Two and three dimensional analysis of a slope failure in a lignite mine

Abstract

With about 9.3 billion tons of reserve, lignite is a major source for energy production in Turkey. The Afşin–Elbistan lignite basin, containing 47% of the overall lignite reserve in Turkey, lies in the Afşin and Elbistan districts as a part of Kahramanmaraş. The new Çöllolar opencast mine is in the Afsin–Elbistan lignite basin, and this mine will be the second mining activity in the basin after the active Kışlaköy opencast mine. The new mine will meet the coal consumption of the Elbistan B power plant. Input parameters for slope stability analysis are essential, and so they must be determined accurately and precisely. Field investigations, laboratory testing and back analyses are vital instruments for the input parameters. This study presents the results of slope stability analysis via finite difference code and a limit equilibrium software for the soil slopes of the Elbistan–Çöllolar lignite mine. The basic input parameters, cohesion and friction angle, were determined in the soil mechanics laboratory. By back analyses of a large scale slope failure, mobilized friction angles for a critical weak clay layer under the lignite seam were determined accurately by using the 2D limit equilibrium method and 3D finite difference models. Results of the friction angles were compared in order to check the effectiveness of commonly used 2D approaches in handling the slope problems. Differences in the results of the mobilized friction angles for the weak clay layer were more than 30%. The 3D models indicated that the mobilized friction angle during the major slope failure was substantially lower than the friction angle generated by the 2D limit equilibrium method.

Introduction

Turkey contains about 9.3 billion tons of lignite reserve, and, with about 24 million tons of yearly production, Turkish Coal Enterprises (TKI) produces about 60% of the yearly lignite production capacity. Providing coal for thermoelectric power plants, TKI is responsible for about 21% of Turkey's electric power production. The Afşin–Elbistan lignite basin, with 47% of the overall lignite reserve of Turkey, is in the Afşin and Elbistan districts, as a part of Kahramanmaraş (Yörükoğlu, 1991).

The lignite basin is divided into six sectors: the Kışlaköy, Koridor, Çöllolar, Hurman River, C and D sectors (Fig. 1). A contract was signed in April 2007 by the Ciner Group firm, Park Enerji and EÜAŞ (Electricity Production Co., Inc.) for 25 years of lignite production in the Çöllolar sector in order to meet the coal consumption of the Afşin–Elbistan B thermal power plant.

For the open pit lignite mines, production scheduling is critical in order to establish proper lignite feed to the power plants. Also, an overall slope failure may lead to human and equipment losses as well as a lag in production.

Conventional slope stability analyses have been used to investigate the equilibrium of a mass of soil bounded below by an assumed potential slip surface and above by the surface of the slope. Forces and moments tending to cause instability of the mass have been compared with those tending to resist instability (U.S. Army Corps of Engineers, 2003).

When a slope fails it can provide a useful source of information on the conditions for the slope at the time of failure as well as an opportunity to validate stability analysis methods. Because of slope failure, the safety factor is considered to be unity (1.0) at the time of failure (Duncan and Wright, 2005).

Hammah et al. (2004) conducted finite element analysis for safety factor calculation of slopes by the shear strength reduction method. Limit-equilibrium and finite-element-analysis results were compared and were considered to be consistent.

Cala et al. (2004) used a modified shear strength reduction method for analyzing complex problems with different geological units. This method with the use of finite difference models enables a user to determine different safety factors for different benches in models with complex geological units. Cala et al. (2006) studied slope stability analysis by using finite difference method computer programs, FLAC and FLAC3D. The shear strength reduction method was utilized, and 3D analysis resulted in higher safety factor values than 2D solutions. They concluded that application of 2D models, especially for certain cases involving different geological units and soft subsoil layers may lead to a very conservative approach.

Rock slope deformation was analyzed by using the FLAC3D program at the Antaibao open pit coal mine in China. Stability was evaluated, and excavation design was optimized by He et al. (2007).

The Mineral Research and Exploration Institute of Turkey (MTA) conducted an up-to-date study (Akbulut et al., 2007) for slope stability of the EÜAŞ Kişlaköy opencast mine for new permanent slopes. The MTA conducted geotechnical investigations by core drilling, sampling and laboratory analysis. A 2D limit-equilibrium back analysis was conducted, and weak slip layer Mohr–Coulomb parameters were determined. This study is important, as this mine is near the Çöllolar mining area in the same lignite basin.

Cheng et al. (2007) found higher factor of safety values with FLAC3D when they increased the problem domain size for a slope with a soft soil band like clay. Boundary effect was more pronounced when cohesive strength of the soil was kept small and friction angles were changed between 5° and 25° for the slope.

Wei et al. (2009) proposed that failure mechanism in slopes can be greatly affected by the domain size. Engineers may either adopt a large domain size at the expense of computer time, or persue a trial and error analysis to determine the proper domain size.

When the literature is investigated, it is seen that numerical modeling is not as widely used for back analysis of failed slopes. In this paper the 3D finite difference analysis is utilized for determination of the shear strength parameters of a particular material having a great impact on the stability of the overall slope and to overcome the difficulty when laboratory material properties show a wide range in values and might not properly represent conditions in the field.