Caustic soda flakes or lye: when a plant should buy which, and the freight arithmetic

The answer, up front

1. One Indian Standard covers both forms. IS 252 : 2013, Caustic Soda — Specification (Fourth Revision), supplies the material either as lye or as a solid in the form of flakes, blocks, sticks or pellets. There is no separate Indian Standard for flakes.
2. Flakes are not the purer form, and they are not the dirtier one either. On the standard’s own table the two forms carry the same sodium hydroxide minimum on a dry basis, and the solid’s impurity ceilings look twice as loose only because they are written against material that is twice as concentrated. Normalise them and nine of the ten match to within about three percent. One row is a genuine difference, and most buyers have never looked at it.
3. The 97 and the 99.5 you see quoted for flakes are the same material on two different bases. Comparing one producer’s dry-basis figure against another’s as-received figure is the commonest arithmetic mistake in this market.
4. The freight question is real, and it is a ratio, not a rate. At the standard’s own floors, moving one tonne of sodium hydroxide as Grade 1 lye means moving about 2.13 tonnes of liquid; as solid it means moving about 1.03 tonnes. That is about 2.06 times the mass on the road for the same delivered chemical.
5. And the ratio does not decide it. The solid saves freight and then charges for it at the plant: a dissolving tank, agitation, an exotherm, an airborne exposure limit low enough that the recommended figures are written as ceilings, and an assay that drifts downward in storage. Lye saves all of that and charges for heating, metallurgy and tankage instead.

This is a buyer’s guide to the two commercial forms of caustic soda, not legal advice, not a specification and not a design document. Velora Chemicals is a bulk supplier and trader sourcing from qualified manufacturers, not a manufacturer. Nothing here states or implies that Velora holds any licence, registration or consent. Every standard and instrument named below should be read in its current form before any of its numbers goes into a purchase order or a plant design.

One standard, two forms, and an amendment that changed the answer

The current edition is IS 252 : 2013, Caustic Soda — Specification, Fourth Revision. It carries two amendments, Amendment 1 of 2015 and Amendment 2 of 2018, and the second of those matters more than the base text for anyone writing a purchase order.

Clause 3.1 of the 2013 text is where the two forms meet: the material is supplied as lye, or as a solid in the form of flakes, blocks, sticks or pellets, and shall be free from foreign matter, dirt and other visible impurities. Flakes are a solid form named inside IS 252, not a product with a standard of its own. So a purchase order that says “flakes to IS 252” is citing the right document; one that hunts for a separate flake standard is looking for something that is not there.

What Amendment 2 did is the part most commonly missed. As published in 2013, IS 252 defined no grades at all — the foreword records that the technical grade of the 1991 edition was deleted, on the stated ground that Indian production is now entirely by the membrane cell process, leaving a single unlabelled set of requirements rather than the Pure and Technical pair. Amendment 2 of 2018 inserted clause 3.1.1 and created two lye grades, and in this standard the grade is the concentration:

• Lye Grade 1 — sodium hydroxide 47 percent by mass, minimum.
• Lye Grade 2 — sodium hydroxide 30 percent by mass, minimum.
• The solid form has no grades. It has one set of requirements.

Two things follow. First, a buyer in India who says “Grade 1” is naming a strength, which is not how the word grade works in the international trade, where a producer may offer one concentration in several grades distinguished by cell technology and salt content rather than by strength. Second, anyone still writing a purchase order against “IS 252 technical grade” is citing the 1991 edition, and specifying something the current standard does not contain.

The 1991 edition was titled Caustic Soda, Pure and Technical. The current title is simply Caustic Soda — Specification. If a tender document carries the old title or the old grade names, it is working from a superseded edition — and freely circulating copies of the 1991 text are easy to find. IS 252 also amalgamated IS 1021 : 1964 at an earlier revision, so a reference to IS 1021 is dead as well.

The table that answers the purity question, and it answers it the other way round

The belief that flakes are the purer form is almost universal among buyers, and on the Indian standard it is wrong. Here is Table 1 of IS 252 : 2013 as amended in 2018, for the three specified products.

Read the second row first. Lye Grade 1 and the solid carry the identical sodium hydroxide minimum on a dry basis: 99.50 percent. Once the water is taken out of the comparison, the standard asks exactly the same thing of both forms. The 47 and the 97 in the first row are statements about how much water is in the package, not about how good the chemical is.

Then read down the rest, and resist the obvious conclusion. Nine of the ten remaining characteristics are set at exactly twice the lye Grade 1 figure for the solid — carbonate 0.40 against 0.20, chloride 0.03 against 0.015, sulphate 0.030 against 0.015, silicate 0.02 against 0.01, iron 20 ppm against 10, copper 2 against 1, manganese 1 against 0.5, insolubles 0.05 against 0.025, mercury 0.02 ppm against 0.01. It is tempting to read that as the solid being allowed to be twice as dirty. It is not, and the reason is in the first row.

