AP1000 remains attractive option for US market, says MIT

The delays and cost overruns faced by the Vogtle AP1000 project have "decimated US energy utilities' interest in large nuclear power plant construction projects. The general energy sector also utilises the experience at Vogtle as an indicator of nuclear energy's high cost and infeasibility of the role it can play in future energy markets," the report notes. It says several unique project parameters have led to the inflation of the Vogtle total project cost (currently standing at USD28 billion), including high interest rates, lack of detailed design prior to start of construction, construction management turnover and first-of-a-kind (FOAK) issues.

Construction of the two Vogtle AP1000 units began in 2013: unit 3 in March and unit 4 in November. Southern Nuclear and Georgia Power, both subsidiaries of Southern Company, took over management of the project to build the units in 2017 following Westinghouse's Chapter 11 bankruptcy. Vogtle unit 3 is now expected to be in service by the end of the first quarter of 2023, with unit 4 following by the fourth quarter of 2023.

Four Westinghouse AP1000 reactors are already in commercial operation at Haiyang and Sanmen in China.

"Almost 10 years after the start of construction in the US, the AP1000 is now a proven technology with four operating plants in China," MIT said. "The next AP1000 plant has potential to provide a viable product in the US and oversees if its original nth-of-a-kind (NOAK) capital cost and construction schedule projections by Westinghouse can be realised."

The 'should cost' decouples the impact of cost by competency of the reactor vendor, supply chain logic and construction execution from the design architecture, the report said. It is also essential to ensure comparison of like costs (FOAK or NOAK) when evaluating future technology selections, in recognition that most publicised costs are NOAK exclusive of owner's and financing costs.

MIT estimates the overnight capital cost for Vogtle 3 and 4 at USD7956/kW. It says the 'should cost' of the next AP1000 overnight capital cost in the USA to be USD4300/kW and USD2900/kW for the following 10th unit (online by around 2045), deployed in series, based on 2018 dollars.

"Contrary to perhaps the consensus verdict among US utilities, the AP1000 NOAK estimated overnight capital cost, which still requires large capital investment to realise, continues to make AP1000 an attractive option for nuclear energy development globally," it said. Globally, while some one-time FOAK issues will be reintroduced, "the lower labour rates, owner's costs and indirect costs relative to the US, can make AP1000 an affordable technology for displacing existing carbon emitting power plants."

The report says constructing more than 10 reactors consecutively allows achieving low costs for the direct portion of the capital cost (minimises rework), but particularly reduces indirect cost (sharing of engineering and management experience among units).

Competitive with SMRs

The study includes a comparison of the AP1000 technology relative to other near-term nuclear technologies, including large light-water reactors (LWRs) and small modular reactors (SMRs).

The AP1000 is an attractive technology for large scale decarbonisation, MIT said, since it "features a compact design in terms of amount of concrete and steel used per MWe generated compared to leading large LWRs and SMRs (i.e., lowest direct 'should cost') while still generating more than 1000 MWe of carbon free electricity (i.e., low O&M cost)."

It said overnight costs per kWe of SMRs are estimated to be 1.4-1.75 times the cost of the next AP1000 plant "because of the lack of economy of scale".

According to MIT, "SMRs are an attractive option for certain markets where small additional capacity of carbon free energy is needed. However, if multiple SMRs are housed in a single reactor building (e.g., NuScale), then no measurable reduction in overall onsite labour input compared to AP1000 is expected and the capital cost will be higher than a large reactor due to the large volume of concrete and steel per MWe produced."