7700104
IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 22, NO. 3, JUNE 2012
Electrical Insulation Characteristics of PPLP as a HTS DC Cable Dielectric and GFRP as Insulating Material for Terminations S. H. Kim, J. H. Choi, W. J. Kim, K. L. Kim, H. G. Lee, Y. S. Kim, H. M. Jang, and S. K. Lee, Member, IEEE
Abstract—A high-temperature superconducting (HTS) DC cable system has attracted a great deal of interest from the view point of low loss, dense structure and large capacity compared to HTS AC cable system. It has a HTS cable and a termination. A HTS DC cable system consists of a conductor, cooling system and electrical insulation. To realize the HTS DC cable system, it is important to study not only high current capacity and low loss of conductor but also optimum electrical insulation at cryogenic temperature. The electrical insulation technology of HTS DC cable and termination must be solved for the long life, reliability and compact of cable. In this paper, we will discuss mainly on the electrical insulation characteristics and the insulation design of 220 kV class HTS DC cable. Voltage-time (V-t) characteristics of laminated polypropylene paper (PPLP) in were studied. Furthermore, the surface flashover characteristics of glass fiber reinforced plastic (GFRP) for termination insulators under DC and lightning impulse voltage were studied. Index Terms—Cable insulation, dielectric breakdown, dielectric materials, superconducting devices.
I. INTRODUCTION
R
ECENTLY, a HTS cable system has attracted a great deal of interest from the view point of energy. It has HTS cable and termination. HTS AC cable system has AC loss, but HTS DC cable has perfectly zero resistance. HTS DC cable has advantages such as the ultimately lower loss, more compact dimensions, and large capacity compared to AC cable. So far, many papers have been published on the HTS AC cable system [1], [2]. However, the research and development of HTS DC cable system is rather few. A HTS DC cable system consists of a conductor, cooling system and electrical insulation. Especially, the electrical insulation technology of HTS DC cable must be solved for the long
Manuscript received September 09, 2011; accepted December 15, 2011. Date of publication December 23, 2011; date of current version May 24, 2012. This work was supported in part by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (2011-0013667). S. H. Kim, J. H. Choi, and W. J. Kim are with the Department of Electrical Engineering, Gyeongsang National University and Engineering Research Institute, Jinju 660-701, Korea (e-mail:
[email protected]). K. L. Kim and H. G. Lee are with the Department of Materials Science and Engineering, Korea University, Seoul 136-701, Korea (e-mail:
[email protected]). Y. S. Kim is with the Electrical Safety Research Institute, Korea Electrical Safety Corporation, Gapyeong 477-814, Korea (e-mail:
[email protected]). H. M. Jang and S. K. Lee are with the Advanced R&D Center, LS Cable & System Ltd., Gumi 730-030, Korea (e-mail:
[email protected]). 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.2011.2181470
Fig. 1. A manufactured mini-model DC cable.
life, reliability and compact of cable. A composite insulation system of nitrogen coolant and polymer films has been developed as an electrical insulation system of HTS cable. This type covers the conductor with multi-layered thin insulation tapes to prevent contraction due to cooling and thermal loss. PPLP has been widely used as an insulating material for AC HTS cable and has already been applied to HTS cable insulation. Although have not the DC breakdown characteristics of PPLP in been revealed satisfactorily until now, which make the insulation design of cable very difficult. The insulating structure of HTS cable termination is consisted , and . There is variety of a four layers of air, of thermal, mechanical and electrical problems to be solved for such purpose. The insulation characteristics of insulators for termination have not revealed satisfactorily until now, which make the design of electrical insulation of HTS cable termination very difficult [3]. In practical termination, the surface flashover characteristics of solid insulator at cryogenic temperature are the most serious problem in insulation. Therefore, it is important to understand the DC surface flashover characteristics of solid at first stage. insulators in In this paper, we will discuss mainly on the electrical insulation characteristics and the insulation design of 220 kV class HTS DC cable. The breakdown characteristics of PPLP mini-model cable under DC and lightning impulse voltage pressure, V-t characteristics, were studied as functions of polarity effect. Furthermore, the surface flashover characteristics of GFRP for termination insulators under DC and lightning impulse voltage were studied. II. EXPERIMENTAL APPARATUS AND METHOD HTS mini-model DC cable shown in Fig. 1 was manufactured. The samples used as cable insulator was PPLP. Figs. 2(a) and 2(b) show the electrode systems of the mini-model DC cable and sheet sample for breakdown test. The structure of mini-model DC cable is almost the same as that of mini-model AC cable [4]. The mini-model cable becomes 400 mm in length and 1 mm in thickness. And effective length of cable is 50 mm. At the end of cable, the stress cone made PPLP prevent surface flashover on the cable terminal.
