Over the last three decades most major industrialised nations have looked to expand their nuclear power generation capability. But throughout this time, the US, where no new nuclear reactor has been built since 1979, the year of the Three Mile Island accident, has remained committed to coal.

But now it seems as if this hiatus will soon end. In July, the Constellation Energy group submitted the first application to build a new US reactor in three decades, and roughly 20 more utilities have some serious nuclear sites. What is more, two US utilities, Dominion and Entergy, have already ordered core components in advance of anticipated applications.

It may be significant that both utilities chose GE-Hitachi Nuclear Energy as their vendor. The two industrial giants formed their joint venture in May. GE-Hitachi is the latest in a series of corporate alliances focused on commercial nuclear power.

It is also the first one essentially led by an American firm: the JV is split into Japanese and global divisions, and the latter unit is majority owned by GE.

Given the longstanding relationships between utilities, national governments and businesses focused on nuclear technology, reactor vendors are likely to have at least some home-court advantage.


The vast resources required and risks involved in reactor projects make corporate alliances highly appealing. Indeed, the reactor fabricator industry now comprises three major Western-Japanese alliances:

  • Areva (formerly the French Framatome) and Mitsubishi Heavy Industries officially partnered this July
  • Toshiba bought a controlling stake in Westinghouse in February 2006, and the Shaw Group took a 20% interest that October
  • GE and Hitachi

Notwithstanding their new JV, GE and Hitachi have cooperated for decades. In the 1980s, GE and Hitachi jointly developed reactors for use in Japan, with Hitachi manufacturing critical components for a GE-developed design. Four such units entered service in the mid-1990s and several more are being built in Asia. In 2000, both companies established GNF, a JV to develop and produce nuclear fuel.


Like most JVs, the GE-Hitachi alliance seeks to capitalise on complementary but distinct skills. According to Andy White, head of GE-Hitachi Nuclear Energy, “there are relatively few overlapping areas in our respective nuclear businesses”. Generally speaking, American and European firms bring design expertise, while Japanese firms are especially strong at construction and fabrication.

“The emergence of many utilities looking to build new units within the same timeframe could create a serious bottleneck for the industry.”

Regarding the Entergy deal, White stated that much of the design, engineering and procurement for the project will occur at a North Carolina GE centre, which is currently tripling in size.

The GE-Hitachi alliance also plans to spend roughly $400m for nuclear plant designs and certification. According to John Krenicki, head of GE Energy, these designs have been in development for over a decade, and should be complete by 2010. A 15-year, $1bn design cycle may sound hefty but the Dominion project alone is worth at least $1.5bn, according to Japanese sources.

The high stakes of operational problems, along with low volumes, high unit costs and large fixed investments, give the reactor business feast-or-famine characteristics. Indeed, Hitachi took a loss last year because of costs associated with repairing faulty turbines it had installed at two Japanese nuclear plants. In 1Q07, by comparison, Hitachi’s power and industrial systems units posted 22% higher revenues and 123% higher operating profits, helped significantly by rising nuclear power sales
in Japan.

Although the nuclear plants crippled by the July earthquake in Japan used GE and Hitachi designs and equipment, the primary fault lies with TEPCO, the utility – among other things, the quake’s fault line apparently lay directly beneath the site and was not discovered prior to construction.


Consequently, reactor design globally has historically proceeded in relatively discrete phases known as ‘generations’. Four generations of reactors are generally defined on technological rather than chronological grounds.

Generation I (1950s-1960s) reactors were prototypes; outside the UK, none are still operating. Generation II (1960s-1990s) reactors comprise the bulk of operational units worldwide. Most of these reactors were derived from designs that were originally developed for naval use, and are therefore inherently suboptimal for land use.

Generation III (1980s-2000s) reactors represent the current state of the art, in which commercial land reactor designs were created from scratch. The industry makes a further distinction in the form of the ‘Gen 3+’ family, which is supposed to provide Gen II benefits with shorter implementation times and lower capital costs.

“Reactor design has historically proceeded in relatively discrete phases known as ‘generations’.”

Generation IV designs are still in the conceptual stage and will not be operational before 2020 at the earliest.


Freed of legacy constraints, Gen III designs are simpler and more robust. This makes reactors easier to operate and increases not only ongoing output reliability but also operating life (to roughly 60 years).

