The operability of the in-core counting subsystem was investigated at the NTR in several ways

a. Scrams were initiated at 30 kW of reactor power after various lengths of time at this

power level, and the output of the subsystem was monitored to determine if there were

any deleterious effects. Detector assembly No. 1 was operated at the core centerline 11 ft

Ф = 7 x 10 1 nv, = 2x 10 R/h) for a period of 5 minutes, after which the reactor was scrammed. Shown in Figure 8-5 is the decay curve with a 1/4-inch-thick lead shield on the detector. This curve has a final slope of 84 sec/ecade (193 sec/decade) at 4 minutes after the scram. Shown in Figure 8-6 is the decay curve after a similar time at the same power level but without the lead shield on the detector.

This curve also exhibits a final slope of 84 sec/ecade (193 sec/decade).

Counter No. 2 was operated in the same flux conditions without lead shielding for approximately 1 hour. The final slope of the decay curve after scram in this instance was 82.2 sec/ecade (189 sec/decade) (see Figure 8-7).

After fissioning has ceased, the decay curve of the delayed neutrons from the fission process contains five major groups of neutrons. The longest-lived group is from Krypton 87 (whose precursor is Bromine 87) with a 55.6-second half-life. The 55. 6-second half-life corresponds to a 80. 3 sec/ecade (185 sec/decade) slope on the decay curve. The close agreement of the experimentally determined slopes with the calculated slope demonstrates a proportionality between the neutron flux in the core and the output of the counting subsystem.

b. In addition, counter No. 1 was operated in its normal operation region prior to scram, and was immediately inserted to the core centerline upon indication of the scram

(see Figure 8-8). In this case, the indicated count rate followed the insertion and then continued to decay in a manner similar to that previously described. The slope of this decay curve, defined at 3 minutes after the scram, is approximately 65 sec/ ecade (150 sec/decade). Agreement of the slopes is not as good in this case as in the previous case. This is probably due to data not being available at long enough times after scram. However, the ability of the subsystem to follow the insertion is demonstrated.

c. A retract and insertion test was performed using detector No. 2 at the NTR facility.

The detector assembly was manually retracted from the core centerline at a reactor power level of 3 watts (7 x 10 nv at the detector) and then reinserted to the core centerline. The velocity of withdrawal and insertion was approximately 3 ft/min. The distance withdrawn was approximately 64 inches. No spurious counts or noise tran­sients were evident throughout the course of the experiment. The counting subsystem output as a function of time during this experiment is shown in Figure 8-9.

Figure 8-5 Count Rate Versus Time After Scram. Detector Assembly No. 1 (Counter) With 1/4-inch-thick Lead Shield.