International Conference on Renewable Energies and Power Quality (ICREPQ14) Cordoba (Spain), 8th to 10th April, 2014 Renewable

Energy and Power Quality Journal(RE&PQJ) ISSN 2172-038 X, No.12, April 2014

Evaluation of Medium Voltage Distribution Network with/without Distributed Generation using ATP Software
J. O. Rezende1, L. M. Peres1, G. C. Guimares1, A. J. de Moraes 1and M. A. Tamashiro1
Faculty of Electrical Engineering Federal University of Uberlndia Campus Santa Mnica Uberlndia, Minas Gerais (Brazil) Power System Dynamics Group Phone/Fax number:+553432394758, e-mail: jaqueline.oliveirarezende@gmail.com, larissaa_mp@yahoo.com.br, gcaixeta@ufu.br, ajmoraes@ufu.br, tamashiro@ifto.edu.br
1

Abstract.

This research has as main objective to analyze the contributions of a distributed generation unit when connected to a medium voltage distribution feeder (13.8 kV). The behavior this electrical system is evaluated following the occurrence of some disturbances. Before this, all the modeling of the electrical system was were performed using the ATP software. Three types of disturbances were simulated: phase-to-ground short circuit, islanding of part of the electrical system and load increase on the feeder. To easy the comparisons, all studies were carried out with and without the presence of the distributed generator taking into account the following variables: busbar voltage profile, output power provided by the substation and by the distributed generator when present.

Key words
ATP software, distributed generation, medium voltage distribution network, output power, voltage profile.

1. Introduction
Electricity is an essential element for the development of various sectors of the economy. The demand for this resource has shown an impressive growth over the years. The main way to reach this demand is to use large power generating plants. This type of generation is usually located away from the centers of consumption. Thus, it is necessary to use the transmission and distribution systems to provide electricity to consumers [1]. Furthermore, the construction of large power plants needs to fulfill various requirements such as high financial investments, detailed environmental studies, among others. These factors often hinder the construction of new generating units and contribute to the delay in its accomplishment [2]. In this context, it is notorious that various sectors are seeking new ways to supply the demand for electricity.

Thus, one way of generation is the distributed generation (DG) and what is being widely researched. Many experts predict that the installation of these new sources of power generation will increase due to the need to provide more clean energy [3]. Distributed generation when compared to large power plants has as main features: location near to the consumer, lower construction time and lower energy distribution costs [4]. Moreover, the connection of distributed generation units contributes to postponement or reduction of investments in construction to reinforcement of the electrical system and also decreases the power demand from the electric company [5]. In Brazil, the distributed generation is regulated by Resolution n 482 of 17 April 2012 and the Module 3: Access to the Distribution System, contained in PRODIST (Procedures for Distribution Network in the National Electric System), both under the responsibility of the National Electric Energy Agency (ANEEL). The first resolution establishes the conditions for access to the distribution system, the power compensation system, the energy measurements and the liability for damage to the electrical system [6]. The second treats the specifications for enabling access, project requirements, deployment of new connections, metering system, requirements for operation, maintenance and connection safeness and presents a contract model to adhere to the power compensation system [7]. Thus, this research aims to analyze the impact of a distributed generator connected to a 13.8 kV feeder when disturbances occur in this system. The phenomena studied are short-circuit-to-ground, islanding of part of the electrical system and load increase on the feeder. To evaluate the distributed generator contributions, all the simulations are performed with/without the connection of the distributed generation unit in the electrical system.

2.

Electrical System

Electrical system analysed of the Bandeirantes electricity company has a distribution feeder of 13.8 kV. In this system are present substation and six loads, whose data were taken from reference [2]. Subsequently, the feeder will be connected to a distributed generator.

ATP ("Alternative Transients Program") software was used to perform the simulations. It is need the appropriate modeling of the data and the parameters of the electrical system for the use of this software, as will be presented in the next section. Figure 1 shows the configuration of the electrical system studied with all data represented in the ATP software, identifying the presence of a distributed generator equipped with voltage regulators and speed.