Every one of those limits is a percentage of the material as received, and the solid is 2.06 times as concentrated as Grade 1 lye. So doubling the limit while the concentration rather more than doubles leaves the two forms carrying almost exactly the same impurity per tonne of sodium hydroxide delivered. The arithmetic is the same for all nine rows, because the ratio is the same in all nine: a factor of 2 in the ceiling against a factor of 2.064 in the concentration. Per tonne of sodium hydroxide the solid is about 3 percent tighter than Grade 1 lye on every one of those nine characteristics — iron, for instance, works out at about 20.6 grams per tonne of sodium hydroxide for the solid against about 21.3 for Grade 1 lye. That is a difference of no commercial consequence whatever. Whatever the committee intended, the effect is that the standard has not made one form cleaner than the other on those nine rows. We do not know the drafting rationale and we are not going to invent one.

One row is a real difference, and it is the one nobody looks at. Chlorates and perchlorates are capped at 10 ppm in the solid against 40 ppm in either lye grade — four times tighter in absolute terms, and, once normalised for concentration, about eight times tighter per tonne of sodium hydroxide. For a buyer whose process is chlorate-sensitive, that single row is the only line in the table that tells the two forms apart, and it points at the solid.

And the normalisation puts Grade 2 lye in an unflattering light. Grade 2 carries the identical absolute ceilings to Grade 1 while being specified at 30 percent instead of 47 percent. So per tonne of sodium hydroxide delivered, Grade 2 is permitted about 1.57 times the carbonate, chloride, sulphate, silicate, iron, copper, manganese, insolubles and mercury of Grade 1 — and about 13 times the chlorates of the solid. On this table Grade 2 is the dirtiest of the three products per unit of chemical, not merely the most dilute.

The normalised figures in the three paragraphs above are our arithmetic on the table’s own cells, not statements in the standard. Divide each ceiling by the sodium hydroxide fraction of that product and compare. We set it out because the unnormalised comparison is the one everybody makes, and it points the wrong way. What we are not saying is that the two forms perform identically in a given process — only that IS 252 does not make one of them cleaner.

What these figures are and are not. They are Table 1 of IS 252 : 2013 as amended by Amendment 2 of 2018, read from a full copy of the standard and its amendments and corroborated against the Bureau of Indian Standards product summary page and one Indian producer’s published specification. We did not establish whether a third amendment or a fifth revision has been issued since 2018. Confirm the current text with the Bureau before writing any of these numbers into a contract. Where a standard has been made mandatory by a Quality Control Order — as caustic soda has — BIS operates a view-only facility for reading it free of charge, so this is a cheap check rather than an expensive one.

The 97 and the 99.5: the same flake quoted two ways

This is the single most expensive misreading in the caustic soda market, and the standard itself is the reason it happens. IS 252 reports sodium hydroxide twice for the same material: once as received, and once on a dry basis. For the solid those two numbers are 97 percent minimum as received and 99.50 percent minimum on a dry basis.

Indian producers then publish on whichever basis they prefer. One major producer publishes its flakes at 99.50 percent minimum on a dry basis and its prills at 99.40 percent minimum on a dry basis; another publishes the same class of product as minimum 97 percent by weight; a third titles its product sheet at 98 percent. Those are not three qualities of flake. At least two of the three are quoting different rows of the same table for the same product. A buyer who ranks suppliers by that figure is ranking their choice of basis.

The conversion is defined, not guessed. Clause A-14 of IS 252 sets out the dry-basis calculation: add the determined percentages of sodium hydroxide, carbonate as Na2CO3, chloride as Cl, sulphate as SO4 and silicate as SiO2, treat that sum as the dry solid content, and express the sodium hydroxide as a share of it. The standard works an example: components totalling 50.4 percent with sodium hydroxide at 50.2 percent gives a dry-basis figure of about 99.6 percent.

Note that the two rows are cumulative, not interchangeable. A conforming solid has to meet both: 97 percent as received and 99.50 percent on the dry basis. They are calibrated close together, and on our arithmetic they do not fall the same way for the two forms. A solid at exactly 97.0 percent as received, carrying every listed impurity at its ceiling, computes to about 99.51 percent on the dry basis — just inside the 99.50 minimum. Grade 1 lye at exactly 47.0 percent, carrying every listed impurity at its ceiling, computes to about 99.49 percent, which is just outside it. If that arithmetic is right, then for Grade 1 lye at the floor the dry-basis row is the binding one, and it caps the listed impurities a little below where the individual ceilings sit. We flag it rather than lean on it: it is our computation from the table’s own cells, the margin is a few hundredths of a point, and we do not know the standard’s rounding convention. The buyer’s point does not depend on it: the two rows are separate requirements, and neither can be converted into the other without the impurity determinations. So a supplier quoting 99.5 and a supplier quoting 97 may be offering the identical specification, and may not be. Ask which row, and ask for the other one.

So the first question to ask of any caustic soda assay figure is which basis it is on. Two more basis traps sit alongside it. Producer density tables carry a percent Na2O column next to the percent NaOH column, and the two are far apart — one published table gives 48 percent NaOH as 37.20 percent Na2O — so lifting the wrong column understates the sodium hydroxide by 22.5 percent, which is simply the ratio of the formula weights and is therefore always the same. And IS 252 : 2013 expresses chloride as Cl where the 1991 edition expressed it as NaCl, and sulphate as SO4 where the older edition used Na2SO4, so a figure from an older document is not comparable with a figure from the current one.

