Current IntelPower LibraryShop main store

How-to

Closed delta transformer fault current calculation

Protection: breakers, fuses and SCCR Published 6 min read

Short answer

Compute each delta branch from its own kVA and impedance, then take phasor differences at the line terminals: for a 150 kVA leg with two 100 kVA legs at 240 V and 3.3% Z each, the bolted three-phase terminal fault is highest at each end of the 150 kVA winding and lower at the junction of the two 100 kVA windings. Size secondary equipment on the higher terminal value.

On this page
  1. What these numbers assume
  2. Step-by-step calculation
  3. Why the bank-total shortcut fails
  4. What the single impedance figure does not give you
  5. Using the result for equipment duty
  6. Checks before you file the study
  7. Questions asked

For this bank — one 150 kVA center-tapped winding and two 100 kVA windings in a closed 240 V delta, with 3.3% impedance quoted by the manufacturer — a bolted three-phase fault at the secondary terminals gives 21.9 kA at the junction of the two smaller windings and 27.5 kA at each end of the tapped winding. You get there by calculating each delta branch from its own kVA and its own impedance, then combining the branch-current phasors at the line terminals. Summing the winding kVA and applying one balanced three-phase formula understates the two high terminals by about 7% and overstates the light one by about 17%. Size secondary breakers, fuses and panel short-circuit ratings on the higher value, and treat the whole set as transformer-limited screening values until real source and conductor impedances go into the study.

What these numbers assume #

The scenario is a padmount closed-delta bank: the large winding is center-tapped to serve 120/240 V single-phase loads, the other two are lighter, and a single impedance figure is quoted for all three. The calculation is a bolted three-phase fault directly at the secondary terminals, which assumes:

  • zero impedance in the fault and its connections;
  • an infinite primary source, with no utility or primary-conductor impedance;
  • the full secondary line-to-line voltage across every delta branch;
  • the quoted impedance applied to each winding on that winding's own kVA base;
  • the same impedance angle in all three branches.

Under those assumptions the results are upper bounds: any real source impedance lowers them, and they say nothing yet about motor contribution.

Keep the terms straight, because a wrong word here produces a wrong number. Branch current flows inside one side of the delta. Line current flows in an external phase conductor. The center tap carries nothing in a three-phase fault; it enters only when the fault includes the neutral or ground.

Step-by-step calculation #

  1. Map the bank. Record which terminal pair the tapped winding spans, which terminal is common to the two lighter windings, and the winding voltages from the nameplate. Work from the nameplate and the transformer test record, the same paperwork-first approach that evaluating dry-type transformer hot spot temperature rise takes. Expect the full line-to-line voltage across each branch and half of it from either end of the tapped winding to its midpoint; confirming that by measurement is energized work, so de-energize and test for absence of voltage first and follow NFPA 70E, 2024 edition.
  2. Compute the rated branch current from each winding's own rating: I = S ÷ V. The tapped winding gives 150,000 VA ÷ 240 V = 625 A. Each lighter winding gives 416.7 A from its own rating at the same voltage.
  3. Divide by the per-unit impedance. With the quoted 3.3% = 0.033: 625 A ÷ 0.033 = 18,939 A on the large branch, and 416.7 A ÷ 0.033 = 12,626 A on each small branch. These are symmetrical rms winding currents.
  4. Convert to ohms if you model in software. Z = 0.033 × 240² ÷ S, so 0.033 × 57,600 ÷ 150,000 = 0.012672 Ω for the large branch and 0.033 × 57,600 ÷ 100,000 = 0.019008 Ω for each small one. Cross-check: dividing the branch voltage by 0.012672 Ω reproduces 18,939 A.
  5. Take the phasor difference at the light terminal. A line conductor joins two delta branches, so its current is the phasor difference of the two branch currents, which sit 120° apart when the impedance angles match. At the junction of the two lighter windings: √3 × 12,626 A = 21,869 A.
  6. Take the phasor difference at the other two terminals. Each joins the large branch to a small one, and |I1 − I2| = √(I1² + I2² + I1 × I2) because cos 120° = −0.5: √(18,939² + 12,626² + 18,939 × 12,626) = √(358,685,721 + 159,415,876 + 239,123,814) = √757,225,411 ≈ 27,518 A. That magnitude formula assumes equal impedance angles; when per-winding X/R data shows they differ, subtract the branch currents as complex quantities, the approach covered in complex phasor addition methods for parallel currents.
  7. Check the pattern. The large branch must carry 1.5 times a small one, two terminals must land at the higher value and one at the lower. Three equal answers mean you modeled a balanced bank, not this one.
Line terminal Branches meeting there Available current
Junction of the two lighter windings 12,626 A and 12,626 A 21.9 kA
Either end of the tapped winding 18,939 A and 12,626 A 27.5 kA

Why the bank-total shortcut fails #

Adding the ratings gives 350 kVA, and the balanced three-phase route gives 350,000 VA ÷ (√3 × 240 V) = 842 A, then 842 A ÷ 0.033 ≈ 25.5 kA at every terminal. That single number is 2.0 kA low at the two ends of the tapped winding and 3.6 kA high at the light terminal, and it hides the phase asymmetry that relay coordination needs. Straight arithmetic addition fails the other way: 18,939 A + 12,626 A = 31.6 kA, but branch currents displaced 120° never add in a straight line.

