Investigating the consolidation behaviour of brown coal

The extensive brown coal deposits in the Latrobe Valley are being extracted at three large open cuts (mines) for electric power generation. The extraction of the coal to depths of more than two hundred metres requires the groundwater pressures in the underlying aquifers to be significantly reduced to prevent heave of the mine floors. Depressurisation has caused widespread subsidence. In the 1980’s a significant effort was made to assess the scale of the subsidence and to model the progress of the subsidence using standard models of consolidation based on Terzaghi’s one dimensional consolidation equation. A preliminary review of the historical data on consolidation behaviour of brown coal showed that the standard assumptions underpinning Terzaghi’s consolidation equation are probably not valid for this material and that other processes are likely to be influential in controlling the rate of consolidation.
   The current research has re-investigated the behaviour of brown coal under one dimensional loading conditions through a series of laboratory experiments that explore the time dependent characteristics of consolidation on samples taken from the working faces of the brown coal open cuts. Standard and non-standard oedometer tests were carried out under a range of loading conditions, including cyclic loading, applicable to the effective stresses that operate in the field. The non-standard tests were carried out to investigate the early time behaviour of consolidation and to examine the long term secondary consolidation rates. Early time behaviour includes an immediate settlement component that can be explained by the presence of occluded gas in the coal and by compressibility of the coal solids. While gas samples could not be extracted from the coal for analysis, short term consolidation experiments under varying ambient pore water pressure conditions provided evidence that gas is partially responsible for the instantaneous settlement and the degree of saturation proves the existence of the gas. The cyclical tests show that the gas contribution to immediate settlement does not dissipate with repeated cycling. This observation and the observation of the occurrence of gas in the in–situ coal at low pore pressures provides evidence to suggest that the immediate settlement contribution would also be applicable under field conditions. A mathematical model of consolidation incorporating the contributions of occluded gas and coal solid compressibility has been developed and incorporated in a numerical model. A non-linear optimisation technique was employed for the model fitting to obtain values for the coal compressibility and the gas content for each sample. The results showed strong consistency in the gas saturation values and coal compressibility.
   The new numerical model was used to predict ground subsidence observed at sites within the mining influenced zone and compared with the monitored data. Good agreement between the observed and predicted results was obtained. The fit is marginally better than the historical modelling using Terzaghi’s consolidation model but produces significantly different forward predictions of the rate of consolidation for the thicker coal seams. The model also confirms that there will be significantly different path of recovery of subsidence during rebound of the groundwater pressures.