Activated carbon: what iodine number and CTC really tell you

The answer, up front

Iodine number, carbon tetrachloride activity and hardness are relative indicators of how far the activation process went and how well the granule holds together. None of them is a measure of how much of your contaminant the carbon will remove. That is not our opinion. It is what each test method says about itself, in its own scope and significance text.

Three practical consequences follow, and they are the reason this article exists. First, the three numbers are in different units and on different scales, so they cannot be ranked against each other: iodine number is milligrams per gram, CTC activity is a percentage by mass, and hardness is a dimensionless index. Second, “hardness 95” from two suppliers is not the same statement, because at least five different methods are in commercial use and no published conversion exists between them. Third, and least expected: the Indian Standard for powdered activated carbon does not specify an iodine number at all.

There is also a commercial number that decides the economics and is usually printed as a typical value rather than a specification. Carbon is priced per tonne and used per cubic metre. Two carbons at the same price per tonne can differ by more than half in the bed volume they fill, and nothing in the iodine number tells you which is which.

This is a buyer’s reading guide to specification numbers. It is not a design document, not a safety procedure and not legal advice. Where it names a statutory duty it says whose duty it is. Where a figure could not be confirmed in a current primary source, it says so rather than rounding the uncertainty away.

Start with the Indian standards, because the structure is not what it looks like

Almost every specification sheet and tender document we have read gets this wrong in the same way, so it is worth settling before any number is discussed. IS 877 is not a specification. Its title is Activated Carbons, Powdered and Granular — Methods of Sampling and Test. It contains test procedures and no requirements table, so a limit written “as per IS 877” is not a limit against anything. The specifications are two separate standards, split by form.

Both of the standards a buyer is most likely to cite were revised recently, and the free copies that turn up first are the withdrawn ones. That is the trap in a sentence. IS 877 : 1989 and IS 2752 : 1995 are both fully readable without payment, on the public archive of Indian Standards, while the 2024 and 2025 editions sit behind the Bureau’s own portal. The path of least resistance leads directly to a superseded edition, which is how a specification written in 2026 ends up quoting a 1995 table.

What we have and have not read, stated plainly, because every figure below depends on it. We have not read IS 877 : 2024 or IS 2752 : 2025. The Indian figures in this article come from the withdrawn IS 2752 : 1995 and from the Bureau’s own wide-circulation draft of the fourth revision, circulated in October 2024 for comments closing in December 2024, which states that it incorporates Amendment No. 1 and updates the reference and packing clauses. A draft is not a standard. Do not write any figure below into a purchase order without reading the published current edition. On IS 8366 we could not establish the current status at all: a second revision was circulated in October 2024 alongside the IS 2752 one, and the IS 2752 draft was published as IS 2752 : 2025, so an IS 8366 : 2025 is plausible and unconfirmed.

One further caution about the Bureau’s portal, learned the hard way on another article in this series. The portal’s amendment fields are not a reliable negative. We have seen an entry read “No amendment issued” and “Amendment : 0” against an amendment that was plainly established by gazette notification. A portal page showing no amendment does not establish that none was issued. For this product we reached no gazette document at all, so we state no amendment position for IS 877 : 2024 or IS 2752 : 2025 in either direction.

The grade numbering in IS 2752 runs the opposite way to intuition

IS 2752 defines two types by end use, and Type 1 is the respirator and solvent-recovery grade while Type 2 is the water-treatment grade. On the two rows where both types carry a figure, Type 1 is the demanding one. A buyer who assumes Type 1 is the ordinary water grade, or that Type 2 is the premium, inverts the table.

We are printing the method column as well as the two requirement columns, because leaving it out is how this table gets misread — and we misread it ourselves before checking. In the Bureau’s own layout the requirement columns are followed by two method-reference columns, one holding IS 877 clause numbers and one holding Annex A subclauses. A flattened copy drops a clause number into the Type 2 cell, and a clause number looks exactly like a requirement.

Note what the Type 2 column actually does on two of those rows: it sets no requirement at all. Carbon tetrachloride adsorption and retentivity index are Type 1 characteristics, and the water-treatment type simply has none. That is a sharper point than a low figure would be, and it has a direct consequence: a quotation offering water-treatment carbon “to IS 2752” and printing a CTC number is printing a figure the standard does not require of that type — which is useful information, freely given, but not a conformity statement.

And watch the direction of each limit, because the table does not run one way. Five rows are minima, where a higher figure is the better carbon. Three are maxima: moisture, ash and half dechlorination value. The last of those is the one that catches people, because it is the only maximum where the two types differ, and the water-treatment type is allowed the worse value — 7 cm against 4. A buyer skimming a mixed-direction table for the bigger number will pick wrong at least once.

Two honest limits on that table. The Type 2 ash cell could not be resolved. The text layer of the Bureau’s draft extracts it as a figure that cannot be right for a water-treatment carbon, and the copies of the withdrawn edition available to us disagree with each other, so we print nothing rather than a guess. Type 1 ash is legible as 5 percent by mass maximum in both. And the draft’s Foreword states that the revision incorporates Amendment No. 1 without saying what that amendment changed, so we cannot tell you whether any figure above is pre-amendment or post-amendment. The wording “hardness number” is worth one further note: it is measured by the IS 877 method against a 600-micron sieve, which is not the ASTM ball-pan test and does not map onto it.

IS 2752 also does something unusual for a national product specification and worth knowing about: it sets a minimum surface area. Most producer data sheets print surface area as a typical value or omit it. If your supply is being offered against IS 2752, that row is a genuine, quotable requirement rather than a courtesy figure — subject to the caveat above about reading the published 2025 edition. What IS 2752 does not set a figure for, so that nobody claims it does: no apparent or bulk density, no pH, no methylene blue value, no phenol value and no heavy-metals panel. A sheet presenting any of those “as per IS 2752” is attributing a requirement to a standard that does not contain one.

Particle size is the instructive exception, and the distinction is worth a buyer’s attention. The standard does carry a particle size clause; what it says is that the particle size shall be as agreed between the purchaser and the supplier. That is not the same as silence. It means the standard has handed the parameter to your contract, so “particle size as per IS 2752 and as agreed” is a coherent clause while “particle size as per IS 2752” alone is not. Several of the characteristics the standard leaves out work the same way in practice: if it matters, it has to be in the purchase order, because nothing else is going to put it there.

And the powdered standard specifies a different set of things entirely

IS 8366 defines three types, again by end use, and the characteristics are almost disjoint from the granular standard’s. There is no iodine number, no CTC activity, no hardness and no density in the powdered specification. What stands in for adsorption capacity is a methylene blue titration the standard calls decolourizing power.

