The problem to be resolved in this work was the use of radial control drums as the sole active reactivity control system for nuclear thermal propulsion, which results in significant rocket performance changes during full-power operation. This can result in large inefficiencies in propellant usage, inaccurate estimations in Isp and thrust, and can be a dangerous operation requiring continuous active control of the reactor given the unstable nature of current nuclear thermal rocket reactor designs.
The innovation described here eliminates the active movement of the radial control drums during full-power operation. The innovation mixes ppm quantities of burnable neutron poison into the existing structural material of the nuclear reactor for nuclear thermal propulsion to passively control the reactivity of the core.
The innovation consists of adding ppm quantities of Gd (enriched or natural isotopic composition) to existing components in the nuclear thermal rocket. By controlling the spatial self-shielding, it was possible to attain a linear depletion rate of the neutron poison that matches the reactivity changes due to the production of xenon-135 (stable and meta-stable states) and other fission products, and the depletion of the fissile material. The result is that a flat reactivity profile is attained without any operator input, removing the need for radial control drum movement during operation.
In order to counter the drop in reactivity found during full-power operation due to fuel depletion and fission product accumulation, BORGalloy (Burn able-poison Operating a Reactor with Gadolinium alloy), a burnable neutron poison doped alloy, was introduced into the reactor. A burnable neutron poison is an isotope that has a large neutron absorption cross-section that is converted into a non-neutron-absorbing isotope with the absorption of a neutron. The premise of the concept is that as the neutron poison is depleted, there will be a resulting increase in the core reactivity, which, if done correctly, can be tailored to match the reactivity reduction from the fissile depletion and fission product buildup.
The poison is Gd dispersed in minute quantities in the outer tie tube. The poison was selected because of its extremely high absorption cross-section and its conversion to an isotope that has a comparatively much lower absorption cross-section. When the poison is introduced into the core such that it has minimal self-shielding (maximum exposure to the core’s neutron flux), it can be rapidly depleted and result in an appreciable change in reactivity. Additionally, the low self-shielding ensures that the depletion rate remains relatively constant for all burns, eliminating the need to replace the poison at the beginning of each burn.
Various locations were explored, including the moderator sleeve, the inner and outer tie tubes, and the fuel matrix. Of these, the outer tie-tube was selected as the location of choice. This is due to its thinness and its reduced role as a structural element when compared with the inner tie-tube. The thinness of the component reduces the spatial self -shielding of the poison, and the reduced need to provide structural support minimizes the chance that additions of Gd to the material will reduce its strength below acceptable levels. While the fuel matrix was seen to be promising in terms of self-shielding, particularly for the graphite composite matrix, it was decided that the outer tie-tube was preferred due to the exponential increase in development costs associated with fuel development.
With the identification of the burnable poison, it was implemented into two LEU fueled NTP cores to flatten the reactivity profile during full-power operation: LEU tungsten fueled (SCCTE) and LEU graphite composite fueled (SULEU). Through the variation of the Gd content, it is possible to achieve a near-flat reactivity change during full-power operation for the TMI-1, MOI, and EOI for both cores. It was found that 20 ppm to 200 ppm Gd is required to achieve near-flat reactivity profiles for TMI-1, MOI, and TEI, depending on the isotopic enrichment of the Gd.
It is important to note that while the use of burnable poisons in terrestrial reactors is a well-known technology, their implementation typically involves the burning of the poison over the course of months, rather than minutes. Consequently, the rapid burn-up behavior of BORGalloy, while computationally demonstrated, needs experimental validation to demonstrate that a noticeable and rapid depletion of the poison is achievable, and that the poison can in fact be distributed in the material in a uniform and predictable manner within calculated tolerance levels.