Tuesday, December 25, 2007

PBMR seeks Mitsubishi investment in new reactor

Demonstration plants are expensive

South Africa has asked for funding from Mitsubishi Heavy Industries in Japan to finance all or part of 24 small nuclear power installations, at 160 Mw each, to be built by the pebble bed modular reactor (PBMR) project. They will need a lot of money because the rollout of this completely new reactor design goes far beyond building a prototype.

The challenges for South Africa are;

  • reduce the high cost and design of the demonstration reactor to commercial scale so it can be built in large numbers for domestic use
  • market it for export at a competitive price point
  • insure a profitable return to Mitsubishi for its investment

The report on the pitch to the Japanese industrial giant was published by Reuters and Business Report, a South African news outlet. They both cite the Nihon Keizai Shimbun, a Japanese newspaper, which reported earlier this month that Alec Erwin, South Africa's minister of public enterprises, had said he hoped the company would make a decision in a few months, and that Mitsubishi Heavy Industries had said it would consider the government's request. That's business talk for "can we make money on this?"

Vimla Maistry, spokesperson for the department of public enterprises, responding to a reporter's inquiry, said it had "always been government's policy to attract investment from strategic partners that had both funding and other advantages to offer the PBMR project. The turbines for the PBMR plant will be supplied by Mitsubishi, so Mitsubishi therefore is a desirable strategic partner."

Mitsubishi already in the picture

In 2005 South Africa has awarded Japan's Mitsubishi Heavy Industries (MHI) two contracts worth $15 million to help build a demonstration advanced nuclear reactor. MHI is said to be focused on supplying parts for the core barrel assembly, an integral component within the reactor pressure vessel of the new reactor design, and on turbines for power generation. It will also provide professional services for the design process that will form part of the reactor's demonstration plant to be built at Koeberg, South Africa's existing nuclear facility near Cape Town.

South Africa plans to build a PBMR as it races to find new energy sources to meet growing domestic demand for electricity. The country is facing a potential energy deficit. Assume demand grows, the South Africans have to ramp up their electric generation capacity and the apparent strategy is that it will be with home grown technology hence the commitment to the new reactor design

South Africa also plans to deploy the PBMR for export. In 2006 the South African firms involved in developing the reactor submitted white papers to the U.S. Nuclear Regulatory Commission to begin the dialog of getting the design accepted in the U.S. A full submission of a PBMR reactor design to the NRC is said to be several years in the future. Even if no PBMRs are ever sold in the U.S., approval of the design by the NRC is the gold standard for selling the reactor elsewhere.

Big plans will create big costs

South Africa has aggressive plans to build up its nuclear energy portfolio. According to Business Report, it could cost between $9.9 billion (R67 billion) and $13.8 billion to build 24 reactor installations, which together could generate 3,960 megawatts. That's expensive power coming in at $3,500/Kw at the upper end of the cost estimate.

The country's top nuclear managers say nuclear technologies are likely to become progressively more expensive as a result of bottlenecks in the global supply chain. Pricing of the PBMR has not been determined, but it will have to be offered at a lower price per kilowatt to compete with larger, conventional plants.

At $1,000/Kw, or less than half the price of a conventional reactor, a 200 MWe PBMR would cost $200 million. If it is priced near the current limit of $2,500/Kw, the units will cost $500M and won't be as competitive. If utilities are going to spend that much per kilowatt, they are going to want to be able to meet much greater levels of demand than what's available from a 200 Mw plant.

Additional competitive issues are that the design is not licensed by the NRC, and it would take longer than conventional light water reactors to complete a review because the technology is radically different. Utilities will need references to guide them to develop operating procedures and train a workforce. A utility buying a PBMR would have to estimate the additional costs of deploying a first-of-kind reactor which is very different than conventional plants. The sheer scope of regulatory and technical overhead in deploying a completely new reactor design will likely stretch out the time frame for its commercial acceptance.

Tom Ferreira, a PBMR spokesman, said: "We are pretty confident that the demonstration plant will cost R16 billion." [$2.36 billion]. Though he didn't say it, the fact is the demonstration plant will have to identify substantial cost reductions in the production model to attain a competitive standing in the global nuclear export market.

Recent efforts to price the cost of a demonstration plant in the U.S. for a high temperature gas cooled reactor [large graphic] using a "pebble bed" type design have also come up with higher than expected costs. Earlier this month Platts reported that demonstrating the commercial viability of a very high-temperature gas-cooled nuclear reactor capable of producing both electricity and hydrogen could cost between $3.8 billion and $4.3 billion. The estimates come from three contract teams competing in the Department of Energy's Next Generation Nuclear Plant (NGNP) project. INL conceptual designs have considered options for 300 MWe and 600 MWe plants.

The Department of Energy has been working on the idea the NGNP, or something like it, is to be built in Idaho for a long time according to a report in Time Magazine from January 1989. [Hat tip to Robert Hargraves]. You have to ask the question -- when is the government going to stop planning and estimating, and actually build something? Imagine how much cheaper the NGNP plant would have been if they'd gone ahead in 1989?

Here's a novel idea. Why don't the PBMR folks and the Department of Energy get together and build one prototype in Idaho and save everyone a boatload of money? Construction of NGNP at Idaho isn't scheduled to begin for at least another 10 years. In 2006 lab research managers told Congress the NGNP would be ready for operation in 2021. Maybe a partnership with South Africa, China, or both, would speed things up?

Place your bets

The bet the South Africans are placing is that it will be more cost effective to build a PBMR than conventional nuclear power. The key difference is that it could take two years to build a pebble bed unit, while a conventional nuclear plant could take four to six years or longer. China is also developing the pebble bed technology and may be betting on the same commercial factors.

Before a PBMR can operate commercially a demonstration plant will be built in South Africa. Last September the project suffered a two year delay when it was determined design documents weren't sufficient to support manufacturing. Startup of construction work was pushed back from 2007 to 2009.

A PBMR spokesman told Business Report that if environmental approval and licenses were received in 2008, construction on the fuel and demonstration plants could begin in 2009 and they could be commissioned in 2013 or 2014.

1 comment:

Rod Adams said...


If I were a betting man, I would choose to put my money on a company led by an aggressive entrepreneur without competing technologies to develop and market.

Hmm - wonder where I could find such a company?

Rod Adams
Adams Atomic Engines, Inc.