SFRA and the Operator’s Diagnostic Gap

Introduction

Transformer fleet operators rely on a set of established diagnostic tools to assess mechanical and electrical health: dissolved gas analysis, insulation resistance and power factor testing, and sweep frequency response analysis, among others. Each test targets a specific failure mode and each has a specific operating constraint. This paper focuses on SFRA specifically, examines the two structural limitations that govern how it is used in practice, and describes how continuous multi-signal condition monitoring addresses those limitations without displacing SFRA's diagnostic role.

The intent of this paper is not to argue that continuous monitoring and SFRA should be run side by side by the same fleet operator. It is to show that the two methods serve different actors in a transformer's lifecycle, answer different kinds of questions, and that the transformer operator's ongoing diagnostic need is met by continuous monitoring, not by SFRA.

SFRA: Technical Overview

SFRA measures the transfer function of a transformer winding across a swept frequency range, typically from 20 Hz to 2 MHz. A low-voltage signal is injected at one terminal and the response is measured at another, producing a signature curve determined by the winding's distributed resistance, inductance, and capacitance network. Because that network is a direct function of physical geometry, turn spacing, and dielectric properties, any change in winding shape or position alters the signature.

Different frequency bands carry different diagnostic sensitivity, summarized below.

SFRA is governed by IEEE C57.149 and IEC 60076-18, with foundational work established by CIGRE working group A2.26. It is well validated against known mechanical fault types, including winding buckling, axial displacement, core displacement, and shorted or open turns.

The Comparative Nature of SFRA Results

SFRA has no absolute fault threshold. A signature curve, on its own, does not indicate whether a winding is healthy or damaged. Diagnostic conclusions require one of three comparisons.

• Time-based comparison, against a signature taken when the unit was known to be healthy.
  
• Type-based comparison, against a sister or identical unit.
• Phase-based comparison, across the three phases of the same transformer.

Each approach depends on the availability and quality of a comparison point. A unit commissioned without a baseline signature, or a unit with no true sister unit in the fleet, is harder to interpret with confidence. Even with a valid comparison, interpretation still requires expert judgment. A deviation in the curve does not always map cleanly to a specific fault location or severity. This is a known and accepted characteristic of the test, not a flaw unique to any particular application of it, and it is addressed directly in the governing standards.

The Frequency Constraint

SFRA is an offline test. The transformer must be de-energized, isolated, and made safe for low-voltage signal injection. In most fleets, this restricts testing to a small number of occasions across a unit's operating life.

• Commissioning, to establish a baseline signature.
• After a suspected through-fault event, such as a nearby short circuit.
• After transport or relocation, when mechanical shock is a concern.
• During a scheduled maintenance outage.

For a critical unit, this may mean an SFRA test every several years. For a large fleet, scheduling constraints, outage costs, and crew availability compound the interval further. Between tests, the transformer's mechanical condition is not observed by this method at all.

The Visibility Gap

The combination of Sections 3 and 4 defines a practical limitation that is independent of SFRA's technical accuracy. A transformer can show a clean SFRA signature at one test, develop a mechanical condition shortly afterward, and continue operating for years without another SFRA measurement to detect it. The test's accuracy is not in question. Its temporal coverage is the constraint.

This gap is most consequential for progressive conditions, conditions that develop gradually rather than appearing suddenly. Loss of clamping pressure from insulation and pressboard shrinkage, or gradual bushing and lead connection degradation, are examples of failure modes that develop over months or years, well within the typical interval between SFRA tests.

Continuous Multi-Signal Monitoring

Continuous condition monitoring addresses the interval problem directly by observing the transformer at all times, under real operating conditions, rather than at scheduled snapshots. A four-signal approach, fusing triaxial vibration, thermal, magnetic field, and local weather data, provides several complementary views of the transformer's condition.

1. Triaxial vibration
   Captures the mechanical response of the winding and core structure to load-driven electromagnetic forces, which oscillate at twice line frequency and its harmonics. Changes in vibration amplitude, harmonic content, or resonant frequency indicate changes in mechanical stiffness, including the loss of clamping pressure that precedes winding displacement.
2. Thermal monitoring
   Tracks temperature distribution and rate of change, providing context for load-dependent behavior and identifying localized heating that can accompany connection degradation or insulation issues.
3. Magnetic field sensing
   Detects anomalies associated with circulating currents, core grounding faults, and stray flux conditions that are not always visible to vibration or thermal signals alone.
4. Local weather data
   Provides the environmental context, ambient temperature, and loading conditions needed to distinguish a genuine condition change from a normal, expected response to load or weather variation.

