A Novel Method For Corona Discharge Detection Under Low-Pressure Conditions Using Laser Speckle Pattern

Abstract

Monitoring the insulation condition is essential for the efficient functioning of any electrical apparatus. Deterioration of the insulation is predominantly caused by various operational stresses at harsh environmental conditions that lead to electrical discharges. Corona discharge is a common and crucial phenomenon in electrical apparatus. Corona discharge detection is considered an essential predictive maintenance method that ensures the sound operation of electrical equipment. Discharge activities significantly depend on pressure conditions and pose a risk to electrical equipment working under low pressure. This paper proposes an acousto-optic effect-based novel, noninvasive method of corona detection where corona discharge is emulated in a laboratory-based setup under low-pressure conditions and detected using the proposed and conventional IEC 60270-based methods simultaneously. It is observed that the increment in the discharge magnitude and repetition rate, as suggested by the phase-resolved discharge analysis obtained from the conventional method, is correlated with the change in the shape of the speckle pattern and the intensity of the speckle grain. Hence, the above methodology can be used as a new tool for analyzing the discharges of electrical power apparatus, especially installed in low-pressure environments such as aircraft operating in high altitudes. Later, for simplification and real-time applications, the above-mentioned experiment is conducted using an optical fiber-based model where two optical fibers forming a fiber-to-fiber coupling through air are developed, and the modulation of the coupled light by the corona discharge-generated acoustic wave is observed. The outcomes establish that the proposed method detects corona discharge reliably, and the sensitivity of detection is comparable with the standard IEC 60270-based method, with a noninvasive, cost-effective, and all-dielectric approach.