The mine and the quarry plan
The quarry is the only part of a cement plant that cannot be replaced, upgraded or relocated. It is also, since 2015, the fastest-rising cost line in the business — and the one least likely to appear in a plant improvement plan.
On this page · 16 sections
- 01The one idea
- 02The parts — four decisions, all taken before you arrive
- 03First principles — why a plant cannot burn the rock it owns
- 04What must be true
- 05The quantitative anatomy — what a tonne of limestone costs before you dig it
- 06What goes wrong here
- 07Scenarios from the field
- 08The numbers that matter
- 09Build the model — landed limestone cost, and what a grade decline is worth
- 10Upgrade paths — what can actually be done to a hole in the ground
- 11The frontier, and who is running it
- 12What’s changing now
- 13Check yourself
- 14Mini case — the plant that was benchmarked on its birthday
- 15Go deeper
- 16Carry forward
§ 01 The one idea
Every other lesson in this course is about a decision that can be revisited. A separator can be replaced, a fuel mix changed, a clinker factor moved. The deposit cannot. Its grade, its overburden, its distance from the kiln and its remaining life were fixed before the plant existed, and they set the floor under everything downstream.
What has changed is the price. Before January 2015, limestone leases were granted administratively and a plant paid royalty and little else. Since then every new lease is auctioned, and the winning bid is a percentage of the value of everything the mine will ever dispatch, for the life of the lease. In FY25 the pan-India average winning bid was 63%, and the range across regions ran from 14% to 156%.
“When was this lease granted?” Before 12 January 2015 or after. That single date determines the DMF rate, whether an auction premium is payable at all, and therefore whether the plant’s statutory limestone cost is around ₹106 a tonne or around ₹241. Two plants running identical equipment on identical rock can be more than twice apart on this line, permanently, for reasons no operating decision can touch.
§ 02 The parts — four decisions, all taken before you arrive
| Decision | What it fixes | Can it be revisited? |
|---|---|---|
| Where the deposit is | Lead distance to the kiln, and the plant’s position inside its own freight radius (F5 §03). | No. |
| What grade it is | The raw mix, the additives you must buy, and the burnability the kiln will fight for forty years (A3, C8). | Only by blending or buying — both of which are the cost. |
| How much overburden sits on it | Mining cost per tonne, and it usually worsens with depth. The Indian average is about 0.95 t of overburden per tonne of limestone. | No, and the mine plan already knows how it changes. |
| When the lease was granted | DMF at 30% or 10% of royalty, and whether an auction premium of 14–156% of dispatch value applies for the life of the lease. | No. This is the largest single number on the page and it is a date. |
§ 03 First principles — why a plant cannot burn the rock it owns
Limestone is calcium carbonate diluted by whatever else was deposited with it. Pure calcite is 56.03% CaO — that is the arithmetic from F3 §03 run the other way, 56.08 divided by 100.09. Every point of CaO below that is silica, magnesia, alumina, iron or clay.
Now there are two thresholds, and the gap between them is the whole problem.
MgO forms periclase, which hydrates slowly and expands after the concrete has set. It is the reason for a hard ceiling rather than a preference, and Indian deposits are magnesia-rich: NCB states that most deposits now available for cement are “magnesia rich (MgO >5%) and silica rich (SiO₂ >15%) low-grade materials.”
Alkalis (Na₂O and K₂O, together R₂O) do not stay put. They volatilise in the burning zone, condense in the preheater, and build the coating rings that stop kilns — the volatile cycles of C10. A high-alkali deposit is a reliability problem disguised as a chemistry number.
Silica is not straightforwardly bad — the mix needs it. What matters is whether it is finely disseminated or present as coarse quartz grains, because coarse quartz is what makes a high-LSF mix hard to burn (A3, C8). A grade table cannot tell you this; only a petrographic study can.
So the practical consequence: a reported reserve figure is not an inventory of usable rock. India’s 227.6 billion tonnes of limestone resources include only 19.0 billion tonnes of reserves — 8% — and about 30% of even those reserves sit under forest and other regulated land unavailable for cement. When a client tells you they have forty years of limestone, the follow-up is: at what grade, at what strip ratio, and with what approvals.
