Insulation Design and Reliability Evaluation of ±80kV HVDC XLPE Cables

• Jung, Chae-Kyun (Power Transmission Laboratory, KEPCO Research Institute) ;
• Park, Hung-Sok (Power Transmission Laboratory, KEPCO Research Institute) ;
• Kang, Ji-Won (Power Transmission Laboratory, KEPCO Research Institute)
• Accepted : 2014.02.23
• Published : 2014.05.01

Abstract

This paper describes insulation design and its reliability evaluation of ${\pm}80kV$ HVDC XLPE cable. Recently, the construction of HVDC transmission system, which is combined overhead line with underground cable, has been completed. This system is installed with existing 154 kV AC transmission line on the same tower. In this paper, the lightning transient analysis is firstly reviewed for selection of basic impulse insulation level and nominal insulation thickness. Then the electrical performance tests including load cycle test and superimposed impulse test based on CIGRE TB 496 are performed to evaluate the reliability of newly designed HVDC cable. There is no breakdown for ${\pm}80kV$ HVDC XLPE cable during electrical performance test. Finally, this system is installed in Jeju island based on successful electrical performance test (Type test). After installation tests are also successfully completed.

1. Introduction

HVDC is a preferred technology for transmitting large amount of power across long distance. HVDC results in overall higher efficiency and reliability than AC system transmitting the same amount of power. One of the biggest advantages of HVDC transmission is efficiency. HVDC transmission can supply more electricity with less power losses than AC system.

In Korea, the construction of hybrid transmission system has been completed. This system consists of existing 154 kV AC transmission line and newly installed ±80kV HVDC transmission line on the same tower. HVDC transmission line is combined overhead line and underground cable.

In previous papers [1-3], the lightning analysis and basic impulse insulation level were investigated including surge arrester type and capacity for ±80kV HVDC transmission line. Specially, the new DC cable with nano-composite XLPE insulation was developed to improve improve the insulation efficiency as well as to decrease the impact on the space charge.

In this paper, firstly, the basic impulse level and insulation thickness are reviewed in detail based on the transient analysis results for ±80kV HVDC cables. Then, the performance tests including load cycle test and superimposed impulse test were performed based on CIGRE TB 496 [4] for evaluating the reliability of newly designed HVDC cable. Finally, the qualified cable was installed in HVDC test area in Jeju. The test after installation was also successfully completed in 2012. The final commissioning test of converter station is now in progress.

2. ±80kV HVDC Transmission Line

± 80 kV HVDC transmission line in the hybrid transmission system is connected from Hanlim converter station to Geumak converter station in Juju island. The length of overhead DC line section with ACCC 160 mm2 is 4 km, underground cable section with nano-composite XLPE insulation is 0.5 km. Fig. 1 shows the diagram of power system in Jeju island. In this figure, the red line means ±80kV HVDC transmission line.

Fig. 1.Power system diagram of Jeju island

Fig. 2 shows the picture and diagram of ±80kV HVDC XLPE cable. This cable consists of conductor (①), inner semiconducting layer (②), nano-composite XLPE insulation (③), outer semiconducting layer (④), metal sheath (⑤) and oversheath (⑥)

Fig. 2.±80kV HVDC XLPE cable

Fig. 3 shows the detailed configuration of the system. As shown in this figure, the system consists of a 154 kV AC transmission line and a ±80kV HVDC transmission line on the same tower. In addition, the HVDC transmission line is connected to underground cables.

Fig. 3.Detailed system configuration

3. Basic Impulse Insulation Level

In previous papers [2-3], the surge arrester for ±80kV HVDC transmission line has been already selected by various lightning overvoltage analysis. The required maximum continuous operation voltage (MCOV) is 59.4 kVrms. Therefore, the surge arrester with the rated voltage of 78 kVrms and MCOV of 63.1 kVrms is selected for ±80kV HVDC transmission line. The nominal discharge current is also selected by10 kA based on IEC 60099-4[5]. These kinds of surge arrester have been installed at cable head (joint between overhead lines and underground cables) and converter station inlet.

