Sampling, scanning and image analysis of ore is going high-tech in all phases of mining – but reliable results still depend on getting the basics right

By Russell A. Carter, Contributing Editor

Sampling is mining’s version of the “quick look” feature available in most modern software systems. In the computing world, it’s a way to check the content and format of a file without interrupting work flow. In the mining sector, sampling can provide a glimpse into mineral deposit characteristics, particle size within a slurry, or whether a concentrate meets customer specifications, without disrupting process flow.

Because of its value as an indicator of what lies underground, in a crusher pile or a tailings stream, sampling is deeply tied to decision-making in almost every phase of mineral production. And, as multidisciplinary teams increasingly collaborate on common projects from diverse global locations, while management strategizing demands ever-more rapid data input, speedy but accurate collection and processing of samples is crucial to eliminate potentially expensive delays and information gaps, while allowing mining’s shift toward digital enablement to advance smoothly.

Equipment and system suppliers are expanding and filling gaps in their mechanical sampling, X-ray scanning and real-time mineral analysis portfolios to help the industry handle specific challenges. These include the need to handle — and sample — more material due to declining ore grades; reduce energy consumption during comminution by culling out less-profitable rock as early in the process as possible; and shrink the volume of tailings produced by conventional processes. Acquisitions also are in the mix as vendors seek to move the starting point for their services, equipment and software farther upstream in the mining sequence.

For example, earlier this year, FLSmidth announced it completed its acquisition of IMP Automation Group, an Australian company specializing in developing automated sampling, sample preparation and analysis solutions. The company said integration of IMP would enable it to support the expanding market for automated laboratories, which has experienced recent growth due to a combination of high exploration activity and increased focus on productivity, automation and digitalization.

“We also see great potential from the joining of our new colleagues to further enhance the development of our digital solutions for mining,” said Tina Knudsen, FLSmidth’s general manager for sampling, preparation and analysis–mining. “For instance, data collected from online analyzers and the laboratory can be used to optimize the entire flow sheet for mining operations. Using this data to augment our process optimization initiatives is an exciting prospect.”

Recognizing a growing need for faster, more flexible sampling collection, analysis and data-accessibility features and services at both the exploration and production phases, industry suppliers are adopting and refining techniques and software tools to provide clients with reliable data derived from on-site examination of drill cores and cuttings. Drilling services and equipment provider Boart Longyear, for example, now offers TruScan, a sampling-as-a-service program available as an independent service to mining clients or bundled as part of a drilling contract. Because the technology is designed to fit into the normal workflow of its drill crews, it doesn’t affect drilling productivity, according to the company.

“Mining companies and geologists are rapidly learning the advantages of having access to tools and services that provide quick and accurate geological data right on-site,” said Mike Ravella, vice president of geological data services at Boart Longyear. “The time to collect the geological data is drastically reduced, significantly lowering the cost of mineral exploration.”  (See Same-day Core Analysis Can Speed Up Project Progress, p. 29 (inserted below))

Same-day Core Analysis Can Speed Up Project Progress

Boart Longyear’s TruScan on-site core and chip scanning technology is designed to help increase the mining client’s understanding of their orebody faster and at a lower cost, by providing geologists with same-day continuous analysis of drill core.

According to the company, TruScan provides non-destructive, accurate, high-density elemental concentration data by using innovative XRF technology specifically built to scan rock. High-definition wet, dry and continuous scan photos (to scale) of the retrieved rock core can be quickly viewed by the geologist and aids in the logging and interpretation of the geology. This means more accurate conceptual geological models can be built while the drill is still on the borehole.

The company claimed TruScan is more accurate than traditional XRF technologies because it utilizes matrix calibration as well as novel hardware and software. TruScan detects elemental ranges from lithium through uranium as well as other key physical rock properties, which far exceed the ranges of traditional XRF technologies.

Its broader capability also enables TruScan to have strong correlation with lab analyses for a larger range of elements. When TruScan is used in pulps, comparison between TruScan and any lab analysis is within the range of error seen when comparing one lab to another. When averaging the continuous scan data from rock core or chip scanning, TruScan has been proven to reveal a very strong correlation with lab assay along the same scanning interval.

“TruScan also reveals the enormous heterogeneity in elemental concentrations along the same interval. This accuracy, paired with detailed data density, can be very useful for geological interpretation and conceptual models,” said Mike Ravella, vice president of geological data services at Boart Longyear.

TruScan is specifically calibrated to all elemental concentrations observed for each specific rock type. The process requires first sending samples to a lab for analysis to allow for calibration of the TruScan scanning tools. This typically occurs before the TruScan unit arrives at the mine site to maximize the efficiency of the technology.

