Over-Current Testing of HTS Tapes

Document Sample
Over-Current Testing of HTS Tapes Powered By Docstoc
					                                                                                                                                     3LB06        1

                                              Over-Current Testing
                                                 of HTS Tapes
                                    Jinwun W. Lue, Robert C. Duckworth, and Michael J. Gouge

                                                                             objective of the present study is to find this pulse heating
   Abstract—      High-temperature      superconducting     (HTS)            thermal limitation of HTS tapes.
transmission cables are subjected to short-circuit fault currents               Another concern is whether the magnitude of the over-
10 to 30 times the normal operating current and lasting up to 15             current is high enough to damage the tape
cycles. These over-currents will drive the HTS conductor normal
and generate heat during the fault. A concern is whether the fault
                                                                             electromechanically. As the HTS tape performance improves,
current will either electromechanically or thermally damage the              fewer and fewer tapes are needed to carry the cable design
HTS conductor and degrade it or burn-out the tape altogether.                currents. When there is a fault over-current, the peak current
Electromechanical and thermal limitations of over-current pulses             might be high enough to deform the HTS tape. Therefore,
were measured on BSCCO and YBCO tapes in a liquid nitrogen                   another purpose of the present study is to find how high a
bath. With pulse lengths as short as 35 ms, it is found that single          current an HTS tape can take without damage – the
BSCCO and YBCO tapes can be pulsed to at least 1 to 1.2 kA
without being damaged electromechanically. Longer pulses at
                                                                             electromechanical limitation of the HTS tapes.
moderate (450-750 A) over-currents indicated that HTS tapes
can be heated transiently to over 400 K range without suffering
degradation. Thus, it is likely that other considerations of the                             II. EXPERIMENTAL PROCEDURE
cable rather than the HTS tape itself would set the limit for
                                                                                A series of experiments with over-current pulses were
short-circuit fault protection.
                                                                             performed on BSCCO and YBCO tapes. Each of the 20-cm
   Index Terms — High-temperature superconductor, Power                      long HTS tapes was laid straight on a G-10 cylinder and
transmission cable, Short circuit                                            covered with layers of CryoflexTM dielectric tapes to simulate
                                                                             an HTS cable construction and tested in an open liquid
                                                                             nitrogen bath. To study the effect of reducing the peak
                          I. INTRODUCTION                                    conductor temperatures via electrical and thermal stabilizers,

H     IGH temperature superconducting (HTS) transmission                     samples with an extra Cu-strip (3.05 mm x 0.25 mm) laying
      cables are subjected to short-circuit fault currents in a              on top of the HTS tape were also tested. Table 1 shows the
utility setting. For ac applications, fault over-currents 10 to 30           seven different sample configurations tested in the
times the operating current and lasting up to 15 cycles are                  experiments. The Cu-plated BSCCO tape (sample 1) made by
expected, depending on the load and circuit breakers in the                  American Superconductor Corporation (AMSC) has a 15-µm
installation. The over-current will drive the HTS conductor                  thick layer of copper plated around the tape. The stainless
normal and thus generate heat during the fault. A concern is                 steel-BSCCO tape (sample 3, also made by AMSC) has two
whether the fault current will over-heat the HTS conductor                   25-µm thick layers of stainless steel laminated on both sides of
and degrade the superconducting properties or burn-out the                   the silver-alloy BSCCO tape. The YBCO tape (sample 5, also
HTS tape altogether. A few over-current tests have been                      made by AMSC) has a 50 µm thick layer of copper laminated
performed on prototype HTS cables resulting in moderate                      on the HTS side of the YBCO tape. Sample 7 consisted of two
temperature rises that showed no appreciable degradation to                  strips of copper and was used to verify/calibrate the
the cables [1,2]. But none of these cables were tested to find
the pulse heating limitation on the HTS conductor. One                                                   TABLE I
                                                                                        SAMPLES TESTED FOR OVER-CURRENT LIMITATIONS
   Manuscript received October 5, 2004. This work was supported in part by                                                                   Ic
the U.S. Department of Energy, Office of Electric Transmission and            Sample             Configuration              Dimensions     (amp)
Distribution, Superconductivity Program for Electric Power Systems, under
contract No. DE-AC05-00OR22725 with UT-Battelle, LLC.                            1     Cu-plated BSCCO                   4.27 x 0.25 mm      117
   M. J. Gouge is with Oak Ridge National Laboratory, Oak Ridge, TN 37831        2     Cu-plated BSCCO + Cu strip                            114
USA. (phone: 865-576-4467; fax: 865-576-7770; e-mail: gougemj@
                                                                                 3     Stainless steel laminated BSCCO   4.75 x 0.345 mm     126
   J. W. Lue, was with Oak Ridge National Laboratory, Oak Ridge, TN              4     ss-BSCCO + Cu strip                                   129
37831 USA. (e-mail:                                              5     Cu-laminated YBCO                 10 x 0.149 mm       161
   R. C. Duckworth is with Oak Ridge National Laboratory, Oak Ridge, TN
                                                                                 6     Cu-laminated YBCO + Cu strip                          148
37831 USA. (e-mail:
                                                                                 7     2 strips of Cu                    3.05 x 0.25 mm       -
                                                                                                                                         3LB06               2

