How to Recover Lost Value When Grind Size Is Out of Spec

For most mining companies, losses from improper grind size are one of the most common yet costly issues. Losses from recovery drop when grind size is outside the optimal target range, whether it is too coarse or too fine. In addition to lower recovery rates, energy use increases, and reagent costs rise. Since the damage accumulates gradually, it is often overlooked, unlike broken machinery, which is more visible.

This article examines the reasons from poor control of grind size, the costs associated with poor grind size control, and the methods mining companies are using to improve control of grind size, including the use of inline sensors, artificial intelligence, and grinding circuit control.

Grind Size Impact on Recovery

Grind size influences the degree of mineral liberation, which is the separation of valuable minerals from the waste rock, and this forms the basis of flotation recovery. The relationship between grind size and recovery rates is very sensitive to change. The points below illustrate the impact of grind size on flotation recovery and reagent consumption.

Too Coarse: This is when the grind size is too large, resulting in the loss of valuable minerals which are trapped in the composite particles. Subsequently, the flotation bubbles are unable to capture the minerals, and reagents are wasted.

Too Fine: On the other end of the spectrum, if the grind size is too fine, ultra-fine particles which are below 15-20 microns will float poorly. This will result in excessive reagent consumption and poor froth stability.

Research has shown that there can be several percent improvements in recovery by having a consistent and achievable target grind size – converting to millions of dollars in recovery per annum (JXSC, n.d.; 911Metallurgist, n.d.). One study of flotation circuits found that variations in ore grades can explain 68.9% of variability in rock-to-metal ratios, and grind size is a variable that can be controlled by operators to affect that variance in a significant way.

In a Zambian copper operation, recovery improved after changing target grind size from 75μm to 65μm, demonstrating the fine margins within which concentrators operate (GSSRR, 2015). Even a small adjustment in grind size can lead to inefficiencies across flotation, dewatering, and downstream processing.

Key Statistic

Most of the energy in a concentrator is used in comminution (50-60%). Because of that, the financial benefits from even a small improvement in recovery from improved control of grind size are significant.

Grind Size Issues and Root Causes

Failing to maintain target grind size has many root causes, both in systems and collateral equipment. The most common reasons for ineffective target grind size and thus recovery are:

Ore variability and grade inconsistencies: Factors such as hardness, density, and moisture content of the ore can affect how material is fragmented in the SAG or ball mill. An unexpected tough ore domain can shift P80 by 20-30μm in a single shift.

Lack of Real-Time Monitoring: The historical approach most sites use is continuous particle size feedback, which has led to operators being overly cautious and reactive to the point of running too coarse with the intent of avoiding overgrinding.

Disconnected Systems: Sensor data collected in the plant historian often goes untapped for actionable control. Recovery information never gets to the mill operator as the flotation and grind circuits are separated.

Legacy Control Strategies: Conventional PID controllers optimize only a single variable in a complex system. Changes in feed consistency, liner wear, or water addition become entirely invisible, and operators are left to fix adjustments post-chase.

What the Cost of Recovery Loss on Grind Size Is

The impact of sub-optimally grinding is large and complex in the short run. This is why it is easy to underestimate. For example:

Recovery of revenue lost. For a mid-sized concentrator, losing one percentage point to top-tier flotation recovery can cost millions annually.

Energy waste. Re-circulation is required in both under- and over-grinding. This process is energy inefficient as it consumes a large amount of energy (kWh per tonne). AI-assisted grinding circuit optimization has proven energy savings of 5–10% in multiple documented cases.

Reagent over-consumption. Ineffective recovery consumes excessive amounts of collecting agents and frothing agents in an over-grinding process, which leads to a direct loss in operating costs.

Accelerated liner wear. Running mills under off-spec conditions, specifically overload caused by hard ore, unavoidably accelerates liner and media wear, increasing maintenance costs, and elevating the risk of unplanned downtime.

Industry analysis shows that plants operating at optimal grind size outperform peers in terms of unit cost and recovery rate, the two main drivers of concentrator profitability.

Ways to Recover Value from Off-Spec Grind Size

1. Install Inline Particle Size Analyzers

Real-time measurement of grind size using acoustic and laser-based sensors has been made possible by removing the 4-8 hour delay associated with laboratory sampling. While using real-time data, cycle control at the target P80 levels is achieved by continuously and automatically adjusting mill speed, load, water addition, and feed rate.

2. Use AI-Driven Optimization of Grinding Circuits

AI and machine learning in the context of real-time sensor data, historical mill behavior, and ore feed characteristics can predict the level of control drift and grind size, thus allowing for proactive control as opposed to reactive control. Unlike traditional PID controllers, AI systems can optimize multiple control variables (feed rate, mill speed, water addition, and classifier settings) at the same time to minimize energy usage while achieving the targeted P80.

