The vocabulary trap
The dangerous words in a cement plant are not the ones you have never heard. They are the ones you think you know, which mean two different things to two people standing in the same room — and the numbers that differ by basis rather than by fact.
On this page · 16 sections
- 01The one idea
- 02The parts — four families of trap
- 03First principles — why you cannot add heat to electricity
- 04What must be true
- 05The quantitative anatomy — the conversion table
- 06What goes wrong here — the six collisions
- 07Scenarios from the field
- 08The numbers that matter
- 09Build the model — the converter, and the drill
- 10Upgrade paths — where units decide the business case
- 11The frontier, and who is running it
- 12What’s changing now
- 13Check yourself
- 14Mini case — the saving that halved twice
- 15Go deeper
- 16Carry forward
§ 01 The one idea
A consultant loses a room in one of two ways. The first is saying a number that is wrong. The second, more common and more damaging, is saying a number that is right on a different basis — and being corrected by a process engineer who does not then trust anything else you say.
Almost every disagreement about cement numbers is a basis disagreement, not a fact disagreement. India’s capacity is 668 MTPA and also about 700 Mt and also 620+ Mt. The sector burns 726 kcal/kg and also 740. A tonne of cement costs ₹4,004 and also ₹4,763. In every one of those pairs, both numbers are correct and published, and the gap is definitional.
Before you say a number out loud, say its basis to yourself. Per tonne of what? Measured by whom? Which edition? Which tonnes are in the denominator? Four questions, two seconds, and they are the difference between being the person who brings clarity and the person who has to be corrected.
§ 02 The parts — four families of trap
| Family | Example | What it costs you |
|---|---|---|
| One word, two meanings in the same building | “Clinker factor” | 0.68 t clinker per t cement (commercial) and 1.5–1.6 t raw meal per t clinker (process). A factor of two apart, both in daily use. |
| One number, two bases | Cost per tonne, capacity, lead distance | ₹760–870/t on cost (F3 §05). ~80 MT on national capacity. 110 km on lead distance. |
| Unit conversions | kcal/kg ↔ GJ/t ↔ kWh/t ↔ ₹/t | The vendor quotes MJ/t, the plant runs on kcal/kg. Factor of 4.1868. Get it wrong once and the whole business case is off by 4×. |
| Vintage and edition | Benchmarks, standards, agency forecasts | CII moved the Indian average from 726 to 740 between editions. IS 8112 and IS 12269 no longer exist as live standards. |
§ 03 First principles — why you cannot add heat to electricity
A joule is a joule, so it is tempting to convert everything to one unit and add it up. Do not. A plant’s thermal energy and its electrical energy are different goods with different prices and different carbon, and the ratio between them is not a physical constant — it is a commercial one.
The physics part is fixed: 1 kWh = 3.6 MJ = 860 kcal. But that is the energy delivered. To make that kilowatt-hour from coal you burn roughly three times as much primary energy, because a thermal power station converts at about 35%. So electricity carries about three times the primary energy and roughly two and a half times the CO₂ of the same delivered heat — and, in India, about 4.7 times the money.
This course converts 1 kcal/kg of clinker into ₹1.90 per tonne of clinker (A7 §11 derives it). That number is not physics. It is the kiln fuel cost per thousand kilocalories, and Indian producers disclosed the following for the same quarter, Q1 FY27: JK Cement ₹1.53 · JK Lakshmi ₹1.65 · Ambuja ₹1.66 · ACC ₹1.67 · JSW ₹1.80 · Shree ₹1.95 per ’000 kcal.
That is a 27.5% spread between the cheapest and the dearest, driven by fuel mix, petcoke share, alternative fuel rate and location. ₹1.90 sits near the top of the disclosed range, above the ₹1.71 mean. A6 §05 shows why the spread is a fuel-mix story rather than a procurement-skill one. Ask the client for their own figure — it is on their monthly cost sheet — and use it. On a 6,000 tpd line, the difference between ₹1.53 and ₹1.95 is about ₹0.08 crore a year for every kcal/kg, which compounds fast across a fifty-item register.
