In the Adriyala project, shield-leg pressure data from all shields and shearer location with time stamp from a mechanized longwall face are monitored and recorded in a surface computer. The data contain random noise due to power failure, leakage of legs, bad sensor readings, and other maintenance or geomechanical problems. An algorithm has been developed to detect mining cycle parameters such as setting and final pressures and duration of the cycle from these data for the analysis of periodic roof weighting phenomena. The intensities and persistence of periodic roof weighting are vital information for mine management to make judicious decisions on longwall operation. This paper presents the details of the algorithm, which has identified 96 percent of mining cycles from the noisy data belonging to five consecutive months or 410 m of face advancement. Based on the analysis, two types of periodic roof weighting phenomena are detected: (1) normal, with peak intensity of 75 to 90 percent of yield pressure, interval of occurrences of about 10 to 15 m and persistence of 3 to 4 m, and (2) severe, with peak intensity of more than 90 percent of yield pressure, interval of occurrences of 25 to 40 m and persistence of more than 10 m. It is envisaged that these phenomena have direct relationships with main roof thickness, its overhung length and material properties.
Full-text paper:
Mining, Metallurgy & Exploration (2021) 38:945–958, https://doi.org/10.1007/s42461-020-00361-z
