Darlington SMR Cost Reassessment: Why Modularization Has Not Eliminated First-of-a-Kind Risk?
Foundation structures and large construction equipment have already appeared at the Darlington small modular reactor construction site in Canada, but the commercial operation date and cost still require reassessment. Ontario Power Generation disclosed on November 10 that the first unit is expected to miss the previously set target of commercial operation in October 2030 and cannot be completed within the original estimate of CAD 7.7 billion. This provides an important reality check for small modular reactors: reducing unit power can change the construction organization approach, but it cannot automatically mature the first-of-a-kind design and engineering interfaces.

Modularization changes the construction method; it cannot replace complete design
The "modularization" of small modular reactors (SMRs) mainly refers to improving delivery through standardized design, factory manufacturing, and repeat construction. It does not mean that an entire nuclear power plant is a device that can be plugged in directly after leaving the factory. The BWRX-300 selected for Darlington belongs to the boiling water reactor (BWR) pathway. Beyond the reactor, there are still steam turbines, cooling water, control, power, and auxiliary facilities. How modules are transported, connected, and accepted all require specific engineering documentation.
This disclosure lists the incomplete first-of-a-kind engineering design as a factor affecting construction and supply chain performance. It should be noted that this is not news that the project has stopped construction, nor has a new final budget been provided. The owner is reassessing design maturity, procurement and construction assumptions, and market conditions, with the assessment expected to be largely completed by mid-2027. This date is the target for completing the recalculation and cannot be treated as a new commissioning date for the nuclear power plant.
The deeper the factory manufacturing, the more important the interface definition
Why does design maturity affect cost? Equipment suppliers need clear loads, dimensions, interfaces, and inspection requirements, while construction contractors need confirmed layouts and work sequences. If the parameters of a system change, the related equipment, supports, piping, and even building penetrations may all need to be re-coordinated. Not all changes will cause large-scale rework, but when multiple parties are simultaneously waiting for engineering input, the waiting of site personnel and equipment itself also consumes money.
Modularization can transfer part of the on-site labor to the factory, but at the same time it increases the importance of upfront interface definition. If factory-completed modules are inconsistent with site conditions, the room for modification may be smaller than with traditional piece-by-piece construction. Transportation boundaries, lifting paths, and assembly tolerances need to be jointly determined before manufacturing. Therefore, truly replicable modules require not only identical 3D models but also validated manufacturing, transportation, installation, and inspection processes.
The CAD 7.7 billion also cannot be directly treated as the price of a single reactor unit itself. The owner explicitly stated that the original estimate included systems and services shared by subsequent units at the entire site. Dividing this figure entirely by the power of the first unit would concentrate part of the common infrastructure cost onto the first unit; conversely, if shared works are completely removed in comparisons, the actual funding required for the first project would be underestimated.
Whether subsequent units can reduce costs depends on how experience is transferred
For subsequent units, sharing water intake, roads, and site facilities does create room for cost dilution, but this benefit has implementation conditions: subsequent units need to actually be built, the design needs to remain sufficiently stable, and the supply chain needs to continue producing. Experience from completing the first-of-a-kind unit does not automatically become a discount for the next unit. Only when rework decreases, labor hours fall, and procurement batches are formed will learning effects appear in real costs.
Prolonged construction periods for nuclear power projects also change the cost of capital. Beyond capital expenditure (CAPEX), the longer funds are tied up during the period without electricity sales revenue, the greater the pressure on financing arrangements usually becomes. However, interest rates, payment milestones, and risk-sharing terms differ, so the impact also differs. Without a new funding plan, a delay cannot be simply converted into a definite increase in electricity prices.
The levelised cost of electricity (LCOE) also depends on net generation after commissioning, service life, maintenance, and decommissioning expenditures. Small reactors can reduce the absolute scale of a single project, but they do not guarantee lower construction cost per kilowatt; more flexible phased investment also needs to match actual electricity demand growth. Only by comparing large and small reactors in the same region, under the same financing conditions and power supply requirements, can the comparison be meaningful.
Delivery credibility is more important than announcing another date
The licensing stages also need to be separated. A construction license allows specific construction activities and pre-fuel-loading system inspections, while an operating license relates to subsequent commissioning and operation. Progress already made at the construction site does not mean that fuel-loading conditions are simultaneously met. Organizing the two work streams in parallel can help reduce waiting, but it does not mean that engineering completion can replace regulatory confirmation.
The Darlington project also reminds grid planners that construction targets should not be directly counted as already available capacity. If new power sources come later than demand growth, the gap needs to be filled through other power sources, demand response, or cross-regional trading. Backup arrangements need to be prepared in advance, but it cannot be presumed from this how much additional gas-fired generation will definitely be added, because the outcome is also affected by hydropower, wind and solar, transmission capacity, and load changes.
For suppliers, this adjustment is not only negative. More complete design inputs and realistic production schedules can help reduce rush work and duplicate manufacturing. The question is whether the reassessment can form a stable basis for execution, rather than allowing budget and schedule to drift repeatedly. The owner, designer, and manufacturer need to break down the remaining work into checkable deliverables and clarify responsibilities for interface changes.
The value of the Darlington SMR will ultimately be proven by both first-unit delivery and subsequent replication. The most meaningful result at present is not announcing another more attractive date, but converting the incomplete design into constructible documents and forming credible cost and schedule on that basis. Only when the actual labor hours, rework, and commissioning experience of the first unit can be reliably transferred to subsequent units will modular cost reduction move from a promise into a measurable engineering process.
(This article is based on Ontario Power Generation's quarterly disclosures, project introductions, and Canadian Nuclear Safety Commission licensing materials, and was independently written after AI-assisted organization. The CAD 7.7 billion is the original estimate for the Canadian project cited in the company's 2026 disclosure and includes shared systems and services; the price base year was not stated in the press release, and it has not been converted into per-reactor cost or LCOE.)
Disclaimer: Information republished from partner media, institutions or other websites is provided for reference and communication purposes only. It does not imply endorsement of its views or verification of its accuracy. Please contact us if any content infringes rights or requires correction.