§ 04 What must be true
A quarry is the one asset a plant can never replace, so the conditions it has to meet are worth stating as conditions rather than as aspirations. These six do not change with the deposit, the country or the decade. A quarry that meets them supplies a stable plant for forty years. A quarry that fails one of them exports the failure downstream, usually to the kiln, where it is diagnosed as something else.
| What must be true | Why | Target | How you verify it |
|---|---|---|---|
| 1. The chemistry delivered is the chemistry the mix was designed for | Every corrective in A3 is sized against an assumed limestone analysis. A deposit that drifts away from it eats the feeder range first and the fuel bill second. | CaO 44–52% MgO <3.5% R₂O <0.6% |
Compare the last twelve months of delivered analysis against the analysis the corrective system was designed on. Most plants have never done this comparison, and the design figure is usually in a commissioning file nobody opens. |
| 2. Variability is controlled at the face, not corrected downstream | Pre-blending removes variation at a square root (A2 §03), so halving the incoming spread is worth more than any downstream stage can buy. | — | Ask whether faces are sequenced on chemistry or on haul-road convenience. If the mine reports on tonnes and the lab reports on chemistry, nobody owns the line between them — and that gap is where A5’s slow drift comes from. |
| 3. Reserve life is reserve, not resource | Resources include material below cut-off grade, under villages and inside forest blocks. Nationally the resource-to-reserve ratio is twelve to one. | reserves = 8.4% of resources |
Ask for the mine plan’s reserve figure with its cut-off grade, its stripping assumption and its lease expiry — three numbers, not one. “Forty years of limestone” is not an answer until it has all three. |
| 4. The strip ratio in the plan matches the strip ratio being worked | Overburden is pure cost and it usually worsens with depth. The Indian average is about 0.95 t of overburden per tonne of limestone, but it is deposit-specific. | plan vs actual within 10% |
Compare planned against actual for the last three years. A widening gap means the mine is taking the easy rock first, which brings the hard rock forward without anyone deciding to. |
| 5. The statutory position is known and separated from the operating position | Royalty, DMF, NMEDT and any auction premium are legislated, not managed. Mixing them into a cost benchmark turns vintage into apparent performance. | ₹106/t legacy ₹241/t auctioned |
The lease grant date, the DMF rate and the winning premium percentage. All three are matters of record. Until the line is split, every comparison against another plant is meaningless. |
| 6. The haul is short, and stays short | Haulage is diesel, tyres, road maintenance and cycle time, and it lengthens every year as the face retreats. | — | Current one-way haul distance against the distance at commissioning, and the distance the mine plan implies in ten years. In-pit crushing and conveying becomes arguable when that third number is large, not when the first one is. |
Row 1. A plant will tell you its limestone grade as a single average figure. Ask for the standard deviation and the twelve-month trend, and ask what analysis the corrective feeders were sized on. Grade decline is a slow variable that nobody is accountable for, and by the time it is visible in the kiln it has been running for years — which is exactly the A3 §14 case, arriving from upstream instead of from a quality decision.
§ 05 The quantitative anatomy — what a tonne of limestone costs before you dig it
| Element | Rate | Basis and source |
|---|---|---|
| Royalty | ₹80/t ₹90/t LD grade | Second Schedule, MMDR Act 1957. Effective 1 September 2014 — not revised since. Specific rate, not ad valorem, and uniform across states. |
| DMF · District Mineral Foundation | 30% or 10% of royalty | DMF Rules 2015. 30% if the lease was granted before 12 January 2015, 10% if on or after. The statutory test is the grant date, not the word “auction” — a legacy lease that was later extended stays at 30%. |
| NMEDT · exploration trust | 3% of royalty | Raised from 2% and renamed by the MMDR Amendment Act 2025, effective 1 September 2025. Anyone still modelling 2% is a year out of date. |
| Auction premium | 14–156% | Of the value of mineral dispatched, monthly, benchmarked to the IBM average sale price — a price a captive plant never actually transacts at. FY25 pan-India average 63%. Reserve price typically 5%; bids are uncapped. |
| State mineral-bearing land tax | ₹160/t | Tamil Nadu only, Act 9 of 2025, in force 20 February 2025. Karnataka’s equivalent Bill proposes ₹20–25/t. Both now sit under the MMDR Amendment Act 2026 — see §12. |
| GST on royalty | 18% | Reverse charge, SAC 997337. Creditable input tax for a cement plant, so it is a working-capital item, not a cost. It is routinely mis-stacked as a cost — do not do that. |
| Mine closure assurance | ₹3 lakh/ha | Rule 27, MCDR 2017, Category A. A bank guarantee, not an expense — a balance-sheet item. |
| Forest NPV, if applicable | ₹6.7–15.9 lakh/ha | Pre-2022 band of ₹4.38–10.43 lakh raised 1.53× from 6 January 2022. Compensatory afforestation is additional. |
India’s ex-mine value of limestone, from the Ministry of Mines monthly series, is about ₹239 a tonne (April–December FY26: 346.28 Mt worth ₹8,283 crore). A specific royalty of ₹80 against that is an effective ad valorem rate of about 33.5%.