Three things in that table deserve a sentence each. The ash ceilings run backwards from intuition — the vegetable-oil grade is permitted the most ash and the pharmaceutical grade the least — so anyone reasoning that food contact must mean low ash will write the table upside down. pH is a band, not a ceiling. Writing “pH max 8” drops the lower bound, which is the one an acid-washed carbon fails. And the most-quoted figure in the standard is explicitly non-binding: the table carries a footnote that the decolourizing power requirement by the methylene blue test is for guidance only, and that specific requirements for each use may be agreed between purchaser and supplier.

That footnote sits in tension with the standard’s own structure, because decolourizing power is listed as a critical characteristic for all three types. So presenting 50, 70 and 205 mg/g as a hard specification misstates the standard, and dismissing them as unimportant misstates it too. The honest reading is that the standard expects you to agree a number, and offers those as a starting point. Agree one.

A specific equivalence to avoid, because it sends buyers to the wrong product. Decolourizing power by methylene blue is not an Indian molasses number. Methylene blue is a far smaller probe molecule than the colour bodies a molasses test is standing in for, and the two report on different parts of the pore structure. A sugar or edible-oil buyer told that the Indian decolourizing power is India’s molasses number will specify the wrong carbon.

Iodine number: what it is, and the sentence its own method prints

ASTM D4607-14(2021) is the governing method, and it defines the iodine number as the milligrams of iodine adsorbed by one gram of carbon under the specified test conditions. Iodine is adsorbed from aqueous solution, and the method’s own scope describes what it is for: measuring the relative activation level of unused or reactivated carbons.

Then come the disclaimers, which are ASTM’s words and not ours. The iodine number “is a relative indicator of porosity”. It “does not necessarily provide a measure of the carbon’s ability to absorb other species”, in the method’s own words. It can approximate surface area for some carbon types, but that relationship “cannot be generalized” because it varies with raw material, processing conditions and pore volume distribution. And the measured value can be shifted by adsorbed volatiles, sulphur and water extractables. The American Water Works Association’s granular carbon standard repeats the same three cautions almost word for word.

So never convert an iodine number into square metres per gram. The habit of treating one milligram per gram of iodine number as one square metre per gram of surface area was tested directly in the peer-reviewed literature, and it gave significantly larger errors than a fitted two-variable model — which itself still predicted BET area with a mean relative error around twelve percent across the authors’ test set. That twelve percent is the fitted model’s prediction error, not an uncertainty attaching to the iodine number itself. But it is the margin you inherit if you do the conversion, and you would be inheriting it on a target quantity that is itself model-dependent. The authors say as much: the method does not substitute for proper characterisation of the textural properties.

Why a high iodine number can be the wrong carbon

The mechanism is a matter of molecular size, and once seen it is obvious. The iodine molecule presents a footprint of about 0.4 square nanometres and can enter the smaller micropores. Methylene blue presents about 2.08 square nanometres and can only enter large micropores and mesopores. A carbon whose porosity is overwhelmingly in the narrowest pores will score superbly on iodine and may be unable to take up a large molecule at all.

That is precisely why the standards offer other tests. The AWWA granular standard provides optional surrogate tests — a tannin value and an acetoxime number — and the powdered standard adds tests for taste-and-odour compounds and for total organic carbon adsorption. ASTM separately publishes an aqueous-phase isotherm practice and a rapid small-scale column test, both of which run on your contaminant rather than on a probe molecule. If colour bodies, humic matter or high-molecular-weight organics are what you are removing, those are the tests that answer your question and the iodine number is not.

What we could not find, and say so rather than filling the gap: no primary source we reached gives a numeric pore-width window for what the iodine test probes. The verified statements are qualitative — the probe footprint, and that it reaches the smaller micropores. Published figures of the form “the iodine number measures pores of 1.0 to 1.5 nanometres” are not something we could source, and we do not repeat them. We also found no published case study of two equal-iodine carbons performing differently with data attached, so the argument above is what the standards say, not a documented demonstration.

One procedural point that matters if you are commissioning the test rather than reading the result. The determination is bracketed, not a single shot. The method requires the residual iodine concentration in the filtrate to land inside a specified window, reported in the literature as 0.008 to 0.04 normal, and the test is repeated with different carbon masses if it does not. A laboratory quoting an iodine number from one determination with no residual figure has not necessarily run the method.

CTC activity: a percentage, not a milligram figure, and the test method is alive

ASTM D3467-04(2020) governs carbon tetrachloride activity, and the first thing to fix is the unit. CTC activity is the percentage ratio of the mass of carbon tetrachloride adsorbed at saturation to the mass of the carbon sample. It is a percentage by weight. So “CTC 60” on a quotation with no unit means sixty percent by mass, and it cannot be compared with, added to or ranked against an iodine number in milligrams per gram. They are different kinds of quantity.

What it measures, in ASTM’s own significance text, is the activation level, and it is “basically a measure of the pore volume”. A review authored from the United States National Institute for Occupational Safety and Health puts the same point as the activity reflecting the micropore volume and being a relative measure of how far the activation process was carried. And ASTM adds the familiar caution, and the wording is worth having exactly: the activity number “does not necessarily provide an absolute or relative measure of the effectiveness” of the carbon on other adsorbates or under other operating conditions.

A persistent error worth correcting: ASTM has not withdrawn the carbon tetrachloride test. D3467 is active in its 2020 reapproval, despite carbon tetrachloride’s status as an ozone-depleting substance. What exists is a recommended substitute, not a withdrawal. ASTM D5742-16(2023), the butane activity method, states in terms that it “can be used as a non-ozone depleting substitute for the carbon tetrachloride activity test in Test Method D3467”, and it publishes a figure titled as an experimental butane-versus-carbon-tetrachloride correlation. Butane activity is also a percentage. The method notes that it was not designed for powdered carbon — but also that it has been used successfully with powdered carbon where the flow rate, time or sample volume are adjusted to keep the pressure drop acceptable, so a powdered grade quoted on butane activity is not automatically out of order. Ask how the adjustment was made.

Two butane standards exist and they are different tests. D5742 is butane activity, a mass percentage at saturation, and it is the substitute for the CTC test. D5228-16(2023) is butane working capacity, the difference between loading at saturation and loading retained after a specified purge, and it is expressed per unit volume of carbon rather than per unit mass. A data sheet calling D5228 the butane activity method has conflated them, and comparing a per-volume figure with a per-mass one is a unit error rather than a naming quibble.

On the numerical relationship between the two: the NIOSH-authored review states that CTC activity is proportional to normal-butane activity and gives a factor of about 2.55 to 2.57. Calgon Carbon publishes paired CTC and butane figures for its own gold-recovery grades — CTC 50 against butane activity 19.5, and CTC 45 against 17.5 — which imply ratios of 2.56 and 2.57. Those are Calgon Carbon’s published values for its own grades, not an industry constant, and whether the factor holds across raw materials and activity levels is not something we could establish. Treat it as the right order of magnitude for a sanity check and not as a conversion to write into a contract.

