In need of urgent assistance? Call +86-13427815151 

新闻与咨询

New industry Technology regarding to Bussmann fuse, ABB breakers, Amphenol connectors, HPS transformers, etc. 

Fuse I2t Explained: Pre-Arcing vs Clearing and How to Build an I2t Budget

Time:2026-11-25   Author:As Beam   Browse:

Fuse Fundamentals · I²t Method

Every fuse datasheet gives two I²t numbers, and neither is the one most designs actually need. This guide explains what pre-arcing and clearing I²t measure, how the clearing value changes with fault current, and how to build an I²t budget that protects a semiconductor instead of merely protecting the cable.

I²t is the energy a fuse lets through, expressed as the integral of current squared over time. It appears in every selection table, it is quoted to three significant figures, and it is routinely misapplied — because the number printed in a table is measured under specific conditions that a real fault rarely reproduces.

The purpose of this article is to make the two published values usable: to show which one applies to which check, why the clearing value rises with fault current, and how to compare a fuse against a semiconductor's withstand rating without relying on a rule of thumb.

The two numbers and what they measure

ValueMeasured toWhat it tells you
Pre-arcing I²tThe instant the arc strikesHow much energy is needed to melt the element
Clearing I²t (total)Extinction of the arcThe energy the downstream device actually absorbs

The difference between them is the arcing energy, and it is not small. In the 170M 1000 V range, a 400 A part has 42,000 A²s pre-arcing and 250,000 A²s clearing — the arc period accounts for five times the melting energy. A design that checks the semiconductor against pre-arcing I²t will therefore conclude that a fuse is acceptable when it is not.

Rule: compare the protected device against the clearing value, never the pre-arcing value. Pre-arcing tells you when the fuse starts to operate; clearing tells you what the device experiences.

Why the printed clearing value is not a constant

The catalogue quotes clearing I²t at a stated voltage and at a stated prospective current, usually a high one. At lower fault currents the fuse takes longer to operate and the arc burns longer, so the total I²t rises — sometimes by a factor of two or three compared with the tabulated figure.

This is why let-through curves exist alongside the tables. The table gives a single point, usually at the maximum breaking capacity; the curve gives the whole relationship. For a semiconductor check, the relevant fault current is the actual prospective current at the installation point, not the fuse's maximum rating.

Fault current relative to ratingOperating regimeWhat to use for I²t
Below ~3 × InOverload, long melting timeTime-current curve; I²t is large and rising
~3 to 10 × InTransition, partial current limitationLet-through curve, both axes
Above ~10 × InFull current limitationTabulated clearing I²t is representative

The practical message: the tabulated number is a limiting-case value for a hard fault. Design for it and you are conservative for hard faults, but you still need the curve to know how the fuse behaves at the softer faults that occur more often.

Building an I²t budget

A semiconductor has a published I²t withstand — the energy it can absorb without junction damage, usually given for a specific pulse width. The fuse must let through less than that, with margin. Three factors erode the margin:

  1. The device's withstand falls at longer pulses. A diode rated for a 10 ms surge has a lower I²t capability at 20 ms. Match the pulse width to the fuse's clearing time at the fault current, not to the headline figure.
  2. Asymmetry. A fault starting at a voltage zero produces a DC offset and a first peak roughly double the symmetrical value. Fuses are tested with this offset; the device's withstand figure may not include it. Where the manufacturer publishes both, use the asymmetrical values on both sides of the comparison.
  3. Ageing and pre-load. A fuse that has carried current for years, or that is pre-heated by an adjacent fuse, operates slightly earlier. The effect is small but it reduces the margin rather than increasing it.

A defensible budget looks like this: fuse clearing I²t at the actual prospective current, multiplied by 1.2 for scatter, must be below the device's I²t withstand at the corresponding pulse width, divided by 1.5. Designs that use a straight equality pass the calculation and fail in service.

Worked example: 170M 400 A at 1000 V

A rectifier leg carries 400 A at 1000 V. The prospective fault current at the fuse is 30 kA. Two candidate parts exist in the 690 V and 1000 V classes at this current.

PartSizePre-arcingClearingClearing × 1.2
170M2621 (690 V)0018,000 A²s125,000 A²s150,000 A²s
170M3975 (1000 V)1*42,000 A²s250,000 A²s300,000 A²s

If the rectifier stack has a 10 ms withstand of 400,000 A²s, the 690 V part passes with margin (150,000 < 267,000, where 400,000/1.5 = 267,000) and the 1000 V part also passes (300,000 < 267,000 is false — so it does not). The 1000 V part is outside the budget at this fault level.

