1410 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 18, NO.

4, OCTOBER 2003

High-Frequency Impacts in a Converter-Based

Back-to-Back Tie; The Eagle Pass Installation
Lars Paulsson, Member, IEEE, Björn Ekehov, Stefan Halén, Tomas Larsson, Member, IEEE, Lars Palmqvist,
Abdel-Aty Edris, Senior Member, IEEE, David Kidd, Member, IEEE, Albert J. F. Keri, Senior Member, IEEE,
and Ben Mehraban, Senior Member, IEEE

Abstract—Voltage sourced converter-based transmission con- very high-frequency harmonics. They are a result of the very
trollers, which use gate turn off switching components, introduce steep voltages, which are generated due to the PWM switching.
a new type of high voltage stress on insulation systems and mate- Very high-frequency harmonics have generally low energy
rials. Fast switching components like IGBTs may generate or cause
high-frequency voltages, which are superimposed upon the power and low voltage amplitude. However, they may excite local
frequency (50/60-Hz) voltage. Insulation within the station may resonances and cause harmonics with high amplitudes, which
thereby be continuously exposed to a new type of voltage stress, may stress some insulation systems excessively and eventually
for which it has not been designed or tested. The high-frequency cause failures.
voltages may, under special circumstances, cause insulation dam-
ages at magnitude levels, which are well below the dielectric voltage
withstand at power frequency. Compact insulation systems with II. EAGLE PASS VSC BTB
resistive/refractive grading of the electrical field are particularly
critical. The Eagle Pass BtB is located in Texas near the Mexican
Several compact type cable terminations rated 24 kV with border and connects the U.S. AEP Texas Central Company
resistive/refractive stressgrading have failed in the voltage soured (AEP-TCC) transmission grid with the Mexican Comisión
converter-based back-to-back installation at Eagle Pass substation Federal de Electricidad (CFE) transmission grid [2].
in the State of Texas. Through field measurements, laboratory The BtB scheme consists of two separate 36-MVA VSCs con-
tests, and simulations, the root cause of the failures has been
identified. The compact cable terminations have been replaced nected to a common dc capacitor link. Each VSC is, via phase
by another type of termination, which is not sensitive to high-fre- reactors and step up transformers, connected to the AEP-TCC
quency voltages. and CFE 138-kV grids. Due to restrictions in space, the BtB sta-
The voltage sourced converter-based back-to-back installation tion is built very compact. The phase reactors and the harmonic
in Eagle Pass is a joint ABB, EPRI, and AEP field demonstration filters are located outdoor and connected via cables to the indoor
project, which has been recently commissioned.
switchgear.
Index Terms—BtB, cable, cable termination, EPDM, IGBT, The Eagle Pass BtB is designed to provide independent con-
insulation, partial discharges, PWM, resistive grading, stress-
trol of the voltages at the U.S. and Mexican grid sides (voltage
grading, voltage sourced converter, VSC, XLPE.
control mode) and to provide control of the real power trans-
ferred between the two grids (real power control mode). The
I. INTRODUCTION rating of the medium voltage ac sides of the BtB is 17.9 kV. The
dc side is ungrounded, while the ac sides are high impedance
T HE voltage sourced converter (VSC) technology is based
on the use of semiconductor switching components, which
enable control of both turn-on and turn-off gating. A VSC in-
resistance grounded via 50-M voltage dividers, used for mea-
surement purposes.
All equipment on the 17.9-kV ac sides is rated 24 kV or higher
verts a dc voltage into an ac voltage with controllable magni-
and in accordance with standards for ac applications. There are
tude, phase angle, and frequency. At Eagle Pass back-to-back
42 short cables of lengths ranging from 4 to 32 m. The main
(BtB), the inversion is achieved using pulse width modulation
phase cables consist of two cables in parallel and with cable in-
(PWM) switching technique of IGBT-based valves [1].
sulation of XLPE (cross-linked Polyethylene) type. The filter
On and off switching creates harmonic distortions with fre-
and charging resistor cables are of ethylene propylene rubber
quencies of an order in proportion to the switching frequency.
(EPR) type. Some of the cable ends and cable terminations are
For the Eagle Pass BtB, the first group of significant harmonics
located indoors in the 24-kV switchgear cubicles. There are al-
appears around the 21st harmonic, which is the converter
together 84 cable terminations of which 48 are the indoor type
switching frequency. However, in addition to the significant
and 36 the outdoor type. All cable terminations were initially of
main harmonics, there is a generation of a broad spectrum of
the resistive/refractive stressgrading type, which is the common
type being used at ac voltage levels up to 36 kV.
Manuscript received April 23, 2003. Surge arresters connected to the low voltage side of the
L. Paulsson, B. Ekehov, S. Halén, and T. Larsson are with the ABB, Västerås
S-721 64, Sweden.
transformers are protecting the ac system against overvoltage
L. Palmqvist is with the ABB, Sweden. transients.
A.-A. Edris is with the Electric Power Research Institute, Palo Alto, CA
94304 USA.
D. Kidd is with the American Electric Power, Tulsa, OK 74103 USA. III. CABLE TERMINATION FAILURES
A. J. F. Keri and B. Mehraban are with the American Electric Power,
Columbus, OH 43215 USA. Three cable failures occurred during the commissioning stage
Digital Object Identifier 10.1109/TPWRD.2003.817724 of the Eagle Pass BtB installation. All cable failures occurred
0885-8977/03$17.00 © 2003 IEEE
PAULSSON et al.: HIGH-FREQUENCY IMPACTS IN A CONVERTER-BASED BACK-TO-BACK TIE 1411