In this paper, EMTP simulation is used for lightning analysis and insulation design of ±80kV HVDC cables. The EMTP model is illustrated in Fig. 4. Four story tower model and back flashover model are used for lightning analysis. The arching horns modeled by TACS and MODELS are also applied in this simulation model. The applied lightning current and waveform are 80 kA and 2/70 ㎲. This lightning surge strikes on grounding wire of HVDC overhead transmission line section at 1 km from cable head.

Fig. 4.Captured EMTP Simulation

The lightning overvoltages according to surge arrester installation are compared in Table 1. The lightning overvoltage with surge arrester is significantly decreased from 647 kV to 182 kV. At this moment, the surge arrester does not exceed the nominal discharge current of 10 kA. The discharge current measured at the positive pole of cable head is 8.99 kA. Fig. 5 shows the lightning overvoltage with surge arrester in cable head as well as converter station inlet.

Table 1.( ) : discharge current of surge arrester

Fig. 5.Lightning overvoltage with surge arrester

The basic impulse insulation level is calculated based on EMTP simulation result and several standards of IEC 60071-1 [6], IEC 60071-5 [7] and CIGRE TB 496 [4].

In this paper, DC voltage source is not considered for EMTP simulation because the calculation error is occurred by complicated impedance matrix. Therefore, for superimposed DC source, the Bahder’s coefficient is applied as expressed in Eq. (1) [6-8].

where, K is Bahder’s coefficient. It generally has a coefficient 0

where, Uc is calculated lightning overvoltage when the DC rated voltage is not considered for simulation. Uk is the overvoltage considering DC source superimposition. As shown in Table 2, the overvoltages considering superimposition of opposite DC rated voltage are 222 kV and 219 kV, respectively. They are 40 kV higher than the overvoltage without DC superimposition. Therefore, the maximum overvoltage by DC source superimposition of 222 kV can be selected to coordinatedly withstand overvoltage for the basic impulse insulation level of ± 80 kV nano-composite HVDC XLPE cable.

Table 2.Lightning overvoltage considering DC superimposition

Then, the safety margin of 20 % considering accuracy of EMTP model and other uncertain factors based on IEC 60071-5 [7] is applied for the calculation of the basic impulse insulation level. Finally, the standard basic impulse insulation level for ± 80 kV nano-composite HVDC XLPE cable can be selected by IEC 60071-1 [6].

The maximum lightning overvoltage considering safety margin of 20% is 266.4 kV. From this result, the standard basic impulse insulation level of 325 kV for ± 80 kV nanocomposite HVDC XLPE cable can be finally selected.

4. Insulation thickness

The insulation thickness for ± 80 kV nano-composite HVDC XLPE cable is calculated considering impulse withstand voltage (Uimp) based on the standard basic impulse protection level (Ubil) of 325 kV and design stress for impulse voltage (ELimp). Eq. (3) shows the calculation for impulse withstand voltage and the minimum insulation thickness can be calculated by Eq. (4).

where Uimp is calculated impulse withstand voltage (kV), Ubil is standard basic impulse protection level, A is the tolerance of arrester protection level (recommended by CIGRE TB 189), K is Bahder’s coefficient, U0 is the DC rated voltage, K1 is the repetition deterioration coefficient, K2 is the temperature coefficient, and K3 is the overall safety factor for uncertainties.

where t is minimum insulation thickness (mm) and ELimp is the design stress for impulse voltage.

The calculated impulse withstand voltage (Uimp) of 411.13 kV can be calculated by Eq. (3). Finally, the minimum insulation thickness of 10.8 mm is also calculated by Eq. (4). Therefore, the nominal insulation thickness of 12 mm is selected in this paper for ± 80 kV nano-composite HVDC XLPE cable.

