Mineral economics and reserve estimation

In short

Mineral economics studies how an exhaustible, fixed-location, grade-variable natural asset is valued, taxed and brought to production; reserve estimation is the quantitative step that converts borehole and sampling data into tonnage and grade, and then into the part of it that can be mined at a profit. Every mining decision — lease bid, plant capacity, pit limit, mine life — rests on these two. In India the statutory frame is the MMDR Act, 1957 (as amended, notably 2015, 2021, 2023 and 2025), the Mineral Conservation and Development Rules, 2017, the Mineral (Evidence of Mineral Contents) Rules, 2015 and UNFC classification maintained by IBM and GSI. It carries heavy weight in GATE (Mining Engineering), university papers on mine economics and mine planning, and in interviews at Coal India, NMDC, Hindustan Zinc, Tata Steel, Vedanta, JSW, Hindalco, GSI, IBM and mining consultancies.

Explained

Why minerals are an economic special case

• Exhaustible and non-renewable: every tonne mined reduces the asset, so a mine is a depleting capital asset, not a perpetual business — hence depletion allowances and finite mine life.
• Fixed in place: value depends on where the deposit is (transport, power, water, port, forest/tribal land), so a deposit and the same mineral elsewhere are not substitutes.
• Grade and tonnage are inseparable: a deposit has no single size — tonnage rises as cut-off grade falls (grade-tonnage curve), so "reserves" are an economic statement, not a geological one.
• Long lead time, high sunk capital, price-taking: exploration to production commonly takes 7-15 years, capital is largely irreversible, and output sells at world or notified prices the producer cannot set.
• Inelastic short-run supply, cyclical prices: capacity cannot respond quickly, so demand shocks move price sharply — the mineral cycle.
• Lasky's law: for many porphyry-type deposits, cumulative tonnage rises roughly geometrically as average grade falls arithmetically, A = K − k·log G.
• Hotelling's rule (theory): in equilibrium the net price (price minus marginal cost) of an exhaustible resource rises at the rate of interest; real mineral prices have not followed it closely because technology and new discoveries keep offsetting depletion.

Demand, supply, price and the value chain

• Demand drivers: steel and construction (iron ore, coal, limestone), power (coal), aluminium (bauxite), batteries and electronics (Li, Co, Ni, graphite, REE), fertiliser (rock phosphate).
• Demand is derived and price-inelastic in the short run: mineral cost is a small part of the finished product, so buyers absorb price rises; substitution and recycling bite only over years.
• Price formation: exchange-traded (LME for Cu, Al, Zn, Pb, Ni), index/benchmark-priced (iron ore indices, coking coal), negotiated long-term contracts, or administered — in India, coal grade-wise notified prices by CIL and the monthly Average Sale Price (ASP) published by IBM for royalty and premium computation.
• Value chain and value addition: exploration → mining → beneficiation → smelting/refining → fabrication; the further downstream, the lower the price volatility and the higher the value per tonne.
• Measures used: reserve-to-production (R/P) ratio = reserves / annual output (years); import dependence; cost curve position (C1 cash cost per tonne decides who survives a price trough).
• India's position: surplus in iron ore, bauxite, limestone, chromite and thermal coal; import-dependent in coking coal, copper concentrate, rock phosphate, nickel and most critical minerals — the basis of the 30-mineral critical mineral list (2023), KABIL's overseas acquisitions and the National Critical Mineral Mission (announced 2025).