§ 04 What must be true
This lesson is about numbers that are wrong for reasons that have nothing to do with measurement. So its conditions are conditions on a number, not on a plant. Any figure that fails one of these cannot be compared with any other figure, however carefully either was measured.
| What must be true of a number | Why | The trap | How you check it |
|---|---|---|---|
| 1. The denominator is stated | Per tonne of what — clinker, cement, cementitious, raw meal, fuel? These differ by 30–50% routinely and by more where the clinker factor is low. | kWh/t | Say the denominator out loud. A petcoke grinding penalty is 53% per tonne of fuel and 0.7 kWh per tonne of clinker — the same measurement, two denominators, two different conversations. |
| 2. The boundary is stated | Does the figure include the quarry? Captive power? Packing? Drying? A “raw material preparation” benchmark of 25–35 kWh/t covers quarry to silo; a raw mill figure of 10.6 covers one machine. Both are correct. | section vs whole plant |
Draw the boundary on paper before accepting the number. If the source does not draw one, the number is unusable for comparison regardless of how precise it looks. |
| 3. The edition and date are stated | Benchmarks are republished and revised. Protocols change versions. A figure quoted without its edition is a figure nobody can check, and it will usually be the undated version that is in circulation. | — | Ask for the version and the publication year. One benchmarking body currently publishes two different coal grinding figures — one on a dated PDF, one on an undated web page (A6 §05). The undated one is the one most often quoted. |
| 4. Heat and electricity are kept apart | A kilowatt-hour of electricity and a kilowatt-hour of fuel heat are not the same thing economically or thermodynamically. Converting between them requires a stated conversion convention, and there are several. | kcal/kg vs kWh/t |
Never add them. If a saving is quoted in a single energy unit across both, ask which conversion factor was used and at what efficiency. |
| 5. The basis of a cost is stated | Company-deck cost per tonne and consolidated-accounts cost per tonne differ by hundreds of rupees for the same company in the same quarter, because they include different things. | ₹/t | Rebuild the number from the filing. If two sources differ by ₹700–900 a tonne, they are almost certainly both right on different bases. |
| 6. A rate is distinguished from an average and from a peak | A plant that ran at 100% substitution for forty hours did not run at 100% substitution. A monthly average, an annual average and a demonstration are three different claims, and all three are published as “achieved”. | — | Ask over what period. The question “for how long?” disposes of more bad benchmarks than any other single question in this course. |
| 7. A plant figure is distinguished from a group figure | Group averages are published constantly and plant figures almost never. A group clinker factor, substitution rate or cost stack describes twenty plants, none of which is the one you are standing in. | — | Ask whose number it is. No clinker factor, no cost stack and no thermal substitution rate has ever been published for a single named plant. Anyone who offers you one has given you a group number. |
Seven questions, in order, applied to any number before it enters a deck: per tonne of what; inside what boundary; from which edition; heat or electricity; on which cost basis; over what period; whose plant. It takes about ninety seconds and it is the single highest-return habit in this course.
The reason it matters is not pedantry. A recommendation built on a mis-based number survives internal review, because everyone in the room uses the same shorthand and nobody re-derives it. It fails in front of the plant head, who knows which denominator his plant reports on.
