S. ESEN & H.A. B LG N Department of Mining Engineering, Middle East Technical University, Ankara, Turkey

Transcription

1 EFFECT OF EXPLOSIVE ON FRAGMENTATION S. ESEN & H.A. B LG N Department of Mining Engineering, Middle East Technical University, Ankara, Turkey ABSTRACT: In this study, effect of change of explosive type on fragmentation is investigated. Two blast tests in which ANFO is used in the first and BARANFO 50 is used in the second one as the main blasting agent were conducted at Lafarge Yibita Lalahan Quarry. Both blasts were carried out at the same bench of the quarry having the same structural geology and same surface blast design parameters. The only varying parameter is the blasting agent. Both tests were monitored by continuous velocity of detonation recorder. Digital images were acquired from the muckpile after each blast test by using the same image sampling technique. Images were analyzed by using SPLIT software and size distributions of muckpiles obtained from both blasts were determined. SPLIT is a digital image processing software developed to compute size distribution of rock fragments from digital images. As a result of this research, it was shown that finer fragmentation is obtained by utilizing BARANFO 50 as the main blasting agent at a constant blasthole pattern. It is also shown that it is possible to expand the blasthole pattern as an alternative if the previous coarser fragmentation is preferred. Moreover, comparison of explosives should not be performed based solely on the purchase price. Since fragmentation affects all post-blast mining operations (loading, hauling, crushing, grinding, etc.), explosive performance should also be assessed by fragmentation evaluation using softwares which are proven to be accurate similar to SPLIT.

2 2 1. INTRODUCTION It is necessary to couple all the parameters, namely explosive and rock properties and surface blast design for an efficient blasting. Minor changes in the controllable parameters which are explosive type and surface blast design can have a major effect on the resultant fragmentation. Once the blast has been carried out, it is necessary to analyze the obtained results, as its interpretation will give hints for the successive modifications of the blast parameters for the following rounds. The factors that must be considered to evaluate blast results are: - Fragmentation, - Geometry of the muckpile, - State of the remaining rock, - Assessment of the results obtained by blast monitoring systems, - Environmental problems due to blasting (ground vibration, airblast, fly rock and dust). Fragmentation is one of the most important concepts of Explosives Engineering. Blasting is the first step of the size reduction in mining and it is followed by crushing and grinding unit operations. The efficiency of these unit operations is directly related to the size distribution of muckpile. Therefore, reliable evaluation of fragmentation is a critical mining problem. In this study, two blast tests were conducted at the same rock environment and surface blast design parameters. ANFO and BARANFO 50 were utilized as main blasting agents in the first and second tests respectively. The reason to carry out these tests was to investigate the effect of explosive type on fragmentation. Split-Desktop software is used to quantify the size distribution of fragmented rock. 2. METHODS DEVELOPED TO DETERMINE THE SIZE DISTRIBUTION OF FRAGMENTED ROCK Methods to quantify the size distribution of fragmented rock after blasting are grouped as direct and indirect methods. Sieving analysis of fragments is the only technique in direct method. Although it is the most accurate technique among others, it is not practical due to being both expensive and time consuming. For this reason, indirect methods which are observational, empirical and digital methods have been developed. Observational method which depends on expert s common sense is a widely used technique. An engineer assesses the fragmentation and other blasting results subjectively. This method is not a scientific method as it does not give any information about the size distribution [1], [2], [3]. Blasting parameters are considered to determine the size distribution in some empirical models such as Larsson s equation, SveDeFo formula, KUZ-RAM model, etc. [3]. The most popular method to quantify the fragmentation is the determination of the size distribution using digital imaging processing techniques. This method being cheap and useful consumes less time and does not interrupt the production at the site. Due to these reasons, it is