That is very high by international standards, and it is why the specific-rate structure matters so much: royalty does not fall when limestone value falls. A Ministry of Mines consultation note of February 2025 proposed leaving auctioned mines at ₹80/₹90 and raising non-auctioned mines by 50%, to ₹120/₹135. No notification implementing this had been issued as at August 2026 — but it is a live exposure for any plant on a legacy lease, and it is the kind of thing that appears in a budget without warning.
The resource base, and how little of it is usable
| Measure | Quantity | Note |
|---|---|---|
| Total limestone resources | 227.6 bn t | UNFC assessment as on 1 April 2020 — still the latest vintage, six years old. |
| Reserves | 19.0 bn t | 8.4% of resources. The other 92% is “remaining resources”, which is not the same as rock you can mine. |
| Cement (Portland) grade | 69% of resources | 157 bn t. Other grades 13%, unclassified 12%, blast furnace 6%. |
| Under forest and regulated land | ~30% of reserves | Unavailable for cement manufacture (CRISIL, July 2025). |
| Proved mineable cement-grade reserve | 8,558 Mt | CMA, December 2021, citing a 2015 assessment. Against cumulative demand to 2030 of ~4,664 Mt at 7% growth. |
| Stripping ratio, national average | 0.95 t OB / t | CMA, February 2022. Deposit-specific in practice — observed mine-level figures range roughly 0.2:1 to 2:1. |
§ 06 What goes wrong here
| The failure | Where it surfaces |
|---|---|
| Grade decline is managed as a mining schedule, not as a chemistry problem. The mine plan tracks tonnes and strip ratio; nobody tracks what the falling CaO is doing to the raw mix. | Two machines downstream, as kcal/kg. A declining LSF has to be corrected with sweetener limestone or a higher-lime mix, and the kiln pays for it. The quarry and the kiln have separate managers, separate KPIs and separate budgets, and nobody owns the line between them. |
| The auction premium is modelled as a one-off. | It is per tonne dispatched, every month, for the life of the lease, indexed to a price the plant does not set. A 63% premium on a ₹239 IBM price is ₹151 a tonne, forever. Any valuation of an auctioned block that treats the bid as a capital cost is wrong by an order of magnitude. |
| Reserve life quoted from resources rather than reserves. | Nationally the ratio is 12:1. At plant level the same optimism appears as “forty years of limestone” that turns out to include material below cut-off grade, under a village, or inside a forest block. |
| Assuming a won block is a working mine. | 101 auctioned mineral blocks have been operationalised since 2015, against roughly 684 auctioned — about one in seven, and only 29 of the 101 are limestone. A preferred-bidder announcement is the start of a multi-year approvals process, not a capacity addition. |
| Stacking GST on royalty as a cost. | It is creditable. Including it inflates the limestone line by 18% of royalty and makes every comparison wrong. This is a real error that appears in circulated models. |
| Comparing limestone cost across plants without asking the lease date. | Two plants on identical rock can be ₹106 and ₹241 a tonne apart on statutory payments alone. That is not performance; it is vintage. Benchmarking it as performance is the classic error in this lesson. |
§ 07 Scenarios from the field
Dalmia substituted LD slag — a steelmaking by-product carrying both lime and iron — for part of the limestone in the raw mix. The result on Line 2: 10 kcal/kg of clinker, ₹11 per tonne of clinker, and NOx down 50–70 mg/Nm³. Stated ROI: 4.67 months of kiln running.
Read what happened there. The saving did not appear in the mining cost line, where the decision was taken. It appeared as heat rate, because slag arrives already decarbonated — the CO₂ was driven off in the blast furnace, at someone else’s expense. This is why the quarry and the pyro line cannot be advised on separately, and why a raw-materials manager optimising ₹/tonne of raw mix will systematically undervalue exactly this class of move.