1051-8223/$26.00 © 2011 IEEE
KIM et al.: PPLP AS HTS DC CABLE DIELECTRIC AND GFRP AS INSULATING MATERIAL FOR TERMINATIONS
Fig. 2. Electrode systems for breakdown test. (a) Mini-model cable; (b) sheet sample.
7700104
Fig. 4. Weibull distribution of DC breakdown strength as a function of the number of breakdown events.
Fig. 3. Electrode system for surface discharge. Fig. 5. Dependence of DC breakdown voltage on pressure of
A sheet sample has upper hole and three multi-layers. The hole of butt-gap is 2 mm in diameter. It was made artificially at the center of sample. PPLP sheet was fixed with GFRP sample holder. The electrodes made from SUS were the plane electrode of upper 25 mm and lower 75 mm. The sample used as termination insulator was GFRP. Fig. 3 shows the electrode system for surface flashover. A non-uniform electrode system with a stainless steel triangle (tip radius: 60 degree) and planar configuration. The surface flashover distances were 5, 10 and 15 mm, respectively. The samples with electrodes were dried by using the drier to completely remove the moisture, and cleaned carefully with ethyl alcohol. The sample with electrodes was fixed with GFRP sample holder. The cryostat is made of SUS and of a double structure with a vacuum layer and multi layer insulation. The sample is attached to the lower part of the flange bushing. After introducing the commercial-grade , it was cooled with enough time. And it was pressed with evaporated nitrogen of 0.1–0.4 MPa. A DC and impulse voltage was applied between the electrodes of sample. The DC power supplier used was 100 kV maximum voltage. A ramp DC voltage of about 2 kV/s is applied manually. The lightning impulse power supplier used was 1.2/50 standard waveform and 400 kV maximum voltage. Impulse voltage of 70% of breakdown is initially applied, and then a series of step inputs of 4 kV is applied. The sample was replaced with new one after each breakdown test. In the case of breakdown and surface flashover test, 10–20 times voltages were measured at each given condition. The maximum probability voltage is obtained from 0.1% value of Weibull distribution. Also, the V-t characteristics of mini-model cable under 0.3 MPa were measured.
.
III. RESULTS AND DISCUSSION One of the prominent characteristics of DC breakdown voltage is its the maximum breakdown strength of Weibull probability and polarity effect. Fig. 4 shows the DC breakdown strength of PPLP sheets under atmospheric as a function of the number of breakdown events. When a negative or positive voltage is applied to the upper electrode (25 mm), we term this “negative polarity” and “positive polarity”, respectively. The number of breakdown events was 10, 20 and 30 times. As evident from this figure, the positive polarity indicates lower than the negative polarity. Namely, is 121–122 kV/mm at positive polarity and 129–131 kV/mm at negative polarity, respectively. However, this experimental result is higher than that reported by Tsuyuki et al. [5]. It was confirmed that PPLP is excellent insulator for HTS DC cable as well as HTS AC cable. Also, P-value of Weibull probability is large (above 0.05). Despite of more than 20 times events, is relatively insensitive to the breakdown events. Therefore, we performed 20 times of breakdown number, and we used the breakdown strength of 0.1% Weibull probability. Fig. 5 shows the pressure dependence of DC breakdown voltage of PPLP sheets. As seen in the figure, breakdown voltage of negative polarity is higher than that of positive polarity. Also, the breakdown voltage increases slightly with increasing the pressure, tending to saturate at about 0.3 MPa. The saturating pressure in DC is similar to that of AC [6]. This indicates that the bubble formation and partial discharge in has a strong influence on the breakdown voltage as has already been reported [7]. Judging from these results, we studied the
7700104
IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 22, NO. 3, JUNE 2012
on the polarity of DC voltage. V-t characteristics are generally expressed as the voltage-time equation (V: voltage, n: life time index). From above experiment, the relation between applied voltage and the time to breakdown can be expressed by (1); (1)
Fig. 6. Weibull probability plot on DC and impulse breakdown voltage of minimodel cable under 0.3 MPa. (a) DC voltage; (b) impulse voltage.
