Transformers, CT/VT & Condition Testing: A Practical Guide

A current transformer (CT) is, functionally, a current source: its whole job is to reproduce a scaled-down version of the primary current in its secondary winding, essentially independent of what impedance (burden) is connected to that secondary. This has a critical and somewhat counterintuitive consequence: a CT's secondary must never be left open-circuited while primary current is flowing. With nowhere for the secondary current to go, the CT tries to force that current through infinite impedance, and the resulting secondary voltage can spike to dangerous, insulation-damaging levels — this is precisely why CT secondaries are always short-circuited (via a shorting terminal block) before any connected relay or meter is disconnected for maintenance.

A voltage transformer (VT), by contrast, is a voltage source: its job is to reproduce a scaled-down version of the primary voltage across its secondary terminals, and it expects to see a relatively high-impedance burden. The equivalent hazard is reversed: a VT secondary must never be short-circuited, since a voltage source driven into a short circuit draws very high current, which can damage the VT winding or blow its secondary fuse (and, either way, remove the voltage signal that protection relays downstream may be relying on). Every burden budget, every accuracy check in this category's CT and VT calculators sits on top of this fundamental difference in how the two devices behave.

A CT's rated burden (in VA) is the maximum load its secondary can drive while still maintaining its rated accuracy class — every relay coil, meter, and length of secondary lead wire connected to that CT consumes some of that burden budget, and the cumulative total (including lead resistance, which itself depends on cable length and cross-sectional area) has to stay within the CT's rating. For protection CTs specifically, the Accuracy Limit Factor (ALF) describes how far above rated current the CT is still guaranteed to reproduce current accurately (e.g. an ALF of 20 means up to 20 times rated current) — a critical figure because protection relays need an accurate current signal precisely during the high-current fault conditions they're meant to detect, not just at normal load.

Knee-point voltage is the practical boundary of that accurate region: beyond it, the CT's magnetic core saturates and secondary output stops scaling linearly with primary current, which can seriously degrade protection performance (particularly for differential and earth-fault schemes, which rely on CT outputs staying proportional even during severe faults). A CT sized with an inadequate knee-point voltage for its actual burden and required ALF can saturate exactly when accurate current information matters most.

VT sizing checks a similar burden budget to CTs (all connected relay coils, meters, and instruments against the VT's rated output), but its most distinctive consideration is the voltage factor — how much continuous overvoltage above nominal the VT must be able to withstand, and for how long, during a system earth fault. This isn't fixed; it depends entirely on the power system's earthing arrangement. In a directly (solidly) earthed system, an earth fault is cleared quickly and the healthy phases see only a modest, brief overvoltage, so a lower voltage factor suffices. In systems without automatic fault clearing (some resistance-earthed or Petersen-coil-earthed systems, where a single earth fault can be tolerated and left connected for an extended period rather than tripped immediately), the healthy-phase overvoltage can persist far longer, so the VT has to be rated for a correspondingly higher voltage factor and duration — a VT correctly sized for one earthing arrangement can be significantly undersized if the same design is reused on a differently-earthed system.

These are genuinely different problems even though they both concern transformers. Sizing a new transformer starts from connected load, applies growth and safety margins, and selects a standard kVA rating (ANSI/IEEE or IEC preferred sizes) — with ambient and altitude derating applied against the transformer's rated nameplate conditions. Analyzing an existing transformer instead starts from open-circuit and short-circuit factory or field test data to determine that specific unit's actual core loss, copper loss, efficiency at various loading levels, and voltage regulation — quantities that describe how a particular transformer performs, not how to select one in the first place.

A transformer's standard nameplate rating assumes specific reference service conditions (per IEC 60076: 40°C maximum ambient, 30°C monthly average, 20°C yearly average) — operating outside those conditions requires derating. Ambient derating above the reference point is commonly approximated at around 1.25% capacity reduction per °C (a widely cited rule of thumb consistent with IEEE C57.96 dry-type guidance), while the standards' own formal loading guides (IEC 60076-7 for oil-immersed, IEC 60076-12 for dry-type) instead model this with a full hot-spot-temperature and insulation-aging thermal calculation dependent on cooling class and load-cycle shape — meaningfully more detailed than the simple percentage rule, and the correct reference for anything beyond preliminary sizing. Altitude derating is a separate, independent effect — thinner air at elevation both reduces convective cooling efficiency and reduces dielectric (insulation breakdown) strength around bushings and in air-cooled sections — and applies its own separate correction (roughly 1K permitted temperature-rise reduction per 400m above 1000m for naturally-cooled ONAN transformers, and per 250m for forced-cooled ONAF/OFAF types).

A transformer's total loss at any given loading is the sum of its core (iron) loss — roughly constant regardless of load, since it depends on the magnetizing voltage which stays close to rated regardless of load current — and its copper loss, which scales with the square of the load fraction (loss = Pfe + x²·Pcu, where x is the fraction of rated load). Because these two loss components scale so differently with load, efficiency itself varies with loading level and typically peaks somewhere below full load rather than at 100% — a transformer quoted as '98% efficient' without a stated loading point is an incomplete figure. Voltage regulation (the difference between no-load and full-load secondary voltage, expressed as a percentage) is estimated from the same open-circuit/short-circuit test data using Kapp's approximate formula, and likewise varies with both loading level and load power factor.

Busbar and switchgear rating actually combines three physically distinct phenomena that are easy to conflate into a single 'is it rated high enough' question. Continuous ampacity is a steady-state heat-balance problem — resistive (I²R) heating balanced against convective and radiative heat loss to the surroundings, solved for the current that holds the busbar at its maximum allowable operating temperature indefinitely. Short-time thermal withstand is a completely different, much faster phenomenon: how much fault current the busbar can carry for a brief, defined duration (typically 1-3 seconds) during a short circuit before adiabatic heating (too fast for meaningful heat loss to the surroundings) raises conductor temperature past a safe limit. Electrodynamic force is not a thermal phenomenon at all — it's the mechanical force between parallel current-carrying conductors during high fault current, which can physically bend or dislodge busbars and their supports if the mounting isn't braced to withstand it. A busbar comfortably rated for continuous ampacity can still be dangerously inadequate for short-time withstand or electrodynamic force, because each check is governed by different physics and has to be verified independently.

Once transformers, motors, and other rotating or wound equipment are in service, periodic insulation testing is one of the standard ways to catch developing insulation problems before they cause a failure. Insulation Resistance (IR) is a single absolute reading — commonly checked against IEEE 43's kV+1 rule of thumb (minimum acceptable megohms roughly equal to rated kV plus 1) — while Polarization Index (PI) is a ratio of two readings taken 1 and 10 minutes into the same test, capturing how much the resistance trends upward as the insulation polarizes. A winding can have an adequate absolute IR reading while still showing a weak PI trend (often a sign of surface contamination or moisture), which is exactly why both are run together rather than relying on either alone — see the Protection & Relaying guide's discussion of grading margins for the broader theme of why single-number checks often need a companion check to be trustworthy.