1702
IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 19, NO. 3, JUNE 2009
30 m HTS Power Cable Development and Witness Sample Test Victor E. Sytnikov, Vitaly S. Vysotsky, Alexander V. Rychagov, Nelly V. Polyakova, Irlama P. Radchenko, Kirill A. Shutov, Sergey S. Fetisov, Alexander A. Nosov, and Vasily V. Zubko
Abstract—In the framework of the Russian R&D Program for superconducting power devices, an experimental 30 m HTS power cable has been developed. Three 30 m phases with nominal current of 2 kA and 20 kV operating voltage were delivered as the result of the project. Variations were used in basic HTS materials, cryostats and current leads for the cable design in this research project. All phases were made of 1G HTS Bi-based tape. The details of the design of the cable are discussed. Before the full length cable test a 5 m witness sample has been cut, heavily instrumented and tested. Results of the witness sample tests and analysis of fault current behavior of the cable are presented. Index Terms—Critical currents, fault currents, HTS power cables, V-I characteristics.
I. INTRODUCTION HE Russian R&D Program for superconducting power devices is underway, supported both by government and electric power companies. In this program R&D of HTS power cables is considered as most advanced and close to commercialization. A 5 m piece of the full scale power cable has been tested before as a first prototype [1]. Technological experiments were performed during 5 m cable developments to work out industrial cable production. On the basis of these developments and tests, an experimental 30 m three phase (3 30 m) HTS power cable has been produced as a prototype for longer cables. The cable has been delivered to the specially developed test facility for HTS power devices at the R&D Centre for Power Engineering in Moscow. The proposed parameters of this power cable are 2000 A and 20 kV. The cable behavior at overloads during faults in a grid was evaluated by numerical simulations. A 5 m witness sample has been cut from the full length cable, heavily instrumented and extensively tested at the low voltagehigh current test facility at the Russian Scientific R&D Cable Institute (VNIIKP) [1]. The next step in the Program is the production and delivery a three phase of 200 m HTS power cable. This cable will be first tested at the R&D Centre for Power Engineering in Moscow and then installed to the substation “Marfino” in the Moscow city distribution grid.
T
Manuscript received August 18, 2008. First published June 05, 2009; current version published July 15, 2009. This work was supported by the Federal Grid Company “United Energy Systems” of the Russian Federation. V. E. Sytnikov, V. S. Vysotsky, A. V. Rychagov, N. V. Polyakova, I. P. Radchenko, K. A. Shutov, S. S. Fetisov and A. A. Nosov are with the Russian Scientific R&D Cable Institute (VNIIKP), 111024 Moscow, Russia (e-mail:
[email protected]). V. V. Zubko is with the IHEP, 142281 Protvino, Moscow, Russia. Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TASC.2009.2019046
In this paper we present details of the 30 m cable design peculiarities, fault behavior analysis and the results of the witness sample tests. II. THE 30 m CABLE DESIGN We considered the 30 m cable development as mainly a research project necessary to work through the technology and production of HTS power cables. That is why variations were used in basic HTS materials, cryostats and current leads for the cable design. The basic design of the cable is 2-layers of HTS tapes, cold insulation, and copper, non-superconducting shield. The non-superconducting shield design has been accepted due to limited budget in this project. The sketch of the cable and its photo are shown in Fig. 1. Some details of the cable parts are listed in the Table I. Cold dielectric design was chosen because we consider using superconducting shield in future projects to eliminate a major source of concern of impacted abutters—external electromagnetic fields. Cold dielectric cables are much simpler to connect to cryogenic systems also, because their cryostats are under zero potential and do not need to be insulated from cryogenics. All modern advanced cable projects use cold dielectrics [2], [3], while in some old projects a warm dielectric was used [4]. The former was made of stainless steel spiral covered by two layers of the copper wires to provide HTS protection at fault. has been selected after fault Copper cross-section behavior analysis (see below). The superconducting cable layers were made of two types of 1G HTS Bi-based tape that were selected after extensive technological