5. Electrical Performance Test

The electrical performance test (Type test) should be performed on test cable based on CIGRE TB 496 for evaluating the reliability of newly designed ±80kV HVDC XLPE cable [4]. Electrical performance tests are made before supplying on a general commercial basis of cable system in order to demonstrate satisfactory performance to meet the intended application. Once accessory with respect to materials, manufacturing process, design electrical stress levels, which might adversely change the performance characteristics [9].

Fig. 6 shows the flow chart for electrical performance test of ±80kV HVDC XLPE cable. The order of electrical performance test is 1st load cycle test, polarity reversal test, 2nd load cycle test, superimposed impulse voltage test including switching and lightning impulse, and DC test. Fig. 7 shows the system diagram for electrical performance test. The test is performed in Underground Test Lab in KEPCO Power Testing Center. Fig. 8 shows the real picture of installed test cable for performance test.

Fig. 6.Flow chart for electrical type test

Fig. 7.System diagram for electrical type test

Fig. 8.Real picture of installed test cable for type test

The conditions for load cycle tests including 1st 24-hour load cycles, polarity reversal cycles and 2nd 48-hour load cycles are as follows:

- Eight 24-hour load cycles at negative polarity at 1.85·U0 (148 kV) - Eight 24-hour load cycles at positive polarity at 1.85·U0 (148 kV) - Eight 24-hour load cycles with polarity reversal cycles at 1.45·U0 (116 kV) - Three 48-hour load cycles at positive polarity at 1.85·U0 (148 kV)

24-hour load cycles consist of an 8 hours heating period and a 16 hours cooling period. During the last 2 hours of heating, the conductor temperature shall be at 70℃ which is suggested by manufacturer. 48-hour load cycles consist of a 24 hours heating period and a 24 hours cooling period. During the last 18 hours of heating period the conductor temperature should be more than 70℃. 48-hour load cycles are only required as part of the test procedure to ensure that electrical stress inversion is well advanced within the cycle. For a polarity reversal, starting with positive voltage, the voltage polarity shall be reversed three times every “24 hours” load cycle (evenly distributed) and one reversal shall coincide with the cessation of loading current in every “24 hours” load cycle. The recommended time duration for a polarity reversal is within 2 minutes.

During all load cycle tests, there are no breakdown of ± 80 kV HVDC XLPE cable and interruption of testing. The load cycle test is successfully completed. Figs. 9 and Fig. 10show the results of polarity reversal cycle test and 2nd 48 hour heat cycle test.

Fig. 9.Polarity reversal cycle test results

Fig. 10.2nd 48 hour heat cycle test results

Then, the superimposed impulse voltage tests are performed. The switching surge withstand test and lightning impulse withstand test are performed on ± 80 kV HVDC XLPE cable. During the impulse tests, conductor temperature of more than 70℃ shall be reached for a minimum of 10 hours before the voltage impulses are applied and shall be maintained throughout the duration of the test. Fig. 11 shows the test circuit diagram for superimposed impulse test. The voltage and temperature maintained for superimposed impulse test are shown in Fig. 12. Also, the conditions for switching and lightning impulse test are as follows:

Fig. 11.Circuit diagram for superimposed impulse test

Fig. 12Maintaining voltage and temperature for superimposed impulse test

- Switching impulse test (10 times each) ·Test objective at Uo, 10 consecutive impulses to -144 kV±3% ·Test objective at -Uo, 10 consecutive impulses to 144 kV±3% ·Waveform (front/tail) : 200-300㎲ / 1000-4000㎲ - Lightning impulse test (10 times each) ·Test objective at Uo, 10 consecutive impulses to -325 kV±3% ·Test objective at -Uo, 10 consecutive impulses to 325 kV±3% ·Waveform (front/tail) : 1-5 ㎲ / 40-60㎲

During superimposed impulse tests, there are no breakdown of ±80kV HVDC XLPE cable and interruption of testing. The superimposed impulse test is successfully completed. Figs. 13 and Fig. 14 show the waveforms of the first and last negative switching impulse test and positive lightning impulse, respectively.

