system analysis tool with specific reference to the PV inverters
of the power grid with PV generators is presented using the open include power flow analysis, time-series analysis with desired
source simulation tool OpenDSS. Specifically, the COM interface time interval, GIS data integration, PV generator models, PV
capability of OpenDSS is used to interface with Matlab to inverter control, PV insolation data, voltage regulation and
simulate the advanced functionalities of Volt/VAR and active transient analysis.
power control and at the same time observe the effect of these
controls in the power grid. The IEEE 13 bus feeder is used to With high penetration of PV, the voltage levels in the grid
demonstrate the usefulness of the proposed simulation begin to oscillate and cause a major concern to the utilities and
methodology interfacing both Matlab and OpenDSS. the customers. One of the solutions is to implement dynamic
Volt/VAR control with advanced functionalities integrated into
Keywords— PhotoVoltaic Systems, Frequency Regulation, the PV inverter. A number of strategies have been described in
Volt/VAR control, OpenDSS [4-6], to mitigate the concern of fluctuating voltage.
I. INTRODUCTION The variable nature of PV is due to the fact that it is
Solar power has been expanding at a rapid pace during the dependent on weather conditions, seasonal conditions, seasonal
past eight years. The cost per kWh has been dropping with the changes and geographical changes. Such a dynamic behavior
average price of a completed PV system recording a drop of has also an adverse effect over the system frequency and causes
33% since 2011 [1]. As solar is becoming cost-competitive with the active power output to fluctuate which is the other cause of
other forms of energy generation, there has been an increase in concern for the utilities. Several strategies have been proposed
the solar PV installations worldwide. Many distribution circuits in [7-12] to mitigate the effect on frequency. High penetration
have already reached high penetration levels of PV. There are of PV makes it necessary to have an effective real-time
several undesirable effects of high PV penetration that include communication and control between the utilities and PV plant
voltage excursions, harmonic distortion and system power loss operators. Utilities having a simulation tool can analyze the
characteristics [1]. effect of PV on the feeder and can evaluate different control
methodologies. This will help to implement a control that would
Quasi Static Time Series Simulation (QSTS) is a useful take care of the concerns distributed PV causes on the feeder.
method for analyzing the impact of distributed PV on the grid
[2]. The PV power output is modelled as a time series and a This paper focuses on a simulation methodology that
power flow solution is performed for every time instant, utilizes the Common Object Model (COM) capabilities of
assuming that the system reaches a quasi-steady state. The OpenDSS, an open source platform to perform Quasi Static
results can then be plotted as a function of time. Time Series Analysis with dynamic Volt/VAR and active
power control. Matlab has been used as an external program to
Voltage regulation devices have evolved over time and invoke OpenDSS through COM. The handles provided after
provide complex and customized control modes for many solving the circuit have been used to achieve the Volt-VAR
different applications. However, with high PV penetration and control and in curtailing the active power. Simulation strategies
variable nature of PV, operation of distribution system and adopted to visually identify the effects of Volt-VAR control and
voltage regulation devices is challenging. The variable nature power curtailment under frequency regulation on the smart
of solar also affects the voltages on the secondary low voltage inverters are further enunciated.
(LV) side of distribution transformers, causing load/equipment

 

  
 II. QUASI STATIC TIME SERIES ANALYSIS USING IEEE 13 an open source and it provides a platform to model the feeder
NODE FEEDER integrated with distributed source of energy and perform a
Quasi Static Time Series Analysis and finally analyze the same
A. System Description
in real-time. Real-time simulations help in analyzing the feeder
A number of power flow simulation softwares are available models that could help make momentary changes in the
that provide us with ease of solving complex radial networks equipment in real-time and mitigate the concerns of voltage
both balanced and unbalanced. But it becomes necessary to fluctuations, harmonic distortions, frequency regulation. The
provide with a standard that could be used in all currently most important feature of OpenDSS is that it can solve
available platforms and compare the results. IEEE has provided
unbalanced radial feeders and it uses Newton Raphson method
a set of defined test feeders that can be used in any platform and
to perform power flow analysis. It provides a user friendly GUI
the results could be compared. The present paper uses an IEEE
13 Bus Node Feeder for analysis. Fig. 1 gives a one line diagram that helps modeling the feeder in a clear way with each element
of the IEEE Node Test Feeder. This feeder is very small and yet being defined separately. It supports different solving modes
displays some very interesting characteristics. It actually that help users to perform this analysis at a wider range. As it
displays an approximate model of the small radial feeder for is an open source software, programmers are free to make
the most common feature of distribution system. changes in the source code and utilize the properties of
OpenDSS as desired.
Three PV generators each rated at 500 KW are connected at
bus 675 to simulate Volt/VAR control and at bus 680 to The COM interface capability can help control OpenDSS
simulate active power curtailment. The PV power profile of all through an external program like Python, Matlab, C#,
the three generators and for both the simulations is assumed to
GridLabD, Lab View and make use of the functionalities of
be the same. Fig. 2 gives the PV power profile used for three
PV generators. In Fig. 2, 30 different points have been these external programs to perform studies at a higher level.
considered. QSTS is performed for a period of 24 hours taking
30 different intervals at a step size of 0.8hours for analysis. C. Methodology
This paper focuses on a simulation methodolgy shown in
Fig. 3 that utilizes the COM capabilites of OpenDSS to perform
QSTS and simulate the capabilities in Volt/VAR and frequency
regulation. Matlab command environment is used as to invoke
OpenDSS through COM. It can be seen in Fig 3 that data can
be transmitted to and from OpenDSS and Matlab through the
COM interface. With the help of this tool, QSTS can be
performed and the feeder can be analyzed with PV integrated
on it. Matlab utilizes the data received from OpenDSS to vary
the parameters to perform Volt/VAR and Active Power control.

