Professional Documents
Culture Documents
AbstractThere are advantages of installing distributed generation (DG) in distribution systems: for example, improving
reliability, mitigating voltage sags, unloading subtransmission
and transmission system, and sometimes utilizing renewables.
All of these factors have resulted in an increase in the use of
DGs. However, the increase of fault currents in power systems
is a consequence of the appearance of new generation sources.
Some operating and planning limitations may be imposed by the
resulting fault currents. This paper discusses a model of inverter
based DGs which can be used to analyze the dynamic performance
of power systems in the presence of DGs. In a style similar to
protective relaying analysis, three-dimensional plots are used to
depict the behavior of system reactance ( ) and resistance ( )
versus time. These plots depict operating parameters in relation
to zones of protection, and this information is useful for the
coordination of protection systems in the presence of DG.
Index TermsDistributed/dispersed generation, fault calculation, inverters, power distribution, power system protection.
I. INTRODUCTION
CCORDING to the demands of clean energy, high reliability, and enhanced power quality for sensitive loads, the
demand of distributed resources is gradually rising. Moreover,
the distributed resources may decrease or defer the investment of
system upgrades (i.e., transmission line and transformer ratings)
due to increasing power demand. There are many technologies
for distributed resources beyond the conventional synchronous
machine DGs such as fuel cells, wind turbines, solar cells, and
microturbines. These DGs often require power electronic interfaces to interconnect with the utility grid.
Installing DG at a customer site enhances certain aspects of
the power quality of the owners by mitigating the voltage sag
during a fault. Moreover, DG improves the owner reliability as
the back up generator can often be started within 2 minutes. Although there are many advantages of installing DGs, a few operating conflicts cannot be ignored. Conflicts from installation
of DGs, such as changes in coordination of protective devices,
Manuscript received September 22, 2005. This work was supported in part
by the Salt River Project (SRP), Phoenix, AZ and in part by the Power Systems
Engineering Research Center (PSerc). Paper no. TPWRD-00559-2005.
N. Nimpitiwan was with the Department of Electrical Engineering at Arizona
State University, Tempe, AZ 85287-5706 USA. He is now with the Department
of Electrical Engineering at Bangkok University, Pratumthani, Thailand (e-mail:
natthaphob.n@bu.ac.th).
G. T. Heydt and R. Ayyanar are with the Department of Electrical Engineering
at Arizona State University, Tempe, AZ 85287-5706 USA (e-mail: heydt@asu.
edu; rayyanar@asu.edu).
S. Suryanarayanan is with the Center for Advanced Power Systems at
Florida State University, Tallahassee, FL 32306 USA (e-mail: sid.suryanarayanan@ieee.org).
Digital Object Identifier 10.1109/TPWRD.2006.881440
TABLE I
INTERCONNECTION SYSTEM RESPONSES TO ABNORMAL VOLTAGE [4]
635
Also, high frequency terms occur as a result of the transformation. The input of the transformation is the power frequency of
the grid system which is detected by a phase locked loop (PLL)
at the point of common coupling (PCC). The advantage of the
to
transformation is the ability to obtain low frequency
control signals and rapid calculations. In this discussion, it is
assumed that the high frequency terms that occur in the
transformed variables are filtered out.
,
, and
mainly
The time domain variables
consist of power frequency components (i.e., 60 Hz comtransformation is time varying with dc and
ponents). The
,
, and
power frequency components. Therefore,
generally contain dc, 60 Hz, and 120 Hz components.
This is a consequence of the property that the Fourier transform
(FT) of a product is the convolution of the FTs of the component multipliers. If a low pass filter is applied to the vector
, then
results. The latter vector
is nearly dc, under near balanced sinusoidal steady state conditions. For purposes of designing and modeling a controller,
,
,
,
,
, and
rms voltages of
are used. These rms values are denoted as
,
,
, ,
, and . The same notation is used for rms current.
A proportional plus integral (PI) controller minimizes the
, the difference between the reference signal
error signal
and the instantaneous voltages
and
.
The phase controller provides a phase difference,
, beand
(see Fig. 2). A model of the phase controller
tween
is shown in Fig. 4. Inputs to the phase controller, and , are
reference frame. The phase of and can
transformed to
be calculated as,
636
Fig. 5. Output current of the inverter based DG (stand alone) with a load change
at t = 0:15 s.