Fig. 1. Eagle Pass VSC BtB. Single line diagram showing converters, dc
capacitors, phase reactors, harmonic filters, step up transformers, and line
breakers. Indoor switchgear including charging resistors is shown within
dashed lines. Medium voltage cables are shown in bold.
Fig. 2. Voltage measured between phase and ground on the CFE side and in
back-to-back operation. The upper curve shows the actual waveform and the
shortly after energization and they all were phase to ground in- lower diagram shows the harmonic spectrum as an average over 1 min.
side the cable terminations. Initially, the cause for the failures
was not clearly understood, but it was assumed to be related
transformers. Measurements were made at several locations on
to bad workmanship during installation of the cable termina-
both the U.S. (AEP-TCC) side and the Mexican (CFE) side.
tions. After the third failure, all of the cables and cable termi-
During back-to-back operation, very high frequencies and
nations were replaced with the same basic type of equipment as
high amplitude harmonics were found superimposed upon the
the original installation, but made by another manufacturer. The
power frequency. An example of the results from measurements
cable termination manufacturer performed the installation of the
on a phase cable at the point of connection to the transformer is
new cable terminations in order to ensure proper workmanship.
shown in Fig. 2 below.
A fourth cable failure occurred three days after the new cables
The dominating harmonics are 1.26 kHz (21st, IGBT
were energized. This failure, which also occurred inside a cable
switching frequency), and 12.4 kHz (207th, approx. ten times
termination, initiated a closer investigation.
the switching frequency). Other main harmonics are 180 Hz
Failures had occurred in different phases and at different ends
(3rd) and 3.78 kHz (63rd, three times the switching frequency).
of the cables. It became obvious that the cable terminations were
The 12.4-kHz harmonic, which was unexpected, varied be-
the root cause of the cable failures. Both the original and the
tween 13% and 40% of the power frequency voltage, depending
replacement cable terminations were of standard ac type with
on the operating mode of the BtBand the angledifference between
internal resistive/refractive stressgrading.
the U.S. and Mexican grids. The maximum measured amplitude
The replacement cable terminations were tested, along with
including all harmonics was about 37 kV peak. The 3rd, 21st,
many other type tests, according to IEEE standards for 25-kV
and 63rd harmonics had fairly constant amplitudes for different
rating. During the ac dry withstand test, 55 was applied
operating modes and were in accordance with design values.
for 6 h. According to IEC Standards, the short duration power
There was no significant difference between the three phases,
frequency withstand voltage is 50 for 24-kV rated equip-
between parallel cables in each phase or between the cables con-
ment. The guaranteed impulse voltage withstand is 150 kV. Both
nected to the reactors or to the transformers. In Fig. 3, it is shown
original and replacement cables were rated 24 kV. The protec-
an example of a corresponding measurement at the same loca-
tive level of the surge arresters at Eagle Pass is 65 kV peak and
tion as in Fig. 2, but with the AEP-TCC side blocked and dis-
the arresters are located at the low voltage side of the step up
connected from the U.S. grid.
transformers.
The dominating harmonics are 3rd and 21st, with amplitudes
in agreement with design values.
IV. FIELD MEASUREMENTS Measurements of the cable screen currents confirmed the
voltage amplitudes and the harmonic spectrum. Proper current
After the fourth cable failure, extensive field measurements balance between the two parallel cables in each phase was
were performed. The objective of the measurements was to an- confirmed by measurements. Transients from operation of
alyze the actual voltage stress on the cable and the cable termi- external breakers on the CFE side were also measured. The
nation insulation. Capacitive voltage dividers rated 50-kVrms, transients were low and could not have caused any cable
680 pF, and 5-MHz bandwidth were used together with a four- insulation breakdown.
channel oscilloscope of type LeCroy 9354 AL with 250-MHz It was clear from the measurements, that the system in BtB
analog bandwidth. operation had a local resonance near 12 kHz and that it was
Voltage measurements were made on the main phase cables excited by the IGBT switching. The resonance is determined
between the cable conductor and the cable screen at the ends by the cable capacitance, the stray capacitance, and the circuit
of the cables, where the screens are grounded. The cables reactance as seen from cable grounding on the CFE side to cable
were connected to ground at the phase reactors and at the grounding on the AEP-TCC side. See Fig. 1.
1412 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 18, NO. 4, OCTOBER 2003