The freight arithmetic

Now the part that people actually mean when they ask which form to buy. The percentages in IS 252 and in producer specifications are percent by mass, so the arithmetic is direct and there is nothing clever about it.

Work in tonnes of sodium hydroxide delivered, which is the only unit in which the two forms are comparable at all. At the standard’s own minima:

So at the Grade 1 floor, about 2.06 times as much material travels for the same delivered chemical — 2.13 tonnes against 1.03, of which 1.13 tonnes is water and the small specified impurities rather than sodium hydroxide, so a little over half of a Grade 1 lye consignment by mass is not the chemical you are buying. At the Grade 2 floor it is 3.23 times. That is the whole of the freight argument for flakes, and on a long haul it is not a small argument.

Two cautions about how far that figure can be pushed. First, the ratio is a ratio and not a cost. We publish no rupee-per-tonne or rupee-per-tonne-kilometre figures here, because there is no primary source for them: road freight for corrosive liquid in bulk is privately negotiated, and the constraint on how much goes on one vehicle that we were able to source is the statutory ceiling on axle weights and gross vehicle weight notified under the Motor Vehicles Act, alongside the vehicle’s own permit and registered laden weight. What that ceiling means commercially is that a part load does not get cheaper in proportion — the vehicle is the unit, so per-tonne freight on an underfilled tanker rises against the fill fraction. A buyer with the delivered rates for both forms in front of them can do this arithmetic properly in ten minutes; a buyer working from a rate per tonne of material cannot do it at all.

Second, the mass ratio is the start of the comparison and not the end of it. The 97 percent is also the as-received floor on the day the bag is filled, and for the solid the assay moves in one direction afterwards. The rest of this article is about the costs that sit on each side of the ratio.

If the tanker is measured by volume rather than weighed, do not convert it yourself. Amendment 2 of 2018 replaced clause 3.2 of IS 252 so that the density composition table in IS 4016 governs. IS 4016 : 2002 correlates density with composition for aqueous sodium hydroxide, in kilograms per cubic metre at solution temperatures from 10 to 40 degrees Celsius in five-degree steps, and gives both the mass percentage and the grams of sodium hydroxide per litre at temperature. Density is strongly temperature dependent, and a volume taken at one temperature against a table read at another is a commercial error, not a rounding error. The equivalent producer figures for 50 percent solution are of the order of 765 to 768 grams of sodium hydroxide per litre at about 15.6 degrees Celsius, which is roughly 1.31 cubic metres of solution per tonne of sodium hydroxide; use them to sanity-check an invoice, not to replace the table the standard points at.

What lye costs at the plant

Strong caustic soda solution is not a difficult material, but it is not an undemanding one either, and two of its demands are expensive to retrofit.

It crystallises at an ambient temperature

This is the constraint that catches plants out, because the number is not low. Producers of 50 percent solution publish it in their own handbooks and data sheets, and they do not quite agree:

• One United States producer’s handbook states that 50 percent liquid caustic soda begins to crystallise at 54 degrees Fahrenheit, about 12 degrees Celsius.
• Another states that the 50 percent solution begins to freeze at about 60 degrees Fahrenheit, which is about 15.6 degrees Celsius and which that handbook rounds to 15.
• A European producer’s data sheet says 50 percent already begins to solidify at approximately 10 to 12 degrees Celsius, and separately recommends heatable storage tanks and storage above 25 degrees Celsius.
• An Indian producer’s safety data sheet for 45 to 50 percent lye gives a melting point of 12 degrees Celsius.

Note that every one of them says begins to. This is not a sharp freezing point but the onset of crystallisation against a hydrate phase boundary, which is why the honest form of the figure is a range of roughly 10 to 16 degrees Celsius for 50 percent solution, attributed to the producer who published it, rather than a single number. Whether that matters at a given site depends on the site’s own ambient minimum and on how exposed the tank, the pipework and the unloading line are — which the buyer knows and we do not.

A lower concentration crystallises lower, which is one of the real reasons a buyer might prefer 47 percent over 50 percent rather than simply buying the strongest available. We found no primary source stating the figure for 48 percent or for 47 percent, so we give that as a direction and not as a number.

And crystallisation is not the only cold-weather problem. One producer states that the viscosity of 50 percent caustic increases rapidly below 65 degrees Fahrenheit, about 18 degrees Celsius; another publishes it numerically for 50 percent solution at 79 millipascal seconds at 20 degrees Celsius, falling to 44 at 30 degrees and 25 at 40 degrees. Pumps, strainers and unloading hoses can be defeated well above the temperature at which anything crystallises. IS 4264 : 1967, the code of safety for caustic soda that IS 252 points at, says the same in its own words: pipelines exposed to low temperatures should be arranged so that they can be heated to prevent the solution crystallising.