§ 05 The quantitative anatomy — the conversion table
| From | To | Multiply by | Note |
|---|---|---|---|
| kcal | kJ | 4.1868 | The one number to memorise. Everything else follows. |
| GJ/t clinker | kcal/kg clinker | 238.8 | 3.10 GJ/t = 740 kcal/kg. European literature uses GJ/t; Indian plants use kcal/kg. |
| kcal/kg clinker | kcal/kg cement | × clinker factor | 740 kcal/kg clinker at 0.68 = 503 kcal/kg cement. Cement-basis and clinker-basis heat rates are not comparable. |
| kcal/kg clinker | ₹/t clinker | × fuel ₹/’000 kcal | See the warning above. ₹1.53–1.95 disclosed range. |
| MTPA clinker | tpd clinker | ÷ operating days | At 330 days, 1 MTPA = 3,030 tpd. Ask the plant for its own days; 330 is a convention, not a standard. |
| tpd clinker | MTPA cement | × days ÷ clinker factor | A 10,000 tpd kiln at 330 days and 0.68 anchors a ~4.9 MTPA cement complex. |
| kWh/t cement | kcal/kg clinker in money | × 4.7 | At ₹6.0/kWh, ₹1.90/’000 kcal, clinker factor 0.68. Recompute for the client. |
| ₹/t cement | ₹ crore/yr | × Mt/yr × 0.1 | ₹100/t on 3 Mt/yr = ₹30 crore/yr. The arithmetic that turns a process finding into a board item. |
Terms that will be used at you without explanation
| Term | What it is | The number to expect |
|---|---|---|
| LSF · lime saturation factor | How much lime the mix carries relative to what the silica, alumina and iron can chemically hold. The master control on burnability. | 92–98 |
| SM · silica modulus | Silica against alumina plus iron. High SM means less melt, harder burning, more coating trouble. | 2.3–2.7 |
| AM · alumina modulus | Alumina against iron. Sets the melt viscosity and how the coating behaves. | 1.3–1.6 |
| Free lime | CaO that did not react. The kiln’s report card, measured hourly. | 0.5–1.5% |
| Litre weight | Mass of a fixed volume of sized clinker. A crude, instant, extremely useful burn indicator. | 1,250–1,400 g/l |
| R90 / R45 | Residue on a 90 or 45 micron sieve — coarseness. Not the same as Blaine. | R90 12–18% (raw meal) |
| Blaine | Specific surface area of cement, cm²/g. Fineness, and the main power lever in the cement mill. | 2,800–3,600 cm²/g |
| False air | Air leaking into the system where it was not invited. Heats nothing, costs fan power, dilutes everything. | <5% per section |
| SEC · specific energy consumption | kWh per tonne. Always ask: per tonne of cement or of clinker? | 70–80 kWh/t cement |
| SHC / STEC | Specific heat consumption, kcal/kg of clinker. | 670–740 kcal/kg |
| TSR · thermal substitution rate | Share of kiln heat from alternative fuel. In India now a statutory obligation, not an ESG metric. | India ~5–8% |
§ 06 What goes wrong here — the six collisions
| The collision | How to defuse it |
|---|---|
| “Clinker factor” means two things. Commercially it is 0.68 t of clinker per tonne of cement. In the process department it is 1.5–1.6 t of raw meal per tonne of clinker — also called the clinker factor, also in daily use. | Say the ratio out loud with both units: “point six eight tonnes of clinker per tonne of cement”. Never say the bare number. |
| India’s capacity has three published values. CRISIL gives 668 MTPA as at March 2025. IBEF gives about 700 Mt on a grinding basis. NITI Aayog gives 620+ Mt. | All three are correct. The differences are grinding versus clinker basis and which small plants are counted. Name your source and its basis, or use the company’s own capacity. |
| Lead distance is two different measures. Ambuja and ACC report primary lead — plant to depot — at 249 and 254 km. UltraTech and Shree report a fuller cement lead at 360 and 445 km. | Never compare lead distances across companies without checking. And never divide freight per tonne by lead distance: the numerator carries packing, handling, depot and clinker transfers (F3 §06). |
| SEC per tonne of cement is not comparable across plants without normalising for clinker factor. CII’s own top-ten shows it: rank 1 at 56.1 kWh/t on a 0.63 clinker factor; rank 10 at 72.9 kWh/t on 0.81. | A plant making more blended cement looks more efficient on kWh/t of cement while doing the same grinding work. Ask for clinker factor alongside every SEC figure. |
| Benchmarks move between editions. CII’s 2023 survey put the Indian thermal average at 726 kcal/kg; the 2025 edition puts it at 740. The sector did not get worse — the sample changed. | Cite the edition and year every time. “CII v7.0, May 2025, gives 740 average and 670 best” is unarguable. “The Indian average is 726” invites a correction. |
| Withdrawn standards are still quoted daily. IS 8112 (OPC 43) and IS 12269 (OPC 53) are no longer live standards — both were absorbed into IS 269:2015, which now covers OPC 33, 43 and 53 together. | The old numbers are still printed on bags and used in tender documents. Know that they are historical designations, and cite IS 269 when it matters. IS 1489 Part 1 (PPC) was revised in 2023. |
§ 07 Scenarios from the field
A team benchmarking an Indian plant at 735 kcal/kg reported it as below the national average, citing CII’s 726 figure. The plant head produced the current edition, which gives 740, and the finding inverted in the room: the plant was above average, not below.
