newsletter - emtp
TRANSCRIPT
NewsletterVOLUME 1, NO. 6April 2009 EDITION
ElectroMagnetic Transients Program
(continued...)
INSIDE:
1010 Sherbrooke St. West, Suite 2500Montreal, Quebec, Canada H3A 2R7
www.ceati.com • [email protected]
Phone +1 (514) 904-5546Fax +1 (514) 904-5038
A Word from the Editor
Software News
SOFTWARE • Software News (1) TECHNICAL CORNER • New EMTP-RV Equivalent Circuit Model
of Core-Shielding Superconducting Fault Current Limiter Taking Into Account the Flux Di�usion Phenomenon
(2)
(13) •
SUPPORT • Training Success
(28)
The next version of EMTP-RV will be released this month. What’s new in version 2.2:1. Full compatibility with Vista.2. New documentation system including new navigation features.3. Various improvements and additions to models. The data handling features for numerous models have been simpli�ed to allow for easier loading with separately calculated data.4. New capability to store complete circuits in libraries. A circuit appearing in a library folder now becomes listed in the library Parts Palette and can be dragged and dropped into a design just like standard parts. This is a very powerful feature that provides easy access to user circuits and allows you to maintain more complex models through libraries.5. Subcircuits are now given the Model or Physical attribute in the Subcircuit Info menu. A model subcircuit is primarily intended to de�ne the operation of the device represented by its parent symbol. A physical subcircuit is primarily used to contain some of the system. The devices inside the subcircuit represent actual physical elements of the system. The physical subcircuit may contain Model subcircuits. This distinction allows propagating computed data into Physical subcircuits for visualization purposes.6. Several new scripting methods, including dynamic modi�ca- tion of device symbol using a separately stored symbol draw ing.7. New and improved ScopeView package.
A new HVDC benchmark (for 50 Hz and 60 Hz networks) originally developed by Professor Vijay Sood (University of Ontario Institute of Technology) is now available upon request. This work was prepared in collaboration with Sébastien Dennetière (Électricité de France) and École Polytechnique de Montréal.
Daniel KatsmanEMTP-RV Sales O�ce at CEATIEmail: [email protected].: 1-888 - 781-EMTPInternational Tel.: +1-514-904-5546
Hello and welcome to the May 2009 issue of the EMTP-RV Newsletter!
In this edition, we are very pleased to present a couple of note worthy articles from L’École Polytechnique de Montréal and a collaborative e�ort from Concordia University, University of Ontario Institute of Technology and IREQ . We would like to thank these organizations for contributing to the newsletter and sharing their experiences with other EMTP-RV users.Of course, the EMTP-RV newsletter would not be possible without the e�orts of our editorial board which reviews and selects the articles. Members of the board include: Alain Xémard (Electricité de France), Harish Sharma (Electric Power Research Institute), Anish Gaikwad (Electric Power Research Institute), Sébastien Dennetière (Electricité de France), Teresa Correia de Barros (Energias de Portugal) and Mario Paolone (University of Bologna).
We would also like to take this opportunity to share with you the success of our training seminar held in Montreal, Quebec (September, 2008), see back page
Mho Relay Model For Protection Of Series Compensated Transmission Lines
1
Technical Corner NEW EMTP-RV EQUIVALENT CIRCUIT MODEL OF CORE-SHIELDING SUPERCONDUCTING FAULT CURRENT LIMITER TAKING INTO ACCOUNT THE FLUX DIFFUSION PHENOMENON Authors: Mouhamadou Dione, Frédéric Sirois, Francesco Grilli
In order to successfully integrate superconducting fault current limiters (SFCL) into
electric power system networks, accurate and fast simulation models are needed. This led us to develop a generic electric circuit model of an inductive SFCL, which we implemented in the EMTP‐RV software. The selected SFCL is of shielded‐core type, i.e. a HTS hollow cylinder surrounds the central leg of a magnetic core, and is located inside a primary copper winding, generating an AC magnetic field proportional to the line current. The model accounts for the highly nonlinear flux diffusion phenomenon across the superconducting cylinder, governed by the Maxwell equations and the non‐linear E‐J relationship of HTS materials. The computational efficiency and simplicity of this model resides in a judicious 1‐D approximation of the geometry, together with the use of an equivalent electric circuit that reproduces accurately the actual magnetic behavior for the flux density (B) inside the walls of the HTS cylinder. The HTS properties are not restricted to the simple power law model, but instead, any resistivity function depending on J, B and T can be used and inserted directly in the model through a non‐linear resistance appearing in the equivalent circuit.
Introduction GIVEN the growing demand for electric power and the increased need for power system
interconnection, fault currents levels are more and more likely to exceed the short‐circuit rating of switchgear equipments and other power system components (bus bars, current transformers, etc.) [1]‐[3]. To reduce the risk of damages to these costly electrical equipments and associated system outages, fault current limiting devices are considered as serious candidates to be inserted into the grid. Among the technological possibilities are superconducting fault current limiters (SFCL), which present the advantage of very low losses in steady state operation, and high limiting impedance under fault conditions. SFCL for high voltage networks, together with medium and high voltage cables, are recognized as the two most promising applications of HTS materials in power systems in the short term [4].
Within the past 15 years, substantial work has successfully been done on the proof of concept of SFCL, at increasingly higher voltages [5]. The most recent projects within the DOE superconductivity program target 138 kV class limiters [4], and manufacturers get more and
2
more maturity with the combination HTS materials, cryogenic temperatures and high voltages. In order to further progress towards the integration of SFCL in power systems, it becomes important to develop circuit models of SFCL that could be integrated in power system transient analysis software, such as EMTP‐RV [6]. This will allow assessing off‐line the real impact of integrating SFCL in power systems (such as the protection coordination, selectivity of fault detection schemes, etc.), and determine the aspects requiring further work before SFCL can safely be integrated in power systems. Such models must be fast, which discards accurate but time consuming finite element models. In addition, these are difficult to couple directly with power system simulators.