Resources and reserves: classification

• Mineral resource — a concentration of material of intrinsic economic interest, with reasonable prospects for eventual economic extraction; mineral reserve — the economically mineable part of a resource after applying mining, metallurgical, economic, marketing, legal, environmental, social and governmental ("modifying") factors.
• CRIRSCO family (JORC, NI 43-101, SAMREC, IMIC): resources are Measured → Indicated → Inferred (decreasing confidence); reserves are Proved (from Measured) and Probable (from Indicated). Inferred resources can never be converted directly to a reserve.
• UNFC — the three-axis code used for statutory reporting in India: E economic viability (E1 economic, E2 potentially economic, E3 intrinsically economic/not evaluated), F feasibility/field-project status (F1 feasibility study or mining report, F2 pre-feasibility study, F3 geological study), G geological knowledge (G1 detailed, G2 general, G3 prospecting, G4 reconnaissance).
• Read as a three-digit code: 111 = Proved Mineral Reserve, 121/122 = Probable Mineral Reserve, 211 = Feasibility Mineral Resource, 221/222 = Prefeasibility Mineral Resource, 331/332/333 = Measured/Indicated/Inferred Mineral Resource, 334 = Reconnaissance Mineral Resource.
• National Mineral Inventory (NMI): IBM compiles India's resources mineral-wise and state-wise in UNFC codes, splitting "reserves" (111-122) from "remaining resources"; lease-holders file annual and monthly returns under MCDR, 2017.
• Mineral (Evidence of Mineral Contents) Rules, 2015: prescribe the minimum exploration level before a block is auctioned — broadly G2 (or better) for a mining lease and G4/G3 for a composite licence (prospecting licence-cum-mining lease). These thresholds have been amended more than once (2021, 2023), so verify the current Rule before quoting a level.
• Public reporting: the Indian Mineral Industry Code (IMIC) gives India a CRIRSCO-style code for public/investor reporting by a Competent Person, alongside the UNFC-based statutory system.

Reserve estimation: conventional methods

Common steps for all methods: 1) compile and validate assay/drill data, 2) build geological boundaries and apply a cut-off, 3) composite samples to equal lengths, 4) treat outliers (top-cut), 5) compute volume, 6) convert to tonnes using bulk density, 7) compute tonnage-weighted average grade, 8) apply dilution and recovery, 9) classify by data density.

• Method of included (polygon) area: each borehole governs a polygon around it (perpendicular bisectors — a Voronoi/Thiessen net); block volume = polygon area × that hole's thickness. Simple; treats each hole as locally representative; poor with irregular spacing.
• Method of triangles: holes joined into triangles; volume = triangle area × mean of the three thicknesses; grade = thickness-weighted mean of the three. Better averaging, but needs similar thicknesses.
• Cross-section (vertical/horizontal section) method: ore area measured on parallel sections; volume between adjacent sections by average end area or the prismoidal/cone (Smith's) rule. Standard for bedded coal, iron ore benches and open-pit design.
• Isopach / contour (iso-value) method: volume from thickness or grade contours and the area within each contour; suited to flat-lying bauxite, laterite and coal seams.
• Blocks: geological block (blocks on natural boundaries, each estimated from its own holes) and mining block (blocks sized to mining units/benches).
• Statistical/level plan methods for narrow veins; Sichel's t estimator for lognormal gold data.
• Bulk density and tonnage factor: tonnage factor is the volume occupied by one tonne (m³/t) = 1/bulk density; use *in-situ* (moist, with voids), not grain, density — one of the biggest sources of systematic error.

Geostatistical estimation

• Why: grade is neither purely random nor perfectly continuous — it is *regionalised*. Geostatistics quantifies spatial correlation and gives an estimate plus an estimation variance, which conventional methods cannot.
• Semi-variogram γ(h) plots half the mean squared difference of sample pairs against separation h; its features are the nugget effect C₀ (micro-variability plus sampling error), the sill C₀+C (total variance) and the range a (distance beyond which samples are uncorrelated). Models: spherical, exponential, Gaussian, linear.
• Anisotropy: range differs with direction (longer along strike than across) — modelled as geometric or zonal anisotropy.
• Ordinary kriging (OK): best linear unbiased estimator — weights λi chosen to minimise estimation variance subject to Σλi = 1. Variants: simple kriging (known mean), indicator kriging (grade-tonnage above cut-offs, highly skewed gold), co-kriging (correlated second variable).
• Nearest neighbour and inverse distance weighting (IDW): quick, non-statistical; IDW weight ∝ 1/dⁿ with n commonly 2. No variance estimate, and smoothing or bullseyes if n is badly chosen.
• Support and selectivity: sample-support estimates are more variable than block-support estimates (volume-variance relation); kriging smooths, so recoverable tonnage above cut-off must be corrected (change-of-support, e.g. affine/indirect lognormal correction, uniform conditioning).
• Conditional simulation: many equally probable grade models used for risk/uncertainty analysis rather than a single best estimate.
• Limitations: needs enough samples for a reliable variogram (rule of thumb >30-50 pairs per lag), assumes stationarity within a domain, and is only as good as the geological domaining and sample quality.

