Sensitivity Analysis of Ventilation Models – Where not to Trust Your Simulation

"Ventilation models are a vital tool to ensure ventilation designs can provide airflow to all work areas in a mine. Airway resistance and variations in natural ventilation pressures will affect simulated airflow and heat distribution through a mine. Validating ventilation model resistances and pressures to the exact underground conditions can be difficult and the variances can contribute to significantly different simulation results depending on the structure of the ventilation design. The study will use sensitivity analysis to explore examples of variability of simulated airflow due to small resistance, heat and natural ventilation changes and thereby help the user identify both trustworthy and potentially unreliable parts of a mine ventilation model. INTRODUCTION Mine ventilation network modeling is a vital tool for ventilation professionals and has become increasingly relied upon during the design and operation phase of mine ventilation systems. However, they provide only an approximation of reality, and caution must always be exercised when drawing conclusions or making decisions based on simulated data. Ventilation modelling should be performed by persons with a sound fundamental knowledge of ventilation and ideally should be validated with field data, but in large mines, or mines which are yet to be mined, accurate input of the simulation variables can be difficult, and errors and inaccuracies may be difficult or impossible to avoid. This study discusses and investigates the sensitivity of the mine ventilation simulation results to input errors. In addition, there are limitations to what steady-state airflow and heat simulations can model, regardless of the accuracy of the model inputs. In some situations, steady-state solvers cannot capture the dynamic changing behavior of a ventilation system. This study will also describe and apply techniques to analyze the natural variability of mine conditions on ventilation models. The consequences of poor accuracy in some parts of a mine model may impact on the health and safety criteria of ventilation designs. For example, if an airway is required to flow in a particular direction and quantity, then despite a ventilation model showing correctly designed conditions, small errors in input values may result in incorrect prediction of the direction or quantity of airflow. This may create unexpected buildup of heat or gases during operational activities, causing safety issues or operational delays. The ability to know which parts of a mine ventilation model have increased potential for poor simulation results will allow engineers to consider the ventilation design and make changes if required to better guarantee the required design flows."