Two limestone blocks in the same district of Chhattisgarh, auctioned in the same era: Kesla-1 went to Century Cement at 10.15%. Kesla-2 went to Dalmia at 96.15% — nine and a half times the premium. Guma, also Chhattisgarh, went at 138.25%. Gojoli in Maharashtra went at 5.2%.
Thestudy that catalogued these noted the obvious problem: Kesla-2 has about 3.3 times the limestone of Kesla-1, which does not explain a 9.5× difference in premium. Greenfield blocks averaged 47%, brownfield 26%.
The consultant’s read: these bids are strategic, not valuation-driven. A producer bidding to deny a competitor a foothold inside its freight radius (F5 §03) will bid past any discounted-cash-flow number, and the cost sits on the plant’s limestone line for fifty years. When a client’s limestone cost looks indefensible, check whether the block was bought for the rock or for the position.
An earlier version of this lesson’s outline said “only 36 of 212 auctioned mineral blocks nationally are operational”. That is wrong. The two numbers come from the same speech but are not a numerator and a denominator: 212 is the number of blocks auctioned in FY 2025-26 alone, a single-year record; 36 is the number operationalised during FY 2025-26, drawn from earlier auction vintages.
The correct cumulative figure is 101 blocks operationalised since 2015 (of which 29 limestone) against roughly 684 auctioned — about 15%, or one block in seven. The direction of the original claim was right and the arithmetic was nonsense.
And the correction had to be corrected. The first attempt at this fix put the cumulative figure at 484 — a straightforward addition error on the Ministry of Mines fiscal-year bands, which are 108 blocks in FY16–21, 364 in FY22–25 and 212 in FY26. Those sum to 684. The auditor caught it before publication. A PIB year-end document of December 2025 gives 585 auctioned cumulatively; that is consistent with 684, since it counts only nine months of FY26. Cite the May 2026 PIB release for the operational count and the fiscal-year bands for the auction count.
§ 08 The numbers that matter
| Metric | India | Source, edition and period | Your client |
|---|---|---|---|
| Royalty, limestone | ₹80 / ₹90 | MMDR Act Second Schedule, effective 1 Sep 2014. LD grade (SiO₂ <1.5%) is ₹90. | — |
| DMF | 30% or 10% | DMF Rules 2015. Split on lease grant date, 12 Jan 2015. | — |
| NMEDT | 3% | MMDR Amendment Act 2025, from 1 Sep 2025. Was 2%. | — |
| Auction premium, FY25 | 63% pan-India | CRISIL, Jul 2025. By region: East 98%, Central 72%, North 55%, South 44%, West 22%. | — |
| IBM average sale price | ~₹239/t | Derived from Ministry of Mines, Apr–Dec FY26: 346.28 Mt / ₹8,283 cr. FY25 equivalent ₹234. | — |
| Statutory stack, legacy lease | ₹106/t | Derived. 80 + 24 + 2.40. Before any mining cost. | — |
| Statutory stack, auctioned | ₹241/t | Derived at the FY25 average premium. Before any mining cost. | — |
| Limestone share of raw-material cost | ~85% | CRISIL, Jul 2025. Raw material is 17–20% of total cost (F3 §05). | — |
| Stripping ratio | 0.95 : 1 | CMA, Feb 2022, national average. Highly deposit-specific. | — |
| Cut-off, regulatory | 34% CaO 5% MgO max | IBM threshold notification, 25 Apr 2018. | — |
| Cement specification | 44–52% CaO | CMA via IBM Yearbook 2022. MgO <3.5%, R₂O <0.6%, TiO₂ <0.5%. | — |
| Lease term | 50 years | s.8A MMDR, 2015 amendment. No regulatory minimum reserve life for a new plant could be sourced. | — |
What is not published, and should not be invented. No credible source gives a cost per tonne of limestone mining in India — not IBM, not the CMA, not a rating agency, not a company disclosure. No sourced cost comparison exists between surface miners and drill-and-blast for Indian cement limestone, and no source publishes the share of Indian cement limestone mined by surface miner. If a proposal quotes you any of those three, ask where it came from.
