Why CEC and NEC disagree on a buried 600 A feeder

Run the same problem through both codes and you get two different conductors. Copper, directly buried, trefoil, one circuit, 20 °C earth, 600 A load, 75 °C terminations. The Canadian Electrical Code lands on 750 kcmil. The National Electrical Code lands on 500 kcmil. That is a full size and a half apart, and it is not a typo in either book.

I ran into this while cross-checking a sizing tool against both codes, assumed I had a bug, and went looking for it. There was no bug. The two codes get to underground ampacity by genuinely different routes, and the difference only shows up when your terminations are rated below the conductor's insulation.

Where the numbers come from

Underground ampacity does not come from the ordinary raceway tables in either code. Both send you to a set of engineered tables built on a specific buried geometry, thermal resistivity, and load factor.

• CEC — Appendix D, Diagram D10 and Tables D10A/D10B. Trefoil, directly buried, 20 °C earth.
• NEC — Informative Annex B, Table B.2(10), Detail 9 for one circuit. Same physical idea: trefoil, buried, 20 °C earth, RHO-90, 100 % load factor.

Now look at the columns those two tables publish, because that is the whole story.

The CEC's Appendix D underground tables are published at 90 °C. There is no 75 °C column. The NEC's Annex B buried table is the opposite — it publishes 60 °C and 75 °C, and there is no 90 °C column.

What happens at a 75 °C termination

Both codes agree on the governing principle. CEC Rule 4-006 and NEC 110.14(C) say the same thing in different words: the ampacity you are allowed to use is limited by the lowest-rated termination in the circuit. Land a 90 °C conductor on a 75 °C lug and you may not use the 90 °C value.

The codes then diverge on how you get the 75 °C number.

Under the NEC, you simply read the 75 °C column that Annex B already prints. For 500 kcmil that is 626 A. It clears 600 A, so 500 kcmil is the answer.

Under the CEC, there is no 75 °C column to read. You have a 90 °C table value and a 75 °C termination, so you convert — the conventional ratio being the 75 °C/90 °C relationship from the ordinary ampacity tables, about 0.886. Applying it:

602 A × 0.886 = 533 A — which does not clear 600 A.

So 500 kcmil fails, and you go up. The next size that survives the same treatment is 750 kcmil: 731 A × 0.886 = 648 A. Hence the CEC answer.

Which one is right?

Both, within their own book. That is the uncomfortable answer, and it is worth sitting with rather than explaining away.

The NEC's 75 °C column is a real, tabulated engineered value for a buried conductor at that conductor temperature. The CEC's 0.886 conversion is a ratio borrowed from a different table's structure and applied to a buried case. It is conservative, and conservative is a defensible place for a designer to sit — but it is a derivation, not a published buried ampacity at 75 °C.

Two practical consequences:

1. Do not port a result across codes. If a US drawing shows 500 kcmil buried at 600 A and you are stamping to the CEC, you have not verified anything by noticing the NEC agrees. It doesn't — it answered a different question.
2. If the CEC number hurts, buy the terminations out of the problem. The entire penalty here lives in the 75 °C cap. Specifying 90 °C-rated terminations on both ends, where the equipment allows it, moves you from 750 kcmil to 500 kcmil for the cost of the lugs. On a long feeder that is not a small saving.

One more caveat about Annex B

Annex B is informative. It is guidance, not a mandatory NEC requirement — the mandatory low-voltage basis remains Table 310.16. If you lean on B.2(10) for a buried run, say so on the calculation and be ready to defend it, because an inspector is entitled to ask why you did not use the mandatory table. In practice Annex B is the only thing in the NEC that models the buried case honestly, which is exactly why it exists.

And if your arrangement falls outside what either table covers — an unusual burial depth, a duct bank the details do not show, more circuits than the table tabulates — neither code has an answer for you. That is a Neher-McGrath or IEEE 835 calculation, and no amount of interpolation makes it otherwise.

You can work either path through the conductor sizing tool; it will show you the derating chain and tell you which table it read.