In this paper, we propose a compromise between finite element models and circuit models. Indeed, in the particular case of the inductive SFCL, we show that it can be accurately modeled by a 1‐D partial differential equation that can be reproduced by an equivalent electric circuit directly within the power system analysis software. This has the advantage to consider the fine physical behavior of the device without having to couple two software programs. In addition, the modeled device can easily be scaled‐up in term of power simply by changing its physical parameters (dimensions, resistivity properties, etc.). Temperature and field dependences of the HTS material are also taken into account in the model.
Note that even if the particular inductive SFCL considered here (BSCCO‐based) is not currently the preferred solution, mostly for economical reasons, it is easy to implement at small scale in a laboratory, and will be used in the next stage of this project to develop a parameter identification methodology using real time power system simulators (with power hardware‐in‐the‐loop), in which both the virtual model and the physical device will be used side by side.
Shielded‐core inductive SFCL modelling General description The shielded‐core inductive SFCL has been well studied over the last 15 years [7]‐[20], so
it is a perfect topology to test the approach proposed below. It consists of a transformer with a primary winding made of copper and connected in series with the protected line, and a one‐turn secondary winding consisting of a bulk Bi‐2212 HTS cylinder. Fig. 1 presents its typical installation in a radial electric circuit (simplest case).
+
Vs
+
S
+
RL
+ XsRs
SFCL
Fig. 1 Simple installation of a SFCL in a radial circuit. The switch S is there only to simulate a short‐circuit of the load. The SFCL equivalent circuit is shown on Fig. 7
During steady state operation, the applied AC load current generates a time varying magnetic field shielded by the HTS cylinder. Under a fault condition in the network, the current increases enough to make the flux penetrate the iron core, which increases much the
3
impedance of the secondary since the primary winding passes from an air‐core inductor to a magnetic core reactor. This high series impedance inserted therefore limits the fault current.
SFCL electromagnetic model Finite element simulations show that the electromagnetic behavior of the SFCL
configuration shown in Fig. 3 can be obtained with very good accuracy using a simple 1‐D flux diffusion model, as long as 1) there is no air gap between the top and bottom parts of the windings and the HTS tube, 2) no saturation occurs in the magnetic core (even though saturation effects could be easily introduced without significant error if needed), and 3) the magnetic permeability of the core is sufficiently high ( 200 300≥ − ) in order to neglect the end effects and the flux return path outside the primary coil (considered as a magnetic short‐circuit). Under these assumptions, we end up solving a 1‐D axisymmetric partial differential equation inside the HTS cylinder, i.e.:
0
1 = ,z zB Brr r r t
ρμ
⎛ ⎞∂ ∂∂⎜ ⎟∂ ∂ ∂⎝ ⎠
(1)
which was obtained by combining Maxwell's equations = /E B t∇× −∂ ∂r r
and =H J∇×r r
together with the constitutive equations = ( , )E J B Jρr r r r
and 0=B Hμr r
.
Fig. 4 Alternate grid for B and J which forms the basis of the model used in this paper
Fig. 3 Hollow Bi‐2212 HTS cylinderFig. 2 Constructional scheme of shielded‐core SFCL
4
To solve the diffusion equation numerically, the HTS cylinder (Fig. 2) can be divided into k sections where each section represents a cylindrical shell of thickness = ( ) /ext intr r r kΔ − .
Considering an alternate grid on which B and J are constant on each elementary section rΔ (see Fig. 4), an electric analogy can be drawn (Fig. 5) if we suppose 1 1B I≡ − , 2 2B I≡ − .
Therefore, using a finite difference approximation for the space derivatives,
2 1 1 2
0 0 0 0
1= = = ,B B I IB IJr r r rμ μ μ μ
− −∂≈
∂ Δ Δ Δ (2)
and,
( )( )
( )1 1
1 1= = ,
2
i
i i i i
rEBrEEr rr r t r r+ +
Δ∂∂∇× − ≈
+∂ ∂ −
r (3)
in which 1 = / 2i ir r r+ − Δ and =1, 2i . Considering =V rE , (3) leads to the voltage drop
across inductance kiL :
( )1( ) = = ,
2 2i i ir r dBrrE V
dt++ Δ ⎛ ⎞Δ Δ −⎜ ⎟
⎝ ⎠ (4)
which is also given by the classical equation:
= = .k ki ii i
dI dBV L Ldt dt
Δ − (5)
Combining (4) and (5) yields:
1 21 = ,
2 2k r r rL + Δ⎛ ⎞⎛ ⎞
⎜ ⎟⎜ ⎟⎝ ⎠⎝ ⎠
(6)
and
2 32 = ,
2 2k r r rL + Δ⎛ ⎞⎛ ⎞
⎜ ⎟⎜ ⎟⎝ ⎠⎝ ⎠
(7)
where ( )1 = 1intr r k r+ − Δ , 21=2intr r k r⎛ ⎞+ − Δ⎜ ⎟
⎝ ⎠ and 3 = intr r k r+ Δ . Hence
13=4 2
kint
rL r k r⎡ ⎤ Δ⎛ ⎞+ − Δ⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦ (8)
Fig. 5 Electric analogy for the flux and current densities
5
21= .4 2
kint
rL r k r⎡ ⎤ Δ⎛ ⎞+ − Δ⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦ (9)
The resistance kR can easily be calculated since the current I is only defined in 2r ,
2 2 2 2= / = / = /R V I r E I r J Iρ , where = ( , , )J B Tρ ρ is the resistivity of the thk cylindrical shell
modeled by an elementary block. Thus
0
( , , ) 1= .2
kk
intJ B TR r k r
rρ
μ⎡ ⎤⎛ ⎞+ − Δ⎜ ⎟⎢ ⎥Δ ⎝ ⎠⎣ ⎦
(10)
Finally, the complete diffusion of the flux inside the HTS cylinder is reproduced with the
assembly of the k elementary blocks. The most interesting perhaps is the simplicity of integration of the resistivity function.
