Overview #
When brand partners ask us to support their sustainability reporting or product carbon footprint declarations, the first question we ask is: have you defined your system boundary? Without that, any LCA number is meaningless — and we have seen brands submit EPDs to retailers with scope boundaries so narrow they excluded the substrate entirely. This article walks through how we apply the ISO 14040/14044 framework to print and packaging production, which data sources we use at each life cycle stage, and where the critical quality control checkpoints sit in our LCA workflow. It is most relevant to brand owners in the EU, UK, and Australia who face mandatory or retailer-driven carbon disclosure requirements, and to product managers preparing for compliance with the EU Packaging and Packaging Waste Regulation (PPWR) 2025 revision.
ISO 14040/14044 Framework: How We Structure an LCA for a Packaging Job #
Every LCA we conduct follows the four-phase structure mandated by ISO 14040:2006 and ISO 14044:2006 — goal and scope definition, life cycle inventory (LCI), life cycle impact assessment (LCIA), and interpretation. On our production floor, this is not an abstract exercise. It maps directly to the bill of materials, machine run data, and waste logs we already capture for each job.
Phase 1 — Goal and Scope Definition
We define the functional unit as 1,000 finished packaging units delivered to the brand’s nominated warehouse. System boundary is cradle-to-gate as a default, but we can extend to cradle-to-grave if the brand provides end-of-life data for their market. The cut-off criterion we apply follows ISO 14044 Section 5.3.3: any input or output contributing less than 1% of total mass and less than 1% of total energy is excluded, provided cumulative exclusions do not exceed 5%.
Phase 2 — Life Cycle Inventory (LCI): Our Data Hierarchy
This is where production floor data becomes the differentiator. We use a three-tier data hierarchy:
| Data Tier | Source | Typical Uncertainty (±%) | Application in Our LCI |
|---|---|---|---|
| Tier 1 — Primary | Our own meter readings, substrate weight logs, ink consumption records | ±3–8% | Electricity, substrate mass, ink volume per job |
| Tier 2 — Secondary (industry average) | Ecoinvent 3.10, World Food LCA Database, FEFCO LCA data | ±15–25% | Upstream raw material extraction, chemical manufacturing |
| Tier 3 — Proxy | Published EPDs from paper mills (e.g. APP, Sappi, Stora Enso) | ±10–20% | Specific board grades where mill EPD is available |
We prioritise Tier 1 data for all processes within our factory gate. For a typical 350gsm SBS folding carton job, our measured electricity consumption runs 0.18–0.24 kWh per 1,000 sheets on our Heidelberg XL 106 offset press. Ink consumption averages 2.8–4.2 g/m² for a 4-colour process job, which we log by weighing ink fountains before and after each run.
Phase 3 — Life Cycle Impact Assessment (LCIA)
We report against the EN 15804+A2 impact categories as a minimum for packaging EPDs, which aligns with the Product Environmental Footprint (PEF) methodology required under the EU Green Claims Directive. The primary indicator is Global Warming Potential (GWP100) in kg CO₂-equivalent, calculated using IPCC AR6 characterisation factors. For packaging with direct food contact, we additionally report Ozone Depletion Potential (ODP) and Human Toxicity — Cancer (HT-c) to satisfy FDA 21 CFR and EU 10/2011 compliance documentation requirements.
Critical LCI Data Collection: Process Parameters and Quality Checkpoints #
Running a credible LCI requires the same discipline as running a quality-controlled print job. We have defined pass/fail thresholds for every data input, and any value outside tolerance triggers a data review before the LCA model is finalised.
Electricity and Energy
Our factory is sub-metered at machine level. Acceptable data quality threshold: meter reading interval ≤ 30 days, with a maximum data gap of 72 hours before we interpolate. Annual energy intensity for our offset printing hall runs 380–420 kWh per tonne of substrate processed. If a job-level figure falls outside 280–550 kWh/tonne, we flag it for review — it usually indicates a short run with high make-ready waste or an unplanned press stop.
Substrate Mass and Waste
We weigh substrate input and waste separately for every job. Typical make-ready waste on our sheet-fed offset lines is 2.5–4.0% of total substrate input for runs above 5,000 sheets. For runs below 2,000 sheets, make-ready waste can reach 8–12%, which significantly affects the per-unit carbon footprint — a point we always flag to brand partners ordering short-run sampling quantities.
Ink and Coating Consumption
Solvent-based coatings are the highest-uncertainty input in our LCI. We measure coating weight by substrate area using a wet film gauge, targeting 4–6 g/m² for aqueous coatings and 2–3 g/m² for UV coatings. VOC emissions from solvent-based inks are calculated using the mass balance method per GB/T 23985-2009, with a default solvent retention factor of 0.05 (5% retained in dried film).
