Icw, Ipk, Icc, Icu, Ics: is this panel rated for this site?

A panel is good enough for a site when the prospective fault current at its incoming terminals is no more than its Icw for the time the upstream protection takes to clear, with prospective peak no more than its Ipk, or let-through peak if a current-limiting device is upstream; or no more than its Icc with the named protective device fitted. Icu and Ics do not answer that question, because they rate a circuit-breaker's ability to break a fault and not the assembly's ability to survive one. A 1000 kVA transformer at an assumed 5% impedance gives about 26.7 kA at its terminals, 23.6 kA at the end of 50 m of four 300 mm² per phase, and 16.4 kA at the end of a 50 m sub-main of one 185 mm² per phase, so the route can decide the answer.

These sums check a claim or spot a problem; a competent engineer should own the fault study, and the verification evidence comes from the assembly maker.

What each letter on the plate is telling you

The definitions follow ABB's application papers, which are supplier guides that summarise IEC 61439 and IEC 60947-2.¹² ABB's workbook converts Icw thermally: 40 kA for 1 s equates to 80 kA for 0.25 s, provided the peak stays inside Ipk.³ That I²t rule covers heating only. A non-limiting breaker does not reduce the first peak: 80 kA prospective peaks at about 176 kA at n 2.2, against 84 kA of Ipk on a 40 kA Icw at n 2.1. So the conversion is usable only with a current-limiting device upstream and the maker's data.

ABB's rule for a breaker is that Icu or Ics must exceed Ik, and its making capacity Icm must exceed the peak, at the breaker's position;² which of Icu or Ics is a decision for the specification. Neither Icw nor Ipk says how the panel behaves with a person standing at it during an arcing fault, so ask separately if it matters. That is our judgement, not a quoted rule.

How much fault current the site has

Ask first. BEAMA says prospective short-circuit current should be provided free of charge by the distribution network operator on enquiry.⁴ To check the letter, the sum needs the kVA and percentage impedance from the rating plate.

The impedances below are an assumption: ABB's typical values by size, not a UK maker's guarantee.² Replace them with your plate value. Every fault figure below scales with it, and a lower impedance raises them: 4.75% on 1000 kVA gives 28.1 kA, not 26.7 kA, and the 10 m case in the illustration then sits on both limits. Ask whether the plate value is measured or guaranteed, and what tolerance applies.

Ik = S / (√3 x U x Z), with an infinite source and no motor contribution. A UK maker describes its range as typically 11000 V to 433 V no-load, with 1333 A on its 1000 kVA drawing,⁵ so 433 V is used here. A finite source lowers the figure: on 1000 kVA, 24.7 kA at an assumed 250 MVA on the 11 kV side and 22.2 kA at 100 MVA.

Two people can get different answers from the same plate. For 1000 kVA at 5%, the 433 V basis gives 26.7 kA, 400 V gives 28.9 kA, and adding the voltage factor of 1.1 from ABB's worked example gives 31.8 kA.² That 19% spread is before anyone has made a mistake, so ask which basis a figure uses. The UK's often-quoted 16 kA is a practical maximum at the cut-out of a 230 V single-phase supply up to 100 A,⁴ and does not read across to a three-phase board.

What the three-phase figure leaves out

Parallel supplies: two 1000 kVA units on one bus give about 53.3 kA with no cable, and BEAMA tells specifiers to count them.⁶

Motors: HSE tells dutyholders to consider contribution from connected loads.⁷ ABB's rule of thumb is 4 to 6 times the rated current of motors connected directly at the fault point, counted when their rated currents sum to more than Ik/100.² As an illustration, 500 kW of direct-on-line motors adds about 3.5 to 5.2 kA.

Line to neutral: Rittal notes that neutrals may be tested at no less than 60% of the three-phase current.⁸ ABB's method gives a delta-star transformer equal sequence impedances,² so a line-to-neutral fault at the terminals equals the three-phase figure with an infinite source (26.7 kA) and exceeds it with a 250 MVA source (25.3 kA against 24.7 kA). Cable then cuts it faster, to 17.9 kA at 20 m on one 185 mm² per phase with a same-size neutral. Close to a transformer, a 60% neutral is not covered by the phase comparison.