Industry Results
AI-optimized closed-loop control in grinding circuits has delivered 2-5% improvement in throughput and 5-10% reduction in energy consumption for grinding without additional capital costs for new equipment.

3. Integration of Mill Data and Flotation Results

The greatest improvement opportunity is closing the feedback loop between grinding size and downstream recovery metrics. When mill operators receive flotation recovery data in near real time, they can understand the grind size deviation's real economic impact and make decisions based on revenue impact instead of just focusing on the process targets.

It will also improve geometallurgical mapping: linking the characteristics of the ore block upstream and the expected grind and recovery, so teams in the plant can be proactive, instead of just reacting to the geometallurgical mapping.

4. Standardization of Operating Envelopes by Ore Type

Not all ore is the same, and different domains within the same deposit can behave very differently in the grinding circuit. Control strategies that are specific to ore type should be developed. These should include specific settings for mill speed, load, and water addition based on the ore's hardness and fragmentation profiles in order to minimize operating judgments during transitions and to avoid over grinding when shifts in blending occur.

How NTWIST Assists in Achieving Consistent Grind Size and Maximum Recovery

At NTWIST, we have AI-driven tools for process optimization that are aimed at reducing the gap between variability in ore and the performance of the mill. By integrating real time sensing data, ore hardness and fragmentation data, and flotation KPIs into a single decision making framework, we assist operators in consistently achieving the target grind size and confidently recovering the lost value.

TheMillMaxapplication addresses issues with the stability of comminution processes by providing the ability to predict and acting to keep the particle grind size to spec for the entire shift for metallurgists and plant managers.

Find out how we Stabilize Throughput and Maximize Recovery:millmax

In Closing

When the grind size is out of spec, the mine loses more than recoverable material, it also loses predictability, efficiency, and profitability. The damage is cumulative and often goes unnoticed until a reconciliation is done and the data is reviewed weeks later. Recovering that lost value is achieved by a more proactive approach than just correcting the problem after the fact. By utilizing integrated, real-time sensor data, artificial intelligence (AI) based control and optimization of the grinding circuits, and a closed loop control of both the milling and flotation processes, the operations personnel can maintain the grind size variability to a very narrow range and provide significant measurable benefits that directly improve the financial bottom line.

People Also Ask

Q1. What is grind size recovery loss in mining?
Grind size recovery loss is a term referring to the loss of valuable minerals that can be recovered. This occurs when a grinding circuit is producing grind (or particle) size that is either too coarse (meaning the minerals are unliberated) or too fine which leads to the loss of the fine particles to a slime (which includes or consumes reagents) and or the over consumption of reagents. In either of these scenarios, the flotation circuit is recovering less metal than it is capable of, which leads to potential revenue loss that is ultimately directed to the tailings dam rather than the concentrate.

Q2. How much recovery can be lost when grind size is off spec?
The effect is ore type and circuit design-specific, but even 5-10 microns flotation recovery at target P80 can cost several percentage points. In copper operations, one percentage point drop below best-in-class performance means loss of several million dollars for mid-scale concentrator on an annual basis. This makes grind size one of the most valuable control variables for plant teams.

Q3. What causes grind size to drift out of its target range?
Most commonly, this is due to ore variability (hardness, density, moisture), lack of real-time measurements of particles, a system where the sensors control data mostly remain unconnected, and classical PID controllers that, in a sense, are blind to the different, interconnected variables why grind size shift (feed rate, water addition, liner wear, etc.) and of the role that these variables play.

Q4. How does AI improve grinding circuit optimization?
AI optimizes the control of four or more grind size related variables (feed rate, water, mill speed, and classifier position) at the same time using correction algorithms to keep the circuit within the goals given its target limits. In contrast to PID controllers, AI control takes into account the non-linear and time-variant characteristics of these variables and ore, liner, and environment changes. This control is refined to the level that in most cases throughput gain is documented to be 2-5% and grinding energy consumption is reduced by 5-10% in level of the control.

Q5. What is the relationship between grind size and flotation recovery?
Froth flotation recovery statistically follows a particle size distribution and is a function of particle size. There is an ideal recovery window between 20-150 µm (depending on the specific ore being processed) as these minerals are fully liberated and are attracted to the hydrophobic/air bubbles. Particles below this size become ultra-fine and are hydrodynamically difficult to recover. On the other end of the spectrum, larger composite or granular particles also become difficult to recover as they aren't fully liberated and are rejected. Therefore, the most important factor to optimizing flotation performance (before improving reagents, modifying flotation circuits, etc.) is just having the right grind size.