Development of a fuzzy model to predict flyrock in surface mining M. Rezaei, M. Monjezi ⇑ , A. Yazdian Varjani Faculty of Engineering, Tarbiat Modares University, Tehran, Iran article info Article history: Received 21 April 2009 Received in revised form 4 August 2010 Accepted 3 September 2010 Keywords: Flyrock Gol-E-Gohar iron mine Fuzzy model Statistical model abstract Flyrock is one of the most hazardous side effects of blasting operation in surface mining. This phenom- enon can be considered as the main cause of casualties and damages. Inaccuracy of the available flyrock prediction empirical methods has caused utilizing of new methods such as fuzzy systems. In this paper a Mamdani fuzzy model was developed to predict flyrock in the Gol-E-Gohar iron mine of Iran. In this regard, a database including 490 datasets of the mine blasting operation was prepared from which about 20% was kept for testing the models. Performance of the fuzzy model was compared with that of the con- ventional statistical method. It was observed that efficiency of the developed fuzzy model is much better than the statistical model. Also, sensitivity analysis showed that powder factor and rock density are the most and least effective parameters on the flyrock, respectively. Ó 2010 Elsevier Ltd. All rights reserved. 1. Introduction As compared to other industries, mining industry is associated with high rates of occupational injuries and fatalities. To reduce costs related to the potential hazards remedial measures should be deemed (Sari et al., 2009). Despite improvement of safety level in the blasting operation, there are still unpleasant reports to both the people and structures (Kecojevic and Radomsky, 2005; Verakis and Lobb, 2003; NIOSH, 2000). Flyrock, propelled rock fragments by explosive energy beyond the blast area, is one of the undesirable phenomena in the mining blasting operation (IME, 1997). Any mismatch between distribu- tion of the explosive energy, mechanical strength of the rock mass and charge confinement can be cause of flyrock (Bajpayee et al., 2004). Other investigations have revealed that the major factors responsible for flyrock are insufficient burden, improper blasthole pattern, unsuitable loading charge, geological anomalies, inade- quate stemming, and inappropriate delay time (Kecojevic and Radomsky, 2005). In various researches, Langefors and Kishlstrom (1963), Holme- berg and Persson (1976), Roth (1979) and Persson et al. (1994) have explained influential parameters on the flyrock. An empirical model for predicting flyrock has been developed by Lundborg (1974, 1981). Further studies on the phenomenon have been per- formed by Ladegaard-Pederson and Persson (1973), Fletcher and D’ Andrea (1986), Siskind and Kopp (1995), Shea and Clark (1998), Bajpayee et al. (2000) and Rehak et al. (2001). Performance of the available flyrock predictive models is not satisfactorily due to complicated nature of the problem. These models mostly have been developed considering only some of the relevant parameters. Therefore, flyrock prediction may require application of the other new methods such as fuzzy system. This system is one of the most competent artificial intelligence subsys- tems that can cope with the complicated and ill-defined problems in a flexible and reliable way. In the last two decades an increase of the fuzzy system applica- tions in the field of mining has been observed (Iphar and Goktan, 2006). Chuang (1995) proposed a fuzzy model bridging the dis- crepancy between the values of the in situ shear strengths of soils and laboratory test results. Mishnaevsky and Schmauder (1996) showed that fuzzy set theory could efficiently be applied to exam- ine the damage evolution in heterogeneous rocks. Also, Habibagahi and Katebi (1996) employed the method to develop a rock mass classification based on the Bieniawski classification. Jiang et al. (1997) applied the approach to characterize roof conditions in long wall mining. Cebesoy (1997) and Bascetin (1999) used fuzzy tech- nique for the selection of surface mine equipments. Wu et al. (1999) employed the method to describe the damage threshold of a rock mass under dynamic pressure of explosion. A methodol- ogy for slope stability analysis using fuzzy system was proposed by Dodagoudar and Venkatachalam (2000). Klose (2002) described a simple approach for geological interpretation of the seismic data utilizing fuzzy method. Tzamos and Sofianos (2006) used fuzzy ap- proach for extending the Q system. In this study, a fuzzy model was developed to predict flyrock in the Gol-E-Gohar iron mine of Iran. For validation of the proposed fuzzy model, regression analysis was performed for the same datasets. 2. Case study The Gol-E-Gohar iron mine is situated 55 km southwest of Sir- jan between 551150E and 551240E longitudes and 29,130 N and 0925-7535/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.ssci.2010.09.004 ⇑ Corresponding author. Tel.: +98 21 82884312. E-mail address: monjezi@modares.ac.ir (M. Monjezi). Safety Science 49 (2011) 298–305 Contents lists available at ScienceDirect Safety Science journal homepage: www.elsevier.com/locate/ssci