Cost estimating
NEC vs CEC for Cost Estimators: What Actually Changes Across the Border
The US runs on the National Electrical Code; Canada runs on the Canadian Electrical Code. For a cost professional moving an electrical estimate between them — or a US owner reviewing a cross-border team — the useful question isn't whether the codes differ. It's what changes in the number, and what transfers. Most of it transfers.
Two codes govern the electrical work I price. In the United States it is NFPA 70, the National Electrical Code (NEC); in Canada it is CSA C22.1, the Canadian Electrical Code (CEC). They are close cousins — the CEC borrows the NEC's structure and much of its physics — but they are not identical, and a cost professional working across the border needs to know which differences move an estimate and which are noise.
Here is the honest answer, from someone who builds electrical estimates under both: the code is the small recalibration. The market is the large one. Get that ranking wrong and you will produce a tidy, code-correct estimate that is wrong by a wide margin.
What barely changes: the quantities
Both codes use the same AWG / kcmil conductor system, and their tabulated ampacities are broadly aligned — identical at the common small sizes (14 AWG at 15 A, 12 AWG at 20 A, 10 AWG at 30 A) and diverging only modestly at larger sizes. The NEC anchors everything to one master table, Table 310.16; the CEC splits the same information across Tables 1–4 by conductor material and installation method. Same physics, different filing system.
What that means for cost: your conductor, containment and termination quantities carry across the border with only minor adjustment. A cost professional who can build a defensible cable and containment take-off in Toronto can build one in Dallas — the quantities are not where the border shows up.
Where the tables diverge — and why it rarely moves the total
The differences are real but marginal to an estimate:
| Item | NEC (US) | CEC (Canada) |
|---|---|---|
| Ampacity reference | Table 310.16 (combined) | Tables 1–4 by material & install |
| Bundling derate bands | 4–6, 7–9, 10–20, 21–30… | Table 5C: 4–6, 7–24, 25–42, 43+ |
| Voltage drop target | ~3% branch / 5% total (informational) | ~3% / 5% total (a Rule, 8-102) |
| Grounding electrode | 8 ft rod minimum | 3 m (≈10 ft) rod |
| Continuous load | 125% sizing (80% rule) | Equivalent 125% / derate |
Different bundling bands change a derate factor here and there; different grounding details change a handful of line items; the voltage-drop philosophy is effectively the same number on both sides (a mandatory Rule in Canada, a recommendation in the US). None of it moves the order of magnitude. If your estimate swings materially when you switch codes, the code isn't what changed.
For a US owner: a cross-border cost professional's method and quantities transfer intact. The border shows up in the rates, not the take-off.
What actually moves the number — and it isn't the code
The real recalibration is the market, and it dwarfs the code differences:
Labor rates and the shop split. US electrical labor is a different market — and the union versus merit-shop divide, region by region, can swing installed cost more than any code table. Davis-Bacon and prevailing wage apply on federal and many public projects, and must be priced deliberately, not assumed.
Regional deltas. Northern Virginia, Dallas–Fort Worth and Phoenix — the three great data-center markets — do not price alike. Labor availability, permitting, utility posture and local escalation differ enough that a single "US rate" is a fiction.
Material pricing and escalation. US material indices, tariff exposure and escalation curves are their own animal, and the long-lead items that dominate an electrical budget — switchgear, transformers, standby generation — sit on a continental supply chain that has been anything but stable.
What's already shared: the labor basis
Here is the part that surprises people. Estimators on both sides of the border build labor from the NECA Manual of Labor Units — a US institution. The labor method is more aligned across the border than the code is. A Canadian electrical estimator is already, in the most cost-relevant sense, speaking American.
The discipline is code-agnostic
Strip it back and the things that make an electrical estimate defensible do not depend on which code governs. An elemental build-up, an honest estimate class — whether you express it as AACE 18R-97 or CIQS Class A–D — a $/MW density read on mission-critical work, a named-risk contingency instead of a padding percentage, whole-life costing: all of it is portable. Only the benchmarks change.
So the move from CEC to NEC is a small code recalibration sitting on top of a genuine market recalibration. Respect the second, and a cost professional crosses the border with the discipline intact — the take-off transfers, the judgment transfers, and only the rates need re-basing to the region in front of you.
Estimate in either code
My free toolkit is built on the code-agnostic disciplines above — elemental build-up, estimate-class ranges (AACE 18R-97 and CIQS A–D), named-risk contingency and a $/MW density check. The cable sizing tool follows CEC conventions, but the method is identical under the NEC.
Open the free toolkit →