TABLE II
CHARACTERISTICS OF AN INVERTER BASED DG FOR SIMULATIONS
Fig. 6. Output voltage of the inverter based DG (stand alone) with a load
change at t = 0:15 s.
The cut-off frequency of the low pass filter has to be set at the
appropriate value to attenuate measurement noise. However, the
value must be high enough to provide a good transient response
of the phase controller. The output of the PI controller is added
and used to calculate the required amplitude
to the phase of
and phase of the modulating signal. The last step in the phase
difference controller is to transform the modulating signal back
reference frame. The transformed modulating signal
to the
is used to control the PWM signal generator.
A simulation of stand-alone operation is utilized to demonstrate the cited model, and this simulation assumes a local load
and an inverter based DG. The inverter based DG is disconnected from the grid system (Fig. 1) to show the transient response of the DG due to changes in load. In the case of standalone operation, the inverter requires only the amplitude controller to set the voltage at the reference point. For purposes of
this demonstration, the output power of the inverter varies depending on the load. The inverter based DG parameters used in
the simulation are shown in Table II.
The demonstration of the stand-alone simulation time is from
to 0.3 s with simulation time step,
. The three
phase output voltage is set at 1.0 p.u. or 12.47 kV. To demonstrate the dynamic performance of the inverter based DG, after
reaching the steady state, the load is changed from 3 MW/0.1
. Figs. 5 and 6 show the
MVAr to 5 MW/0.1 MVAr at
current of the inverter based DG and the output line to neutral
voltage, respectively. Note that, in Fig. 6, a voltage swell occurs
during the change of the load. In this case, the voltage swell
is approximately 5% of the nominal operating voltage. Fig. 7
in the amplitude controller
depicts the modulation index
which is used to control the switching of PWM generator.
Fig. 8. Harmonic content of the output current from inverter based DG in standalone operation.
637
domain
equation solvers in both the time domain and in the
(e.g., trapezoidal method, Adams method, modified Rosenbrock
method and the RungeKutta method). This approach allows
one to accurately model the effects of electronic switching. The
results of simulations, such as voltages and currents measured
in the system, can be recorded for further analysis.
The simulation of synchronous machine DGs is derived from
voltage equations and flux linkage equations, expressed in a rotational reference frame. Details of the synchronous machine
model are discussed in [10]. In general, simulation time steps in
the 50 microsecond range are recommended for a 60 Hz system.
The simulation interval in the illustrative examples shown is 01
second. Longer simulation times are unlikely to be needed for
fault current analysis, but could be easily accommodated.
To demonstrate the impact of DGs in a subtransmission
system, a test bed representative of an actual distribution
system is produced for illustrative purposes and this test system
is shown in Fig. 10. Characteristics of the test bed are shown
in Table III. The test bed system is connected to a 230 kV
transmission system at Bus1 which itself is modeled as a
Thevenin equivalent obtained from conventional short circuit
analysis. The voltage level at the 230 kV supply is stepped
down to 69 kV. The taps of the 230 kV substation transformer
and the 12 kV distribution transformers usually operate higher
than 1.0 p.u. to mitigate the effect of voltage drop in the
distribution primary system. The Thevenin equivalent positive and negative sequence impedance of the 230 kV bus is
ohms per phase for the demonstration below.
Note that capacitors are installed at the load sites to improve
the power factor. The per unit base used in all illustrative cases
is 100 MVA.
IV. SIMULATION RESULTS
In this section, the simulation technique is used to investigate
and compare the impact of the presence of both synchronous
machine DGs and the inverter based DGs in the subtransmission
system. The subtransmission system (Fig. 10) is used as a test
bed to illustrate the simulation technique.
Assume that DGs are installed at 12.47 kV buses at four locations: Bus16, Bus24, Bus25 and Bus27 (Fig. 10). Illustrative
cases are separated into three parts: no DGs in the system in
Case 1, the system with synchronous machine DGs in Case 2
and the system with inverter based DGs in Case 3. The generation capacity of each synchronous machine DG in Case 2 and
inverter based DG in Case 3 are 5 MW. Parameters of all synchronous machine DGs applied in the simulations (in Case 2)
are identical and shown in Table IV. Similar to the synchronous
machine DGs, all machine parameters of the inverter based DGs
(in Case 3) are identical and are described in Table II. Summary
of the illustrative cases is provided in Table V. In these particular cases, the penetration level in the system is approximately
7% based on the total load of the test bed system (300 MVA).