There is a quiet consequence of specifying against CTC that affects your supplier rather than you. Running the test means a laboratory handling carbon tetrachloride — an ozone-depleting substance which is also the subject of its own Indian code of safety, IS 5311 : 2026, a first revision superseding the 1969 edition. A supplier who has moved to butane activity has a defensible reason, and the right response is to ask for the correlation rather than to insist on the CTC number.

Hardness: five different numbers, one word

This is where two quotations most reliably look comparable and are not. Ball-pan hardness, abrasion number and attrition are three different measurements, and the abrasion number itself comes in two variants used on different raw materials. No published conversion exists between any of them.

ASTM D3802-23 gives the ball-pan hardness number, and it is dimensionless — the method sets no unit. Read what ASTM says about it, because it is unusually candid. The property is “loosely described as hardness”; “the test does not actually measure in-service resistance to degradation”; and one of the uses it names is establishing the comparability of lots ostensibly of the same grade of carbon. And the standard records that ASTM Committee D28 has questioned the ball-pan, stirring bar and dust elutriation methods as tests of degradation resistance, and that none is fully satisfactory for all applications. A hardness number predicts lot-to-lot consistency. It does not predict your backwash losses.

The method’s scope covers granular carbon, where granular means at least ninety percent of the sample retained on a 180-micron sieve, equivalent to 80 US mesh. That definition is a floor on coarseness, not a ceiling, so it does not by itself exclude a pellet — and the AWWA granular standard expressly covers extruded carbons alongside granular ones. What is worth asking on a pelletised product is whether the method was designed with that shape in mind, because the companion particle-size method, ASTM D2862-16(2022), carries its own warning that for extruded carbons the validity of the results may be affected as the length-to-diameter ratio of the particles increases.

The clearest evidence that these are not interchangeable comes from a single producer’s own data sheets. Calgon Carbon prints “Hardness Number 95 min” on a coconut-shell granular grade, and “Abrasion Number 75 min” and “Abrasion Number 80 min” on two bituminous coal grades. Same producer, same catalogue, two different properties on products a buyer would naturally compare. Comparing 95 against 80 and concluding the coconut grade is harder is not a comparison at all — and we cannot tell you why that producer made the choice it did, because no source we read explains it. What we can say is that the AWWA standard records a current practice of using the stirring abrasion test for lignite-based and wood-based carbon and the Ro-Tap test for bituminous and coconut-based carbon — which is described as practice rather than imposed as a requirement, and which does not account for a ball-pan number appearing on a coconut grade. That the explanation does not close is itself the point: ask for the method, because you cannot infer it.

Indian practice adds a fifth. IS 877 determines hardness against a 600-micron sieve — a sieve-retention method, not a ball-pan test. And it has moved: Amendment No. 1 of January 1998 to the 1989 edition substituted 12.7 millimetres for 12.5 millimetres in the hardness clause, an apparatus dimension — so hardness results either side of 1998 are not strictly comparable even within the Indian method. If you are comparing a historical certificate with a current one, that is worth knowing before you conclude the material has changed.

And a gap we will not fill with a guess. The Bureau’s portal records two amendments against IS 877 : 1989, and the copy available to us carries only Amendment No. 1, with its own amendments table blank. So a second amendment exists and we cannot tell you its date, its clause or what it changed. We mention it because the same discipline is why this article states no amendment position for IS 877 : 2024 or IS 2752 : 2025: if you are relying on a method clause by number, confirm the amendment status of the edition you are citing rather than the edition you can download.

Higher hardness is not simply better. Calgon Carbon’s own gold recovery literature describes pre-attrition as a deliberate process step, breaking off sharp edges that would otherwise readily adsorb gold. And the same document notes that smaller particles tend to adsorb more quickly while larger particles have slower kinetics — so hardness, particle size and rate trade against one another and a hardness number carries no kinetic information at all.

The number that decides the money, and it is usually a typical value

Activated carbon is bought by weight and used by volume. A vessel is sized in cubic metres; an invoice is denominated in tonnes. Apparent density is the conversion between them, and the ASTM method that measures it says so explicitly: the number matters for designing vessels that will hold the carbon and for ordering material to fill existing vessels.

The density figures in that table are the American Water Works Association’s, published with its own caution that they show a range of products rather than specific grades and that manufacturer data should be consulted for a particular product. Two details in the interests of accuracy: three of those four rows are published in that standard as several mesh-dependent bands rather than as one span — wood as two, coconut as two and bituminous as six, with only lignite given as a single range — so the figures above are the outer envelopes across the bands, which is our compression and not the standard’s presentation; and the bed volume column is our arithmetic on those figures, not a published claim. One tonne divided by the density, with the usual unit conversion, and nothing else in it.

On the extremes of what individual producers actually publish, the spread is larger than most buyers expect. Cabot’s PetroDarco 4x10 is published at an apparent density of 400 kilograms per cubic metre on vibrating feed, which is 2.5 cubic metres per tonne. Calgon Carbon’s Filtrasorb 600 is published at 0.62 grams per cubic centimetre tamped, which is about 1.61 cubic metres per tonne. That is roughly fifty-five percent more bed volume per tonne from the lower-density product. At the same price per tonne those are not the same purchase, and at different prices per tonne the comparison cannot be done at all until both are converted to cost per cubic metre of bed — or better, to cost per unit of contaminant removed.

Which direction is better depends entirely on which constraint binds. If you are filling a fixed vessel, the denser carbon means more tonnes per changeout and a higher invoice for the same job. If you are buying a tonnage and want bed volume, the lighter carbon delivers more of it. Neither is a quality statement, and that is the trap inside the trap: a low apparent density can mean a highly activated carbon with excellent pore volume per gram, or a low-density raw material, or simply more fines and poorer packing. The density number alone does not distinguish those three.

Two method cautions on density, both of which produce wrong comparisons quietly. First, there are two different standard methods: ASTM D2854-09(2025) for apparent density applies to granular carbon only, by the same 80-mesh definition, while powdered and fine-mesh material has its own method, ASTM D8176-18(2023) for mechanically tapped density. Second, producers print different quantities under similar names. We have read “apparent density, tamped”, “apparent density, vibrating feed” and “density, backwashed and drained” on producer sheets, the last two on the same sheet with different values. Tamped, vibrating-feed and backwashed-and-drained are not the same measurement.