test [1]. For one phase the CT-OP tapes from Sumitomo Electric Industry Co. were used (self field, 77 K, critical current 100–105 A, 42 tapes in total). Two phases were made of American Superconductor Co. Hermetic wires (self field, 77 K, critical current 115 A, 40 tapes in total). Different tapes were used to verify the survival of superconducting properties of different HTS tapes during all technology routes for the power cable production. The central former and the HTS layers cabling were done by use of the specially upgraded cabling machines at the VNIIKP workshop. The direction and twist pitches of HTS tapes were calculated by using computer codes for HTS power cables optimization developed in VNIIKP. The optimization methods are based on theories in [5], [6]. In total about 110 m of the cable core were produced. The insulation of the cable core by the conventional cable paper has been done at the cable factory “Kamkabel” at the city of Perm (some 1500 km from Moscow). The non-superconducting copper shield from copper tapes has been made at the “Kamkabel” factory as well. After insulating and shielding, the cable core was cut into pieces of the certain length and terminations were soldered to the superconducting layers to
1051-8223/$25.00 © 2009 IEEE
SYTNIKOV et al.: POWER CABLE DEVELOPMENT AND WITNESS SAMPLE TEST
1703
Fig. 1. Sketch of the cable phase design and the photo of the cable. TABLE I SOME DETAILS OF THE CABLE DESIGN Fig. 2. Current leads connected to cryostats.
connect the cable core with current leads. Then, the three cables cores were inserted into the three flexible cryostats. The 30 m flexible cryostats were produced by Nexans Co in Hannover (Germany). We used two of three cryostats with O.D. of 110 mm and one with O.D. of 92 mm to compare the heat losses in cryostats (see Table I). Three pairs of current leads were developed and produced by two different Russian institutions. The current leads were developed for a 4 kA maximum current and 20 kV nominal’ voltages. All current leads were tested at cryogenic conditions at 70 kV during 1 min. The photo of the current leads connected to the cryostats is shown in Fig. 2. The cable has been delivered and installed at the special test facility for tests of HTS power devices (Fig. 3) mentioned above. This test facility can perform tests at voltages up to 120 kV and long time currents up to 2 kA. It is equipped with all necessary control and measuring devices to test powerful electro technical devices. The cryogenic system for the test facility is being delivered by the Stirling Cryogenics and Refrigeration BV, the Netherlands. Right now the cryogenic power will be up to 3.4 kW at 77 K with the possible future upgrade up to 7 kW. The cryogenic system permits variation of the cold power, pressure and temperatures in wide ranges. The important thing about the test facility is that there are installed electrical reactors as loads that permit testing of different superconducting power devices under the full load. It means the possibility to model the power devices behavior in real conditions in a grid.
Fig. 3. The 30 m three phase cable installed at the test facility.
Currently, cryogenics of the test facility is being installed. The full scale cable tests under load will be started in the coming months. III. EVALUATION OF THE CABLE’S BEHAVIOR AT THE FAULT It is important to evaluate the electrical and thermal behavior of an HTS cable during fault conditions to estimate possible overheating. In addition, a major concern from the utility perspective is the time required to cool the cable down to its normal operating temperatures and return the cable back to service. We performed the numerical, electro-magnetic and thermal analysis of our cable behavior during a suggested fault of 30 kA for 1.5 s. It appears that during the first cycle, 90% of the fault current will redistribute from the HTS superconducting layers to the former. During fault the maximum currents are about 1000 A in each HTS layer while in the each of the copper layer of the former, currents are about 13 kA as shown in Fig. 4. Thus, our former effectively secures a fast current transfer from HTS layers to the former at faults. For the thermal analysis, we suggested the liquid nitrogen flow as much as 50 liters/min at 72 K initial temperature before the fault. We considered two flow in the central channel together cases. First one is with flow outside the copper shield. The second case is with
1704
Fig. 4. Currents during a 1.5s fault with the initial amplitude
IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 19, NO. 3, JUNE 2009
30 kA.