 temperature measurements and thermal analysis procedures.                                 -8
                                                                                      2.5 10                                                 120
    A 3-kA, 30-V dc power supply was used for both the V-I
                                                                                                                   860 A for 180 ms
 measurement of the HTS conductors and for supplying the                               2 10
                                                                                                                   860 A for 340 ms          100
 over-current pulses. A thin-film RTD was laid under the HTS
 tape or between the HTS tape and the Cu-strip at the center of

                                                                           ρ (Ω-m)
                                                                                      1.5 10                                                 80

                                                                                                                                                   RTD (Ω)
 the conductor to measure the temperature of the conductor.
 The voltage of the middle 10 cm (i.e. at the 5 and 15 cm                              1 10
 points of the 20-cm long tape) of the conductor was monitored
 to find the voltage excursion of the conductor during the                             5 10                                                  40
 current pulse. First, the V-I curve of the sample was
 measured. Then a short over-current pulse was applied to the                              0                                                 20
                                                                                                2   3        4         5         6        7
 conductor. The V-I curve was measured again to see if there                                                    Time (s)
 was any degradation to the HTS tape. This process continued               Fig. 2. Conductor resistivity and RTD measurement for the two pulses.
 until the HTS tape was damaged or the power supply reached                to reach the peak, indicating some cooling between the
 its limit. The second series of tests on a given sample                   conductor sample and the RTD. The cool-down part of the
 configuration was to set the over-current level somewhat                  RTD curve can be fitted to an exponential decay function. It
 lower than the electromechanical or power supply limit and to             was found the cooling time constant was about 9 s.
 increase the pulse length to increase the heating to the                  Extrapolating the exponential fitting curves to the end of the
 conductor. A thermal limitation was found for each of the                 current pulses showed a temperature of 157 K for the 180-ms
 HTS samples.                                                              long pulse and 270 K for the 340-ms long pulse. This
                                                                           indicates that the RTD did not accurately measure the peak
                                                                           temperature rise even with the extrapolating correction.
                 III. VERIFICATION WITH STRIPS OF CU                          Integrating the V-I products over the pulse lengths in Fig. 1
    A thin-film RTD at the center of the sample was used to                gives 309 J/cm3 of heat generation density for the 180-ms shot
 measure the temperature rise during the over-current pulses.              and 955 J/cm3 for the 340-ms shot. The specific heat of copper
 The resistivity of the conductor increases as the temperature             as a function of temperature is available. Integrating the
 rises. Because of the known temperature dependence of the                 copper specific heat from 77 K to 184 K gives 280 J/cm3 of
 copper resistivity and specific heat, sample 7 of two strips of           heat absorption by the conductor for the 180-ms shot and
 copper was used to verify the temperature measurements and                integrated to 343 K gives 815 J/cm3 for the 340-ms shot. Both
 the thermal analysis. Figure 1 shows two shots of an 860-A                are slightly lower than the V·I integral. The energy absorption
 peak pulse for pulse lengths of 180 and 340 ms. The voltage               factor (F, see equation (1) below) by the conductor itself
 traces coincide with each other in the beginning, and reach               appears to be 91 % for the 180-ms shot and 85 % for the 340-
 0.48 V at an x-axis time of 2.5 s and dropped off for the 180-            ms shot. These results show that there is some cooling during
 ms long current pulse and continue to 1.0 V for the longer                the pulse and there is additional cooling for the RTD
 340-ms current pulse.                                                     measurement. Therefore, the technique of using the resistivity
     Figure 2 shows the conductor resistivity changes and the              rise of the conductor gives a better indication of the conductor
 RTD responses of these two shots. The resistivity peaked at               temperature, and will be used for the remainder of this paper.
 9.3 x 10-9 Ω-m for the 180-ms long pulse which corresponds
 to a conductor temperature of 184 K, and reached 2.0 x 10-8
 Ω-m for the 340-ms long pulse which corresponds to a                                IV. ELECTROMECHANICAL LIMITATION OF HTS TAPES
 conductor temperature of 343 K. The RTD shows a                              The power supply was remotely controlled by a square
 broadening of the temperature rise and a delay of about 0.7 s             wave voltage pulse to supply a pulsed current to the sample.
                                                                           Because of the dc nature and voltage limitation of the power
        1200                                                 1.2           supply there is an inherent current rise and decay time. The
                   860 A for 180 ms
        1000       860 A for 340 ms                          1             actual start up time of the current pulse depends on the
                                                                           magnitude of the control voltage also. This made it somewhat
        800                                                  0.8           difficult to obtain desirable short pulses. The shortest pulse
I (A)

