Abstract: In the CO₂+O₂ in-situ leaching (ISL) uranium mining process, the long-distance pipeline transportation of gas-liquid two-phase mixtures from the surface to the injection wellhead frequently suffers from gas-phase retention and gas blockage, which severely compromises oxygen delivery efficiency and operational safety. Such mixtures are typically generated by surface dissolved oxygen devices, among which the 16-6 jet-type unit is the most widely adopted in uranium mining operations across China. To systematically investigate the stability characteristics of gas-liquid two-phase flow from such devices during long-distance pipeline transportation and to develop quantitative diagnostic criteria for gas blockage risk assessment, a field experimental system was established at a uranium mine in Inner Mongolia, China, enabling in-situ measurements under actual pressurized operating conditions. High-frequency pressure fluctuation signals and high-speed camera images were simultaneously acquired at two observation stations separated by 300 m, namely, immediately downstream of the dissolved oxygen device and at the injection wellhead, respectively. The pressure signals were analyzed using a multi-scale approach integrating fast Fourier transform (FFT), continuous wavelet transform (CWT), and discrete wavelet transform (DWT), while high-speed imaging provided direct visualization of bubble dynamics and flow pattern evolution. The results reveal that the outlet of the 16-6 jet-type dissolved oxygen device exhibits a highly turbulent flow field characterized by dense bubble swarms with irregular shapes and vigorous motion. The bubble size distribution is notably broad, with a considerable proportion of large bubbles, indicating intense bubble breakage and coalescence driven by strong shear forces. Moreover, the power spectral density exhibits a dominant peak in the intermediate-frequency range, and the high-frequency turbulent kinetic energy accounts for a substantial share of the total energy, which serves as the dynamic source for effective oxygen dissolution but also introduces significant flow uncertainty. In contrast, after 300 m of pipeline transportation, the flow characteristics undergo substantial evolution: the high-frequency energy attenuates dramatically, and the energy distribution shifts predominantly toward the low-frequency region, indicating that the flow regime transitions from turbulence-shear dominated to pipeline-transport dominated. Concurrently, the pressure fluctuation standard deviation decreases markedly, and the autocorrelation half-life extends considerably, suggesting improved flow regularity. Nevertheless, the elevated low-frequency oscillations and the persistent moderate-frequency energy indicate that secondary bubble coalescence remains a potential risk at the wellhead under elevated downhole pressure conditions. On the basis of these findings, two quantitative diagnostic indicators for gas blockage risk were established: the high-frequency energy ratio (HFE), which reflects bubble coalescence activity, and the autocorrelation coefficient (AC), which characterizes flow regularity. Accordingly, HFE values above a specified threshold coupled with low AC values correspond to high-risk conditions typical of the dissolved oxygen device outlet, whereas low HFE combined with high AC indicates stable flow at the wellhead. Consequently, a practical early-warning strategy integrating real-time pressure monitoring, threshold-based risk assessment, and proactive operational adjustments is proposed to prevent gas blockage incidents. This research provides a theoretical foundation and technical support for intelligent monitoring and safe operation of gas injection systems in ISL uranium mining.