The monitoring and prediction of mining subsidence in the Amaga, Angelopolis, Venecia and Bolombolo Regions, Antioquia, Colombia

Amaga, Angelopolis, Venecia and Bolombolo are small towns located in Antioquia, in the Central Cordillera of the Colombian Andes. Mining has been practised in this region for a period of at least 100 years. This mining has mainly been small-scale, poorly mechanised and restricted to shallow room and pillar workings. Recently, the semi-mechanisation of some mines has enabled coal to be extracted using longwall mining methods. However, this has resulted in subsidence that has caused severe damage to structures, residential property, and agricultural land, and also induced landslides. In the British Isles, there are several reliable methods that can be used to predict the likelihood and magnitude of mining subsidence. The British Coal Corporation and the University of Nottingham have developed one such method, the "Subsidence With Influence Function Technique (SWIFT)." Based on mining subsidence observations undertaken in the coalfields of Britain over a period of approximately 50 years. The SWIFT program was used to predict the magnitude of subsidence, above a longwall panel, at the Industrial Hullera mine in Colombia. The results were then compared with subsidence profiles obtained from precise levelling and field monitoring. In each case, the SWIFT program overestimated the magnitude of mining subsidence by 0.17-0.20 m. However, the morphology of the subsidence profile, area-of-influence and location of maximum subsidence were similar. This overestimation of the predicted subsidence was attributed to the occurrence of strong, igneous rocks, such as rhyolite sills, in the Colombian coal measures. These strong, competent horizons act as cantilever beams during subsidence, causing bed separation and therefore reducing the magnitude of subsidence, in spite of these differences, mining subsidence can be predicted with a reasonable degree of accuracy and precision using the SWIFT technique, provided the software is calibrated and used in conjunction with local expertise. (C) 2001 Elsevier Science B.V, All rights reserved.