GE-Hitachi opted to pursue its first BWRX-300 project in Canada because CNC’s regulations allow construction permit applications to be submitted with much of the design still incomplete. Last year, GE-H and TVA tried to get the NRC to “align” its requirements for a construction permit application with CNC’s less stringent standards, but NRC refused.
The jury is very much still out on whether the BWRX-300 will prove feasible to build on-time and on-budget, but what we know so far is not encouraging. The $20.9 billion price tag amounts to about US$15 billion, or US$12.5 million/MW. Adjusted for inflation, that is more than the projected cost of Vogtle 3&4 at the time NRC issued the COLs to Southern Co. in 2013. Westinghouse’s design was also still incomplete at the time, which is blamed for many of the delays and cost overruns. The fact that GE-H and OPG appear to be following the same playbook provides no reason for optimism. There are plenty of reasons they would feel compelled to move forward without heeding the warnings, including pressure on OPG to justify its commitment to nuclear when renewables and storage are clearly more economical and able to be deployed quickly and on time. OPG is making similar decisions regarding license extensions for the old Pickering reactors, as well. And TVA management appears to have made a similar decision under pressure from the White House and the TVA board to submit the construction permit application for the Clinch River BWRX-300 project despite the indications that the reactor design is not complete enough to meet NRC’s standards. (Or else, why did TVA ask NRC to “align” its standards with the CNC’s?)
Also, OPG’s unit-by-unit breakdown of the BWRX costs appears to be manipulating numbers to paint a picture through rose-tinted lenses. For instance, OPG removes “shared infrastructure” from the costs of each reactor, even though none of the reactors can operate without that. Then, after subtracting the “shared” costs, it shows the unit costs of each reactor rapidly declining from the first to the second and third reactors, and cost reductions starting to level off with the fourth. That implies the “economies of scaling” taper off after three reactors, which is extremely speculative. The economics of SMRs are predicated on “nth-of-a-kind” economic models that predict cost reductions to levels that would be competitive with other generation sources once manufacturing scales to a certain number of units. But no one believes that the “n” number is 4. Typically, it is in the dozens or even 100 or more. If, as it appears, GE-H intends to build the four reactors basically in parallel (similar to Vogtle), then there is no reason to believe the costs of manufacturing and building each reactor will differ as significantly as OPG portrays. They seem, in fact, to be copying a similar narrative DOE tried to spin about the costs of Vogtle 3 and 4 in a report published last summer, which claimed unit 4 cost significantly less than unit 3 — even though they were completed on essentially the same schedule. In reality, the whole point to building multiple reactors at the same site is to achieve cost reductions through shared infrastructure and lower costs for land acquisition and other construction and permitting costs. But, historically, we have seen plenty of cases in which subsequent units increased in costs rather than decreased.
It would be more honest for OPG and GE-H to present the unit costs of the BWRX-300 project with the “shared” infrastructure included. In reality, the four reactors are being proposed as one project. Looked at that way, they are building 1,200 MW of nuclear generation for about US$15 billion. That is pretty close to the final costs of each of the new Vogtle reactors, but at the start of construction. We can only anticipate costs escalating from here.