Prompt fission neutron uranium logging (I): Direct uranium quantification method theory

摘要: Prompt fission neutron uranium logging (PFNUL) is an advanced method for utilizing pulsed neutron bom- bardment of the ore layer and a fission reaction with uranium (235U) to detect the transient neutrons produced by fission and then directly measure and quantify uranium; however, the stability and lifetime performance of pulsed neutron sources are the key constraints to its rapid promotion. To address these problems, this study proposes a PFNUL technique for acquiring the time spectrum of dual-energy neutrons (epithermal and thermal neutrons) from the upper and lower detection structures and establishes a novel uranium quantification algo- rithm based on the ratio of epithermal and thermal neutron time windows (E/T) via a mathematical-physical modeling derivation. Through simulations on well logging models with different uranium contents, the starting and stopping times of the time window (∆t) for uranium quantification in the dual-energy neutron time spec- trum are determined to be 200 and 800 μs, respectively. The minimum radius and height of the model wells are 60 and 120 cm, respectively, and the E/T values in the time window show an excellent linear relationship with the uranium content. The scale factor is KE/T = 1.92 and R2=0.999, which verifies the validity of the E/T uranium quantification algorithm. In addition, experiments were carried out in the Nu series of uranium standard model wells, and the results showed that under different neutron source yields, the E/T-based uranium quantification method reduced the relative standard deviation of the scale factor of the uranium content from 33.41% to 1.09%, compared with a single epithermal neutron quantification method. These results prove that the E/T value uranium quantification method is unaffected by the change in the neutron source yield, effectively improves the accuracy and service life of the logging instrument, and has great scientific and popularization value.