Fig. 3. Voltage measured between phase and ground on the CFE side and with
the AEP-TCC side blocked. The upper curve shows the actual waveform and
the lower diagram shows the harmonic spectrum as an average over 1 min. Fig. 5. Example of the test voltage. The power frequency is 50 Hz and the
high-frequency component is 10.6 kHz. The voltage magnitude is 39.2 kV peak.
(voltage scale 10 kV/div, time scale 5 ms/div).

TABLE I
SUMMARY OF RESULTS FROM THE LABORATORY TEST

Fig. 4. Circuit for testing the cable terminations. C1 is blocking low

frequencies while L2 is blocking high frequencies. L1 is an inductance for
providing high-frequency resonance.

V. LABORATORY TESTS
It was anticipated that the high-frequency voltages were the maximum voltage amplitude at the higher frequencies. The
cause of the cable failures. Laboratory tests were performed as high-frequency source was a generator with variable speed and
verification. The tests were carried out at the ABB Corporate a high-frequency transformer. The room for the testing was a
Research Center in Västerås, Sweden. heat chamber where high outdoor temperature conditions like
Three test objects were prepared, each consisting of a short at Eagle Pass could be simulated. Fig. 5 shows an example of
cable (1.5 m) with cable terminations at each end (Fig. 4). One the test voltage.
end of the cables was provided with exactly the same type of All three cable terminations of resistive/refractive type failed
termination, which had failed lately at Eagle Pass. The other within 470 h. All cable insulation break-down occurred at the
end of the cables was provided with an APIT (ABB brand name) same physical location inside the terminations as was the case
termination. This type of cable termination, which is generally for all the terminations, which failed in the field at Eagle Pass. A
called geometric type, has an insulation characteristic expected comprehensive summary of the test results is shown in Table I.
not to be dependant on frequency. The cables came from the
same manufacturing batch as the replacement cables presently VI. FAILURE EXPLANATION
installed at Eagle Pass. Both types of terminations were exposed A simplified cross section diagram of a compact cable termi-
to exactly the same voltage stress. nation with resistive/refractive stressgrading is shown in Fig. 6.
The test circuit consisted of The termination is a slip over termination (SOT) type and is
• power frequency (50-Hz) supply; shown for indoor application, without sheds.
• high-frequency (7.5–10.6 kHz) supply. This type of cable termination can be made very slim and is
Both supplies were connected to the test object at the same easy to apply on to the cable. The end of the cable has to be pre-
time to obtain a high-frequency voltage superimposed upon the pared and then the termination is slipped over it. The interface
power frequency voltage [3]. between the cable and the termination is filled with high quality
The voltage amplitudes were controlled independently for silicon grease in order to avoid cavities, which could lead to
the two supplies. The high-frequency supply could be varied partial discharges (PD). For outdoor use, the terminations are
between 7.5 and 10.6 kHz, however, with some restrictions in provided with integrated sheds.
PAULSSON et al.: HIGH-FREQUENCY IMPACTS IN A CONVERTER-BASED BACK-TO-BACK TIE 1413

Fig. 6. Schematic outline of a slip over cable termination (SOT) for indoor
use. Equipotential lines and percent figures indicate the voltage distribution at Fig. 8. Electrical field strength along the stressgrading layer as a function of
power frequency. distance from the edge of the semiconducting layer of the cable, at different
frequencies.