Parameters for ρ are calculated from currents 1I , 2I and I (corresponding to B and J ), as
shown in Fig. 5, except for the temperature, which must be computed in a thermal model coupled with the current electric analogy (not shown here). In order to complete the circuit analogy, it is necessary to add an additional element that accounts for the inside part of the HTS cylinder filled with the iron‐core. Therefore, an inductance 2= / 2int r intL rμ is used to model the
magnetic effect of the core, and a resistance coreR′ accounts for the iron losses. Finally, we use
the expression = ( )extB I tγ to relate the current I flowing in the primary winding with
magnetic field extB applied at the outside wall of the HTS cylinder. The proportionality constant
γ was determined from 2‐D finite element simulations. The entire model, based on the electric analogy, is given in Fig. 6, in which the applied field corresponds to a controlled current source, i.e. a extI B≡ .
Fig. 6 SFCL electromagnetic model based on electric analogy
6
HTS model The resistivity scρ used in our model was taken from an empirical model proposed in
[21] with adapted material parameters for the BSCCO‐2212 HTS tubes available in our lab [22]. In order to model the transition of the HTS resistivity from the superconducting to the normal state, an experimentally based value 6( ) = 3.5 10 (1 0.01( 77)) n T T mρ −× + − Ω , representing the
normal state resistivity was added in parallel with the usual power law model, i.e.
1
0( ) = .( , ) ( , )
n
c c
E JJJ B T J B T
ρ−
(11)
The resulting non‐linear resistivity scρ is given by the following equation,
( ) ( )( , ) = .
( ) ( , , )n
scn
T JJ TT J B Tρ ρρ
ρ ρ+ (12)
Coupling of SFCL model with electrical circuit The coupling of the electromagnetic SFCL model with the power system electric circuit is
achieved using a controlled voltage source. First, the conservation of power is applied. In fact, the instantaneous electric power ( sP ) consumed by the controlled voltage source of the SFCL
model corresponds to the instantaneous power delivered by the controlled current source aI in
the electric analogy ( aP ). It is then possible to determine the SFCL voltage based on
= = ( ) ( ) = ( ) ( )a s a aP P V t I t V t I t , where ( )aV t and ( )V t are respectively the voltage across the
controlled current source in Fig. 6 and the controlled voltage source of the SFCL model in Fig. 7, and ( )I t is the current flowing into the SFCL. It is important to note that the power equation naturally takes into account both the magnetic energy and the losses in the limiter. We can therefore calculate the voltage of the controlled source directly and without ambiguity using:
( ) ( )( ) = ( ) = ( ) = ( ).( ) ( ) /
a exta a a
ext
I t B tV t V t V t V tI t B t
γγ
(13)
Complete SFCL equivalent circuit Fig. 7 shows the complete SFCL equivalent circuit (the EMTP subcircuits are shown in the
enclosed attachment). The remaining parameters to be defined, i.e. the coil resistance ( coilR )
and the leakage inductance ( fL ), are either computed with standard formulas or measured
experimentally if a prototype is available.
Fig. 7 Equivalent electric circuit of our SFCL: the controlled voltage source accounts for all the magnetic energy and losses in the limiter
7
Thermal model A simple equivalent electric circuit model is used to simulate the temperature
dependence (unique temperature T for the whole HTS cylinder). This model accounts for the fact that the increase of the temperature is due to the power losses ( scP ) in the HTS cylinder
obtained by summing the losses of each elementary section of the model. The thermal equation solved is:
0= ( ),sc sc scdTP C V h T Tdt
+ − (14)
where scC , scV and 0T are respectively the thermal capacity, the volume of the cylinder
and the temperature of the refrigerant liquid (nitrogen in the case considered below), and h is an effective convection coefficient between the cylinder and the nitrogen bath.
Application example Dimensions of SFCL The main components of the shielded‐core type SFCL considered here are (see Fig. 3
and Fig. 2): • Bi‐2212 hollow cylinder: 50 mm (height), 14.4 mm (inner radius) and 2.8 mm (wall
thickness) • Primary copper winding: 50 mm (height), 17.2 mm (inner radius), 10.7 mm (winding
thickness) and 254 turns • Iron core: 50.5 mm (height), 27 mm (radius) and = 290rμ . These dimensions are based on a small prototype available at our lab for which we
measured = 0.1 coilR Ω , =1.6 fL mH and = 100 coreR Ω (with 2/core coreR R γ′ ≈ and = 0.0035γ ). Simulation results In order to evaluate the behavior of the developed SFCL model, short‐circuit simulations
were performed in EMTP‐RV, based on the circuit of Fig. 1. The supply voltage SV was set to
100 V peak, and = 0.38 SX Ω , = 0.13 SR Ω (i.e. = 4%Z and / = 3X R , which is typical of a
distribution system), = 10 LR Ω and the switch (S) was closed at time = 50t ms to create the
short‐circuit. The resulting current waveform is given in Fig. 8. We remark that as soon as one cycle after the application of the fault, the current is limited to 30 A peak (roughly ˆ3 nominalI ),
with a first peak of ˆ8 nominalI≈ ( 80≈ A). Note that the prospective fault current was 325 A peak
(symmetric). Therefore, the SFCL behaves as expected. Nevertheless, a comparison with measurements will be required to fully validate the model. The main challenge here is to find a good ( , , )J B Tρ model for BSCCO‐2212 over a wide range of J , B , and T values.
8
Conclusion In this project, a 1‐D partial differential equation representing the behavior of a
shielded‐core type SFCL has been reproduced by an equivalent electric circuit within EMTP‐RV. The model accounts for the nonlinear flux diffusion within the superconducting hollow cylinder. The model is simple to implement, and can easily be scaled up to larger physical dimensions and power rating simply by changing a few geometric parameters. It would be equally applicable to shielded‐core SFCL using YBCO thin film tubes, as the latter may have better thermal properties. Furthermore, the temperature and field dependences of the HTS material are directly taken into account, so there in no limitation in the resistivity model that can be used. In a near future, a small scale inductive SFCL will be tested in similar conditions for the purpose of validating our model.