Hot concentrated caustic attacks the tank and then contaminates the product

The metallurgy has a ceiling, and it is lower than most people expect. From the published producer handbooks:

• The ideal storage temperature for caustic soda solutions is given as 80 to 100 degrees Fahrenheit, roughly 27 to 38 degrees Celsius.
• Above 120 degrees Fahrenheit, about 49 degrees Celsius, one handbook states that there is accelerated corrosion of carbon steel and iron contamination of the caustic, and that cracking can occur if concentrated caustic is processed in steel equipment that has not been stress relieved. A second producer puts the same threshold the other way round: annealing to prevent stress corrosion cracking is not required unless temperatures above 120 degrees Fahrenheit are expected.
• Above that, the published ladder is 316L and 304L stainless steel to 200 degrees Fahrenheit, then nickel, with Monel, Inconel or Hastelloy also used; epoxy linings are given an upper limit of about 225 degrees Fahrenheit.
• Steam tracing on piping is described as not normally recommended, because caustic in traced piping can readily exceed 140 degrees Fahrenheit under static conditions, which is above the 120 degree Fahrenheit carbon steel threshold set out two bullets above.

The 120 degree Fahrenheit threshold is worth reading twice, because it protects the product as well as the vessel. A buyer who has specified lye Grade 1 for its 10 ppm iron ceiling, and then stores it hot in unlined carbon steel, has arranged for the tank to put the iron back.

The incompatibility list is also longer than the usual warning about aluminium. The published lists between them cover aluminium, tin, lead, zinc, copper and their alloys including brass and bronze, with glass and enamel added by one producer; the international chemical safety card notes that the solution corrodes aluminium, tin, lead and zinc producing flammable hydrogen gas, and that it reacts with ammonium salts to form ammonia. IS 4264 : 1967 adds that caustic soda also slowly corrodes iron, copper and Monel metal, so the alloy ladder is a ranking rather than a permission. Two specific prohibitions from one producer’s handbook are easy to breach by accident because a plant would naturally fit both: sight glasses should never be used on 50 percent storage tanks or transfer lines, and valves should never be installed in the vent line.

A chart of the caustic stress corrosion cracking zones — temperature against concentration, with regions labelled for no stress relief, stress relief of welds and bends, and consideration of nickel alloys — is published in producer literature, and in one case is sourced to a corrosion data survey of 1971. We could not read the numeric boundaries of those zones, and we therefore publish none of them. The two 120 degree Fahrenheit statements above are text, not chart, and they are the figures to work from — for concentrated caustic, which is the case they are written for. A plant design decision here belongs to a materials engineer with the current chart in front of them, not to a buying note.

What flakes cost at the plant

The solid avoids every paragraph above. It has its own list, and it is not shorter.

Somebody has to dissolve it, and the dissolving step is the hazard

The rule is unambiguous and it is stated to the same effect by an international chemical safety card, by an Indian producer’s safety data sheet, by two producer handbooks and by IS 4264 : 1967: add the caustic slowly to water, with agitation, and never add water to the caustic. One producer adds the two qualifiers that matter most — the water should be cold, and rapid addition to unagitated water is dangerous, with the risk greater in warm water.

And the hazard is not that it gets warm. IS 4264 : 1967 states that considerable heat is generated on adding water to solid caustic soda or its strong solutions, and that flakes should be added to a well-agitated tank of water, bulk addition producing rapid heat with local boiling and spurting. The United States emergency response data for sodium hydroxide solution puts it as dilution releasing heat possibly enough to cause local boiling and spattering; one producer handbook warns of dangerous mists; the international safety card warns that the heat generated on contact with moisture or water can ignite combustible material. The thing that leaves the tank is hot caustic and aerosol.

How much heat, for the dilution case, is quantified consistently by two independent producers working the same example: diluting 50 percent solution at 120 degrees Fahrenheit with water at 80 degrees Fahrenheit down to 20 percent lands at approximately 142 to 143 degrees Fahrenheit, about 61 to 62 degrees Celsius. One of the two also publishes a heat of dilution table in British thermal units per pound, which stops at 48 percent. For dissolving anhydrous solid, every primary source we found says only that the reaction is exothermic or that considerable heat is generated, with no figure, so we publish no figure for it.

Two practical details from the 1967 Indian code that a plant will otherwise learn the hard way. Where solid caustic has been supplied in a steel drum, the drum should be peeled off the cake before dissolving, to avoid iron contamination — the same contamination a buyer may have changed form to avoid. And flakes spilled on the floor should be swept up carefully and the area then hosed down, not hosed first.

One more arithmetic point, because it defeats batch sheets: volumes are not additive. A producer handbook’s own worked example takes 3,000 gallons of 50 percent solution and 6,917 gallons of water and notes that the final volume is slightly less than the sum. Size a dissolving tank on mass, then check the volume, not the other way round.

