Open Pit Gold Mining Process
and Workflow Overview

A gold deposit advances toward production through scoping study, pre-feasibility study (PFS), and definitive feasibility study (DFS). Each stage narrows geological, engineering, and economic uncertainty at progressively higher cost. A DFS for a major open pit project runs $30–60 million and takes two or more years.

The resource classification system governing these studies (JORC in Australia, NI 43-101 in Canada, SAMREC in South Africa, S-K 1300 in the United States) sorts mineral resources into Inferred, Indicated, and Measured categories based on geological confidence. In practice, confidence is a function of drill spacing relative to the spatial continuity of grade, and spatial continuity is described by the variogram.

At the Turquoise Ridge joint venture in Nevada (Barrick/Newmont), the sediment-hosted Carlin-type gold mineralization has variable continuity depending on structural position and host rock alteration. In some zones, gold grade is reasonably continuous along favorable stratigraphic contacts over tens of meters. In other zones, particularly near fault intersections, high grades cluster in pods with dimensions smaller than the drill spacing. The variogram that describes the first zone and the variogram that describes the second zone are different animals, and the resource estimate has to handle both within the same deposit.

Cutoff grade connects geology to economics. It is calculated from gold price, mining cost per tonne, processing cost per tonne, and metallurgical recovery. The gold price assumed in a feasibility study is the most powerful lever available for changing the apparent size and value of a project. A study run at $1,950/oz gold will show a larger pit, a lower strip ratio, a longer mine life, and a higher NPV than the same deposit studied at $1,600/oz. The geology does not change between these two studies. The arithmetic around it changes.

Pit optimization software (Geovia Whittle, Datamine NPV Scheduler, Maptek Evolution) generates a set of nested pit shells at progressively higher revenue assumptions. The Lerchs-Grossmann algorithm, published in 1965 and still embedded in every commercial pit optimizer, finds the ultimate pit boundary that maximizes undiscounted profit subject to slope angle constraints. Scheduling algorithms then sequence extraction across the nested shells, grouping them into pushback phases that balance ore access and waste deferral.

Strip ratio is the ratio of waste tonnes to ore tonnes and is the dominant driver of mining cost. At the KCGM Super Pit in Kalgoorlie, life-of-mine strip ratios have been in the 5–6:1 range. At Barrick’s former Cortez Hills open pit in Nevada, the strip ratio varied by cutback phase, with earlier phases at 2–3:1 and deeper phases exceeding 7:1. This escalation with depth is geometric: each new bench at the bottom of the pit requires removal of a larger annulus of waste to maintain slope angles, so the incremental strip ratio of each successive bench increases even if the overall ratio remains stable for a time.

Slope design varies by geotechnical domain. Rock mass quality, assessed through systems like RMR or GSI, controls the achievable inter-ramp angle. Fresh, competent granodiorite with widely spaced, tight joints might sustain 52–55 degrees. A domain intersecting a major fault with clay gouge and high pore pressure may require 38–42 degrees. The design is not a single number for the entire pit. It is a set of sector-specific angles, each backed by oriented core logging, structural mapping, hydrogeological data, and sometimes numerical stress modeling. Slope stability radar (GroundProbe SSR, IDS IBIS-FM) provides real-time wall displacement monitoring during operations, triggering exclusion zones if displacement rates accelerate beyond threshold values.

Scheduling optimization in its standard NPV-maximizing form pulls high-grade ore forward and defers waste, which improves early cash flows and raises the NPV. The trade-off is that deferred waste accumulates as a “waste debt” that comes due in the middle years when the mine transitions between early low-strip pushbacks and later high-strip pushbacks. The resulting cash flow profile has a valley, sometimes lasting 18–24 months, where free cash flow drops or turns negative as the fleet works through the deferred stripping. How honestly the DFS schedule represents this valley, versus smoothing it with optimistic assumptions about fleet utilization or contractor availability during the critical period, varies widely between projects.

This part of the workflow receives disproportionate attention here because the blast is where the cheapest energy in the entire process chain is applied to rock breakage, and the quality of that breakage propagates forward through every subsequent step.