Surface area, and why a bigger number is not a better carbon

Surface area is the figure that appears in marketing and it is the one most often misunderstood, because for a highly microporous carbon the BET surface area is an apparent quantity rather than a physical one. The International Union of Pure and Applied Chemistry’s 2015 technical report, which the current measurement standard is built on, is explicit: for the isotherm type a microporous carbon gives, the BET area “must not be treated as a realistic probe accessible surface area”, and is better regarded as “an apparent surface area, which may be regarded as a useful adsorbent fingerprint”.

The reason is mechanical, not philosophical. In a microporous solid it may be impossible to separate monolayer and multilayer adsorption from micropore filling, so the model that converts gas uptake into an area is being applied outside the regime it describes. The number is reproducible and useful for comparing one carbon with another by the same method. It is not an inventory of accessible surface, and the familiar line about a gram of carbon holding the surface area of several football pitches is exactly the error the report warns against.

Higher is not better here either, and for a reason that is easy to miss: raising surface area generally means burning off more carbon. A more highly activated product has more pore volume per gram and less mass per unit volume, which brings you straight back to the density arithmetic above. Producers publish wide and only partly overlapping ranges for the same broad product family — Haycarb typically 500 to 1500 square metres per gram, Jacobi Carbons 800 to 1600, Kuraray 500 to 2500 — and a product published at 600 square metres per gram is sold successfully on a decolourisation performance figure instead. The spread between producers is wider than the spread that matters inside any one application.

A citation error we have seen on specification sheets. ASTM D6556 is not an activated carbon standard. It is a carbon black method, from a different ASTM committee, measuring nitrogen surface area and the statistical thickness surface area available to rubber. For activated carbon the correct citation is ISO 9277:2022, the third edition, which supersedes the 2010 edition and is based on the 2015 IUPAC recommendations. Also worth knowing when reading two producer sheets side by side: some publish pore classes as radii and some as diameters. Jacobi Carbons gives micropores as radius under one nanometre; Haycarb gives them as diameter under two. Those are the same classification expressed two ways, not a disagreement — and IUPAC itself calls the boundaries to some extent arbitrary.

Molasses number: the number you cannot compare at all

If a quotation for a decolourising carbon carries a molasses number, it is worth knowing what kind of thing it is. There is no ASTM method for the molasses number, and no standards-body method of any origin that we could find. It is an in-house test, and it reports a relative result against a reference carbon chosen by the laboratory running it.

One producer, DESOTEC, states the consequence in terms on its own technical page: “it is impossible to compare molasses numbers of different suppliers” under normal conditions, because the molasses standard solution, the test pH, whether the solution is buffered, the reference carbon and the test variant all differ between laboratories. And the direction of better flips with the variant: where the result is expressed as relative colour removal, higher is better; where it is expressed as the carbon dose needed to reach a given decolourisation, lower is better.

The incomparability is visible in the published sheets. Calgon Carbon prints “Molasses Number 200 min” on a bituminous grade. Cabot prints “Molasses decolorizing efficiency (RE) min. 65” on another product. Those are different scales entirely, neither sheet cites a method, and no conversion exists. What the molasses test does report on is the macropore and transport pore structure — the part of the pore network the iodine test says nothing about — which is why it survives commercially despite being unstandardised. If colour removal is your application, the right move is not to compare molasses numbers between suppliers but to run a trial on your own liquor.

Ash, moisture and the basis question

Ash is where the quoted number hides the thing that matters. The ASTM total ash method’s own significance text says that the amount and composition of ash may affect the capabilities of the carbon in specific end uses. Composition is doing as much work in that sentence as amount. A three percent ash carrying a high acid-soluble iron content can be worse for your process than a six percent ash that is inert, which is why separate ASTM methods exist for acid-extractable content and specifically for acid-soluble iron, and why producers sell acid-washed grades for applications needing low pH or low acid-soluble iron. If iron, or leachable metal generally, is your concern, the total ash figure is not the parameter to specify.

Moisture has a commercial edge that is easy to forget. The ASTM moisture method notes that moisture content is needed in relation to the net weight of the carbon, and that moisture typically increases during shipping and storage. A carbon bought as received at five percent moisture delivers about five percent less carbon per tonne than its dry mass suggests, and the increase happens after the certificate was issued. Where a moisture ceiling matters, specify where and when it is measured, not just the number.

The basis problem, stated honestly including the part that cuts against a confident answer. Indian practice under IS 877 reports ash, water-soluble matter and acid-soluble matter on a moisture-corrected dry basis, and takes the methylene blue sample as received. On iodine number we could not establish a dry-basis convention from any primary source. The indications point that way — the Indian draft procedure dries the sample to constant weight before the determination, and the NIOSH review describes the CTC method as working on dry carbon — but we have not read a method text that states it. What we can say is what the producer sheets do: of the sheets we read, none stated a basis for the iodine number, and the only basis stated anywhere was “as packaged” against moisture. So the basis is something to ask for, because the data sheet will not tell you.

Particle size, and what 8 by 30 actually promises

A mesh designation is a nominal trade description with producer-specific tolerances, not a definition. One producer, DESOTEC, publishes its convention for 8 by 30 as at least ninety-three percent by weight larger than 30 mesh, which is 0.60 millimetres, and at least ninety percent smaller than 8 mesh, which is 2.36 millimetres. Note that those two tolerances are different numbers and that neither is a hundred percent.

Now compare a producer sheet. Calgon Carbon’s own 8 by 30 bituminous grade is specified as at most fifteen percent retained on 8 mesh and at most four percent through 30 mesh — a materially looser oversize tolerance than the ninety percent convention above. Two carbons both honestly described as 8 by 30 can have different size distributions, and size distribution drives pressure drop, backwash behaviour and contact time. If the hydraulics matter, specify the sieve limits rather than the trade designation.

The powdered and granular boundary is defined, and it is the same line in three ASTM methods: granular means at least ninety percent of the sample retained on a 180-micron sieve, 80 US mesh, and that definition governs the particle size, apparent density and ball-pan hardness methods alike. Powdered carbon has its own sizing method, by air-jet sieving, rather than the granular one. And for an Indian buyer there is a further wrinkle: the Indian sieve series is not the US mesh series. IS 877 works in 500, 230, 125, 63 and 53-micron IS sieves, and IS 8366 specifies percentages passing 150, 125 and 75-micron sieves. A mill certificate reporting percent passing 125 micron is not directly comparable to a data sheet reporting percent through 40 US mesh.

Raw material and activation route: what is established and what is folklore

The received wisdom is that coconut-shell carbon is microporous, with high iodine number, high CTC and high hardness, while coal-based carbon is more mesoporous and therefore better for colour. We went looking for a producer-published, side-by-side, numeric comparison of pore volume by feedstock, and did not find one. Jacobi Carbons publishes pore classes and activation routes but does not link a raw material to a pore distribution. Haycarb asserts that coconut properties “perform better compared to bituminous or lignite-based carbons” and publishes no coal or lignite figures to support it. Kuraray gives no per-feedstock numbers.