Thus, n value under DC voltage is estimated to be 500 at positive polarity. Considering the design margin, n value of DC can be obtained as 400. However, the measurement of n value under DC voltage in is quite few. Several papers have been published on the n value under AC voltage [8]. The n value of insulating materials such as PPLP, Kapton, aromatic polyamide (Nomex), polyethylene (PE), and Kraft paper in is determined in a range of 14–100. The n value of DC voltage is it means that the degradation of PPLP is relatively insensitive to the DC voltage compared to the case for AC voltage. Therefore, we adopted as degradation coefficient of HTS DC cable. The dependence of DC breakdown voltage on pressure of is similar for that of AC. However, of DC is remarkably higher than that of AC. Especially, the degradation rate of PPLP under DC voltage is remarkably slow. In conclusion, the breakdown characteristics of DC seem to be better than that of AC. Since the breakdown voltage of negative polarity was higher than that of positive polarity, the insulation design was determined to the breakdown voltage of the positive polarity. Generally, the required insulation thickness of AC cable was basically derived by dividing the design stress of the design voltage. However, the study of the required insulation thickness of DC cable is quite few. In this experiment, it was used the design method reported in CV cable [9]. The required lightning impulse voltage can be calculated by (2); (2)
Fig. 7. V-t characteristics of mini-model cable under 0.3 MPa.
breakdown voltage under 0.3 MPa for the insulation design of HTS DC cable. Fig. 6 shows the Weibull probability plot on DC and impulse breakdown voltage of mini-model cable under pressure of 0.3 MPa. In the case of DC voltage, has been obtained as 120 kV/mm at positive polarity and 130 kV/mm at negative polarity, respectively. Also, in the case of impulse, has been obtained as 112 kV/mm at positive polarity and 136 kV/mm at negative polarity, respectively. For the mini model cable using 125 PPLP, of AC and negative impulse is reported as 47.5 kV/mm [10]. of DC is remarkably higher than that of AC. The V-t characteristics of mini-model cable under pressure of 0.3 MPa are shown in Fig. 7. It is seen that the time to breakdown decreases as the applied voltage increases. And the life time index n of V-t characteristics is slightly dependent
represents the operation voltage, for tolerWhere, ance (1.1), for abnormal overvoltage coefficient included Bahder’s coefficient (2.5). Based on (2), the required lightning impulse voltage of HTS DC cable is calculated as 605 kV. Also, the design stress can be calculated by (3); (3) Where, represents maximum impulse breakdown strength, for design margin (1.2). Based on (3), the required lightning impulse strength of HTS DC cable is calculated as 93 kV/mm. The insulation thickness can be calculated by (4) and (5); (4) (5) Where represents impulse design stress in r, for impulse design voltage in r, for inner radius of the insulation layer and for outer radius of the insulation layer. Also,
KIM et al.: PPLP AS HTS DC CABLE DIELECTRIC AND GFRP AS INSULATING MATERIAL FOR TERMINATIONS
TABLE I DESIGN STRESS AND INSULATION THICKNESS
7700104
of negative/positive voltage under DC voltage is 1.03. Also, the impulse surface flashover voltage quickly increases with increasing of the electrode length. The negative surface flashover voltage in is nearly equal. The ratio of negative/positive voltage under impulse voltage is 1.1. IV. CONCLUSION We have studied on the electrical insulation characteristics of PPLP and the insulation design of the 220 kV class HTS DC cable was performed. Also, the surface flashover characteristics of GFRP under DC and lightning impulse voltage were studied. The results may be summarized as follows: 1) It was confirmed that PPLP has more excellent electrical properties for HTS DC cable as well as HTS AC cable. 2) The breakdown voltage increases slightly with increasing the pressure, tending to saturate at about 0.3 MPa. The saturating pressure in DC is similar to that of AC. 3) In the case of positive polarity, has been obtained as 120 kV/mm under DC voltage and 112 kV/mm under impulse voltage. of DC is remarkably higher than that of AC. 4) The degradation of PPLP in is relatively insensitive to the DC voltage compared to the case for AC voltage. The degradation coefficient of DC cable is used as 1.03. In conclusion, the breakdown characteristics of DC seem to be better than that of AC. 5) The insulation thickness of the 220 kV class HTS DC cable was calculated for 9 mm. 6) The DC surface flashover voltage of negative polarity is slightly higher than that of positive polarity in GFRP.
