Native sulphur deposits mined using the underground melting method are characterized by a complex structure, which is the result of the many geologic processes which led to their formation.
The resource utilization rate and the consumption of hot water per ton of sulphur are the main criteria of mining effectiveness. They depend on the porosity and permeability of the rocks forming the deposit, the content and mode of occurrence of sulphur (ore texture), and the distribution of rocks with these varying features. Good recognition of geological and hydrogeological deposit features, exploitation results, is important for formulating the rules of controlling the course of exploitation in order to achieve the best recovery of sulphur with the lowest possible water consumption and to reduce operating costs.
Sulphur deposits are characterized by great local and directional variations in their structure and hydrogeological parameters. This makes the melting process irregular. The flow of hot water and melted sulphur is facilitated in certain directions. As a result, the shape, and distribution and form of exploited parts of the deposit are highly variable. Full information about the deposit is necessary for the proper understanding and prediction of processes that occur in the deposit during sulphur melting, for forecasting its effects, and for controlling the exploitation process. This information is obtained through the lithological description of core samples from exploratory and exploitation boreholes, geophysical borehole logging, and surface seismic surveys.

This publication presents the research aimed at developing statistical models, on the basis of which it was possible to prepare credible forecasts of unit cost and coal net output for longwalls in 5 hard coal mines in P oland. The argument has been verified that there is a dependence between the level of nuisance and the level of costs, as well as longwall production results.

A research procedure has been developed for that purpose, which aimed at developing two statistical models connecting the nuisance due to geological and mining conditions with costs and longwall production results. The multiple linear regression technique has been used to develop statistical models. The set of data taken into account in the analyses comprised 120 longwalls mined in the years 2010–2019. Two models have been developed – one for forecasting unit costs, the other for forecasting coal net output. Subsequently, the models’ forecasting ability has been verified on a sample of historical data. A relative forecast error for 75% of observations has been in the range of (–25%; +37%). That result has been considered satisfactory. Subsequently, using those models, forecasts of unit costs and coal net output have been prepared for 220 longwalls planned for mining in the years 2020–2030. Those forecasts have been prepared in the stipulated ranges of geological and mining nuisance influencing mining process, by means of dedicated W Ue and W Ut factors. The nuisance models for forecasting purposes have been developed using the AHP (Analytic Hierarchy Process) method. The research hypothesis has been confirmed on the basis of the obtained results. An increase in the level of nuisance leads to an increase in the unit costs for longwalls and the deterioration of production results. Unit operating costs for longwalls in specific ranges of nuisance may differ by up to 30%, being in the range of 52.0–120.3 zł/Mg. Likewise, the coal daily output of longwalls may be even 22% lower, having the average level in the range of 1.89–3.61 thousand Mg/d.