Conference PaperPDF Available

Modeling and simulation of a hybrid single-phase/three-phase system in modelica

Authors:

Abstract and Figures

Power system studies seldom consider the interaction between transmission and distribution systems. This sort of analysis, however, have been gaining importance due to the progressive growth of renewable energy penetration in the distribution networks. In this context, the current study combines a positive-sequence transmission system model with a three-phase distribution system model. The connection between both systems is attained by a hybrid three-phase to single-phase interface element. The system model is written in Modelica language, and simulated using OpenModelica. A test system is built o top of the IEEE14 test system, where two load buses are expanded into three-phase distribution systems. Results of studied system are validated against the power system simulator, Simulight. Results also renders the presently analyzed hybrid model very promising for complimenting modern power systems studies.
Content may be subject to copyright.
1
Modeling and Simulation of a Hybrid
Single-Phase/Three-Phase System in Modelica
Marcelo de C. Fernandes1, Jana´
ına G. de Oliveira1, Luigi Vanfretti2, Maxime Baudette3and Marcelo A. Tomim1
1Federal University of Juiz de Fora, Electrical Engineering Department, Juiz de Fora–MG, Brazil
2Rensselaer Polytechnic Institute, Department of Electrial, Computer, and Systems Engineering, Troy–NY, USA
3KTH–Royal Institute of Technology, Department of Electric Power and Energy Systems, Stockholm, Sweden
Email: marcelo.castro@engenharia.ufjf.br
Resumo—Power system studies seldom consider the interaction
between transmission and distribution systems. However, this sort
of analysis have become really important and necessary due to
the progressive growth of renewable energy penetration in the
distribution side of power system. In this context, the current
study uses a formulation which combines a transmission system
modeled in positive sequence component with a distribution
system in three phase modeling. The connection between both
systems is made by a hybrid three-phase single-phase element.
A model of IEEE14 buses connected to two three-phase buses
is built. System model is written in Modelica language using
software OpenModelica and in Simulight. Results of studied
system are presented, validating the hybrid single-phase/three-
phase formulation and the models developed in Modelica. The
hybrid model shows promising potential for studying modern
power systems.
Keywords– Modelica, Power system simulation, distribu-
tion, transmission.
I. INT ROD UC¸˜
AO
Nos ´
ultimos anos, uma importante transic¸˜
ao tem sido ob-
servada no sistema energ´
etico mundial. A preocupac¸˜
ao com
impacto ambiental proporcionada pelo uso constante de com-
bust´
ıveis f´
osseis, em conjunto com o interesse das nac¸ ˜
oes em
diversificar suas matrizes energ´
eticas impulsionam investimen-
tos em fontes alternativas de energia. Muitos governos vem
propondo diversas pol´
ıticas para incentivar essas fontes de
forma a garantir um futuro sustent´
avel. A mesma tendˆ
encia de
transic¸˜
ao ´
e observada no setor de gerac¸˜
ao de energia el´
etrica,
onde as fontes renov´
aveis tˆ
em ganhado destaque. Espera-se
que essa participac¸˜
ao continue crescendo, j´
a que a performance
das fontes renov´
aveis cresce, enquanto o custo diminui [1].
Muitas das vezes, as fontes renov´
aveis s˜
ao conectadas `
as
redes de baixa ou m´
edia tens˜
ao, na forma de gerac¸˜
ao dis-
tribu´
ıda. Esses sistemas, tradicionalmente, n˜
ao s˜
ao estudados
e simulados em conjunto com redes de alta tens˜
ao. Por´
em,
o aumento r´
apido de energia provinda de fontes distribu´
ıdas
gera a necessidade de se estudar esses sistemas conectados
[2]. Outros fatores, como a necessidade de considerar linhas
de transmiss˜
ao n˜
ao transpostas, endossam a necessidade de
possuir modelos e ferramentas h´
ıbridas que possam realizar a
integrac¸˜
ao entre distribuic¸˜
ao e transmiss˜
ao [2].
Nesse sentido, diversas ferramentas, t´
ecnicas e abordagens
vˆ
em sido propostas na literatura [3]–[7]. Com o intuito de
resolver o problema da integrac¸˜
ao Distribuic¸˜
ao-Transmiss˜
ao,
t´
ecnicas de simulac¸˜
ao h´
ıbrida vˆ
em ganhando destaque [8]–
[10]. Na essˆ
encia, elas consistem na modelagem do sistema
de transmiss˜
ao em componentes sim´
etricas, do sistema de
distribuic¸˜
ao em componentes de fase e na interface desses
sistemas sendo realizada por um elemento h´
ıbrido. Em [8],
[9] a modelagem em componentes sim´
etricas ainda pode ser
simplificada se as componentes de sequˆ
encia zero e negativa
possu´
ırem valores desprez´
ıveis. Assim, somente a sequˆ
encia
positiva ´
e modelada, e o sistema de transmiss˜
ao ´
e modelado
em seu monof´
asico equivalente.
Inserido neste contexto, o presente trabalho tem como
objetivo a modelagem e simulac¸˜
ao do sistema h´
ıbrido proposto
em [8] na linguagem Modelica. Desta forma, uma validac¸˜
ao
da formulac¸˜
ao h´
ıbrida pode ser estudada e apresentada. Al´
em
disso, os objetivos secund´
arios s˜
ao estabelecidos como sendo
a construc¸˜
ao e disponibilizac¸˜
ao de alguns modelos trif´
asicos
em linguagem Modelica, e a an´
alise de softwares baseados em
Modelica para estudos de sistemas de potˆ
encia.