Cut-off grade, dilution and modifying factors

• Cut-off grade (COG): the grade at which revenue just covers the relevant costs. Break-even COG uses operating costs; internal (mill) cut-off compares mill-or-waste for already-mined rock (ignores mining cost); marginal/incremental cut-off ignores sunk and fixed costs.
• Lane's approach: the optimum COG maximises NPV and depends on which of mine, mill or market is the bottleneck, plus the opportunity cost of delaying cash flow — so optimum cut-off generally *declines* over mine life (high grading early).
• Dilution — waste unavoidably mined with ore, lowering grade; ore loss/mining recovery — ore left behind. Both must be applied before calling a resource a reserve.
• Stripping ratio (SR) = waste : ore; break-even SR fixes the economic pit limit; overall SR decides the capital in waste handling.
• Other modifying factors: metallurgical recovery and concentrate payable terms, mining method and minimum mining width, geotechnical slope angles, water, forest and environmental clearance, R&R and land, infrastructure and evacuation, mineral price and exchange-rate assumptions.

Project appraisal, valuation and the Indian fiscal frame

• DCF is the standard: build an annual after-tax free cash flow over mine life and discount it. Criteria: NPV (absolute value added — the primary criterion), IRR, payback period, profitability index. Non-discounted payback ignores both the time value of money and everything after payback.
• Discount rate: typically the weighted average cost of capital plus a risk premium; Indian mining studies commonly use 10-15% real, but it is always project- and client-specific.
• Older mine-valuation formulas still examined: Hoskold (two rates — a speculative rate on the investment and a safe rate for the redemption/sinking fund) and Morkill; both are special cases of discounting an annuity from a wasting asset.
• Capacity and life: Taylor's rule gives a first-pass optimum mine life from reserve tonnage; cross-check against market absorption, infrastructure and lease period (mining leases under the MMDR Act run 50 years, with transfer/extension provisions).
• Risk handling: sensitivity (spider) plots on price, grade, capex, opex and recovery; scenario analysis; Monte Carlo; real-options thinking for the value of deferring or expanding.
• Indian statutory costs on output: royalty at rates in the Second Schedule of the MMDR Act, mostly ad valorem on IBM's ASP (iron ore 15%, limestone and most others specific or ad valorem — check the current Schedule); DMF contribution (as a prescribed percentage of royalty — lower for leases granted by auction, higher for earlier leases); NMET 2% of royalty; the auction premium as a quoted percentage of the value of mineral dispatched; plus GST (slabs of 5%/18%/40% since 22 September 2025, with mineral royalty treated as a supply attracting GST under reverse charge), and income tax under the Income-tax Act, 2025 (in force from 1 April 2026, which renumbered all sections).
• Policy and legal frame: National Mineral Policy, 2019; MMDR Amendment Act, 2015 (auction-only grant, DMF, NMET); 2021 amendment (captive mines allowed to sell part of output, removal of end-use restrictions); 2023 amendment (exploration licence for deep-seated and critical minerals, central auction of 24 critical/strategic minerals); Offshore Areas MDR Amendment Act, 2023; MMDR Amendment Act, 2025 (recovery of critical minerals from tailings/overburden and a widened exploration-and-development trust — rules were still being notified, so verify details); Coal Mines (Special Provisions) Act, 2015 and commercial coal auctions on a revenue-share basis.

