Development of geological structural criteria for predicting unstable mine roof rocks
This study was designed to investigate the cause of roof falls in room-and-pillar coal mines and to determine geologic methods for predicting unstable roof conditions. The study area was located in Harlan County, Kentucky and the Bailey Creek and Highsplint mines were selected for study. The Bailey Creek mine operates in the Harlan coal bed and the Highsplint mine operates in the Darby coal bed, about 200 feet higher stratigraphically. Both coal beds are in the Mingo Formation of the Breathitt Group of Pennsylvanian age, deposited in a fluvio-deltaic environment.
The mines are drift mines and correlations were made laterally between surface outcrops and within the mines using lithologic and trend analysis criteria as follows:
1. Lithologic criteria involved comparing changes in roof strata. Outcrop studies showed that the strata are cyclic or repetitious vertically. Five vertical sequences were recognized and correlations made with roof falls. Results indicated that roof falls are associated with zones of weakness within or between vertical cycles. These zones include rider coal seams, carbonized plant debris on bedding planes, thin-bedded brittle rocks, cross-bedding and zones of poor sandstone cementation. Abrupt lateral facies changes occur. In-mine studies were also conducted. In the Highsplint mine, 9l measured rib and coal sections provided the basis for a roof classification in four categories or conditions: Conditions A and B (Laminated Shale), Condition C (Rolling Sandstone with Shale in Pockets) and Condition D (Zone of Major Sandstone Roll(s)). A structure contour map of the Darby coal bed, and isopach maps of the Darby coal bed and immediate roof rocks were used to project axial trends of the sandstone roof.
2. Trend analysis studies were conducted to investigate cross-bedding, channel trough axes and sole marks. Computer programs were used to analyze the data, and a comparison between surface trough axes and sandstone roll axes means within the Highsplint mine indicated a discrepancy of about 35°. Computer programs were used to analyze jointing, coal cleat, and "slips" (slickensides). The program groups into 4- or 6-sets. The mean directions of the four joint sets for the area are: Set I - N65°E to N75°E, Set II - N27°W to N15°W, Set III - N25°E to N32°E and Set IV - N65°W to N58°W. Surface joints were compared with surface drainage segments and with joints, "slips," and sandstone rolls inside the mines. A direct relationship between joint and "slip" frequency and direction with sandstone roll trends was found. Correlations between surface joints and joint-controlled roof falls underground were not found in areas of the mines under overburden in excess of 600 feet.
Studies were made of samples obtained from ten NX cores taken from the roof in the Highsplint mine, on both sides of a roof-fall zone, to investigate trends. Petrographic methods were used to evaluate composition and texture for 27 thin-sections of the sandstone roof. Correlation between cores showed that the roof-fall zone represented a marginal position between sand-filled channels. Twelve shale and siltstone core samples from the immediate roof were studied using x-ray diffraction. The dominant clay minerals include illite, chlorite and kaolinite. Sharpness ratios (10A illite peak), and the chlorite 14A peak height were compared between cores.
Studies at the Bailey Creek mine indicated that roof falls involving the main roof were related to cross-bedding in a sandstone channel. Failure was caused by the heading direction being driven subparallel to the channel direction and to the strike of cross-bedded sandstones filling the channel. Monitoring of the mine from April, 1976 to September, 1976 indicated improved conditions when the heading direction was changed to avoid both the channel direction and frequent joint directions.
Thesis76b.H955.pdf
31.39 MB
Adobe PDF
666ebab3f040a92aa1354d4dac8be42a