Fig. 8. Surface flashover characteristics of GFRP. (a) DC voltage; (b) impulse voltage.
represents DC design stress in r, for DC design voltage in r. In the case of , the total insulation thickness under impulse voltage was determined to be 9 mm. On the other hand, considering a degradation coefficient (1.03) and the DC design stress (96 kV/mm), the total insulation thickness under DC voltage was determined to be 3 mm. In conclusion, the total insulation thickness of 9 mm in 220 kV class HTS DC cable was adopted. The results of design stress and required insulation are shown in Table I. In order to development the HTS termination, it is necessary to study the surface flashover characteristics of GFRP in at first stage. Fig. 8 shows the surface flashover voltage versus electrode length of GFRP under DC and impulse voltage, respectively. The surface flashover voltage is measured at the atmospheric . As shown in this figure, the DC surface flashover voltage linearly increases with increasing of electrode length. The DC surface flashover voltage of negative polarity is slightly higher than that of positive polarity in . The ratio
REFERENCES [1] J. Cho, J. H. Bae, H. J. Kim, K. D. Sim, S. Kim, H. M. J, C. Y. Lee, and D. W. Kim, “Development of a single-phase 30 m HTS power cable,” Cryogenics, vol. 46, pp. 333–337, 2006. [2] J. F. Maguire, J. Yuan, W. Romanosky, F. Schmidt, R. Soika, S. Bratt, F. Durand, C. King, J. McNamara, and T. E. Welsh, “Progress and status of a 2G HTS power cable to be installed in the Long Island Power Authority (LIPA) grid,” IEEE Trans. Appl. Supercond., vol. 21, no. 3, pp. 961–966, June 2011. [3] T. Shimonosono, S. Nagaya, T. Masuda, and S. Isojima, “Development of a termination for the 77 kV-class high Tc superconducting power cable,” IEEE Trans. Power Delivery, vol. 12, no. 1, pp. 33–38, Jan. 1997. [4] J. W. Choi, D. S. Kwag, J. H. Choi, H. J. Kim, J. W. Cho, and S. H. Kim, “Electrical insulation design of a 154 kV-class HTS power cable,” Superconductivity and Cryogenics of Korea, vol. 11, no. 2, pp. 25–28, 2009. [5] K. Tsuyuki, S. Washida, O. Tanda, T. Masuda, K. Kato, T. Nakajima, and S. Mukoyama, “Study on the electrical insulation of high-Tc superconducting DC power cable,” Cryogenic Engineering of Japan, vol. 35, no. 7, pp. 350–356, 2000. [6] H. J. Kim, D. S. Kwag, J. W. Cho, K. C. Seong, K. D. Sim, and S. H. superKim, “Insulation studies and experimental results for high conducting power cable,” IEEE Trans. Appl. Supercond., vol. 15, no. 2, pp. 1727–1730, June 2005. [7] M. Hara and H. Okubo, “Electrical insulation characteristics of superconducting power apparatus,” Cryogenics, vol. 38, no. 11, pp. 1083–1093, 1998. [8] M. O. Pace, I. Sauers, D. R. James, and A. R. Ellis, “Aging of lapped tape insulated model cable at cryogenic temperature,” in 2002 Annual Report Conference on Electrical Insulation and Dielectric Phenomena, IEEE, 2002, pp. 27–30. [9] K. Terashima, Y. Murata, H. Muto, T. Uozumi, and M. Yoshida, “Study of design method of XLPE cable for DC voltage,” T. IEE Japan, vol. 119-B, no. 2, pp. 212–222, 1999. [10] J. W. Choi, H. G. Cheon, J. H. Choi, H. J. Kim, J. W. Cho, and S. H. Kim, “A study on insulation characteristics of laminated polypropylene paper for an HTS cable,” IEEE Trans. Appl. Supercond., vol. 20, no. 3, pp. 1280–1283, June 2010.