                                                                   V (V)

        600                                                  0.6           length that can be obtained with an appreciable flat top was
                                                                           about 35 ms. In a series of tests to find the electromechanical
        400                                                  0.4           limitation of HTS tapes, the pulse length was set from 35 to
        200                                                  0.2           60 ms. For sample 1 (Cu-plated) and 3 (stainless steel
                                                                           laminated) BSCCO tapes the maximum peak current obtained
          0                                                   0            was ~1 kA. After this high current pulse the V-I curve was
           2.2    2.3   2.4    2.5 2.6        2.7    2.8   2.9
                              Time (s)                                     found to be the same as the initial V-I. Thus, it is seen that
Fig. 1. Two 860–A pulses on sample 7 ( 2 strips of Cu).                    ~115-130 A critical current (Ic) BSCCO tapes can take a short
                                                                                                                                                                            3LB06                       3

                                                                                    -8                                                                                              -8
             1400                                                            7 10                                    800                                                   8 10
                                                                                    -8                                                                                              -8
             1200                                                            6 10                                    700                                                   7 10
             1000                                                            5 10
                                                                                                                     600                                   ρ               6 10
                                                                                                                                    I                                               -8

                                                                                         ρ (Ω-m)

                                                                                                        I (A)
                 800                                                         4 10                                    500                                                   5 10

                                                                                                                                                                                         ρ (Ω-m)
     I (A)

                 600                                                         3 10
                                                                                    -8                               400                                                   4 10
                 400                                                         2 10
                                                                                    -8                               300                                                   3 10
                                                                                    -8                               200                                                   2 10
                 200                                                         1 10
                                                                                                                     100                                                   1 10
                      0                                                     0
                          2   2.1     2.2    2.3     2.4         2.5     2.6                                          0                                                     0
                                            Time (s)                                                                   2.4         2.6    2.8         3        3.2       3.4
                                                                                                                                           Time (s)
                                    Initial ramp                                                   Fig. 4. Resistivity change of the Cu-plated BSCCO sample during a 450-A,
                      60            After pulsing to 1.36 kA                                       320-ms over-current pulse.
             V (µV)