Fig. 7. Circuit for calculating the voltage distribution and losses along the
stressgrading layer at different frequencies.

A stressgrading layer is moulded into the silicon rubber and Fig. 9. Losses in the stressgrading layer as function of the distance from the
it defines, together with the local stray capacitance inside the edge of the semiconducting layer of the cable, at different frequencies. Note:
cable, the stress distribution along the termination. The length the 12.4-kHz losses are 10 times the figures in the diagram.
of the stressgrading layer, its distribution of resistivity, and its di-
electric constant, in simplified terms, are chosen to give a linear
voltage distribution outside the termination at 50/60 Hz.
However, at higher frequencies, the coupling between the re-
sistance of the stressgrading layer and the stray capacitance in-
side the cable insulation results in a quite different distribution
of the voltage (Fig. 7). The higher the frequency, the higher
the electrical field strength near the edge of the outer semicon-
ducting layer of the cable. Rough calculations have been made Fig. 10. Same cable termination as in Fig. 6, but with equipotential lines
to illustrate the changing conditions. indicating the voltage distribution at high frequencies. Circle shows area with
high stress.
A very simplistic model based on 1000 series connected R/C
elements was used to calculate the electrical field along the
field strength in any small cavity. The field strength in such cav-
stressgrading layer at different voltage frequencies. The capac-
ities can far exceed 3 kV/mm, which is the approximate level for
itance (C) was given by the cable characteristics and kept con-
starting partial discharges in air. The PD will cause local tem-
stant, while individual resistor values were calculated to
perature increase and eventually lead to break down of the insu-
yield a linear voltage distribution along the stressgrading layer
lating material.
at 60 Hz.
Even if reality is much more complex, the calculations clearly
The electrical field strength (V/mm) and the losses
illustrate the increased insulation stress (Fig. 10) and risk for
(W/squarecm) in the semiconducting layer were calculated sep-
local PD at high frequencies. All failures experienced at site,
arately for the main frequencies measured at Eagle Pass BtB.
and in the test laboratory, have occurred at the edge of the cable
The actual 10.4-kVrms phase to ground voltage was chosen for
semiconducting layer.
the 60-Hz calculation. For 1.2 kHz (20th) and 12.4 kHz (207th),
calculations were made with voltage amplitudes 30% and 40%
of the power frequency voltage, respectively. This corresponds VII. CORRECTIVE ACTIONS
to the worst conditions measured at site. The results are shown All cable terminations at the Eagle Pass BtB installation have
in Figs. 8 and 9. been replaced by geometric type of cable terminations (APIT).
For the highest frequency, both the electrical field strength The geometric terminations are not considered to have any fre-
and the losses become very high near the edge of the semicon- quency dependency below the Megahertz range. The geometric
ducting layer of the cable. Further field amplification, due to dif- terminations are of a classical design structure, and are gener-
ferences in dielectric constants, may cause very high electrical ally used for high and extra high voltages, but can also be used
1414 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 18, NO. 4, OCTOBER 2003

velops into a conducting channel through the XLPE insulation

in the cable.
As can be concluded, the failure time lapse depends on the
frequency and amplitude of the harmonics. However, tolerances
of the characteristics of the stressgrading layer of the cable ter-
mination, workmanship at installation, environmental tempera-
ture, etc., may also influence the time lapse.
Cable terminations as well as other medium voltage equip-
ment are type tested only in accordance with standards and test
procedures for conventional power frequency ac applications.

