Adjustment Of Mine Ventilation System Parameters

Mine ventilation survey data includes generally redundant, usually inconsistent, often incomplete, and always noisy measurements. Yet, these data are essential for operational decisions and represent the foundation of any mine ventilation systems model used for future engineering analysis and design. For example, the basic ventilation network geometry can be measured with a reasonable precision, the network flows can be measured, but with significant errors, and the branch resistance factors can not be directly measured at all. The network-based, constrained weighted least squares approach, presented in this paper, advances a method for the sys tem-wide adjustment of ventilation branch measurements and the consistent estimation of global system parameters (e.g. branch flows, head losses, and resistance factors). The constraints guarantee that the adjusted values satisfy the governing physical laws: conservation of flow, conservation of energy, and the appropriate energy dissipation rule (e. g. Atkinson's RQ2). The weights allow for the proper inclusion of data with widely disparate precision, as well as offering a simple method for retiring old data as the mine changes through time. The post-adjustment sensitivity analysis presents two potentially useful forms of analysis: the identification of previously unspecified leakage branches, and the semi- automatic recommendation for future ventilation measurement campaigns. The overall efficiency of the algorithm is exemplified by the fact that a ventilation network with hundreds of branches and thousands of measurements can be routinely adjusted on a microcomputer.