What is verifiable is a pattern in which numbers producers choose to publish, which is weaker evidence but still useful. From one producer’s catalogue: the coconut granular grade carries an iodine number and a hardness number and no CTC; a bituminous grade carries an iodine number, an abrasion number and a molasses number; another bituminous grade carries an iodine number and an abrasion number. A lignite-class product from a different producer carries no iodine number at all and is specified on molasses decolorizing efficiency with a surface area of 600 square metres per gram. Which numbers appear tracks which market the grade is sold into. That is a commercially real signal and it is not a measured pore-structure comparison, and we are not going to present it as one.

On activation route the producer statements are consistent and worth knowing because the route changes the transport position later in this article. Steam activation is described as the common route for coconut shell and coal, in rotary kilns for coconut shell and vertical shaft furnaces for coal, with carbonised material exposed to steam at around 900 to 1100 degrees Celsius. Chemical activation with phosphoric acid is described for sawdust, wood and peat, carbonising at a far lower 400 to 500 degrees Celsius. Both of those are producer descriptions, attributed, and no producer source we read quantified the product differences between the two routes.

One industry-body document circulating as a source of “typical coconut carbon specifications” writes iodine as a maximum and hardness as a maximum, which is the opposite direction to every producer data sheet and to the Indian standard minima. It cites no test method for any parameter and states no basis. A buyer copying that wording into a purchase order would be specifying the reverse of what they want. Check the direction of every limit you copy.

Reactivated carbon, and the disclosure question

Thermal reactivation is a legitimate and environmentally sensible route, and the standards treat it as a different product with its own rules rather than as a cheaper version of the same thing. The AWWA standards for virgin granular carbon and for virgin powdered carbon both exclude reactivated material from their scope, which has a sharp consequence: a quotation claiming compliance with the virgin granular standard is, by that standard’s own scope, claiming virgin carbon.

The reactivation standard itself is careful in ways a buyer should borrow. It does not claim equivalence to virgin carbon. What it asks is that reactivated material have particle size distribution, effective size and uniformity coefficient comparable to the virgin product, so that filtration and adsorption behaviour are maintained. It records that adsorptive capacity declines with service time and that lightly loaded carbon recovers a higher percentage of its capacity than heavily loaded carbon, with results depending on carbon type and water quality — and it recommends testing a representative sample, ideally with the plant’s own water.

Three items in that standard’s foreword are worth writing into your own purchase order — and note that a foreword in these standards is expressly not part of the standard, so these are guidance you should make contractual rather than requirements you can assume. Commingled material should be clearly identified as commingled, and used only where regulations allow and the purchaser approves. Make-up carbon should come exclusively from virgin material or from reactivated material from the same potable water facility, to avoid introducing leachable inorganics from another source, and make-up can significantly affect size distribution. And the purchaser carries disclosure duties of its own — it must state whether the spent carbon has been exposed to particular contaminant classes, supply a representative sample and the spent carbon’s history, and state whether a potable-water conformity requirement applies. Reactivation is a two-sided disclosure, not a one-sided warranty.

A figure not to transfer, because it is widely quoted out of context. The United States Environmental Protection Agency’s cost manual states that the adsorption capacity of regenerated carbon is typically about half that of virgin material — but that figure is for in-situ low-temperature steam regeneration of solvent-recovery beds, at 250 to 350 degrees Fahrenheit, which the AWWA standard separately says may not be as effective as high-temperature reactivation and treats as a different process. It is not a verified figure for off-site thermal reactivation of water-treatment carbon, and we do not apply it to one. The same manual separately uses a fifty percent rule for design working capacity, which is a different fifty percent again and is easily merged into one false claim.

The hazards buyers do not expect, and one piece of firefighting advice that is wrong

This section is short and deliberately conservative, because the most repeated statements about this product in general circulation are the least well sourced. Nothing here is a procedure; it is what the producers and regulators publish, attributed.

Oxygen depletion in a confined space

Wet activated carbon removes oxygen from the air around it, and in a closed or partly closed vessel the depletion can reach hazardous levels. Every major producer we read states this, in close to those words — Calgon Carbon, Jacobi Carbons, Cabot, CarboTech, DESOTEC and Arq Solutions. Several go further and tell you what to do: Jacobi Carbons states that whenever workers enter a vessel containing activated carbon the oxygen content should be determined; Chemviron instructs the reader to check oxygen content before entering; Cabot says the space should be evaluated for oxygen and carbon monoxide by a qualified person. Jacobi also locates the hazard in storage, not only in vessels: wet carbon in a closed area can deplete oxygen from the air.

Three honest qualifications, because the gap between what is published and what is repeated is wide. First, the literature is about wet carbon. One producer’s safety data sheet draws the line explicitly, stating that the product does not displace oxygen in the ambient atmosphere but slowly adsorbs oxygen from a confined space when wet. We did not establish that dry carbon creates an oxygen-deficient atmosphere, and we do not say it does. Second, no producer document we read explains the mechanism — one states the effect and says in terms that it does not explain it — so the common explanations in circulation are inference. Third, we searched the public accident databases of several regulators and found no incident report attributing a death to activated-carbon oxygen depletion. The hazard rests on consistent producer assertion. That is a reason to take it seriously and a reason not to illustrate it with somebody else’s incident: the oxygen-deficiency fatalities that surface on a search for this are nitrogen-inerting cases with no carbon in them.

What the confined-space rules themselves add is useful, because it shows how many triggers one vessel can pull. Under the United States occupational standard a hazardous atmosphere includes an oxygen concentration below 19.5 percent or above 23.5 percent, airborne combustible dust at or above its lower flammable limit, and an atmospheric concentration of a substance with a published dose or permissible exposure limit which could result in exposure in excess of that limit. A loaded carbon vessel can meet three of those at once — low oxygen from the carbon, combustible dust from the carbon, and a toxic contaminant from whatever the carbon adsorbed. The same standard requires the pre-entry test to run in a set order: oxygen, then flammables, then toxics.

Self-heating, bed fires and the water question

The authoritative document is a chemical safety alert issued by the United States Environmental Protection Agency in May 1997 on fire hazards from carbon adsorption deodorising systems, and the mechanism it describes is not obvious. Adsorbing certain chemicals releases a large amount of heat, creating hot spots in the bed; if flammable vapour is present and a hot spot reaches its autoignition temperature, a fire starts wherever oxygen is available. The substances it names are organic sulphur compounds such as mercaptans, plus ketones, aldehydes, some organic acids, and high vapour concentrations generally.

