Geoèkologiâ

Assessing reliability of engineering geological models still requires further studies.. Much more attention is paid to the methodology for assessing the reliability and quality of models in oil and gas geology. A comparison of the methods used in this field with the methodology for assessing the quality of machine learning models showed the similarity of principles and approaches. Therefore, the algorithms for engineering geological modeling can be justified and calibrated using tools for evaluating the quality of machine learning models. The article systematizes and analyzes the metrics used in solving problems and classification, and describes the methodology of using cross-validation techniques to assess the quality of algorithms. The practical experience of constructing a computer stratigraphic-genetic model using various algorithmic approaches is described: on the basis of triangulation constructions, using ordinary kriging with automated variogram composition and using the custom heuristic algorithm that considers the history of sedimentation and technogenic transformation of the territory. It is shown that the problem of three-dimensional geological modeling can be considered both as a classification and regression problem. The error index of the stratigraphic-genetic model is proposed based on the calculation of the average absolute errors in determining the spatial position of geological layers. The proposed approaches are applicable to testing methodologies of engineering geological modeling in a broad sense, the verification of predictive models of any kind being the most difficult issue. It is emphasized that the development and filling of databases of various engineering and geological data should be intensified, including field and laboratory data, the results of their processing, forecast estimates and conclusions based on them, monitoring measurements of various kinds, remote sensing data, etc. The possibilities of processing and analyzing big data in engineering geology will allow us to move from subjective expert estimations to the application of modern approaches to modeling complexly formalized objects and phenomena using the capabilities of machine learning and artificial intelligence.

The limitations of the existing deterministic approaches to the quantitative assessment of slope stability are analyzed. The basic principles of probabilistic analysis in mathematical simulation of slope stability are considered. The possibility of obtaining probabilistic assessment of the development of landslide deformations is shown, which can later be used in the analysis of geological risk. An assessment of the objectivity of probabilistic modeling of slope stability is carried out.

The construction of the nuclear power plant (NPP) requires conducting a large number of engineering and hydrogeological surveys, as well as assessment of the design decisions’ safety. A deep excavation pit at the Paks II NPP construction site requires execution of the cut-off wall due to extremely high groundwater saturation of the alluvial deposits. However, lithological anisotropy and the presence of dislocation zones did not allow identifying the appropriate depth for the cut-off wall construction. Unfortunately, engineering geological boreholes with a 20-meter distance between them and surface and borehole geophysical surveys could not identify the hydrogeological units. Thus, to conceptualize the hydrogeological settings, an analysis of the groundwater head distribution and the large-scale pumping tests were conducted. The interpretation of the geological data and the distinguishing of the hydrogeological units were carried out iteratively using the hydrogeological numerical model. The flow model could represent the hydraulic head distribution, the response of the lithologically heterogeneous layers to the water fluctuations in the Danube river, and the pumping tests carried out at the different depths. The results of the hydrogeological modeling revealed the aquitard to be continuous throughout the territory; however, its top’s depth changes from 30–35 to 90 m within the construction site of the Paks II NPP. This complex geometry of the aquitard is controlled by the plicated dislocation zone, which cuts the construction site in half and is revealed as the right wall of the graben.Correct hydrogeological stratification enabled us to ensure waterproof activities such as the cut-off wall construction using the hydrogeological model when excavating a deep pit for the Paks II NPP. This also minimizes the hydrodynamic impact on the closely located NPP Paks in operation.

Safe development of primary diamond deposits in Western Yakutia requires constant monitoring of the hydrogeological regime of the exposed aquifer complexes within the quarry and mine fields of the deposits, as well as in the adjacent areas of drainage water injection. Over the entire period of development of the Alakit-Markhinsky, Daldynsky, Mirninsky and Nakynsky kimberlitic fields, about 400 million m³ of highly mineralized drainage water from quarries and mines were involved in the pumping-injection process. Complex cryohydrogeological conditions of the territory, i.e., lithological-facial specifics, continuous distribution of permafrost, structural confinement of kimberlite fields, fault-block structure of individual pipes, influence the dynamics of occurring changes and make the cryohydrogeological conditions of each individual pipe unique and having no complete analogues. In order to predict successfully and implement subsequently technical solutions aimed at controlling inflows of all types formed within mine and quarry fields, the Yakutniproalmaz Institute elaborated a program for the development, constant maintenance and updating of hydrogeological “digital twins” of all key mining deposits. The developed models take into account the influence of both natural factors and the applied schemes for opening and draining deposits, which impose their own limitations. The filtration problem was solved out using the licensed program FEFLOW, which performs modeling of hydrogeological conditions by the finite element method in a multilayer strata for areas of arbitrary configuration with boundary conditions of type I, II, III changing according to the known law in the presence of filtration heterogeneities in plan and section, as well as vertical filtration. The creation and constant updating of permanent digital models made it possible not only to acquire a modern tool for forecasting water inflows, but also helped to improve the planning process in terms of drilling drainage and injection wells, purchasing pumping equipment, etc. Deviation of predicted values from those actually observed within the short-term forecast for the period of use 2021–2023 ranged from 5 to 10%.

Gorevskoe zinc-lead ore deposit is located within a unique terrain: it follows the left bank of the Angara River and lies underneath its current river bed, starting 38 km far from the Angara River mouth. This plays a significant role in the ore field hydrogeology and has a great impact on its further development. The complexity of ore field hydrogeology is defined by a number of natural factors, i.e., the occurrence of both aquifers as well as low-permeable and water-proof layers; tectonic failures (if any) and their functioning in terms of hydrogeology dynamics; variability in terms of water-bearing rocks hydraulic properties; water bodies on the surface and their connection with groundwater flows. All the above mentioned features allow us to classify Gorevskoe deposit hydrogeology as rather complex and complicated. The main task of the work performed was to evaluate and forecast the water inflow to Gorevskii open-pit mine and to determine the groundwater level at the end of mining operations. Because of the deposit rather complex hydrogeology, it was decided to perform the task on the basis of numerical geohydraulic modelling using finite difference method (FDM) in the environment of Visual MODFLOW software tools. The paper describes the following steps of works performed within the project: input data acquisition, processing and analysis; creation of concept-based hydrogeology model; development of hindcasting numerical model; model calibration in the context of environmental conditions and actual mining operations hosted within the open-pit; sensitivity analysis and advanced simulation to develop the forecasting numerical model. When performed, the forecast value in terms of groundwater inflow to the pit has been calculated. Thus, pit groundwater level at the end of mining operations has been developed. Following a number of scenarios applied, model sensitivity analysis has been made and this provided the basis to identify the factors that have the most essential impact on water inflow to the pit.