O trabalho apresenta uma breve introduc¸˜
ao `
a linguagem e
seus benef´
ıcios na Sec¸˜
ao II e a modelagem de componentes
trif´
asicos e h´
ıbridos na Sec¸˜
ao III. Na Sec¸˜
ao IV o sistema estu-
dado ´
e descrito e na Sec¸˜
ao V os resultados s˜
ao apresentados.
Conclus˜
oes s˜
ao desenhadas na sec¸˜
ao VI.
II. LI NG UAGEM MODELICA
Sistemas El´
etricos de Potˆ
encia (SEPs) s˜
ao complexos e
possuem uma vasta gama de fenˆ
omenos dinˆ
amicos a eles
associados. Devido a isso, diversas abordagens para a mo-
delagem e simulac¸˜
ao desses sistemas foram propostas. O
grande problema ´
e que modelagem dinˆ
amica ´
e, muitas das
vezes, inconsistente entre as plataformas usadas, devido a
simplificac¸ ˜
oes e suposic¸ ˜
oes [11]. Como exemplo, o emprego
da convencional modelagem em diagramas de bloco pode
resultar em dois modelos incosistentes entre si [11]. Esse
problema pode ser resolvido com uma linguagem que permita
a modelagem n˜
ao amb´
ıgua de fenˆ
omenos dinˆ
amicos.
Sistemas s˜
ao descritos utilizando um conjunto de equac¸ ˜
oes
diferenciais e alg´
ebricas na linguagem Modelica, o que permite
grande flexibilidade na modelagem [12]. Al´
em disso, modelos
n˜
ao-causais podem ser criados, ou seja, modelos nos quais
a relac¸˜
ao entrada/sa´
ıda n˜
ao ´
e definida explicitamente. Essa
caracter´
ıstica atribui grande flexibilidade aos modelos desen-
volvidos nesta linguagem [12].
2
Outra importante raz˜
ao que motiva a modelagem de elemen-
tos de SEPs na linguagem Modelica ´
e a possibilidade de inter-
face com outras poderosas ferramentas, como Matlab/Simulink
e Mathematica, utilizando os padr˜
oes FMI (Functional Mock-
up Interface) [13]. O FMI pode ser utilizado como uma
interface entre dois softwares, permitindo a co-simulac¸˜
ao de
sistemas complexos.
Por fim, estudos mostram que a linguagem Modelica ´
e
vi´
avel para simulac¸ ˜
oes de sistemas el´
etricos de larga-escala
[14] e, por isso, uma crescente utilizac¸˜
ao da linguagem no
desenvolvimento de bibliotecas para simulac¸ ˜
oes de sistemas
de potˆ
encia pode ser observada. Como destaque podemos citar
a biblioteca Open Instance Power System Library (OpenIPSL)
para simulac¸ ˜
oes fasoriais no dom´
ınio do tempo [15]. Ela
apresenta a modelagem de diversos componentes el´
etricos e
n˜
ao-el´
etricos, necess´
arios para a simulac¸ ˜
ao coerente de SEPs,
e ser´
a utilizada neste trabalho.
III. MOD EL AGE M
Essa sec¸ ˜
ao descreve como os elementos trif´
asicos e h´
ıbridos
foram modelados. A modelagem descrita baseia-se nas re-
ferˆ
encias [9], [16], [17]. Nota-se que o equacionamento utiliza
o benef´
ıcio da n˜
ao-causalidade, evidenciando a vantagem de
se utilizar a linguagem Modelica.
A. Elemento H´
ıbrido
Este elemento ´
e respons´
avel pela interface entre o sis-
tema modelado em sequˆ
encia positiva (transmiss˜
ao) e o sis-
tema modelado em componentes de fase (distribuic¸ ˜
ao). Sua
formulac¸ ˜
ao foi proposta em [8], [9] e consiste na utilizac¸ ˜
ao
de um elemento πpassivo, como uma linha de transmiss˜
ao
ou um transformador. A Figura 1 apresenta um diagrama
representando um modelo trif´
asico de um elemento π.
+
Vabc
k
Iabc
k
Yabc
ser
+
Vabc
m
Iabc
m
Yabc
shtkYabc
shtm
Figura 1: Representa c˜
ao trif´
asica de um elemento π.
O equacionamento do diagrama apresentado na Figura 1
pode ser representado pela equac¸ ˜
ao matricial (1) abaixo.
Iabc
k
Iabc
m=Yabc
ser +Yabc
shtk
Yabc
shtk
Yabc
ser Yabc
ser +Yabc
shtmVabc
k
Vabc
m(1)
Realizando a transformac¸ ˜
ao das matrizes Iabc
keVabc
kpara
as componentes de sequˆ
encia e admitindo que as correntes
e tens˜
oes de componentes de sequˆ
encia negativa e zero s˜
ao
desprez´
ıveis, ´
e poss´
ıvel escrever a equac¸ ˜
ao (2).
Vabc
k=T1·v+
k=
1
α2
α
v+
k
i+
k=T2·Iabc
k=1
3h1α α2iIabc
k
(2)
onde α=ej2π
3.