What the single impedance figure does not give you #

Percent impedance is a magnitude. It carries no resistance-to-reactance split, so it cannot yield X/R, DC offset, first-cycle peak or any asymmetrical duty, and one quoted value does not establish that all three branches share an impedance angle. Confirm what that value is defined on: each winding on its own base, which is the case calculated here; the tapped winding only; or a three-phase bank test. If the answer is anything but per-winding data, the results for the lighter branches are screening values. Request the transformer test report with per-winding %Z and X/R, use the higher terminal duty as the working value until it arrives, and finalize equipment ratings only on the documented bases.

Using the result for equipment duty #

Decisions by case:

  • A secondary breaker or fuse at either end of the tapped winding needs an interrupting rating at least equal to the higher terminal value before any motor contribution is added. NFPA 70, the National Electrical Code, 2023 edition, requires in 110.9 that equipment intended to interrupt fault current carry an interrupting rating at least equal to the current available at its line terminals, and 110.10 requires circuit impedances and component short-circuit ratings to be coordinated so fault clearing stays within withstand limits.
  • A device at the light terminal is governed by the lower value on that one phase, but one rating for the whole bank is simpler: apply the higher value everywhere unless the per-phase engineering is actually done.
  • An industrial control panel on this secondary takes the higher terminal value as the available fault current input to the short-circuit current rating constructed under UL 508A, Supplement SB.
  • Protective relaying keeps the two terminal values as separate phase quantities; an average blunts coordination on the light phase.
  • Arc flash runs on the modeled current at the work location plus the device clearing time per IEEE 1584-2018, the guide for arc-flash incident energy calculations; a transformer-limited terminal current alone does not set incident energy.

Confirm the adopted code edition and any local amendments with the authority having jurisdiction, and the ratings themselves with the equipment documentation.

Checks before you file the study #

  • The model with source and conductor impedances set to zero reproduces both terminal values from the table above.
  • The ohms cross-check in the conversion step holds in both branches.
  • Restoring utility, primary and secondary conductor impedances lowers the utility-fed contribution, and any motor or generation contribution is listed separately at the fault location.
  • The one-line diagram, transformer test data and protective-device schedule name the same winding identities and the same impedance base.
  • Every device record shows its location, phase-specific available current, the symmetrical duty and the rating it is compared against.

Questions asked

Which terminal value sizes the secondary breakers?

The higher one. It appears at both line terminals of the tapped winding, and most secondary schedules carry one rating for the whole bank. The lower value only governs equipment on the single phase at the junction of the two smaller windings.

Are these numbers symmetrical or asymmetrical?

Symmetrical rms. Percent impedance fixes only a magnitude; the DC offset and first-cycle peak depend on each winding's X/R, which a single percent-impedance figure does not contain. Get R and X per winding from the test report before comparing any device against peak or asymmetrical duty.

What changes when the real upstream source is not infinite?

Every value drops. These results are transformer-limited upper bounds; utility source impedance, primary cable and secondary conductor impedance all reduce them, while running motors add current for the first few cycles. A final study models those impedances as complex quantities.

What about a line-to-line fault on this bank?

On a balanced system a bolted line-to-line fault comes out near 87% of the three-phase value. On this unequal bank, compute the two involved branches the same way and take their phasor difference at the faulted terminal; a pair that includes the tapped winding still gives the higher result.

Does the center tap add fault current?

Not to a bolted three-phase fault, because the tap is outside that path. Faults from a line terminal to the tap are single-phase faults on the tapped winding and need their own analysis; the three-phase results here do not cover them.

Keep reading

  • Explainer

    Evaluating dry-type transformer hot spot temperature rise

    A 128.2°C average winding rise proves only the average, not the hot spot: the resistance method and a point RTD measure different quantities, so the 219°C combination is invalid. The 202.6°C full-load reading is a 180.8°C local rise, which projects to 220.8°C at the 40°C rating ambient and exceeds the 180°C hot-spot rise a class 220 system allows. Decision: run the simultaneous full-voltage, full-load heat run.

    Power distribution and transformers

  • Explainer

    Converting a 350 kVA soft start limit to starting current

    A 350 kVA soft start limit is about 421 A of line current at 480 V: divide 350,000 VA by √3 × 480 V. The figure is apparent power, so no power factor enters the conversion, and it carries no duration — that comes from the torque-speed margin. The 12.5 kV fuse is a coordination problem; motor protection belongs at 480 V.

    Motors and motor starters

  • How-to

    Powering a 600 V three-phase machine from 240 V single phase

    The nameplate works out to 20.8 kVA of apparent power. Fed from the single-phase shop service, that is about 87 A before converter and transformer losses, so the machine needs a rotary phase converter sized to its total demand plus a reverse-connected step-down transformer, unless tracing shows the internals run on 200 to 220 V. Verify loaded voltage balance before releasing the machine to production.

    Power distribution and transformers