A mineral processing plant has one job: separate value from waste every second of every shift. When the circuit drifts and nobody sees it in real time, recovery slips, reagent bills balloon, and concentrate grade walks out the door with the tailings. Most plants don’t lose money in one big event.

They lose it in 2% recovery drops, 30-minute mill stalls, and grind size creeping coarser while everyone watches lagging shift reports. By the time the daily metallurgical balance lands, the metal is already gone. These ten KPIs are the ones a mineral processing operations leader should see updating live, with alerts firing on the deviations that actually matter.

Mill Throughput

• Why it Matters: Throughput drives revenue. Every tonne not milled is a tonne of unrecovered metal sitting in the stockpile or the pit.
• What it Measures: Tonnes of dry ore processed per hour through the grinding circuit, measured at the mill feed weightometer.
• What Happens if Missed: Hidden bottlenecks compound. Operators chase phantom issues while the real constraint, often a single cyclone or pump, stays invisible.
• Formula: Throughput (t/h) = Σ dry tonnes fed / Σ operating hours
• Indicator Type: Current. It reflects live grinding circuit performance against design capacity.
• Unit of Measure: Tonnes per hour (t/h)
• Ideal Visualization(s): KPI blocks for current throughput, KPI trend with real-time alerts on sustained dips, Pareto chart when ranking mill lines by throughput deviation
• Frequency: Real-time, 1-minute updates
• Data Required: Mill feed mass flow, ore moisture content, mill operating status

Metallurgical Recovery

• Why it Matters: Recovery is the difference between a profitable plant and a tailings dam full of money. Even 1% lost across a year is millions.
• What it Measures: Percentage of valuable mineral in the feed that reports to final concentrate, calculated by the two-product or three-product formula.
• What Happens if Missed: Recovery losses compound silently across shifts. Without live visibility, root cause hunts start days after the metal is already in the tailings.
• Formula: Recovery (%) = (C × c) / (F × f) × 100
• Indicator Type: Current. Modeled from online analyzers it updates live; based on shift assays it lags by hours.
• Unit of Measure: Percent (%)
• Ideal Visualization(s): KPI blocks for live recovery, KPI trend with real-time alerts on recovery drops, SPC trend for shift-over-shift variability
• Frequency: Real-time when modeled from online analyzers, otherwise every shift
• Data Required: Feed grade, concentrate grade, tailings grade, feed mass flow, concentrate mass flow

Concentrate Grade

• Why it Matters: Grade determines smelter payable terms and freight efficiency. Low-grade concentrate gets penalized at the smelter and costs more to ship per tonne of metal.
• What it Measures: Percentage by mass of the valuable mineral or metal in the final concentrate stream.
• What Happens if Missed: Operators chase recovery and let grade slip, or chase grade and bleed recovery. Without live grade visibility, the trade-off is guesswork.
• Formula: Grade (%) = mass valuable mineral / total concentrate mass × 100
• Indicator Type: Current. Online analyzers deliver live readings; shift composites lag the operating decision by hours.
• Unit of Measure: Percent (%) or grams per tonne (g/t) depending on commodity
• Ideal Visualization(s): KPI blocks for current grade, KPI trend with real-time alerts on grade slippage, XY scatter plot for grade versus recovery trade-off
• Frequency: Real-time with online analyzer, every 30 to 60 minutes with composite sampling
• Data Required: Online analyzer readings, concentrate mass flow, feed grade for reconciliation
• Pro Tip: The grade-recovery curve isn’t static. Watch how it shifts with feed mineralogy, reagent dose, and froth depth. The operating point should move with the ore.