                                                                                                   (sample 1), Fig. 4 shows a 450-A, 320-ms pulse shot. The
                      20                                                                           resistivity of the tape increased sharply during the pulse and
                                                                                                   reached a peak of 7.6 x 10-8 Ω-m which corresponds to a
                          0                                                                        temperature of 725 K. After this shot the V-I measurement
                                                                                                   showed no difference from the initial V-I. The next shot at
                         80         100      120           140         160      180                450 A for 330 ms raised the peak temperature to about 800 K.
                                                   I (A)                                           The Cu-plated BSCCO tape was significantly degraded: the Ic
                                                                                                   dropped from 119 A to below 50 A.
Fig. 3. Cu-laminated YBCO tape resistivity increase during a 1.36-kA pulse
(top) and V-I curve measured after the shot as compared to the initial V-I                            For the Cu-laminated YBCO sample 5, Fig. 5 shows that a
(bottom).                                                                                          shot of 675 A for 200 ms that raised the average tape
over-current pulse to at least 1 kA without being damaged                                          temperature to 370 K. After this shot there was no change in
electromechanically. For the Cu-laminated YBCO tape                                                the V-I characteristics. The next shot of 330 ms increased the
(sample 5), there was no change in V-I curve after pulsing to                                      peak temperature to 670 K. After this shot, the Ic of the YBCO
1.23 kA. Figure 3 (top plot) shows the resistivity increase of                                     tape dropped from 160 A to 90 A–a large degradation.
the conductor when the peak current was increased to 1.36 kA
and the V-I curve measured after the pulse shot as compared                                                          1000                                                       6 10

to the initial V-I (bottom plot). There was a slight degradation                                                                                                                         -8
                                                                                                                      800                                                       5 10
- the Ic decreased from 161 A to 157 A. The resistivity
increase indicates the YBCO tape was heated to an average                                                             600                                                       4 10

                                                                                                                                                                                              ρ (Ω-m)
                                                                                                             I (A)

temperature of 370 K. At the same time, the electromechanical                                                         400                                                       3 10

load on the YBCO tape produced a buckling compressive                                                                                                                                    -8
                                                                                                                      200                                                       2 10
stress of 0.38 MPa (56 psi). The thermal limitation test
described later indicates that this small degradation was more                                                             0                                                    1 10
likely from the electromechanical load.                                                                               -200                                                      0
                                                                                                                               2    2.2     2.4        2.6       2.8        3
                                                                                                                                                  Time (s)
                                                                                                                 1200                                                           1.2 10
                      V. THERMAL LIMITATION OF HTS TAPES
                                                                                                                 1000                                                           1 10
   The resistivity of the Cu-plated BSCCO, the stainless steel
laminated BSCCO, and the Cu-laminated YBCO tapes were                                                                800                                                        8 10
                                                                                                                                                                                              ρ (Ω-m)

all measured from room temperature to about 70 K with a GM
                                                                                                     I (A)

                                                                                                                     600                                                        6 10
cryocooler. Superconducting transition temperatures of 109 K
were found for the BSCCO tapes and 91 K for the YBCO                                                                 400                                                        4 10
tape. The resistivity above the superconducting transition can                                                       200                                                        2 10

be fitted quite well with linear functions of temperature. These
                                                                                                                       0                                                        0
linear temperature dependences were extrapolated above 300                                                              2.2        2.4     2.6        2.8            3     3.2
K in the present experiment to determine the average                                                                                            Time (s)
temperature reached in the conductors.
    To find the pulse heating limit on the HTS samples, pulse                                       Fig. 5. Cu-laminated YBCO tape resistivity increase from a 200-ms pulse
currents between 450 and 750 A were used and the pulse                                              (top) and a 330-ms pulse (bottom).
duration was increasing in successive shots to increase the
Joule heating in the conductor. For the Cu-plated BSCCO tape
                                                                                                                                                                                3LB06     4

                 VI. EFFECT OF ADDITIONAL STRIP OF CU                                                                            1.1
   When a strip of Cu with comparable cross section to the                                                                        1

                                                                                                           Absorption Fraction
HTS tape is added in parallel to the HTS tape as in samples 2,                                                                   0.9
4, and 6, higher peak currents can be drawn from the power
supply. In these samples, the peak current reached about 2 kA.
However, about 60 % of the total current flows through the                                                                       0.7
Cu-strip during the over-current pulse. So the peak current in                                                                   0.6         tau = 1 s
                                                                                                                                             tau = 2 s
the HTS tape was actually lower than the tests with the HTS                                                                      0.5         YBCO w/Cu
tape alone. Figure 6 shows that the peak current obtained for                                                                                YBCO w/ Cu + strip
sample 6 went to 2 kA as compared to the 1.36 kA shown in                                                                              0   100    200   300       400   500   600   700
Fig. 3. The conductor temperature only reached 230 K as                                                                                             Pulse Length (ms)
                                                                                                     Fig. 7. Heat absorption fraction for the YBCO samples, with and without an
          2500                                                             2.5 10
                                                                                     -8              additional strip of Cu.
                     YBCO + Cu-strip
          2000                                                             2 10
                                                                                                     and the YBCO core was assumed to be 0.7 of Ag. The Cp
                                                                                                     integrals were performed on selected shots for each sample
          1500                                                             1.5 10
                                                                                                     and are compared with the V·I integrals over the pulse length.