The dust limits are low, and the recommended ones are ceilings

This is the exposure route that pumped lye largely does not have. The published occupational limits for sodium hydroxide are low, and the two recommended ones are expressed as ceilings: the United States NIOSH recommended limit is a ceiling of 2 milligrams per cubic metre, the enforceable OSHA permissible limit is an eight-hour time-weighted average of 2 milligrams per cubic metre, the international chemical safety card gives a threshold limit value of 2 milligrams per cubic metre as a ceiling value and notes that occupational exposure limits must not be exceeded at any time, and the NIOSH immediately-dangerous level is 10 milligrams per cubic metre. The pocket guide lists pneumonitis among the symptoms alongside eye and skin burns, and prescribes N100, R100 or P100 filters for the air-purifying options up to that level. Note the asymmetry: the two recommended limits are ceilings, not to be exceeded at any time, while the enforceable American one is an eight-hour average.

And India has a statutory figure of its own, at the same level. The Second Schedule to the Factories Act, 1948, which section 41-F makes the schedule of maximum permissible threshold limit values of chemical substances in the work environment, lists sodium hydroxide at 2 milligrams per cubic metre, in the eight-hour time-weighted average column. That is the number that binds an Indian factory, and the American and international figures above are quoted only because they carry the supporting detail on filters, symptoms and the immediately dangerous level.

One detail we could not settle. The sodium hydroxide entry in that Second Schedule is printed with a letter suffix, and the reproductions available to us do not define it — one of them carries a legend defining a different letter as denoting a ceiling limit, and defines nothing for the one against sodium hydroxide. So we state the 2 milligrams per cubic metre, and we do not state whether the Indian limit is formally a ceiling or an average. The two reproductions we read also disagree on the serial number, which is why none is given. Read the bare Act. The practical point is unaffected: the Indian statutory figure and the recommended international ones are at the same level, and it is a low one.

The safety card is explicit about why the solid is the problem: harmful concentrations of airborne particles can build up quickly when the material is dispersed as dust, and the prevention line is to prevent dust dispersion. IS 4264 : 1967 carries a clause that applies to the flake buyer and not to the lye buyer: among the personal protective equipment for personnel handling caustic soda, it requires respirators or protective hoods when handling flakes, granules or powder — in addition to the goggles, elbow-length rubber or PVC gloves, boots, apron and alkali-resistant hat it requires for caustic soda generally, and the safety showers and eye-washing facilities it requires to be readily available.

A caution on comparing solid forms: one producer publishes its spherical bead product as dust-free with a very low dust rating, and states in the same bulletin that loose dust surrounds flake forms, that the beads cake less, and that they dissolve about twice as fast as flake. Those are claims for beads, and they are the producer’s own. Bead and prill data should not be used to describe flakes, and Indian producers do sell flakes and prills as separate products with separate specifications.

The assay drifts, and it drifts one way

Solid caustic soda is described by the international safety card as a white hygroscopic solid that absorbs moisture from the air, to be kept in the original container, dry and tightly closed; one producer calls it hygroscopic and deliquescent, meaning it absorbs atmospheric moisture until it dissolves in it. Separately, it reacts with carbon dioxide in air to form sodium carbonate: a producer bulletin says so directly, an Indian producer’s sheet says containers must be tightly closed to prevent the conversion of sodium hydroxide to sodium carbonate by the carbon dioxide in air, and the United States emergency response entry for the solution notes that it slowly absorbs carbon dioxide from the air to give solid products as crusts or precipitates.

Both mechanisms push the as-received assay down and the carbonate figure up, and both are ordinary consequences of a bag being opened in a humid shed. The commercial consequence is that the 97 percent is a day-one floor, and a buyer who computed sodium hydroxide per rupee from the label and never revisited it is working from a number the material does not stay at. We found no quantified drift rate in any primary source and we are not going to estimate one.

What we did find, on either side: one producer rotates its own bead inventory first-in first-out so that all product is consumed within 180 days — that is one producer’s guidance for its own bead product in its own packaging, not a general figure and not an Indian one. Against that, a European producer states that properly stored 50 percent solution has a virtually unlimited shelf life, while diluted solutions should be stored for as short a time as possible in cool conditions. On the storage question the two forms are not close.

The packaging the standard calls for bears on exactly this. Clause 5.1 of IS 252, as replaced by Amendment 2 of 2018, requires the solid in laminated HDPE bags with polyethylene liners, sealed or tied against moisture, or in steel drums, or as agreed between purchaser and supplier; lye goes in tankers or tank cars. The standard specifies the bag type and never a bag weight. The 25 and 50 kilogram packs that everyone quotes are trade practice with no Indian Standard behind them, so if a pack size matters to a plant, it is a line in the purchase order and not an assumption.

What a buyer can actually demand, and where the standard stops

IS 252 is a product specification, and a specification binds nobody of itself. What binds, for caustic soda, is the prohibition in the Bureau of Indian Standards Act read with the 2018 Order, which is the subject of the next section — and it is not a thing a buyer complies with. A buyer specifies it and then verifies. Three levers are worth knowing about, and two of them are easy to miss.