§ 09 Build the model — landed limestone cost, and what a grade decline is worth
When Tamil Nadu introduced its ₹160/t mineral-bearing land tax in early 2025, brokers estimated the EBITDA impact at ₹80/t of cement for Ramco (52% of its clinker capacity in TN) and ₹34/t for Dalmia (23%).
Build it from first principles: 1.5 t of limestone per tonne of clinker × a 0.68 clinker factor is 1.02 t of limestone per tonne of cement. So ₹160 × 1.02 = ₹163/t of cement on TN-sourced volume. Ramco at 52%: ₹85. Dalmia at 23%: ₹37. Against the published ₹80 and ₹34.
Two lines of arithmetic reproduce a brokerage estimate to within 6%. That is the standard to hold yourself to on this line: if your model cannot reproduce a published number you can check, it will not survive one you cannot.
The mining-cost field is a placeholder, not a benchmark. No credible source publishes a cost per tonne of limestone mining in India — not IBM, not the CMA, not a rating agency, not a company. The ₹180 default is there to make the model run. Replace it with the client’s own figure from the cost sheet before quoting anything.
And notice what the defaults say. Statutory payments on an auctioned block at the FY25 average premium come to about ₹241 a tonne — against an IBM average sale price for limestone of about ₹239 a tonne. The statutory charges alone, before a single rupee of mining cost, now roughly equal the price at which limestone changes hands. A producer bidding at the FY25 average is committing to a fifty-year cost structure in which the levies exceed the merchant value of the rock. That is not a modelling artefact; it is what a 63% premium on a specific-royalty regime does, and it is the strongest argument in this lesson for treating bid discipline as a strategic control rather than a procurement one.
How the grade-decline term is built, and its limits. A drop in CaO must be made up somewhere. The model corrects it two ways: by raising the limestone-to-clinker ratio to hold lime input constant, and by adding the extra decarbonation heat that the additional carbonate requires. It uses the F3 §03 chemistry (44% of limestone mass leaves as CO₂, roughly 1,650–1,800 kcal per kg of CaCO₂ decomposed at typical kiln conditions) and the A7 conversion of kcal/kg into ₹/t. It is a first-order estimate and deliberately excludes the second-order effects — harder burning, more additive purchases, shorter refractory life — which in practice are often larger than the term shown. Use it to establish that grade decline is a material number, not to size a project.
§ 10 Upgrade paths — what can actually be done to a hole in the ground
| Tier | Intervention | What it preserves, and the honest assessment |
|---|---|---|
| 0 | Chemistry-led mine scheduling | Sequence the faces to hold the raw mix steady rather than to hold tonnes steady. Costs nothing but a change of KPI, and it is the intervention with the best ratio of value to capital in this lesson. It fails for organisational reasons: the mine reports on tonnes. |
| 1 | Sweetener and corrective additive sourcing | Buying high-grade limestone, laterite or bauxite to correct a declining mix. Preserves everything; it is pure operating cost. The question is always whether it is cheaper than the fuel it saves — which is A3 §09. |
| 1 | Industrial by-products in the raw mix | LD slag, red mud, fly ash, marble slurry. The best-evidenced move in this lesson — Dalmia’s 10 kcal/kg and ₹11/t of clinker at a 4.67-month payback. Preserves the quarry, the mill and the kiln. Constrained by supply and by transport distance, not by technology. |
| 2 | Surface miner replacing drill-and-blast | Removes primary crushing and the blasting permit problem, and gives finer, more selective extraction. No sourced Indian cement cost comparison exists — make the case on selectivity and community relations, and demand the vendor’s own site references. |
| 3 | Beneficiation of marginal-grade limestone | The classic “we already own it” upgrade: the reserve is on the books, it is simply below cut-off. Treat any proposal here as first-of-a-kind in India. NCB has studied dry beneficiation for decades and offers it as a laboratory service; no operating commercial Indian cement beneficiation plant could be verified for 2025–26, and no published cost per tonne exists. Price the risk accordingly and insist on a reference plant. |
| 4 | A new lease | Auction premium of 14–156% for fifty years, and roughly a one-in-five chance of the block reaching operation. This is a corporate development decision, not a plant improvement. |
§ 11 The frontier, and who is running it
Quarrying is the part of the plant where the gap between what is claimed and what is running is widest. Almost everything published as cement quarry automation turns out, when traced, to be an aggregates operation or a metals mine. That is a useful fact in itself: if a vendor tells you autonomous haulage is proven in cement, ask which cement plant.
