Spectral Shift Absorbers for Fast Spectrum Space Nuclear Reactors

Abstract

The space nuclear reactors being considered to support the Jupiter Icy Moons Orbiter (JIMO‐1) mission‐1 sometime in the next decade are compact and fast spectrum with void fractions ranging from 20–40%. In order to secure launch approval, it has to be demonstrated that these reactors will remain sufficiently subcritical when submerged in water or wet sand and subsequently flooded with water, following a launch abort accident. The resulting shift in the neutron spectrum towards thermal increases reactivity, potentially making the reactors supercritical. Incorporating “Spectral Shift Absorbers” (or SSAs), elements such as boron, europium, gadolinium or rhenium, which have significantly higher absorption cross‐sections for thermal versus fast neutrons, can offset the reactivity increase. It has always been the assumption that the worst‐case submersion accident is with a fully flooded reactor; however, this work shows that, depending on the type and amount of SSA in the reactor, a submerged but unflooded reactor core could be more reactive, raising a safety concern. This condition, referred to as “reactivity inversion”, is investigated for the following SSA elements: boron, boron‐10, samarium‐149, europium, europium‐151, gadolinium, gadolinium‐155, gadolinium‐157 and rhenium, by varying the ratio of SSA to fissile atoms in the reactor. The effect of placing a coating of SSA material on the outside of the reactor vessel is also investigated. Gadolinium‐157 is the most effective SSA material investigated as a core additive, yielding the greatest increase in the safety worth of a submerged space reactor with a comparatively small decrease in excess reactivity. Samarium‐149 yields a similar increase in safety worth, but with a much larger decrease in excess reactivity. Natural europium and gadolinium are also promising: however, the decrease in the excess reactivity is greater with europium than with gadolinium. Boron and boron‐10 are not particularly attractive as core additives, but boron‐10 yields the greatest benefit as a coating. Rhenium is useful only if the desired safety worth is small (< $3) and at a significant mass penalty.