Fig. 13.The first and last negative superimposed switching impulse waveform

Fig. 14.The first and last positive superimposed lightning impulse waveform

± 80 kV HVDC XLPE cable is installed in Jeju island based on successful electrical performance test including load cycle tests and superimposed impulse voltage test. The construction has been already completed. After installation tests including DC withstand voltage test, nondestructive test and TDR (Time Domain Reflectrometer) test are also successfully completed. Fig. 15 shows a DC withstand voltage test of real HVDC cables in Jeju. Fig. 16 shows the TDR test results during after installation test. As shown in this figure, the inserted signal from cable head is reflected back from termination of converter station inlet. There is no abnormal signal between the first and the second reflections. The time difference between these two reflections is 5.37㎲. Therefore, the propagation velocity of real HVDC cables can be calculated as 186.2 m/㎲. This velocity can be applied for fault location when the fault occurs on this HVDC cable.

Fig. 15.DC withstand voltage test in real HVDC cable system

Fig. 16.TDR test result during after installation test

6. Conclusions

In this paper, the insulation design and its reliability evaluation of newly developed nano-composite ± 80kV HVDC XLPE cable is reviewed. The results are summarized as follows;

1) The maximum overvoltage including safety margin and DC superimposition is 266.4kV. The basic impulse insulation level of 325 kV can be finally selected based on IEC 60071-1. 2) The calculated minimum insulation thickness is 10.8 mm. The nominal insulation thickness of 12 mm is selected for ± 80 kV nano-composite HVDC XLPE cable. 3) The electrical performance test is performed based on CIGRE TB 496 for evaluating the reliability of newly designed ±80kV HVDC XLPE cable. 4) During all load cycle tests including 1st and 2nd load cycle and polarity reversal, there are no breakdown of ± 80 kV HVDC XLPE cable and interruption of testing. 5) During all superimposed impulse tests including switching and lightning impulse tests, there are no breakdown of ±80kV HVDC XLPE cable and interruption of testing. 6) ±80kV HVDC XLPE cable is installed in Jeju island based on successful electrical performance test. The construction and after installation tests including DC withstand voltage test, non-destructive test and TDR are also successfully completed.

References

1. EPRI Technical Report, "AC to DC Power Testing Line Conversion", 2009. 12.
2. C. K. Jung, J. W. Kang, K. Y. Shin, D. H. Kim and D. I. Lee, "Lightning Analysis on HVDC/HVAC Hybrid Transmission System in Korea", 2011 ISH Conference proceedings, 2011.8.
3. C. K. Jung, H. S. Park, J. W. Kang, D. H. Kim, J. W. Shim, B. S. Moon and D. I. Lee, "Basic Impulse Insulation Level Review of ${\pm}80kV$ Nano-composite DC XLPE Cables", CIGRE Colloquium on HVDC and Power Electronic Systems, 2012. 3.
4. Electra 496, "Recommendation for testing DC extruded cable systems for power transmission at a rated voltage up to 500 kV", CIGRE WG B1.32, 2012. 4.
5. IEC 60099-4, "Metal-Oxide Surge Arresters without Gaps for AC Systems", 2009.
6. IEC 60071-1, "Insulation coordination-Part 1: Definition, Principles and rules", 2006. 1.
7. IEC 60071-5, "Insulation coordination-Part 5: Procedures for high-voltage direct current(HVDC)", 2002. 6.
8. Electra 86, Overvoltage on HVDC Cables, CIGRE J/W Group 33/21/14.16, 1994. 8.
9. B. SANDEN, et al., "Recommendations for testing HVDC extruded cable systems for power transmission at a rated voltage up to 500 kV", CIGRE Colloquium on HVDC and Power Electronic Systems, 2012. 3.