Fig. 1 Schematic of the electrical system in the ATP software.

3.

Modelling of Electrical System

To represent the electrical system in the software ATP was necessary to model all components of the system, as is detailed below. A. Modeling of Substation The power substation was represented as a constant voltage source (infinite bus) in series with an inductance. For the module voltage was considered the nominal feeder (13.8 kV). It was used a free software called UFUFlow [8] for processing a program load flow and determine the phase angle of this voltage source. The values of active power and reactive power from the substation to the representation of the electrical system were found in reference [2]. These values are 7 MW of active power and 2.3 MVAr reactive power. Thus, after processing software UFUFlow, the value of the angle of the substation bus voltage was determined in 3.23. B. Modeling of Loads The demands of the loads were also taken from reference [2], which also shows the values of active power and

reactive power. These loads were represented in software by ATP corresponding values of resistance and inductance (constant impedance). Thus, after the calculations, the values obtained for the loads can be seen in Table I.
TABLE I - Loads Values
Bus CG2 CG3 CG4 CG5 CG6 CG7 P[kW] 870 800 1160 1260 470 2170 Q[kVAr] 275 255 370 400 150 690 R[] 193,86 196,51 132,85 121,48 323,83 69,54 L[mH] 162,52 166,16 112,39 102,28 274,14 58,65

C. Modeling of Distribution Lines The lengths of the distribution lines for the modeling by ATP software were also obtained from reference [2]. The conductor used in modeling is the CAA 336.4 and technical data were obtained from the manufacturer Nexans's catalog [9]. The values found for the resistance, inductive reactance and capacitive reactance were, respectively, 0.2051 /km, 0.3770 /km and 0.2266 M..km. By having the parameters of the line, then, it can be represented by the model described by their PI resistance and inductance arranged in series, and their

overall capacitance placed in parallel so that each half occupies one end of the line. The parameters of each branch of the line can be seen in Table II.
TABLE II- Parameters of Line Distribution
Branch SE1 - CG2 CG2 - CG3 CG3 - CG4 CG4 - CG5 CG5 - CG6 CG6 - CG7 CG7 - CG8 Lenght [m] 1000 3000 1000 500 500 1000 1000 R [] 0,2051 0,6153 0.2051 0,1025 0,1025 0,2051 0,2051 L [mH] 1,000 3,000 1,000 0,500 0,500 1,000 1,000 C [F] 0,0117 0,0351 0,0117 0,0058 0,0058 0,0117 0,0117

TABLE VII - Initial Values of Active and Reactive Powers Provided

With DG P [MW] Q [MVAr] 5,04 0,92 2,51 3,49 Without DG P [MW] Q [MVAr] 7,00 2,6 -

SE1 DG

5.

Case Studies

D. Modeling of Distributed Generator The distributed generator used in the simulation has the power of 5 MVA, 60 Hz frequency and 4 poles. The other parameters of the generator were taken from reference [10] and can be seen in Table III. It was used the synchronous machine controlled type SM59, provided by ATP, to represent the generator.
TABLE III - Parameters of Distributed Generator
H 0,68 s Xd 0,23 pu Xd 1,31 pu Xq 0,75 pu Xd 0,16 pu Xq 0,17 pu Tdo 3,58 s Tdo 0,012 s Tqo 0,08 s

