Performance Evaluation of High-Voltage Cable Accessories under Accelerated Thermal and Electrical Aging

1. Introduction

Cable accessories are widely recognized as the weakest links in underground cable systems. During normal operation, they endure continuous electrical stress from the system voltage and thermal stress from Joule heating of the conductor. Under fault conditions or load fluctuations, these stresses become even more severe. Thermal aging causes polymer chain scission and crosslinking breakdown in insulation materials (e.g., cross-linked polyethylene – XLPE, ethylene-propylene rubber – EPDM), while electrical aging promotes partial discharge (PD) activity and electrical treeing. When applied simultaneously, thermal and electrical stresses interact in a non-linear manner, drastically reducing the service life of cable accessories.

2. Materials and Methods

2.1 Test Specimens

Premolded cable joints and heat-shrinkable terminations rated for 35 kV (U₀/U = 20/35 kV) were selected. The main insulation material was silicone rubber (SiR) for the joints and XLPE for the cable body. Each accessory was installed on a 1.5-meter-long XLPE insulated cable according to the manufacturer’s instructions to ensure consistent quality.

2.2 Accelerated Aging Setup

A custom-built aging chamber allowed simultaneous application of:

· Thermal stress: Heating tapes wrapped around the conductor at the accessory location, controlled to maintain conductor temperatures of 90°C, 110°C, and 130°C (representing normal, overload, and extreme conditions).

· Electrical stress: AC voltage at 1.5 × U₀ (30 kV) applied continuously between conductor and earth. This voltage level accelerates electrical tree initiation without immediate breakdown.

The ambient temperature inside the chamber was kept at 40°C to simulate a warm environment. A control group was aged with only thermal stress (110°C, no voltage) and another with only electrical stress (30 kV, ambient temperature 40°C) for comparison.

2.3 Testing Schedule and Measurements

Aging cycles lasted up to 2000 hours. Every 500 hours, specimens were removed (cooled to room temperature, voltage removed) and subjected to:

· Partial discharge inception voltage (PDIV) measured according to IEC 60270.

· Dielectric dissipation factor (tan δ) at 0.5 U₀ (10 kV) using a Schering bridge.

· Visual inspection for cracks, discoloration, or surface tracking.

· AC breakdown test (only on a separate set of identical specimens at each interval) using a ramp rate of 2 kV/s until failure.

3. Results and Discussion

3.1 Partial Discharge Inception Voltage (PDIV)

The PDIV of unaged specimens was approximately 24 kV. Under thermal-only aging at 110°C, PDIV decreased slowly, reaching 21.5 kV after 2000 hours (a 10.4% drop). Under electrical-only aging, PDIV dropped to 19.8 kV (17.5% drop). Under thermal-electrical co-aging at 110°C + 30 kV, PDIV fell dramatically to 15.2 kV after only 1500 hours, and further to 12.8 kV at 2000 hours (46.7% drop). This indicates a strong synergistic effect: the heat weakens the polymer’s resistance to electrical discharge, allowing PD to initiate and propagate at much lower voltages.

3.2 Dielectric Dissipation Factor (tan δ)

At 10 kV, the initial tan δ was 0.0035. Under co-aging conditions, tan δ increased to 0.0078 after 1000 hours and reached 0.021 after 2000 hours – nearly six times the initial value. The rise in tan δ suggests increased dipolar polarization and conduction losses due to the formation of carbonyl groups (from oxidation) and charge carriers released by partial discharges. In contrast, thermal-only aging produced tan δ of 0.0095 after 2000 hours, and electrical-only gave 0.012.

3.3 AC Breakdown Strength

Breakdown strength (BDS) was measured at 500-hour intervals on separate specimens. Initial BDS was 28 kV/mm. After 2000 hours:

· Thermal-only: 24.5 kV/mm (12.5% reduction)

· Electrical-only: 21.2 kV/mm (24.3% reduction)

· Thermal-electrical co-aging: 15.6 kV/mm (44.3% reduction)

The steep decline in BDS under co-aging can be attributed to the combined effects of thermal oxidation (creating micro-voids) and electrical treeing (forming conductive channels). SEM images of co-aged samples revealed large cavities and network-like cracks that were absent in single-stress aged samples.

3.4 Microscopic Analysis

FTIR spectra of co-aged SiR showed a significant increase in the absorption peak around 1710 cm⁻⊃1; (carbonyl group), indicating severe oxidative degradation. The peak intensity was 3.2 times higher than in thermal-only aged samples, confirming that the electric field promotes oxygen diffusion and reaction. SEM images showed that co-aging led to a rough, porous surface with numerous electrical tree channels, while thermal-only aging mainly produced a smoother degraded layer.

3.5 Failure Mode Analysis

After 2000 hours of co-aging, 3 out of 5 specimens failed during the final withstand test – two due to electrical tree breakdown at the stress cone interface, and one due to thermal runaway caused by increased tan δ. These failure modes are rarely observed in single-stress aging within the same timeframe, further proving the necessity of combined-stress evaluation.

4. Practical Implications for Asset Management

The results demonstrate that laboratory tests using only thermal or only electrical aging significantly underestimate the degradation rate of high-voltage cable accessories. For utilities and maintenance teams, this means:

· Lifetime predictions based on single-stress aging models (e.g., Arrhenius or inverse power law) are overly optimistic. A co-aging factor of 2–3 should be applied for realistic estimates.

· Condition monitoring should focus on PD activity during periods of high loading (when both thermal and electrical stresses peak). On-line PD sensors can detect early-stage degradation.

· Material selection for replacements in high-ambient-temperature or heavy-load corridors should prioritize formulations with superior thermo-electrical resistance, such as nano-filled silicone rubbers.

5. Conclusion

Accelerated thermal-electrical aging tests on 35 kV cable accessories reveal a clear synergistic degradation effect that cannot be reproduced by single-stress aging. Key findings include:

· PDIV decreased by nearly 47% after 2000 hours of co-aging at 110°C/30 kV, compared to less than 18% under electrical-only aging.

· tan δ increased six-fold, indicating severe dielectric loss and oxidation.

· Breakdown strength dropped by 44%, mainly due to combined micro-void formation and electrical treeing.

· Chemical analysis confirmed that the electric field accelerates oxidation reactions initiated by heat.