[en] Analysis of the hypothetical upper limit accident in the Savannah River Test Pile showed that the offsite thyroid dose from fission product release would be <2.2 x 10^-3 of the 10-CFR-100 guideline dose for 95 percent of measured meteorological conditions. Offsite whole body dose would be negligible. The Test Pile was modified to limit the length of test piece that can be charged to the pile. These modifications reduce the potential offsite dose to <2.2 x 10^-5 of the regulatory guidelines. Assessment of Test Pile safety included calculations of transients initiated by a variety of reactivity additions that were either terminated or not terminated by safety systems. Reactivity addition mechanisms considered were abnormally driving control rods out of the pile and charging abnormal test pieces into the pile. The transients were evaluated in the adiabatic approximation in which three-dimensional calculations of static flux shapes and reactivity were superimposed on point reactor kinetics calculations. Negative reactivity feedback effects appropriate for the pile and the temperature dependence of material properties, such as specific heat and thermal conductivity, were included. The results show that, for the worst initiators, safety systems can prevent the temperature rise from exceeding 1⁰C anywhere in the Test Pile. If the safety systems do not function, the pile temperatures will increase until the transient is ended by the inherent negative reactivity effects, including the melting of some fuel