• Marking. Clause 5.2 as amended requires the material name and description, the manufacturer’s name, the net mass and the lot or batch number, with the caution CORROSIVE-HANDLE WITH CARE and a corrosion hazard label to IS 1260 (Part 1).
• A certified test report per tanker, for lye. Amendment 2 of 2018 requires, for lye tankers specifically, a certified test report showing the grade, the manufacturer, the date of supply, the tanker number and the quantity. It sits in the amendment and not in the 2013 base text, which is one reason a buyer working from the base text would never know to ask for it. It is the most useful single thing in the amendment and it is the lye route’s advantage on documentation.
• Strength is contractual. Clause 3.2 of the 2013 text made relative density subject to agreement between purchaser and supplier; Amendment 2 replaced it so that the IS 4016 density composition table applies. That fixed the method, not the value. Beyond the Grade 1 and Grade 2 floors, the standard will not set the strength of your lye — the purchase order has to.

And since the grade gives a minimum assay and a set of maximum impurity ceilings, neither is a target and neither is a band. If a process needs an upper limit on strength as well as a lower one — and in dosing and dilution an over-strength delivery can be as unwelcome as an under-strength one — that limit has to be written in, because the standard does not supply it. We have set that argument out at length elsewhere in this series.

An edition warning that matters more than most, and it is about a safety document. Clause 4 of IS 252 : 2013 points at IS 4264 for handling and storage, and the edition available to us, and the edition every clause reference in this article is taken from, is IS 4264 : 1967, Code of safety for caustic soda. The Bureau of Indian Standards now records that edition as withdrawn and superseded by IS 4264 : 2023, Caustic Soda — Code of Safety, published 22 November 2023. We have not read the 2023 edition. So every clause of IS 4264 cited in this article should be read as the 1967 text and re-checked against the current code before it is relied on, and the 1967 copy available to us carries visible scanning damage besides. Where the same point is also made by a producer handbook or by the international chemical safety card, we have said so, and those are the sources to lean on. This is worth generalising: a dated reference inside a current standard is a reference to the edition named, not to the current edition of the referenced document.

The regulatory position, in the third person

Four things apply to caustic soda in India that do not apply to every bulk chemical, and each of them attaches to a particular party. None of what follows is a statement about Velora.

Caustic soda is under a Quality Control Order, and both forms are inside it

The instrument is the Bureau of Indian Standard (Caustic Soda) Order, 2018, notified as S.O. 1468(E) of 3 April 2018 by the Department of Chemicals and Petrochemicals under sub-sections (1) and (2) of section 16 of the Bureau of Indian Standards Act, 2016. It is not titled a Quality Control Order, which is how it is almost always described, and getting the title right is a reasonable test of whether anybody has opened it. It requires caustic soda to conform to IS 252 : 2013 and to bear the Standard Mark under a licence from the Bureau. Its only exemption is for goods meant for export that conform to the specification required by the foreign buyer.

The Order’s entry names caustic soda with no qualifier as to form, and the Bureau’s own product manual for IS 252 scopes a licence as covering solid, and lye in Grade 1 and/or Grade 2. So flakes and lye are both within it.

The prohibition itself is in section 17(1) of the BIS Act, 2016, and the drafting matters: it is written against no person, and the acts it prohibits are to manufacture, import, distribute, sell, hire, lease, store or exhibit for sale goods covered by a section 16 order without the Standard Mark under a valid licence, or bearing the Mark where the goods do not conform. Two opposite errors are common here. One is to say that a Quality Control Order is the manufacturer’s problem and does not concern anyone downstream — section 17(1) catches distributing, selling and storing. The other is to read it as creating a licence for traders: the licence under the scheme is the manufacturer’s, and nothing in S.O. 1468(E) or in sections 16 and 17 creates a separate licence or registration for a reseller. The accurate statement is about the goods: caustic soda sold or stored in India is required to conform to the standard and to carry a valid Standard Mark, and the penalty provision in section 29(3) is drafted against any person.

We could not establish the amendment history of S.O. 1468(E). We found no amending, extending, suspending or superseding order, and the department’s own page for the Order lists none — but we could not run a full-text gazette search, and the department’s published Quality Control Order list is itself stale, with its latest notification entry from August 2022. Absence of evidence is not evidence of absence, and we do not state that the Order has never been amended. Note also that the Order names the 2013 edition of IS 252 specifically, which is a detail to watch if the standard is ever revised.

Caustic soda is a hazardous chemical under the 1989 rules, but it carries no threshold quantity

This one is routinely got wrong in both directions, including by us. Sodium hydroxide is listed by name in Schedule 1, Part II of the Manufacture, Storage and Import of Hazardous Chemical Rules, 1989, at serial number 571. Rule 2(e) defines a hazardous chemical as a chemical that either satisfies the criteria in Part I of Schedule 1 or is listed in column 2 of Part II. Those are alternative routes, and the named list is the one that catches caustic soda.

That correction matters because Part I’s criteria are acute toxicity, flammability and explosives, and corrosivity is not among them — reasoning which is correct as far as it goes and which we have relied on elsewhere in this series, and which does not settle the question, because the Part II list is independent of it. Caustic soda is a hazardous chemical under those rules.

