Assessment of helical-cruciform fuel rods for high power density LWRs

[摘要] In order to significantly increase the power density of Light Water Reactors (LWRs), the helical-cruciform (HC) fuel rod assembly has been proposed as an alternative to traditional fuel geometry. The HC assembly is a self-supporting nuclear fuel configuration consisting of 4-finned, axially-twisted fuel rods closely packed against one another in a square array. Within the LWR core, HC fuel would in theory possess several inherent advantages over traditional fuel, potentially allowing for operation at a higher power density. Chief among these advantages are a larger surface-to-volume ratio, a shorter radial heat conduction path, and improved mixing characteristics. In previous work, computational models of the HC fuel assembly have been of limited accuracy due to the absence suitable correlations. To address needs within these subchannel analysis models, experimental measurements of rod bundle coolant mixing have been conducted with 4x4 arrays of HC test rods. The tests used the technique of a hot water tracer injection (at 95°C) into a bulk flow of cold water (at 25°C). Downstream temperature measurements were used to judge the rate of lateral cross-flow within the HC rod bundle. These tests were conducted at atmospheric pressure, and encompassed a range of mass fluxes from 1000 kg/m2s to 3500 kg/m2s, HC rod twist pitches of 200cm, 100cm, and 50cm, and different hot water injection velocities and mixing lengths. Data from over 300 tests was analyzed, yielding a best fit correlation for use with any twist pitch, rod length, or coolant flow rate. Compared to the bare rod bundle, this correlation implies an enhancement in the intensity of turbulent interchange of 40% brought about by the HC geometry, and a 1.6% forced diversion of axial flow per subchannel, per quarter-turn along the rod length. These parameters fit all data points considered within a standard deviation of 24%. Stochastic error was limited to ±16% by the use of precise temperature sensors. By applying this empirical mixing model to the subchannel representation of a BWR core featuring the HC rod design, a need to increase the flow area of the edge subchannels was demonstrated. This prompted a slight re-design of the HC fuel rod cross-section in order to make room for small spacer protrusions at the duct wall, to increase flow to peripheral subchannels. The modification was accomplished by reducing fin length, but increasing the inner diameter to maintain the reference fuel volume. The water rod region was also adjusted to maintain the reference assembly hydrogen to uranium atom ratio. With this modification, the model predicted a 24% allowable power uprate for the 200cm twist pitch HC core. Inlet and exit enthalpies were maintained from the reference cylindrical-rod core. When applied to a PWR core of HC rods, also with a fixed power to flow ratio, this empirical mixing model predicted an allowable power uprate of 47%, using traditional CHF correlations for cylindrical fuel. In subcooled conditions, CHF is known to be more sensitive to peaked areas of non-uniform heat-flux than in saturated two-phase flow conditions. Therefore power density gains will likely be dependent on the degree to which the rod twist would disrupt of nascent pockets of vapor; this effect should be further investigated experimentally. In order to further ascertain the potential gain in power density for the new design, an experiment must be carried out to obtain CHF data for the HC rod bundle. Two facilities with this aim were designed in great detail for BWR conditions: the first would operate using high pressure water at 7MPa, and the alternate would use a relatively low pressure refrigerant at equivalent conditions. The appropriate scaling laws were applied, which resulted in the choice of R134a as the simulant fluid. The R134a facility was found to be possible to construct at a greatly reduced cost.

[发布机构] Massachusetts Institute of Technology