| Status | Technology | Site, company, country | The number |
|---|---|---|---|
| OPERATING | Autonomous loading, underground | Rüdersdorf — Cemex, Germany, with sensmore. One Aramine L140B LHD. | Autonomous load-haul-dump cycles described as part of daily production after about three years of development. The only genuine autonomous machine at a cement limestone operation anywhere — and it is underground, not in a surface pit. Reported June 2026. |
| OPERATING | Battery-electric mine fleet | Małogoszcz, Kujawy and Radkowice — Holcim, Poland. SANY SKT105E trucks, LiuGong electric loader. | Eleven electric machines across three sites with 4.4 MW of charging, including a 1,200 kW Megawatt Charging System satellite at Małogoszcz. Second phase completed August 2026; about 20% of the Polish mining fleet electric by end-2026. No diesel-saving figure disclosed. Cement is genuinely ahead here. |
| OPERATING | Cross-belt control of the stockpile | Al-Rashadiya — Jordan Cement, Jordan. Two cross-belt analysers on the pre-blend feed, updating every minute. | Kiln Feed Uniformity Index brought inside the group standard, raw material cost down 0.08 JD/t, and the cement-to-clinker ratio raised from 1.14 to 1.16. The best-quantified cross-belt case in the public domain — and note the benefit landed on the clinker factor, not on the mine. |
| OPERATING | Analyser on the quarry crusher | Suwannee American Cement (Votorantim), Branford, Florida, USA. Sodern CNA on the raw mill feed from 2003, second unit downstream of the portable quarry crusher from 2007. | ~98% availability sustained for nearly twenty years. Reduced waste material and lower LSF variability. The twenty-year availability figure is the useful one: this is not experimental equipment. |
| OPERATING | Drone survey and digital mine planning | Rüdersdorf — Cemex, Germany, with Propeller. | Stockpile inventory data collection cut from about four hours to about one, results within 24 hours, and an inventory discrepancy found and corrected. Open-source software was tried first and abandoned on scalability. |
| AGGREGATES | Autonomous haulage at scale | Lake Bridgeport — Heidelberg Materials, Texas, USA, with Pronto. Mixed Cat and Komatsu 70-tonne class. | Over 2 Mt of limestone hauled in under eight months. Real, operating, impressive — and an aggregates quarry, not a cement plant quarry. All six Heidelberg autonomy sites are aggregates. The target is about 30 autonomous vehicles in 2026 and over 100 by end-2028, so this will reach cement; it has not yet. |
| AGGREGATES | Autonomous drilling | Luck Stone — Virginia and the Carolinas, USA. Epiroc SmartROC D65 MkII. | Complete drill patterns with no operator in the cab, GMG Level 4 full autonomy, September 2025. Billed as the first fully autonomous surface drill in the quarry market. No cement quarry has published a case. |
| STALLED | Autonomous electric haulers at a cement quarry | Gabenchopf, Siggenthal — Holcim, Switzerland, with Volvo Autonomous Solutions. | Announced November 2021 as the first autonomous haulage in the cement sector. Every traceable source still dates to November 2021. No commissioning, no tonnage, no fleet number, in five years. Treat as a trial that did not progress. |
| LETTER OF INTENT | Electric and autonomous fleet | SANY and Holcim — no sites named. | 100 electrified machines and 20 autonomous mining trucks, CHF 100 m over five years, March 2026. No sites, no models, no dates. It is an LOI, and it is already being quoted as a fleet. |
§10 treats beneficiation of marginal-grade limestone as a first-of-a-kind proposition in India. The wider search makes it stronger than that: no operating commercial limestone beneficiation plant for cement could be verified anywhere in the world — not in India, not elsewhere. What exists is laboratory and pilot flotation work, one old Indian announcement, and vendor pages.
That matters because the ten percentage points of CaO between the regulatory threshold and the cement specification (§03) sit on the books of every plant with a declining deposit, and beneficiation is the obvious answer. If a vendor proposes it, ask for the reference plant, anywhere on earth, with its capacity and its cost per tonne. As of October 2026 that question has no answer.