Future works will consider the coupling of the physical devices and the model with a real time simulator to develop a new parameter identification methodology, and for investigating new protection schemes required for the integration of those innovative devices into the power system.
References [1] M. Steurer, M. Noe, and F. Breuer, ``Fault current limiters ‐ R & D status of two
selected projects and emerging utility integration issues'', in Proc. IEEE General Meeting, Denver, CO, June 2004, pp. 1423‐1425.
[2] H. Yamaguchi, K. Yoshikawa, ``Current limiting characteristics of transformer type superconducting fault current limiter'', IEEE Trans. Appl. Supercond., vol. 15, no. 2, June 2005, pp. 2106‐2109.
[3] “Fault Current Limiters – Utility Needs and perspectives”, EPRI, Palo Alto, CA:2004,
Fig. 8 Short circuit simulation result with ˆ ˆ3 = 30lim nominalI I A≈ peak (first peak 80≈ A). The
prospective fault current was 245 A peak
9
1008696. [4] DOE‐EPRI, ``Demonstration of a superconducting fault current limiter'', EPRI, July
2008, 1009035. [5] M. Noe et al, ``Conceptual design of a 110 kV resistive superconducting fault current
limiter using MCP‐BSCCO 2212 bulk material'', IEEE Trans. Appl. Supercond., June 2007, pp. 1784‐1787.
[6] J. Mahseredjian, S. Dennetière, L. Dubé, B. Khodabakhchian and L. Gérin‐Lajoie, ``On a new approach for the simulation of transients in power systems'', Electric Power Systems Research, vol. 77, issue 11, September 2007, pp. 1514‐1520.
[7] V. Sokolevsky, V. Meerovich, G. Grader, G. Shter, ``Experimental investigations of a current‐limiting device based on high‐Tc superconductors'', Physica C, vol.209, 1993, pp. 277‐280.
[8] L. S. Fleishman, Y. Bashirov, V. A. Aresteanu, Y. Brissette, J. R. Cave, ``Design considerations for an inductive high Tc superconducting fault current limiter'', IEEE Trans. Appl. Supercond., vol. 3, no. 1, March 1993, pp. 570‐573.
[9] W. Paul, Th. Baumann, J. Rhyner, ``Tests of 100kW high‐Tc superconducting fault current limiter'', IEEE Trans. Appl. Supercond., vol. 5, no.2, June 1995, pp. 1059‐1062.
[10] M. Ichiharu, M. Okazaki, ``A magnetic shielding type superconducting fault current limiter using a Bi2212 thick film cylinder'', IEEE Trans. Appl. Supercond., vol. 5, June 1995, pp. 1067‐1070.
[11] D. Ito, M. Urabe, T. Yasunaga, and N. Jindo, ``Applicability of high‐Tc superconducting bulk to superconducting fault current limiting device'', IEEE Trans. Magn., vol. 32, July 1996, pp. 2728‐2730.
[12] J. R. Cave, D.W.A. Willen, R. Nadi, W. Zhu, A. Paquette, R. Boivin, Y. Brissette, ``Testing and modelling of inductive superconducting fault current limiters'', IEEE Trans. Appl. Supercond., vol. 7, no. 2, June 1997, pp. 832‐835.
[13] W. Paul, M. Lakner, J. Rhyner, P. Unternahrer, Th. Baumann, M. Chen, L. Windenhorn, and A. Guerig, " Test of a 1.2 MVA high‐Tc superconducting fault current limiter'', in Inst. Phys. Conf. Ser., vol. 158, 1997, pp. 1173‐1178.
[14] V. Meerovich, V. Sokolovsky, J. Bock, S. Gauss, S. Goren, G. Jung, ``Performance of an inductive fault current limiter employing BSCCO superconducting cylinders'', IEEE Trans. Appl. Supercond., vol. 9, no. 4, Dec. 1999, pp.4666‐4676.
[15] E. M. Leung, ``Superconducting fault current limiters'', IEEE Power Engineering Review, August 2000, pp. 15‐18.
[16] G. Zhang, Z. Wang, M. Qiu, ``The improved magnetic shield type high‐Tc superconducting fault current limiter and the transient characteristic simulation'', IEEE Tans. Appl. Supercond., vol. 13, no. 2, June 2003.
[17] H. Yamaguchi, T. Kataoka, K. Yaguchi, S. Fujita, K. Yoshikawa, K. Kaiho, ``Characteristics analysis of transformer type superconducting fault current limiter'', IEEE Trans. Appl. Supercon., vol. 14, no. 2, June 2004, pp. 815‐818.
[18] J. Langston, M. Steurer, S. Woodruff, T. Baldwin, and J. Tang, ``A generic real‐time computer simulation model for superconducting fault current limiters and its application in system protection studies'', IEEE Trans. Appl. Supercond., vol. 15, no.2, June 2005, pp. 2090‐2093.
10
[19] S. Kozak, T. Janowski, G. Wojtasiewicz, J. Kozak, and B. A. Glowacki, ``Experimental and numerical analysis of electrothermal and mechanical phenomena in HTS tube of inductive SFCL'', IEEE Trans. Appl. Supercond., vol. 16, no. 2, June 2006, pp. 711‐714.
[20] H. Yamaguchi, T. Kataoka, ``Current limiting characteristics of transformer type superconducting fault current limiter with shunt impedance and inductive load'', IEEE Trans. Appl. Supercond., vol. 18, no. 2, June 2008, pp. 668‐671.
[21] J. Duron, F. Grilli, B. Dutoit, ``Modelling the E–J relation of high‐Tc superconductors in an arbitrary current range'', Physica C, vol. 401, 2004, pp. 231‐235.
[22] F. Sirois, J. Cave, Y. Basile‐Bellavance, ``Non‐linear magnetic diffusion in a Bi2212 hollow cylinder: measurements and numerical simulations'', IEEE Trans. Appl. Supercond., vol.17, no. 2, June 2007, pp. 3652‐3655.
