Icu Ics Icw Ratings in Air Circuit Breakers Explained for Engineers

What are Icu, Ics, and Icw ratings in air circuit breakers? Icu, Ics, and Icw are IEC 60947-2 breaking-capacity ratings for low-voltage air circuit breakers that define, respectively, the ultimate short-circuit breaking capacity, the service short-circuit breaking capacity (expressed as a percentage of Icu, typically 50–100%), and the short-time withstand current a breaker sustains without tripping for a defined duration (0.05–1 s). Misapplying these ratings — specifying a breaker where Ics falls well below system prospective fault current, or ignoring Icw when upstream protection requires time-graded selectivity — leads to breaker failure under fault conditions, loss of discrimination, or failed IEC type-test compliance. This guide covers the IEC 60947-2 test sequence behind each rating, the engineering role of Icw in time-graded distribution systems, when Ics = 100% Icu is operationally critical, ABB Emax 2 frame rating comparisons, and a method for estimating required Icu in transformer-fed systems.

What Icu, Ics, and Icw Actually Mean (and Why Engineers Confuse Them)

In our experience reviewing switchgear specifications across pharmaceutical plants, data centers, and steel mills, the single most misunderstood line item on an ACB datasheet is the short-circuit performance block. Designers see "Icu = 65 kA" stamped on a breaker frame and assume the breaker can handle 65 kA repeatedly. It cannot. Not even close.

The three ratings exist because IEC 60947-2 recognizes that a breaker's behavior during a fault depends on what you expect it to do after the fault clears. Survive once and be replaced? Reclose immediately and keep operating? Carry through-current for a defined duration to allow downstream selectivity? Each question has its own rating.

Icw is a different animal entirely. It applies almost exclusively to ACBs and certain heavy-duty MCCBs.

Here is the practical hierarchy you'll see on almost any ABB Emax 2 or SACE E1.2 datasheet: Icu ≥ Ics ≥ Icw. A typical E1.2B frame, for example, lists Icu = 42 kA at 415 V, Ics = 42 kA (100% of Icu), and Icw = 42 kA for 1 second. The ratings happen to converge for that frame size; on larger frames like E2.2 or E4.2, you'll see Ics drop to 75% or 50% of Icu, and Icw stay constant regardless.

How IEC 60947-2 Tests These Ratings (and What the Test Sequence Means in Practice)

The test procedures in IEC 60947-2 Annex C are deliberately punishing. They are designed to expose weaknesses that bench testing at lower currents would miss. A breaker that passes Ics testing has done something genuinely impressive — it has interrupted a near-bolted fault at maximum prospective current, then re-closed onto that same fault, then interrupted it again, then carried full rated current for a thermal soak, and finally passed a dielectric withstand test at 2× Ue.

The O–t–CO Sequence for Icu

For Icu verification, the breaker performs one Open operation onto the fault current, waits a defined time (t ≈ 3 minutes), then performs a Close-Open operation onto the same fault. After this, the breaker is inspected. It does not need to function again. Visible damage to arc chutes, contact erosion, and insulation degradation are acceptable as long as the enclosure remains safe and no parts are ejected.

The O–t–CO–t–CO Sequence for Ics

Ics adds a second close-open cycle. After the second CO, the breaker must:

Pass a temperature-rise test at rated In. Pass a dielectric test at 2 Ue + 1000 V. Demonstrate that the overload release still trips correctly within tolerance.

Engineers often overlook the dielectric requirement. It is the most demanding part of the test, because by the third interruption the contacts are pitted and the arc chutes are coated in metallic vapor. A breaker that passes this test is genuinely robust.

The Icw Soak Test

Icw is tested differently. The breaker is closed, the test current is applied for the rated short-time duration (1 s or 3 s for most ACBs), and the breaker must not trip and must not weld. After the soak, the breaker must still open electrically and pass a follow-on temperature-rise test. The thermal energy involved is staggering — at 65 kA for 1 second, the I²t energy is 4.225 × 10⁹ A²s. The contact silver-tungsten alloy and the busbar copper must absorb this without softening.

This formula matters because downstream cable and busbar thermal ratings must exceed the upstream Icw I²t. If your ACB has Icw = 50 kA for 1 s, that's 2.5 × 10⁹ A²s of thermal energy that downstream conductors must survive. A 240 mm² copper cable has a thermal withstand of roughly 1.36 × 10⁹ A²s — already insufficient.

Why Icw Exists: Time-Graded Selectivity in Real Distribution Systems

A common mistake is to ask "why would I want a breaker that doesn't trip instantly on a 50 kA fault?" The answer lives in the philosophy of selective coordination, which is fundamental to any well-designed industrial distribution system.

Consider a pharmaceutical plant we worked on in Ireland. The main incomer is an ABB 1SDA070861R1 E1.2B 1600 ACB feeding a main bus. From that bus, six feeders supply distribution panels via MCCBs. If a fault occurs on one feeder, the design intent is that only that feeder's MCCB trips — the main ACB must hold the fault current long enough for the downstream device to clear.