Key points

• Resource vs reserve - a resource is geological; a reserve is the economically mineable part after modifying factors.
• Inferred resource - cannot be converted to a reserve, only to Indicated first.
• UNFC code - three digits E-F-G; 111 is a Proved Mineral Reserve, 334 a Reconnaissance Mineral Resource.
• MEMC Rules, 2015 - set the minimum G-level of exploration needed before a block can be auctioned.
• Tonnage = volume × in-situ bulk density; tonnage factor = 1/density in m³/t.
• Average grade must be tonnage- (or thickness-) weighted, never a plain arithmetic mean of assays.
• Polygon, triangle, cross-section, isopach - conventional volumetric methods; simple, no error estimate.
• Variogram - nugget C₀, sill C₀+C, range a; describes how grade correlation decays with distance.
• Kriging - best linear unbiased estimator, minimum estimation variance, Σ weights = 1, gives a confidence measure.
• IDW - weight ∝ 1/dⁿ (n usually 2); fast but no variance and prone to smoothing.
• Grade-tonnage curve - tonnage rises and average grade falls as cut-off grade is lowered.
• Break-even cut-off grade = cost per tonne ore ÷ (recoverable value per unit grade per tonne).
• Lane's rule - the NPV-optimal cut-off falls over mine life and depends on the binding capacity constraint.
• Stripping ratio = waste/ore; break-even SR defines the economic pit limit.
• Dilution lowers grade, ore loss lowers tonnage - both reduce reserve value and both must be stated.
• Taylor's rule - optimum mine life ≈ 6.5 × (reserve in Mt)^0.25 years, a first approximation only.
• NPV is the deciding criterion; IRR and payback are supporting measures.
• Royalty, DMF, NMET and auction premium are output-linked Indian statutory costs that directly raise cut-off grade.

Formulas and examples

Example 1 — triangle method, tonnage and weighted grade. Three bauxite boreholes form a triangle of plan area 12,000 m² with ore thicknesses 6, 8 and 10 m and Al₂O₃ grades 48%, 45% and 42%. In-situ bulk density 2.4 t/m³.

• Mean thickness = (6+8+10)/3 = 8 m → V = 12,000 × 8 = 96,000 m³.
• Tonnage = 96,000 × 2.4 = 230,400 t (tonnage factor 1/2.4 = 0.417 m³/t).
• Thickness-weighted grade = (48×6 + 45×8 + 42×10)/(6+8+10) = (288+360+420)/24 = 1,068/24 = 44.5% Al₂O₃.
• The plain average of 45.0% would overstate grade because the thickest hole is the leanest.

Example 2 — cross-section volume, three rules. Two parallel sections 50 m apart show ore areas A₁ = 2,500 m², A₂ = 3,100 m²; mid-section area A_m = 2,800 m². Iron ore density 3.9 t/m³.

• Average end area: V = 50 × (2,500+3,100)/2 = 50 × 2,800 = 140,000 m³.
• Prismoidal: V = (50/6)(2,500 + 4×2,800 + 3,100) = 8.3333 × 16,800 = 140,000 m³.
• Cone rule: √(2,500×3,100) = √7,750,000 = 2,783.9 → V = (50/3)(2,500+3,100+2,783.9) = 16.667 × 8,383.9 = 139,732 m³ (≈0.2% lower).
• Tonnage (average end area) = 140,000 × 3.9 = 546,000 t.

Example 3 — IDW block grade and the effect of power n. Four holes around a block: grades 1.20, 0.80, 1.50, 1.00% Cu at 50, 80, 120, 150 m.

• Weights 1/d²: 4.000e−4, 1.5625e−4, 0.6944e−4, 0.4444e−4; Σ = 6.7013e−4.
• Σ(g/d²) = 4.800e−4 + 1.250e−4 + 1.0417e−4 + 0.4444e−4 = 7.5361e−4.
• g* = 7.5361/6.7013 = 1.12% Cu (with n = 1 the answer is 1.12%; with n = 0 it is the plain mean 1.125% — and as n rises further the estimate is pulled towards the nearest hole's 1.20% Cu).