Gen III designs are also standardised by reactor type. By creating fixed modular templates for various reactor sub-systems, vendors and customers can reduce licensing and construction times, reducing capital cost.

Exploiting advances in fuel fabrication, Gen III designs achieve a better burn rate, which both improves fuel economy and reduces waste creation.

Most critically, Gen III designs use passive safety systems based on universal physical laws such as gravity, rather than the active systems prevalent in Gen II designs. Unlike passive systems, active systems could themselves malfunction, and thus lead to a catastrophic ‘failure cascade’ (such as Chernobyl). In contrast, GE’s Gen 3+ reactors will have ceiling tanks for cold water, which, in emergencies, can cool the core without relying on pumps.


Generations are also different from reactor types, which, in terms of currently feasible technologies, are classified on the basis of whether:

  • the heat exchange medium is light water, heavy water, gas, or some other substance
  • the circulating water performing the actual heat exchange is boiling or pressurised, and if the latter, at what temperature range
  • the fuel required is uranium, plutonium or mixed oxide (MOX)
  • the reactor is a breeder (which produces fuel as a by-product of the fission reaction)
“The elevated volatility that characterises current financial markets can actually work to the advantage of reactor vendors.”

To illustrate this classification scheme, consider GE’s historical reactor offerings. In Gen II, GE chose the boiling water reactor (BWR), using light water, as its signature design type. The BWR evolved into the Gen III advanced BWR (ABWR), a fully certified 1,500MW-range reactor that GE describes as the foundation of its nuclear portfolio.

GE’s Gen 3+ offering is the simplified BWR (ESBWR), which is currently in the US certification process. As its name implies, this design is simplified (e.g. 25% fewer pumps) and presumably more economical due to a claimed 36-month timeframe for ‘first concrete to first core load’.


The importance of shorter decision-to-operation cycles is underscored by the timelines of the Dominion and Entergy projects. In 2003, Dominion applied for an early site permit for a new reactor at its Virginia plant. Official NRC approval is expected (but not guaranteed) by the end of this year. In early May of this year, however, Dominion reached an agreement with GE to secure certain critical ‘long-lead’ components, such as large forgings and turbine island parts.

Why strike a deal before receiving approval, and even before filing for the Combined Operating License (COL), which Dominion plans to do later this year?

According to Mark McGettrick, Dominion’s CEO, the emergence of many utilities looking to build new units within the same timeframe could create a serious bottleneck. “We believe the market for these components will be tight in the near future, so getting to the head of the line makes good sense,” stated McGettrick.

Although Dominion will sign a formal contract with Hitachi-GE by September—still prior to approval—actual reactor construction would not begin until 2010, and the reactor would not become operational until 2014 or 2015.

Entergy, which evidently did not request an early site permit, faces an even longer timeline. The utility signed a deal this July with GE-Hitachi to provide long-lead components for its facilities in Mississippi and, possibly, Louisiana. Entergy plans to submit an application for the former by year-end, and the latter by mid-2008.

According to Entergy, committing now is necessary to guarantee access to fabrication capacity should approval come as expected. Even so, construction is slated to begin in 2017.

“The vast resources required and risks involved in reactor projects make corporate alliances highly appealing.”

Given the time involved, the extra output needs to be sizeable. Dominion’s Virginia site currently has two 910MW reactors; at 1,500MW, the proposed ESBWR would increase total facility output by 80%.

Furthermore, adding new reactors to existing facilities, which by definition have COLs, is easier from a regulatory perspective than obtaining approval for a brand new reactor site. Entergy’s facilities also house operating reactors, which were built in the 1980s.


Some observers have recently voiced concern that rising prices for raw materials such as copper, nickel, and steel could raise construction costs. However, not all commodity price increases are meaningful in the long term, notes Ron Lawson of Logic Advisors, a commodities boutique.

“Speculators and hedge funds have pushed the [base metals] markets around recently,” observes Lawson, who notes that the hot-money search for oil-market replays in particular and alternative investments in general can turn around very quickly, as illustrated by the current fallout from the US subprime lending meltdown.

Ironically, the elevated volatility that characterises current financial markets can actually work to the advantage of reactor vendors. Given the timeframes discussed earlier, ‘current prices just aren’t relevant’, according to a source at a large engineering firm.

With confirmed orders three years in advance of actual physical needs, reactor vendors that stay patient can potshot the market for raw materials, buying on dips and stepping aside on spikes.