Production drill rigs (Epiroc PV-271, PV-351; Cat MD6250, MD6310) drill 250 mm or 311 mm holes on patterns ranging from 5×6 m in harder rock to 8×9 m in softer or more fractured material. Hole depth matches the bench height, typically 10–12 m in gold operations, plus 1–2 m of sub-drill to ensure clean floor breakage. Drilling accuracy matters. A hole that deviates from its planned position by 0.5 m at the collar and 1.0 m at the toe changes the effective burden and spacing, altering the energy distribution in that part of the blast. Multi-hole deviation surveys, either downhole or collar-based, are routine on operations that take fragmentation seriously and rare on operations that treat drilling as a commodity activity.

In dry ground, bulk ANFO (ammonium nitrate and fuel oil mixed at approximately 94:6 by weight) is the standard explosive, delivered by mobile manufacturing units (MMUs) that blend the product at the blast hole. In wet ground, where water degrades ANFO performance or prevents it from detonating entirely, emulsion-based products are used: either heavy ANFO (a blend of emulsion and prilled ammonium nitrate) or straight pumped emulsion. The cost difference is material. ANFO costs roughly $0.80–1.20 per kg; emulsion products cost $1.50–2.50 per kg depending on formulation and delivery. On a large bench blast consuming 200–400 tonnes of explosive, the difference between an ANFO blast and an emulsion blast can be $50,000–150,000 per shot. Operations that can selectively load ANFO in dry holes and emulsion in wet holes within the same blast, using dual-product MMUs, optimize this cost. Operations that default to emulsion across the entire pattern because a subset of holes are wet pay more than necessary for the dry holes.

Electronic detonators (Dyno Nobel DigiShot, Orica i-kon III, Austin Powder E*Star) allow each hole to fire at a unique, programmable time with ±1 ms accuracy. The blast designer sets the initiation sequence to control three things: fragmentation, muck pile shape, and environmental effects (vibration and flyrock). Delay timing between holes within a row (typically 15–25 ms) controls how the burden slab detaches from the rock mass. Delay timing between rows (typically 40–65 ms) controls how the broken rock displaces toward the free face. A well-sequenced blast produces a loose, well-fragmented muck pile with a geometry that suits the excavator. A poorly sequenced blast produces a tight, blocky muck pile that slows digging and sends oversized material to the crusher.

The fragmentation distribution from the blast has consequences that extend well past the loading face. Coarse fragmentation means the primary crusher processes more material near its maximum feed size, increasing the probability of bridging and reducing crusher throughput. It means the SAG mill receives coarser feed and has to do more breakage work, drawing higher power per tonne and reducing throughput. At operations like Cadia Valley in New South Wales (originally Newcrest, now Newmont after the 2023 merger), the mine-to-mill program demonstrated quantitatively that investing in blast energy to shift the fragmentation curve finer reduced total comminution energy and increased plant throughput. Explosive energy costs $0.10–0.30 per tonne of rock broken. SAG mill energy costs $2–5 per tonne of rock ground. The energy cost differential between breaking rock with explosives versus breaking rock with a SAG mill is roughly one to two orders of magnitude. Every tonne of size reduction work that can be shifted from the mill to the blast saves money.

Capturing this value requires cost accounting that spans the organizational boundary between the mining department (which pays for drilling and explosives) and the processing department (which pays for energy and grinding media). Under conventional budgeting, each department manages its own cost center. A blast engineer who increases powder factor (kg of explosive per tonne of rock) to improve fragmentation overruns the mining department’s drill-and-blast budget. The processing department, receiving finer and more uniform feed, sees lower SAG mill energy, higher throughput, and potentially better recovery. The net effect is positive at the mine level. The blast engineer’s budget is negative. Changing this requires either a total cost accounting framework that evaluates the drill-to-doré cost chain as a unit, or at minimum a transfer pricing mechanism that credits the mining department for delivering fragmentation quality that benefits the plant.

