Development of a Statistical Basis for the Estimation of Local Airborne Contamination Concentration in Underground Mines

Abstract

Airborne contamination is an inescapable feature of modern underground mining environments. Contamination can enter an underground mine's ventilation system from equipment, blasting or other rockbreaking processes, and the desorption of gases from rock. The monitoring of airborne contamination is split into two types: personal and area monitoring. Personal monitoring can be incredibly expensive and time-consuming for industrial hygiene professionals, and critical health data can take up to three weeks to be available for decision making. Area monitoring is generally lower in cost, but area data is only considered valid within the immediate vicinity of the monitor. This dissertation lays the groundwork for estimating airborne contamination concentrations across a portion of an underground mining environment using statistical methods. Although area monitoring data are not directly admissible for use as personal monitoring data, there necessarily exists some relationship between observed concentration in one part of an underground mine's ventilation system to any other part of the underground mine's ventilation system. Local airborne contamination concentrations can be estimated based on knowledge of other local concentrations. The work required to make this estimation possible can be split into a few main topics: data interpolation and estimation methods, the statistical relationships between different locations in a ventilation system, and the optimal placement of air quality monitors underground. Ultimately, this culminates in a statistical means for estimating airborne contamination distributions that is transferable to any major airborne contaminant that can be readily monitored underground. By using a statistical means, the resulting method remains transferable from contaminant-to-contaminant and mine-to-mine without the need to retrain a machine learning model.