Al´
em disso, ´
e conveniente dizer que:
A1×1=T2·Yabc
ser +Yabc
shtk·T1
B1×3=T2·Yabc
ser
C3×1=Yabc
ser ·T1
D3×3=Yabc
ser +Yabc
shtm
(3)
Finalmente, ´
e poss´
ıvel substituir as equac¸ ˜
oes (2) e (3) na
equac¸ ˜
ao (1) e obter a equac¸ ˜
ao (4), que representa o elemento
h´
ıbrido de modelo aproximado proposto em [8]. Este estudo
utilizar´
a apenas a formulac¸ ˜
ao aproximada, j´
a que o sistema
estudado n˜
ao apresenta os dados de sequˆ
encia negativa e zero,
que s˜
ao fundamentais para a formulac¸ ˜
ao completa. O diagrama
do elemento h´
ıbrido πcom a formulac¸ ˜
ao aproximada est´
a
representado na Figura 2.
i+
k
Iabc
m=A B
C Dv+
k
Vabc
m(4)
+
v+
k
i+
k
Yabc
ser
+
Vabc
m
Iabc
m
Yabc
shtkYabc
shtm
Figura 2: Representac¸ ˜
ao de um elemento πh´
ıbrido.
B. Linha de Transmiss˜
ao
A linha de transmiss˜
ao tamb´
em ´
e modelada atrav´
es de
um elemento πequivalente como o apresentado na Figura
1. Por´
em, a matriz de admitˆ
ancia s´
erie entre os terminais
kem,Yabc
ser ,´
e separada em Gabc
ser +jBabc
ser . Al´
em disso, os
elementos shunt s˜
ao considerados apenas como tendo sua parte
imagin´
aria jBabc
ser . Por fim, os elementos das matrizes podem
ser entrados individualmente, permitindo a representac¸ ˜
ao de
linhas desbalanceadas e desequilibradas.
3
+
Vabc
k
Iabc
k
Gabc
ser +jBabc
ser
+
Vabc
m
Iabc
m
jBabc
serkjBabc
serm
Figura 3: Diagrama do modelo trif´
asico πequivalente para a
linha de transmiss˜
ao.
Desta forma, analisando o diagrama acima e considerando
(5), a equac¸ ˜
ao matricial para a linha de transmiss˜
ao pode ser
escrita como demonstrada em (6) abaixo.
LT =Gabc
ser +j(Babc
ser +Babc
sht )Gabc
ser jBabc
ser
Gabc
ser jBabc
ser Gabc
ser +j(Babc
ser +Babc
sht )
(5)
Iabc
k
Iabc
m=LT Vabc
k
Vabc
m(6)
C. Carga
As cargas foram equacionadas atrav´
es do modelo ZIP para
que uma abordagem mais gen´
erica pudesse ser obtida. Desta
forma, tanto modelos de potˆ
encia ativa e reativa constantes
ou modelos de impedˆ
ancia constante podem ser representados.
Modelos de potˆ
encia constante podem ser aplicados em um es-
tudo de fluxo de potˆ
encia, enquanto os modelos de impedˆ
ancia
constante podem ser aplicados em estudos de estabilidade
transit´
oria.
¯
ia
Pa+jQa
Pc+jQc
Pb+jQb
¯
ib
¯
ic
¯vn
¯va
¯vb
¯vb
Figura 4: Representac¸ ˜
ao da carga conectada em Y aterrado.
Consideremos, ent˜
ao, o diagrama da Figura 4 para uma
carga trif´
asica conectada em Y aterrado. Para uma carga desse
tipo, podemos escrever a equac¸ ˜
ao (7), onde o sobrescrito ()
refere-se `
a operac¸ ˜
ao de conjugado complexo.
Sa=Pa+jQa= ¯van ·¯
i
a
Sb=Pb+jQb= ¯vbn ·¯
i
b
Sc=Pc+jQc= ¯vcn ·¯
i
c
(7)
No modelo ZIP as potˆ
encias ativas e reativas podem ser
escritas como func¸ ˜
oes da tens˜
ao terminal ¯vxn. Existem dois
modelos difundidos para representar essa func¸ ˜
ao: os modelos
exponencial e polinomial. Neste estudo, o modelo polinomial
ser´
a adotado. Assim, para a fase xpodemos escrever a equac¸ ˜
ao
(8) abaixo.
Pxvxn) = P0
xαp
x+αi
x1
¯vxn +αz
x1
¯vxn 2
Qxvxn) = Q0
xαp
x+αi
x1
¯vxn +αz
x1
¯vxn 2
αp
x+αi
x+αz
x= 1
(8)
Finalmente, podemos escrever a equac¸ ˜
ao final (9) para o
modelo ZIP de cargas trif´
asicas conectadas em Y aterrado.
(Pa+jQa)γa= ¯van ·¯
i
a
(Pb+jQb)γb= ¯vbn ·¯
i
b
(Pc+jQc)γc= ¯vcn cot¯
i
c
(9)
onde
γa=αp
a+αi
a1
¯van +αz
a1
¯van 2
γb=αp
b+αi
b1
¯vbn +αz
b1
¯vbn 2
γc=αp
c+αi
c1
¯vcn +αz
c1
¯vcn 2
(10)
IV. SISTEMA ESTUDADO
O sistema estudado consiste no tradicional IEEE 14 barras
[18] com algumas modificac¸ ˜
oes, e seu diagrama est´
a apre-
sentado na Figura 5. Como pode ser observado, ao inv´
es de
haver uma carga conectada `
a barra 11, um transformador com
modelagem h´
ıbrida foi conectado `
a ela. Este elemento ser´
a
respons´
avel ent˜
ao pela interface entre o sistema modelado em
sequˆ
encia positiva e o modelado de forma trif´
asica. A este
transformador est´
a conectada `
a barra trif´
asica 650 que, por
uma linha de transmiss˜
ao trif´
asica, est´
a conectada `
a barra
trif´
asica 632. Uma carga trif´
asica modelada em Y aterrado
est´
a conectada `
a barra 632.