And the way air gets in is the part nobody anticipates: daily temperature cycling. Daytime heat pushes vapour out of the tank, and night-time cooling creates a slight vacuum that draws air back in. If the carbon is already hot, that incoming air supplies the oxygen. The same alert recommends that a qualified engineer or technician supervise the design, construction and operation of such a system; where testing or known reactions indicate a fire potential, designing so air cannot enter over the bed, for instance with vacuum breakers; fitting flame arrestors between carbon containers and containers of flammable chemicals; keeping carbon containers separated from containers of flammable or combustible substances and able to be removed rapidly if one overheats; diluting the inlet air or time-weighting the inlet concentration so heat can dissipate; pre-wetting the carbon; testing an unknown vapour stream against the carbon for heat release before service; and inspecting systems frequently for hot spots and fires. A 1995 fire and explosion at a chemical terminal, involving three tanks of crude sulphate turpentine connected to activated carbon deodorising drums, is the documented incident behind that alert.

Now the advice that is wrong, and it is the instinctive one. A solid water stream is condemned: Cabot’s sheet reads “DO NOT USE a solid water stream as it may scatter and spread fire” and recommends a fog if water is used. Arq Solutions goes further and says to use water only when necessary, because water will displace surface oxygen and the subsequent drying can cause localised heat build-up — which is to say applying water can set up the next self-heating event. So the one firm negative is the jet: no source we read endorses a solid stream and one condemns it in capitals.

On what to use instead, we want to be careful about the status of what we are relaying. The lists of water spray, fog, dry chemical, carbon dioxide, foam and sand are the suitable-media entries on producers’ safety data sheets. They are not a statement that those agents will put out a carbon bed, and the regulator document above gives no extinguishing guidance at all. The distinction matters because of what the next paragraph establishes: a burning carbon bed is a deep-seated fire, and dry chemical, carbon dioxide and sand are surface agents. Applying one and seeing no flame is not the same as having put the fire out. What the published investigation of a real adsorber fire — Hoyle, Astbury and Chen, in the Institution of Chemical Engineers Symposium Series No. 141 of 1997 — actually supports is isolation rather than surface extinguishment — and the response to a carbon bed fire is a matter for a competent person working to the producer’s own sheet and your site’s emergency plan, not for a paragraph on a supplier’s website.

Which brings us to the points that decide whether anyone gets hurt after the flames are gone. Carbon is difficult to ignite and tends to burn slowly, smouldering without producing smoke or flame, and smouldering hotspots within a bed can persist for a long time after a fire. Three producers independently warn that smouldering carbon in an enclosed space can build carbon monoxide past its lower explosive limit of 12.5 percent in air. And the investigation found that reintroducing air caused rapid reignition, while stopping airflow or substituting nitrogen halted the exothermic activity — which is the clearest published indication that cutting off the oxygen supply, not spraying the surface, is what actually arrests it. “The fire is out” is the dangerous judgement with this material, and closing up a vessel on that judgement is the dangerous action.

Keep it dry and pre-wet the carbon are both published instructions, for different situations, and conflating them is how someone floods a silo. Producer storage rules say store dry, cool, well ventilated, in the original closed container, away from ignition sources and away from strong oxidisers. The EPA alert recommends pre-wetting as a heat-control measure for an operating vapour-phase adsorber bed. Those are storage and operation respectively. Note also that the producers give two different reasons for keeping it dry — wet carbon is the oxygen-depletion hazard, and wet-then-dry cycling releases heat — and that “it spoils the product” is not the stated reason in any producer document we read.

Dust, incompatibility and the card that is not for this product

On incompatibility, the honest position is narrower than what gets written. Every producer safety data sheet says to keep activated carbon away from strong oxidising agents, and Cabot adds that it may react exothermically with them and should not be stored together with them. But almost none of them names a chemical. The United States chemical reactivity database names only exotic oxidisers — fluorine, ammonium perchlorate, the halogen fluorides, potassium peroxide. One producer, Arq Solutions, names sulphuric and nitric acid. The only named, authoritative pairing we found that is relevant to a water treatment plant is from the United States Environmental Protection Agency’s office of water: powdered activated carbon should not be mixed with potassium permanganate.

What we did not find is any source naming hypochlorite, hydrogen peroxide or ozone as an incompatibility with activated carbon — not the reactivity database, not the EPA document, which discusses calcium hypochlorite separately without pairing it with carbon, and not any producer sheet. Those are the three oxidisers a water treatment plant actually stores, so the absence is awkward, and it would be easy to write a confident sentence here that no source supports. We are not going to. The generic instruction is real and it is what the producers publish: segregate carbon from your oxidiser store. Treat that as the rule, and do not treat the absence of a named pairing as permission.

On dust, two things. No substance-specific occupational exposure limit for activated carbon was set by any of the bodies we checked — and Calgon Carbon says the same on its own safety data sheet, that no specific exposure limits have been established for activated carbon by the American or Canadian occupational bodies or by ACGIH. We checked those and the Indian Factories Act Schedule and no further, so read that as the scope of our search rather than as a worldwide negative. The American occupational limit that applies, 15 milligrams per cubic metre total and 5 respirable, is the limit for particulates not otherwise regulated, and it applies because nothing else does. NIOSH publishes no recommended limit and ACGIH publishes no threshold limit value for this substance. Where a producer sheet prints a tighter figure attributed to a named authority, that is a producer figure and we could not verify it against the authority.

Carbon dust is explosible, and the one producer who publishes test data puts it at the weak end of the range. Chemviron publishes, for a steam-activated product, a dust explosion class of St-1 with a Kst of 21 bar metres per second, a minimum ignition energy above one joule, a minimum cloud ignition temperature of 590 degrees Celsius and a minimum explosible concentration above 60 grams per cubic metre. Arq Solutions independently classes it St-1 without publishing the figures. That is a real explosion hazard requiring proper treatment, and it is not a high-energy dust. Note the counterweight: the public reactivity database describes the material as highly flammable with dust explosive on exposure to heat or flame, which reads very differently from a producer measuring a 590-degree ignition temperature and a minimum ignition energy above a joule.

A card that is not for this product. We could not find an international chemical safety card for activated carbon, and say that as a search result rather than a proven negative, because the list we searched truncated. What we did find is the trap: card 0471 is titled CARBON BLACK, a different substance by a different route, and it carries a suspected-carcinogen classification, a nanoparticle note and a polycyclic-aromatic note that do not belong to activated carbon. Compounding it, the CAS number printed on every activated carbon safety data sheet we read is the one for elemental carbon, so a CAS lookup can land on graphite or carbon black. Check what card your supplier’s safety data sheet cites. And one more, from the coal and lignite side: the regulated respirable hazard in those carbons may be crystalline silica, which two producers list in their own exposure-limit sections, and which a nuisance-dust framing hides.