Fig. 11. Schematic outline of a geometric type of cable termination. IX. SUMMARY AND CONCLUSION
Equipotential lines indicate the voltage distribution at any frequency.
Due to a high frequency and high amplitude internal oscilla-
tion in the Eagle Pass BtB installation, cable terminations have
for medium voltage applications. The electrical field distribu- failed after a very short time in service. The original cable ter-
tion is determined only by the stray capacitance in the cable and minations installed were of the resistive/refractive stressgrading
in the cable termination itself. Since there is no resistive grading type, which are commonly used up to 36 kV.
interacting with the stray capacitances, the voltage distribution The failure mechanism of these terminations has been ex-
is mainly independent of the frequency. plained and also verified by laboratory tests. All terminations
The highest field strength occurs inside homogeneous high at Eagle Pass have been replaced by geometric type cable ter-
quality insulating material. For the cable terminations installed minations, which are mainly insensitive to frequencies below
at Eagle Pass, this material is ethylene propylene (EPDM) the Megahertz range. As the high-frequency voltage was a re-
rubber (see Fig. 11). sult of an unforeseen local oscillation, RC filters have also been
The geometric type of cable terminations, which are now in- installed to damp the oscillation.
stalled at Eagle Pass, were tested in the laboratory simultane- Since installation of the geometric-type cable terminations
ously with the resistive/refractive type of terminations. For the in 2002, the Eagle Pass VSC BtB has been operating satisfac-
same test conditions and test periods, while all of the resis- torily, providing support to the AEP and CFE networks. The
tive/refractive type of terminations failed, none of the geometric knowledge which has been gained from the cable failures and
type of terminations were affected in any respect which could its remedy will be very important and beneficial for the appli-
influence the insulation performance. cations of VSC technology in the future.
The geometric type of cable termination naturally has a Under high-frequency harmonic conditions, insulation sys-
larger cross area than the resistive type and requires somewhat tems or insulation materials may be exposed to a type of voltage
more space for installation. For outdoor use, both types of stress, for which they have not been designed or tested. This is
terminations are provided with sheds for extended creepage a new situation, which has to be considered in connection with
and protection. VSCs and other systems which generate very high-frequency
In addition to replacing all of the cable terminations, RC fil- harmonics. The stress can be directly caused by the harmonic
ters have also been installed at the Eagle Pass BtB in order to voltages or indirectly by energizing local resonances in the elec-
damp the 12.4-kHz oscillation. trical circuit. Such excessive stress may, in some cases, lead to
a considerably reduced length of equipment life.
Based on the field measurements and the laboratory tests, fur-
VIII. PHYSICS OF THE FAILURES ther R&D work in order to fully understand and quantify the
impact of high-frequency voltage harmonics for various types
Field experience, laboratory tests, and calculations have all of equipment containing composite insulation is recommended.
pointed to the same root cause of the failures of the cables. The The amount of high frequency harmonic voltages tends to in-
high-frequency harmonics have caused local high voltage stress crease in new types of electrical installations while at the same
and temperature increase of the stressgrading layer of the termi- time the insulation becomes more compact and complex.
nation. This, in turn, has caused rapid break down of the cable
insulation and as a result, caused phase to ground fault.
The sequence of failure development is as follows: Local high REFERENCES
temperature, due to high-frequency losses in the stressgrading [1] A. Edris, Å. Petersson, M. Bahrman, A. Mcqueen, R. Haley, and D.
layer, leads to a dry out of the silicon grease in the interface Kidd, “Emerging application of Voltage Source Converter technology in
between the cable insulation and the stessgrading layer of the back-to-back asynchronous tie,” in VII SEPOPE, Curitiba, Brazil, May
termination. At high temperatures, the silicon grease will start 2000.
[2] T. Larsson, Å. Petersson, A. Edris, D. Kidd, R. Haley, and F. Aboytes,
migrating through the stressgrading layer and into the silicon “Eagle pass back-to-back tie: a dual purpose application of Voltage
rubber. As soon as any cavity becomes present, the elevated Source Converter technology,” in Proc. IEEE Power Eng. Soc. Summer
electrical field strength causes partial discharges in the cavity, Meeting, Vancouver, BC, Canada, 2001.
[3] M. Paede, W. Pfeiffer, and R. Plessow, “Testing of insulation systems
which, in turn, will cause further increased local temperature. with respect to high-frequency voltage stress,” Europe. Trans. Elect.
The high temperature will result in electrical treeing, which de- Power, vol. 12, no. 5, pp. 337–346, 2002.
PAULSSON et al.: HIGH-FREQUENCY IMPACTS IN A CONVERTER-BASED BACK-TO-BACK TIE 1415