The case studies allowed to evaluate the impact of the distributed generator connected to a distribution feeder of 13.8 kV, in the three disturbances situations. All simulations were performed with and without this connection to evaluate the contribution of generator. A. Short Circuit Phase to Ground The short-circuit phase-to-ground solid was simulated for being what else happens in the electrical system. Therefore, it was chosen a point located 984.25 ft to bus CG5 and 656.17 ft to bus CG6 in phase C, line CG5 CG6 of Figure 1. The short circuit started at 2s. Table VIII shows the final values of the phase voltages after end of the disturbance, with presence of the distributed generator.
TABLE VIII - Voltages Phase: System with Short Circuit and DG
Bus SE1 CG2 CG3 CG4 CG5 CG6 CG7 CG8 Va [kV] 7,97 7,81 7,38 7,26 7,21 7,17 7,10 7,10 [pu] 1,000 0,979 0,926 0,911 0,905 0,899 0,891 0,891 [kV] 7,97 7,77 7,27 7,13 7,07 7,03 6,63 6,63 Vb [pu] 1,000 0,975 0,912 0,895 0,887 0,882 0,832 0,832 [kV] 7,97 6,58 2,46 1,09 0,41 0,13 0,82 0,82 Vc [pu] 1,000 0,825 0,309 0,137 0,051 0,016 0,104 0,104

E. Modeling of voltage and speed regulators Voltage and speed regulators were modeled in the software according to the references [11] and [12], with the parameters obtained from reference [10]. These data can be seen in Table IV and Table V, respectively.
TABLE IV - Parameters of Voltage Regulator
Ka 400 pu Ta 0,02 s Kf 0,03 pu Tf 1s Ke 1 pu Te 0,8 s Vmax 6,6 pu Vmim 0 pu

TABLE V - Parameters of Speed Regulator

Flyballs 0,01 pu Reg 0,05 pu Tfb 0,2 s T1 0,3 s T2 1,0 s T3 1,0 s T4 0s T5 0,1 s

4.

Electrical System in Steady State

The first simulation was done for the initial steady state condition of the electrical system without disturbance, to voltages values (Table VI) and potencies (Table VII), with and without the presence of the distributed generator.
TABLE VI - Initial Values of Line Voltage and Phase

Table VIII shows that the short circuit occurred in phase C, once this stage has the lowest voltages values. It is also observed that the three phases of the bus CG6, CG7 CG8 have the lowest voltages values, when compared to predisturbance values, shown in Table VI. With the presence of short-circuit, substation went to supply to 16.7 MW of active power and 27.6 MVAr reactive power. The distributed generator provides 1.16 MW of active power and reactive power -1.37 MVAr. The behaviors of the substation and power generator can be seen, respectively, in Fig 2 and Fig 3.
30 *10 6 25 20

Bus SE1 CG2 CG3 CG4 CG5 CG6 CG7 CG8

With DG Vlinha[kV] Vfase[kV] 13,80 7,97 13,76 7,94 13,68 7,89 13,68 7,89 13,69 7,90 13,72 7,92 13,80 7,97 13,80 7,97

Without DG Vlinha[kV] Vfase[kV] 13,80 7,97 13,62 7,86 13,17 7,60 13,04 7,53 12,99 7,50 12,96 7,48 12,91 7,45 12,91 7,45

15 10 5 0 -5 0 2 4 6
m:QT

[s]

10

(f ile SimulaoCongresso.19_CurtoFaseTerra.pl4; x-v ar t) m:PT

Fig. 2 Active power (red) and reactive power (green) in the SE1 electrical system with DG and short circuit.

4 *10 6 3

this machine has enough capacity to supply power to loads islanded. Line voltages with islanding, considering distributed generator connected to the electrical system, are presented in Table X.
TABLE X - Line Voltages - System with Islanding and with DG

-1

-2 0 2 4 6
m:QT

[s]

10

(f ile SimulaoCongresso.19_CurtoFaseTerra.pl4; x-v ar t) m:PT

Fig. 3 Active power (red) and reactive power (green) distributed generator with short-circuit.

