A complaint arrives, say, six months after a cartridge lot shipped: several mouthpieces loosened during filling. The supplier’s release report passed, the buyer’s incoming inspection found nothing unusual, and the production records look normal. What can the team examine now?
An unlabelled bag of spare parts has no controlled evidentiary value.
A useful retain is a traceable, protected and retrievable piece of evidence from a defined production lot. It gives a brand and its hardware supplier something physical to compare with the complaint unit, the approved specification and later production. Its evidentiary value is limited to what the retained units and associated records actually show; it does not establish that every unit in the original lot conformed. Even so, physical evidence gives the team something to examine instead of relying on memory.
For buyers sourcing empty vape cartridges, the program should be agreed before routine orders begin—not improvised after the first field issue.
Give each sample type one job
Teams often call every stored unit a “golden sample.” That collapses several controls that answer different questions.
An approval or golden sample represents an agreed design state. It helps answer: Is this production unit consistent with what we approved? It should be tied to a drawing, bill of materials, color standard and revision.
A limit sample illustrates the boundary of an appearance or workmanship decision. It may show an acceptable color shift, molding witness, printing position or surface mark. It does not replace a written defect catalogue because physical samples fade, become damaged and are interpreted differently.
A lot-retention sample comes from a particular production lot. It helps answer: What can we still observe or test from the lot that produced the reported units?
An AQL inspection sample supports the original accept/hold decision. Units consumed by torque, pull, disassembly or filled-oil testing are not automatically useful retains. The AQL sampling plan and the retained-sample plan may draw from the same lot, but their purposes and custody records should remain distinct. Routine release samples should be issued and consumed separately; designated retains should remain sealed until an authorized investigation or planned review requires them. If tested specimens are worth keeping, identify them as tested material rather than returning them to the untouched-retain inventory.
Start with a lot that can be reconstructed
The sample cannot represent a lot that was never clearly defined. Record the product code and revision, manufacturing site and line when relevant, production dates, purchase order, supplier lot, buyer receipt lot and quantity covered by the release decision.
Avoid silently combining output across a material change, tool repair, line transfer or long production interruption. If units before and after an intervention cannot reasonably be treated as one uniform lot, they should not share one retain identity merely because the cartons left on the same truck.
The selection record should also show where the retained units came from. Pulling the first convenient tray from the top carton gives excellent traceability for that specific carton, but weak coverage across the broader lot. A more defensible plan distributes selections across production time, pallet or carton positions, lines, tools or mold cavities when those variables can affect the failure mode.
A small retain remains too limited to represent every possible defect statistically. Distributed selection reduces avoidable selection bias and preserves useful location information within that limitation.
Decide quantity from future examinations
There is no responsible universal rule such as “keep ten units” or “retain two percent.” Start with the questions the units may need to answer.
List the non-destructive checks first: identity, construction, finish, dimensions, thread fit, center-pin position, resistance and visual comparison. Then list destructive or consumptive work that might be justified: mouthpiece pull, seal disassembly, sectioning, material identification, filled compatibility work or third-party laboratory analysis.
The retained quantity should allow the planned examinations, preserve untouched units for a second review, include a reasoned backup allocation for inconclusive or repeat work, and reflect the lot’s risk and complaint window. Document the rationale instead of applying a universal multiplier. A high-volume repeat order with several critical interfaces may justify a different quantity from a small engineering build. Where the buyer and supplier may each need independent evidence, agree which party retains what and how sealed units can be transferred.
Do not count a unit already damaged during release testing as an untouched retain.
Preserve the condition, not just the part
Empty hardware can change in storage. Elastomers can take compression set, plated surfaces can corrode, adhesives can age, and packaging can abrade finishes. A sample stored loose in a hot, humid cabinet may eventually describe the cabinet better than the production lot.
Keep retains in original saleable packaging when practical, or use packaging demonstrated to be at least as protective for the intended period. Define temperature, humidity, light, cleanliness and handling limits appropriate to the hardware and the investigation purpose. Restrict physical access with a controlled key, badge, locked cabinet or comparably effective measure, and use seals where opening itself must be evident. If the production units shipped with plugs, caps, trays, labels or barrier bags, decide whether those elements form part of the evidence.