                                                                                     ρ (Ω-m)
  I (A)

          1000                                                             1 10
                                                                                -8                   Figure 7 shows the resulting heat absorption fractions for the
                                                                                                     two YBCO samples. Near adiabatic heat absorption by the
          500                                                              5 10
                                                                                                     conductor itself is observed for pulse lengths up to the 100-ms
                                                                                                     range. For longer pulses, the results fit well with a heat
            0                                                               0
                 2           2.1       2.2         2.3           2.4     2.5                         diffusion process to the surrounding coolant and insulating
                                             Time (s)                                                tapes with decay time constants of 1-2 s.
Fig. 6. The addition of a strip of Cu raised the peak achievable current
while decreased the resistivity increase of the Cu-laminated YBCO sample.
                                                                                                                                                 VIII. CONCLUSION
compared to 370 K without the added Cu strip. Thus, the                                                 Electromechanical and thermal limitations of over-current
shunting function of the additional Cu increases the                                                 pulses were measured on BSCCO and YBCO tapes. With
electromechanical limitation by about a factor of two. Longer                                        pulse lengths as short as 35 ms, it is found that the ~115-130
pulse lengths can also be tolerated without over-heating the                                         A Ic BSCCO and 150-160 A Ic YBCO tapes can be pulsed to
HTS tape. In sample 2, the pulse length at 710-A current was                                         at least the 1-kA range without electromechanical damage.
extended to 600 ms (as compared to sample 1 of 450 A for                                             Longer pulses at moderate over-currents indicated that both
320 ms) without degrading the BSCCO tape. In sample 6, the                                           HTS tapes can be heated to over 400 K without suffering
pulse length at 845-A current was extended to 640 ms (as                                             degradation. However, the heating of the tapes accelerates as
compared to sample 5 of 655 A for 330 ms) with only a small                                          the temperature and the resistivity of the tape gets higher.
Ic degradation from 148 A to 142 A in the YBCO tape.                                                 Thus, a prudent design peak temperature of the HTS tape for
                                                                                                     short-circuit, over-current faults would be 200-300 K. Other
                                                                                                     considerations of the HTS cable, such as nitrogen pressure
                               VII.   HEAT DISSIPATION                                               surge, vapor generation and thus high voltage breakdown may
   During the over-current pulse the V-I product gives the                                           further limit the design temperature. When an additional
Joule heating power. Integrating V·I over the pulse length, τ                                        copper strip of about the same cross section as the HTS tape
gives the total heat generated in the pulse. This heat is mostly                                     was added, both the over-current magnitude and duration
absorbed by the conductor itself. Integrating the volume                                             limitations were found to be about doubled. This is apparently
specific heat, Cp of the conductor from the initial temperature                                      due to the shunting function and the added heat capacity of the
(77 K), To to the peak conductor temperature, Tp from the                                            copper strip.
resistivity peak gives the total heat absorbed by the conductor.
Thus, there is a heat balance equation:
                                                             τ                                                                                     REFERENCES
                     v ∫ p C p dT

                                               =         F   ∫
                                                                   V ⋅ I dt ,                  (1)   [1]   J. W. Lue et al., “Fault current tests of a 5-m HTS cable,” IEEE Trans.
                                                                                                           On Applied Superconductivity, Vol. 11, no. 1, pp. 1785-1788, 2001.
where v is the conductor volume between the voltage taps, and                                        [2]   J. A. Demko et al., “Performance tests of an HTS power transmission
                                                                                                           cable splice,” Adv. Cryog. Eng., Vol. 47, pp. 599-605, 2002.
F is the absorption fraction. An F of 1 corresponds to
adiabatic condition. The specific heats of the HTS samples are
calculated as the composite specific heat from their
components. The specific heats of Cu, Ag, and Ni are taken
from CryodataTM and fitted with polynomial functions. The
specific heat of the BSCCO core was assumed to be 0.6 of Ag,

Shared By: