Banned Substances Lists and MRSL Compliance for Waterproof Bags: Making a Free-From Claim Defensible

Why a restriction only works when it sits at the input

Nearly every buyer who writes a banned substances list into a specification is thinking about the finished article, and nearly every failure of MRSL compliance happens because the list was applied at the wrong end of the process. The reason is physical: in wet processing, the majority of what is used does not remain in the goods. Scouring agents, leveling agents, carriers, fixatives and most auxiliaries leave with the bath. A product test on the finished fabric will therefore be clean, and the facility will have used chemistry the buyer prohibited, and both statements will be true at the same time.

There is a second reason that is commercial rather than chemical. Product testing is retrospective: it tells you about material that already exists. Input control is prospective: it tells you what will happen next Tuesday, which is the only question that matters when a reorder is on the line. A supplier who has never been asked about inputs will produce a product test within a week, because product tests are cheap and easy to arrange. The same supplier will need a month to produce an inventory with purchase reconciliation, and that delay is the most honest signal in the whole exercise.

The third reason is the one regulators increasingly rely on. Environmental and occupational exposure happens at the facility, not at the customer. A substance can be perfectly safe at the residue level found in a bag and still be unacceptable in the wastewater from the mill that made the fabric. Any framework that only measures the article is structurally blind to the harm it was written to prevent, which is precisely why the manufacturing restricted substances list was invented separately.

None of this means product limits are pointless. They are the legal backstop and they catch the cases where a restricted substance does persist in the article. What it means is that a programme running only product tests has a compliance file that proves nothing about the process and cannot predict the next shipment. The frameworks are published openly and are worth reading rather than summarising: the current list and its conformance guidance sit with the ZDHC Roadmap to Zero programme, and the European substance restrictions are published by the European Chemicals Agency.

How the ZDHC MRSL is structured and what the levels buy you

The manufacturing restricted substances list is organised by chemical family rather than by product, and it applies to formulations used in processing: dyes, pigments, auxiliaries, finishes, coatings, inks, adhesives and cleaners. Conformance is assessed per formulation and graded by the strength of the evidence behind the claim. Understanding the grading is the whole skill, because the number attached to a formulation is a statement about evidence, not about how clean the chemistry is.

• The lowest grade rests on a written statement from the formulator that listed substances are not intentionally used. It is cheap, fast and unverified.
• The next grade adds full formulation disclosure and a safety data sheet reviewed by an accepted third party, which establishes that the paperwork is internally consistent.
• The third grade adds an on-site assessment of how the facility stores, labels, dispenses and records chemicals, which is the first grade that says anything about behaviour.
• The highest grade adds analytical verification of the formulation itself, which is the strongest statement available about that one product.

The single most useful question to ask is not which grade a certificate carries but what proportion of purchased chemical volume is covered by any conformance at all. A supplier with three certificates at the highest grade and forty products in the store has a chemical programme covering perhaps eight per cent of its inputs, and the honest way to report that is as a percentage rather than as a certificate count. Suppliers rarely volunteer the denominator, which is exactly why it should be asked for explicitly.

Two decay mechanisms apply to every grade and both are invisible without a register. Formulations are reformulated, and a certificate issued against a formulation that has since changed now describes a product that no longer exists. Lists are revised, and a certificate issued three years ago was assessed against a list that has since grown. A compliance file with no dates is a file about a process that may no longer be running, and the fix is administrative rather than technical: record the issue date, the list version, and the formulation revision date for every entry.

There is a practical rule of thumb for which grade to demand. For a promotional order with no label claim and no regulated market, a declaration with disclosure behind it is proportionate. For anything sold into a market with enforceable chemical law, or anything carrying a claim on the packaging, the on-site assessment grade is the minimum defensible position, because it is the lowest grade that tests behaviour rather than paperwork. The analytical grade is worth paying for on the specific formulations that carry the highest risk, which for this product category means coatings, inks and adhesives.

The evidence hierarchy, from a bare declaration to a verified formulation

Every compliance claim sits somewhere on a ladder, and the useful discipline is to name the rung rather than accept the claim. The rungs differ in who produced the evidence, whether anyone outside the supplier saw it, and whether it describes a document or a physical reality. Moving up one rung usually costs money; moving up two usually costs time as well.

The ladder explains a pattern that confuses buyers constantly: a supplier can hold strong evidence at the top rung for one formulation and nothing at all for another, and both facts can be true simultaneously. Compliance is not a property of a facility; it is a property of a formulation, on a date, against a list version. Reading a submission as a set of rungs rather than as a single verdict is what makes the file usable.

One rung deserves a caution because it is frequently oversold. Finished-product testing is the rung buyers instinctively trust, and for legal limits it is indispensable, but as evidence about process chemistry it is the weakest rung on the ladder. The specimen is selected by the supplier unless the buyer specifies otherwise, most processing chemistry has already left the article, and the result says nothing about the reorder. Insisting on a random pull from a sealed production carton, with the pull recorded, moves it from marketing evidence to real evidence at no additional laboratory cost.

How specific a free-from claim has to be before it is credible

This is the practical heart of the subject. A statement of the form "this bag does not contain X" can be made at five different levels of precision, and only the top two will survive contact with a retailer’s compliance team, an advertising standards complaint, or a customs enquiry. The test is simple: can the statement be checked against a document that names a formulation, a material and a date. If not, it is a slogan.

The pattern in the right-hand column is always the same, and it is worth memorising because it converts any slogan into a specification in about thirty seconds. Name the substance or the family. Name the components it applies to. Say whether the claim is about intentional addition or about analytical detection. Name the list or the method and its version. Add a date. That is five fields, and a claim missing any one of them can be challenged.

The distinction between intentional addition and detection deserves emphasis because it is the one that decides how a claim behaves over time. "No intentionally added" is a statement about a recipe, and it remains true if nobody changes the recipe — but it says nothing about contamination from shared equipment, from recycled feedstock, or from a previous production run. "Not detected" is a statement about a specimen, and it is stronger for that specimen and weaker for everything else. Serious programmes carry both: the formulation claim for continuity, and periodic analytical checks to catch contamination the recipe cannot see.

Where a claim appears on packaging or a product page, one more discipline applies. The claim has to match the evidence exactly, including the components covered. A hangtag saying PFC-free over a file that only covers the shell fabric is a misrepresentation even if the document is genuine, and the exposure sits with the brand that printed it rather than with the mill that issued it. Our review of PVC-free waterproof materials sets out what the material substitution actually involves when a claim is made real.

The four ways a no-fluorine claim gets quietly diluted

Fluorinated chemistry is the clearest worked example in this category, because the performance stakes are high, the regulation is moving quickly, and the claim is easy to make and hard to verify. Four dilution mechanisms account for almost every misleading no-fluorine statement seen in the market, and all four are detectable by asking one question each.

• Scope dilution: the shell fabric is verified and the rest of the assembly is assumed. Ask which components the evidence names, and compare that list with the bill of materials rather than with the marketing copy.
• Analyte dilution: a targeted screen for a dozen compounds is reported as though it were a total fluorine measurement. Ask which analytes, by which method, and whether total fluorine or total organic fluorine was measured.
• Threshold dilution: a pass means nothing without the number. Ask for the limit applied and the detection limit achieved, since total organic fluorine limits for textile articles in several jurisdictions now sit in the low hundreds of parts per million and tighten on a published schedule.
• Continuity dilution: the evidence covers the first order and the reorder uses a different mill, a different finish or a different applicator. Ask whether the claim is re-verified per production run, and who signs for it.

The scope mechanism deserves a concrete picture, because it is the most common and the least malicious. A fabric mill switches its durable water repellent to a non-fluorinated product and issues a declaration. The bag factory buys that fabric, then adds a zipper whose tape was treated elsewhere, a printed logo applied with an ink from another supplier, and a seam tape from a third mill. Every one of those three additions is a separate chemical decision, and none is covered by the mill’s declaration. The bag is assembled from one verified material and three unverified ones, and the hangtag says PFC-free.

There is a fifth mechanism that is not a dilution but a genuine trap, and it is worth knowing because it catches careful buyers. Non-fluorinated finishes are now widely used, and cross-contamination from shared equipment at a finishing house is a real and documented phenomenon. A mill running both chemistries on the same range can produce a non-fluorinated fabric that measures above a strict total fluorine threshold. The controls are segregation, flush procedures and periodic verification, and the question to ask is whether the finishing range is dedicated or shared.

Finally, the performance side has to be said out loud, because a claim made without it will be discovered in the field. Non-fluorinated repellents have improved substantially but generally remain weaker on oil repellency and on durability to abrasion and laundering than the long-chain chemistry they replaced. The honest framing is a duty-band decision about how much repellency the bag actually needs, which is what our explainer on DWR, PU and silicone coatings works through in material terms.

Which substance families a finished-product test will and will not find

A product test is not a general detector; it finds what it was told to look for, in the specimen it was given, by the method it was told to use. Knowing which families are reliably caught by routine article testing and which are invisible to it is what decides where a programme has to rely on input control instead. The split is not obvious and it surprises people in both directions.

The rows that say "rarely" or "only if requested" are the ones that determine where money should go. A routine article panel is cheap because it covers what everyone tests. The families that matter most in a welded waterproof build — organotins in coatings and films, flame retardants in foam, and the full fluorine family — are mostly outside it. Asking for them costs more per test and is usually the difference between a file that looks complete and one that is.

The metal rows are the exception and the good news: metals in hardware, pigments and plating are reliably detectable and cheap to screen, which makes hardware the most efficient place to spend a testing budget. The reason is that the analyte is an element rather than a molecule, so the test does not have to guess which compound is present. Our guide to plasticiser migration in PVC and the one on flame retardant treatments cover two of the harder families in more depth.

Building your own list rather than inheriting someone else’s

Most brands adopt a list by reference: a retailer’s restricted substances list, a sector initiative’s manufacturing list, or a previous employer’s spreadsheet. That is a reasonable start and an insufficient finish, because every list encodes assumptions about materials and markets that may not be yours. A list written for cotton apparel does not map cleanly onto a welded TPU bag, and a list written for the European market does not cover a North American obligation.

• Start from the markets you actually sell into, and list the legal obligations first. Voluntary lists sit on top of law, never underneath it.
• Add the substance families your construction actually uses. A welded bag has coatings, films, adhesives and inks, which is a different risk profile from a sewn cotton garment.
• Name a version and a date for every list you reference. Unversioned references are unenforceable.
• Decide your own position on families where the law is unsettled, such as fluorine, and write it down rather than leaving it to whoever is answering an enquiry that week.
• State the limit and the method for each entry, because a list without methods is a wish.

The versioning discipline is worth a paragraph on its own because it is the cheapest control in the entire subject and the one most often skipped. Restricted substance lists are revised on a published cycle, sometimes twice a year, and the candidate list of substances of very high concern in the European Union has grown to several hundred entries over time. A contract that says "compliant with the current restricted substances list" without naming a version is a contract that can be satisfied by whatever the supplier had on file, and the buyer discovers the gap at the worst possible moment.

There is a scale argument for a short, enforced list over a long, aspirational one. A list of forty entries that every supplier can evidence is worth more than a list of four hundred that nobody can, because the value of a list is in its enforcement rather than in its length. The pragmatic construction is a short mandatory core, universally evidenced, plus a longer watch list reviewed annually and tested on a rotating basis. Programmes that do this tend to find real issues, because the rotating tests are looking at something new each cycle rather than re-confirming the same six parameters forever.

The three documents that must agree with each other

Any chemical claim can be reduced to a small set of documents, and the decisive test is not whether each document looks convincing in isolation but whether they agree with one another. Three documents carry almost all the evidentiary weight, and a discrepancy between any two of them is more informative than a missing third.

The pairing column is where the work is, and it is work a buyer can do in an afternoon without any technical background. Take one month of purchase invoices for processing chemicals and compare the trade names and quantities against the inventory. Products bought but not listed are the hidden part of the chemical programme. Then take the conformance certificates and check that every trade name on them appears in the inventory, which establishes whether the certified chemistry is the chemistry on the line. Neither check requires a laboratory.

Discrepancies are normal and should not be treated as bad faith. Inventories lag behind purchases, certificates are issued for products that were later discontinued, and formulations are renamed. What matters is whether the supplier can explain the discrepancy quickly and specifically, because that demonstrates the records are maintained rather than written. A supplier who cannot reconcile a purchase to an inventory entry in a day is telling you the inventory is a document for auditors, and that single finding is worth more than any certificate in the file.

Wastewater data is the useful fourth document, and it earns its place because it cannot be staged in the way a curated inventory can. Composite sampling across a full production shift integrates everything that went down the drain during that window, including products that were never listed. It is not a substitute for the other three, but where it disagrees with them, the disagreement is the most informative thing in the file. The water-side interface is covered in our guide to water use and discharge in production.

Purchase records: turning a declaration into a checkable statement

A declaration is a sentence someone wrote. A purchase record is an event that happened. The reconciliation between the two is the single cheapest and most under-used control in chemical compliance, and it requires no laboratory, no auditor and no chemistry. It requires only that the supplier be willing to show invoices, which is a much lower bar than any other form of evidence.

• Ask for one month of purchase records for processing chemicals, and match trade names and quantities to the inventory.
• Ask which products are bought in the largest volume, and whether those specific products are the ones carrying conformance evidence. Volume-weighted coverage is the honest metric.
• Ask whether laboratory, maintenance, printing and cleaning chemicals are included. Excluding them is the most common way an inventory under-reports.
• Ask for the safety data sheet revision date for each entry. A sheet older than three years usually means the formulation predates the current list.
• Repeat the reconciliation annually. A one-off reconciliation proves the records existed once; a repeating one proves they are maintained.

The volume-weighting point changes how a submission should be read, and it is the question suppliers least want to answer. Ten certificates covering ten per cent of purchased volume looks better in an email than two certificates covering ninety per cent, and the first version wins most comparisons because certificate counts are what get reported. Asking for the percentage of purchased volume covered converts a beauty contest into a measurement, and it usually changes the ranking of suppliers.

There is a contractual version of this that costs nothing and does a lot of work: require the supplier to notify the buyer before introducing any new processing chemical into the product’s process. That notification is the trigger for a declaration or a screen, and it converts input control from an annual audit into a continuous one. Most facilities already have a new-product approval step for other reasons, so the ask is usually to add a line rather than to build a process.

Testing as a check on the system, not a substitute for one

Product testing is where buyers feel safest and where compliance files are weakest, because a test result is concrete while a system is abstract. The correct relationship is the reverse of the intuitive one: testing is what you do to check that the input controls are working, not what you do instead of them. A programme that tests but does not control inputs is running a lottery with extra paperwork.

• Test to verify a system, on a schedule: a rotating panel across the bill of materials catches drift that a fixed annual panel misses.
• Specify the sampling. A random pull from a sealed production carton, recorded with photographs, is materially stronger evidence than a submitted sample.
• Specify the components and the shades. A test on the shell says nothing about the print, and a test on a light shade says nothing about a dark one.
• Treat every test result as a sample of one, and record it against a lot reference so that a trend can be seen.
• Escalate rather than repeat. Two consecutive borderline results on the same component are a signal about a supplier, not about a batch.

The rotation idea deserves expansion because it is the highest-value change most programmes can make. Testing the same six parameters every year produces a file that confirms the same six things and discovers nothing. Rotating the panel so that each cycle covers a different slice of the bill of materials, and a different set of substance families, means that over three years the whole product has been looked at rather than one corner of it repeatedly. The total spend is identical; the information is far greater.

There is also a legal reason to keep testing regardless of how strong the input controls are. Several regimes place the obligation on the importer or the brand, and that obligation is discharged by testing against a defined method, not by holding supplier declarations. The declarations determine whether the product is likely to pass; the testing is what the file contains if anybody asks. Both are needed, and the sequencing is set out in our guide to REACH and CPSIA compliance testing.

How a substitution quietly voids a compliance file

Most chemical compliance failures in established programmes are not caused by a bad original specification. They are caused by a substitution made during a reorder, when a material was unavailable, a price moved, or a supplier changed, and an equivalent was accepted. The equivalence was never verified, nobody recorded it, and the file now describes a product that is no longer being made.

• The material swap: a fabric, film or tape from a different mill with the same nominal specification and different chemistry.
• The colourway swap: a new shade introduced without re-testing, which matters most for dye-related parameters.
• The component swap: hardware or trim bought from a different source, which is the highest-risk substitution for metals.
• The process swap: a coating applied by a different applicator, or a weld parameter changed, which alters residues and heat history.
• The silent swap: a supplier reformulates a product and keeps the trade name, so nothing on the paperwork appears to have changed.

The fifth one is the genuinely difficult case and the reason continuous controls beat periodic audits. A reformulation under an unchanged trade name defeats any check that compares names, which is what most document review does. The defensible responses are periodic analytical verification of high-risk formulations and a contractual requirement to notify formulation changes, and neither is foolproof. What they do is create a record that the buyer asked, which matters if the question is ever litigated.

The control is one clause and it should be in every purchase order: no material, component, colourway or process may be substituted without re-submission of the relevant evidence and, where the change affects a joint, re-validation of the welding or bonding parameters. It costs nothing to write. It converts a silent substitution into a decision somebody has to make and, in most cases, that is enough to prevent it happening by default. The same clause appears in our guide to contract terms buyers should insist on, where it sits alongside the commercial protections.

What compliance costs and where the money actually lands

Chemical compliance costs are diffuse, which is why they are resented, and they land in three identifiable places. Material premiums come first: screened or certified fabrics, tapes, inks and hardware cost more than commodity equivalents, and the premium is proportionally larger at low volumes because the upstream mill applies its own minimum rather than yours. Testing comes second and scales with the breadth of the panel rather than with the size of the order. Administration comes third, and it is the cost most often skipped, which is why files decay.

The administration line deserves defending, because it is the one that gets cut first and causes the most damage downstream. Somebody has to collect declarations, chase renewals, record expiry dates, maintain the inventory reconciliation and file test reports against lot references. That is a few hours a month for a small programme and a part-time role for a large one, and skipping it produces a file that is two years out of date and a scramble before every shipment.

Against that sits the cost of an unsubstantiated claim, which is a different order of magnitude entirely. A retailer that audits a claim and finds the evidence missing can delist, and several markets have advertising regulators that act on environmental claims with real penalties. A recall or a withdrawal costs more than a decade of compliance administration. The asymmetry is so large that the compliance spend is best understood as insurance rather than as overhead, and programmes that frame it that way stop arguing about it.

There is one cost that is genuinely optional and worth naming so buyers can decide consciously: the analytical verification of individual formulations. It is the strongest evidence available and it is not cheap, and for a low-volume promotional order with no label claim it is hard to justify. For a product that will carry an environmental claim on the packaging, or that will be sold by a retailer with its own compliance audit, it is the evidence that makes the claim survivable. Deciding which products fall into which category is a real decision, and it should be made before the first quotation.

Turning the list into contract language that survives a reorder

Everything in this guide collapses into a short set of lines that belong on the tech pack and the purchase order. A requirement that lives in an email is a preference, and preferences are renegotiated under time pressure. Written into the order, the same requirement becomes a condition of payment, which is a different thing entirely.

• Name the lists and their versions: which manufacturing list, which product limit set, which edition, and which date.
• Name the scope: every component in the bill of materials, including tape, print, foam, thread, labels and hardware, with no exclusions left to assumption.
• Name the evidence and its timing: declarations before bulk, test reports before shipment, and whether testing is per batch, per colourway or per season.
• State the coverage metric: the proportion of purchased chemical volume covered by conformance, reported as a percentage.
• State the substitution clause, including re-validation of weld parameters where a material change affects a joint.
• State the claim discipline: no free-from statement on product, packaging or listing unless the evidence names that material and that analyte set.
• State record retention and the right to audit, including the option of a semi-announced visit.

The coverage metric line is the unusual one, and it is the line that most improves the quality of what comes back. Suppliers respond to what is asked for. Ask for certificates and you receive a folder. Ask for the percentage of purchased volume covered and you receive a number, which can be compared between suppliers and tracked over time, and which cannot be improved by adding certificates for products nobody uses.

The claim discipline line is the one that protects the brand rather than the product. A claim on a hangtag is a representation the brand must be able to substantiate, and the substantiation must match the claim in scope. Requiring that the evidence name the material and the analyte set before any claim is printed is the cheapest possible protection against a misrepresentation that would otherwise be discovered by a competitor, a regulator or a journalist rather than by the brand.

If you want this applied to a specific programme, send the bill of materials, the markets you sell into and any claim you intend to print, and the list and evidence pack can be built against those three before sampling begins. You can see how a programme moves from first enquiry through sampling into bulk production; every style starts at 500 pieces minimum, with samples in 6–10 working days and bulk in 35–50 days, quoted FOB Xiamen.

Frequently Asked Questions

Q1. What is a banned substances list?

A purchasing rule naming the chemical formulations that may not be bought or used in production, regardless of whether any residue ends up in the finished article. It controls inputs rather than testing outputs.

Q2. How is an MRSL different from an RSL?

A manufacturing restricted substances list governs what may be used inside the facility. A product restricted substances list governs what may remain in the article. The same substance can be legal in the product and banned from the process.

Q3. Why does restriction have to happen at the input?

Because most processing chemistry leaves with the wash water rather than staying in the article. A finished-product test will be clean while prohibited chemistry was used, and both facts can be true simultaneously.

Q4. What do ZDHC conformance levels mean?

They grade the strength of the evidence behind a formulation claim, from a written statement through third-party document review, on-site system assessment and laboratory verification of the chemistry itself.

Q5. Is a certificate count a good measure of compliance?

No. Ask what percentage of purchased chemical volume is covered by conformance. Certificates are issued per formulation, and a facility uses dozens, so a count without a denominator is nearly meaningless.

Q6. How specific does a PFC-free claim need to be?

It should name the substance family, the components covered, whether the claim is about intentional addition or detection, the method or list and its version, and the date. Anything less can be challenged.

Q7. What is the difference between no intentionally added and not detected?

The first is a statement about a recipe and holds while the recipe is unchanged. The second is a statement about a specimen and says nothing about the next one. Serious programmes carry both.

Q8. Why does a PFC-free claim often not cover the whole bag?

Because the evidence usually comes from the fabric mill and covers the shell finish, while the zipper tape, print, seam tape and webbing are separate chemical decisions made by separate suppliers.

Q9. Does a passing product test prove chemical compliance?

No. It proves a specimen passed. Without input control, a programme is testing outputs and hoping, and a specimen selected by the supplier is weaker evidence than a random pull from a sealed carton.

Q10. Which substance families are usually missed by routine testing?

Organotins in coatings and films, flame retardants in foam, and most of the fluorine family. A routine panel covers what everyone tests, which is not the same as what matters in a welded build.

Q11. How do I check a supplier declaration without a laboratory?

Reconcile one month of chemical purchase invoices against the inventory list, then check that conformance certificate trade names appear in that inventory. Discrepancies are more informative than missing documents.

Q12. Why is wastewater data useful?

Because composite sampling across a production shift integrates everything that went down the drain, including products that were never listed. It is much harder to stage than a curated inventory.

Q13. What is the biggest cause of compliance failure in established programmes?

Substitution during a reorder, when a material, colourway, component or applicator changes and the evidence on file is never updated. A substitution clause in the purchase order prevents most of it.

Q14. How often should restricted substance lists be reviewed?

At least annually, and preferably whenever the referenced list is revised. Candidate lists of substances of very high concern grow on a published cycle, so an unversioned contract reference is unenforceable.

Q15. Should I build my own list or adopt one?

Adopt as a starting point and then modify. Lists encode assumptions about materials and markets; one written for cotton apparel does not map cleanly onto a welded coated bag sold into a different jurisdiction.

Q16. What does chemical compliance actually cost?

It lands in material premiums, in testing breadth rather than order size, and in the administration of keeping declarations and expiry dates current. The administration is the line most often cut and the one that causes files to decay.

Q17. Is analytical verification of formulations worth paying for?

For a low-volume promotional order with no label claim, usually not. For anything carrying an environmental claim or sold by a retailer with its own audit, it is the evidence that makes the claim survivable.

People Also Ask

What is a banned substances list?

A purchasing rule naming chemical formulations that may not be used in production, controlling inputs rather than testing the finished article.

How does an MRSL differ from an RSL?

An MRSL restricts what a facility may use during processing; an RSL restricts what may remain in the finished product.

What do ZDHC conformance levels prove?

The strength of the evidence behind a formulation claim, from a self-declaration up to laboratory verification of the chemistry.

How specific should a free-from claim be?

It should name the substance, the components covered, the method or list version, whether it covers intentional addition or detection, and the date.

Why can a PFC-free claim be misleading?

Because it usually covers only the shell finish, while the zipper tape, print, seam tape and webbing are separate chemical decisions.

What causes most chemical compliance failures?

Substitution during a reorder, where a material or component changes and the evidence on file is never updated.