Ravella said, “It is well worth the effort in the beginning, and extremely cost effective for the long term, to complete a calibration process. This maximizes the accuracy of TruScan for the client’s particular rock type going forward.”

Because it has been previously calibrated, TruScan does not require the operator to have knowledge of XRF analysis. TruScan can be operated by the drill crew, removing the need for additional staff on-site. No additional staff onsite both minimizes the cost to deploy this technology and keeps on-site environmental health and safety related risk low.

TruScan technology has been used on exploration sites for gold, copper, iron ore and other base metals. Boart Longyear noted that, because of the significant reduction in time needed to analyze geological data, mining companies can make more timely and accurate decisions on where to drill next or to what depth, which becomes a considerable cost savings.

“The majority of our clients today are major mining companies, but we fully expect TruScan technology to make sense for junior and midtier mining companies as well,” Ravella said. “The technology is extremely diverse, field robust, and saves mining clients time and money.”

(back to main story)

In a similar vein, geoscience software supplier Seequent announced an investment and partnership with Minalytix, a Canadian mining data software company. Seequent said it will work with Minalytix to maximize the value of the MX Deposit platform for Seequent customers. The cloud-hosted solution is designed to simplify how drill hole and sample data are collected, managed and shared. Users can log and access data from anywhere in the world.

According to Minalytix, MX Deposit’s real-time activity feed allows team members to post comments and questions either globally or to a specific group of people. Images and comments can be tagged to specific drill holes, samples, tables, records or values within a record.

Picking up Speed

Industry digitalization commonly requires awareness and acceptance of disruptive technologies. The extent to which these technologies are implemented, and the useful application of the new data sources that they provide, largely determine the effectiveness of digital initiatives. Although the mundane tasks of collecting, handling, processing and analyzing field or plant samples are far from disruptive, they’ll be joining fast company in the years ahead. It will be incorporated into new technologies aimed at helping producers avoid unnecessary drilling, blasting and haulage of waste rock, while speeding up the overall process-data flow needed to remove waste, boost feed grade and improve productivity.

The need for speed is driving development and commercialization of systems employing various technologies, ranging from innovative digital signal processing applications, to sensor “fusion” techniques, to robotics, that may yield potential savings in time and expense by:

Allowing real-time onboard ore analysis during drilling, on-site analysis of blasthole chip piles before a production pattern is charged with explosives or similarly quick analysis of shovel-bucket contents during loading operations.

• One of the stated goals of X-Mine, a collaborative project among European research organizations, X-ray scanning and imaging equipment vendors and mining companies, is aimed at developing novel sensing tools that can support core scanning and analysis while drilling.

• Canadian miner Teck partnered with MineSense, another Canadian company specializing in industrial IoT technology for real-time, sensor-based ore data and sorting solutions. In 2017, Teck began a full-scale trial of MineSense’s bucket-mounted ShovelSense technology at its Highland Valley copper mine in British Columbia, and after obtaining good results said it intended to expand trial usage to other mines.

• In a related development, MineSense and Siemens came to an exclusive cooperation agreement on the real-time measurement of ore grade and ore characteristics for conveyors. The resulting joint solution, according to Siemens, will provide unprecedented real-time measurement of ore grades and byproducts by MineSense’s BeltSense in combination with a single material and quality management system, Siemens Simine MAQ. This will enable customers to significantly increase efficiency via a single view of quality across the whole conveying process, said Siemens.

Offering ore-sorting systems that can reduce ore dilution to drastically cut the amount of waste rock that a plant has to handle.

Interest in this technology appears to be growing rapidly among producers of everything from diamonds — where the earliest applications of ore-sorting technology originated — to precious metal producers, iron ore miners and rare-earth extraction operations. A few examples:

• Hecla Mining has been testing a Rados XRF (X-ray fluorescence) sorting machine at its San Sebastian Au/Ag mine in Mexico, with results favorable enough to warrant ordering another machine for its Casa Berardi gold mine in Quebec. Fire-weed Zinc tested an XRF machine at its Macmillan Pass zinc project in Canada’s Yukon, with early results from bulk-tonnage testing showing the potential to upgrade feed grade from 2.5% Zn to 5% Zn, with metal recoveries ranging from 80%-85%.

• NextOre, a company spun out of the Australian research organization CSIRO, raised $2 million in a private funding round primarily to ramp up manufacturing and sales of its flagship bulk ore-sorting sensor system, according to a mid-June announcement. The company said its copper-detecting analyzers, which employ magnetic resonance technology, have been installed at mines in Latin America and Australia, with more installations scheduled for 2019. While NextOre is focusing initially on copper sorting applications, it said MR technology is applicable to other commodities including iron ore and gold.

