The study of clustering processes is fundamentally significant in hydraulic engineering to comprehend the interactions between turbulence and particles. Previous studies on bubble clustering were carried out in plunging jets, hydraulic jumps and dropshafts. The present paper focuses on the bubble clustering process in a large-size dropshaft, and three criteria for cluster identification were applied. They were based upon the analysis of water chord between two adjacent air particles. When two bubbles are closer than a characteristic length scale, they can be considered as a cluster. The characteristic water length scale may be related to the water chord statistics, such as the mean or the median water chord, or to the air chord length of th...