What a run of cable does

Three-phase levels at the panel terminals on a 1000 kVA, 5% transformer:

Single-core copper values are ABB's, converted from 80 °C to 20 °C with ABB's 3.95 x 10⁻³ per kelvin, because a cold cable gives the highest fault current.⁹² Take real figures from your cable maker. A bigger conductor adds under 1 kA at 50 m because reactance barely falls with size, while parallel runs do halve it.

Read the single-run rows as sub-mains behind a breaker. That breaker sets the clearing time and any let-through, and its clearing time can lengthen as current falls, so recheck the Icw time at the lower current. A fault on the cable feeding a close-coupled panel may be cleared only by the HV protection, so ask for that time too. Both points are engineering judgement.

Peak: use the formula close to a transformer

Peak governs the force on busbars and supports. ABB publishes the factor n that it says comes from Table 7 of IEC 61439-1:¹ 1.5 up to 5 kA, 1.7 up to 10 kA, 2.0 up to 20 kA, 2.1 up to 50 kA, and 2.2 above 50 kA. ABB warns that near transformers the power factor can be lower than the table assumes.¹ The usual formula, Ip = √2 x k x Ik with k = 1.02 + 0.98 x e^(-3R/X), reproduces ABB's worked example of 33 kA at X/R 6.6 giving 76.6 kA.²

Transformer resistance is load loss divided by rating: R% = load loss in kW / kVA x 100, and the test certificate gives the load loss. The GB ecodesign limits for a liquid-immersed 1000 kVA unit are 10.5 kW (Tier 1) and 7.6 kW (Tier 2, from 1 July 2021), so R is at most 1.05% and 0.76%.¹⁰ A lower R raises the peak and leaves the RMS alone: at the 1000 kVA terminals, 1.05% gives 57.9 kA, 0.76% gives 61.8 kA (2.32 x Ik), and 0.5% gives 65.8 kA.

The table falls short of the calculated peak wherever the transformer dominates the circuit. At the terminals, with R at the Tier 2 limit and the impedances assumed above, n x Ik is about 7% low at 315 kVA, 9% low from 500 kVA to 1250 kVA, and 12% low at 1600 kVA and 2000 kVA. It is worst just below a band edge: a 315 kVA unit with a plate value of 4.75% gives 8.8 kA at n 1.7, so 15.0 kA against 19.7 kA, and a panel plated 10 kA with Ipk 17 kA passes on RMS and fails on peak. Cable closes the gap only once it adds enough resistance. On 1000 kVA with R at the 0.76% limit that takes about 12 m of 1 x 185 mm², 23 m of 1 x 300 mm², 45 m of 2 x 300 mm² or 90 m of 4 x 300 mm². The limit is a maximum, and at R 0.5% those become 18 m, 56 m, 111 m and 221 m. By 50 m of 1 x 185 mm² the table is safe, at 32.9 kA against a calculated 29.4 kA. Close to a transformer, compare prospective peak by the formula, not by n.

The check, in order

1. From the site, get the prospective short-circuit current at the panel's incoming terminals, its basis, its source (DNO letter, transformer plate or a fault study), and the modes of operation of the supplies.⁶ If you only have the origin value, ask for cable length, size and runs, and work it down. Do not take credit for cable that could be replaced or bypassed.
2. Ask what is upstream, its clearing time at that fault level, and whether generators, parallel transformers or large motors can feed the fault.
3. From the nameplate and the maker's documents, read Icw with its time, Ipk, and Icc with the device named, for the circuit concerned, and the neutral and protective conductor ratings as well as the phase rating. HSE's point on fault conditions extends to earthing conductors.⁷
4. Compare. Icw route: prospective RMS at or below Icw, clearing time inside the stated time, prospective peak at or below Ipk, or let-through peak if a current-limiting device is upstream. Icc route: prospective RMS at or below Icc with the named device fitted.
5. Check each breaker inside the panel separately, at its position. ABB makes that a separate condition.¹
6. Record the fault level, the date and the source, and recheck if a transformer, a supply or the route changes. Regulation 5 of the Electricity at Work Regulations bars equipment from use where its strength and capability may be exceeded in a way that may give rise to danger.¹¹

As an illustration with invented round figures, take a panel plated Icw 25 kA for 1 s and Ipk 52.5 kA on that 1000 kVA transformer and a 1 x 185 mm² sub-main. At 5 m it fails on both, at 25.2 kA and 55.7 kA. At 10 m it passes with 23.9 kA and 50.7 kA, but at 4.75% it sits on both limits (25.0 kA and 52.5 kA), and at 4.5% it fails (26.2 kA and 54.4 kA). At 20 m it passes at 21.6 kA and 43.0 kA on the transformer alone, while 500 kW of direct-on-line motors takes it to 25.0 to 26.8 kA, on or over the limit.

