The article describes the influence of anomalous values and local variability on the structure of variability and the estimation of deposit parameters. The research was carried out using statistical and geostatistical methods based on the Pb accumulation index in the shale series in part of the Cu-Ag ore deposit, LGCD (Lubin-Głogów Copper District). The authors recommend the use of a geostatistical tool, the so-called semivariogram cloud to determine the anomalous values. Anomalous values determined by the geostatistical method and removed from the dataset have resulted in a significant reduction of the relative variability of data, which is still very large in the case of the analyzed parameter or parameters with similar statistical features such as extreme variability and strongly asymmetric distribution. Calculations of the resources of this element can be treated only as estimates and formally classified to category D. The hypothetical assumption of the absence of sampling errors, resulting in a decrease in the magnitude of local variation, leads to a certain reduction of the median error of resource estimates. However, they are still high (> 35%). This is due to the large natural variability of the accumulation index of Pb on the local observation scale. The current method for collecting samples from mine workings of the Cu-Ag deposits in the Lubin-Głogów Copper District (LGCD), aimed at the proper assessment of copper resources, the Cu content, and at estimating the quality of copper output, makes it impossible to achieve an accuracy of estimates of Pb resources similar to that obtained for the main metal. Theoretically, this effect can be achieved by a strong concentration of the sample collection points and thanks to a multiple increase in the samples weight; this, however, is unrealistic for both economic and organizational reasons. It is therefore to be expected that the assessment of Pb resources and other accompanying elements of similar statistical features (e.g. As), located in parts of the deposit where mining activities are to be carried out, will be subject to significant errors.
The occurrence of gas confined in shales allows us to consider it as a component of the host rock. During drilling wells, the gas is released into the drilling fluid from finely ground gas-bearing rock particles. The amount of gas released can be determined on the basis of mud-gas logging; in addition, it is possible to determine the gas-content in shales expressed by the volume of gas released per mass unit of rock [m3/ton]. The gas content in the Ordovician and Silurian shales (Sasin formation and Jantar member respectively) in two selected wells in northern Poland was determined using this method. It has been found that clearly distinguishable, highly gas-bearing sections, which are separated by very poorly gas-bearing ones, can be determined in the well log. The increased gas content in shales can be observed in zones generally enriched in TOC. No direct correlation between TOC and gas-bearing capacity was found however, but the structure of TOC variability and the gas-bearing capacity described using variograms is identical. Correlations of the distinguished gas-bearing layers in the wells under consideration suggest a multi-lens or multi-layered reservoir model. The lack of natural boundaries in the shale gas reservoir means that they must be determined arbitrarily based on the assumed marginal gas-bearing capacity. In the case of several gas-bearing zones, numerous variants of interpretation are possible. In any case the low, best and high estimated resources may be evaluated, assigned to each borehole in the area with radii equal to the range of variogram of gas content in horizontal part of the well.