Because these signals are captured continuously, the system develops a model for each individual unit under its own real operating conditions, rather than relying on a single offline measurement taken years earlier. Deviations from that baseline, and the rate at which they develop, become the basis for anomaly detection and trending.

What Continuous Monitoring Does Not Replace

A precise and honest scope is necessary for this argument to hold up. Continuous multi-signal monitoring does not replicate SFRA's diagnostic resolution. SFRA, once a deformation has occurred, can localize the affected region of a winding with a level of geometric specificity that a fleet-level anomaly detection system is not designed to match. SFRA also does not depend on load conditions being present, since it is a controlled, repeatable injection test, and it provides a self-contained comparison across phases without requiring a lengthy monitoring history.

That specificity is real, and it has real value. The question this paper addresses next is who that value belongs to. Disc-level geometric resolution is decisive information for whoever is about to physically work on the winding. It is largely inert information for whoever is deciding, from a control room or an asset management platform, which of several thousand units to worry about this week.

SFRA's Correct Home: Manufacturing and Repair, Not Fleet Operations

A transformer's lifecycle involves distinct actors with distinct diagnostic needs, and SFRA's value concentrates in two of them.

The manufacturer, at factory acceptance testing, uses SFRA to confirm a newly built unit left the factory with correct winding geometry, establishing an as-built baseline against a known-good reference at effectively zero operating hours. This is the cleanest possible application of the test: a true baseline exists, the unit is already de-energized as a matter of course, and the outcome directly gates shipment.

The repair facility, when a unit has been pulled from service for a suspected or confirmed mechanical issue, uses SFRA before and after rebuild work to confirm precisely which section of the winding requires attention and whether reclamping or reforming actually restored the correct geometry. This is the one context in the transformer's operating life where disc-level specificity converts directly into an action taken with a wrench, on the unit in front of the technician, in real time.

The transformer operator occupies a different role entirely, and SFRA's structure does not match it. Operators do not decide which disc to reclamp; that decision, and the mechanical work behind it, belongs to the repair facility once a unit has already been pulled from service. What operators decide is which unit, among many, needs attention now, whether it can continue in service or should be derated, and when it should be scheduled for inspection or replacement. Those are trend and severity decisions made across a live fleet, and they require the kind of ongoing, load-and-weather-correlated data that a single offline snapshot cannot produce.

There is a practical compounding factor as well. The comparative value SFRA depends on, a valid baseline, a true sister unit, or a clean phase-to-phase comparison, is often incomplete for units already in service across a large, heterogeneous fleet. A manufacturer's factory acceptance test starts with a guaranteed baseline. An operator running SFRA on a twenty-year-old unit acquired through a fleet transaction may have no baseline at all, which weakens the very comparison the test relies on to mean anything.

The operational conclusion follows directly. SFRA should not be run as a routine, calendar-driven fleet-health practice by the operator. It should be reserved for the junctures where it has a genuine actor and a genuine action attached to it: a factory acceptance step at manufacture, and a pre- and post-repair verification step at a repair facility, invoked by cause, whether that cause is a new unit entering service, a suspected through-fault, or a condition flagged by continuous monitoring that warrants pulling a unit for physical work. Used sparingly and by cause in this way, SFRA delivers its full value. Used routinely by an operator as a fleet monitoring tool, it delivers a precise answer to a question the operator is not the right party to act on.

The Dynamic View: Load, Weather, and Operating Condition

The most structurally distinct advantage of continuous monitoring is one that no offline test, regardless of resolution, can provide: observation of the transformer's actual behavior under the full range of conditions it experiences in service. SFRA is performed at zero load, in a controlled state chosen for repeatability. It cannot show how a winding responds as load current rises through a summer peak, as ambient temperature swings between a Mongolia winter and an Arizona summer, or as thermal cycling accumulates across thousands of load cycles over a year.