What it does not have is a threshold quantity. Sodium hydroxide is not in Schedule 2 and not in Schedule 3, so the quantity-triggered machinery — advance site notification, safety reports, on-site and off-site emergency plans — is not engaged by holding caustic soda, however much of it is held. What remain are the duties that do not depend on a threshold, and the significant one points at the buyer rather than at the supplier. The rule on collection, development and dissemination of information applies to an industrial activity in which a Schedule 1 chemical is or may be involved, and it requires the occupier who has control of that activity to arrange to obtain or develop a safety data sheet, to ensure the information is accurate and reflects the scientific evidence behind the hazard determination, to update it as significant new hazard information becomes available, and to keep it available on request. That is the receiving site’s own duty, not something discharged by the sheet that arrives with the drum. There is a separate general duty on an occupier during an industrial activity to identify major accident hazards and take adequate steps. And the container labelling requirement — contents, the name and address of the manufacturer or importer, and the physical, chemical and toxicological data — is drafted against the container and names no party at all, which is worth noticing before either side assumes the other owes it.

Nor is a storage-only site obviously outside this. The definition of an industrial activity in those rules extends beyond operations in a listed installation to isolated storage and to pipelines. Whether a given site is inside it is a question about that site, and the answer is not supplied by the fact that caustic soda carries no threshold quantity.

We read those rules in a commercial consolidation rather than in the gazette, and the schedules rather than a clean text of every rule. So we give the duties in outline, quote none of them, and cite no sub-rule number. Check the allocation against the consolidated text published by the environment ministry before acting on this paragraph — and note that we corrected it once already in drafting, in the opposite direction, which is a fair indication of how easy it is to get the wrong way round.

Both forms are 18 percent GST, and the September 2025 rationalisation did not touch them

Solid sodium hydroxide and sodium hydroxide in aqueous solution sit in different customs subheadings — 2815 11 for the solid and 2815 12 for the aqueous solution — and a buyer who infers from that that they carry different tax is wrong. Both fall to 18 percent, through the residual entry for inorganic chemicals in Schedule II of the rate notification of 17 September 2025, which took effect on 22 September 2025 and supersedes the 2017 notification that advisory pages still cite. There is no flake-versus-lye distinction anywhere in the rate notification.

The September 2025 rate cuts did not include caustic soda. The Chapter 28 reductions announced then were sulphuric acid, nitric acid and ammonia for the fertiliser sector, from 18 percent to 5 percent, together with some health-sector and micronutrient items. The published press release of the 57th GST Council meeting of 8 October 2026 records no recommendation touching inorganic chemicals either — and in any case a Council recommendation takes legal effect only through the notifications that follow it. Because the 18 percent arrives through a residual entry rather than a named one, it is worth re-checking rather than assumed if a targeted entry is ever created.

On imports, the commercially useful fact is a documented decision rather than a duty. A Directorate General of Trade Remedies final finding of 16 December 2021 recommended anti-dumping duty on caustic soda from Japan, Iran, Qatar and Oman — expressly covering the product in all forms, lye and solids, and expressly rejecting a request to exclude flakes — and the Department of Revenue decided not to accept that recommendation, by an office memorandum of 23 February 2022. We found no anti-dumping duty on caustic soda in force, and we could not open a consolidated measures-in-force list to prove the negative, so we put it that way round. We state no basic customs duty rate, because we could not reach the tariff schedule on an official source.

Road movement: who owes what

Caustic soda in either form moves as a corrosive, and the duties under the dangerous-goods provisions of the Central Motor Vehicles Rules, 1989 fall on named parties. Nothing we verified turns on whether the cargo is solid or liquid — the regime is keyed to the goods being of a dangerous or hazardous nature and to the applicable class label, not to physical state. In outline, and attributing each duty to the party who carries it:

Separately, rule 9 of those rules requires a person driving a goods carriage carrying goods of a dangerous or hazardous nature to hold a transport-vehicle licence, to be able to read and write an Indian language and English, and to hold a certificate from a three-day course at an institute recognised by the State Government, with the authorisation endorsed on the driving licence. The course syllabus set out in that rule covers defensive and advanced driving alongside product safety and emergency procedures.

Two honest limits on the paragraph above. We could not open the gazetted text of the dangerous-goods rules themselves; the duty allocation is taken from two State transport department publications of those rules, which agree on substance and differ on a detail. We therefore quote none of it verbatim and state no dimension or number of panel faces. And a buyer asking the consignor’s questions does not discharge the consignor’s duty, nor does a seller discharge the transporter’s. Who the consignor is on any given movement is a matter of the contract, and it should be settled in the contract rather than assumed from the delivery term.