Fig. 6. VCC at the DC mode. Left—for the cable, right recalculated per tapes with taking into account the real current distribution among layers. Ic 4:2 kA.
the recovery to the initial temperature is much slower when there is no inner bore in the former. It could be important for the real operational mode of the HTS power cable. The results of the fault analysis demonstrated that the selected former and each) provide the copper shied cross-sections ( proper protection of the experimental HTS power cable. IV. WITNESS SAMPLE TEST RESULTS
Fig. 5. Heating and recovery of the HTS power cable during a fault. The temperatures of the HTS layers, insulation, former and copper shield are shown for the end of the cable (LN outlet).
when no flow is in the central bore, i.e. no cooling from the inner channel, the cooling is possible outside of the copper shield only. Cross-section of the copper is considered the same in both cases. Rather ambiguous are the data about coefficient of heat conductivity of the paper insulations impregnated by coolant. For our estimations we used heat conductivity coeffi[7]. Then we took into account the heat cient transfer from the central spiral and corrugated tube causing the flow. Our estimations gave us the value of the turbulence of . Similar data heat transfer coefficient to were found in [8]. We calculated also the heating of the LIPA cable using the parameters above and our estimations coincided with those in [9]. The results of the numerical simulations demonstrated the importance of the inner cooling channel in the central former. In flow is going through the inner spite that 8% of total channel the cooling recovery time changes significantly. The calculated temperatures are shown in Fig. 5 for HTS layers, former, insulation and copper shield for two cases: with flow though the central cooling channel (“with central cooling”) and without it. One can see that while the temperature rise in layers during fault is small 5 K (still below 77 K and for this type of HTS) and practically the same for both cases,
The tests of witness samples are a good method to verify the long design before the full scale tests start. We tested a 5 m witness sample from the phase made of Sumitomo Electric HTS tapes. The sample from AMSC tapes has been tested before [1]. The witness sample has been cut from the long cable core after returning from insulating and shielding. Thus, the witness sample has undergone the entire technological route of the cable production. Before the test the 5 m sample was heavily instrumented as before [1]. Two sets of Rogowsky coils were installed on each layer. Many potential taps were attached to each layer along the length and to current leads. Hall sensors (HS) were placed to measure all magnetic field components (axial, radial and transversal) in different locations inside and outside of the cable. Thermocouples were attached under insulation directly to layers to measure a temperature rise [10]. All of the above instrumentation permitted us to collect a lot of the information about the sample behavior during all tests. Two sets of tests of the sample have been performed: at DC mode and AC mode. During all tests the current distribution among all layers was measured at the same time with voltage distribution along cable layers. The temperature and magnetic field were controlled at selected points. The DC tests results are shown in Figs. 6–8. The critical currents were determined from the voltage current characteristics criteria. The measured VCC are shown (VCC) with 1 in Fig. 6. Example of the current run and DC current distribution are shown in Fig. 7. One can see that the critical current is reached at the inner layer when total current is 4.2 kA. If to take into account the real current distribution (see Fig. 7, right) the current per tape in the layers is 100–105 A (Fig. 6, right). This is 100% of the average critical currents of tapes used. It confirms that the cabling technology used is properly developed to keep intact superconducting properties of tapes during entire technological route.
SYTNIKOV et al.: POWER CABLE DEVELOPMENT AND WITNESS SAMPLE TEST
Fig. 7. Example of a current run up to 5.5kA (left) and measured current distribution among layers at DC mode (right). Right graph: solid lines-currents in layers, dashed lines—current in a layer divided by the total current.