Grind Size (P80)

• Why it Matters: Grind size sets the ceiling on recovery. Too coarse and minerals stay locked, too fine and energy and reagents are wasted on overgrind.
• What it Measures: The particle size at which 80% of the cyclone overflow passes, measured by online particle size analyzer.
• What Happens if Missed: P80 drifts coarser as liners wear or feed hardens. Recovery quietly drops for hours before assay results catch up.
• Formula: N/A
• Indicator Type: Leading. Grind size changes show up in recovery several minutes to hours later.
• Unit of Measure: Microns (µm)
• Ideal Visualization(s): KPI blocks for live P80, KPI trend with real-time alerts on coarsening, histogram of the full size distribution
• Frequency: Real-time, 1 to 5 minute updates from the particle size analyzer
• Data Required: Online particle size analyzer output, mill discharge density, cyclone feed pressure
• Red Flag: P80 coarsening with stable mill throughput and load. That’s usually cyclone wear or feed density drift.

Specific Energy Consumption

• Why it Matters: Grinding consumes 40 to 60% of plant electricity. A few kWh/t saved across a high-tonnage mill is a major operating cost win.
• What it Measures: Electrical energy consumed per tonne of ore milled across the comminution circuit.
• What Happens if Missed: Energy creep gets blamed on ore hardness when it’s often worn liners, suboptimal ball charge, or pump inefficiency.
• Formula: SEC (kWh/t) = Σ kWh consumed / Σ tonnes milled
• Indicator Type: Current. It moves with load, liner condition, and ore competence in real time.
• Unit of Measure: Kilowatt-hours per tonne (kWh/t)
• Ideal Visualization(s): KPI blocks for live SEC, KPI trend with real-time alerts on sustained rises, Pareto chart when ranking mills by SEC deviation from baseline
• Frequency: Real-time, 1-minute updates
• Data Required: Mill motor power draw, mill feed tonnage, mill operating status
• Pro Tip: Benchmark SEC against ore Bond Work Index, not against a static target. The ratio of actual to theoretical SEC tells you how efficiently the mill is running today.

Reagent Consumption per Tonne

• Why it Matters: Flotation reagents are one of the largest variable operating costs. Overdosing is wasted money, underdosing is lost metal.
• What it Measures: Mass of each reagent (collector, frother, depressant, modifier) consumed per tonne of ore processed.
• What Happens if Missed: Operators dose to a fixed setpoint regardless of ore type. Recovery drifts and reagent costs creep with nobody connecting the two.
• Formula: Dose (g/t) = reagent mass flow (g/h) / feed tonnage (t/h)
• Indicator Type: Current. Live dosing rates show whether circuits are over-fed or starved against present ore conditions.
• Unit of Measure: Grams or kilograms per tonne (g/t or kg/t)
• Ideal Visualization(s): KPI blocks per reagent, KPI trend with real-time alerts on dose deviation, Pareto chart when ranking reagents by cost contribution
• Frequency: Real-time, every 1 to 5 minutes
• Data Required: Reagent flow rates, reagent density, feed tonnage, ore type indicator
• Red Flag: Reagent dose climbing while recovery stays flat. That signals the circuit isn’t responding and there’s a deeper chemistry or feed mineralogy issue.

Cyclone Overflow Percent Solids

• Why it Matters: Classification sets up flotation. Wrong overflow density skews the particle size reporting to flotation and degrades both grade and recovery downstream.
• What it Measures: Mass percentage of solids in the hydrocyclone overflow stream feeding the flotation circuit.
• What Happens if Missed: A drifting nuclear density gauge or a worn apex masks a classification problem for hours. Flotation gets blamed for grinding’s mistakes.
• Formula: % Solids = mass solids / total slurry mass × 100
• Indicator Type: Current. It reflects the live operating state of the classification circuit.
• Unit of Measure: Percent solids by mass (%)
• Ideal Visualization(s): KPI blocks for current % solids, KPI trend with real-time alerts on deviation, Pareto chart when ranking cyclones by deviation from target
• Frequency: Real-time, 1-minute updates
• Data Required: Cyclone overflow density, slurry flow rate, water addition rate, cyclone feed pressure
• Pro Tip: Pair this with cyclone feed pressure. Pressure drifting up with stable flow usually means apex wear or roping, both of which kill classification.

