Measurement of argon concentrations in a TRIGA Mark-III pool

Argon-41, the principal radioactive effluent from a pool type reactor during normal operation, is produced by the 40A (n,γ) reaction. The reactant, 40A, is introduced into the pool water by contact with the air. Reduction in radioactive argon release can be accomplished by reducing the concentration of dissolved 40A and retaining the 41A within the pool. However, little data were available concerning the mechanisms of argon introduction, production, retention, and release from a reactor pool. Experiments have therefore been performed at the Torrey Pines TRIGA Mark-III Reactor to develop techniques to sample dissolved argon and to provide data on argon concentrations in the pool for release modeling studies. Significant results for argon dissolved at different pool depths can only be obtained if the water samples are sealed at the point of collection. A special handling tool was developed to perform this remote operation. Pool samples were counted for 41A soon after collection with a NaI spectrometer. After allowing one day for decay of 41A, the concentration of 40A in the water sample was determined by neutron activation analysis. In each case, the 1.29 MeV gamma-ray peak of 41A was used. Interference from the 1.37 MeV 24Na peak was considered and its effect subtracted after determining 24Na content from the 2.75 MeV 24Na peak and a sodium standard. A Ge(Li) detector was tried and found to eliminate the problem, but it introduced an unacceptable geometrical effect dependent on bubble size within the sample bottles. Samples were taken from the 27 ft deep TRIGA pool at various locations. Results were obtained for samples taken on several different days along the same vertical line about 3-1/2 ft from the reactor centerline. Temperature measurements along this vertical traverse indicated a sharp temperature gradient at about 15 ft below the surface (∼6 ft above the top of the reactor). The 40A concentration increases slightly below this point, but the distribution is relatively more uniform throughout the pool than assumed by the usual argon release model. The 41A count rate distribution, a relative measure, suggests that the dissolved 41A distribution is also relatively uniform throughout the pool, an apparent consequence of the recirculation rate provided by the pool cooling system. Although unsubstantiated, reductions in 40A concentration might be achieved by redesign of the pool skimmer and the diffuser pump to reduce air input into the lower pool region. (author)