Ira said:
Does anyone know why neutral and ground are not bonded in a household
electrical sub-panel?
The only explanation I could find so far was this:
If the both the neutral and the ground become open, with respect to the main
panel, at the sub-panel AND the neutral and ground are still short to each
other at the sub-panel then neutral and ground and any conductor that ground
is attached to, such as the metallic case of a power tool, will carry a
voltage determined by the resistance of appliances plugged into the circuit.
Are there other scenarios that anyone is aware of?
Let me think as I write. The purpose of safety ground is to have a
low voltage (with respect to the Earth or any electrical path to
Earth) to put between a person and the dangerous voltage available
within the system. The neutral is grounded at (or as close as
possible) to the source to make sure that the highest voltage in the
system is minimized, or at least, determinate. At the point where the
neutral is grounded, (bonded to the safety ground conductor) the
impedance of the ground electrode must be low enough that expected
current through this path do not produce dangerous safety ground
conductor voltage with respect to local Earth. This is why good
conductors, like water piping must be well tied into the ground
conductor system, since they may be the first contact points with the
local Earth potential. The voltage drop on the safety ground
conductor is minimized by having it carry no load current, except in
the case of a line to safety ground fault. By having a neutral
conductor carry load current, it is expected that the far end of that
conductor will have some voltage with respect to Earth, any time there
is load current.
Now, applying these principles, what is the case for a sub-panel?
First of all, the sub-panel, is, by definition not the closest
available point to the power source, so it is not the preferred place
for the ground to neutral bond. If the neutral conductor between the
main panel and the were to become high resistance, and the ground and
neutral were bonded at the sub-panel, Then the safety ground
conductor at the sub-panel would be expected to have a voltage
proportional to load current. During an ordinary line to neutral
fault, this would be half of line voltage, even though the safety
ground conductor were still connected. If both neutral and safety
ground conductors were broken, any load at all would apply full line
voltage to the distributed (from the sub-panel) safety ground.
So, unless the bonded node is also re grounded with a suitably low
resistance ground electrode system to mitigate the above problems, it
is much less risky to rely on the bond back at the main panel. It is
still a good idea to add an additional ground electrode to the local
safety ground node, since the ground potential at this locality may be
a bit different that it is back at the main panel, so a compromise may
be useful.
Of course, for sub-panels far enough away from the main panel, this
compromise may become intolerable (for instance, during a lightning
strike near one panel or the other) and at some point (that I don't
know how to define) it is better to re isolate the mains and redefine
the neutral connection to ground, creating a new main panel.