O modelo do IEEE 14 barras ´
e constru´
ıdo utilizando a bibli-
oteca em modelica OpenIPSL [15] para simulac¸ ˜
oes fasoriais
no dom´
ınio do tempo. Os modelos h´
ıbridos e trif´
asicos utili-
zados est˜
ao descritos nesse artigo na Sec¸ ˜
ao III, Modelagem.
Os geradores e condensadores s´
ıncronos utilizados no es-
tudo possuem modelagem de ordem VI e tamb´
em s˜
ao parte
da biblioteca OpenIPSL. Os parˆ
ametros das m´
aquinas, assim
como os parˆ
ametros dos reguladores autom´
aticos de tens˜
ao
(AVRs) est ˜
ao presentes no trabalho [19].
Os dados do transformador modelado como elemento
h´
ıbrido de interface, da linha de transmiss˜
ao trif´
asica e
da carga conectada `
a barra 632 podem ser observados nas
equac¸ ˜
oes abaixo. ´
E necess´
ario lembrar que, para o estudo
dinˆ
amico, a carga ´
e considerada como de impedˆ
ancia cons-
tante, ou seja αz= 1 para todas as fases. Al´
em disso, o valor
de potˆ
encia ativa e reativa base deve ser medido sob uma
determinada tens˜
ao terminal. Os dados utlizados para a carga
4
Elemento
H´
ıbrido
Sistema
Trif´
asico
1
2
34
56
7
8
9
10
11
12
13
14
G1
G2
C1
C2
C3
650
632
Figura 5: Modified IEEE 14-bus test feeder diagram.
assim como os parˆ
ametros do elemento h´
ıbrido e da linha de
transmiss˜
ao s˜
ao dados nas Tabelas I e II e nas equac¸ ˜
oes (11),
(12), (13).
Tabela I
Parˆ
ametros da Carga Trif ´
asica.
Fase P[kW ]Q[kV Ar]Tens˜
ao terminal
A1318,05 677,6 0.9678 p.u.
B937,61 347,43 1.0245 p.u.
C1429,46 677,85 0.9559 p.u.
Tabela II
Parˆ
ametros do Transformador.
Descric¸ ˜
ao Valor
Resistˆ
encia 0,2p.u.
Reatˆ
ancia 1,6p.u.
Conex˜
ao (Alta-Baixa) Y
Gabc
ser =
0.1982 0.0841 0.046
0.0841 0.1735 0.0219
0.046 0.0219 0.1535
p.u. (11)
Babc
ser =
0.5712 0.2112 0.1579
0.2112 0.5413 0.1206
0.1579 0.1206 0.5106
p.u. (12)
Babc
shtk=Babc
shtm= 0 p.u. (13)
Para a validac¸ ˜
ao do estudo aqui apresentado, o mesmo sistema
´
e desenvolvido e simulado em dois ambientes: Simulight c
e OpenModelica [20]. O primeiro software j´
a possui a
formulac¸ ˜
ao h´
ıbrida e ´
e utilizado para a comparac¸ ˜
ao dos
resultados obtidos. O segundo ´
e um software livre em lingua-
gem Modelica e ´
e utilizado para a validac¸ ˜
ao da formulac¸ ˜
ao
h´
ıbrida. O fenˆ
omeno estudado ´
e um curto trif´
asico e todos os
parˆ
ametros para a execuc¸ ˜
ao da simulac¸ ˜
ao est˜
ao descritos na
Tabela III abaixo.
Tabela III
Parˆ
ametros do Sistema e de Simulac¸ ˜
ao.
Descric¸ ˜
ao Valor
Parˆ
ametros do Sistema
Potˆ
encia Base 100 MV A
Frequˆ
encia do Sistema 60 Hz
OpenModelica
Passo 0,001 s
Tolerˆ
ancia 104
Simulight
Passo 0,0001 s
Tolerˆ
ancia 104
Parˆ
ametros do Evento
Tempo Simulado 20 s
Localizac¸˜
ao do Curto Barra4
Impedˆ
ancia do Curto j0,6p.u.
Instante do Curto 12 s
Durac¸˜
ao do Curto 100 ms
V. RE SU LTADO S
Os sistema ´
e simulado em dois softwares, como mencionado
previamente, e as figuras apresentam a comparac¸ ˜
ao entre os
resultados obtidos com os dois softwares. As Figuras 6, 7
e 8 apresentam os resultados para a variac¸ ˜
ao da magnitude
tens˜
ao nas barras 1,6e11, que possuem a modelagem em
5
sequˆ
encia positiva. As barras 1e6s˜
ao barras que possuem
m´
aquinas s´
ıncronas a elas conectadas. A barra 11 ´
e a barra de
sequˆ
encia positiva conectada ao dispositivo h´
ıbrido. Percebe-se
que ambas as simulac¸ ˜
oes, no Simulight e no OpenModelica,
apresentam o mesmo comportamento e que as curvas est˜
ao
sobrepostas. Al´
em disso, nota-se que o resultado da simulac¸ ˜
ao
´
e apresentado somente para o intervalo de simulac¸ ˜
ao de 10 s
a20 s. Desta forma, as oscilac¸ ˜
oes orginadas pelo dist´
urbio em
12 spodem ser analisadas mais facilmente.