Neither number was wrong. CII’s 2023 and 2025 surveys have different samples, and the average moved up because the composition changed, not because plants regressed. The finding was destroyed by a missing four-digit year. This is the cheapest possible error to avoid and one of the most expensive to make.
European retrofit costs do not convert to India at the exchange rate. On waste heat recovery, the gap between ECRA’s European reference figure and an actual Indian project cost runs five to eight times — local fabrication, local erection, local civil works and a different regulatory burden. A business case built on a European €/t figure converted at the spot rate will be wrong by most of its value.
The same trap runs in reverse on energy: a vendor quoting a guarantee in MJ/t against a plant that measures in kcal/kg is 4.1868 apart, and the two parties can sign a contract without discovering it.
An earlier draft of this course wrote: “at ₹1.90 per ’000 kcal, that is ₹1.90 per tonne of clinker.” A reader challenged it, correctly — the sentence looks like it drops a factor of a thousand. It does not: 1 kcal/kg is 1,000 kcal per tonne, and ₹1.90 per 1,000 kcal gives exactly ₹1.90 per tonne. The two thousands cancel.
But the reader was right about the writing. A conversion whose correctness is invisible will be doubted in a client meeting, and you will lose the argument even though you are right. The lesson now shows the cancellation explicitly. Show the units on every line of a conversion, always.
§ 08 The numbers that matter
| Quantity | Value | Basis you must state with it |
|---|---|---|
| India thermal energy, average / best | 740 / 670 kcal/kg | CII v7.0, May 2025. v6.0 gave 726 / 675. Cite the edition. |
| India electrical energy | 70–80 kWh/t | Per tonne of cement, and only comparable at the same clinker factor. |
| India installed capacity | 668 MTPA | CRISIL, as at 31 Mar 2025. IBEF ~700 Mt (grinding basis); NITI 620+ Mt. |
| India clinker factor | 0.68 | CRISIL FY25, blending ratio 1.47, sample ~70% of production. |
| Raw meal to clinker | 1.5–1.6 | Tonnes of raw meal per tonne of clinker. Also called the clinker factor. |
| Kiln fuel cost | ₹1.53–1.95 | Per ’000 kcal. Disclosed Q1 FY27 range across six producers. Use the client’s own. |
| Power cost | ₹5.6–6.0/kWh | ACC ₹5.6, Ambuja ₹6.0, Q1 FY27. Falls with green share — Ambuja was at 34%. |
| Energy conversion | 4.1868 | kcal to kJ. And 1 GJ/t = 238.8 kcal/kg. |
§ 09 Build the model — the converter, and the drill
Every one of those outputs changes when the client’s fuel or power price changes. That is the point of the tool. A rule of thumb carried from the last engagement is the most reliable way to be wrong with confidence.
