The FuTure oF nuclear Fuel cycle - MIT Energy Initiative

urnups, but the gap between the necessary fuel burnup for a technically viable reactor and
the experience base has narrowed. The TerraPower reactor is a modified sodium fast reactor
but the concept would be applicable to lead-cooled fast reactors (see below). Beyond technical
feasibility is the separate question of economic viability of the fast reactor.
The concept of a once-through sustainable reactor with in-situ breeding is not new. The
concept of a reactor that could breed its own fuel inside the reactor core was initially proposed
and studied in 1958 by Saveli Feinberg [1]. This concept is often referred to as the
breed and burn reactor concept. Driscoll et al. published further research on the concept in
1979 [2], as did Lev Feoktistov in 1988 [3], Edward Teller/Lowell Wood in 1995 [4], Hugo
van Dam in 2000 [5], Hiroshi Sekimoto in 2001 [6] and Yarsky et al. in 2005 [7]. Previous
studies have shown that the primary technical challenge is fuel cladding that must survive
very high neutron fluences and fuel burnups to achieve a steady-state breed-and-burn condition.
Recently, TerraPower, LLC has developed methods to achieve steady-state breedand-burn
cores that significantly reduce the neutron fluence and burnup, and consequent
core materials degradation. Their proposed traveling wave reactor (TWR) with lower materials
requirements was achieved by a combination of core design features and engineering
accommodations that would need to be demonstrated in a prototype TWR.
Various concepts are under exploration from a small modular design rated at hundreds of
MW(e), to a large monolithic power plant for baseload electrical power production of a
gigawatt or more to address the potential range of applications. The first versions of TWRs
are based on elements of sodium-cooled fast-reactor technology [8-10]. The core design
that emerged as most promising has an approximate cylindrical core geometry composed
of hexagonally shaped fuel bundles (assemblies) containing a combination of enriched
and depleted uranium metal alloy fuel pins clad with a sodium thermal bond in ferriticmartensitic
steel tubes. Depending on power density, after a predetermined time (e.g., one
to two or more years) of core operation, the reactor is shut down in order to move highburnup
assemblies to low-power regions of the core, replacing them with depleted uranium
assemblies. This fuel shuffling accomplishes three functions: (1) controls the power
distribution and burnup so that core materials remain within operating limits, (2) manages
excess reactivity, in conjunction with control rods and (3) extends the life of the reactor
core because core life is largely determined by the number of depleted uranium assemblies
available for shuffling.
All the fuel movements are accomplished in a sealed reactor vessel, which contains enough
depleted uranium assemblies to support reactor operation for the plant lifetime. After the
initial small number of enriched fuel assemblies initiate reactor operation, enough fissile
material in depleted uranium assemblies will have been bred so that the core continues to
run on depleted uranium until the end of plant life. The feed material (except for the startup
enriched uranium fuel assemblies) can be depleted uranium (over a million tons in world
inventory), natural uranium, or LWR SNF after conversion to a metallic form (without
separation of radionuclides). The fuel cycle costs would be expected to be lower than for
any other reactor.
The technical issue is whether the fuel can obtain a high burnup required to breed enough
fuel to maintain reactor operation. The experimental data base on fuel cladding fluence
limits does not go to the required burnup; thus, it is not known whether existing materials
could meet these requirements or new materials would be required.
196 MIT STudy on The FuTure oF nuclear Fuel cycle