Medição de resistência elétrica muito baixa em corrente alternada

Measurement of very low electrical resistances in alternating current

Márcio Antônio Sens
Eletrobras Cepel
E-mail: [email protected]
Abstract: Electric cables for power transmission lines become increasingly large
cross-sections. In Brazil cables up to 2500 kCM (1267 mm²) are used in the
transmission from the Amazon Region to the Southeast Region. In these sections
certainly, the Skin Effect becomes also significant and must be verified the electrical
resistance in ac current. Since the samples are usually 5 to 10 meters, the expected
electrical resistance is of the order of few tens of micro-ohms. The paper will show
that the steel core may have some contribution to the resistance at high currents, of the
order of 2 to 3 kA.

Keywords: Skin Effect; ac resistance; measurement; aluminum; cable.

1 INTRODUCTION neighboring circuits also negligible, it can be

Concerns about the differences between the considered to have only one inductance with an
electric resistance in direct current and the equivalent series electrical resistance. The
electric resistance to alternating current for wires measurement principle adopted here was that of
and electric conductive cables of large transverse Rs and Xs.
sections, at the frequency of 60 Hz, are not new. In order to obtain these equivalent parameters, it
Much theoretical speculation and numerical is sufficient to measure the voltage and current
simulations have been made and disseminated during the current application period. By
since Lord Kelvin in 1887. Records of effective mathematical processing we obtain the apparent
measures of such effects, however, few are power - S (VA), the power dissipated in heat - P
found [ 1 ], [ 2 ]. The ac resistance measurements, in (W) and the reactive power - Q (var), and with
fact, constitute great technical difficulties. these, by dividing the respective power by the
In order to evaluate the resistance to quadratic current we obtain the impedance - Z,
alternating current of electrical conductor wires the resistance - Rac and the reactance - Xs,
and cables of large transverse sections, in respectively.
samples of up to 20 meters or in small samples of 2 EXPERIMENTAL TESTING SYSTEM
superconducting cables, the following are In the experiments dealing in this paper, single-
required: phase 166 kVA current sources were used, with
 Application and measurement of currents continuous adjustments of up to 5000 A, at
at high levels, exceeding 1000 A at 60 Hz; 60 Hz. For accurate measurement of high current
 Measurement of low voltage levels, less levels, three current transformers - CTs are
than 10 volts or less than 1 volt; available, a multi-ratio CT of 100 to 6000 A-
 The test object at stable temperature, in the 5 A (± 50 ppm), another of a multi-ratio, from
air or under liquid nitrogen. 100 to 12000 A-5 A (± 50 ppm) and a single ratio
For an electrical cable distant enough from other CT of 1000-1 A (± 2 ppm). All CTs are provided
metals, in a closed loop, with negligible parallel with tertiary compensation windings, to be in
capacitance and magnetic coupling with other parallel with the respective secondary windings.
1
XII Congresso Internacional de Metrologia Elétrica/Fortaleza - CE
For accurate measurement of the electric This difficulty was overcome by the use of a
currents, the current transformers - CTs were fast measuring system, synchronized with a fast
connected to triaxial shunts of class ± 0.03 %, current drive system, never exceeding 70 ms (1 to
through 4 wire probe connections at the power 4 cycles of 60 Hz). Thus, the sample does not
analyzer. receive enough energy for the significant
For the measurement of low voltage levels, temperature rise during the test duration and can
the sample segment under test was connected remain at (23 ± 2) ºC or at (77 ± 1) K. The tests
directly to a low voltage module of 1.58 V or, in can be repeated, because the surface power in the
case of overload, to an inductive voltage divider sample will be less than 1 μW / mm².
of 10-1 V, with the use of coaxial cable or The test circuit described in Figure 1, where
twisted pair of measurement cables. the thickest line represents the aluminum cable
Now the most important and relevant subject under test, in length generally in the order of 10
of the present paper — with high currents, there to 15 meters, but with potential draws of
is still a way to keep the object under the distances from one to 5 meters. A longer cable
temperature of the surrounding medium, for length than the segment under evaluation
repeated and increasing applications of currents becomes necessary for the sample to pass through
on the same sample, every five minutes of test. the high current power supply.

5 to 5000 A
5V
I1 = Norma Triax-Shunt ± 0.03% up to
0 to 30 A 500 kHz
600 V Olman Multi-CT V1 = Adapter Module of 1.58 V
100 A to 12 kA-5A Under Test Object 1 to 7 m

10Vmax

SCR Ratio Transformer

Gate
Drive

Sig. Out
Sync-out
Ext. Timing
GenRad 1396B
I1 V1 Ext. Trig.