The intent is to illustrate a modest penetration of DGs and the
resultant fault current response.
After analyzing the results of simulation with various fault
situations (i.e., various type of faults, fault locations), it is found
that the installation of DGs (at the specific locations in the test
bed system) may disrupt the operation of the protective relays
638
TABLE III
CHARACTERISTIC OF THE TEST BED SYSTEM
located at Bus16 (12.47 kV bus). For this particular case, misoperation of the protective relay can be illustrated by a three phase
bolted fault at the midpoint of the subtransmission line between
Bus14 and Bus23 (69 kV). Assume that the fault occurs at
and is cleared at 500 ms or 100 ms after fault occurs. This
100 ms fault is a commonly selected duration time for fault anal, resistance
, and reactance
ysis. The impedance
(i.e., seen from the line terminals) are measured at various locations to monitor the consequences of the installation of DGs.
For the purposes of this presentation, the notation and procedures of power system protective relaying are used with regard
to , , and . These quantities are ratios of rms volts to rms
amps, and the rms is carried out over one cycle (i.e., 1/60 s).
The measurement at Bus16 is detected at the PCC or the point
of DG connection on the secondary side of 69/12.47 kV distribution transformer. Note that these measurements are the input
to the protective relays at Bus16 (12.47 kV). The analysis is
performed by assuming that all the DGs have constant power
output. Loads in the system at 12.47 kV buses are considered as
constant impedance loads.
Fig. 10 shows the one line diagram of the test bed system used
for the simulation case studies. The several plots in Fig. 11 will
be discussed.
For the purposes of illustration and comprehension, two
graphic devices are used: a two-dimensional plot of versus
and three-dimensional plots of
versus time. A
three-dimensional plot can be used to depict the behavior of
TABLE IV
SYNCHRONOUS MACHINE PARAMETERS
TABLE V
SUMMARY OF THE ILLUSTRATIVE CASES
and versus time. The three-dimensional plots in some commercial software are equipped with the ability to zoom-in,
zoom-out and rotate graphs to any desired orientation. This
ability cannot be completely demonstrated in a paper document.
versus time are illustrated from
However, the plots of ,
two distinct vantage points in Fig. 11. Although interesting in
appearance, these plots may be difficult to interpret and, thereversus
fore, a two-dimensional plot may be used, namely
. The concepts and applications of graphical illustrations on
protective relaying evaluation and testing are discussed in [11].
639
Fig. 11. Plots of X R and X R-time at Bus16 (12.47 kV) illustrating the visualization afforded by rotation around the vertical (reactance) axis. (a) Plots of
X R and X R versus time at Bus16 (12.47 kV) of the test bed system without DG, Case 1. (b) Plots of X R and X R versus time at Bus16 (12.47 kV)
of the test bed system with synchronous machine DGs, Case 2. (c) Plots of X R and X R-time at Bus16 (12.47 kV) illustrating the visualization afforded
by rotation around the vertical (reactance) axis.
640
Fig. 12. Comparison of the fault currents (p.u.) from three phase to ground
bolted fault at the middle of the line between Bus14Bus23 (69 kV), Cases 13
(per unit base: 100 MVA, 69 kV).
V. CONCLUSIONS
Fig. 13. Plot of time to operation of the distance relay at bus Bus16 (12.47 kV)
versus reach of the relay, Case 2 and 3 (per unit base: 100 MVA, 12.47 kV).
seen to start at
,
and move toward the secondary
protection boundary [see the leftmost plot of Fig. 11(b)]. The
operational trajectory is seen also in the center and rightmost
plots of Fig. 11(b). Fig. 11(c) depicts similar results for Case 3.
Conclusions can be drawn from the simulations as: installation of DGs in the distribution system increases the fault current throughout the system. This situation may change the way
protective relays react to the faults. Circuit breakers, fuses and
setting of relays may need to be adjusted to the new appropriate
range. For a person skilled in the interpretation of the three-dimensional plots, the rotational and zoom features may be used
to quickly determine protective relay settings.
