If you've ever flipped to NEC Table 250.66, picked a size, and moved on, you're not alone — but it's worth knowing what that table is actually doing for you.
What Table 250.66 is
NEC Table 250.66 sizes the grounding electrode conductor — the conductor that connects your system (usually at the service) to the grounding electrode system. You look up the size of your largest ungrounded service-entrance conductor, and the table gives you the grounding electrode conductor size. Copper or aluminum, a couple of columns, done.
But notice what the table doesn't ask you: it doesn't ask how much fault current your system can deliver, or how long a fault might last before the protective device clears it. That's because the table is a conservative, prescriptive shortcut. It's built on the assumption of typical service sizes and typical fault levels, and it bakes in a safety margin so you don't have to do the math on every job.
The assumption underneath
Here's the idea the table is standing on: a grounding conductor has to survive a fault without melting or getting so hot it damages its insulation or connections. That's really two things at once — the current has to be low enough, or short enough, that the conductor doesn't overheat.
IEEE 80, the guide for substation grounding, treats this the same way, just more explicitly: a conductor's ability to carry fault current depends on its cross-section, its material, how hot it's allowed to get, and how long the fault lasts. The thermal relationship is basically I-squared-t — current squared times time — against the conductor's ability to soak up that heat. Double the fault current and you need four times the thermal capacity, or a device that clears four times faster.
So when NEC 250.66 gives you, say, a 2/0 copper grounding electrode conductor for a 500 kcmil copper service, it's effectively saying: for the fault levels you'd normally see on a service that size, this conductor will ride through a fault without failing. The table hides the I-squared-t thinking, but the thinking is still there.
A small worked example
Say you're sizing a grounding electrode conductor for a service with 500 kcmil copper ungrounded conductors. Table 250.66 points you to 2/0 AWG copper.
Now sanity-check it the IEEE 80 way. Suppose the available ground-fault current at the service is about 10 kA, and your main breaker clears a fault in roughly 0.1 seconds. A 2/0 AWG copper conductor (about 67 mm²) has far more thermal capacity than that short 10 kA burst needs — the I-squared-t of the fault (10,000 squared times 0.1) sits well below what that size of copper can absorb before hitting its temperature limit. The table's answer and the thermal check agree, which is exactly the point: the table is the quick route to a size that the physics already supports.
The takeaway isn't that the table is wrong — it's that it's a shortcut with assumptions baked in. If your fault current is unusually high or your clearing time is unusually slow, the table's assumptions start to stretch, and that's when doing the thermal check yourself is worth the effort.
The practical habit
For most services, Table 250.66 is fine — use it, and move on. But whenever the numbers look out of the ordinary, take a minute to check the fault current and the clearing time behind your installation. If you'd like to run the numbers rather than guess, try the grounding conductor sizing calculator — it walks through the sizing with the formula and the standard clause shown, so you can see what the result is based on.
Tables are great servants, but they work best when you know what they're assuming.
Top comments (0)