Cadia published results showing that blast optimization lifted plant throughput by a measurable percentage at an incremental explosive cost far below the value of the additional gold produced. Other operations have replicated these results. The reason it remains uncommon in practice is not that the economics are unclear. The economics are compelling. The reason is that implementing it requires the mining manager and the processing manager to agree on shared KPIs, which is an organizational design problem, not an engineering problem.

There is a second dimension to blast quality that gets less attention than fragmentation: blast damage to the remaining bench floor. An overcharged or poorly timed blast can fracture the rock below the design floor elevation, creating a zone of damaged, dilated rock that allows groundwater infiltration, weakens the bench for subsequent mining, and, in the worst case, contaminates the grade control sample interval with material from below the intended mining horizon. Sub-drill length and explosive energy in the bottom of the hole control this. The standard practice of 1–2 m sub-drill with full column loading occasionally damages the floor in geological conditions (weak bedding planes, clay-filled fractures) where a reduced toe charge or air deck at the bottom of the hole would be more appropriate. At Goldstrike’s Betze-Post pit on the Carlin Trend, where the ore-waste contact in some areas is a relatively sharp lithological boundary, blast damage below the floor can mix barren waste material into what the grade control interprets as the top of the next ore bench, causing dilution that is invisible until the reconciliation data accumulates months later.

Grade control drilling happens after each blast, on the newly exposed bench floor. RC rigs drill sample holes at 5×5 m or 10×10 m spacing, producing cuttings that are logged, split, and assayed. The grade control geologist interprets the results, draws ore-waste polygons, and assigns each polygon a grade and a routing destination: mill, heap leach, low-grade stockpile, or waste dump.

The spacing of grade control drilling determines the resolution of the ore-waste boundary. At 10×10 m, the smallest feature the grade control can reliably resolve is roughly 10–15 m across. Ore pods or waste inclusions smaller than this fall between the holes and are either included in ore (diluting it) or excluded from ore (losing it). Tightening the pattern to 5×5 m improves resolution at the cost of more drilling, more samples, more assay cost, and more time. Which spacing is appropriate depends on the geological style of mineralization. Disseminated gold in a broad alteration halo, like the oxide cap at Yanacocha, can be adequately controlled at 10×10 m. Narrow high-grade structures with sharp boundaries need 5×5 m or tighter to avoid losing ore or diluting feed.

Assay turnaround is the rate-limiting step. An on-site fire assay lab can turn results in 12–18 hours. LeachWELL or similar accelerated cyanide leach assays can reduce this to 4–8 hours with some sacrifice in precision for refractory samples. An off-site commercial lab takes 48–96 hours. During the turnaround period, the excavator either idles or digs using the exploration-stage block model, which was estimated from drill holes spaced 25–50 m apart. The block model is less accurate at the bench scale than the grade control, which is the whole reason the grade control drilling exists.

The mines that compress grade control turnaround to under 24 hours, and that wire the assay results directly into the dispatch system so the excavator operator receives updated dig boundaries before the next shift, recover more gold per tonne mined than mines that operate on 72-hour turnaround. This is not a marginal effect. At Barrick’s (now Nevada Gold Mines’) operations along the Carlin Trend, grade control turnaround time was identified as a key operational variable and systematically reduced during the 2010s by investing in on-site lab capacity and digital integration between the lab, the geology department, and the fleet management system.

Hydraulic excavators (Liebherr R 9800, Cat 6060, Komatsu PC8000) or electric rope shovels (Cat 7495, P&H 4100XPC) load trucks in three to five passes. Truck capacity ranges from 220 tonnes (Cat 793F) to 400 tonnes (Liebherr T 284, BelAZ 75710 at 450t). Fleet management systems (Modular DISPATCH, Wenco, Cat MineStar) assign trucks to shovels and destinations based on grade control polygon tags and real-time fleet position.

