Less speculative stock does not automatically mean less impact

Pigment printing can remove a wet fixation stage, but the wording of that advantage must remain at stage level. The technical review describes pigment fixation as a dry process operated only by dry heat.[S01] That statement supports a distinction in fixation chemistry and operation. It does not establish zero water use across fabric preparation, printer cleaning, garment manufacture or use, and it does not establish zero energy demand.

Heat is part of the account rather than a footnote to it. IFC guidance identifies drying and curing operations, together with activities involving wet treatments, as particularly significant for heat consumption in textile manufacturing.[S17] A comparison that credits a digital route with avoided washing while ignoring curing would therefore miss a physical burden that the source tells us to examine. The same caution applies to pretreatment drying, transfer pressing and any thermal process used to restore or finish a rejected article.

GFT would record electricity and fuel at the process stage, not only at the utility bill. The logging plan should separate printer operation, pretreatment, fixation or curing, drying, ventilation used for process control, compressed air where relevant, and shared ancillary loads. Startup, idle operation, warming, cleaning and maintenance periods belong to the observed production system. Excluding them would make intermittent order-triggered production look artificially efficient.

Allocation should follow physical causation where practical. A curing batch shared by several orders needs a rule based on occupied capacity, processing time, garment mass or another measured driver. The analyst should document part-load operation rather than assuming that a full-load energy intensity applies to every small consignment. If only a whole-line meter exists, the result remains a line-level observation and should not be presented as a precise garment measurement.

The material ledger must accompany the energy ledger. It should record pretreatment chemicals, inks, binders, cleaning agents, purge and flushing waste, transfer media where used, rejected prints and remake inputs. These records reveal whether lower water demand coincides with higher chemical, material or thermal demand. They also provide the quantities needed for later characterisation without claiming that every kilogram of input has the same environmental effect.

IFC guidance calls for environmental monitoring to cover normal and unsettled conditions, with representative frequency, trained personnel and calibrated, maintained equipment.[S17] For the proposed service, this means retaining failed startups, interrupted cures, cleaning events and reprints in the observation window. A tidy week of successful orders is not a representative baseline if the operating route regularly includes recovery work.

The engineering decision should examine trade-offs directly. A route with less process water may require more fixation energy or pretreatment. A small batch may avoid setup materials but operate a thermal stage at low occupancy. Neither direction should be assumed. The measured result can favour one route on one impact category and another route elsewhere, without forcing a single environmental winner.

Order-by-order production changes more than the timing of printing. It can change consignment size, packaging, transport frequency, approval courier movements, return handling and the number of physical attempts needed to retain an order. These are measurable operating events. They should appear in the same cohort ledger as fabric and energy, rather than being dismissed as downstream exceptions.

The on-demand wool-knitting LCA illustrates why denominator choice matters. Its abstract reports cradle-to-gate comparisons, a cradle-to-grave calculation per wear and a revenue-normalised comparison of on-demand knitting with mass production.[S25] Those different views can answer different questions. The study concerns Merino wool jumpers, knitting and recycling, so none of its numerical effects should be transferred to printed cotton garments in Cameroon. GFT uses it only to support the practice of declaring the functional unit and denominator.

A retained-order account should include the original consignment, any replacement consignment and the reverse movement of a returned article. The evidence can come from carrier records, packaging issue records, delivery distances and actual vehicle or service categories where available. If those data are missing, a scenario must say so. A nominal local-delivery factor borrowed from another market would not convert the estimate into an observation.

Remakes deserve a complete physical record. The remake consumes another set of materials and process time even when the customer pays nothing additional. The original rejected garment also remains a physical output requiring a destination. The ledger should distinguish internal rejection before dispatch, customer return after dispatch and a quality failure discovered during use. These events can have different logistics and end-of-life consequences.

Exhibit 46 maps the decision boundary. It is a system map, not a quantified Sankey: every arrow needs evidence before it can carry an impact value.

The map also prevents a false stock credit. Lower finished-garment stock may coincide with more blank stock, more frequent consumable deliveries or more returned articles. The decision should compare total physical movements and final destinations across routes. Only then can the analyst test whether less speculative stock corresponds to lower impact within the stated boundary.

Personalisation changes product specification; circularity concerns what happens to materials, products and resource demand over time. The Ellen MacArthur Foundation states that circular business models must decouple revenue from production and resource use, and warns that such models do not always deliver environmental benefits.[S20] A business selling additional customised new garments has not demonstrated that decoupling merely because each garment responds to an order.

UNEP presents on-demand production as one option for addressing overproduction, alongside better demand and stock management. It also states that overconsumption has no clear, agreed definition and must be considered in social as well as functional context.[S19] This supports a measured hypothesis, not a universal claim. Producing against an order may reduce unsold stock while still encouraging additional purchases, rapid novelty or short use.

Emotional attachment is similarly a testable proposition. A buyer may value a chosen motif or individual fit, but that preference does not establish longer use. The pilot can record repeat wear only through an appropriate and consented method, alongside observed durability, repair events, resale or handover, and reasons for disposal. Self-reported intention at checkout should not be treated as realised service life.

GFT would evaluate personalisation against three separate questions. Does the garment remain physically usable? Does the wearer continue to use it? Does that use displace another purchase? A positive answer to the first does not prove the second, and the second does not automatically prove the third. This sequence keeps product design, behaviour and counterfactual consumption distinct.

Rebound belongs in the same review. Easier design iteration and shorter release cycles may increase the number of products offered or purchased. The environmental account should therefore observe total garments and total resource use per customer cohort as well as intensity per retained order. A falling intensity can coexist with a rising total when volume grows. Detailed material recirculation routes belong to the circularity programme, but the order data here determines whether circular claims have a credible starting point.