From waste to value: How SRK consulting and TOMRA Mining are redefining pebble processing

Pebble recirculation is standard practice in mining operations. This approach is costly, energy-intensive, and can limit overall plant performance. In some operations, pebbles can represent up to 30% of the mill feed, increasing complexity and reducing efficiency. SRK Consulting and TOMRA Mining explain in this article that pebble streams are not uniform, and that selecting what to process can improve performance. In documented cases, this has delivered up to a 6% increase in throughput, along with clear economic benefits.

In many large-scale mining operations, particularly those processing hard, competent ore, grinding circuits are quietly dealing with a significant inefficiency that often goes unnoticed. In many cases, this inefficiency is simply accepted as part of the process,  despite its cost. Between 5% and 30% of the material entering the mill can report back as larger, hard pieces known as “pebbles,” particles that do not break easily and are recirculated within the process, either directly or after crushing.

These pebbles do not simply move through the system. They occupy space inside the mill, consume additional energy when processed again, and can limit throughput and stability. In high pebble rate circuits, the effect is considerable, for every ton of pebbles returned to the mill, between 0.4 and 0.7 tons of fresh feed may effectively be displaced.

Even after crushing, pebbles do not disappear; they continue to impact mill performance. In many cases, this has simply been accepted as part of the process.

Beyond the technical impact, this inefficiency comes at a real cost. Higher energy consumption, reduced throughput, and increased wear on grinding and crushing equipment all translate into lost revenue and higher operating costs. Pebble crushers, often required to manage this stream, are also subject to high mechanical stress, frequent maintenance, and operational challenges, particularly when grinding media enters the system.

This has been particularly well documented in large copper operations, but the same challenge is also seen in gold and other mineral processing operations treating hard, competent ore.

Despite this, pebbles are still often treated as a uniform stream, assumed to carry enough value to justify reprocessing. However, closer analysis shows a different reality. On average, pebbles may appear to have a similar value to the feed. In reality, each particle can be very different; some contain little or no value, while others carry significantly higher metal content.

This gap between assumption and reality is starting to change how grinding circuits are viewed. Instead of asking how to process all pebbles, operators are now questioning whether all of them should be processed at all.

The overlooked problem in grinding circuits

While pebbles are widely accepted as part of the grinding process, their behaviour inside the circuit is often not fully understood. As Adrian Dance from SRK Consulting Canada’s Vancouver office explains, pebbles are the result of how different ore types respond to grinding. Some particles break easily, while others, typically harder and more resistant than the original feed, remain at “critical size”, where they are difficult to break further. When these accumulate, they interfere with grinding efficiency.

Even after crushing, pebbles continue to circulate through the mill, occupying volume and reducing the system’s ability to process fresh ore. The higher the pebble rate, the greater the impact on throughput and stability.

At the same time, this recirculation increases operating costs, adding energy consumption, water use, and wear across both grinding and crushing equipment.

Not all pebbles are equal

Pebbles are typically evaluated as a single stream, often reported at around 60% of the feed grade, reinforcing the assumption that they should all be reprocessed.

However, detailed studies show no clear relationship between particle size and metal content, meaning that screening alone offers little opportunity for upgrading or waste rejection.

At the particle level, the picture is very different. Pebble streams can contain a wide distribution of metal values, with individual particles carrying significantly higher grades than the average feed.

In practical terms, two pebbles of similar size can have completely different value, one below the cut-off grade, the other containing substantially higher metal content.

Test work reviewed at SRK confirms this variability. “A significant portion of the contained metal can be concentrated within a smaller fraction of the mass, with up to 80% of the metal found in around half of the particles, leaving the remaining fraction largely below cut-off grade. This is not just a theoretical concept: selective processing of pebble streams has been shown to increase both feed grade and throughput, reinforcing the economic potential of this approach”, says Adrian Dance

Rethinking a long-standing assumption

Every ton in the pebble stream has already consumed energy and will consume more if it is sent back through the circuit. If pebbles are not uniform, the logic of reprocessing all of them becomes increasingly questionable.

Grinding circuits have traditionally been designed to process all material entering the mill, with pebbles treated as an unavoidable part of the loop. However, growing evidence shows that selectively rejecting part of this stream can deliver measurable benefits, including increased throughput and improved feed grade.

Separating value from waste in pebble streams

If pebble streams contain a mixture of value and waste, the next step is to separate them before they re-enter the grinding circuit.

This is where sensor-based sorting offers a practical solution. As Fernando Romero-Lage, Area Sales Manager TOMRA Mining, explains, “X-ray Transmission (XRT) technology analyses each particle based on its internal atomic density. As pebbles pass along a conveyor, the system scans every particle in milliseconds and separates valuable material from low-grade or barren rock.”

In practice, this allows a significant portion of the contained metal to be recovered from a smaller fraction of the mass, with test work showing that up to 80% of the metal can be concentrated in around half of the material.

“This particle-by-particle approach directly addresses the variability observed in pebble streams. Rather than treating the material as a bulk flow, it allows operations to selectively retain value while rejecting material that would otherwise add cost without contributing to recovery,” Romero-Lage adds.

Pebble streams are particularly well-suited to this type of application. Typically, already screened, washed and conveyed, they present a controlled and stable feed, allowing sensor-based systems to operate efficiently and consistently.

In practical terms, this means that low-value pebbles can be removed before they add unnecessary load to the circuit, while higher-value material continues downstream. The grinding circuit is no longer forced to process everything, only what adds value.

From wasted recirculation to measurable gains

The impact of this approach extends well beyond metal recovery.

In well-documented operations, even simple pebble rejection has delivered measurable economic impact. In a copper operation in Canada, rejecting lower-grade pebbles increased throughput by up to 6%, translating into an estimated $21 million in additional annual revenue. In another copper operation in Peru, improvements in both feed grade and throughput resulted in multi-million-dollar gains, highlighting the economic potential of a more selective approach.

By reducing the volume of low-value material circulating in the system, operations can increase throughput, improve energy efficiency, and stabilise mill performance. With fewer hard, resistant particles returning to the circuit, the grinding process becomes more efficient and predictable.

At the same time, the load on downstream equipment is reduced. Lower recirculating volumes mean less demand on pebble crushing circuits, reduced wear, and fewer maintenance challenges, particularly in operations where crusher reliability is a concern.

More importantly, this approach changes how grinding circuits are fundamentally understood. Pebbles are no longer treated as an unavoidable by-product, but as a stream that can be actively managed and optimised.

In this context, sensor-based sorting enables a shift from processing everything to processing what truly adds value, turning a long-accepted inefficiency into a measurable improvement in throughput, energy efficiency, and overall plant performance.