3 3 preferred widely by explosives engineers. It is the second reliable method after sieve analysis. In this method, image acquired from muck pile, haul truck, leach pile, draw point, waste dump, stockpile, conveyor belt, etc. are delineated automatically by using digital image processing techniques and size distribution of fragmented rocks is determined finally [4], [5]. There are several softwares namely SPLIT, WipFrag, GoldSize, FRAGSCAN, TUCIPS, CIAS, PowerSieve, IPACS, KTH, WIEP, etc. that are commercially available to quantify the size distribution. The accuracy of these systems varies between 2 % to 20 % [6], [7]. 3. THE SPLIT SYSTEM FOR ANALYZING THE SIZE DISTRIBUTION OF FRAGMENTED ROCK 3.1. Description of the Split-Desktop System SPLIT is an image processing program for determining the size distribution of rock fragments at various stages of rock breaking in the mining and processing of mineral resources. The desktop version of SPLIT refers to the user-assisted version of the program that can be run by mine engineers or technicians at on-site locations. The desktop SPLIT system consists of the SPLIT software, computer, keyboard and monitor. There must be a mechanism (software and/or hardware) for downloading digital or video camera images onto the computer. For digital cameras the software that is supplied with the camera is required and for video camera images a frame grabber board is necessary. For higher resolution images and for ease of image selection, than is available by most frame grabbers, a digital camera is recommended. Resolution of the images should be at least 512x512. The first step is for the user to acquire images in the field and download these images onto the computer. The source of these images can be a muck pile, haul truck, leach pile, draw point, waste dump, stockpile, conveyor belt, or any other situation where clear images of rock fragments can be obtained. The SPLIT program first assists the user in properly scaling the images. SPLIT can then automatically delineate the fragments in each of the images and determine the size distribution of the rock fragments. SPLIT allows the resulting size distributions to be plotted in various forms (linear-linear, log-linear, log-log, and Rosin- Rammler). The size distribution results can also be stored in a tab-delineated file for access in separate spreadsheet and plotting programs [5]. The desktop version of the SPLIT program has five major parts. The first part of the program concerns the scaling of images taken in the field. The second part of the program deals with the automatic delineation of the fragments in each of the images that are processed. The third part of the program allows editing of the delineated fragments to ensure high quality results. The fourth part of the program involves the calculation of the size distribution based on information from the delineated fragments. Finally, the fifth part of the program concerns the plotting or export of the size distribution results. Each of the five parts of the program are described below [5], [8].

4 Image Acquisition and Scaling There are many ways that images can be acquired in the field and scaled. For instance, if images are taken along a moving conveyer belt, the scaling of the images is straightforward and can be as simple as measuring the width of the belt. When acquiring images of muck piles, the angle of the slope relative to the axis of the camera needs to be considered. If it is not perpendicular, the scale as represented in the image varies continuously from the bottom of the slope to the top of the slope. There are several ways to correct the scale in muck pile images [5], [8]. The simplest way is to place two objects of known size in the image, one near the bottom of the slope and one near the top of the slope, as shown in Figure 1. To eliminate side-to-side distortion, all pictures should be taken perpendicular to the line of the toe of the slope. Three scales of image which are large scale (6x6 m), medium scale (3x3 m) and small scale (0.5x0.5 m) are required. Equal numbers of images at each scale should be acquired. If one is not interested in the size distribution of the smallest scale of the material and is happy to accept Schuhmann or Rosin-Rammler curve in this range, taking the small-scale images may be omitted. Total number of images acquired from each blast range from 8 to 20 depending on the size of the blast. Figure 1 and Figure 2 show that images acquired at large scale and medium scale respectively. While taking the images, lighting is important. Best lighting is provided in overcast days due to even lighting and fewer shadows Fragment Delineation Once the images have been acquired and scaled, the next step is for SPLIT to delineate the individual rock fragments in each of the images. Lighting corrections and auto thresholding algorithms are used. After preprocessing and auto-thresholding, the Split-Desktop program automatically delineates the fragments using a set of algorithms based on the following 4 steps: gradient filter, shadow convexity analysis, Split algorithm, and Watershed algorithm. Details of these steps are described by Girdner et al. [8] and Kemeny [9]. The result of the automatic delineation is a binary image (2 gray levels, black and white) that contains white particles and a black background. Figure 3 is the binary image that results from the delineation of the muck pile image shown in Figure 1 (some editing has also been performed, which is shown in gray, and is described in the next section). The black areas in these images contain fine material too small to delineate in addition to the unfilled air space between particles. This black area is very important in estimating the amount of fines Editing of the Delineated Binary Image In most muck pile images and in many images from other sources such as haul trucks or leach piles, there are instances when the automatic delineation algorithms in SPLIT will not delineate the fragments properly. This may be due to situations where the lighting is poor, there is an abundance of fines in the image, the image quality is low or other reasons. In these cases, the binary file containing the delineated fragments needs to be edited using hand editing tools in the program. There are three common cases where minor editing is needed. First of all, if there are

