Computational Fluid Dynamics Modelling of Gas Flow Dynamics in Large Longwall Goaf Areas

摘要

Computational fl uid dynamics (CFD) models have been developed based upon fi eld data collectedrnfrom a longwall mine in New South Wales, Australia to study the behaviour of goaf gas fl ow in bothrnactive and sealed goaf areas. A base CFD model was built to represent the goaf situations whenrnthe active longwall retreats near the fi nish-off line. Results from the base model were calibratedrnand compared against fi eld goaf gas monitoring data. The base model was then used to carry outrnparametric studies to investigate a number of operational scenarios and their impact on goaf gasrnbehaviour. CFD modelling results indicate that the overall goaf gas fl ow pattern changes as thernoperating longwall retreats towards fi nish-off line. In all cases, oxygen penetration into the activerngoaf remains high, reaching 15 per cent or higher, even at some 800 m behind the longwall face.rnThe modelling results also indicated that it would be diffi cult for early detection of an activerngoaf heating (on the maingate side) based upon CO readings in the return airfl ow, as the mainrnstream of the gaseous product will be fl owing into the sealed deep goafs in adjacent longwall (LW)rnpanels. Monitoring points (such as tube bundles and bag sampling points) should be selected atrndeep seated seals along the active goaf so that abnormal CO or C2H6 readings can be picked up forrnearly detection and location of potential heating spots; goaf inertisation can be better achieved byrninjecting inert gas such as nitrogen at deeper points (>200 m) behind the operating longwall; therninjection of in-seam drainage methane into the goaf areas will only have a limited impact on goafrninertisation as much of the injected methane will migrate towards deep and higher parts of the goafrndue to its buoyancy effect.