1705
2%. Unlike the DC mode, the AC mode current distribution is determined by inductances of layers that in turn are determined by layer diameters, twist pitches and directions. The twist pitches were calculated by use of the codes developed in VNIIKP based on theories [5], [6]. Then during cabling, all measures were taken to strictly keep the prescribed twist pitches. The test results showed that all was done properly and AC current distribution is quite uniform. The proper cable optimization was confirmed by the magnetic field measurements. The central axial magnetic field is very low, 1.5 mT per kA. It leads to low losses in formers during AC mode [10]. At AC mode we measured the current in the copper shield when it the shield current was about was short circuited. At 2 . This small current adds negligible AC loss values as 300 well. The results of the sample tests shown in Figs. 6–8 demonstrate that design methods and cabling technology are properly developed to provide the full use of superconducting properties of basic HTS tapes used for the 30 m cable. V. CONCLUSION The 30 m three phase (3 30 m) experimental HTS power cable with rated current of 2 kA and voltage of 20 kV has been produced and delivered to the test facility in the framework of the Russian R&D Program for HTS power devices. The numerical simulation of the current distribution between HTS layers and former and heating during a 30 kA, 1.5 s fault demonstrated the proper cable protection designed. The tests of the witness-sample confirmed the proper design and technology of the cable production. Critical current of the cable is equal to the sum of the critical current of tapes. In AC mode current distribution among layers is uniform. The test of the full length 3 30 m cable will start in coming months. The next step of the Program for R&D of HTS cables is development, production and delivery of a three phase 200 m cable with the same voltage and current ratings. This future cable will have a superconducting shield, similar cryostats and after preliminary test will be installed into the distribution grid in Moscow. REFERENCES
(top) Fig. 8. Two examples of AC current runs (left graphs) up to 3:6 kA (bottom) with measured current distribution among layers at and up to 3 kA AC mode (right graphs). In the right graphs currents in layers and currents in a layer divided by the total current are shown.
DC current distribution as shown in Fig. 7 is determined by the difference in joint resistances. In our case the maximum difference of currents in the inner and outer layers is below 10% and it reduced to 5% as the current rises. More than 20 runs with the current above 4.5 kA have been performed without any problems for the sample. Maximum current achieved was 6.5 kA, which is about 50% above the critical current. It demonstrates proper protection of superconducting layers by the former. In Fig. 8 examples of AC current ramps are shown (left graphs) and measured current distribution at AC mode (right graphs). One can see that at AC mode current distribution among layers is much more uniform. The difference is less than
[1] V. E. Sytnikov, V. S. Vysotsky, and A. V. Rychagov et al., IEEE Trans Appl Supercon., vol. 17, no. 2, pp. 1684–1687, Jun. 2007. [2] J. F. Maguire, J. Yuan, F. Schmidt, S. Bratt, and T. E. Welsh, “Installation and Testing Results of Long Island Transmission Level HTS Cable,” presented at the ASC 2008 Conf., Paper 1LB01, unpublished. [3] H. Yumura, Y. Ashibe, and H. Itoh et al., “Phase II of the Albany HTS Cable Project,” presented at the ASC 2008 Conf., The paper 1LB02 [Online]. Available: http://www.superpower-inc.com/pdf/, unpublished [4] D. W. A. Willen et al., IEEE Trans Appl Supercon., vol. 11, no. 2, pp. 2473–2476, Jun. 2001. [5] V. E. Sytnikov, P. I. Dolgosheev, and G. G. Svalov et al., Physica C, vol. 310, pp. 357–362, 1998. [6] V. E. Sytnikov, N. V. Polyakova, and V. S. Vysotsky, Physica C, vol. 401, pp. 47–56, 2004. [7] Heat Conductivity Coefficients of Different Materials [Online]. Available: http://www.ecoplast.ru/termo-index.html (in Russian) [8] Y. Terentyev, Y. M. Pavlov, I. V. Yakovlev, and V. I. Antipov, Cryogenics, vol. 32, pp. 279–282, 1992, (Proceedings ICEC-14). [9] J. F. Maguire, F. Schmidt, F. Hamber, and T. E. Welsh, IEEE Trans. Appl. Supercond., vol. 15, no. 2, pp. 1787–1790, Jun. 2005. [10] V. E. Sytnikov, K. A. Shutov, and V. S. Vysotsky et al., “AC Loss Analysis in the 5 m HTS Power Cables,” presented at the ASC 2008 Conf., Paper 1LPF01, unpublished.