Fig 1. One Line Diagram of IEEE 13 Node Test Feeder

Fig 3. Proposed Model of OpenDSS and Matlab interface

Having OpenDSS perform QSTS on a distribution circuit

Fig 2. Power profile for PV generators combined with the various control methodolgies in Matlab for
mitigating the concerns of high penetration of PV is a valuable
addition for power system planners. Hence, in order to make the
B. Simulation Software open sorce a much better tool, it is interfaced with Matlab in
OpenDSS is one such software developed by EPRI for order to utilize the full capabilities of Matlab and have a much
utilities that offers the COM interface capability. OpenDSS is detailed analysis of the grid and control over the PV generators.
The following two paragraphs explain the simulation of the interface. As the power fluctuates, it becomes necessary to
Volt/VAR and active power control through the proposed control it and keep it within limits as demanded by utilities for
methodology in Fig. 3. curtailment.
1) Volt/VAR Control As OpenDSS does not analyze change in frequency, a
Recently, the possibility of incorporating VAR control simulation strategy is adopted that allows the user to have an
capability with PV inverters for network support have been On/Off control of PV inverters that would help curtail active
reported in a number of papers. High penetration of PV causes power and contribute to frequency control.
a high degree of unpredictability in the operation of distribution
feeder every hour and day due to erratic behavior of PV output. III. RESULTS
It is because of the PV output dependency upon the solar A. Volt/VAR Control
radiation reaching earth’s surface and being tapped by the solar The feeder is solved for a 24 hour period. Voltage profile in
panels. This has a great impact over the flow of power and the Line 684-611 is plotted and shown in Fig. 4.
voltage profile of the feeder. A cloudy day has a low PV output
and a large amount of power is drawn from the grid creating a
low voltage profile in the feeder.
The traditional way of controlling varying feeder voltage is
to use voltage regulators and tap changers that would maintain
the voltage levels during fluctuations or switching of capacitor
banks that would reduce the reactive power in the feeder when
ON or boost the reactive power in the feeder when OFF. In
micro-grids, which are radial the concerns of fluctuating voltage
is even higher.
For utilities, it becomes a fundamental practice to supervise
VAR in the grid and execute Volt/VAR control. This is possible
if customers having standalone PV inverters and PV plants can
contribute to the control.
This paper focuses on demonstrating in simulation dynamic
VAR compensation using OpenDSS-Matlab tool. Dynamic
VAR compensation approach refers to providing compensation Fig 4. Voltage V1 in Line 684-611
by injecting VAR into the grid by customers or PV operators.
This will ensure a good voltage profile in the distribution
network and will be within ANSI limits of +/- 3%. It is observed that the voltage drops below the ANSI limits
in line 684-611 when three PV generators are connected at Bus
2) Active Power Control 675. Therefore, it becomes necessary to provide a dynamic
It is necessary for the utilities to maintain the frequency in VAR compensation during this period to boost the voltage in
the grid and supply power to consumers at 60Hz. As the utilities the feeder to maintain the voltage within limits. Fig. 5 shows
are moving towards high penetration of PV, the frequency starts the voltage V1 in Line 684-611 after VAR compensation and
fluctuating. This happens as the active power of the PV is not Fig. 6 gives the comparison plot.
constant and is dependent on the irradiance value, temperature
of the cell and the weather conditions that are not constant
during the entire day. Hence, ensuring a constant frequency
becomes a major concern for utilities. The other concern steps-
out when PV is generating at peak and the total amount of power
to be delivered needs to be curtailed. Therefore, taking
necessary actions to control PV output becomes important.
This paper focuses on demonstrating in simulation
curtailment of active power output of PV generators to reduce
the total output power of PV being delivered to the grid and
contribute to maintaining the frequency within desired limits.
Power grid operators in such situations can demand the PV
plants to curtail their power generation. Under power
curtailment, it is beneficial in terms of cost for the PV plants to
maintain the active power at a constant level through control of
smart inverters.
An OpenDSS-Matlab interface is used to simulate this
condition and observe the effects of power flow in the Fig 5. Voltage V1 in Line 684-611 after VAR control
distribution grid. The circuit is solved using OpenDSS through
 It can be observed from Fig. 7 that due to varying irradiance
profile the output of PV is not consistent and causes the power
to fluctuate in Line 671-680. It becomes necessary to control it
and keep it within limits demanded by utilities for curtailment.
As observed in Fig. 7, at any point below 2405 kW, active
power remains constant and anything above that value, power
begins to fluctuate at higher rate. Matlab senses the change in
the voltage at Bus 680 and alters the KW value of the three PV
generators in order to respond to curtailment in the Line 671-
680. Fig. 8 gives the summary of the Active Power in Line 671-
680 with PV and after curtailment. It can be noted from Fig. 8
that the PV generator output has been reduced after control. The
plot shows a basic control mechanism and in future case studies,
more advanced controllers for active power control can be
implemented through the Matlab and COM interface.

