Why does a second source need more than a price comparison?

Because the price is agreed between purchasing and the supplier, while the consequences land on whoever signs the product file. A consumer-facing product carries statements about what is in it, where the ingredients come from and what has been assessed. Those statements were written against a particular material from a particular route. A second source may match the specification line for line and still differ in the impurity panel, in the declared origin or in the supporting documents, and each of those differences has to be checked against a claim somebody already made in writing.

Does a matching certificate of analysis mean the materials are equivalent?

No. A certificate reports the parameters the specification asks for, and a specification is a filter rather than a description. Two lots can pass the same filter while differing in what the filter does not look at: the isomer ratio inside a permitted range, the homologue tail, trace residues characteristic of one route, colour stability over time, odour. Equivalence is a conclusion you reach after comparing the things the specification omits, not a box that a passing certificate ticks.

Who decides whether a change needs notifying anywhere?

The regulatory function responsible for the finished product, working from the rules of each market where it is sold, and from the wording of the file that was submitted or held there. This is not a question a supplier can answer, and it is not a question a general article can answer either, because obligations differ by market and by product category. The practical rule is to raise the change with regulatory affairs before the trial material is ordered, not after a plant batch has been made with it.

How long should a dual-source qualification be expected to take?

Long enough that starting it under supply pressure is the wrong time. The laboratory work is usually the shortest part. The long poles are the documents that have to come from the new supplier's own suppliers, any stability study that has to run to its planned end, and the internal review queue that approves a change to a released product. Treat it as a project with a plan rather than as a purchasing task, and begin it while the incumbent source is still comfortable.

What should be kept once the change is approved?

Keep the comparison itself, not only its conclusion. That means the two specifications side by side, the certificates and the retained samples of the lots actually compared, the analytical work done on both, the documents each supplier provided, the dates, and a record of who approved the change and on what basis. If a question is asked about the product two years later, the file has to be able to say which material was in which batch and why the switch was considered acceptable at the time.

Changing solvent supplier: the regulatory workload

Technical article · Eapearl Chemical · 11 March 2026

Adding or switching the source of a solvent in a consumer-facing product looks like a purchasing decision and behaves like a change-control project. The chemistry is usually the easy part; the file that describes the product is where the work actually falls.

A second source is a change, not a purchase

Most organisations discover this in the wrong order. Purchasing finds a competitive offer, orders a drum, the laboratory confirms it performs, and only then does somebody ask whether the product file still says something true. By that point a decision has been made informally and the review becomes an argument rather than an assessment.

The cleaner sequence treats the material as a change to a released product from the first enquiry. That framing costs nothing early and saves the expensive version of the conversation later. It also changes who is in the room: a supplier change to a consumer product is a discussion between purchasing, the formulator, quality and whoever owns the regulatory file, and it should have an owner and a target date like any other project.

Start from what the file already claims

Before comparing suppliers, read what has already been asserted about the current material. In a typical consumer product file that will include the name used on the ingredient statement, the identifiers attached to it, statements about origin and about animal-derived inputs, a safety assessment that relied on a particular impurity picture, and possibly a declaration about the manufacturing route or the absence of specified substances.

Each of those is a commitment that the incoming material has to keep satisfying. A useful exercise is to write the list out as questions addressed to a supplier, because that list becomes the enquiry document. For a glycol ether such as propylene glycol ethyl ether, the questions that usually matter are the isomer description, the identifiers that correspond to the grade being offered, the impurity panel with its limits and methods, and the documentary statements the finished-product file depends on.

Comparing two lots of nominally the same material

The specification is the wrong place to look for differences, because both suppliers are offering material that meets it. Differences live in the space the specification does not constrain.

Isomer description — a propylene series glycol ether is normally a mixture, and the identifiers that apply depend on whether the grade is described as a single isomer or as the mixture. Fix the description in writing rather than relying on a short trade abbreviation.
The impurity panel — not only whether limits are met, but whether the same substances are being looked for. Two suppliers can run different panels and both report full compliance.
Analytical methods — the same parameter measured by different methods can produce values that are not directly comparable, which matters most for water, acidity and trace residues.
Route-specific residues — a different process leaves a different minor pattern, which is invisible against a pass or fail specification but visible in a chromatogram.
Colour and odour behaviour — both can drift with thermal history in transit and both are noticed by consumers long before any instrument flags them.
Packaging and liner — a change of container is a change of contact material, and it belongs in the comparison rather than in the logistics annex.

Run the comparison on real lots that you have retained, not on data sheets. Where the finished product is sensitive to the solvent’s behaviour rather than merely to its assay, include a sample of the incumbent material of known age, so that the new lot is compared with what you actually use rather than with an idealised version of it.

The bridging pack to request from each supplier

A specification with methods stated for every line, not only limits.
Certificates for the lots supplied for trial, with lot identity that can be traced back.
A safety data sheet valid for the market of use, in the required language.
Statements the finished-product file relies on, such as origin, absence of specified substances, or statements on animal-derived inputs, provided as signed documents rather than as an assurance in an email.
A commitment to notify changes to the process, the site or the specification, with a defined notice period.
Confirmation of the packaging offered and of the liner or closure in contact with the product.

The last two are the ones most often omitted and most often regretted. A supplier who agrees in writing to notify a process change gives you the chance to repeat the comparison before the change reaches your plant; a supplier who does not has effectively transferred that risk to you.

Sequencing the change so nothing is released early

Order the steps so that each gate can still be closed without wasted work. Raise the change internally and agree what evidence will be needed. Obtain documents before samples, because a missing declaration can end the exercise before any laboratory time is spent. Run the laboratory comparison, then a production trial under quarantine, then whatever stability or ageing work the product needs. Approve the change formally, update the specification, the approved supplier list and the file, and only then release material into normal production.

Keep the incumbent qualified. Dual sourcing is only useful if both sources remain approved, and an approval that is never exercised quietly decays as specifications and documents are revised on both sides. A small periodic order against the second source keeps it real and keeps the paperwork current.

Where these projects go wrong

The recurring failures are consistent enough to be listed. Ordering trial material against a short abbreviation rather than a full description, and receiving a neighbouring homologue such as propylene glycol monomethyl ether instead of the intended grade. Treating a passing certificate as proof of equivalence. Starting the qualification only when the incumbent has already announced an interruption. Accepting documentary statements informally. Approving the change but forgetting to update the internal specification, so that goods-in continues to test against the old one. Discarding the retained samples that would have answered the first question anyone asks if a complaint appears later.

None of these is a chemistry problem, which is precisely why they survive technical review. The defence is a written comparison, a named owner and a file that still says something true when the material behind it has changed. Specifications, documents and samples for comparison work are supplied against a named application through our contact page.