improved fragmentation through data integration

SME Annual Meeting Feb. 27-Mar. 02, 2011, Denver, CO

1 Copyright 2011 by SME

Preprint 11-091

IMPROVED FRAGMENTATION THROUGH DATA INTEGRATION

R. Owen, Freeport-McMoRan Copper & Gold Inc., Morenci, AZ

INTRODUCTION

An overwhelming amount of data can be collected around the blasting process. This information can include blasting product, pattern design, blast results, and routing. It is only when these data sources are combined in a concise and accurate form that they are of real use in determining the safety and economic implications of each parameter. Two key technological aspects of blasting are drill fleet management (DFM) and size fraction analysis (SFA). As these technologies are integrated as near-real-time measurement and QA/QC tools, the resolution and realization of blasting parameters is significantly increased. The Freeport-McMoRan Morenci Mine has integrated these technologies along with other existing database structures to create a responsive and sustainable tool for reconciliation, forecasting and parameter matching. This ability allows Morenci to customize blast patterns to meet the criteria most critical to each shot, whether the impact is influenced by routing and recovery, equipment maintenance or safety concerns.

DRILL FLEET MANAGEMENT

The current DFM was stabilized in its current form in mid 2009 after database and hardware maintenance met stability requirements. Currently the system is running on 12 production drills. The system receives pattern layout and is capable of semi-autonomous drilling over the entire district. The fleet is capable of drilling in highly altered intrusive units and hard granites through the range of sedimentary and volcanic units that comprise the Morenci District. The system brings a level of accuracy not achievable with earlier paint-marked patterns. Depth control is also increased along with more consistent wear and maintenance patterns due to semi-autonomous drilling.

The primary advantage of the DFM is to increase the realization of blast pattern design. This reduction in variation from design to actual allows for more accurate blast modeling. Manually surveyed patterns characteristically have large enough error to potentially invalidate correlations of blast parameters and blast results. As a primary element of successful blasting, energy distribution is achievable when assisted by the DFM. The current resolution of the DFM is around 6 inches in the horizontal and 2 feet in the vertical.

The vertical component is based upon hardware installed on the drill platform and is more physically maintenance intensive than the GPS and software-based system utilized for the rest of the DFM. This aspect is also the easiest and least expensive aspect of energy distribution to correct. A deep hole can be back filled using blasthole cuttings to reach target depth with ease. The driller manually tapes each blasthole to ensure quality control. This will catch any errors with the on-board system and allow for holes to be extended as needed.

The horizontal component requires the first pass to be accurate. Any adjustments to this parameter require the hole to be re-drilled or abandoned. Either of these options is time consuming and expensive. Current production patterns are completed with mean horizontal variation less than 6 inches. At this level of accuracy, a 10 to 12 diameter blasthole can be expected to deliver the modeled charge to the area designed.

The value add of the DFM does not stop at drilling accurate blast patterns. The data flow from the system gives both operators and engineers a suite of real-time and near-real-time tools for forecasting and adjusting mining strategy. The DFM utilizes a web-based interface

to send blasthole patterns to the drills, view drilling progress in real time and review the status of key parameters for the current and previous shifts.

Blastholes that are completed are sent to a server with a multitude of drilling parameters. This data includes the GPS coordinates for the collar of the hole, as well as multiple down-hole parameters that include the total drill depth. The time in which each phase of down-hole drilling occurs is recorded giving detailed penetration rate and energy consumption per drilling interval. This data is used to supplement geotechnical data and forecasting on drill time requirements. This incremental down-hole data is also valuable as a drill bit tracking source to ascertain bit performance and correlation to bit wear and drilling conditions.

SIZE FRACTION ANALYSIS

The SFA has been developed to capture images of active dig faces for all large production shovels. The images are manually cleaned of unusable images and automatically processed. This data is stored on a database where each image is collated to a GPS coordinate of the location where the image was taken. This combination of information has increased the ability to systematically judge fragmentation results.

The SFA cameras are mounted to the bottom of the shovel cab and are triggered to capture images only when the shovel is in dig configuration. This is achieved by reading the shovel geometry and control inputs to determine the shovels position relative to the bank. By using this Smart Trigger, it is possible to significantly increase the number of quality images taken by each camera. The cameras are equipped with lighting sources that allow the images to be taken 24 hours a day.

The images are transferred to a folder where they are checked for quality. Images that contain obscuration from dust or machinery are discarded. Images that pass this quality control are then analyzed for size fraction distribution and processed into a database. The database names each image and captures the time it was taken. From this time it is possible to determine the exact GPS point where the dipper was located when the image was captured. This position is used to code the images into the blasting block model for comparison to other parameters.