10 11 12 13 14 15 16 17 18 19 20
1.02
1.04
1.06
1.08
1.1
1.12
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
Figura 6: Maginitude de tens˜
ao da barra 1.
10 11 12 13 14 15 16 17 18 19 20
0.98
1
1.02
1.04
1.06
1.08
1.1
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
Figura 7: Maginitude de tens˜
ao da barra 6.
10 11 12 13 14 15 16 17 18 19 20
0.95
1
1.05
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
Figura 8: Maginitude de tens˜
ao da barra 11.
A Figura 9 apresenta as oscilac¸ ˜
oes iniciais de inicializac¸ ˜
ao
provindas de ambas as simulac¸ ˜
oes. Percebe-se que as
oscilac¸ ˜
oes s˜
ao diferentes entre os softwares. Esses valores
tˆ
em sua prov´
avel origem nas inicializac¸ ˜
oes dos reguladores
autom´
aticos de tens˜
ao. Uma alternativa para eliminar esse tipo
de oscilac¸ ˜
ao em Modelica seria a de inicializar a simulac¸ ˜
ao
cerca de 3santes para permitir que os solvers atinjam o regime
permanente [13].
0 1 2 3 4 5 6 7 8 9 10
1.05
1.06
1.06
1.06
1.06
1.06
1.07
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
Figura 9: Efeitos da inicializac¸ ˜
ao na magnitude da tens˜
ao da
barra 1.
J´
a as Figuras 10 e 11 apresentam os resultados para a
variac¸ ˜
ao da magnitude tens˜
ao nas barras 650 e632, que
possuem modelagem trif´
asica. Percebe-se, novamente, que
ambas as simulac¸ ˜
oes apresentam comportamento semelhante.
´
E poss´
ıvel notar que as tens˜
oes est˜
ao desbalanceadas, como
era de se esperar. Al´
em disso, essas barras s˜
ao conectadas
ap´
os o elemento de interface h´
ıbrida. Logo, ´
e poss´
ıvel associar
os resultados convergentes com uma modelagem correta do
componente h´
ıbrido.
10 11 12 13 14 15 16 17 18 19 20
0.9
0.95
1
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
(a) Magnitude de tens˜
ao da fase A.
10 11 12 13 14 15 16 17 18 19 20
0.95
1
1.05
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
(b) Magnitude de tens˜
ao da fase B.
10 11 12 13 14 15 16 17 18 19 20
0.92
0.94
0.96
0.98
1
1.02
1.04
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
(c) Magnitude de tens˜
ao da fase C.
Figura 10: Resultados para a barra trif´
asica 650.
6
10 11 12 13 14 15 16 17 18 19 20
0.88
0.9
0.92
0.94
0.96
0.98
1
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
(a) Magnitude de tens˜
ao da fase A.
10 11 12 13 14 15 16 17 18 19 20
0.95
1
1.05
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
(b) Magnitude de tens˜
ao da fase B.
10 11 12 13 14 15 16 17 18 19 20
0.86
0.88
0.9
0.92
0.94
0.96
0.98
Tempo [s]
Voltage [pu]
OpenModelica
Simulight
(c) Magnitude de tens˜
ao da fase C.
Figura 11: Resultados para a barra trif´
asica 632.
VI. CO NCLUS ˜
AO
O presente estudo desenvolveu modelos para uma bi-
blioteca em Modelica que pode ser utilizada para estudos
dinˆ
amicos que considerem sistemas h´
ıbridos Monof´
asico equi-
valente/Trif´
asico. Esses modelos est˜
ao presentes na biblioteca
OpenIPSL, localizados em ApplicationExamples/ThreePhase.
A linguagem Modelica ´
e demonstrada como sendo adequada
para a modelagem do sistema trif´
asico al´
em de ser apresentada
como uma excelente alternativa para a modelagem de sistemas
el´
etricos de potˆ
encia.
Um elemento h´
ıbrido que realiza a interface entre um sis-
tema em sequˆ
encia positiva e um sistema trif´
asico ´
e modelado
e as equac¸ ˜
oes para o modelo aproximado s˜
ao apresentadas.
Os resultados da simulac¸ ˜
ao nos dois softwares corrobora que
a modelagem ´
e feita de forma correta e seu funcionamento
foi validado com um software alternativo `
aquele no qual
o elemento h´
ıbrido foi simulado no trabalho em que foi
proposto.
A modelagem h´
ıbrida monof´
asico/trif´
asico que foi estudada
neste trabalho ´
e vista pelos autores com uma adequada soluc¸ ˜
ao
para a simulac¸ ˜
ao integrada de sistemas de transmiss˜
ao e
distribuic¸ ˜
ao. A representac¸ ˜
ao de parte do sistema atrav´
es de
seu monof´
asico equivalente permite a reduc¸ ˜
ao de equac¸ ˜
oes
necess´
arias para as simulac¸ ˜
oes. Al´
em disso, o evento estudado
revela que a formulac¸ ˜
ao h´
ıbrida ´
e adequada para representac¸ ˜
ao
de fenˆ
omenos ocorrendo na parte do sistema modelada em seu
monof´
asico equivalente.
