Minimisation of Surface Mining Costs using Artificial Neural Network P. Y. Dhekne 1 , Nawal Kishore 2 , and Romil Mishra 1 1 Faculty, Department of Mining Engineering, National Institute of Technology, Raipur, India and 2 Faculty, Department of Mining Engineering, IIT-BHU Varansi, India Abstract: Costs of surface mining unit operations are controlled by rock fragmentation distribution. The costs can be reduced if the muck pile does not contain oversize fragments which require crushing and grinding. The oversize fragments can be reduced by adjusting the surface mine blast design so that their number in the muck pile is minimum. This paper explains the application of the Artificial Neural Network (ANN) for the minimisation of oversize fragments so that overall cost is minimum. It was observed that the trained neural network model estimated the boulder count with sufficient accuracy and it provides a feasible choice to the field engineers to optimize the blast design so that the boulder-count is the minimum and subsequently the improving the efficiency of downstream operations and their costs Keywords: Artificial neural network, Blasting, Rock fragmentation, Boulder count. 1. Introduction Production of mineral from a mine involves a number of unit operations downstream to drilling and blasting. Drilling and blasting claim around 20% share of the total operating costs, the efficiency of other downstream operations and ultimately their costs – which account for almost 80% of the total operating cost, depend largely on the fragmentation distribution resulting from blasting. This requires the breakage in such a way that the oversize fragments are minimum. Thus, minimisation of oversize fragments (boulder) is always one of the objectives in any production blasting. The objective can be achieved by improving the efficiency of drilling and blasting. Rock fragmentation is a complex phenomenon and it depends upon many factors. These factors can be grouped in four different categories: rock geotechnical parameters such as density, hardness, compressibility; explosive parameters such as density, velocity of detonation; technical parameters such as delay interval, primer strength and location and geometrical parameters such as burden, spacing, and stemming [1]. Concept of Artificial Neural Network (ANN) has been applied to model the fragmentation [2]-[12] etc. ANN is suitable in such a case because a large number of affecting variables and their complicated mutual dependence is not reflected in the output of empirical modeling. Fragment size or as the sieve analysis of the muckpile can be obtained from the developed ANN based fragmentation models. Mining engineers are interested in knowing the boulder count so that they can plan the secondary breakage operations to reduce the downstream operation costs. Therefore an ANN model has been developed to predict the boulder count. The data sets required for the development of the model have been generated from the Limestone quarries. 2. Description Of The Sites The ANN model described in this paper has been developed from the blast records generated from Baikunth, Hirmi, Sonadih and Rawan Limestone quarries. The quarries are situated in Raipur district of Chhattisgarh province of India and are located within a radius of 20 km. The geotechnical properties of the deposits are summarized in Table 1. Int'l Journal of Research in Chemical, Metallurgical and Civil Engg. (IJRCMCE) Vol. 4, Issue 1 (2017) ISSN 2349-1442 EISSN 2349-1450 https://doi.org/10.15242/IJRCMCE.IAE03170002 156