Process Parameter Table — LCI Data Collection Standards
| Process Input | Measurement Method | Typical Value | Acceptable Range | Out-of-Range Action |
|---|---|---|---|---|
| Electricity (offset press) | Sub-meter, kWh/tonne substrate | 400 kWh/tonne | 280–550 kWh/tonne | Review run log, check press stops |
| Substrate input mass | Calibrated floor scale, ±0.5 kg | Job-specific | ±2% vs. purchase order | Reweigh, check delivery note |
| Make-ready waste (>5k run) | Waste bin weight, % of input | 3.2% | 2.5–4.0% | Investigate if >5%; check plate register |
| Aqueous coating weight | Wet film gauge, g/m² | 5.0 g/m² | 4–6 g/m² | Adjust coater blade gap |
| UV ink cure energy | UV meter, mJ/cm² | 180 mJ/cm² | 120–220 mJ/cm² | Recalibrate UV lamp output |
| Water consumption (CTP/wash) | Flow meter, L/tonne substrate | 320 L/tonne | 200–450 L/tonne | Check wash cycle frequency |
Compliance, Certification, and Third-Party Verification #
An LCA produced for internal decision-making has different requirements from one used in external claims. We distinguish three use cases:
Internal Carbon Accounting — follows ISO 14064-1:2018 for organisational GHG inventory. No third-party verification required, but we apply a ±10% data quality threshold on all primary inputs.
Product-Level EPD — must be verified by an accredited third-party verifier against a relevant Product Category Rule (PCR). For paper and board packaging, the applicable PCR is EN 15804+A2 (construction products EPD standard is not applicable; we use the FEFCO/Pro Carton PCR for cartonboard). Verification adds 6–10 weeks to the LCA timeline and typically costs USD 2,500–5,000 depending on scope.
Retailer or Regulatory Submission — increasingly, major EU and UK retailers require PEF-aligned methodology. The EU PPWR (Regulation 2024/1781) mandates recycled content declarations and, from 2030, minimum recycled content thresholds for paper packaging (65% by weight). Our FSC-certified substrate supply chain (FSC-C[chain of custody number available on request]) supports the upstream data quality required for these submissions.
For brands targeting the US market, we align LCA documentation with the FTC Green Guides (16 CFR Part 260) to ensure carbon claims are substantiated and not misleading — a requirement that has become a litigation risk for brands making unverified “carbon neutral” claims.
Specification Notes for Brand Partners #
When you brief us on an LCA or carbon footprint project, we need the following to scope the work accurately: (1) intended use of the LCA — internal reporting, EPD, retailer submission, or marketing claim; (2) system boundary preference — cradle-to-gate, cradle-to-grave, or cradle-to-cradle; (3) the specific packaging SKUs to be assessed, including substrate grade, print specification, and annual volume; (4) any existing mill EPDs or supplier data sheets you hold for your materials.
The most common mistake we see is brands requesting a “full LCA” without specifying the functional unit. A functional unit of “one box” gives a very different result from “packaging 1,000 units of product” — the latter captures the protective function and reduces the per-unit impact when pack count per pallet is optimised. We guide every brand partner through this decision before we open the LCA model.
Our typical LCA workflow: data collection and modelling in 15–20 working days, internal review and draft report in 5 working days, third-party verification (if required) in 6–10 weeks. We issue a draft LCA summary for brand review before finalising.
Frequently Asked Questions #
Q1: What electricity emission factor do you use for your China factory operations, and how does it affect our product carbon footprint?
A: We use the China Southern Grid emission factor of 0.5271 kg CO₂e/kWh (CEA 2023 published value), applied to our measured 0.18–0.24 kWh per 1,000 sheets on our offset press. For a typical folding carton job, grid electricity contributes 8–15% of total cradle-to-gate GWP — substrate production is almost always the dominant hotspot at 55–70%.
Q2: What is your minimum order quantity for a packaging job that includes a full LCA report?
A: We can conduct an LCA for any production run, but the per-unit carbon footprint becomes highly sensitive to make-ready waste below 2,000 sheets — where waste can reach 8–12% of substrate input versus 2.5–4.0% on standard runs. We recommend LCA-supported projects run at 5,000 units minimum to produce a representative per-unit figure. LCA documentation is included at no additional charge for orders above 50,000 units.
Q3: Does your LCA methodology comply with the EU PPWR and Green Claims Directive requirements?
A: Yes. Our LCIA reporting covers the EN 15804+A2 impact categories required under the EU Product Environmental Footprint (PEF) methodology, which underpins both the Green Claims Directive and the PPWR 2024/1781 framework. For EPD-level submissions, we engage an accredited third-party verifier — the process adds 6–10 weeks and USD 2,500–5,000 to the project scope.
Q4: Can you model different substrate options — for example, virgin SBS versus recycled board — and show the carbon impact difference?
A: Yes, comparative LCA across substrate scenarios is one of the most useful outputs we produce. In our modelling, switching from virgin SBS (GWP approximately 0.85–1.10 kg CO₂e/kg board) to 100% recycled greyboard (GWP approximately 0.45–0.65 kg CO₂e/kg board) typically reduces substrate-stage GWP by 35–45%, though this must be weighed against structural performance changes — recycled board at equivalent caliper is generally 10–15% lower in burst strength per TAPPI T807.
Q5: What is the most common data quality problem you encounter when building an LCA for a new brand partner, and how do you resolve it?