Cycle time has four segments: spotting and loading at the shovel, loaded haul (typically uphill at 10–15 km/h on 10% grade), dumping, and empty return (downhill at 30–45 km/h). Fuel consumption during loaded uphill haul ranges from 150 to 280 L/hr depending on truck class, payload, grade, and rolling resistance. Rolling resistance is a function of haul road surface condition. A road that has been graded and watered has a rolling resistance around 2%. A deteriorated road with ruts, loose material, and potholes may reach 4% or higher. That 2% difference costs 15–20% more fuel, 10–15% slower speeds, and higher tire wear. Maintaining haul road quality is one of the highest-return activities on the mine site per dollar spent. During periods of production pressure, the grader and water truck tend to be the first resources pulled from road maintenance to support other activities, which saves nothing and costs measurably in fleet efficiency over the following days and weeks.

Operator variance in cycle time is 8–12% between individuals on the same truck, same route, same conditions. The difference comes from gear selection on grades, approach speed into curves, spotting technique at the shovel (how quickly and accurately the driver positions the truck under the bucket), and dump procedure. Fleet telemetry captures all of this. Translating it into an operator coaching program that lifts the bottom quartile toward median performance generates a fleet capacity increase equivalent to one or two additional trucks on a 40-truck fleet, with no capital spend. At Cat 797F pricing around $5.5 million per unit, that represents $6–11 million in avoided capital. The limitation is organizational: fleet productivity belongs to the mining superintendent’s scorecard, workforce training belongs to HR, and the two functions rarely share a performance target.

Between the pit and the plant, ore sits on a ROM stockpile for hours to weeks. This intermediate step causes more grade loss than most operations quantify.

Ore from blast polygon A, grading 2.8 g/t, is dumped at one point on the stockpile. Ore from polygon B, grading 0.9 g/t, is dumped next to it an hour later. Truck traffic and dozer activity compact and mix the contact zone. Weathering, fines migration, and rainfall further homogenize the surface. When the front-end loader reclaims from the stockpile face to feed the primary crusher, each bucket cuts across both grade zones and delivers a blended product somewhere between 0.9 and 2.8 g/t. The fleet management system, which tracks material by polygon tag, reports that the plant is receiving polygon A material at 2.8 g/t. The plant’s on-line analyzer or shift composite assay shows 1.6 g/t. The 1.2 g/t discrepancy enters the reconciliation as a negative F2 factor and is attributed to “geological uncertainty” or “sampling bias.” In many cases, the geology and the sampling are fine. The problem is physical mixing on the pile.

Operations that have investigated this with GPS-tagged dumping, survey-controlled stockpile sectioning, and dedicated reclaim plans that restrict the loader to a single grade domain per shift have measured and reduced the effect. The cost is modest: some additional survey work, some constraints on loader flexibility, and a fleet management system configured to enforce grade-segregated dump zones. Most operations do not implement it because the stockpile is managed by the mining department as a logistics problem (keep the crusher fed) rather than by the geology department as a grade management problem (preserve grade segregation).

Crushed ore (minus 19 mm or minus 12.5 mm, sometimes agglomerated with 5–10 kg/t Portland cement for clay-bearing ores) is stacked on a lined pad in 8–12 m lifts. Sodium cyanide solution at 200–500 ppm, pH 10–10.5, is dripped through emitters at 8–12 L/hr/m². Gold dissolves and the pregnant solution is collected on the liner and pumped to a recovery plant for carbon adsorption or Merrill-Crowe zinc precipitation.

Column tests during feasibility provide recovery estimates. Field recoveries run 5–15 percentage points below column results because of compaction in lower lifts, preferential solution channeling, and variable ore character that a limited number of column tests cannot fully represent. Tracer studies on operating pads have documented that large volumes of ore receive minimal solution contact even after months of irrigation. Yanacocha, which has operated heap leach pads on oxide, transitional, and clay-bearing ores over decades, has accumulated operational evidence that each ore type requires distinct crush size, agglomeration, and irrigation management, and that the response of a given ore type in the field deviates from its column test prediction in ways that are consistent within ore type and can be calibrated over time, as long as the operation maintains the metallurgical accounting to do so.

