I don't know about IEC, but for the ANSI standards, there's no tolerance. The maximum is the maximum. From a practical viewpoint, you have to take into account the accuracy of your short circuit calculation. There are generally impedances that are neglected in most calculations, such as impedance in breakers, terminations, etc that can tend to reduce the actual fault current slightly. You might be able to sharpen your pencil and come up with a slightly lower worst-case calculation.
But if you are the responsible engineer, then you probably can't just say that it's so close, I'm sure it is fine.
I have been there a couple of times with low voltage installation, 208 V to 600 V.
When the ASCC of a transformer was a little too high for the switchgear, I then factored in the impedance of the conductors from the transformer to the switchgear.
That gave me a documented, acceptable, value.
I don't know about IEC, but for the ANSI standards, there's no tolerance. The maximum is the maximum. From a practical viewpoint, you have to take into account the accuracy of your short circuit calculation. There are generally impedances that are neglected in most calculations, such as impedance in breakers, terminations, etc that can tend to reduce the actual fault current slightly. You might be able to sharpen your pencil and come up with a slightly lower worst-case calculation.
But if you are the responsible engineer, then you probably can't just say that it's so close, I'm sure it is fine.
I have checked with one Vendor and they send me part of tests report and is mentioned values a little bigger than nominal (Ik3" and Ip) ans well related with duration of short circuit in seconds.
However in nameplate will be the mominal values (for example Ik3"= 40 kA (3 sec), Ip= 100 kA).
From the manufacturer, it depends on whether they are testing to meet the rating (e.g. Ik3" = 40 kA) or if they are testing to see when it actually fails (Ik3" > 40 kA).
As a user - design around it passing ONLY the maximum value(s), but no more (i.e. no "higher tolerance"). That is the conservative and safe approach - and, as an engineer, one of your prime responsibilities is the SAFETY of the system with regard to both equipment and personnel.
I don't know about IEC but when close to the nameplate, X/R must be considered in the ANSI/IEEE world. Maximum could end up being less than nameplate. I believe 17 is the breakpoint.
From the manufacturer, it depends on whether they are testing to meet the rating (e.g. Ik3" = 40 kA) or if they are testing to see when it actually fails (Ik3" > 40 kA).
As a user - design around it passing ONLY the maximum value(s), but no more (i.e. no "higher tolerance"). That is the conservative and safe approach - and, as an engineer, one of your prime responsibilities is the SAFETY of the system with regard to both equipment and personnel.
Depending on exactly how the manufacturer did their testing, they might be able to provide a customized rating. Often equipment falls between two preferred rating categories. On a recent project a manufacturer was able to provide a nameplate showing a rating about 5% higher than the preferred rating.
I should have have said custom rating rather than custom nameplate. The device was a voltage regulator, so the nameplate doesn't actually include the short circuit rating. The IEEE standard calls for regulators to withstand 25x load current for 2.0 seconds, and the vendor was able to calculate a rating 28.4x load current for 1.55 seconds for our specific regulator build. That apparently was the transition point between being limited by thermal damage versus being limited by mechanical strength to withstand magnetic forces.