What the plate does not let you conclude

A 70 kA breaker in a 35 kA panel does not make a 70 kA panel. ABB's example shows the real path: a 35 kA Icw assembly on a 60 kA plant is acceptable only if the breaker upstream lets through less than the assembly's I²t and Ipk, and the devices inside are adequate.¹

Equal or larger I²t does not mean equal performance. BEAMA warns that a test at 50 kA for 3 s does not accurately represent performance at 70 kA for 1 s.⁶ The 3 s test has more I²t (7,500 against 4,900 kA²s), but the peak at 70 kA is 154 kA at n 2.2 against 105 kA at 50 kA.

Do not specify busbar kA behind a current-limiting device. BEAMA says to state the prospective current at the incoming terminals and let the maker coordinate.⁶

A machine panel may carry no 61439 assembly rating, but it should still have a short-circuit current rating. Siemens' white paper on Edition 6 of IEC 60204-1 says chapter 7.10 requires one, reached by design rules, calculation or test, and names IEC 61439-1 and IEC 60909-0 among the methods.¹² Ask for the figure, the method and the protection it relies on, and settle which standard your panel is built to first.

When the rating depends on protection upstream

A conditional rating is only as good as the device it names, and that device is often in someone else's switchboard. Ask for these in writing:

• The Icc value and the circuit it applies to.
• The exact protective device: maker, type, frame, rating and trip settings, or fuse maker, type and rating, with its breaking capacity and current-limiting characteristic.³
• The let-through I²t and peak at your prospective current, against the assembly's I²t and Ipk.¹
• That the device's own breaking capacity is at least the fault level where it is installed.
• A label on the panel and a line in the operating file stating that the rating relies on that device.

If the answer is "it's a standard panel", ask for the design verification evidence for that device combination. The temperature rise of the same section is a separate verification.

Notes

1. ABB, Technical Application Papers No. 11: Guidelines to the construction of a low-voltage assembly complying with IEC 61439 Part 1 and Part 2, section 8. Supplier guide. Table 8.2 and the Ipk rule are stated by ABB to come from IEC 61439-1 Table 7; check the licensed standard before relying on them. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

2. ABB, Technical Application Papers No. 2: MV/LV transformer substations, theory and examples of short-circuit calculation, table 2, sections 2.3 and 2.4, and Annex B. Supplier guide, 2008. ABB's worked example uses c = 1.1, and it refers its typical impedances to IEC 60076-5. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

3. ABB, Work book: The standard IEC 61439 in practice, 2CPC 000 313. Supplier guide, 2015, written against the German edition. ↩ ↩²

4. BEAMA, Consumer unit 16 kA conditional rating application, December 2018. ↩ ↩²

5. Bowers Electricals, Distribution Transformer Manufacture and 1000 kVA standard drawing, marked uncontrolled, 2011. Supplier documents, which give no impedance. ↩

6. BEAMA, Specifying a Low-Voltage and Controlgear Assembly, edition 1, July 2026, sections 3 and 7. ↩ ↩² ↩³ ↩⁴

7. HSE, Memorandum of guidance on the Electricity at Work Regulations 1989, HSR25, paragraph 64 (under regulation 4), which lists fault level at the point of supply and contribution from loads such as motors, and paragraph 86 (under regulation 5). ↩ ↩²

8. Rittal technology library, Standard-compliant switchgear and controlgear production: application of IEC 61439, 2013. Supplier guide, which predates the 2021 edition of the standard. ↩

9. ABB, Electrical installation handbook, volume 2, table 1 of the voltage drop section, single-core copper. Supplier guide. ↩

10. legislation.gov.uk, Regulation (EU) No 548/2014 as it applies in Great Britain, Annex I, Table I.1, liquid-immersed units with one winding up to 24 kV and the other up to 3.6 kV. Limits are maximums, so a real unit may have lower R. ↩

11. legislation.gov.uk, Electricity at Work Regulations 1989, regulation 5. ↩

12. Siemens, White paper on the changes in IEC 60204-1 Edition 6. Supplier guide; check the clause in the licensed standard. ↩