Primary crushing (gyratory or jaw), SAG milling, ball milling, and classification produce a slurry at 75–106 microns P80. The slurry enters a CIL or CIP circuit for cyanide leaching and carbon adsorption over 18–24 hours residence time. CIL is used where preg-robbing carbonaceous ore is present (Pueblo Viejo in the Dominican Republic is a prominent example where Barrick designed the plant specifically for preg-robbing ore with high organic carbon content). CIP is used where preg-robbing is not a factor. Loaded carbon is stripped by Zadra or AARL elution, the eluate goes to electrowinning, and the cathode product is smelted to doré, typically 70–90% gold and 5–25% silver, for shipment to a refinery (Rand Refinery in South Africa, Metalor in Switzerland, or similar).

SAG mill throughput is controlled by ore hardness. Bond Work Index, SPI (SAG Power Index), or SMC test parameters (A×b and DWi) quantify hardness by ore domain. A 40% increase in feed hardness, which can happen when the mine transitions from a weathered saprolite zone to fresh granodiorite, can cut SAG throughput by 25–30%. If the mine schedule does not blend hard and soft domains to stabilize feed hardness, the plant oscillates between periods of excess capacity (on soft feed) and bottleneck (on hard feed). Geometallurgical modeling maps hardness, recovery, and reagent demand onto the block model so the mine schedule can manage these variables alongside gold grade.

Carbon attrition inside the CIL/CIP circuit produces fine carbon particles that escape through inter-tank screens into the tailings. These fines carry adsorbed gold with them. At typical attrition rates, the loss is 0.5–1.5% of plant feed gold, amounting to hundreds or thousands of ounces per year on a large operation. Harder carbon products (coconut shell carbon from specific suppliers in the Philippines and Sri Lanka tends to score highest on hardness tests), better screen maintenance, and optimized carbon transfer rates reduce this loss. Some operations track it as a distinct line item in their metallurgical balance. Others let it merge into the overall tailings loss and never quantify it separately.

Dewatering borefields lower the piezometric surface around the pit to maintain stability and dry benches. The water enters the mine circuit for process use, discharge, or re-injection. In arid regions (Nevada, parts of Western Australia), water supply constrains plant throughput. In tropical regions (West Africa, Southeast Asia), managing excess water during wet seasons is the binding constraint.

Tailings go to an engineered storage facility (TSF) after cyanide destruction (SO₂/air or peroxide process). TSF design and governance standards have tightened globally following the adoption of the Global Industry Standard on Tailings Management (GISTM) in 2020.

Waste rock is classified as PAF or NAF by acid-base accounting. PAF material, containing sulfide minerals that generate acid on exposure to air and water, goes to lined or encapsulated storage. The lag between waste placement and the onset of acid drainage, which can span years while carbonate buffering capacity is consumed, creates a monitoring window during which everything looks fine. Kinetic test data from humidity cells and field test pads help predict the timeline, with the caveat that extrapolating a two-year lab test to a century-scale waste dump involves substantial uncertainty.

The pit floods to a lake after closure. Wall rock geochemistry, groundwater quality, and climate determine pit lake water chemistry over timescales of decades to centuries. Closure bonds are sized during permitting and are frequently smaller than the eventual closure cost, particularly where acid drainage develops on a timeline longer than the mine life.

The elapsed time between grade control sample collection and the ore routing decision at the dig face is, among all the operational metrics available on a gold mine, the one that correlates most directly with reconciliation performance. A mine that closes this loop in under 24 hours routes ore and waste based on current assay data. A mine that takes 72 hours routes material based on the exploration-stage block model for three days out of every blast cycle. Both mines report production tonnage to the head office. Their F1, F2, and F3 reconciliation factors tell different stories at year end.

The other high-value integration point is between blast fragmentation and mill performance. The organizations that have captured this value reorganized their cost accounting so the mining and processing departments share a common KPI on total rock-to-gold cost. The organizations that have not captured it keep the departmental budgets separate and observe, without fully explaining, why the plant alternates between good months and difficult months despite stable head grade.

Every other integration point in the workflow, from geometallurgical feed management to slope monitoring feedback into mine planning, from water balance management to tailings chemistry, follows the same pattern: the technical solution is understood; the organizational integration is what determines whether it gets implemented.