Example 4 — cut-off grade and pit limit. Copper ore: mining + processing + G&A = Rs 1,600 per tonne of ore; Cu price Rs 750/kg; mill recovery 85%; waste mining Rs 120/t.

• Recoverable value of 1% Cu in one tonne = 10 kg × 750 × 0.85 = Rs 6,375.
• Break-even cut-off = 1,600/6,375 = 0.251% Cu.
• Block at 0.60% Cu: value = 0.60 × 6,375 = Rs 3,825/t; margin over ore cost = 3,825 − 1,600 = Rs 2,225/t.
• Break-even stripping ratio = 2,225/120 = 18.5 : 1 — mining this block is justified while the waste needed to expose it stays below 18.5 t per tonne of ore.
• If 8% dilution with barren waste occurs, diluted grade = (1×0.60 + 0.08×0)/1.08 = 0.556% Cu, and the break-even SR falls to (0.556×6,375 − 1,600)/120 = 16.2 : 1.

Example 5 — DCF appraisal and mine life. Reserve 25 Mt; capex Rs 500 crore at t = 0; net annual cash flow Rs 120 crore for 8 years; discount rate 12%.

• Taylor's rule life = 6.5 × 25^0.25 = 6.5 × 2.236 = 14.5 years, i.e. about 1.72 Mt/y — the 8-year, higher-rate plan is a deliberate fast-track, so its reserve use must be checked.
• Annuity factor at 12%, 8 yr = [1 − 1.12⁻⁸]/0.12; 1.12⁸ = 2.4760 → 1/2.4760 = 0.4039 → (1 − 0.4039)/0.12 = 4.9676.
• PV of inflows = 120 × 4.9676 = Rs 596.1 crore → NPV = Rs 96.1 crore (positive, accept).
• Payback = 500/120 = 4.17 years (undiscounted).
• IRR: factor needed = 500/120 = 4.1667. At 16% factor = 4.344; at 18% = 4.077 → interpolating, IRR ≈ 17.3%, comfortably above the 12% hurdle.

Example 6 — statutory costs per tonne. Iron ore lumps, ASP Rs 4,500/t, royalty 15% ad valorem, pre-auction lease (DMF at 30% of royalty), NMET 2% of royalty.

• Royalty = 0.15 × 4,500 = Rs 675/t; DMF = 0.30 × 675 = Rs 202.5/t; NMET = 0.02 × 675 = Rs 13.5/t; total statutory levy = Rs 891/t, i.e. 19.8% of ASP — which is why levies must sit in the cut-off grade calculation and not be treated as an afterthought.

Where it is used

• Resource geologists and mine planners at NMDC, Coal India subsidiaries, Hindustan Zinc, Tata Steel, Vedanta, JSW, Hindalco — block models in Datamine, Surpac, Vulcan, Minex or Micromine feeding pit optimisation (Lerchs-Grossmann/Whittle) and life-of-mine schedules.
• Statutory reporting and regulation: UNFC-coded mining plans and MCDR returns to IBM; the National Mineral Inventory; GSI's and NMET-funded exploration reports classified to G-levels.
• Mineral block auctions: state governments and bidders use G2/G3 reports to set reserve tonnage, quote premium percentages and model royalty, DMF and NMET; consultants (MECON, MECL, SRK, SGS) prepare the technical reports.
• Project finance and due diligence: banks, SBI Capital, PFC-type lenders and the IBBI-route resolution of stressed mining assets rely on DCF valuations and Competent Person reports under CRIRSCO/IMIC-style codes.
• Cement, steel and aluminium captive mines: limestone, bauxite and coal reserve estimates fix plant capacity and the 50-year lease plan; grade blending schedules come straight from the block model.
• Critical minerals programme: reserve and economic assessment of lithium (Reasi, J&K), graphite, REE and recovery from tailings under the 2023 and 2025 MMDR amendments and the National Critical Mineral Mission.