5 5 Figure 1. A large scale muckpile image. Figure 2. A medium scale muckpile image. Figure 3. Delineation of the muckpile image shown in Figure 1.

6 6 large patches of fines in the image, SPLIT sometimes mistakes these patches as a single large fragment. Secondly, if there is excessive noise on a fragment (due to bedding, rock texture, etc.), the SPLIT program may divide this fragment into a number of smaller fragments. Thirdly, some of the delineated particles are neither rock fragments nor fines and should not be counted in the final size distribution, such as the balls in Figure 3. The SPLIT program has built in editing capabilities to handle the situations described above. The SPLIT program first makes a stack of images, where one file in the stack is the delineated image pasted over the original grayscale image and the other file in the stack is the original grayscale image. The user can quickly toggle between the original and delineated images to determine which parts of the image need editing. Three kinds of editing are most common: paint bucket filling of fines, erasing unwanted delineations, and identifying non-rock features such as scaling objects. In most cases a skilled user can edit the images in less than 3 minutes [5] Calculation of the Size Distribution Once the individual fragments in the images have been delineated, the next step is to use characteristics of the fragments to calculate their size distribution. These characteristics include the area and dimensions of each fragment and the area of the non-particle regions (black areas). Screen size and volume of each fragment from these characteristics are determined. The second step is to determine a realistic distribution for the fine material. Two options for the distribution within the fines are available in SPLIT, a Schumann distribution and a Rosin- Rammler distribution. Each of these distributions has two unknown parameters and these parameters are determined from two known points in the size distribution, one point at the finesize and the other at 1.5 times the finesize. Figure 4 shows an example size distribution calculated from a muck pile image using the Schumann distribution assumption for fines. Both the full size distribution and various percent passing sizes (P20, P50 and P80) are given Presentation of the Size Distribution Results Once the size distribution has been calculated, it can be plotted in 4 ways: linear-linear plot, loglinear plot, log-log plot, and Rosin-Rammler plot. Figure 4 is an example of a log-linear plot. Next to each plot, the size distribution data is also printed in one of four formats (ISO standard, British standard, US standard, no standard). The P20, P50, P80 and topsize are also shown. The size distribution and percent passing sizes are written to files stored on the hard disk in text format for further manipulation in separate database or plotting programs.

7 7 Figure 4. Size distribution results of the first test Accuracy of the SPLIT System Many calibration studies have been conducted of the SPLIT system in the past several years. For calibration studies involving muck piles, images are taken of rock fragments in real field situations and processed by SPLIT and these rock fragments are also screened using traditional methods. According to the test results, the error in evaluation of the size distribution by SPLIT system is much less than 10 % and average error is about 5% [5], [8], [9], [10]. 4. CASE STUDY Two blast tests required for this study were conducted at Lafarge Yibita Lalahan Quarry. Mechanical and physical properties of limestone are given in Table 1. At this quarry currently used explosive is regular ANFO. Since the limestone is medium to hard rock, it is intended to prove the benefits to be gained by using another explosive having higher shock energy.