To facilitate comparison, distributed generator was disconnected from the system and then applied the same disturbance. For this condition, Table IX shows the final values after phase voltage disturbance on all buses.
TABLE IX - Voltages Phase - System with Short Circuit and without DG
Bus SE1 CG2 CG3 CG4 CG5 CG6 CG7 CG8 Va [kV] 7,97 7,87 7,61 7,53 7,50 7,48 7,45 7,45 [pu] 1,000 0,987 0,955 0,945 0,941 0,938 0,935 0,935 [kV] 7,97 7,87 7,61 7,53 7,50 7,48 7,45 7,45 Vb [pu] 1,000 0,987 0,955 0,945 0,941 0,938 0,935 0,935 [kV] 7,97 6,58 2,46 1,09 0,41 0,00 0,00 0,00 Vc [pu] 1,000 0,825 0,309 0,137 0,051 0,000 0,000 0,000

Bus SE1 CG2 CG3 CG4 CG5 CG6 CG7 CG8

Voltage [kV] 13,80 13,70 13,45 13,39 13,37 13,78 13,80 13,80

Voltage [pu] 1,000 0,993 0,975 0,970 0,969 0,999 1,000 1,000

Looking at Table X, it can be seen that buses islanded, CG6, CG7 and CG8, have voltages within proper limits established by ANEEL [13]. The substation with islanding provided 4.5 MW of active power and 1.49 MVAr reactive power. The distributed generator supplied active power 3.08 MW and 0.78 MVAr reactive power. The graphs of power in SE1 and DG can be seen, respectively, in Fig. 5 and 6.
6 *10 6 5

Table IX indicates the voltage drops only in phase C and maintained the same voltage values in steady state, in phases A and B (due to the infinite bus). Phase C of buses CG6, CG7 and CG8 were de-energized. When the substation supplies power to all system loads, it provides 17.1 MW of active power and 24.2 MVAr reactive power. The graph of active and reactive power can be seen in Figure 4.
30 *10 6 25

0 0 2 4 6 8 [s] 10
(f ile SimulaoCongresso.20_Ilhamento.pl4; x-v ar t) m:P m:Q factors: 1 3 3 offsets: 0,00E+00 0,00E+00 0,00E+00

Fig. 5 Active power (red) and reactive power (green) in the SE1 electrical system with DG and with islanding.

4,0 *10 6 3,5 3,0

20

2,5
15

2,0 1,5 1,0

10

0,5
0 0 2 4 6
m:QT

[s]

10

0,0 0 2 4
m:QT

(f ile SimulaoCongresso.19_CurtoFaseTerra.pl4; x-v ar t) m:PT

[s]

10

(f ile SimulaoCongresso.20_Ilhamento.pl4; x-v ar t) m:PT

Fig. 4 Active power (red) and reactive power (green) in the SE1 electrical system without DG and with short circuit.

Fig. 6 Active power (red) and reactive power (green) of distributed generator with islanding.

B. Islanding The islanding of the power system at time 2s was the second disturbance considered in the simulations. The islanding was performed by opening the circuit breaker on the output bus CG5, leaving loads CG6 and CG7 with total power of 2.77 MVA (2.64 MW and 0.84 MVAr) provided only by distributed generator of 5 MVA. Thus,

At this time are shown in Table XI, the line voltages without distributed generator and islanding at the output of bus CG5. Loads located close islanding were without power. It can be seen from Table XI shows that the voltages in the bus CG6, CG7 and CG8 are equal to zero. Furthermore, the voltages in the other bus of the electrical system have similar values compared to the system voltages in the presence of the distributed generator.

