Future USA development

The planning for further development in the USA at this time is a mixture of expectation and guessing. Modeling development is certain to continue, but the target reactor is uncertain. The next plant may or may not use the FFTR driver fuel design. The planning, therefore, emphasizes fundamentals and flexibility. There are many options to be modeled. The FFTF driver fuel performance in FFTF must be evaluated; both the reference and improved designs. A decision to use the FFTR driver design in the large plant will demand predictions on the effects of axial blankets, constant power (rather than decreasing) throughout life, and power changes, behavior beyond breach and design basis transients in large plants. A decision favoring a lower doubling time oxide design adds the effects of higher strength/lower swelling alloys, increased pin diameter, reduced cladding thickness/diameter, increased smeared density, gap versus pellet density, and reduced pin pitch/diameter. A helium bonded carbide design adds concern about increased potential for fuel-cladding-assembly mechanical interactions. And blanket pin performance predictions, either in a homogeneous or a heterogeneous core, add an increasing power history and enhanced assembly interactions. It is possible that the decision will be to choose a first core and retain all options for later cores. The modeling objective, for whatever options are chosen, is to predict the effect of normal and off-normal design conditions on performance limits (i.e., fuel temperature, pin deformation, pin lifetime). Several significant uncertainties in the mechanisms associated with the performance limits remain and will be addressed. These include gap closure, gap conductance and fuel properties at higher burnup, fuel-fission product reactions, retained gas, breach mechanisms, assembly interactions and behavior beyond breach, plus establishing appropriate criteria. The LIFE system, with its elements of 1D and 2D fundamental modeling, algorithms, and analysis guidelines is appropriate for this work. The thermal and structural solution procedures are stable. Modules for improved or new behavior models and properties can be added readily. The cost of running these codes is not large, compared to Physics, or compared to running tests. Further, the 'reference code' system has proven an excellent means for communication between analysts in the USA. There is room for improvement, but no major change in approach is planned. Both pin and assembly tests in EBR II and FFTF are planned to isolate the effects of design variations