Trabalhos futuros incluem a simulac¸ ˜
ao de sistemas mais
complexos, que possam incluir modelos para a gerac¸ ˜
ao dis-
tribu´
ıda. Al´
em disso, fenˆ
omenos que ocorram na parte do sis-
tema modelada de forma trif´
asica. Desta forma, a ferramenta
da formulac¸ ˜
ao h´
ıbrida pode ter seu real potencial analisado e
explorado.
REF ER ˆ
ENCIAS
[1] International Eenergy Agency (IEA). World Energy Outlook 2017. 2017.
OECD/IEA. Paris, France. Retrieved in November 30, 2017.
[2] Jain, H.; Rahimi, K.; Tbaileh A.; Broadwater, R.P.; Jain, A.K.; Dilek, M.
Integrated Transmission & Distribution System Modeling and Analysis:
Need & Advantages. Published in: IEEE Power and Energy Society
General Meeting, 2016.
[3] Palmintier, B.; Hale, E.; Hodge, B.M.; Baker, K.; Hansen, T.M. Experien-
ces integrating transmission and distribution simulations for DERs with
Integrated Grid Modeling Systems (IGMS). Published in: Power Systems
Computation Conference (PSCC), Genoa, Italy, June, 2016.
[4] Li, Z.; Wang, J.; Sun, H.; Guo, Q. Transmission Contingency Analysis
Based on Integrated Transmission and Distribution Power Flow in Smart
Grids. Published in: IEEE Transactions on Power Systems, Volume: 30,
No. 6, Pages: 3356-3367, Nov. 2015.
[5] Li, Z.; Guo, Q.; Sun, H.; Wang, J. Coordinated Transmission and
Distribution AC Optimal Power Flow. Published in: IEEE Transactions
on Smart Grids, Volume: PP, No. 99, June 2016.
[6] Sun, H.; Guo, Q.; Zhang, B.; Guo, Y.; Li, Z.; Wang, J. Master-Slave-
Splitting Based Distributed Global Power Flow Method for Integrated
Transmission and Distribution Analysis. Published in: IEEE Transactions
on Smart Grids, Volume: 6, No. 3, Pages: 1484-1492, May 2015.
[7] Aristidou, P.; van Cutsem, T. Dynamic Simulations of Combined Trans-
mission and Distribution Systems Using Parallel Processing Techniques.
Published in: Power Systems Computation Conference, August 18-22,
Wroclaw, Poland, 2014.
[8] Jos´
e Mauro T. Marinho and Glauco Nery Taranto. A Hybrid Three-Phase
Single-Phase Power Flow Formulation Published in: IEEE Transactions
on Power Systems (Volume: 23, Pages: 1063:1070, Issue: 3, Aug. 2008)
[9] Jos´
e Mauro Teixeira Marinho. Simulac¸ ˜
ao em Sistemas de Energia El´
etrica
com Modelagem Flex´
ıvel - Monof´
asica e Trif´
asica.(Portuguese) Simu-
lation in Electrical Power Systems with Flexible Modeling — Single-
Phase and Three-Phase Rio de Janeiro, 2008, XVIm 221 p. 29.7cm
(COPPE/UFRJ, D.SC., Electrical Engineering, 2008).
[10] Huang, Q.; Vittal, V. Integrated Transmission and Distribution Sys-
tem Power Flow and Dynamic Simulation Using Mixed Three-
Sequence/Three-Phase Modeling. Published in: IEEE Transactions on
Power Systems, Volume: 32, No. 5, Pages: 3704-3714, Sep. 2017.
[11] Vanfretti, L.; Li, W.; Bogodorova, T.; Panciatici, P. Unambiguous Power
System Dynamic Modeling and Simulation using Modelica Tools. Pu-
blished in: Power and Energy Society General Meeting (PES), Vancouver,
Canada, July, 2013.
[12] Chieh, A. S.; Panciatici, P.; Picard, J. Power System Modeling in
Modelica for Time-Domain Simulation. Published in: IEEE Trondheim
PowerTech Conference, Trondheim, Norway, June, 2011.
[13] Bogodorova, T.; Sabat ´
e, M; Le´
on, G.; Vanfretti, L.; Halat, M.; Heyber-
ger, J.B.; Panciatici, P. A Modelica Power System Library for Modeling
Phasor Time-Domain Simulation. Published in: IEEE PES Innovative
Smart Grid Technologies Europe, Copenhagen, Denmark, October, 2013.
[14] Casella, F.; Bartolini, A.; Pasquini, S.; Bonuglia, L. Object-Oriented
Modelling and Simulation of Large-Scale Electrical Power Systems using
Modelica: a First Feasibility Study. Published in: Industrial and Electro-
nics Society 42nd Annual Conference, Florence, Italy, October, 2016.
[15] Vanfretti, L; Rabuzin, T.; Baudette, M.; MURAD, M. iTesla Power
System Library (iPSL): A Modelica library for phasor time-domain
simulations. Published in: SoftwareX, Volume 5, Pages: 84-88, 2016.
[16] Kersting, W. H. Distribution system modeling and analysis. CRC Press
LLC, 2002.
[17] Arrilarga, J.; Arnold, C. P.; Harker, B. J. Computer Modelling of
Electrical Power Systems. New York: Wiley, 1983.
[18] Power System Test Archive 14 Bus Power Flow Test Case. Retrie-
ved in November 1, 2017. Available in: http://www2.ee.washington.edu/
research/pstca/pf14/pg tca14bus.htm.