TABLE XI - Line Voltages - System with islanding and without DG

Bus SE1 CG2 CG3 CG4 CG5 CG6 CG7 CG8 Voltage [kV] 13,80 13,69 13,45 13,38 13,37 0,00 0,00 0,00 Voltage [pu] 1,000 0,992 0,975 0,969 0,967 0,000 0,000 0,000

6 *10 6 5

0 0 2 4 6 8 [s] 10
(f ile SimulaoCongresso.21_AumentoCarga.pl4; x-v ar t) m:P m:Q factors: 1 3 3 offsets: 0,00E+00 0,00E+00 0,00E+00

Fig. 8 Active power (red) and reactive power (green) in the SE1 electrical system with DG and with load increase.
5 *10 6 4

The substation supplied 4.35 MW and 1.49 MVAr active reactive power. The graph of these powers can be seen in Fig 7 below.
8 *10 6 7 6 5 4 3 2 1 0 0 2 4 6 8 [s] 10
(f ile SimulaoCongresso.20_Ilhamento.pl4; x-v ar t) m:P m:Q factors: 1 3 3 offsets: 0,00E+00 0,00E+00 0,00E+00

0 0 2 4
m:QT

[s]

10

(f ile SimulaoCongresso.21_AumentoCarga.pl4; x-v ar t) m:PT

Fig. 9 Active power (red) and reactive power (green) of distributed generator with increased load.

Fig. 7 Active power (red) and reactive power (green) in the SE1 electrical system without DG and with islanding.

In the case of the simulation considering the increased load on the electrical system without the presence of the distributed generator, the values of the line voltage are shown in Table XIII.
TABLE XIII - Line Voltages - Increase System with Load and without DG
Bus SE1 CG2 CG3 CG4 CG5 CG6 CG7 CG8 Voltage [kV] 13,80 13,61 13,11 12,97 12,91 12,87 12,80 12,78 Voltage [pu] 1,000 0,986 0,950 0,940 0,936 0,933 0,927 0,926

C. Increasing Load An increase of load on the bus feeder distribution CG8 at time 2s was third phenomenon simulated. This new load is 10% of the initial loading feeder, or 673 kW active and reactive power 214 kVAR, which is also modeled as previously described (constant impedance). Table XII shows values of the line voltage for the load increase and the presence of the generator.
TABLE XII- Voltages Line - System with Increase Load and with DG
Bus SE1 CG2 CG3 CG4 CG5 CG6 CG7 CG8 Voltage [kV] 13,80 13,75 13,67 13,67 13,69 13,72 13,80 13,78 Voltage [pu] 1,000 0,996 0,990 0,990 0,992 0,994 1,000 0,999

The analysis of Table XIII shows that the line voltages were lower compared to those in Table XII. The largest voltage drop was 7.4% in bus CG8. The active and reactive power supplied by the substation without connecting of distributed generator rose to 7.63 MW and 2.90 MVAr, respectively, as shown in Fig 10.
8 *10 6 7 6 5 4 3 2 1 0 0 2 4 6 8 [s] 10
(f ile SimulaoCongresso.21_AumentoCarga.pl4; x-v ar t) m:P m:Q factors: 1 3 3 offsets: 0,00E+00 0,00E+00 0,00E+00

The analysis of Table XII shows that even with the increased load, the voltage values remained within the limits considered appropriate by ANEEL [13]. The substation raised its active power to 5.87 MW and its reactive power to 1.28 MVAr. The distributed generator has not altered its active power (2.51 MW), but increased its reactive power to 4.19 MVAr, to ensure the nominal voltage on your bus. Figs 8 and 9 show potencies, respectively, of SE1 and DG.

Fig. 10 Active power (red) and reactive power (green) in the SE1 electrical system without DG and with increased load.

6.

