This is a fundamentally difficult problem that the IEEE 1584 committee is struggling with, too.
There were just 3 different size enclosures used in the original 2002 data set. With so few geometries you can't realistically expect to do much with this.
Overall if you look at the table, there are three parameters that come out of the box model: default working distance, default bus gap, and exponent. In terms of how these affect your results, the larger the bus gap, the greater the incident energy. Although this is an empirical model, the longer the the arc length, the higher the amount of thermal energy emitted. From the formula, both arcing current and incident energy increase exponentially with bus gap but the overall impact is relatively small.
In terms of the exponent, open air equipment incident energy falls off at an inverse squared rate. So doubling distance cuts incident energy by a factor of 4. As the equipment gets to be deeper and has more of a tendency to "focus" the arc, the exponent is reduced so that with medium voltage (1-15 kV) gear it is nearly linear. Doubling the distance doesn't quite reduce the incident energy by a like amount or only just slightly more than half. So this tends to as you put it, push the arc flash boundary out a lot further. This is not realistic but it's what you get with an empirical model.
Hand in hand with that, the working distance is also increased with larger equipment so that has a tendency to reduce incident energy but not enough to overcome the increases due to the "focusing" effect.
|