A mechanical condition can manifest differently, or only become apparent, under specific load or thermal conditions. A loose winding may show a vibration signature change that is negligible at light load and pronounced at peak load, because the electromagnetic forces driving winding motion scale with the square of current. A connection degrading from thermal cycling may show a temperature rise that only appears during rapid load swings. None of this is visible to a test performed once, at zero load, in a controlled state.

This is the comparison summarized below.

The Actionability Gap: Specificity Is Not the Same as Decision Utility

Diagnostic resolution and operational usefulness are two different axes, and it is worth separating them explicitly. SFRA answers a geometric question with real precision: has this winding's shape changed, and if so, in which region. What it does not answer, and structurally cannot answer from a single offline measurement, is the question an operator actually needs answered to make a decision: is this condition stable or getting worse, how fast is it changing, and does it warrant action now or monitoring later.

A snapshot has no trend. An SFRA test performed once, at zero load, produces one point in time. Even a confirmed deformation, precisely located, leaves the operator without the information needed to prioritize it against every other unit in the fleet. A deformation that has been stable for ten years and a deformation that started last month can produce a similar-looking signature deviation, but they represent very different levels of urgency. Distinguishing between them requires a second data point, and in practice a second SFRA test is rarely available on a timescale short enough to establish a rate before a decision is needed.

Disc-level location, while diagnostically satisfying, also does not map cleanly onto the decisions an operator actually makes. Field response to a suspected mechanical condition is almost never disc-specific repair. It is one of a small number of fleet-management actions: continue the unit in service and monitor, reduce loading, schedule an outage for inspection or repair, or plan replacement. Each of these decisions is driven by severity and trajectory, not by which disc within the winding is affected.

This is where continuous monitoring's lower spatial resolution is a smaller practical cost than it first appears. A four-signal system trades some geometric precision for something an offline snapshot cannot produce at all: a rate. Because vibration, thermal, and magnetic field signals are captured continuously against known load and weather conditions, the system observes not just that a condition exists but how it is behaving over time, and whether it correlates with load, ambient temperature, or seasonal patterns. That rate, and its relationship to operating conditions, is the input an operator needs to decide whether a unit can run through the next peak season or needs to move up the inspection queue now.

The comparison below lays out the operator's actual questions against what each method can answer.

None of this diminishes SFRA's diagnostic accuracy. It clarifies who that accuracy serves. SFRA is well suited to confirming and localizing a condition for whoever is about to physically work on a unit already pulled from service, which is the manufacturer at acceptance testing or the repair facility at rebuild. It is poorly suited to telling an operator, managing a live fleet, which units need that decision made in the first place, or how urgently. That earlier and more consequential question, the one the operator actually faces, is answered by continuous monitoring.

Positioning by Role in the Transformer Lifecycle

The distinctions in this paper resolve into a straightforward positioning: the right diagnostic tool depends on which actor is making a decision and what kind of decision it is. Rather than a shared toolkit run by a single party, each actor in the lifecycle has a tool matched to the decision they are actually responsible for.

For the transformer operator specifically, the implication is direct. Continuous multi-signal monitoring is not a supplement to a fleet SFRA program. It is the primary, ongoing diagnostic tool for the decisions an operator actually makes, and it should be treated as such. SFRA remains part of the lifecycle, but as an event-driven activity performed by the manufacturer at the start of a unit's life and by the repair facility at the points where a unit is physically opened up, not as a periodic health check the operator runs on a live fleet.

Conclusion

SFRA is an effective and well-validated test. Its comparative structure and offline requirement are not flaws, but they do mean its value is realized fully only by an actor with a true baseline and a physical action to take: a manufacturer at factory acceptance, or a repair facility confirming a rebuild. Neither describes the transformer operator managing a live, distributed fleet.

The operator's actual questions, is this unit's condition stable, how fast is it changing, does it correlate with load or weather, and which units need attention this week, are trend and severity questions that a single offline snapshot cannot answer regardless of how precisely it localizes a fault. Continuous, non-invasive monitoring across vibration, thermal, magnetic field, and weather signals is built for exactly this job: it observes the transformer under real operating conditions, continuously, and produces the rate and severity information an operator needs to act.

For the transformer operator, continuous multi-signal monitoring is the right solution. SFRA remains a valuable and necessary test elsewhere in the lifecycle, but it belongs to the manufacturer's factory floor and the repair facility's rebuild bay, invoked sparingly and by cause, not run as a routine practice by the operator it was never built to serve.