Interview questions

• Q: What is the difference between a mineral resource and a mineral reserve? A: A resource is a geologically defined concentration with reasonable prospects of eventual economic extraction; a reserve is the part of a Measured or Indicated resource that is proved economically mineable after mining, metallurgical, legal, environmental and economic modifying factors.
• Q: Can an Inferred resource be reported as a reserve? A: No. Confidence is too low; it must first be upgraded to Indicated by further drilling, and only then converted to a Probable reserve.
• Q: What does UNFC code 111 mean? A: E1-F1-G1 — economically viable, supported by a feasibility or mining report, with detailed geological exploration: a Proved Mineral Reserve.
• Q: Why does lowering the cut-off grade increase tonnage but reduce average grade? A: Lower-grade material previously classed as waste enters the reserve; it adds tonnes but drags the weighted mean grade down — the grade-tonnage curve.
• Q: Why is kriging preferred over inverse distance weighting? A: Kriging uses the measured spatial correlation (variogram) to produce the minimum-variance unbiased estimate and reports an estimation variance; IDW uses an arbitrary distance power and gives no measure of confidence.
• Q: What do the nugget, sill and range of a variogram represent? A: The nugget is micro-scale variability plus sampling error at zero separation, the sill is the total variance, and the range is the distance beyond which samples are no longer spatially correlated.
• Q: How do you decide the pit limit economically? A: By the break-even stripping ratio — strip waste while the margin from the ore it exposes exceeds the cost of removing that waste; formally through pit optimisation on a block model with revenue and cost per block.
• Q: Why does the optimum cut-off grade fall over the life of a mine? A: Early high grading maximises discounted cash flow (Lane's theory); as remaining reserves and the opportunity cost of delay shrink, lower grades become worth processing.
• Q: Which Indian levies are linked to the value of mineral dispatched? A: Royalty under the Second Schedule of the MMDR Act (mostly ad valorem on IBM's Average Sale Price), the auction premium, DMF as a prescribed percentage of royalty, and NMET at 2% of royalty.
• Q: What is Taylor's rule used for and what is its limitation? A: A first-pass estimate of optimum mine life and production rate from reserve tonnage; it is an empirical fit ignoring grade, market absorption, capital constraints and lease period, so it is a starting point for trade-off studies only.
• Q: Why is in-situ bulk density so critical? A: Tonnage scales directly with it; using grain or dry density instead of in-situ moist density, or one global value across varying lithologies, biases the whole reserve by several per cent.
• Q: NPV says accept, IRR says marginal — which do you follow? A: NPV, because it measures absolute value added at the firm's cost of capital; IRR can mislead with non-conventional cash flows, multiple sign changes or projects of different scale and life.

Common mistakes

• Reporting a resource as a reserve without applying dilution, ore loss, recovery, costs and clearances — the single most common examination and audit error.
• Taking a plain arithmetic mean of assays instead of a tonnage- or thickness-weighted mean.
• Confusing grain/dry density with in-situ bulk density, or forgetting that tonnage factor is the reciprocal of density.
• Mixing units — m² with ha, m³ with cubic metres of loose (swelled) rock, Rs/t with Rs/kg of metal, percentage with ppm or g/t.
• Forgetting that 1% of a tonne is 10 kg (and 1 g/t = 1 ppm) in cut-off grade calculations.
• Quoting a cut-off grade without saying which costs it covers (break-even, internal/mill or marginal) — the three numbers differ widely.
• Letting high-grade outliers through without top-cutting, or kriging across a geological boundary instead of domaining first.
• Using payback period alone to approve a project, and ignoring royalty, DMF, NMET and premium in the cash flow.

Related topics

Mining geology and exploration · Mineral processing · Surface mining and open-pit design · Mine planning and production scheduling · Mine legislation and safety (MMDR Act, MCDR, Mines Act) · Mineral policy and critical minerals.