The Spallation Neutron Source (SNS) has just completed its second major maintenance period since the start of production operations with beam on target. The first maintenance period dealt with some minor upgrades and corrections needed for more reliable operations, but the latest maintenance period involved significant changes to high-loss areas and consequently higher doses to workers. Predicting doses for first-time maintenance tasks is difficult since there is no "history" (recorded doses for similar tasks performed during previous outages) to refer to. This time I used a straight-forward dose estimating process and ended up with an embarrassingly conservative estimate.
To inject beam into the accumulation ring, H- ion mini-pulses pass through a primary stripper foil and are then (as bare protons) more-or-less gracefully merged into the circulating H+ pulse. The relatively few H- ions that are not successfully stripped in the primary foil pass through a secondary foil and are then diverted to a beam dump. Currently this "Injection Area" is the highest-loss section of the accelerator and so has the highest activation levels when beam operations stop and personnel enter the beam tunnel. For planning purposes prior to the six-week maintenance period, dose rate estimates in the Injection Area were based on measurements made during the preceding (one-day) weekly maintenance days. The exact tasks to be performed during the maintenance period were uncertain since several components were ordered and promised, but not delivered by the start of the maintenance period. Finally, the number of man-hours needed to perform each task was conservatively estimated to be sure the limited technician resources were not over-committed. Adding up all the dose numbers under those assumptions, the estimated total dose for the maintenance period came out to be at least 2 rem. Considering that the entire SNS project had a cumulative dose of only 0.5 rem for calendar year 2006, this estimate drew some attention.
The as low as reasonably achievable (ALARA) dose-reduction process began to have an effect even before the maintenance period started. Accelerator Physics personnel were performing some tuning studies at the end of the operational period, but they reduced the beam intensity for these studies, partly to reduce activation in the Injection Area. This also started the "cool-down" period while beam was still being run. One vacuum chamber that had been ordered did not arrive in time for installation this period, so not all planned tasks could be carried out. For the upgrades that were completed, removal of some of the activated parts significantly lowered the dose rate in the area for later tasks. Finally, work planners did a very good job of scheduling work to take advantage of further cool-down during the outage, and tasks were completed faster than estimated. In the end, the total dose for maintenance work in this outage was a cool 250 mrem, barely more than 10% of the original estimate!! The good news is that the project gained valuable experience at improving on work planning, initial dose estimates, and working together to reduce dose to workers. The only downside is that my initial cumulative dose estimate was too conservative, and we will strive to learn how to predict doses for new tasks more accurately while maintaining effective ALARA planning.