8 8 Table 1. Mechanical and physical properties of limestone. Uniaxial Compressive Strength, MPa 99.0 Indirect Tensile Strength, MPa 8.0 Density, g/cm p-wave velocity, m/s 6204 s-wave velocity, m/s 3156 Laboratory Dynamic Elastic Modulus, GPa Laboratory Dynamic Poisson s Ratio Surface blast design parameters of the quarry are given below: Hole diameter: 89 mm Bench height: 7.5 m Hole length: 9.2 m Burden: 2.60 m Spacing: 3.60 m Stemming length: 2.5 m Ignition system: Electrical Blasthole pattern: Staggered Delay between rows: 30 ms Above parameters were kept same for both tests. Tests were conducted at the same bench of the quarry. In other words, structural geology of the bench did not vary. The only changing parameter was main blasting agent. It is intended to investigate the effect of explosive on fragmentation at the same rock environment and surface blast design parameters. ANFO and BARANFO 50 were used as main blasting agents in the first and second tests, respectively. Technical properties of these explosives are given in Table 2. BARANFO 50 differs from ANFO in that BARANFO 50 consists of a mixture of prill porous and crushed ammonium nitrate. However, ANFO contains completely prill porous ammonium nitrate. BARANFO 50 has different detonation characteristics and energies partitioned during blasting when compared to ANFO due to the size of the ammonium nitrate. Table 2. Technical properties of ANFO and BARANFO 50. ANFO BARANFO 50 Appearance Prill Prill+Crushed Loading density, g/cm Average VOD, m/s Detonation pressure at given hole diameter and rock properties, GPa Effective shock energy, MJ/kg Effective heave energy, MJ/kg Total useful shock energy, MJ/kg Total useful heave energy, MJ/kg Total useful energy, MJ/kg

9 9 34 kg of ANFO is charged into each blasthole in the first test. On the other hand, 39 kg of BARANFO 50 is loaded into each blasthole in the second test due to its higher loading density. Both tests were monitored by continuous velocity of detonation recorder. Same sampling strategy is employed for the images acquired to compare the size distribution of the both muckpiles after blasting. Since the aim of this study is to compare the size distribution of muckpiles, images were acquired from the surface of the muckpile. If it is desired to obtain the real size distribution of the blasted material, loader should advance into the half of the pile and one should acquire the images from this part. 6 large scale and 6 medium scale images were acquired from each test. Totally 24 digital images were acquired by a digital camera having a resolution of 1024x1024. Having downloaded all images from digital camera to computer, they were transferred to SPLIT Engineering LLC by using FTP software. They were analyzed by using Split-Desktop software and the results returned back within three days. 5. DISCUSSION OF RESULTS Figure 4 and 5 show the size distribution results of first and second tests, respectively. Comparison of both shots in terms of resulting fragmentation is given in Table 3. Various percent passing sizes, P20, P50, P80 and top size values are shown in Table 3. P20, P50, P80 and top size in Case 2 are 2.52, 1.93, 2.05 and 1.94 times less than that in Case 1. Mean fragment sizes (P50) are cm and cm for the first and second tests, respectively. Therefore, finer fragmentation is obtained by utilizing BARANFO 50 as the main blasting agent. The change in main blasting agent affects fragmentation greatly at same rock environment and surface design parameters. The reason for finer fragmentation in Case 2 is that BARANFO 50 has much greater effective shock energy (Table 2). Effective shock energy governs fragmentation. The higher the effective shock energy, the finer the fragmentation is. It was observed after blasting that first muckpile profile was tighter than the second one while backbreak, overbreak and fly rock problems were not observed for both test blasts. Table 3. Comparison of both shots in terms of resulting fragmentation. Explosive P20, mm P50, mm P80, mm Top size, mm Case 1 ANFO Case 2 BARANFO Size Reduction Ratio