The simulation results show that the increase of fault current from synchronous machine DGs is higher than that of inverter based DGs of identical ratings. Fault currents from the
three-phase to ground bolted fault in Cases 13 are shown in
Fig. 12. In order to compare fault currents with and without DG,
let the no DG case for the test bed system result in a fault
This paper discusses the impact of installation of DGs in distribution systems from the perspective of increase in fault current and protection coordination. The fault current calculation
matrix is comof the inverter based DGs by applying the
plicated because inverters are nonlinear electronic devices. This
problem can be approached by applying dynamic simulation
contechniques. A model of inverter based DGs based on
trols is discussed and compared with the synchronous machine
counterpart. Simulation techniques are proposed to investigate
the operation of protection systems. Information from the dynamic simulations is useful in protective system coordination.
The use of a novel three dimensional , , depiction is illustrated in which graphic rotation and zoom features are used.
Simulation results show that the increase in fault currents
is often greater in the synchronous machine implementation
versus a comparable inverter based design.
ACKNOWLEDGMENT
The authors would like to thank J. Blevins, A. B. Cummings,
and R. Thallam for their technical expertise and advice.
REFERENCES
[1] N. Nimpitiwan and G. T. Heydt, Fault current issues for market
driven power systems with distributed generation, in Proc. North
Amer. Power Symp., Moscow, Idaho, Aug. 2004, pp. 400406.
[2] R. C. Dugan and T. E. McDermott, Operating conflicts for distributed
generation on distribution systems, in Proc. Rural Electric Power
Conf., 2001, pp. A3/1A3/6.
[3] R. C. Dugan and S. K. Price, Issues for distributed generation in the
US, in Proc. IEEE Power Eng. Soc. Winter Meeting, 2002, vol. 1, pp.
121126.
[4] IEEE Standard for Interconnecting Distributed Resources With Electric Power Systems, IEEE Std. 1547-2003, Aug. 2003.
[5] G. T. Heydt, Computer Analysis Methods for Power Systems. Scottsdale, AZ: Stars in a Circle, 1996.
[6] M. Prodanovic and T. C. Green, Control of power quality in inverterbased distributed generation, in Proc. IEEE Ind. Electron. Soc. Conf.,
2002, vol. 2, pp. 11851189.
[7] M. N. Marwali and A. Keyhani, Control of distributed generation systems-Part I: voltages and currents control, IEEE Trans. Power Electron., vol. 19, no. 6, pp. 15411550, Nov. 2004.
[8] C. J. Hatziadoniu, A. A. Lobo, F. Pourboghrat, and M. Daneshdoost,
A simplified dynamic model of grid-connected fuel-cell generators,
IEEE Trans. Power Del., vol. 17, no. 2, pp. 467473, Apr. 2002.
[9] T. Petru and T. Thiringer, Modeling of wind turbines for power system
studies, IEEE Trans. Power Syst., vol. 17, no. 4, pp. 11321139, Nov.
2002.
[10] C. M. Ong, Dynamic Simulation of Electric Machinery: Using Matlab/
Simulink. Upper Saddle River, NJ: Prentice-Hall, 1998.
[11] A. P. Sakis Meliopoulos and G. J. Cokkinides, Visualization and animation of protective relay operation, in Proc. Power Eng. Soc. Winter
Meeting, 2002, vol. 2, pp. 14101414.
Natthaphob Nimpitiwan (S01) received the Ph.D.
degree in electrical engineering from Arizona State
University, Tempe.
Currently, he is a Faculty Member in the Department of Electrical Engineering at Bangkok
University, Pratumthani, Thailand. His research
interests include distributed/dispersed generation,
modeling/simulation of power systems, artificial
neural networks, and engineering education.
641
Raja Ayyanar (S97M00) received the M.S. degree from the Indian Institute of Science, Bangalore,
and the Ph.D. degree from the University of Minnesota, Minneapolis.
Currently, he is an Assistant Professor with Arizona State University, Tempe. He has many years of
industrial experience designing switch-mode power
supplies. His research interests include topologies for
modern dcdc converters, new pulsewidth modulation techniques for drives, and power electronics applications in power systems.