In-situ recovery (ISR) brings uranium-bearing solution to the surface, where ion-exchange resins capture the dissolved uranium. Mine operators need to know when the resin is approaching saturation, when elution has removed most of the uranium, and how other elements are changing during the process. Standard laboratory methods can provide accurate measurements, but sample collection, preparation and analysis may take hours to days, limiting their use for timely process decisions.
Researchers led by Prof. Daqian Hei at Lanzhou University, together with colleagues at East China University of Technology, the Beijing Research Institute of Chemical Engineering and Metallurgy, and Nanjing Tech University, have developed a compact built-in prompt gamma-ray neutron activation analysis (PGNAA) system for measurements beside a resin adsorption tower. Reported in Advanced Scientific Instruments, the sealed cylindrical measurement head is 237 cm long, 17.5 cm in diameter and weighs about 80 kg. It was installed vertically in a bypass loop. By comparison, the conventional cross-belt PGNAA analyzers considered in the study typically occupy more than 6 m3 and weigh over 6,000 kg.
The instrument uses two independent measurement routes. For uranium, two He-3 proportional counters record thermal and epithermal neutrons. The ratio of epithermal to thermal counts is used to quantify uranium from prompt fission neutrons while reducing the influence of changes in neutron output. For non-uranium elements, a LaBr3(Ce) detector records the lower-energy portion of the prompt-gamma spectrum and a BGO detector records the higher-energy portion. Combining the two spectra retains the better low-energy resolution of LaBr3 and the higher detection efficiency of BGO.
Wet resin and process solution contain abundant hydrogen, and changes in composition, density or moisture can alter the neutron field inside a sample. To account for this matrix effect, the researchers applied the standard single-element response spectrum (SSERS) method to the combined gamma spectrum, using the 2.223 MeV hydrogen capture peak as an internal reference for neutron self-shielding correction. The SSERS regression is used only for non-uranium gamma-spectrum analysis; uranium is measured independently by the prompt fission-neutron channel.
In laboratory calibration with eight uranium standards from 0.161 to 15.28 g/L, the neutron count ratio showed a linear response with R2 = 0.9965. For a 60 s measurement, the reported uranium detection limit was 0.129 g/L at 95% confidence, and the Type A relative uncertainty was below 3.26% within the calibration range. Repeated measurements of mixed standards gave relative standard deviations below 5% for major elements and below 10% for trace constituents.
The system was then demonstrated in a bypass loop at a commercial uranium ISR mine. The team obtained about 100 daily datasets covering a first adsorption stage, elution, resin transformation and a second adsorption stage. During the approximately 80-day first adsorption stage, the reported uranium concentration rose through four stages and reached a maximum of 74.1 g/L before stabilizing near 70 g/L. During about five days of elution, it fell to below 0.3 g/L. Calcium, iron, silicon, magnesium and aluminum generally increased during adsorption and decreased during elution, while sulfur and chlorine showed the opposite trend. These are elemental measurements and do not directly identify ionic species.
The authors note that the current setup samples a side stream rather than the full tower inventory. The results instead show that the compact system can provide on-site trends to support decisions such as when to switch a resin tower or end elution. Future work will focus on reducing the probe diameter for direct insertion into standard process piping.
Experimental study