11
Attachment: sub‐circuits of the presented SFCL model
HTS Cylinder Thermal model
p1
p2
+ C1
14.0374
+ R1
?v
#inv
_K#
v(t)p2
VM
+m
1?v
+
c I2 0/100
#T0#
DC1
SUM1
2
Fm6
POut
In
T
BlocTe
DEV1
POut In
T
BlocTe
DEV2
POut In
T
BlocTe
DEV3
POut
In
T
BlocTe
DEV4
POut In
T
BlocTe
DEV5
+
8.16M
?p
R2
+
#Lint#
L1
f(u)
12
3
45
?s
Fm1
i(t)p1
f(u) 1
Fm7
f(u)1
2
Fm9
SFCL
+1
2
Tfo_
idea
l
+ Rcoil
+ Lf
p1p2 +-
5Blocs
HTS_cylinder
Elementary block (Block Te)
i(t)
p1
+
I
YI Y
0
Rn2 f ( u) 1
Fm 4
f ( u) 1
Fm 5
+#L1n#
L3
+#L2n#
L4
f ( u)1
Fm 6
i(t) p3 i(t) p4
f ( u)1
2 ?s
Fm 16
p(t)
p6
f ( u)1
2
Fm 17f ( u)
1
2
3
Fm 18
P
OutIn
T
P
Out In
T
12
MHO RELAY MODEL FOR PROTECTION OF SERIES COMPENSATED TRANSMISSION LINES
A.B. Shah1, V.K. Sood2 and O.Saad3 1. Dept. of Electrical & Computer Engineering, Concordia University,
Montreal, Quebec, Canada. H3G 1M8. email: [email protected]
2. Faculty of Engineering and Applied Science, University of Ontario Institute of Technology, Oshawa, Ontario, Canada. L1H 7K4. email: [email protected]
3. IREQ, 1800 Montée Ste. Julie, Varennes, QC. J3X 1S1, email : [email protected]
Abstract ‐This paper presents the design of an EMTP‐RV based Mho relay model for the protection of two parallel 500 kV, 280 km long transmission lines. The transmission lines are 40% compensated with fixed series capacitors, installed at the remote end of the lines. An average value current compensation algorithm is used to compensate for the error in the impedance measurement and detect fault location under earth fault conditions. The algorithm detects faults by comparing phase angles between voltage and current signals, using four specially shaped characteristics (three forward zones and one reverse zone) and applying appropriate logic functions. Simulation results for improving the measuring accuracy of distance protection under various fault types and fault locations are presented. Keywords ‐Series‐compensated line, distance relay algorithm, transmission line protection, EMTP‐RV simulation.
1. NOMENCLATURE R0, R1 = Zero and positive sequence resistance respectively of the protected line, L0, L1 = Zero and positive sequence inductance respectively of the protected line, Ictp = Primary current of the current transformer (CT), Icts = Secondary current of the CT, Vcvtp = Primary voltage of the capacitor voltage transformer (CVT), Vcvts = Secondary voltage of the CVT, Zline = Impedance of the line, Zangle = Angle of the line impedance, tzone = Zone delay time, Icomp = Compensation current,
13
Ipn = Phase current, Iavg = Average value of input currents (Average current), kc = Conventional average compensation factor, kmag = Magnitude compensation, krad = Angle compensation, ZR1, ZR2, ZR1a, k1, k2, θ1, θ2, α1, α2 = Comparator design constants [8],
ZR11, ZR12 and ZR13 = Impedances of Zones 1, 2 and 3 respectively, Rf = Resistance in fault path.
2. INTRODUCTION
Utilities find it is cost effective to better utilize their existing transmission assets with
series compensation techniques to compensate for the inductive reactance of long transmission lines [1]. Adding series compensation is one of the simplest and cheapest ways of increasing transmission line capacity, power transfer capability, system stability, voltage regulation[2]‐[3] and lowering losses. However, installation of series capacitors and their over‐voltage protection system with nonlinear Metal‐Oxide Varistors (MOVs) etc., introduces certain difficulties for fault location and protective relaying reach, particularly when distance protection schemes are applied [4]. For this reason, it is necessary for the distance protection scheme to do the impedance measurement with sophisticated algorithms with the series capacitor bank in circuit.
Different algorithms and models have been put forward for the protection of series compensated transmission lines [5]‐[7]. In particular, the Mho relay has a circular characteristic with directionality, good phase selection and a simple criterion.
A transmission line demonstrates a predictable impedance, which increases with the length of the line. A distance relay has a pre‐established impedance setting, which determines the size of the relay's impedance characteristic, which is typically in the form of a circle in the impedance (R‐X) diagram and matched to the length of the line to be protected by the relay. The relay is capable of rapidly detecting faults on the transmission line, indicated by a drop in the measured impedance of the line. This means that the relay is capable of detecting faults when the impedance of the line is inside the impedance characteristic of the relay. The operation boundary of the Mho relay can be adjusted to provide consistent Zone coverage over the area of interest.
In this paper, an EMTP‐RV based Mho relay model and is used to evaluate the performance of a distance protection scheme applied to a 500 kV, two parallel lines, series‐compensated transmission network. Simulation results are presented for single and three phase‐to‐ground faults created at the beginning of the protected line and at the remote end, behind the capacitor of the network.
14
3. METHODOLOGY 1. Test System Model (Fig.1)
The 500 kV test system, modeled with EMTP‐RV [10], is comprised of two parallel lines L1 and L2. The two lines are paralleled at Buses A, B and C. Series compensation capacitors are located just ahead of Bus B. The series capacitors are protected by a parallel metal oxide varistor MOV, spark airgap and breaker. Line L1 of the power system is protected with the distance relay model placed at the beginning of the line, next to Bus A. The distance relay monitors the phase voltage and line current through a capacitor voltage transformer (CVT) and current transformer (CT) respectively.