For the cameras to capture adequate data an average of 1 image per kiloton of shot material is required. The average number of images captured per kiloton is 5.8. The high number of images to shot material supplies the engineer with very clear trends in muck size. The total time for an image to spend in processing is 24 hours or less. This time frame allows operators and engineers to monitor the process of a shovel through a bench with day resolution to determine the quality of fragmentation. Any un-forecasted trends in lithology or geotechnical parameters are easily recognized and can be reconciled in the block model.

BLASTING DATA SOURCES

The SFA reporting allows for very detailed feedback on blasting performance while the DFM allows for blastholes to be placed within a half-diameter error of design. The other parameters necessary for blast design and reconciliation are also captured to create a full process map. This data includes bulk explosives information,



detonation details and accessory utilization. These items are combined with the blasthole location from the DFM to give precise kilocalorie calculations and high explosive tracking.

The bulk explosive data stored in the system include total weight of product loaded into the blasthole, the type of product used, the depth of the blasthole, amount of stemming loaded and any measured water levels in the blasthole. This information is stored for each blasthole. By linking this data to the exact coordinates of each blasthole, it is possible to model the kilocalories applied to the shot. This information is stored on the blasthole in the database and coded to the block model. The addition of blasthole depth and water level can be used to track water infiltration and build water models to forecast bulk explosive product selection and loading times. Acceptance sampling on all blasting products is completed to verify that product placement matches design and conforms to determined standards. This brings confidence to the kilocalorie calculations and reduces variation in the process. Any changes to product density and mixing ratios can have a large impact on blasting performance and are essential for quality control.

The detonation information captured in the database includes the detonator ID, timing and communication logs. The use of electronic detonators has greatly increased the accuracy and quality of blasting. Along with this immediate blasting performance benefit is the ability to track each detonator from delivery to consumption in the blasthole. When multiple detonators are required, the position of each detonator is recorded in the database and can be recalled to report form or visual display on the mine planning software. The timing information is also stored to analyze muck movement, pattern timing and burden relief. Recording the communication logs in the database allows the operators and engineers to know the exact detonator ID and location of any detonators that experienced communication issues. Any detonator issues can be visually displayed in the mine planning software or on the shovel routing map. The safety implications of this database feature are immediately apparent. The database provides the ability to readily track all communication issues for any electronic detonator used on property and track the precise location the detonator was loaded. This information can also be used to quality check blasthole loading and tie-in procedures as well as providing feedback to manufacturers about any detonator issues.

The database also allows blasting accessories to be stored and tracked for cost coordination and forecasting purposes. Each accessory used is linked to a specific blasthole ID and therefore can be related to blasthole condition and fragmentation requirements.

BLOCK MODEL

The blasting block model is the heart of the integrated system. The model allows all stored parameters to be combined into one area. The distribution of blastholes, blasthole parameters from DFM, product utilization, explosive energy, water content and many other parameters are imported. The blasting parameters are combined with geotechnical and grade information. This combination of data allows for detailed modeling of parameters and interactions.

The SFA data is interpolated into the block model where it is bounded by key relational parameters. This will be used to measure the performance of the blasts. Information is stored for several passing rates as well as nearest image and number of images used to code the block. Using these parameters, it is possible to qualify each block as a viable data point and reconcile forecasted results with actual results. This resolution of detail can account for lithology and faulting as well as blast pattern timing and bulk blasting product variance. The SFA can be reported in table and graph form or in visual format for plotting. These reports quickly inform the operators and engineers of areas that met forecast and those areas that have potential for improvement.

The bulk blasting product and ideal kilocalories are coded into the block model. These values are similarly reported monthly but are available on demand. These elements can be run through the mine planning statistical suite to identify correlations between product,

energy yield and geotechnical properties of the block to determine interactions between each.

Very structured tests conducted with varied blasting parameters can be used to define the base interaction of product, energy distribution, energy level and geotechnical zones. This information will be built upon by every shot conducted within the district, using the block model to capture each parameter. By following the results of the original experiment, parameter correlations should become more consistent. If the correlation between parameters starts to drift apart, other parameters should be considered. This iterative process assists blasters in dealing with ever changing mining conditions.

WEB BASED REPORTING

The consolidation and integration of this data allows for web based reporting that is accessible from any computer on the company network. This accessibility of near-real-time reports promotes improved decision making. These reports can be customized to track QA/QC processes, review design achievement and review cost tracking. Where the block model does the design and statistical work, the web base reporting delivers everyday reporting designed to suit the needs of diverse work groups from the same bank of information.

To achieve and maintain high levels of accuracy with the DFM it is important to give driller feedback. The data stored in the DFM is combined with the QA/QC blasthole quality analysis performed by the blasting crew to inform drillers of plugged holes or re-drill requirements. This, along with the drillers accuracy in achieving pattern design and use of automated drilling systems, is used to ensure consistent blast patterns. These reports are generated on a scheduled basis but can be run at any time to coach drillers and planners on best optimizing the system.