That "long enough" is typically 100–300 ms. During that window, the ACB is carrying tens of kA without tripping. This is only possible because the ACB has a defined Icw. The short-time pickup is set above the let-through current of the largest downstream device, and the short-time delay is set to a value (typically 100–200 ms) that allows downstream coordination.

For deeper coverage of this design philosophy, see our breakdown of the IEC 60947-2 standard for Air Circuit Breakers.

Why MCCBs Don't Have Meaningful Icw

Molded-case circuit breakers typically have Icw values of only 5–10 kA for 1 s, or sometimes none at all (declared as "n/a"). That's because MCCBs are designed for current limiting — they trip in 5–10 ms during a fault, using the magnetic blow-off effect to break contacts before peak current is even reached. ACBs deliberately do not current-limit; they hold the fault, then trip on a defined delay.

This is why you see ACBs at the top of any radial distribution tree — they're the only device with the thermal mass and contact design to wait. Selecting an MCCB where an ACB is needed is a coordination disaster waiting to happen, a topic we cover in our guide on ACB nuisance tripping causes and fixes.

Real-World Selection: When Ics = 100% Icu Matters and When It Doesn't

Some engineers argue that Ics = 100% of Icu is overkill and adds unnecessary cost. In my experience, this depends entirely on the criticality of the installation.

Data Centers: Always 100% Ics

Data center loads cannot tolerate breaker replacement after every fault. A 25 MW colocation facility in Frankfurt we specified for had over 80 ACBs in the main distribution. The customer's design rule was non-negotiable: every ACB must have Ics = 100% of Icu, and every breaker must be selectable from the same family for spare parts standardization. We specified ABB Emax 2 frames throughout — the ABB 1SDA070981R1 E2.2B 1600 for distribution and larger E4.2 frames for the main incomers. For more on this design philosophy see our guide on ACBs in data centers.

Industrial Process Plants: Often 50–75% Ics is Acceptable

For a non-critical process plant — a cement mill, a quarry, a sawmill — Ics at 50% of Icu is often economically justified. The probability of a maximum-prospective fault is low, and if one occurs, a short outage to replace the breaker is acceptable. Cost savings on frame size can be 15–25%.

Hospitals and Life-Safety Systems: Always 100% Ics, with Documented Margin

Per IEC 60364 and local codes (including NFPA 99 in North American hospitals), life-safety branch circuits require breakers that can be returned to service without delay after a fault. This effectively mandates Ics = 100% of Icu for any ACB feeding such loads.

Comparing ABB Emax 2 Frame Ratings — A Practical Reference

The Emax 2 family is one of the most widely deployed ACB ranges globally. Here is how the short-circuit ratings stack up across common frame sizes typically used for incomer and feeder applications.

Notice that Icw at 3 s drops below the 1 s value. This is not a marketing trick — it reflects real thermal physics. The thermal mass of the silver contacts and the copper conductors limits how long they can carry fault current before reaching critical temperatures. A breaker rated 42 kA / 1 s might only handle 36 kA / 3 s.

Selection Calculator: Estimating Required Icu for a Transformer-Fed System

The most common scenario in industrial design is sizing the main ACB on a transformer secondary. The prospective short-circuit current at the transformer LV terminals can be estimated from the transformer impedance.

For a typical 2 MVA / 400 V / 6% transformer, this gives roughly 48 kA prospective. With 25% design margin, you'd specify an ACB with Icu ≥ 60 kA — landing you in E2.2B or larger territory. For complete methodology, see our walkthrough on how to size an air circuit breaker.

NEMA and IEEE Perspectives — How North American Standards Differ

Engineers working across regions need to understand that NEMA AB 1 and UL 489 use different terminology. The IEC concept of Icu is roughly equivalent to the UL 489 "interrupting rating," and Ics has no exact UL equivalent — UL listed devices are simply expected to remain operational after interrupting at their rated value, which is closer to the IEC Ics philosophy implicitly.

IEEE C37.13 covers low-voltage power circuit breakers (the ANSI equivalent of ACBs) and uses the term "short-time current rating" which corresponds closely to IEC Icw, typically rated for 30 cycles (0.5 s at 60 Hz). This is shorter than the IEC 1 s standard, which means an IEC-rated ACB with Icw = 50 kA / 1 s is actually thermally more capable than an ANSI-rated equivalent at 50 kA / 30 cycles.

What we typically see in the field is that European EPCs spec to IEC and accept ABB, Schneider, or Siemens equally. North American projects often spec to ANSI/UL, and the brand options narrow considerably. For a comparison of how the major manufacturers stack up across both standards, see our breakdown of ABB vs Schneider vs Siemens ACBs.