The Indian regulatory position, in the third person

This section states regimes and attributes each duty to the party that carries it. It makes no claim about what Velora holds, and where a position is a search result rather than a proven negative it says so.

No Quality Control Order, and what that does not mean

Activated carbon does not appear in the Department of Chemicals and Petrochemicals register of Quality Control Orders, on either its chemicals list or its petrochemicals list, and IS 877, IS 2752 and IS 8366 appear in neither. The Bureau’s own records show voluntary certification against the activated carbon standards. So on the evidence available there is no mandatory certification, no Standard Mark requirement and no registration requirement for imports, and the sections of the Bureau of Indian Standards Act that bite on goods covered by such an order are not engaged. Two honest limits on that: Quality Control Orders are issued by several ministries and we checked one register, and no gazette sweep was run.

And the absence of an order is not an absence of obligation. The regime that actually bites on this product is transport, below, and it runs on a self-heating test rather than on a certification scheme. A page describing activated carbon as unregulated because there is no Quality Control Order would be badly misleading.

Classification and tax

Activated carbon is tariff item 3802 10 00, inside heading 3802, which covers activated carbon, activated natural mineral products and animal black including spent animal black. India does not split the activated carbon line further, so one code covers every feedstock, every form, virgin or reactivated, steam or chemically activated. Import policy for every line in the heading is free, with no policy condition, in the schedule edition we were able to read.

Goods and services tax is 18 percent, and the entry is worth citing precisely because the citation has moved even though the rate has not. The rate sits at Schedule II, serial 89 of notification 9 of 2025, Central Tax (Rate), dated 17 September 2025, effective 22 September 2025, which supersedes the 2017 notification that most specification sheets still cite. The entry is at heading level and its description is the whole heading text, so activated carbon, activated alumina, activated bauxite, activated bleaching earth and animal black all fall in the same entry at the same rate. There is no residual entry picking up part of the heading at a different rate, and the September 2025 restructuring did not change the rate here. The 2025 notification has itself since been amended twice — by notification 19 of 2025 of 31 December 2025 and by notification 1 of 2026 of 30 April 2026, the latter in force from 1 May 2026 — and neither amendment touches this entry — so the citation to give is the 2025 notification as amended, and a reader checking this after our publication date should look for a third.

Two deliberate gaps. We publish no basic customs duty rate. Every authoritative route to it was closed to us this week and the only figure available was on a trade mirror whose own latest duty row is dated 2019. Check your bill of entry. And the tariff schedule we read for the policy position and the eight-digit tree is an older edition; the codes are very unlikely to have moved and we did not confirm it against the current one. Separately, note the classification boundaries, because getting them wrong creates two exposures at once: carbon black is heading 2803, and non-activated charcoal including coconut-shell charcoal is heading 4402 in Chapter 44 — which is the same line the transport rules draw, so a misdeclaration is simultaneously a customs and a dangerous-goods problem.

Anti-dumping, stated as what the record shows

The Directorate General of Trade Remedies case database records no anti-dumping investigation into activated carbon. A search of that database on the word returns no case; a search on “carbon” returns a set of cases none of which is this product. We state it that way, attributed to the database, rather than as an unqualified historical negative, because the database may not reach the earliest cases and we did not check safeguard or countervailing duties separately.

Two confusions are worth heading off, because both are real measures on products that are not this one. India has imposed anti-dumping duty on carbon black used in rubber applications, which is heading 2803 and a different product entirely, and that is almost certainly the source of any belief that activated carbon carries an Indian duty. And there are measures on activated carbon in other jurisdictions: a United States order, which is live, and a European measure on powdered activated carbon from China, which was repealed in 2014 and is not in force — though that the European measure was form-specific is still instructive on a page whose argument is that form decides everything. Writing that activated carbon is subject to anti-dumping duty without naming the jurisdiction imports a foreign measure into an Indian buyer’s file.

Transport, which is the regime that actually bites

Activated carbon is UN 1362, proper shipping name CARBON, ACTIVATED, Class 4.2, packing group III, the class for substances liable to spontaneous combustion. There is one UN number for both activation routes; what differs is the evidence the exemption demands.

We have not read the Code text itself. The wording of the current provision above is taken from a carrier advisory reproducing it, and an industry guideline we read reproduces the superseded provision, under which steam-activated carbon was excluded outright with no certificate named. So treat the shipper certificate as the current position on good but second-hand authority, and read the operative text, or have your forwarder read it, before you organise a shipment around it. The same industry guideline recommends a six-month validity for the laboratory test certificate; that is a recommendation from that document rather than a rule in the Code.

The provision that most safety data sheets cite for this has been repealed. The international maritime code’s Amendment 42-24, mandatory from 1 January 2026, deleted the old special provisions from the activated carbon and charcoal entries and substituted new ones, adding a documentation requirement that the exempting certificate be referred to in the special provision and submitted with the cargo information. A safety data sheet citing the old provision number is citing a dead provision, and we have read a major producer’s current sheet that does exactly that — and which also contradicts itself between two of its own sections on whether the product is classified for transport at all. The substantive exemption did not begin with the new provision; it was restated and tightened by it.

The exemptions differ by mode, and this is where a sample shipment goes wrong. The sea provision and the road and rail provision both accept a certificate for steam-activated material. The air regime’s special provision requires a self-heating test report for both types, so air-freighting a sample on the strength of the sea rule leaves the shipper uncovered. On Indian roads the Central Motor Vehicles Rules carry the dangerous goods scheme, and it keys off class labels — two of its rules are about nothing else, and the driver training syllabus ties those labels to the UN classification. The duty is split between two parties by separate rules, one on the consignor and one on the transporter or owner of the goods carriage, so who carries which obligation on a given movement is a matter of the contract and should be settled there.

On who actually produces the evidence, keep the legal position and the commercial reality apart. The provision names the shipper as the party whose certificate exempts steam-activated material. But the shipper is certifying a fact about how the carbon was manufactured, which only the manufacturer knows — so a trader shipping material it did not make is in practice relying on what the producer told it about the activation route, and should be holding the producer’s own statement of that route. The laboratory limb is cleaner: a negative self-heating test from a laboratory recognised by the competent authority is a precondition to shipping discharged by an accredited third party, not a buyer’s diligence item. Recast as a line on a buyer’s checklist it reads as advisory, and it is not.