Burst Generator Wide Band Power Analyzer

Sig. In
DC-Power Supply NORMA LEM D6100
12 V

Figure 1 - Schematic diagram of experimental testing system.

After the proper connections and adjustments, the SCR Gate Drive, simultaneously a trigger pulse
test is initiated by the Burst Generator, pre-set for is applied to the Power Analyzer LEM D6100 to
current application with duration of one, two, initiate the measurement.
three or maximum four cycles of 60 Hz, in single
A second after the measurement test finished.
mode. With the start of the current, through the
2
XII Congresso Internacional de Metrologia Elétrica/Fortaleza - CE
2.1 Alternating current resistance of samples The results of the variations of the electrical
of aluminum cables for electric power resistances with the applied current for the Duck
transmission cable, and the dissipated power are shown in
Although in the last 20 years many tests were Figure 3. The variations of resistance and the
contracted by Brazilian industries to measure the power dissipated in direct current were used as
electrical resistance to alternating current of references. The 60 Hz resistance of this cable is
electric cables for electric power transmission, 112 μΩ / m, according to the manufacturer [ 4 ].
and some even for rural electrification in steel The results showed that the used cable varies
cables, the results constitute laboratory more the electric resistance to alternating current
confidentiality and can not be published. with the current applied, in the order of 14 %
However, as an internal research to clarify above the resistance in direct current, for the
doubts about the electrical resistance of the current of 290 % of the rated current of the cable,
ACSR cables (aluminum with steel core for at the frequency of 60 Hz, as shown in Figure 2,
for the Rail cable.
reinforcement in the mechanical strength of
rupture) with three layers of aluminum wires, if 800
1.4 kA 2.0 kA 2.4 kA 2.8 kA
116
Used Rail Cable - Losses in 60 Hz

Resistance Rac/Rdc (%)

the electrical resistance at the frequency of 60 Hz
Power Losses (W/m)
700 114
600 954 kCM 112
could vary with the intensity of the applied 500 60 Hz
Direct 110
114 % current
current, for values well above the rated currents, 400 Rac/Rdc 100 % 108

by the inductive effects on the steel core [ 5 ] or by 300

(25 ± 1) ºC
106

mm
29,61mm
200 y = 0x2 + 66x 104
other effects, two similar cable samples were

30
100 R2 = 1 102

used for experimental laboratory investigations. 0

0 2 4 6 8 10
100

A sample of five meters of cable Rail – M.A.Sens-14/05/2017 22:14

Cabo Rail-Resistencia em 60 Hz.xls Applied Current I² (kA²)
954 kCM, for 970 A, used, withdrawn from the
operation and another one of cable Duck- Figure 2 – AC resistance for Rail cable – Used.
605 kCM, for 770 A, new, was selected to be By now, in the case of the new cable, for the
tested. The 60 Hz electrical resistance of both same condition, that is, for 290 % of the rated
samples was measured at very high current current, the Rac rose only 4 % in relation to the
levels, above rated values, up to 290 %. Rdc, as shown in Figure 3 for the cable Duck.
1.0 kA 1.4 kA 1.7 kA 2.0 kA 2.2 kA
Laboratory currents were used at 60 Hz 600 106
Power Losses P (W/m)

New Duck Cable - 60 Hz Loss 605 kCM

Resistance R ac/Rdc (%)

frequency, greater than 2 kA for Duck cable and 500

60 Hz
105
Rac/Rdc
more than 3 kA for the Rail cable. Sample 400 104 % 104
Direct
temperatures were maintained within the 300 Current 103
100 %
(23 ± 2) ºC C

tolerances of the laboratory environment even at 200 B

A M . A. Sens

y = 98,99 µΩ/m
102
24 mm

S t e e

high currents of (23 ± 2) ºC, using low thermal 100 R² = 0,99999354 101

energy for the measurements, lasting only four 0 100

0 1 2 3 4 5 6
M.A.Sens-05/02/2015 16:00
cycles (4/60 s) for each reading, with an interval DistribuicaoDeCorrenteCaboDuke-Rev01.xls Applied Current I²(kA²)

