Partial Discharge Monitoring of Solid-State Transformer Insulation: HFCT Sensors, UHF Couplers, and Acoustic Emission

Partial discharge (PD) is the early-warning signal of insulation failure in high-voltage equipment. A PD event is a localized electrical discharge that does not bridge the insulation gap but does progressively degrade the dielectric material. Detecting PD activity before it escalates to full insulation breakdown is the foundation of modern transformer-asset-management practice. Solid-state transformer (SST) products inherit the PD-monitoring requirement from their iron-core ancestors but face additional engineering challenges because the SST’s converter switching creates electromagnetic emissions in the same frequency range that conventional PD detectors operate.

Three PD detection methods dominate commercial transformer monitoring. High-frequency current transformer (HFCT) sensors detect PD signals on the grounding conductor of the transformer through the high-frequency current pulses they create. Ultra-high-frequency (UHF) couplers detect electromagnetic emissions in the 300 megahertz to 3 gigahertz range that PD events emit. Acoustic emission (AE) sensors detect the ultrasonic pulses that PD events create in the insulating fluid or in the surrounding mechanical structure. Each method has different sensitivity characteristics, different placement requirements, and different cost profiles.

For SST applications, the UHF and acoustic methods are typically preferred over HFCT. The HFCT method is sensitive to the converter switching transients the SST creates by design, which can mask actual PD activity. The UHF method can be tuned to frequency bands away from the SST switching fundamental and harmonics. The acoustic method is fundamentally insensitive to electrical switching transients because it detects mechanical (sound-wave) phenomena rather than electrical phenomena. The trade-off is acoustic-method susceptibility to mechanical noise from cooling pumps, fans, and adjacent equipment.

The major commercial PD monitoring product vendors are OMICRON, Iris Power, FJ Innovation, Insulect, and Rugged Monitoring. Each ships online PD monitoring systems with continuous data acquisition, digital filtering, signal classification, and remote diagnostics through SCADA or Internet-of-Things platforms. The cost of an online PD monitoring installation for a single MV-SST product runs $20,000 to $60,000 depending on sensor count and the analytics package selected. For multi-megawatt SST installations the cost is small relative to the protected asset value.

The strategic implication is that the PD monitoring discipline is the most direct path to closing the multi-decade-reliability gap between commercial SST products and the iron-core distribution transformer benchmark. Whether SST manufacturers ship products with integrated PD monitoring as a standard feature, or whether utility customers install third-party PD monitoring after deployment, is an open commercial question. The integration-first approach is technically superior because the sensor placement can be optimized at manufacture, but adds product cost; the third-party approach preserves manufacturer cost competitiveness at the price of less-optimal sensor coverage.

Why It Matters

For asset managers, SSTs inherit the partial-discharge monitoring requirement from iron-core transformers but add converter-specific noise that complicates detection, so the diagnostic playbook cannot be copied over unchanged. Getting PD monitoring right on SSTs is what lets utilities apply the same condition-based maintenance discipline they rely on today; getting it wrong means flying blind on insulation health across a networked fleet.

Critical Perspective

Editorial correction: This post is part of MGRID’s Solid-State Transformer industry coverage. As of May 2026, that body of work systematically framed manufacturer announcements, funding rounds, and laboratory demonstrations as commercial deployments. The reality is that field-deployed commercial-class SST in revenue service globally is measured in single digits, and almost every product cited in this series is at “announced” or “funded” stage, not “operational.” Readers should treat the specific claims in this post against the standards documented in our [SST Industry Reality Check](/?p=9044) (the per-claim audit table maps marketing language to verifiable deployment status). The corrective article is the canonical reference for SST industry reality; this post remains published with its original framing so the editorial drift is traceable.

Related Coverage

• GE Vernova’s Quiet Solid-State Transformer Roadmap: From McMurray’s 1968 Patent to a 2027 Commercial Product: GE Vernova’s roadmap for solid-state transformers
• Wolfspeed’s 200mm Silicon Carbide Wafer Capacity Is the Single Biggest 2026 Bottleneck on Solid-State Transformer Production: Wolfspeed’s capacity bottleneck for SST production
• Amperesand Targets 30 Megawatts of Commercial Medium-Voltage Solid-State Transformer Shipments to Hyperscale AI Data Centers in 2026: Amperesand targets SST shipments to AI data centers
• How Moving AI Data Centers to 800 VDC Distribution and Solid-State Transformers Sidesteps Phase Imbalance Entirely: AI data centers moving to 800 VDC distribution and SSTs