15
2. The Relay Model (Fig.2)
The diagram of a conventional Mho relay model for a series compensated transmission line is shown. The relay has two 3‐phase inputs, three 1‐phase voltages from the CVT and the three line currents from the CTs, and provides one logical output which gives a Trip indication to the protection system. The Mho relay is comprised of 3 fundamental blocks: • Block A ‐ Fault Detection and Compensation Block, • Block B ‐ Zone Detection and Time Delay Block, and • Block C ‐ Logic Block. A. Fault Detection And Compensation Block A:
16
The fault detection and compensation block receives inputs from the CVT and CT and
derives as outputs either phase‐to‐phase or phase‐to‐ground voltages or currents. Block A has three sub‐blocks, as described below. Data Acquisition sub‐block: A band‐pass filter is used to remove harmonics from the three phase voltages and currents. If Ia, Ib and Ic are the input currents, then the average current is
derived as: Iavg = (1/3)*(Ia + Ib + Ic) (1)
Calculation sub‐block: Since a fault may or may not involve the ground connection, input voltages and currents, after being filtered, are converted into phase‐to‐phase and phase‐to‐neutral values by the calculation sub‐block. Since this sub‐block receives only phase‐to‐ground values, phase‐to‐phase values are obtained by subtracting two voltages or two currents i.e. Vab=Van‐Vbn.
Detection and compensation sub‐block: This sub‐block provides either phase‐to‐phase or phase‐to‐ground voltages and currents as outputs depending on the type of fault. The selection is carried out based on the current flowing through the circuit. During a fault condition, current in each phase varies depending on the type of fault.
The impedance seen by the relay is given by the ratio V/I (=Z). The impedance measured by the relay [8] is influenced by the fault type and also by a number of power system parameters such as MOV rating, series capacitance etc. Here, an algorithm called “average value current compensation” is employed where an average value of the input current is added to the phase currents to obtain the impedance measurement from the relay location to the fault location. The current seen by the relay for impedance measurement is given by: I=Ipn + Icomp (2)
Here, Icomp = kc*Iavg and kc = kmag with angle (krad)
kmag = 22
22
)1()1()10()10(
LRLLRR
+−+−
krad = ⎟⎠⎞
⎜⎝⎛−⎟
⎠⎞
⎜⎝⎛
−− −−
11tan
1010tan 11
RL
RRLL
B. Zone Detection And Time Delay Block B:
After computations on the inputs received from Block A, the output from this block provides data about the phase(s) and the Zone(s) where the fault has occurred. In total, six outputs are obtained from this block: one each for Zones 1, 2 and 3, and one each for phases a, b and c. Block B has three sub‐blocks, described below.
17
Zone detection sub‐block: A phase comparator compares the two input quantities and operates if the phase angle between them is less than or equal to 90° [9]. The 3‐phase input voltages and currents are fed through the phase comparators to detect the Zones. The output signals are based on each phase and Zone such as Zone1_a, Zone1_b, Zone2_a etc. For instance, if the fault occurs on phase a and in Zone 2, then Zone2_a gives the output signal for further processing and the other output signals provide a zero signal. During this process, the relay can detect the Zone where the fault has occurred.
During a fault condition, the voltage and current values will change the impedance seen. The fault Zone indication is based upon whether the impedance measured at the relay location is greater than or less than the protected line impedance. Zone 1 primary impedance magnitude = Length * Zline (5) where
Length = 0.85 of the protected line of 280 km
Zline = 22 )1()1( LR +
Zone 1 primary impedance angle, Zangle = tan‐1 ⎟
⎠⎞
⎜⎝⎛
11
RL
The setting value of each Zone is expressed as a percentage of the line length. Normally, the first Zone covers only up to 80 to 90% of the protected line length. The second Zone covers the remainder of the line left unprotected by the Zone 1setting, plus 50% of the adjacent line section. The third Zone is used for back‐up protection and covers the first and section line sections, plus 20 to 25% of the adjacent line. Faulted phase detection sub‐block: The 3‐phase input voltages and currents are fed to the phase comparators. Output of this block provides the sequence of phases a, b or c through the phase comparators and the impedance trajectory of each phase in the impedance (R‐X) diagram. Selected values for phase comparator design constant parameters k1, k2, θ1, θ2, α1, α2
are shown in the Appendix. Time delay and Zone representation sub‐block: This sub‐block incorporates two functions. The first is a Time Delay function. When signals are received from Zone detection and faulty phase detection, they pass through a logical OR function to determine the Zone where the fault has occurred. After the detection of the faulted Zone, Zone 1 relay trips instantaneously. However, Zone 2 and Zone 3 relays have some intentional time delays added to coordinate with the relays at the remote bus, before providing an output. Time delays may vary depending on the circumstances.
The second function in this sub‐block is the Zone Representation function which draws a distance relay characteristic on an impedance (R‐X) diagram. With internal mathematical calculations, this sub‐block decides the centers and radii of the circles on an R‐X diagram for different Zones according to the data chosen for the system. These circles pass through the
18
origin and have different radii for different Zones. Each circle denotes a particular length of the line. The diameter of the circle is proportional to the impedance of the line or indirectly the length of the line covered by the each Zone. For instance, if the length of the line is 280 km and if Zone 1=0.9 is selected, it means that circle 1 will cover 90% of the protected line length. When a fault occurs within that area, this can be located within Zone 1 circle in the R‐X diagram. C. Logic Circuit Block C:
The output of this block determines the final decision of the relay for tripping a circuit breaker (CB). If the fault is temporary and can be isolated within the reset time of the relay, then this block will not send a trip signal. However, if the fault is permanent, then it will send a trip signal for the CB. This block has two sub‐blocks. Logic sequence sub‐block: The Zones detection signal from Block B is passed through a logical OR function, and the output gives the final Zone decision and identifies where the fault has occurred. Now, as information about the Zone and the faulted phase(s) is available, a logical AND function provides an output based on the combination of faulted Zone with faulted phase. Reclosing sub‐block: A single phase auto‐reclosing scheme is employed, in which only the faulted phase pole of the CB is tripped and reclosed. At the same time, synchronizing power still flows through the healthy phases. For a multi‐phase fault, all the three‐phases are tripped and reclosed simultaneously [9]. When the Zone and faulted phase(s) are decided then it is necessary to determine whether the fault is temporary or permanent in nature before tripping the three phases CBs.
