The headline is the delay, but the more interesting fact is where the physical build actually stands. OPG's Q3 2026 disclosure that Darlington's first BWRX-300 will miss its October 2030 target and its C$7.7 billion estimate arrived, by OPG's own timetable, only about four years before a target that has now been withdrawn. What OPG has not done is reveal what replaces either number: a comprehensive assessment of cost and schedule is expected to be substantially complete only by mid-2027. Until then, the project's central commercial parameter is unknown.
What is actually being built
Four GE Vernova Hitachi BWRX-300 reactors are planned at OPG's Darlington site in Clarington, Ontario, next to four operating CANDU units. Unit 1 would be Canada's first non-PHWR power reactor and, in OPG's framing, the G7's first grid-scale SMR. The BWRX-300 is a roughly 300 MWe natural-circulation boiling water reactor that reuses the ESBWR licensing basis — a design lineage that matters less commercially here than the delivery model does.
There is no EPC contractor — and that is deliberate
A frequent gap in coverage of this project is the missing EPC/EPCI contract. The correct answer is that there isn't one. Delivering the project is an Integrated Project Delivery (IPD) alliance, executed in January 2023 for six years, among:
| Party | Role in the alliance |
|---|---|
| OPG | Owner, licence holder, overall project responsibility |
| GE Vernova Hitachi | Technology developer; design, engineering and major-component procurement |
| AtkinsRéalis | Architect-engineer / integration engineer; engineering and procurement support |
| Aecon (via Aecon Kiewit Nuclear Partners) | Construction services: project management, construction planning and execution |
In May 2025 OPG awarded the execution-phase alliance construction contract to Aecon Kiewit Nuclear Partners, a general partnership of Aecon and Kiewit Nuclear Canada with Aecon as lead. Aecon's share is approximately C$1.3 billion, booked into Construction-segment backlog in Q2 2025.
The commercial consequence of IPD is significant and, for outside observers, awkward: an alliance shares risk and reward among its members rather than fixing a lump-sum price. The size of the overrun now materialising, and who carries it, is therefore not simply readable off a contract value. OPG has not disclosed how overrun exposure is allocated across the alliance — and that allocation is likely a large part of what the mid-2027 assessment is actually about.
The steel build: diaphragm-plate modules, not conventional rebar
The steel story at Darlington is specific and worth understanding, because it is not a standard structural-steel package.
The Unit 1 reactor building is being assembled from diaphragm-plate steel composite modules — steel plate boxes with concrete poured between them — rather than a conventional reinforced-concrete frame. Roughly 1,200 modules are expected in the reactor building overall; the first of eight levels alone comprises more than 200. Additional ring-shaped, pie-slice modules form the inner and outer annular layers and are lowered into the shaft.
The first major placement came on 22 April 2026, when the basemat module — about 37 m in diameter and nearly 2.1 million lb (roughly 950 t) — was lowered by heavy crawler crane into a 35 m-deep reactor shaft. Basemat steel modules were fabricated in multiple Ontario shops by Aecon Kiewit Nuclear Partners and welded together on site at a purpose-built fabrication facility. The stated target is to complete the first reactor building level by end-2026.
The fabrication route for this steel was set out in 2023, when Aecon and E.S. Fox agreed to jointly fabricate components at Aecon's Cambridge, Ontario facility and E.S. Fox's Port Robinson facility, welding sub-assemblies for shipment to site. A separate but related award went to Canam, which supplied and installed the SMR fabrication plant (about 2,014 m²) and a pump house using its Murox prefabricated steel building system under contract to Aecon — reportedly ahead of schedule.
Where procurement actually stands
Given the schedule reset, it is worth being precise about what is locked in and what is not, because a stalled project can create a false impression that packages have re-opened.
| Package | Status |
|---|---|
| Reactor pressure vessel, Unit 1 | Awarded — GE Hitachi to BWXT, 27 Jan 2025; value not disclosed separately |
| Basemat diaphragm-plate steel composite modules | Fabricated and installed (April 2026) |
| Reactor building upper-level modules (first level, ~200 modules) | Fabrication under way; first-level completion targeted end-2026 |
| SMR fabrication plant and pump house steel buildings (Murox) | Supplied and installed by Canam |
| Reactor building primary steel structure (Aecon / E.S. Fox) | Already contracted via the Aecon–E.S. Fox teaming route |
| Remaining ~1,000 reactor building modules (Unit 1) | Within the alliance construction scope — not an open market package |
| Units 2–4 | Concept stage; no schedule or procurement evidenced |
The short point for steel suppliers: the structural steel for Unit 1 is substantially tied up inside the IPD alliance and its named fabricators. The material that is genuinely still to be made and placed is the remaining diaphragm-plate module set — real tonnage and real welding, but not available as a competitive package. Any suggestion that the mid-2027 cost reset will open the reactor-building steel to new bidders is speculation, and not supported by anything OPG has said.
The long-lead equipment list beyond the RPV — where the C$1 billion-plus BWXT Canadian award also bundled 48 Pickering steam generators, so no standalone Darlington figure exists — is not fully public.
Reading the reset
The timing is telling. OPG flagged the miss before a final number existed, which is the transparent choice but also the one that leaves the market without an anchor. There are three things to watch:
- The revised commercial operation date. A slip past 2030 by a modest margin and a slip by several years are very different signals for Ontario's supply adequacy and for the credibility of the BWRX-300 as a repeatable product.
- Whether the C$7.7 billion figure survives in any form. OPG has said the previous estimate is not achievable; it has not said by how much.
- What the reset means for Units 2–4. One of the core selling arguments for a 300 MWe SMR is that serial construction drives cost down. A first-of-a-kind overrun on Unit 1 raises the stakes on whether the subsequent units are delivered as a series or quietly re-scoped.
What this means beyond Darlington
The first BWRX-300 was always going to be the reference case — for GE Vernova Hitachi's order book, for Ontario's nuclear refurbishment-and-new-build programme, and for every utility watching whether small reactors can be built on time in a Western regulatory environment. That reference case is now defined not by a 2030 milestone but by an unresolved cost assessment due in 2027. The steel is going in, but the arithmetic is being rewritten around it.