§ 12 What’s changing now
Current as of August 2026 · refresh every six monthsA new section 9D bars state governments from imposing any tax, cess or levy on mineral rights or mineral-bearing lands except in accordance with conditions the Centre has yet to prescribe. Critically, it provides that any such levy not already deposited with or recovered by a state before commencement is deemed invalid at all material times, notwithstanding any judgment or decree — and that amounts already paid are not refundable.
This reverses the tail of the Supreme Court’s July 2024 Mineral Area Development Authority ruling, which had held royalty is not a tax, permitted state levies, and allowed demands back to 1 April 2005 payable in twelve annual instalments from 1 April 2026 — with interest and penalties waived for periods before 25 July 2024. The Act landed weeks before that instalment window would have opened.
What it means: this is a finance conversation, not a policy one. Any Indian plant that provisioned for state mineral cess back to 2005 has a live balance-sheet question this quarter. And note the asymmetry the legislative research flags: companies that paid bear the cost permanently; those that litigated or withheld are released. Congress-ruled states announced a joint Supreme Court challenge on 20 August 2026, so this is not settled.A Ministry of Mines consultation note proposed holding auctioned mines at ₹80/₹90 per tonne while raising non-auctioned mines to ₹120/₹135. No notification had been issued as at August 2026.
What it means: every plant on a legacy lease — which is every plant whose limestone line currently looks cheap — carries an unpriced ₹40–45/t exposure. It belongs in a sensitivity, not in a base case, and it should be named explicitly rather than left out.FY26 saw the highest single-year auction count since the regime began in 2015. Cumulatively about 684 blocks have been auctioned and 101 operationalised, of which 29 are limestone. The FY25 pan-India limestone premium rose to 63% from 50%.
What it means: reserve accumulation has become a competitive activity in its own right — Ambuja won 587 Mt in FY24 alone and now holds roughly 8 bn tonnes against UltraTech’s 10 bn. But auctioned tonnes are not capacity, and the four-fifths that never open are the reason F5’s regional imbalance does not correct itself.§ 13 Check yourself
§ 14 Mini case — the plant that was benchmarked on its birthday
A group with eleven integrated plants ran an internal benchmarking exercise. Two plants in the same state, comparable size, comparable equipment vintage. Plant A’s limestone cost: ₹295/t. Plant B’s: ₹448/t. Plant B’s mine manager was asked to close a ₹153/t gap.
What the gap actually was. Plant A operates on a lease granted in 2009. Plant B operates on a block won at auction in 2019 at a 71% premium. Statutory payments alone: A pays ₹80 + ₹24 + ₹2.40 = ₹106. B pays ₹80 + ₹8 + ₹2.40 + (0.71 × ₹239) = ₹260. The difference is ₹154 a tonne, which is the entire gap to within a rupee.
Their mining costs were within ₹12 of each other. On the only part of the number either manager controls, Plant B was marginally the better operator.
What the exercise cost. Four months of a mine manager’s attention aimed at a number he could not move, and a credibility problem for the benchmarking exercise that outlasted the engagement. The genuine finding — that the group’s post-2015 blocks carry a structurally higher limestone cost that will not converge, and that this should change how it evaluates future bids — was never reached, because the analysis stopped at the total.
A composite scenario. The rates, the FY25 average premium and the IBM sale price are sourced; the plants are not real ones.
The transferable move. Decompose every cost line into the part that is vintage and the part that is performance before you benchmark it. On the limestone line the split is knowable exactly — the rates are published and the lease date is a matter of record. There is no excuse for getting this one wrong, and it is got wrong constantly.
§ 15 Go deeper
§ 16 Carry forward
- “When was this lease granted — before or after January 2015? That one date is worth about a hundred and thirty-five rupees a tonne on your limestone line, and no operating decision touches it.”
- “The auction premium isn’t a capital cost. It’s a percentage of every tonne you ever dispatch, for fifty years, indexed to a price you don’t set.”
- “Your reserve figure and your usable rock aren’t the same thing. Nationally, reserves are eight percent of resources, and about a third of those sit under forest.”
- “A tonne of limestone is worth about two hundred and forty rupees ex-mine and carries an eighty-rupee royalty. That’s an effective rate of a third, and it doesn’t fall when prices do.”
- “Before we benchmark this line — how much of the gap is vintage and how much is performance? On limestone we can split that exactly, because the rates are published and the lease date is a matter of record.”