Fig 6. Voltage V1 in Line 684-611 with PV and after VAR control

From Fig. 5 it is evident that the voltage levels rise and is

brought within the limits. This happens because the three PV
generators connected at Bus 675 are commanded to inject VAR
to counterbalance the voltage levels and bring them within the
ANSI limits. Fig. 6 gives a summary of the change in voltage
levels with PV and after VAR compensation. Matlab plays an
important role as it senses this drop using COM interface and
taking voltage at Bus 675 as a reference changes the VAR
output of the PV generators in order to boost the voltage. The
proposed method of using COM interface to control the PV
generator as the voltage changes is possible as shown in Fig. 6
and will be a valuable tool for utilities.
B. Active Power Control
Fig 8. Active Power in Line 671-680 with PV and after curtailment
Three PV generators are integrated at Bus 680 and the
circuit is solved for 24 hours using OpenDSS through the
interface. Active Power in Line 671-680 is then measured and
plotted shown in Fig. 7. IV. CONLCUSION
Volt-VAR and Volt-Watt curve were defined by EPRI as
the best methods to control VAR and active power output of the
PV generator. The capabilities of OpenDSS were utilized to
interface it with Matlab and the volt-var control and active
power curtailment were performed through the Matlab by
writing basic program to control the functioning of the PV
generators. An ON and OFF control was defined for curtailing
the active power output of the PV generators in order to respond
to frequency variations.
ACKNOWLEDGMENT
The authors would like to thank their research sponsor CPS
Energy for their generous support. This research was made
possible by the Research Alliance between the San Antonio
electric utility CPS Energy and the Texas Sustainable Energy
Research Institute of The University of Texas at San Antonio.
REFERENCES
Fig 7. Active power in Line 671-680 with PV [1] J. E. Quiroz, M. J. Reno, and R. J. Broderick, “Time series simulation of
voltage regulation device control modes,” Conf. Rec. IEEE Photovolt.
Spec. Conf., pp. 1700–1705, 2013.
[2] B. A. Mather, “Quasi-static time-series test feeder for PV integration [8] H. Xin, Y. Liu, Z. Wang, D. Gan, and T. Yang, “A New Frequency
analysis on distribution systems,” IEEE Power Energy Soc. Gen. Meet., Regulation Strategy for Photovoltaic Systems Without Energy Storage,”
pp. 1–8, 2012. IEEE Trans. Sustain. Energy, vol. 4, no. 4, pp. 985–993, 2013.
[3] M. J. Reno and K. Coogan, “Grid Integrated Distributed PV (GridPV),” [9] D. Kottick, D. Edelstein, and M. Blau, “Battery energy storage for
August, 2013. frequency regulation in an island power system,” IEEE Trans. Energy
[4] R. Ayyanar, “Distribution System Analysis Tools for Studying High Convers., vol. 8, no. 3, pp. 455–459, 1993.
Penetration of PV with Grid Support Features,” PSERC Webinar, 2013. [10] Ota, T.; Mizuno, K.; Yukita, K.; Nakano, H.; Goto, Y.; Ichiyanagi, K.,
[5] W. Sunderman, R. C. Dugan and J. Smith, “Open Source Modeling of "Study of load frequency control for a microgrid," Power Engineering
Advanced Inverter Functions for Solar Photovoltaic Installations,” T&D Conference, 2007. AUPEC 2007. Australasian Universities , vol., no.,
Conference and Exposition, 2014 IEEE PES , vol., no., pp.1,5, 14-17 pp.1,6, 9-12 Dec. 2007.
April 2014 [11] Omran, W.A.; Kazerani, M.; Salama, M.M.A., "Investigation of Methods
[6] W. Zhang, M. Baran, A. De, and S. Bhattacharya, “Fast Volt – VAR for Reduction of Power Fluctuations Generated From Large Grid-
Control on PV Dominated distribution systems," T&D Conference and Connected Photovoltaic Systems," Energy Conversion, IEEE
Exposition, 2014 IEEE PES , vol., no., pp.1,5, 14-17 April 2014. Transactions on , vol.26, no.1, pp.318,327, March 2011.
[7] M. Datta, T. Senjyu, A. Yona, T. Funabashi, and C. H. Kim, “A [12] A. Ahmed, L. Ran, S. Moon, and J. H. Park, “A fast PV power tracking
frequency-control approach by photovoltaic generator in a PV-diesel control algorithm with reduced power mode,” IEEE Trans. Energy
hybrid power system,” IEEE Trans. Energy Convers., vol. 26, no. 2, pp. Convers., vol. 28, no. 3, pp. 565–575, 2013.
559–571, 2011.