Plant Availability

• Why it Matters: Availability is a direct multiplier on annual production. An hour lost on the SAG mill is rarely recovered, since downstream circuits can’t run dry.
• What it Measures: Percentage of scheduled time the plant or a critical asset is actually available to process ore.
• What Happens if Missed: Small repeating stoppages get coded as “minor” and never escalate. A 20-minute trip three times a shift is an hour of lost production nobody is accounting for.
• Formula: Availability (%) = Operating Time / Scheduled Time × 100
• Indicator Type: Lagging. It summarizes recent runtime against schedule.
• Unit of Measure: Percent (%)
• Ideal Visualization(s): KPI blocks per circuit, status history trends for outage causes, Pareto chart when ranking downtime events by cause and duration
• Frequency: Real-time status, with hourly and shift rollups
• Data Required: Equipment run status, stoppage reason codes, scheduled operating hours
• Red Flag: Availability above 95% with throughput below target. The plant’s running but throttled, which is a hidden bottleneck, not an uptime win.

Tailings Slurry Density and Water Recovery

• Why it Matters: Water is a regulated, cost-laden resource. Thicker tailings mean more recycled water, lower pumping costs, and safer dam operation.
• What it Measures: Density of thickener underflow and the percentage of process water recovered for reuse back into the plant.
• What Happens if Missed: Thickener upsets push thin tailings to the dam, raise freshwater intake, and quietly inflate water balance reports until regulators or auditors notice.
• Formula: Water Recovery (%) = recycled water volume / total process water consumed × 100
• Indicator Type: Current. Density and recovery move with flocculant dose, bed level, and feed solids in real time.
• Unit of Measure: Percent solids (%) for density, percent (%) for water recovery
• Ideal Visualization(s): KPI blocks for live density and recovery, KPI trend with real-time alerts on thickener bed level, SPC trend for stability
• Frequency: Real-time, 1-minute updates
• Data Required: Thickener underflow density, overflow turbidity, flocculant dose rate, bed level, recycled water flow
• Pro Tip: Watch flocculant consumption per tonne of solids. A rising flocculant dose with stable feed is the first sign of mineralogy or pH changes hitting settling rates.

Mass Pull

• Why it Matters: Mass pull links grinding, flotation, and the grade-recovery curve. It’s a fast proxy for circuit response when assays lag.
• What it Measures: Percentage of feed mass reporting to concentrate, measured by mass flow before assay results are available.
• What Happens if Missed: Mass pull drifts up and operators read it as recovery success when it’s really dilution. Concentrate grade quietly drops behind it.
• Formula: Mass Pull (%) = concentrate mass flow / feed mass flow × 100
• Indicator Type: Leading. Mass pull moves minutes ahead of any grade or recovery confirmation.
• Unit of Measure: Percent (%)
• Ideal Visualization(s): KPI blocks for current mass pull, KPI trend with real-time alerts on rapid shifts, XY scatter plot for mass pull versus grade
• Frequency: Real-time, 1-minute updates
• Data Required: Concentrate mass flow, feed mass flow, slurry densities
• Red Flag: Mass pull climbing while concentrate grade falls. That’s froth carrying gangue, often from over-aerated cells or too-shallow froth depth.

Why Real-Time Visibility Matters

Mineral processing operates on minute-by-minute trade-offs. A grind size drift, a reagent overdose, a thickener bed creeping up: none of these show up cleanly in a shift report, but each one moves recovery, grade, and operating cost in real time. When the visibility window is 12 hours wide, every shift starts behind, reacting to yesterday’s metal balance instead of today’s circuit.

The plants that pull ahead aren’t the ones with the newest equipment. They’re the ones where the control room, the metallurgist, and the maintenance lead all see the same live numbers, with alerts firing on the deviations that actually matter. Real-time KPIs turn a reactive plant into one that catches problems while the ore is still in the circuit.

How Transpara Can Help

If real-time operational visibility is a challenge you’re facing, you’re not alone. At Transpara, we help teams like yours gain clarity from complex systems without the need to centralize or overhaul your data stack.
Learn more about Transpara
Browse our documentation
Contact us