Storage conditions need not imitate every distribution excursion. The goal is controlled preservation. Separate shelf-life and transport studies are designed to challenge product and packaging under defined conditions; routine retains should not be accidentally aged by uncontrolled storage.
Make the label survive the memory of the team
Each retain container should connect the physical units to a controlled record. At minimum, capture:
- product name, SKU and revision;
- supplier and manufacturing site;
- supplier lot and buyer receipt or production lot;
- purchase order and quantity represented;
- sampling date, sampler and selection locations;
- initial retained quantity and unit condition;
- packaging and storage location;
- required retention review or disposal date;
- related specification, drawing and release record.
Use a durable identifier on both the container and the electronic record. A filename, spreadsheet row or cabinet label alone is fragile. The system should still work after a buyer changes roles or the supplier reorganizes its quality team.
Photographs document seal condition and initial layout. The investigation still needs retrievable physical units. Likewise, a retained unit without its governing revision can create a misleading comparison after a controlled design change. Link the program to the vape hardware change-control record so obsolete and current retains are not confused.
Control checkout like evidence
Define who can open a retain, why, and how consumption is recorded. A simple checkout log should identify the date, requester, investigation or complaint number, units removed, examinations performed, condition on return and remaining balance.
When opening the container could affect later analysis, photograph the seal and use tamper-evident packaging. If samples move to a supplier or laboratory, record the transfer, carrier, dates, received condition and custody. Chain-of-custody discipline is most valuable when the parties disagree; that is the worst time to discover that nobody knows who opened the bag.
Periodically reconcile the database against the cabinet. A procedure that says samples “shall be retained” is not effective if the stated location is empty or the label no longer matches the contents.
Define the triggers before a complaint arrives
Retains may be opened for a field complaint, an unexpected filling-line trend, a supplier dispute, comparison before and after a change, a recurring incoming defect or a customer or regulatory request. Each trigger should point to an investigation owner and an approved test plan.
Start with the reported failure mechanism. A loose mouthpiece complaint calls for identity confirmation, visual examination, fit and pull-force methods—not an indiscriminate full test panel. Compare the complaint unit, untouched retain, production records and relevant current sample when available.
A retained unit that reproduces the failure is important evidence. A non-reproducing result is evidence of a narrower kind and cannot establish that the complaint was false. The defect may be intermittent, carton-specific, distribution-induced or absent from the small retained set. Combine the observation with lot history, returned-unit condition, complaint rate and process records.
Never use a later retain examination to rewrite the original release data. Record it as a new result with its own date, method, equipment and purpose.
Set a reasoned retention and disposal rule
Retention periods in pharmaceutical or dietary-supplement rules are useful examples of disciplined systems, but they should not be copied as universal legal requirements for empty vape hardware. Define the period from applicable market rules, customer contracts, expected storage and use life, warranty or complaint window, distribution time, supplier agreement and the stability of the sample itself.
State when the clock starts: production, release, receipt, last distribution or another controlled event. Define who approves an extension when an investigation remains open.
Disposal also needs evidence. Confirm that no complaint, hold, legal request or investigation requires continued retention; record the quantity and method; and remove the electronic inventory entry without erasing its history. Unused empty-hardware units may follow the company’s approved scrap or recycling route. Units used in filled compatibility work, and samples containing batteries, residual oil or other controlled materials, should follow the applicable local environmental, health and waste procedures rather than ordinary disposal.
Audit whether the program can answer a real question
Choose a recently distributed lot and ask the team to retrieve its retains without advance notice. Verify the identity, quantity, condition, selection record, governing revision and checkout history. Then simulate a narrow complaint and determine whether enough untouched material remains for the planned work.
The audit often exposes practical failures: duplicate lot codes, missing carton positions, expired limit samples, uncontrolled keys, opened bags with no log, or a retention period shorter than the normal complaint cycle.
A useful program preserves the right evidence, connects it to the right lot, and makes that evidence available while a decision can still be improved.
Before approving repeat production, buyers should align the retain plan with the specification, acceptance sampling, change-control agreement and complaint process. Bring those requirements into cartridge-selection and sample-qualification discussions with ILEVA. The brand remains responsible for its market-specific compliance and release decisions.