§ 10 Upgrade paths — where units decide the business case
| Where it bites | What to insist on |
|---|---|
| Vendor performance guarantees | Guarantee in the plant’s own units, at the plant’s own measurement points, with the reference clinker factor and moisture stated. A guarantee in MJ/t against a plant reading kcal/kg is a dispute waiting to happen. |
| European reference capex | Never convert at the exchange rate. On waste heat recovery the European-to-Indian gap is 5–8×. Ask for an Indian comparable, or an Indian EPC quote. |
| Savings claims in kWh | Per tonne of cement or of clinker? At what clinker factor? Measured at the motor or at the incomer? Three questions that routinely halve a claimed saving. |
| Payback calculations | At whose fuel price? A payback computed at ₹1.95/’000 kcal and delivered at ₹1.53 is 27% longer than promised. |
| Capacity claims | tpd at what operating days, and is that clinker or cement? The same plant is “2 MTPA” and “2.9 MTPA” depending on which and on the clinker factor. |
§ 11 The frontier, and who is running it
The frontier here is not technology. It is the rulebook that defines what the industry’s numbers mean — and it is moving fast enough that figures measured two years apart are now on different bases. Three changes matter, and all three bite between now and 2029.
| Status | What | Who, and when | What it changes |
|---|---|---|---|
| IN FORCE | A new CO₂ and energy protocol | GCCA Cement CO₂ and Energy Protocol Version 4.2, published September 2026, superseding Version 3.1 (the former CSI protocol). | Roughly 70% unchanged, 30% new or revised. New dedicated sections on calcined clay and CCUS, new provisions on recarbonation, revised definitions for mineral components, updated kiln-fuel accounting, alignment with ISO 19694-3. Default emission factors themselves may shift reported results. Mandatory from 2028 reports; V3.1 comes off the GCCA website in September 2027. Until 2029, every CO₂ intensity figure needs its protocol version attached. |
| IN FORCE | The industry’s benchmark dataset | GCCA “Getting the Numbers Right”, now GNR 2.0. Latest data year 2024, collected mid-2026. | A two-year non-disclosure period is required by the UK competition authority, which is why GNR is always about two years behind — build that lag into any benchmarking exercise. GNR publishes thermal energy including and excluding fuel drying as two separate indicators, and both clinker-to-cement and clinker-to-cement-equivalent ratios. Comparing the wrong pair is the commonest error in this dataset. |
| IN FORCE | CBAM, definitive regime | EU. Transitional reporting ran 2023–2025; the definitive regime applies from 1 January 2026, first declaration for 2026 due in 2027. | The functional unit for cement goods is tonnes of clinker content, not tonnes of cement. The benchmark for finished Portland cement is zero — free allocation flows through the clinker benchmark (0.666 t CO₂e/t grey clinker, 0.859 white) and the clinker ratio. Slag and fly ash are not CBAM goods, which is a material commercial asymmetry in favour of blended cement. |
| IN FORCE | India’s compliance market | Greenhouse Gases Emission Intensity Target Rules, 2025, notified 8 October 2025. Cement is one of four covered sectors. | Metric is tCO₂e per tonne of cement equivalent, baseline year 2023-24, with notified targets for 2025-26 and 2026-27 — obligations have already begun. 186 named cement facilities are listed, each with its own target. Integrated plants range from about 0.28 to 1.06 tCO₂e/t. A plant’s own notified target is the single most useful public number about it in India, and it is in a gazette notification rather than in any company disclosure. |
| STANDARD | The standard under all of it | EN 19694-3:2016 and ISO 19694-3:2023 — determination of greenhouse gas emissions in energy-intensive industries, cement. | The chain runs EN/ISO 19694-3 → GCCA Protocol V4.2 → GNR. Worth knowing because it tells you where a definition comes from when two sources disagree. There is no equivalent standard for reporting energy intensity in cement — only for greenhouse gases. |
1. “The CBAM factor starts at 100% in 2026.” It starts at 97.5%. The Directive’s “CBAM factor” is the share of free allocation retained; the share actually payable is its complement, 2.5% in 2026. Both sets of numbers circulate and they are opposites. Say which you mean.