Two smaller differences between a bag and a tanker

• Packaged commodity declarations, where the 25 kilogram line is not where people think it is. Chapter II of the Legal Metrology (Packaged Commodities) Rules, 2011 does not apply to packages containing more than 25 kilograms or 25 litres, nor to packaged commodities meant for industrial or institutional consumers. A 50 kilogram bag is over that line. A 25 kilogram bag is not — 25 is not more than 25 — so a 25 kilogram pack is outside Chapter II only through the industrial or institutional consumer route. Those definitions carry three conditions, and the one usually forgotten is the first. The buyer must have bought directly from the manufacturer, from an importer or from a wholesale dealer; the purchase must be for use by that industry or institution; and the package must bear the declaration not for retail sale. For an institutional consumer there is a fourth: not for commercial or trade purposes. The channel condition is the one that bites on a resale, and it is worth checking against the current rule text rather than taken from this summary. A tanker load is not a retail package at all, and the quantity question on it becomes a contractual and weighbridge matter instead.
• Bulk storage at the buyer’s site. We looked for an Indian rule that requires a pollution control board consent purely for storing bulk lye, and for a central specification for bunding or secondary containment around a caustic tank, and we found neither. Consent under the water pollution legislation attaches to establishing an industry, operation, process or treatment and disposal system likely to discharge trade effluent, and to outlets and discharges — not to the act of keeping a chemical in a tank. That is the result of our search and not a clearance: a given site’s own consent conditions, factory approvals and layout requirements may well say something about a new bulk tank, and those are documents only the buyer has.

A proposal to amend rule 3 of those packaged commodity rules, which would have removed the size exemption for bags sold in retail, was published for comment in July 2024. We could not establish that it has been notified, and the later amendment we could check touched other rules. So the exemption above should not be described as removed, and should not be relied on as permanent either.

So which form should a plant buy

Reduce the whole article to the cases where each form wins on its own terms.

The solid is usually right when

• Consumption is small or intermittent relative to a tanker load. The vehicle is the freight unit, so a part-filled tanker carries the worst of both arguments.
• The haul from the producer is long. This is where the 2.06 ratio does its work, and it compounds with distance.
• There is no bulk tankage, and no appetite to build it — which means no heated or traced tank, no caustic-rated pumps, no unloading bay and no secondary containment to arrange.
• Dosing is batch-wise anyway, so a dissolving tank is already in the flowsheet or is a small addition to it.
• The site can handle the dust — a ceiling exposure limit, respiratory protection for the people opening bags, and somewhere dry to keep them.
• The process is chlorate-sensitive, where the solid’s 10 ppm ceiling beats either lye grade’s 40 ppm — about eight times tighter per tonne of sodium hydroxide, and the only row in the table that genuinely separates the forms.

Lye is usually right when

• Consumption is continuous or large enough to turn over a tanker load before it matters.
• The tank, the pumps, the metallurgy and the heating already exist, in which case most of lye’s costs are sunk and most of the solid’s are not.
• A closed transfer is worth more to you than the impurity table is — which, on nine of the table’s ten impurity rows and once normalised for concentration, it may well be, since Grade 1 lye and the solid are the same buy on those nine. The chlorate row is the exception, and it points the other way.
• Documentation per consignment matters, where the per-tanker certified test report required by Amendment 2 is a real lever that the bagged route has no equivalent of.
• Dust exposure is unacceptable — a closed transfer from a tanker to a tank removes the exposure route the solid creates.
• The material will sit for a while. One producer publishes a virtually unlimited shelf life for properly stored 50 percent solution; for the solid the mechanisms — moisture pickup and carbonation — run one way only, and we found no quantified drift rate for it.

And the case for Grade 2 lye at 30 percent is narrow, which is worth saying because it is a specified Indian product. It triples the mass on the road per tonne of sodium hydroxide, for the benefit of arriving pre-diluted and crystallising at a lower temperature. That is a short haul argument, or a specific process argument, and not a general one.

The short version

IS 252 : 2013 covers both forms, and flakes are one of its solid forms rather than a separate standard; Amendment 2 of 2018 created lye Grade 1 at 47 percent minimum and Grade 2 at 30 percent minimum, replaced Table 1, and replaced the clause that had left relative density to agreement. On the standard’s own table neither form is the purer one — the dry-basis sodium hydroxide minimum is 99.50 percent for both lye Grade 1 and the solid, and although the solid’s impurity ceilings read twice as loose on nine of the ten other characteristics, they are percentages of material that is 2.06 times as concentrated, so per tonne of sodium hydroxide the two match to within about 3 percent. The one genuine difference is chlorates, about eight times tighter in the solid per tonne of sodium hydroxide, and the one product the normalisation punishes is Grade 2 lye, which carries Grade 1’s absolute ceilings at 30 percent instead of 47. The 97 percent and the 99.50 percent quoted for flakes are two rows of the same table, one as received and one on a dry basis, which a conforming solid has to satisfy together; comparing one supplier’s row against another’s is how buyers mis-rank suppliers. On freight, work in tonnes of sodium hydroxide: at the standard’s floors a tonne of the chemical travels as about 1.03 tonnes of solid or about 2.13 tonnes of Grade 1 lye, which is about 2.06 times the mass, and the vehicle rather than the tonne is the freight unit. Then set against that what each form costs at the plant: lye crystallises somewhere around 10 to 16 degrees Celsius at 50 percent and gets viscous well above that, and hot concentrated caustic both attacks carbon steel and puts iron into the product above about 49 degrees Celsius; flakes need a dissolving tank, cold agitated water, caustic added to water and never the reverse, respiratory protection, and an assay that only falls once the bag is open. Buy on sodium hydroxide delivered, specify the standard, the amendment, the grade and the strength, name the impurity that matters to your process, and ask for the per-tanker test report if you are taking lye.