Whenever this block receives the information about faulted phase(s) and Zone where the fault occurs, the relay sends a trip signal for the faulted phase(s). The relay checks the status of the fault again after a reset time and depending upon the nature of the fault; thereafter, either the relay sends a trip signal for the three phase CBs or restores the line after the reset time.
19
4. RESULTS
Due to space restrictions, only a small sample of the tests carried out are presented next. Single phase‐to‐ground (a‐g) fault at F1 (Fig.3)
This case shows results from a permanent single phase‐to‐ground fault (a‐g) placed at 280 km from the relay, behind the capacitor (at location F1) with fault resistance Rf =10 ohms.
The fault occurs at time=0.06s, and the simulation is run for a total time period of 0.7s.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7−0.2
0
0.2
0.4
0.6
0.8
1
Operating time of the relay − Phase a−to−ground fault
Time (Second)
Tri
p Si
gnal
Phase aPhase bPhase c 0.5618s
0.5774s0.3816s
Figure 3(a)
Fig. 3(a) shows the trip signals for phases a, b and c. For phase “a” the trip signal is
generated after 0.3816s (including0.3s Zone 2 delay). The relay checks the status of the fault after reset time (0.18s) and due to the permanent fault; all three phase circuit beakers are tripped after 0.5174s.
20
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
−2
0
2pu
(1) Phase a current
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
−2
0
2
pu
(2) Phase b current
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
−2
0
2
Time (Second)
pu
(3) Phase c current
1 pu
1 pu
1 pu
0.06s 0.5618s
0.3816s
0.5774s
0.5774s
0.5774s
Figure 3(b)
21
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7−2
0
2pu
(1) Phase a voltage
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7−2
0
2
pu
(2) Phase b voltage
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7−2
0
2
Time (Second)
pu
(3) Phase c voltage
1 pu
1 pu
0.06s
0.06s
1 pu
0.06s
0.3816s
0.5774s
0.5774s
0.5774s
Figure 3(c)
Fig. 3(b) and (c) show the Line L1, 3‐phase current and voltage waveforms respectively measured at the relay location. When the fault occurs at 0.06s, the current for phase “a” increases and at the same time voltage decreases. The relay trips the phase “a” CB at 0.3816s, therefore, no current passes through the phase “a”. Due to a permanent fault, after reset time (0.18s), the relay trips the three phase CBs at 0.5174s and the protected line is completely disconnected from service.
22
−10 −5 0 5 10 15 20 25 30−20
−15
−10
−5
0
5
10
15
20
25
30
Resistance (R) in Ohms
Rea
ctan
ce (X
) in
Ohm
sImpedance diagram
zone 3
zone 1
zone 2
Phase bImpedanceTrajectory
Phase cImpedanceTrajectory
Reverse Zone
Phase a Impedance Trajectory
Figure 3(d)
The impedance (R‐X) diagram (Fig. 3(d)) shows the three circles covering Zones 1, 2 and 3 and another smaller circle covering Reverse Zone operation. The trajectories of the impedance detection for phases a, b and c are also shown. The trajectory of phase “a” indicates that the fault involved phase “a” and is covered by the Zone 2 circle.
23
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7−2
0
2pu
Voltage across capacitor for phase a
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7−4
−2
0
2
4
pu
Capacitor current for phase a
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7−4
−2
0
2
4
Time (Second)
pu
MOV current for phase a
1 pu
0.0677s
0.3808s
0.5627s
0.5753s0.5668s
0.5668s
0.06s
0.06s
1 pu
Figure 3(e)
Fig. 3(e) shows the capacitor voltage (top trace), capacitor current (middle trace) and the MOV current (bottom trace) for phase “a”. The results show that when the fault occurs at0.06s, the capacitor voltage and current increase. The voltage increase is enough to trigger the MOV to conduct also to protect the capacitor. The capacitor and the MOV take turns conducting currents.
Many results were obtained with the test system. The findings of some of these evaluations are summarized in Table 1. The performance of the relay operation and the algorithm scheme on single phase‐to‐ground (SLG), two phase‐to‐ground (2LG) and three phase‐to‐ground (3LG) permanent faults at two different locations F1 and F2 as shown in Figure 1. At F1 the fault is located at the remote end, behind the capacitor. At F2 the fault is located at the beginning of the protected line. The relay operation and the algorithm is tested with 75 kV MOV reference voltage and different fault resistances (0, 5, 10 and 20 ohms). Type of faults and fault resistance are listed in columns 1 and 2 respectively in Table 1. Columns 3 and 4 show the two different fault locations, which include the zone of operation, number of cycles and secure,
24
insecure or missing operation for each fault case. The data shown in the table indicates that the relay operates securely and correctly for all close‐in faults (F2). For all close‐in faults, the relay operates in Zone 1 and an average tripping time is less than 1 cycle or 16.7ms. The table also shows that the relay operates correctly and securely with 20ohms fault resistance in each fault case and at both locations. In summary, out of 56 permanent faults with different fault resistance, 44 secure operations (relay operates in the expected zone), 12 insecure operations (relay operates in other zone than expected) and 0 missing operations (relay fails to operate) were obtained.
5. CONCLUSIONS
This paper evaluates the performance of a Mho relay model and distance protection algorithm for a 500 kV, 40% series‐compensated transmission system with two lines. An EMTP RV based simulation model is used to test the Mho relay. A scheme based on the average value of current is used to compensate the error in the impedance measurement. The distance protection scheme is based on measuring phase angle of the input signals and comparing them through phase comparators and using four specially shaped characteristics.
The results show that the relay model detects the faults correctly and generates trip signals with regards to the location of the fault. The MOV protects the capacitor against over voltage during fault conditions. Furthermore, it is noted that the operating time of the relay is a function of the distance to the fault.
Finally, for close‐in faults, satisfactory relay performance was obtained and an average tripping time is less than 1cycle. However, the relay may not be as secure on certain unbalanced fault types generated at the remote end, behind the capacitor.