of at least five minutes between measurements or

reapplications of measure currents. Figure 3 - AC resistance for Duck cable – New.
The differentiated behavior between samples of
3 EXPERIMENTAL RESULTS new and used ACSR cable can be observed, by
The results of the variations of the resistances the Figure 2, and Figure 3. The surface layer of
with the current applied to the Rail cable as well aluminum oxide is of high electric resistivity.
as the dissipated power are shown in Figure 2. The oxidized surface layers of the cables are not
The electric resistance in 60 Hz of this cable is the same, but one cable is older in service than
73 μΩ / m, according to the manufacturer [ 4 ]. the other, as shown inside of the pictures.
3
XII Congresso Internacional de Metrologia Elétrica/Fortaleza - CE
 A third cable sample was also assessed, but at Depending on the environmental conditions, the
different temperatures, where the heating of the surface of the aluminum degrades more or less.
sample was obtained by long time measurement Moreover, this layer of aluminum oxide is of
current itself. This is the Grosbeak heat-resistant high resistivity and hard material. With the
cable, which is a two-layered aluminum wires. It oxidized surface layer, the circulation of electric
was tested at temperatures up to 180 °C on current between layers and between adjacent wire
currents up to 1000 A, at 60 Hz, as Figure 4. conductors is reduced, forcing the current
170
through the turns of the cable's aluminum wires.
Specific Resistance Rt (mW/m)

Heat-resistant aluminum cable - Grosbeak with steel The turns way, on the other hand, are longer than
150
core
the cable itself and, therefore, have greater
130 a 20 = (4.413 ± 0.067) x 10-3 ºC-1
R20 = (87.7 ± 1.6) mW /m
electrical resistance than the compacted cable.
25.15 mm Possibly, the greatest increase of the electric
110

60 Hz
90 resistance to alternating current in the aged cables
70 can still be attributed to the magnetic flux density
50
in the steel core, because it is a cable with odd
0 20 40 60 80 100 120 140 160 180 200
M .A .S e ns - 07/12/2005
R e s is te nc ia e m 60Hz-C abo s .xls Temperature T ( ºC )
layers. But for new cables, the finding of this
hypothesis is still a challenge for other
Figure 4 - AC resistance for Grosbeak cable. experimental investigations, such as those carried
Using the data obtained experimentally, the out at Ontario Hydro, Canada [ 3 ].
coefficient of variation of the resistance with the
temperature, with the respective uncertainties, 5 REFERENCES
was calculated as also indicated in Figure 4. [1] G. SCHRÖDER, J. Kaumanns, R. Plath,
"Advanced measurement of ac resistance on
3.1 Validation of results skin-effect reduced large conductor power
Four standard resistors of current rated of 15 to cables", Jicable ´11, Paper A 8.2.
600 A, with rated resistances of 100 μΩ to [2] Gero SCHRÖDER; Dominik HÄRING;
100 mΩ, were evaluated in DC and AC by the Andreas WEINLEIN and Axel
techniques described here, in currents up to BOSSMANN “AC resistance measurements
on skin-effect reduced large conductor
1100 A for all units, even for that one of 15 A.
power cables with standard equipment”, on
The results indicated a linear correlation of the 9th Jicable'15 - Versailles 21-25 June, 2015.
power with the quadratic current, between 35 and [3] J. S. Barrett and O. Nigol - Ontario Hydro-
1100 A, greater than or equal to 0,9999995. CA; C.J. Fehervari and R. D. Findlay –
kCMaster University-Ontario-CA – “A New
4 CONCLUSIONS Model of AC Resistance in ACSR
It was concluded from the results presented here Conductor”, IEEE Transaction on Power
that the test system meets the expectations of Systems, Vol. PWRD-1 N. 2, April 1986.
finding differences between the electric [4] Nexans, Catálogo de Produtos,
resistance in alternating and continuous current disponibilizado em
and that in alternating current, in fact, the http://www.nexans.com.br/eservice/Brazil-
resistance is greater, according to all the literature pt_BR/pdf-family_-
74539/Cabos_de_Alum_nio_Nu_CA_Serie_
dealing with skin effect. However, the trials
kCMIL_.pdf, acesso em 26/02/2017.
revealed an unexpected result. The results
[5] Sens, Márcio Antônio and Ueti, Edson,
showed that new and used cables have distinct
“Envelhecimento de cabos condutores para
behavior under high currents. This differentiated transmissão de energia elétrica” at XXIV
behavior between samples of ACSR cables new SNPTEE - National Seminar On Production
and used is certainly due to the surface aging of And Transmission Of Electrical Energy,
the aluminum of the cable during the use. October 22 – 25/17, Curitiba - PR – Brazil.

4
XII Congresso Internacional de Metrologia Elétrica/Fortaleza - CE