Transformer Short-Circuit Impedance: The Design Logic Behind Matching System Short-Circuit Capacity

A transformer's short-circuit impedance (%Z) is the one dial that sets the fault current it can push into the system and its load-voltage sag. Match it to the system's short-circuit capacity — not a generic table — or the breaker rating and the voltage regulation both lose.

That sentence is the whole design problem. Most specifications treat %Z as a catalog default pulled from a capacity range, then move on. The result is a transformer that either drives fault current past the breaker's breaking capacity, or sags the bus enough to stall the motors it was bought to supply. This article treats %Z as what it actually is: a system-level matching decision between the transformer and the short-circuit capacity of the network it joins.

The Dial That Controls Two Opposite Things

Short-circuit impedance, measured as a percentage, is the voltage you must apply to one winding (the other short-circuited) to circulate rated current, expressed as a percentage of rated voltage. A 6% unit needs 6% of its voltage to push rated current through its own internal impedance.

That single number does two opposite jobs at once:

• It limits fault current. The higher the %Z, the more the transformer resists the enormous current a short circuit tries to draw. A 4% unit can deliver roughly twice the fault current of an 8% unit of the same rating.

• It worsens voltage regulation. The same impedance that blocks fault current also drops voltage under normal load. A higher %Z means a deeper sag from no-load to full-load, and a worse dip during motor starting.

So %Z is not a parameter to minimize. It is a balance, and the fulcrum of that balance is the system short-circuit capacity at the point where the transformer is installed.

Why "System Short-Circuit Capacity" Is the Real Constraint

The common instruction — "choose %Z so the fault current stays under the breaker rating" — is incomplete because it forgets that the transformer does not sit alone. The grid behind it has its own short-circuit capacity, and the two impedances add.

On the transformer's own MVA base, the source impedance in per-unit is roughly the transformer rated power divided by the system short-circuit power at the primary:

Z_source(pu) ≈ S_rated / S_k

and the transformer itself contributes:

Z_xfmr(pu) = %Z / 100

The fault current the secondary can deliver is set by the sum of the two, not by the transformer alone. Two consequences follow that most selection guides leave out:

A strong grid forces a higher %Z. When the upstream network is stiff (large S_k, tiny source impedance), the source contributes almost nothing to limiting fault current. The transformer %Z alone governs, and the worst-case fault current approaches the transformer-limited figure I_fault ≈ I_rated × 100 / %Z. If that exceeds the breaker, the only lever left is a higher %Z. A strong grid is exactly the case where you cannot get away with a low impedance.

A weak grid makes raising %Z pure penalty. When S_k is small, the source impedance already throttles fault current. Pushing %Z up to "improve" fault limiting buys almost nothing — because the source was doing the limiting — while still degrading voltage regulation. The knob only helps where the grid is already strong.

This is why matching %Z to the system short-circuit capacity matters more than matching it to the transformer's own rating. The right %Z is pinned between two limits set by the network: the maximum credible S_k (which sets the worst-case fault current the breaker must clear) and the minimum acceptable voltage at the most sensitive load.

Design for the Grid That Will Exist, Not the One That Exists Today

A frequent and expensive mistake is to size %Z against today's measured short-circuit capacity. System short-circuit capacity is not static — as the network grows, lines are added, and generation comes online, S_k at the connection point rises over the asset's life. A transformer specified against this year's S_k can find itself, five years later, feeding a much stiffer grid through the same %Z, with fault current now above the breaker's original rating.

The specification should therefore state the projected short-circuit capacity over the planning horizon (commonly a 10-year figure from the utility or a system study), and the %Z must keep fault current within breaker capacity at that projected S_k — not merely at the present one. This single forward-looking input prevents the most common mid-life breaker-upgrade crisis.

The Trade-Off Made Quantitative

Two rules of thumb let a specifier sanity-check a choice without a full study:

• Fault current scales inversely with %Z. For a stiff source, doubling %Z roughly halves the transformer-limited fault current. Moving a distribution unit from 4% to 8% can take a low-voltage fault current from the high-30s of kA down to the low-20s — often the difference between a standard and a premium breaker.