• Southern Innovation, another Australian company, offers its SITORO digital signal processing technology, which the company claims can dramatically enhance existing analytical techniques in a number of applications. The company pointed out that the techniques currently used for online material analysis, including XRF, XRT (X-ray transmission) and PGNAA (Prompt Gamma Neutron Activation Analysis), often may be slow, intermittent and inaccurate; have limited capability to detect light elements or low concentrations; provide poor representivity of results; and can impose burdensome radiation safety measures. SITORO, according to Southern Innovation, can enhance the speed and accuracy of each of these techniques.

Southern Innovation has worked with BHP since 2015 on various projects including real-time characterization of iron ore, coking coal and potash on conveyors. More recently, BHP has been testing a SITORO-based, on-rig device called Drill-Scan that can provide real-time analysis of reverse-circulation drill cuttings. Following favorable results from prototype trials, the company reportedly has commercialized the design and sold two units to BHP for further field testing.

Providing faster, more accurate sample results for metallurgical accounting, plant control and audits.

• In a presentation given at the 2018 CIM conference, Tom Strombotne, global product manager–minerals at Thermo Fisher Scientific, noted that when it comes to monitoring flotation operations, for example, modern sampling and online analysis systems need vastly higher capacity than the equipment typically installed in brownfield plants a few decades ago. Sampling in the industry’s newest, largest concentrators must be capable of handling up to 32,000 m3/h throughput. High assay data rates provided by state-of-the-art dedicated — as opposed to centralized — sampler analyzers enable rapid detection of process upsets and feedback for process control, allowing greater recovery gains. He pointed out that designing a large flotation plant with more than 40 dedicated samplers and analyzers — and without a single sample pump — represented a “break through” achieved just during the last decade.

• Scantech, an Australian company that offers on-belt mineral analyzers using a variety of measurement technologies, reported on a Scantech Geoscan-MPGNAA system installed at a Middle Eastern phosphate mine in 2017.

Placed just downstream from the crushing plant, the system’s primary purpose is to facilitate real-time mine grade control. Prior to installation, mine personnel had to wait three hours for laboratory data. Continuous 2-minute analysis of minerals of interest (P2O5, CaO, MgO, and SiO2) as well as moisture in the conveyed material provided by the Geoscan-M system allow the mine to take almost immediate actions for grade control as well as for adjustments in downstream processing.

The Geoscan-M uses a Californium-252 source to excite elemental nuclei in the conveyor flow with neutrons, which in turn emit gamma rays that are detected by a detector array above the conveyor, allowing individual elements to be measured directly.

The unit has a patented no-belt-contact design and no wear components, according to Scantech. This enables minimal maintenance requirements and provides enhanced safety by careful design of shielding and access points. Typically, said Scantech, no isolation area is required around the unit during normal operations.

Getting It Right, From the Start

Although the future of sampling technology looks to be briskly moving along a path toward sophisticated, high-speed semi- or fully-automated systems, it’s still a concept based on probability — and industry sampling practices often make high probability less than a sure bet. A 2016 paper authored by D.E.G. Connelly, director/principal consulting metallurgist at Perth, Australia-based Mineral Engineering Technical Services titled Metallurgical Accounting-Systems & Procedures for Modern Day Mineral Processing Plants paints a less-than-ideal picture of actual sampling-system integrity.

Connelly wrote, “Great responsibility rests on a very small sample. Therefore, it is essential that samples are truly representative of the bulk of the material. Proper collection of a representative sample requires an understanding of the physical characteristics and mineralogy of the material to be sampled as well as determination of the minimum number and size of increments to be taken in order to produce the overall sampling precision required.”

He provided anecdotal examples of problems observed over a number of years within the Western Australia gold industry, noting that, “An overview of common industry practices reveals very few operations have, and maintain, automatic samplers. Manual sampling is commonplace and often a cause for concern, unless the importance of the sampling is explained and reinforced with operators and management. The reality is such that, within twelve months, even new plants let the samplers fall into a state of disrepair.”

To get a better sense of the state of sampling pitfalls and performance standards, E&MJ turned to McLanahan Corp., a Pennsylvania, USA-based company that has supplied crushing, sizing, separation and mechanical sampling equipment for more than a century. Its product line includes cross-belt, falling-stream and auger sampling systems.

Adam Orner, McLanahan’s global product manager–sampling systems, commented via email on sampling practices, system selection and maintenance/training factors that can impact overall sampling system accuracy and reliability. His experience in the field seems to agree with Connelly’s findings.