Lars Paulsson (M’95) was born in Hässleholm, Sweden. He received the M.Sc. Abdel-Aty Edris (SM’88) was born in Cairo, Egypt. He received the B.S. de-
degree in electrical engineering from Chalmers University of Technology, gree (Hons.) from Cairo University in 1967, the M.S. degree from Ain-Shams
Sweden, in 1968. University, Cairo, Egypt, in 1973, and the Ph.D. degree from the Chalmers Uni-
Currently, he is with ABB, Västerås, Sweden, in development and application versity of Technology, Gothenburg, Sweden, in 1979.
of FACTS controllers. He joined ABB, Ludvika, Sweden, in 1969, and has held Currently, he is Manager of flexible ac transmission system (FACTS) tech-
various engineering positions within the ABB Company in Sweden, in the Re- nology at EPRI, Palo Alto, CA, where he has been since 1992. He is Target
public of South Africa and in the U.S. He has also been Technical and Marketing Leader of EPRI’s substations and transmissions asset utilization. He also spent
Manager for Ifö Electric High Voltage AB, Sweden, and worked in private busi- 12 years with ABB, Västerås, Sweden, and in the U.S., in development and
ness and consulting. application of reactive power compensators and high voltage dc transmission
systems.
Dr. Edris is a member of several IEEE and Cigré working groups.

Björn Ekehov was born in Gävle, Sweden, in 1958. He received the M.Sc.
degree in electrical engineering from Lund University of Technology, Sweden.
Currently, he is a Manager with FACTS Business Development at ABB,
Västerås, Sweden. He joined ABB in 1983 and has worked 16 years within the David Kidd (M’91) received the B.S.E.E. degree in electrical engineering from
field of development and design of both main circuit and control and protection New Mexico State University, Las Cruces, in 1987.
of HVDC systems. He has also been Technical Manager for FACTS systems Currently, he is with AEP, Tulsa, OK, in the Texas Transmission Planning
within ABB. group, where he has been working with FACTS technology. He spent ten years
with El Paso Electric Company, El Paso, TX, working in system protection.
Mr. Kidd is involved with various groups within ERCOT.

Stefan Halén was born in Västerås, Sweden, in 1953. He received the M.Sc.
degree in electrical engineering from the Royal Institute of Technology, Stock-
holm, Sweden, in 1987.
Currently, he is with ABB Corporate Research, Västerås, Sweden, in the field Albert J. F. Keri (SM’80) received the Ph.D. degree in electrical engineering
of power apparatus and dielectric diagnosis. He joined ASEA Research and In- from the University of Missouri, Columbia, in 1972, and the MBA degree from
novation, Västerås, Sweden, in 1987. Mr. Halén has also spent seven years with the Ohio University, Athens, in 1985.
R&D working at Vattenfall Utveckling AB Västerås and at the Electric Power Currently, he is a Principal Engineer with AEP, Columbus, OH, in the trans-
Center at the Royal Institute of Technology, Stockholm. mission planning dynamics area. He has been involved with system stability
Mr. Halén is a member of Cigré. analysis, FACTS, superconductivity, over-voltage calculations, harmonics in-
vestigation, field tests, single-phase switching techniques, transmission line loss
reduction techniques, insulation co-ordination, equipment failure analysis, FCL,
and consulting. He holds two U.S. patents and has published many IEEE and
Tomas Larsson (S’92–M’99) was born in Sävsjö, Sweden, on May 9, 1964. He Cigré papers.
received the M.Sc. and Ph.D. degrees in electrical engineering from the Royal He is a member of HKN and Tau Beta Pi, and of several IEEE working groups
Institute of Technology, Stockholm, Sweden, in 1991 and 1998, respectively. and chairman of GS subcommittee.
Currently, he is with ABB Power Systems, Västerås, Sweden, involved in
projects concerning voltage source converters, reactive power compensation,
and flicker mitigation. Dr. Larsson joined ABB in 1998.

Ben Mehraban (SM’02) received the B.S.E.E. and M.S.E.E. degrees in elec-
trical engineering from the University of Missouri, Columbia, in 1969 and 1971,
Lars Palmqvist was born in Malmö, Sweden, in 1952. He received the M.Sc. respectively, and the MBA degree from Ohio University, Athens, in 1985.
degree in chemical engineering from Lund University of Technology, Sweden, He has more than 30 years of experience in engineering, design, coordination,
in 1979. and project management for HV/EHV, GIS, HVDC, and FACTS projects. Mr.
He joined ABB, Sweden, in 1980, and has worked with development and Mehraban has published many IEEE papers and technical articles.
project management of cable accessories for 17 years. Mr. Mehraban is a member of several IEEE-PES working groups.