A: The most frequent issue is missing or estimated substrate mill data — brands often cannot provide a mill EPD for their current board grade, forcing us to use Ecoinvent secondary data with ±15–25% uncertainty. We resolve this by cross-referencing against published EPDs from major mills (Sappi, Stora Enso, APP) for the closest matching grade and GSM, and we flag the uncertainty range explicitly in the LCA report so the brand understands the confidence interval around their carbon number.
Planning a packaging project with sustainability documentation requirements? Contact our team to request a complimentary specification review and sample quote.
© 2026 Ukugi.com. All rights reserved.
Unauthorized reproduction or distribution is prohibited.
The substrate exclusion thing is real — we had a supplier submit a carbon figure for our watch box program last year and it covered print process only, nothing upstream of the pressroom, so the board itself (GD2, roughly 600gsm for the outer) was just… missing from the calculation entirely.
The system boundary point hit close to home. We had a brand submit an EPD to Woolworths in early 2023 and the scope cut off at our gate — excluded outbound freight entirely, which for a Brisbane-to-Perth linehaul run is not a rounding error. The carbon delta when we added Tier 2 transport data from Ecoinvent 3.10 was roughly 23% higher than what they’d declared. Woolworths kicked it back and we spent six weeks in remediation with the brand’s sustainability consultant rebuilding the LCI from scratch.
The substrate exclusion point is exactly where we got burned on a rigid gift box programme for a tea brand — our Yiwu supplier’s EPD documentation covered the greyboard core but stopped at the mill gate, so when the brand’s sustainability team ran their own check against Ecoinvent 3.10, the lacquered art paper wrap (which was carrying most of the carbon) wasn’t in scope at all. Took us two cycles of back-and-forth before we agreed the functional unit had to include the lidding substrate as a separate LCI entry.
The Tier 3 proxy approach using mill EPDs is fine for coated woodfree grades from Sappi or Stora Enso, but we’ve had real problems applying it to virgin kraftliner versus recycled testliner where the Ecoinvent 3.10 datasets diverge by nearly 40% on global warming potential — at that point you’re not reducing uncertainty, you’re just picking the number that suits your scope.
Switching to a cradle-to-gate boundary actually saved us a meaningful conversation with our CFO last year — when we finally ran the full cradle-to-grave numbers on our 350gsm SBS folding carton line for a personal care client, the outbound distribution leg added roughly 18% to the per-unit carbon figure, which fed straight back into a freight consolidation project that cut $0.23/unit off landed cost at 50k MOQ.
Watch the functional unit definition when the job involves mixed-format kits — we ran a 2024 gift set programme for a personal care brand where the LCA was scoped per 1,000 “finished units” but nobody had locked down whether a finished unit meant the outer shipper, the inner carton, or the assembled kit, and we ended up with three different carbon figures before the brand’s sustainability team finally arbitrated it in writing.
The functional unit definition as “delivered to nominated warehouse” sounds clean until you’re running a job where the same SBS carton ships to three different DCs — one in Rotterdam, one in Dubai, one in Melbourne — and suddenly your cradle-to-gate number is doing a lot of heavy lifting to hide what’s actually a wildly variable transport tail. We ended up having to model three separate LCI runs for a 2023 fragrance launch just to satisfy the brand’s internal consistency check, which wasn’t scoped into the original project at all.
One thing that doesn’t get flagged enough in PPWR prep conversations is how the nominated warehouse definition interacts with returnable transit packaging loops — we run a reusable crate program for a Bordeaux négociant and the LCA treatment of the crate’s multi-trip amortisation completely changes the cradle-to-gate figure depending on whether the warehouse is in the EU or a third country, and ISO 14044 gives you almost no prescriptive guidance on which trip count assumption is auditable.
Seal failure on a 90µm BOPP overwrap for a botanical supplement sachet line — roughly 40,000 units destined for a UK health retailer — and we spent three weeks chasing the root cause before realising our LCI data for the heat-seal lacquer had been pulled from an Ecoinvent 3.8 process entry that didn’t match the actual coating weight we were running. Nothing in our Tier 2 data flagged the variance because the acceptable range in our LCI table was set too wide for that substrate class. The carbon figure was wrong by about 18% but honestly the bigger problem was we’d already submitted the PCF declaration to the buyer before the seal failures came back from the DC.
The cradle-to-gate default makes sense for most of what we run, but we hit a real complication last year when a French fragrance client needed to comply with both PPWR and their parent group’s Scope 3 Category 12 reporting simultaneously — the two frameworks wanted incompatible system boundaries from the same LCA study, and ISO 14044 clause 4.2.3 doesn’t give you much guidance on how to reconcile that. We ended up producing two separate goal-and-scope documents for the same folding carton job, which no one had budgeted time for.
The “1,000 finished units delivered to nominated warehouse” functional unit trips people up when you’re running a job with significant reel-to-sheet trim waste — we tracked a 340gsm GC1 carton job last year where trim loss was running at 18% and that waste stream simply wasn’t captured in the LCI until we flagged it at the LCIA phase.