Sampling systems, he explained, are small-scale material handling plants, and correct system design requires incorporation of best practices from both sampling and bulk material (or slurry) handling. “From the sampling perspective, there are basic requirements that must be met to assure that sample collection is correct.  If these basic requirements are met, it is highly likely that the overall sampling process will be successful,” Orner noted.

“The most important concept is the idea of equiprobability — a characteristic of the process that means all portions of the consignment (lot) being sampled have equal probability of being collected as sample. For example, if all of the material contained in a consignment is discharged from a conveyor past the point where a sampler can collect an increment, the concept of equiprobability is being respected. If samples are being removed from a stockpile, it is not likely that the concept of equiprobability is being respected since it is unlikely that that all portions of the stockpile will have an equal probability (or maybe any possibility at all) of being collected as sample.”

Other critical requirements include:

• Collecting a required minimum mass of sample and minimum number of increments (usually varies by application);

• A properly defined or “delimited” sample increment;

• Complete removal of the delimited increment (especially fine particles);

• Appropriately sized cutter openings (usually a minimum of 3x the material size);

• Straight and parallel cutter lips that are reasonably sharp and in good condition;

• A cutter that can completely pass (or contain) an entire increment (usually with some additional capacity for safety);

• An appropriately designed drive unit that assures constant, correct cutter speed during sample collection; and

• Cutter motion that intersects a complete cross section of the material flow.

There are typically no one-size-fits-all sampling solutions, according to Orner. “There are many similar sampling applications, but identical applications are rare. In conversation, we sometimes refer to one particular sampling application in terms of another — usually something like ‘Application A is just like application B, except for…’,” he explained. “Much of what is the ‘best’ choice for a sampling application is determined by process layout, plant operating requirements, and similar items.  Some options for sampling equipment and system configurations will be a better choice depending on the constraints associated with a particular application.

“That said, in hardrock mineral applications, particularly for accounting purposes, falling stream type samplers are most commonly implemented,” he said. “Designing and installing these types of samplers is often not the most cost-effective option from a capital expense standpoint, but in terms of simplicity of operation, effectiveness, reliability, and maintenance requirements, it is difficult to beat falling stream samplers in the long term.”

Orner cautioned that regular inspection and maintenance of sampling equipment, along with education of staff involved with the sampling process, are important considerations that are often overlooked: “Sampling equipment is like any other in that, over time, parts wear and components will need adjustment. Regular inspection to identify potential issues and a maintenance plan to address these issues is key to assuring correct samples are collected and equipment operates reliably. 

“Simply operating sampling equipment and having material in the final sample container does not necessarily mean that the sampling equipment and process is working correctly. Unfortunately, once a sampling system is installed and commissioned, it can become somewhat lost in the background with respect to the overall operation of the facility in question. This can mean that the sampling equipment will receive minimal attention unless a machine is somehow obviously not working correctly, an alarm condition is set, or no sample material is present in the collection container at the end of a lot. 

“You’ve heard the expression about the devil being in the details. In sampling activities this is true more often than not,” he said. “The details are what we typically need to get right to have a successful quality or process control program. This concept is why it is very important to implement a periodic inspection and maintenance/adjustment plan for sampling systems and equipment. Depending on your application and type of equipment, this inspection and maintenance plan can take on a variety of forms, but your sampling equipment manufacturer can help you identify common areas for inspection and maintenance, as well as offer some guidance as a starting point for scheduling purposes.”

And, he added, “Education and training of personnel involved with the sampling process also is important. From the time a sample is physically collected up to the point where data is developed, many people can be involved — from mechanics who maintain the equipment, to operators running the equipment, to lab technicians processing the sample. If you consider that many facilities operate more than one shift, the number of people who may somehow touch the sampling process can be large. Given this, it is very important that all personnel involved in the sampling process have at least a basic understanding of sampling concepts, the importance of correct sampling, and an overview of the complete sampling process at their facility.”

He suggested that training personnel to understand why their particular function is important in the larger scheme, and what the overall plan is trying to accomplish, can encourage personnel to be more aware of their daily tasks and more engaged in making their portion of the process better, thereby improving the overall sampling process.

When it comes to selecting the right system, a reputable supplier will try to correctly match the best type of equipment, system layout, operation, etc. to the customer’s needs. However, he noted, “The best sampling solution is not always the least expensive solution. In cases where a low capital cost is prioritized, compromises are often made that can have a negative effect on the overall sampling process. It’s important to implement a sampling solution that meets your needs and goals — even if it may have a higher short-term cost — or potentially risk losing money over a longer period of time due to decisions made based on poor data derived from poor sampling.”

Share