6. APPENDIX 1. System data: Rated voltage = 500 kV rms, Rated power = 1450 MW, Length of the protected line = 280 km, Ictp = 1074 A, Icts = 5A, Vcvtp = 410 kV, Vcvts = 115V,
R0 = 0.06162 Ω/km, L0 = 1.05 Ω/km, R1= 0.0205 Ω/km, L1= 0.35 Ω/km, Zone1 = 85% of the protected line, Zone2 = 1.5 and Zone 3 = 2.1, tzone1= 0.001s, tzone2 = 0.3s and tzone3 = 0.6s,
Reset time = 0.18s,
25
ZR11 = Zline*Zone1, k1=1, θ1 = Zangle, α1 = π, ZR12 = Zline*Zone2, k2 = 1, θ2 = 0, α2 = 0, ZR13 = Zline*Zone3, ZR1a = Zline/2, θ1a = Zangle + π, Fault resistance (Rf) = 0, 5, 10 and 20 ohms, MOV reference voltage (Vref) = 75 kV,
Series Capacitance = 67.66 μF/phase.
7. ACKNOWLEDGEMENTS
The authors acknowledge financial support from Natural Sciences and Engineering Research Council (NSERC) for this work and the contributions of Dr. V. Ramachandran of Concordia University.
8. REFERENCES
[1] Marc Coursol, Chinh T. Nguyen, Rene Lord and Xuan‐Dai Do, “Modeling MOV‐protected
series capacitors for short‐circuit studies,” IEEE Transactions on Power Delivery, Vol. 8, No. 1, pp. 448‐453, January 1993.
[2] Power Systems Relaying Committee (PSRC) of the IEEE Power Engineering Society: “IEEE
Guide for Protective Relay Applications to Transmission Lines,” IEEE Std. C37.113‐1999. [3] Belur S. Ashokkumar, K. Parthasarathy, F.S. Prabhakara and H. P. Khincha, “Effectiveness
of Series Capacitors in Long Distance Transmission Lines,” IEEE Trans. on Power Apparatus & Systems, Vol. PAS‐89, No. 5/6, May/June 1970.
[4] M.M.Saha, K. Wikstrom, J. Izykowski and E. Rosolowski, “Fault Location in
Uncompensated and Series‐Compensated Parallel Lines,” 2000 IEEE Power Engineering Society Winter Meeting. Conference Proceeding (Cat No.00CH37077), 2000, p 2431‐2436, Vol.4.
[5] F. Ghassemi and A.T. Johns, “Investigation of alternative residual current compensation
for improving series compensated line distance protection,” IEEE Transactions on Power Delivery, Vol. 5, No. 2, pp. 567‐574, April 1990.
[6] M.M. Saha, E. Rosolowki and J. Izykowski, “ATP‐EMTP Investigation of a New Distance
Protection Principle for Series Compensated Lines,” International Conference on Power Systems Transients ‐ IPST 2003 in New Orleans, USA.
[7] Y. Heo, C.H. Kim, K.H. So and N.O. Park, “Realization of Distance Relay Algorithm using
EMTP MODELS,” International Conference on Power Systems Transients ‐ IPST 2003 in New Orleans, USA.
26
[8] V. Cook, “Analysis of Distance Protection,” Research Studies Press, Wiley (Letchworth,
Hertfordshire, England, UK), January 1985, pages 185. [9] Badri Ram and D N Vishwakarma, “Power System Protection and Switchgear,” Tata
McGraw Hill Publishing Company Limited, New Delhi, 1995, pages 456. [10] J. Mahseredjian, S. Dennetière, L. Dubé, B. Khodabakhchian and L. Gérin‐Lajoie: “On a
new approach for the simulation of transients in power systems”. Electric Power Systems Research, Volume 77, Issue 11, September 2007, pp. 1514‐1520.
TABLE 1. Analysis of the relay operation for permanent fault, MOV Vref = 75 kV and Rf = 0 to 20 ohms
Location F1 Location F2
Fault type
Rf
ohms Zone No. of
cycle
Secure
operation
Unsecure
operation
Missing
operation
Zone No. of
cycle
Secure
operation
Unsecure
operation
Missing
operation
SLG 0 2 20 3 0 0 1 1 3 0 0
5 2 19 3 0 0 1 1 3 0 0
10 2 19 3 0 0 1 1 3 0 0
20 2 19(1/2) 3 0 0 1 1 3 0 0
2LG 0 1 13 0 3 0 1 1 3 0 0
5 1 2 0 3 0 1 1 3 0 0
10 1 1(1/2) 0 3 0 1 1 3 0 0
20 2 19 3 0 0 1 1 3 0 0
3LG 0 1 1(1/2) 0 1 0 1 1 1 0 0
5 1 1/2 0 1 0 1 1 1 0 0
10 1 4(1/2) 0 1 0 1 1 1 0 0
20 2 19 1 0 0 1 1 1 0 0
Total 16 12 0 28 0 0
27
VOLUME 1, NO. 6May 2009 EDITION
ElectroMagnetic Transients Program
1010 Sherbrooke St. West, Suite 2500Montreal, Quebec, Canada H3A 2R7
www.ceati.com • [email protected] +1 (514) 904-5546Fax +1 (514) 904-5038
Support
TRAINING SUCCESS!
CEATI is pleased to report the success of the EMTP-RV training seminar held in Montreal, Quebec, in Septem-ber 2008. This event brought together 18 professional engineers and students from around the globe.
We would like to take this opportunity to thank all those who contributed to the event, in particular the lecturers, Dr. Jean Mahseredjian, Luc Gérin-Lajoie, Doug Mader, Luis-Daniel Bellomo, and to extend our congratulations to all those who participated, for their successful completion of the course.
Also note that a three day training course and a user’s group meeting took place in Dubrovnik, Croatia, from April 27-30, 2009. The course covered theoretical backgrounds to the simulation transients, equipment modeling and applications, insulation coordination issues and practical Power System Studies.
The user group meeting discussed the latest developments in EMTP-RV, as well as end-user simulation studies, exchanges around simulation topics, evolution of the software and the future of EMTP-RV.
28