[19] Kodis, S. M. K.; Ca˜
nizares, A. C. Modeling and Simulation of IEEE 14
Bus System with Facts Controllers. Technical report, 2003.
[20] OPENMODELICA c
http://www.openmodelica.org/.
... and new features are being added. For example, its latest version comes with an application ex-ample package made for modeling positive sequence systems with three-phase networks using the hybrid interface proposed in (Marinho and Taranto, 2008), for which some results are reported in (de Castro Fernandes et al., 2018) (in Portuguese). ...
... Models for all machines are also available in the OpenIPSL library. The ThreePhase package was recently added by the authors de Castro Fernandes et al., 2018) and thus, it is useful to present the mathematical modeling of such elements, along with their implementation in the Modelica language. ...
Conference Paper
Full-text available
This paper present mathematical modeling and implemen- tation in Modelica language of a coalesced electric power transmission and distribution system model. To this end, a newly developed feature in OpenIPSL that allows to amalgamate power transmission and dis- tribution networks at the equation level is described and implemented in Modelica. Mathematical models for three-phase components such as lines and loads are also presented and implemented in Modelica. The models are used to assemble two different, small-scale, sample power systems, and three simulations are performed for each of them in a Modelica-compliant software. Dynamic simulations are carried out to perform comparisons between different modeling approaches for a distri- bution feeder and among different load characteristics. Each simulation is linearized using a script in ten different time instants and an eigen- value comparison is performed. The results from all dynamic simulations presented expected results. Voltage behavior observed using positive se- quence models and three-phase model with balanced load increase are similar. In addition, results coming from three-phase unbalanced load increase and positive sequence model diverge. The comparison between eigenvalues values corroborates the dynamic simulation results and show that a misused positive-sequence model may lead to wrong conclusions about system stability. The results are meaningful as distribution net- works gain relevance due to increasing level of distributed generation being introduced in low-voltage grids. The analyses performed in this paper are easily conducted, showing that Modelica language and com- pliant software packages may have key role in the development of new computational tools to study complex emerging power systems.
... In particular, the IEEE 14-bus system is modeled with the Dynaωo library, and its different components are presented. The IEEE 14-bus system is successfully implemented in the literature using other Modelica libraries like in (Adib Murad, Gómez, and Vanfretti 2015) and (Fernandes et al. 2018), and results show that it can be simulated in Modelica simulation environments. Also, several open-source electrical power system libraries are available in Modelica language ((Winkler 2017), (Bartolini, Casella, Guironnet, et al. 2019)). ...
Conference Paper
Full-text available
Power systems modeling and simulation are essential to conduct studies on the electrical transmission system and ensure its security. For this purpose, RTE, the French Transmission System Operator (TSO), is developing Dynaωo, a hybrid Modelica/C++ open-source suite of simulation tools for power systems. Most power systems models for Dynaωo are developed in the Modelica language using the Dynaωo Modelica library. This paper presents a full Modelica standard electrical power system benchmark implemented using the Dynaωo library. The IEEE 14-bus system benchmark is modeled here for steady-state calculation, with an approach that replaces the static load flow. Two test cases are simulated using the OpenModelica environment showing differences in the final steady-state result. We show flexibility in modeling with this library where different system behaviors can be observed and where models with different levels of details can be replaced depending on the application: steady-state calculation, long-term stability, or short-term stability.
... It contains the MonoTri transformer, a set of three-phase component models and some example models. The implementation was successfully validated against the original implementation in Simulight [6], using the IEEE four-node test feeder [7] and the IEEE 14-bus model, and the results have been submitted for publication [8]. ...
Article
Full-text available
This paper presents the latest improvements implemented in the Open-Instance Power System Library (OpenIPSL). The OpenIPSL is a fork from the original iTesla Power Systems Library (iPSL) by some of the original developers of the iPSL. This fork’s motivation comes from the will of the authors to further develop the library with additional features tailored to research and teaching purposes. The enhancements include improvements to existing models, the addition of a new package of three phase models, and the implementation of automated tests through continuous integration.
Article
Full-text available
[[Open-source codes on Github: https://github.com/InterPSS-Project/ipss-plugin/tree/master/ipss.plugin.multiNet]] The interactions between distribution and transmission systems have increased significantly in recent years. However, in traditional power system simulation tools, transmission and distribution systems are separately modeled and analyzed. Hence, it is difficult to analyze the impacts of distribution systems on transmission systems and their interactions in detail. To facilitate the analysis of integrated transmission and distribution (T&D) systems, a novel modeling framework is proposed in this paper, where the transmission system is modeled as one subsystem in three-sequence detail, while each distribution system connected to it is represented as a subsystem and modeled in three-phase detail. With this modeling approach, unbalanced conditions at the boundary between T&D systems and within the transmission system can be represented. The integrated T&D power flow is solved by iteratively solving a three-sequence power flow for the transmission system and a three-phase power flow for each distribution system. Furthermore, a new dynamic simulation algorithm for integrated T&D system is proposed. The main challenge in developing this integrated T&D dynamic simulation is associated with different network representations in the transmission and distribution systems. With a partitioned solution approach adopted for dynamic simulation, the Multi-Area Thévenin Equivalent (MATE) approach is utilized in the network solution step to address this challenge. The proposed algorithms have been tested against an all electromagnetic transient simulation.