• Regulation scales with %Z. A higher impedance drops more voltage at the same load. A 4% unit typically regulates several percent better than an 8% unit at full load, and the gap widens sharply during motor starting, when the locked-rotor current exaggerates the dip across the impedance.

The conflict is concrete: the same %Z that protects the breaker also starves the motor. Industrial plants with large starting motors therefore cannot always accept the high %Z that a strong grid would otherwise demand, and the resolution is usually a coordination study that finds the highest %Z the motor bus can tolerate — not the highest the breaker allows.

Low %Z vs High %Z: What the Specification Actually Buys

Neither extreme is "correct." The answer is the value that satisfies the breaker limit at maximum projected S_k and keeps voltage within load tolerance at minimum S_k — the band the system short-circuit capacity defines.

Impedance Tolerance and Parallel Operation

%Z is not a single exact number in production. IEC 60076-1 and IEEE C57.12.00 both set the impedance tolerance near ±7.5% of the declared value. For units intended to run in parallel, the paired transformers must match each other more tightly — commonly within about ±10% of one another — or circulating current flows and the load shares unevenly, overheating one unit while underloading the other.

This rule intersects with system short-circuit capacity in a specific way: when a new transformer is added to an existing parallel bank, the new unit's %Z must sit inside that ±10% band of the installed units regardless of how the grid behind them has changed. The system SC capacity does not relax the matching rule; it only changes how badly a mismatch shows up in fault current and load share.

What the Specification Must Say

A procurement specification for short-circuit impedance should demand numbers and a study, not an adjective:

1. State the system short-circuit capacity (S_k) at the primary terminal, both present and projected over the planning horizon, supported by an IEC 60909 or IEEE 551 short-circuit study.

2. Specify the %Z value and its tolerance (IEC 60076-1 ±7.5%), not just "standard for the rating."

3. Require a fault-current calculation that includes source impedance, not merely the transformer-limited I × 100 / %Z figure, and confirm the result against the breaker's rated breaking capacity at the projected S_k.

4. Verify voltage drop at the most critical load — usually the largest motor — at full load and during starting, to confirm the chosen %Z does not breach the acceptable sag.

5. For parallel operation, require matched %Z within ±10% of the existing units, with the parallel coordination stated explicitly.

6. Factory test record of the measured %Z, so site acceptance is confirmation rather than discovery.

Six lines turn %Z from a copied default into a defended engineering choice.

FAQ

Q: What is transformer short-circuit impedance (%Z)?A: It is the percentage of rated voltage needed across one winding, the other short-circuited, to circulate rated current. It sets how much fault current the transformer can deliver and how much its voltage sags under load.

Q: How do I match %Z to the system short-circuit capacity?A: Start from the maximum projected short-circuit capacity (S_k) at the connection point. Choose %Z so the resulting fault current — calculated with the source impedance included, not just the transformer alone — stays within the breaker's breaking capacity, then confirm the voltage drop at the critical load is still acceptable.

Q: If the grid is strong, can I just use a low %Z?A: No. A strong grid has very low source impedance, so the transformer %Z alone governs fault current. A low %Z on a stiff network can drive fault current past the breaker rating; that is precisely when a higher %Z is required.

Q: Why does a higher impedance hurt voltage regulation?A: The same internal impedance that limits fault current also drops voltage under normal load. A higher %Z deepens the no-load-to-full-load sag and worsens the dip during motor starting, which can stall large motors.

Q: What impedance tolerance matters for parallel transformers?A: Standards allow about ±7.5% on the declared value, but units run in parallel must match each other within roughly ±10% to avoid circulating current and uneven load sharing.

The Takeaway

Short-circuit impedance is the parameter that decides whether a transformer fits the system or fights it. Specify it from the system's short-circuit capacity and the load's voltage needs — present and projected — not from a generic table. Get the match right and the breaker, the bus, and the motors all stay inside their limits. Get it wrong and the cheapest line on the specification becomes the most expensive failure on the network.