2. “CBAM has a default value per tonne of cement.” There are three different instruments here — benchmark values used in the free-allocation adjustment, default values for unverified embedded emissions, and a standard minimum factor of 0.525 t CO₂ per tonne of clinker in the sector guidance. Conflating them is the standard error.
3. “Protocol v3.1 is current.” V4.2 has been published since September 2026.
4. “The clinker factor is 82%” and “the clinker factor is 72%”, both quoted for the same industry in the same year. One is clinker-to-cement, the other clinker-to-cementitious. Reconcile the denominator first.
§ 12 What’s changing now
Current as of August 2026 · refresh every six monthsIndia’s Carbon Credit Trading Scheme moved cement into a compliance regime with notified emission-intensity targets and a tradable certificate, the CCC. Two terms that barely existed two years ago now appear in board papers.
What it means: ask for the plant’s CCTS baseline and notified intensity target on day one. It is a number the plant now has to know, and it constrains the clinker-factor and fuel decisions in every later lesson.The Solid Waste Management Rules made a minimum thermal substitution rate a statutory requirement rather than a voluntary target.
What it means: when a plant says “our TSR is 6%”, that is now a compliance position, not a sustainability slide. The right follow-up is about RDF supply contracts and quality, not intent.LC³ — limestone calcined clay cement, standardised in India as IS 18189:2023, the only supplementary material a producer can lease rather than depend on another industry for. And RDH — rotodynamic heater, the electrified calcination unit going into Adani’s Boyareddypalli plant.
What it means: both are small today and neither belongs in a five-year cost forecast. Both belong in a strategy conversation, because they are the first two credible routes at the process CO₂ that F3 §03 showed no fuel switch can touch.§ 13 Check yourself
§ 14 Mini case — the saving that halved twice
A vendor proposes a cement mill separator upgrade on a 2.4 Mt/yr plant. The claim: “4 kWh/t saving.” At ₹6.0/kWh that is ₹24/t, or ₹5.8 crore a year. Capex ₹7.5 crore. Payback fifteen months. The plant is enthusiastic.
Question one: per tonne of what? The 4 kWh/t is per tonne of OPC, measured during the trial. The plant makes 78% PPC. On its actual product mix the mill-section saving is about 3.1 kWh/t of cement.
Question two: measured where? At the mill motor. The plant’s own SEC is measured at the section incomer, which includes the separator fan, the bucket elevator and the auxiliaries. Two of those get busier under the new separator. Net at the incomer: about 2.6 kWh/t.
Question three: at whose power price? The vendor used ₹6.5/kWh, a grid tariff. This plant runs 31% green power and its blended cost is ₹5.4/kWh.
The corrected case: 2.6 kWh/t × ₹5.4 = ₹14/t, or ₹3.4 crore a year on 2.4 Mt. Payback twenty-six months, not fifteen.
And the project still went ahead — twenty-six months is a good payback, and the plant knew the number was real because it had been through three questions. The point of this lesson is not to kill projects. It is that a claim that survives interrogation gets funded faster than one that has to be defended after the fact.
A composite scenario. The prices, the green-power share and the OPC/PPC split are drawn from Q1 FY27 disclosures; the plant is not a real one.
The transferable move. Per tonne of what, measured where, at whose price. Three questions, under a minute, and they routinely move a savings claim by a factor of two. Ask them of your own numbers before someone else does.
§ 15 Go deeper
§ 16 Carry forward
- “Per tonne of what — cement or clinker? And measured at the motor or the incomer?”
- “Which edition of the CII benchmark is that? The average moved from 726 to 740 between surveys, and it wasn’t because plants got worse.”
- “Before we use ₹1.90 per thousand kilocalories — what does your cost sheet say? The disclosed range across listed producers this quarter was ₹1.53 to ₹1.95.”
- “A kilowatt-hour per tonne of cement is worth about four point seven kilocalories per kilo of clinker in money. So your mill project and your cooler project are worth roughly the same, and we should sequence them on risk, not on size.”