Conference Paper
Full-text available
This paper discusses the development of, approaches for, experiences with, and some results from a large-scale, high-performance-computer-based (HPC-based) co-simulation of electric power transmission and distribution systems using the Integrated Grid Modeling System (IGMS). IGMS was developed at the National Renewable Energy Laboratory (NREL) as a novel Independent System Operator (ISO)-to-appliance scale electric power system modeling platform that combines off-the-shelf tools to simultaneously model 100s to 1000s of distribution systems in co-simulation with detailed ISO markets, transmission power flows, and AGC-level reserve deployment. Lessons learned from the co-simulation architecture development are shared, along with a case study that explores the reactive power impacts of PV inverter voltage support on the bulk power system.
Article
Full-text available
The iTesla Power Systems Library (iPSL) is a Modelica package providing a set of power system components for phasor time-domain modeling and simulation. The Modelica language provides a systematic approach to develop models using a formal mathematical description, that uniquely specifies the physical behavior of a component or the entire system. Furthermore, the standardized specification of the Modelica language (Modelica Association [1]) enables unambiguous model exchange by allowing any Modelica-compliant tool to utilize the models for simulation and their analyses without the need of specific mediator. As the Modelica language is being developed with open specifications, any tool that implements these requirements can be utilized. This gives users the freedom of choosing an Integrated Development Environment (IDE) of their choice. Furthermore, any integration solver can be implemented to simulate Modelica models. Additionally, Modelica is an object-oriented language, enabling code factorization and model re-use to improve the readability of a library by structuring it with object-oriented hierarchy. The developed library is released under an open source license to enable a wider distribution and let the user customize it to their specific needs.
Article
Full-text available
With the recent rapid development of smart grid technology, the distribution grids become more active, and the interaction between transmission and distribution grids becomes more significant. However, in traditional power flow calculations, transmission and distribution grids are separated, which is not suitable for such future smart grids. To achieve a global unified power flow solution to support an integrated analysis for both transmission and distribution grids, we propose a global power flow (GPF) method that considers transmission and distribution grids as a whole in this paper. We construct GPF equations, and develop a master–slave-splitting (MSS) iterative method with convergence guarantee to alleviate boundary mismatches between the transmission and distribution grids. In our method, the GPF problem is split into a transmission power flow and a number of distribution power flow sub-problems, which supports on-line geographically distributed computation. Each sub-problem can be solved using a different power flow algorithm to capture the different features of transmission and distribution grids. An equivalent method is proposed to improve the convergence of the MSS-based GPF calculation for distribution grids that include loops. Numerical simulations validate the effectiveness of the proposed method, in particular when the distribution grid has loops or distributed generators.
Article
Full-text available
In future smart grids, with distribution networks having loops more frequently, current transmission contingency analysis (TCA) which usually neglects the distribution power flow variations after a contingency may leave out severe outages. With more distribution management systems deployed on the distribution side, a new transmission CA method based on global power flow (GPF) analysis which integrates both the transmission and distribution power flow is proposed in this paper (named as GTCA) to address the problem. The definition and new features of GTCA are first introduced. Then, the necessity of GTCA is physically illustrated. Difference in the results of GTCA and TCA is mathematically analyzed. A GPF-embedded algorithm of performing GTCA is then provided. The data exchange process and the performance with communication interruptions are discussed. As multiple contingencies are considered in GTCA, several approaches are proposed and discussed to reduce communication burdens and improve the computational efficiency. Plenty of numerical tests are performed in several systems to verify the theoretical analysis. With theoretical analysis and numerical verification, it is suggested that GTCA should be performed instead of TCA to avoid potential false alarms, especially in the condition that DNs are more frequently looped in the future smart grids.
Conference Paper
Full-text available
Simulating a power system with both transmis-sion and distribution networks modeled in detail is a huge computational challenge. In this paper, we propose a Schur-complement-based domain decomposition algorithm to provide accurate, detailed dynamic simulations of the combined system. The simulation procedure is accelerated with the use of parallel programming techniques, taking advantage of the parallelization opportunities inherent in domain decomposition algorithms. The proposed algorithm is general, portable and scalable on inex-pensive, shared-memory, multi-core machines. A large-scale test system is used for its performance evaluation.
Article
The current separate transmission and distribution energy management system faces multiple challenges associated with integrating distributed generators (DGs) into future grids. These challenges, such as a voltage rise issue resulting in curtailment of DGs, are difficult to solve via the current separate energy management. Thus, coordination between transmission and distribution is suggested, and a coordinated transmission and distribution AC optimal power flow (TDOPF) is proposed in this paper. A mathematical TDOPF model is established and analyzed in a master-slave structure. A heterogeneous decomposition algorithm (HGD), which is inspired by heterogeneous transmission and distribution characteristics, is proposed to solve the TDOPF in a distributed manner. The HGD is compared to other typical multi-area OPF decomposition algorithms and the differences are discussed. Numerical tests verify the benefit of the TDOPF to both transmission and distribution systems. The TDOPF improves economic operations, mitigates voltage rises and decreases boundary bus mismatches. Hence, more DGs can be accommodated by the grid. In addition, a series of numerical tests indicate that the HGD competitively solves the TDOPF.
Article
The primary objective of this paper is to highlight the need for and benefits of studying the steady state and dynamic response of power systems using three phase integrated transmission and distribution (T&D) system models (hereafter referred to as hybrid models). Modeling and simulation requirements for building and analyzing hybrid models are also discussed. Finally, results from steady state and dynamic simulations of a large hybrid model are presented to demonstrate insights that can be obtained from hybrid models which cannot be obtained from the study of transmission and distribution systems separately.