Lots of confused issues here. The first one to be aware of is that switched devices (breakers, switches, contactors) actually have three ratings, one for closing, opening, and maintaining. The lowest rating is for opening where it has to quench the arc and it is even lower with DC compared to AC. The highest rating is when it maintains only. Although some manufacturers (S&C for instance) list all three you may only see one rating. This is rarely used at low voltage but at medium voltage it is common practice to interrupt low level faults with a switch device but maintain position and let a fuse trip on high level faults as a fuse saver. This is particularly common in class E2 starters. It is unique to switches. Everything else has either an absolute current limit due to magnetic forces such as bus bars or a time-current limit due to thermal effects such as wiring. Some such as bus bars have both but some such as semiconductors naturally have enough resistance that only the thermal limit applies. So you might get multiple ratings with switches but only a single AIC or SCCR rating with most equipment.
Second concept is current limiting. This can be achieved via obviously adding long wiring runs but can be done in a more compact form using a small isolation (1:1) transformer or line reactors. Superconducting current limiters are the ultimate but unfortunately unaffordable (for most uses) current limiters. There are some electronic current limiters that use a combination of a reactor to protect the electronics and an SCR bridge. So far this is most common in solid state breakers. In fact current limiting can be done very inexpensively in a very small package using a one turn inductor which is really just a hair-pin turn section of bus. This is exactly the principle behind current limiting breakers. Fuses use a more complicated spring loaded cone approach.
Many of these devices are nonlinear, specifically current limiting fuses, breakers, and superconducting current limiters. They do nothing below a certain current. Above that current they clamp but the clamping rate is limited so we get a let through curve. The exception is SCLs which are a true clamp and to a limited extent ferro-resonant so-called constant voltage transformers (though you would not use them that way).
Third concept is dynamic resistance which is a circuit breaker concept. As a breaker opens the arc resistance rapidly rises. This inherently solves a coordination problem with breakers. According to time-current curves typically all breakers trip at the same speed in instantaneous tripping. In practice though the dynamic resistance of the smaller, ever so slightly smaller breaker will inhibit a larger one no matter what the time-current curves say. But since it relies on an effect that is hard to predict or model the only way to guarantee this works is with testing. Hence it only works with breakers that are in tested combinations and we must rely on manufacturers for this or test ourselves. This would be where that “engineering supervision” exception applies. I have the equipment to do the testing but nobody ever asks to do this.
Third is series resistance which dovetails into dynamic resistance as well. In this case we rely on an upstream breaker to provide current limiting despite the fact that we are exceeding SCCR. This is different from what I guess we can call “static” current limiting which is what you get with let through curves, transformer, reactor, and wiring impedance. This is just the standard SCCR calculation in action. No series rating needed because we are not relying on dynamic resistance at all.
There are many methods but I’ll mention one of the most prominent, the ANSI method. The IEEE Buff book covers most of them and their assumptions and limitations. The ANSI method was developed specifically for power distribution and prior to widespread availability of calculators. It works very well with just slide rules and paper. It makes use of the fact that at distribution levels, inductance dominates so it largely ignores resistance. This drops the need for complex number math. It uses relatively simplified assumptions about motors and shielded cables so that you can do the calculations with just a set of tables, name plate data and some simple calculations. IEC also contains a similar method. This method has the advantage of fixing the major issues with the infinite bus calculation in that it provides much lower, more realistic results and avoids under predicting the effects of large inductors without the issue of overestimating sources that the simplified IEC method suffers from. But it is intended to upper bound SCCR for short circuit rating purposes. A far more accurate method is to include the full complex math and just do it as a straightforward impedance calculation. This is the most accurate and used by the power system software in arc flash models where over-predicting SCCA results in under-predicting arc flash effects. In practice my observations is that ANSI overpredicts anywhere from 10-100%. In computer software we can simply select the more accurate model and take advantage of lower SCCA values.
Then we get into some old discredited series rating methods. The first is “cascade ratings”. These were kind of arbitrary rules used prior to formalized series rating tests prior to the 1970s. No reason to even mention them except historically. They were things like maintaining a ratio of say 3:1.
The up over down method was promoted until “Interplay of Energies in Circuit Breaker and Fuse Combinations” was published in 1993 which discredited this method. In the up over down method start with a let through chart. Starting with the SCCR on the line side of the fuse move up to the fuse rating then to the left to get the instantaneous current limiting rating. This is valid but the next step isn’t. Now move to the left but stop at the ultimate current limiting line on the left then go straight down vertically to read the downstream SCCA.
The problem with SCCA is that we know the time it takes for the breaker to interrupt current. That is taken right from the time current curve. But we don’t know the precise time that the breaker contacts actually open. We cannot know this because it depends on the magnetic forces involved (dynamic resistance again). If the contacts open before the fuse clears then the resulting current seen by the fuse drops and clearing times are extended. So except for peak let through energy the degree of protection is indeterminate. The up over down method results in higher SCCR ratings than the Listed ratings which alone should tell you not to rely on it.
Here is an example.
So on a 60 A class J fuse on their charts at 60 kA line side short circuit we start at the bottom and move up to the 60 A line then to the left to see peak let through of 8 kA. This is real, does not rely on and is not affected by series ratings. This is what the contacts see if they open before the fuse does. Now in the down (discredited) method we stop at the diagonal line and slide down to the bottom line again to predict only 3 kA symmetrical amps. So if this method was accurate a 5 kA breaker is all that is needed! The reality though is that although the J fuse is very fast it probably won’t clear fast enough for a 15 A, 5 kA miniature DIN rail breaker with a clearing time of one cycle. So the breaker will see 8 kA, not 3 kA.
So using “up over” gives us peak short circuit current, which is the intended (not discredited) use of the let through chart. Series rating will of course be much less. In the above example it will be between 3 and 8 kA.