The regimes that are silent, and the ones that are not

Activated carbon is not a listed hazardous chemical under the Manufacture, Storage and Import of Hazardous Chemical Rules, 1989. Neither the term nor “charcoal” appears anywhere in the Rules, and it is in none of the threshold schedules. So none of the threshold-triggered machinery — the site notification, the safety report, the on-site emergency plan — is engaged by this chemical in any quantity. That is a statement about the product, not a clearance for a site. Those duties attach to the occupier of a site by reference to all the chemicals it holds, so a warehouse storing activated carbon alongside, say, hydrogen peroxide or caustic soda is assessed on those.

There is no Indian statutory occupational exposure limit for activated carbon. The Second Schedule to the Factories Act, 1948 has no entry for it, and no entry for carbon black, nuisance dust or particulates not otherwise regulated either. The nearest entries sit outside the numbered series, in an unnumbered dust block, and concern coal dust, cement and silica. Two cautions: the published reproductions of that Schedule disagree with each other on its length and highest serial number, so any claim of the form “entry number N of the Second Schedule” is unsafe for this product until an official copy is checked; and state Factories Rules were not checked. India’s silence here is India’s silence, not a finding that there is no dust hazard — the American particulates-not-otherwise-regulated limit does apply in that jurisdiction.

On food use activated carbon is a permitted processing aid, and the entry is worth quoting narrowly. It sits at serial 1 of Table 12, generally permitted processing aids, in the Food Safety and Standards Authority of India’s material on processing aids under the Food Products Standards and Food Additives Regulations, 2011. The functional purpose is printed as adsorbent, decolourizing agent, and the residue level is good manufacturing practice — in a column headed in milligrams per kilogram, so the non-numeric entry is a deliberate choice of control rather than an omission. There is no numeric residue limit and no purity specification for activated carbon anywhere in that instrument.

The product categories are where we have to be careful, because two documents on the Authority’s own site do not say the same thing and we are not going to paper over it. The consolidated appendix to the Regulations gives the categories as sugars, oils and fats, juices and alcoholic beverages — the last of those added by an amendment that came into force on 1 August 2023. The separate active list of processing aids allowed for use by food business operators, published as an annexure under section 16(5) of the Act, carries the same serial, the same functional purpose and the same residue control, but adds non-alcoholic beverages and sugar syrup treatment. Neither document says how it relates to the other, and we could not establish which governs. What we can say is which categories survive either way: the four appear in both, so they are safe to rely on, while the two additional ones are defensible only by citing the section 16(5) active list specifically. If your application is in the additional pair, read the current list yourself before you build a specification on it — and expect the appendix and the active list to be at different stages of consolidation.

Read the allocation carefully, because this is a permission that is easy to attribute to the wrong party. The list is headed as the active list of processing aids allowed for use by food business operators, so it makes the use lawful and the permission belongs to the operator using the material. Nothing in either document says anything about suppliers, or that a supplier may certify or claim the permission. It is not a product clearance and it does not describe the seller’s own position, which is a separate question under the same Act.

For potable water there is a gap worth knowing about. The Indian drinking water standard specifies water quality, not treatment chemicals, and prescribes no treatment chemical at all; the only mention of activated carbon in the Bureau’s product manual for it is as a laboratory reagent for preparing odour-free water. The Indian standard that actually addresses carbon for water treatment is IS 2752’s Type 2, and it is voluntary. We looked for an Indian equivalent of the American potable-water material conformity standard that water tenders routinely ask for, and did not find one — which makes that requirement a contractual one in India rather than a legal one. That is what we looked for and did not find, not a proven negative.

Packs, and spent carbon

On packaged-commodity declarations, a 25 kilogram bag is not “more than 25 kilogram”. The relevant chapter of the Legal Metrology (Packaged Commodities) Rules, 2011 does not apply to packages containing more than 25 kilogram or 25 litre, so a 500 kilogram jumbo bag is outside it on size alone — but a 25 kilogram bag fails that test by exactly nothing, and the separate 50 kilogram line is confined to cement, fertiliser and agricultural farm produce. A 25 kilogram bag therefore escapes the chapter only through the industrial-or-institutional consumer route, which carries three conditions and not one: the package must bear the declaration “not for retail sale”; the goods must be for use by that industry or institution rather than for resale, with the institutional limb additionally requiring use not for commercial or trade purposes; and the purchase must come directly from the manufacturer, an importer or a wholesale dealer. That third condition — the sales channel — is the one everybody forgets, and it makes the exemption a property of the transaction rather than of the buyer. And the current form of those definitions is not the 2011 original: they were substituted by the Legal Metrology (Packaged Commodities) Amendment Rules, 2017, in force from 1 January 2018 — so this is not the rule as it has stood since 2011, nor even since 2017.

On spent carbon, the Indian answer is precise and both of the sweeping versions of it are wrong. Schedule I to the Hazardous and Other Wastes (Management and Transboundary Movement) Rules, 2016 lists spent carbon from the production or formulation of drugs, pharmaceuticals and health care products, and spent carbon or filter medium from the purification of organic compounds and solvents. The term “activated carbon” appears nowhere in that Schedule and there is no generic entry. So spent carbon from water treatment, sugar, edible oil or metal recovery is not a listed Schedule I waste, and would become hazardous only by meeting one of Schedule II’s leachable-concentration, total-threshold or hazard-characteristic tests. Saying “spent activated carbon is hazardous waste” without naming the process is wrong, and so is saying it is not. One discipline we owe this instrument that we have applied to every other on the page: the 2016 Rules have been amended since, and we did not establish whether any amendment touched either of those two entries. Check the current consolidated Schedule I before you rely on the entry numbers.

The underlying principle is one the producers state better than the rules do, and it is the right way to think about a loaded bed. Activated carbon is an adsorbent medium, and the hazard classification of spent material is generally determined by the adsorbate. Whatever your bed took out of the stream is in the carbon when it comes out, and the physical and chemical properties of saturated carbon may differ from the virgin material. Classify the spent carbon by what it adsorbed, and settle who classifies it before the changeout, not after.

The short version

Four questions turn a quotation into a specification, and none of them costs anything at the enquiry stage. By which method, to which edition, and on what basis — because most producer data sheets we read cite no test method for any individual property, so the number arrives with nothing behind it. Which figures are specifications and which are typical values — because the sheets that do distinguish them put density, surface area and pH on the typical side, and a typical value will not support a rejection. What the apparent density is, measured how — because that is the number that converts your tonnage into bed volume and your price per tonne into something comparable. And what the carbon has to do, tested on your own stream — because an isotherm or a small-scale column test on your contaminant answers the question the iodine number explicitly declines to answer.

And one piece of information hiding in plain sight: the absence of a number is information. Which parameters a producer chooses to publish tracks which market the grade is sold into. A grade with no CTC figure, or no iodine number at all, is not a defective data sheet — it is a grade specified for a different job.