Note what the check did not involve: the current rating, the voltage rating, or the watts. Those were settled earlier. The I²t budget is a separate gate, and a part can pass the first three and fail this one.

The mistake hiding inside “same current”

Because the voltage class drives element mass, two fuses with the same current rating in different classes have different I²t — and the difference is in the unsafe direction when substituting a higher-voltage part into a lower-voltage design.

Current690 V clearing1000 V clearing1250 V clearingRatio 1250/690
200 A28,000 A²s40,500 A²s41,000 A²s1.46×
400 A125,000 A²s250,000 A²s335,000 A²s2.68×
630 A325,000 A²s755,000 A²s660,000 A²s2.03×

At 400 A the substitution nearly triples the let-through energy. A design that was inside its semiconductor's withstand on the 690 V part can be outside it on the 1250 V part, with no change to current, voltage margin or physical rating that would flag the problem.

Parallel operation and why it does not halve I²t

Two fuses in parallel do not simply double the current rating and halve the I²t. They share current imperfectly, and when one operates first the other must carry the full current before it opens. The result is a clearing I²t higher than half of a single fuse's value, and a risk of unequal ageing.

Manufacturers publish parallel-operation guidance where they support it, usually with a derating factor and a requirement that the two fuses be from the same batch and mounted symmetrically. Where that guidance does not exist, parallel fuses are not a legitimate way to reach a higher current.

DC circuits: the same method, different numbers

The I²t method applies unchanged to DC, but three inputs change. The clearing value is published at the DC voltage, not the AC one; the arc period is longer because there is no natural current zero; and the system time constant affects the shape of the fault current.

For a battery system with a long time constant, the current rises slowly and the fuse's clearing time lengthens, raising the total I²t above the tabulated value. Where the manufacturer publishes only an AC clearing value, that value must not be used for a DC design.

Reading a let-through curve

The let-through chart has prospective current on the horizontal axis and let-through I²t on the vertical, usually both logarithmic. Two curves appear: the fuse's own let-through, and the “no fuse” line showing what the circuit would deliver without protection.

Reading it takes four steps:

  1. Locate the prospective current at the installation point on the horizontal axis.
  2. Read up to the fuse curve and across to the vertical axis. That is the clearing I²t at that fault level.
  3. Compare with the device withstand at the same pulse width.
  4. Repeat at the lowest fault current the circuit can produce, because that is where the fuse is slowest and the I²t is highest.

The fourth step is the one skipped in most designs. A fuse selected against a hard 30 kA fault may pass comfortably, then operate at a 3 kA arcing fault lasting 40 ms and let through far more energy than the table suggests.

I²t for cables is a different calculation

Semiconductor protection is about the first few milliseconds. Cable protection is about seconds. The same fuse cannot be optimised for both, which is why drives use an aR fuse for the semiconductor and a separate device for the cable.

AspectSemiconductor checkCable check
Time scale1–20 ms0.1–5 s
ParameterClearing I²t versus device withstandTime-current curve versus cable damage curve
Failure mode if wrongJunction damage, short life, silent degradationInsulation damage, fire risk
Typical deviceaR high-speed fusegG fuse or MCCB

The k²S² relationship governs the cable side: the fuse's total clearing I²t must not exceed k²S² for the conductor, where k depends on the insulation and conductor material. For a 35 mm² copper/PVC cable, k is around 115, giving a permissible I²t of about 154,000 A²s — a figure that the 400 A aR fuse at 250,000 A²s exceeds. That is why the cable needs its own protection in a drive circuit rather than relying on the semiconductor fuse.

Choosing between adjacent ratings

A common decision is whether to use a 200 A or a 250 A part when the calculated load sits at 190 A. The current check alone would allow either; the I²t check usually decides.

Consideration200 A part250 A part
Thermal margin at 190 A5% — too tight in a hot enclosure24% — comfortable
Clearing I²t (1000 V class)40,500 A²sapproximately 63,000 A²s
Watts at In40 Wabout 52 W
Semiconductor marginBetterWorse by roughly 50%

So the choice is a trade: the larger part protects the fuse against nuisance operation, the smaller part protects the semiconductor. Where the enclosure is hot and the load is steady near 190 A, the 250 A part is usually correct; where the device's I²t margin is thin, the 200 A part is correct and the thermal problem must be solved another way.