10 10 Figure 5. Size distribution results of the second test. 6. CONCLUSIONS Two blasts carried out within this research study were conducted by utilizing two different blasting agents, namely, ANFO and BARANFO 50 under the same test conditions (rock environment and surface design parameters). Images acquired after each shot were analyzed by using Split-Desktop software. According to the results of the software, mean fragment size of the muckpile obtained by utilizing BARANFO 50 as a main blasting agent is cm and that by using ANFO as a main blasting agent is cm. By using BARANFO 50 as a main blasting agent, P20 value reduced from 9.08 cm to 3.61 cm and, P80 value reduced from cm to cm. The most important result is that finer fragmentation is obtained by utilizing BARANFO 50 as an alternative blasting agent when two shots at Lalahan Quarry are compared. The reason is that BARANFO 50 has 55 % higher effective shock energy at this rock environment. If the quarry management accepts the previous coarser size distribution as satisfactory, in this case, it seems possible to expand the blasthole pattern to keep the fragmentation the same. It

11 11 would reduce the drilling cost for quarries where cost of drilling is high. Size distribution should also be determined while trying to expand the blasthole pattern. This study shows that the minor changes in blast parameters affect fragmentation greatly and size distribution should be quantified by using softwares which are proven to be accurate similar to SPLIT. Since fragmentation affects the efficiencies of post-blast mining operations (loading, hauling, crushing and grinding), fragmentation should be assessed properly which would lower the overall production cost of the company significantly. 7. ACKNOWLEDGEMENT We would like to acknowledge the help provided by the Lafarge Yibita Lalahan Quarry for the case study, BARUTSAN Company for the supply of explosives, accessories and blast monitoring systems and SPLIT Engineering LLC for the Split-Net Service. REFERENCES [1] Wu, X. and Kemeny, J.M., A Segmentation Method For Multi-Connected Particle Delineation, Proc. of the IEEE Workshop on Applications of Computer Vision, IEEE Computer Society Press, Los Alamitos, California, USA, pp , [2] Kemeny, J.M., Devgan, A., Hagaman, R.M., and Wu, X., Analysis of Rock Fragmentation Using Digital Image Processing, Journal of Geotechnical Engineering, Vol. 119, No. 7, pp , [3] Jimeno, C.L., Jimeno, E.L. and Carcedo, F.J.A., Drilling and Blasting of Rocks, A. A. Balkema, Rotterdam, Netherlands, [4] Higgins, M., BoBo, T., Girdner, K., Kemeny, J. and Seppala, V., Integrated Software Tools and Methodology for Optimization of Blast Fragmentation, Proceedings of the Twenty- Fifth Annual Conference on Explosives and Blasting Technique, Nashville, Tennessee, USA, Volume II, pp , February [5] Kemeny, J., Girdner K. and BoBo T., New Advances in Digital Image Analysis Software to Quantify the Size Distribution of Fragmented Rock, MINNBLAST 99, pp , [6] Franklin, J.A., Kemeny, J.M. and Girdner K.K., Evolution of Measuring Systems: A review, Proceedings of the Fragblast-5 Workshop on Measurement of Blast Fragmentation, A.A. Balkema, Montreal, Quebec, Canada, pp , August [7] Rustan, P.A., Automatic Image Processing and Analysis of Rock Fragmentation Comparison of Systems and New Guidelines for Testing the Systems, FRAGBLAST International Journal of Blasting and Fragmentation, Vol. 2 No. 1, pp , [8] Girdner, K., Kemeny, J., Srikant, A., and McGill, R., The Split System for Analyzing the Size Distribution of Fragmented Rock, Proceedings of the Fragblast-5 Workshop on Measurement of Blast Fragmentation, A.A. Balkema, Montreal, Quebec, Canada, pp , August 1996.

12 12 [9] Kemeny, J., A Practical Technique for Determining the Size Distribution of Blasted Benches, Waste Dumps, and Heap-Leach Sites, Mining Engineering, Vol. 46, No. 11, pp , [10] Liu, Q., and Tran, H., Comparing systems - Validation of Fragscan, WipFrag, and Split, Proceedings of the Fragblast-5 Workshop on Measurement of Blast Fragmentation, A.A. Balkema, Montreal, Quebec, Canada, pp , August 1996.