Accession Number : ADA319142

Title :   Identifying and Separating from a Sequence of Mine Tremors Series of Events Generated by a Single Process.

Descriptive Note : Final rept.,

Corporate Author : ACADEMY OF MINING AND METALLURGY KRAKOW (POLAND)

Personal Author(s) : Lasocki, Stanislaw

PDF Url : ADA319142

Report Date : SEP 1996

Pagination or Media Count : 254

Abstract : Generation of induced seismicity in underground mines is controlled by various local and mine wide factors, and the seismic series are usually composed of events of different origins. Although it is well known that seismicity in mines is a multimodal phenomenon, it is not easy to identify its various components. A set of statistical methods to study complexity of seismic series is proposed. These are the descriptive statistics analysis to recognize basic features of the seismic series, the extreme value method to identify modes of the energy distribution of events, to estimate the limiting energy values for these modes and to distinguish those event generating processes which have well separated modes of energy and the analysis of deflection to identify generating processes that are responsible for certain dominant directions in the seismic series. The dominant directions of the distribution of epicenters are described by the deflections of straight lines connecting the epicenters of every two consecutive events in a series, measured from the NS direction. The deflection then forms another set of time series. The series of deflections are analyzed by a nonparametric kernel estimation of the deflection probability density function; a novel, objective quantitative approach. The proposed methods were applied to the sets of events from Wujek coal mine, that occurred from 1988 to 1993 in different areas of the mine. Altogether 14297 events was analyzed. Their distribution is not random both in space and in energy. The analyses resulted in identification of likely processes of seismic event generation. The analysis of deflection applied in the pseudo-multivariate way can be used to separate, from the seismic series, events originated by these identified processes.

Descriptors :   *SEISMOLOGY, *MINES(EXCAVATIONS), MULTIVARIATE ANALYSIS, TIME SERIES ANALYSIS, PROBABILITY DENSITY FUNCTIONS, ESTIMATES, KERNEL FUNCTIONS, NONPARAMETRIC STATISTICS, GEOPHYSICS, TREMORS.

Subject Categories : Mining Engineering
      Seismology

Distribution Statement : APPROVED FOR PUBLIC RELEASE