Common Specification Mistakes (and How to Avoid Them)

Mistake 1: Specifying Only Icu

A purchase order that says "ACB, 1600 A, Icu = 50 kA" is incomplete. Without Ics, you don't know if the breaker is meant to be replaced after the first fault or kept in service. Without Icw, you don't know if it can support time-graded coordination. Always specify all three.

Mistake 2: Ignoring Voltage Dependency

Icu is voltage-dependent. The same Emax 2 E2.2B frame might offer Icu = 66 kA at 415 V, but only 50 kA at 690 V. Engineers often miss this when designing 690 V systems for large motor drives or offshore platforms. Always reference the Icu at your system voltage, not the catalog headline value.

Mistake 3: Confusing Peak (Ipk) with RMS (Icu)

Icu is RMS symmetrical. Ipk is the peak current the breaker can withstand mechanically — typically 2.1× to 2.2× the RMS Icu, accounting for DC offset. Both are listed on datasheets. Don't compare them directly.

Mistake 4: Assuming Cascading Replaces Proper Sizing

Cascading (back-up protection) allows a downstream breaker with lower Icu to be backed up by an upstream device. This is permitted under IEC 60947-2 Clause 8.3.4, but only when verified by combined test results from the manufacturer. You cannot calculate cascading — it must be tabulated and certified. For more on this and related coordination issues, our piece on how air circuit breakers actually work provides the foundation.

Practical Decision Framework for Procurement

For procurement managers building purchase specs, here's the framework we use on EPC projects:

Step 1: Calculate prospective Isc at the breaker location using the transformer impedance method or a software study (ETAP, EasyPower, DIgSILENT). Add 25% margin for future system growth.

Step 2: Determine criticality. Data center, hospital, or continuous-process plant? Specify Ics = 100% Icu. Non-critical industrial? Ics ≥ 50% Icu is acceptable.

Step 3: Identify selectivity needs. Are there downstream breakers that must clear faults independently? If yes, specify Icw ≥ maximum prospective Isc, with duration matching your coordination delay (typically 1 s for ACBs in radial systems).

Step 4: Verify voltage. Ratings drop significantly at 525 V and 690 V. Always cross-check the manufacturer's table at your Ue.

Step 5: Standardize on a single family. Spare parts, training, and maintenance procedures all benefit. For most distribution projects in the 630–1600 A range, the ABB E1.2B family covers feeders well — the 1SDA070701R1 (630 A), 1SDA070741R1 (800 A), 1SDA070781R1 (1000 A), and 1SDA070821R1 (1250 A) all share the same Icu/Ics/Icw of 42 kA, allowing fully selective coordination across the range.

Step 6: Cross-reference with downstream protection. Smaller branch protection often comes from the Miniature Circuit Breaker collection, ground-fault protection from the Residual Current Device range, and control logic from the Relay collection — all of which must coordinate beneath the ACB's short-time band.

Field Anecdote: When the Wrong Rating Caused a €400,000 Outage

A few years ago, we were called in after an automotive paint shop in northern Italy lost power for 14 hours. The cause: a 2500 A ACB on the main bus had failed during a cable fault on a downstream feeder. The breaker had cleared the fault — Icu was sufficient — but it could not be reset, and no spare was on site.

The investigation revealed the original specification had called for Ics = 50% of Icu. The procurement team, optimizing cost, had selected exactly that. When the fault occurred at roughly 38 kA, the breaker absorbed the energy, performed the O–CO sequence within its Ics envelope, but the contacts welded slightly during the second close. It opened correctly but could not close again.

Replacement breaker lead time was 6 weeks. The plant ran on diesel generators at reduced capacity for two weeks while a temporary unit was sourced. Total downtime cost: approximately €400,000.

The lesson? Ics = 100% Icu is not a luxury for any plant where downtime costs more than the breaker premium. The cost difference between Ics = 50% and Ics = 100% on a 2500 A frame is typically €1,500–€3,000. The downtime premium should make the decision obvious.

Conclusion: Specify All Three, Always

Icu, Ics, and Icw are not interchangeable values that engineers can pick and choose between. Each describes a different physical capability — interrupt-and-survive, interrupt-and-keep-working, and carry-fault-without-tripping. A robust specification names all three at the operating voltage, with Icw duration explicit. Anything less leaves the design exposed to coordination failures, unplanned outages, and warranty disputes when a breaker fails to perform as the loose specification implied.

For procurement teams, the practical advice is simple: standardize on Ics = 100% Icu for any installation where downtime costs more than the breaker premium, which is almost always. Specify Icw to match your selectivity scheme, typically 1 second at the maximum prospective fault current. Verify ratings at your actual system voltage, not the catalog headline. And cross-reference your selected breaker against the manufacturer's coordination tables before signing the purchase order.

For the complete selection methodology, sizing calculations, and coordination strategies that tie these ratings together, see our master Air Circuit Breaker engineering guide, or browse the full Air Circuit Breakers collection at Stoklink for ABB Emax 2 frames covering 630 A through 6300 A applications.