Conclusion

References
[1] S. Filho, Anlise Regulatria das Condies de nterconexo da Gerao Distribuda: Requisitos para os Procedimentos de Distribuio, Dissertao (Mestre) Departamento de Engenharia Eltrica, Universidade Federal de Itajub, Itajub, 2005. E. S. Lora, J. Haddad, Gerao Distribuda - Aspectos Tecnolgicos, Ambientais e Institucionais , Rio de Janeiro: Intercincia, 2006. C. J. Mozina, Impact of Green Power Distributed Generation, IEEE Industry Applications Magazine, July/Aug 2010. F. A. M. Moura; J. R. Camacho; M. L. R. Chaves, G. C. Guimares. Grid Integration of Renewable Sources in the Distribution Network: An Analysis Through ATP-EMTP. IEEE PES Conference on Innovative Smart Grid Technologies (ISGT 2012), 2012, Washington. L. M. Peres, M. L. R. Chaves, G. C. Guimares, F. A. M. Moura, A Computational Contribution to Analyse the Connection of an Independent Power Producer to the Grid, International Conference on Renewable Energies and Power Quality and Renewable Energy and Power Quality Journal, v.11, p.506, 2013. Resoluo Normativa N 482, 17 de Abril de 2012, Agncia Nacional de Energia Eltrica, Abril de 2012. Procedimentos de Distribuio de Energia Eltrica no Sistema Eltrico Nacional (PRODIST): Mdulo 3 Acesso ao Sistema de Distribuio, Agncia Nacional de Energia Eltrica, Dezembro de 2012. G. C. Guimares, Manual do UFUFlow: Programa de Fluxo de Carga (LoadFlow). Mtodo de NewtonRaphson, Universidade Federal de Uberlndia, 2000. Catlogo, Alumnio - Condutores Nus, Nexans. G. C. Guimares, Computer methods for transient stability analysis of isolated power generation systems with special reference to prime mover and induction motor modeling, PhD Thesis, University of Aberdeen, Scotland, GB 1990. P. M. Anderson, A. A. Fouad, Power System Control and Stability, Second Edition. Iowa, IEEE Press, A John Wiley & Sons, Inc., Publication, 2003. L. M. Peres, Estudos do Desempenho Dinmico de Geradores Sncronos de Produtores Independentes em Sistemas de Gerao Distribuda via ATPDraw, Dissertao (Mestre) Faculdade de Engenharia Eltrica, Universidade Federal de Uberlndia, Uberlndia, 2013. Procedimentos de Distribuio de Energia Eltrica no Sistema Eltrico Nacional (PRODIST): Mdulo 8 Qualidade da Energia Eltrica, Agncia Nacional de Energia Eltrica, Fevereiro de 2012.

The study allowed to evaluate the impact of a distributed generator connected to a distribution feeder of 13.8 kV with respect to the occurrence of disturbances in the electrical system, such as short-circuit phase-to-ground, islanding of the system and increased load on the feeder. The simulations of short-circuit showed that, without the presence of the distributed generator, the phase voltages C on the bus CG6, CG7 and CG8 were eliminated. In addition, without the generator, the active power supplied by the substation was higher and the lower reactive power. In islanding, it was observed that the presence of the distributed generator was decisive in ensuring the supply of the loads islanded. The voltage values were close to those of nominal loads, since the total power thereof is lower than the nominal power of the distributed generator. The power supplied by the substation had the same value, either with or without the connection to the generator. The analysis of simulations, considering the increased load, the voltage on all busses were higher in the presence of the distributed generator. And the power supplied by the substation was lower for the same condition. This fact contributes to the possibility of postponement or reduction of reinforcement work on the electrical system. It is concluded, therefore, with distributed generator, the values of the short-circuit voltage are smaller, with islanding are equal and with load increase is greater. Also considering the generator connected to the system, the power supplied by the substation for the short-circuit has lower values for the active power, larger values for the reactive power, powers are equal with islanding and smaller with increasing load. This work performed the first studies concerned to evaluation of medium voltage distribution network with/without distributed generation, using a simple electrical system. This fact permits a better comprehension of the phenomena cause by the connection of a distributed generator to the distribution grid. Future works will involve a more complex power systems topology. Comparisons with results presented by other researchers will also be conducted in order to improve and deep the acquired knowledge about this subject.

[2] [3] [4]

[5]

[6] [7]

[8] [9] [10]

[11] [12]

[13]

Acknowledgement
The authors acknowledge the funding support of the Brazilian research agencies: CAPES, CNPq and FAPEMIG.