The ability to consistently achieve design on blast patterns is critical to understanding and optimizing the performance of each blast. The web based reporting is able to generate detailed reports that review the design parameters of each blasthole for each pattern for each district zone and compare the achieved parameters. This comparison can be reported directly after a blast pattern is shot, and again once the material has been mined. The initial report compares the design versus actual while the follow up report will compare the actual performance of the shot. These pieces are continually fed back into the block model where the iterative process of blast optimization is continued. These reports are built in drop down form to accommodate high level overview of district zone performance, blast pattern resolution or specific to each blasthole. This range of detail can easily fit the needs of upper management through engineering and blasting technician.

The design versus actual achievement often hinges upon the ability to use products as specified. A QA/QC process on blasting products used is a vital part of the integrated system and must be followed closely. The quick access to reports on QA/QC allows all members of the blasting process to review the daily results and identify trends that need to be addressed. Most processes vary over time depending on supply requirements, climate changes and blasting conditions. This variance must be understood and an acceptable range should be established. The web based reporting will use these tolerance parameters to flag aspects that fail to meet the design requirements. These reports can even be sent out automatically if parameters fail to meet required tolerances. Automatic reporting can often increase response time to failing systems and ensure blasting performance is not affected.

The ability to customize these reports to suit many needs allows the system to be transplanted to any mine site and allows interaction between sites. As data becomes more centralized and near-real-time reporting from sites and business headquarters grows, the ability to view on-demand reporting will continue to develop. The web based reporting used at Morenci is already in these advanced stages and continues to adapt to more data and more global requirements for process and product tracking.



CONCLUSION

This model of integration improves the ability of blasting to respond to geotechnical, economical and safety issues with high precision. Blast testing conducted under this blasting model has the advantage of quick data analysis with more detailed results. This ability to see impact on multiple variables with less time allows for greater integration of these test results into the production environment. By fully integrating these blasting aspects, blasting performance will better suit both upstream and downstream processes.

The ability of each aspect of the mining process to adapt in a timely manner to evolving needs is critical. This integrated blasting model allows blasting costs and results to vary as the mine progresses. As mines are expanded, steepened and new mines are designed, geotechnical parameters become more important. Integrated data allows the blast model to build correlations between geotechnical parameters and their impacts on both blast performance from a processing standpoint and highwall stability standpoint. This information then allows the blast design to be best optimized for the current economic situation. The ability to move blasting results to adapt to economic situations means real cost savings with minimum lag time.

Near-real-time reporting and complete data analysis ability are critical when evaluating the safety concerns associated with blasting. The energy expended with each blast is a key concern with every pattern. Inability to control energy adversely affects blast results and may put personnel and equipment in harms way. By reviewing the design and actual blast parameters of each blasthole, it becomes possible to analyze well-shot blastholes and less than optimal shot blastholes to evaluate what parameters can be optimized. Blast video along with recognized success or failure is less effective without good data collection to analyze.

Understanding the interaction of each variable involved in the outcome of a blast is a daunting prospect. The amount of data that can be collected on each blasthole, for each blast pattern, on each bench, for each pushback in every corner of a mining district is immense. The need for a robust and manageable database is crucial. The information must be integrated quickly and with confidence. The information then should be accessed in a clear and concise manner. The integrated blast model facilitates this with ability to grow as technology is advanced and becomes available.

The upstream and downstream process improvements are still being realized. Each mine site may have different requirements, and the requirements may change as equipment is changed. Blasting can accommodate crusher throughput limitations, contributing to the mass reduction from the initial blast. This can easily be evaluated from a cost standpoint allowing for the money to make the largest impact: more crusher capacity or more money spent in the initial blast. In leaching operations the reduction of ROM material may realize immediate and significant profit increases. From total recovery to recovery times, the size of material in a ROM leach pad is very influential. In milling processes the ability to impart micro-fractures and small initial input size can greatly reduce the energy and time requirements to realize particle size. Waste material may not require small size fractions, but fragments that are too large may cause unnecessary damage to equipment or loading hardships for personnel. By optimizing these and many more downstream processes it may be possible to reduce or streamline the types and amount of blasting products purchased. Improved maintenance may reduce work hours required to keep equipment running and increase production.

Blasting is the pillar of open pit mining and is directly influenced by many factors, both in design and terrain, and can influence much of the ultimate mine design and cost realization. The technology exists to reduce analysis time and build sufficient models to accurately predict rock fragmentation. The integration of data sources and the application of this data into the process is the key to success. The future of blasting is here and will only continue to develop as new technologies are developed.

INTRODUCTIONDRILL FLEET MANAGEMENTSIZE FRACTION ANALYSISBLASTING DATA SOURCESBLOCK MODELWEB BASED REPORTINGCONCLUSION

improved fragmentation through data integration

Documents