Why field I²t measurement is unreliable

Instruments that record fault current and estimate I²t exist, and their readings after a fuse operation are frequently misleading: the recorder's bandwidth is lower than the fuse's speed, the current sensor saturates, and the recording window may miss the first half-cycle where most of the energy is delivered.

A recorded value that is lower than the catalogue figure should not be taken as evidence of better performance. Use it to confirm that a fault occurred and roughly how big it was; use the datasheet for the actual design value.

What I²t does not tell you: peak current

Two fuses can have the same clearing I²t and very different peak let-through currents. The peak matters because it is what stresses the semiconductor's voltage blocking capability through di/dt and stray inductance, and because it sets the mechanical and magnetic forces on busbars.

For a current-limiting fuse, peak let-through is published separately, usually as a curve of peak current against prospective current. In the 170M 1000 V range, at 100 kA prospective, a 400 A part limits the peak to a small fraction of the prospective value — but the exact figure comes from the let-through chart for that part, not from the I²t table.

Both numbers are needed: I²t answers “will the junction survive”, peak answers “will the busbar and the device's blocking margin survive”. A selection that checks only I²t is half a selection.

DC case: 800 V battery string

A battery string at 800 V DC with a prospective fault current of 12 kA and a time constant of 5 ms. The fuse under consideration is a 400 A square-body part with a published DC clearing value of 300,000 A²s at 800 V DC.

Because the time constant is long, the current rises slowly and the fuse's operating time lengthens. Applying the published value without adjustment underestimates the let-through. Where the manufacturer provides a DC let-through curve, read it at 12 kA and the 5 ms time constant; where only the tabulated value exists, apply a margin or consult the manufacturer.

The comparison is then against the battery module's or the DC bus capacitor's withstand, which is usually expressed for a short pulse. Two very different components — a semiconductor and a capacitor bank — both need to be inside the fuse's let-through envelope, and the smaller of the two governs.

Series connection: what it does and does not do

Two fuses in series do not halve the I²t and do not double the voltage rating in any useful sense. Whichever operates first determines the let-through, and the second one then interrupts the residual current. The practical effect is a slightly lower total I²t than a single fuse of the lower rating, but never the sum of two.

Where a designer needs a higher voltage rating than a single part provides, series connection is sometimes proposed. It is a legitimate technique only where the manufacturer publishes guidance for it, because the voltage sharing between two fuses after one operates is not defined by the datasheet.

A practical checklist

  1. Find the prospective fault current at the fuse position, not the transformer's rated current.
  2. Read the clearing I²t for the exact part at that voltage class.
  3. Adjust for fault current using the let-through curve if the fault is not a hard one.
  4. Apply margin of 1.2 on the fuse side and 1.5 on the device side.
  5. Check the pulse width matches: the fuse's clearing time at that current versus the device's I²t rating period.
  6. Re-check after any substitution — voltage class, indicator variant, or a different manufacturer's equivalent.

FAQ

Is a lower I²t always better?

Usually, because it means the protected device absorbs less energy. But a very low I²t part may operate on normal inrush, so the characteristic has to suit the load as well as the fault.

Why do two parts with the same current have different I²t?

Because the element geometry and mass differ — voltage class, indicator variant and body size all change the element. I²t is a property of the part, not of the ampere rating.

Can I use the pre-arcing value if my device's withstand is specified as pre-arcing?

Only if the device data sheet explicitly says pre-arcing and the comparison is like for like. In practice, semiconductor withstand figures include the full pulse, so clearing is the correct comparison.

Does ambient temperature change I²t?

It changes the rating and the operating time slightly; the published I²t is measured under defined conditions. For a hot enclosure, derate the current first, then check I²t at the reduced current.

Summary

I²t is the only fuse parameter that describes what the protected device experiences, and the published value is a single point on a curve. Use the clearing figure, adjust it for the actual fault current, apply margin on both sides, and re-check whenever the part number changes — because the difference between a 690 V and a 1250 V fuse at the same current is a factor of nearly three in energy, and nothing on the label warns you.

New industry Technology regarding to Bussmann fuse, ABB breakers, Amphenol connectors, HPS transformers, etc. 


TAG:   fuse I2t pre-arcing I2t clearing I2t let-through energy semiconductor fuse selection I2t coordination fuse let-through curve