EPD and Carbon Footprint Implementation Roadmap: From a Four-Layer Framework to a Manufacturer's Action Checklist

TL;DR — Continuing the argument of the previous article (LCA is a compliance system, not science), this one tackles the implementation problem. CBAM entered its definitive phase as of 2026/1/1, and manufacturers without credible carbon footprint data are assigned higher default values—effectively a penalty on those who haven’t done LCA. Going from zero to a first EPD takes roughly 8–18 months and NT$500,000–5,000,000, advancing through the four-layer framework (method/standards/data/certification), with choosing a PCR taking priority over choosing a database. 2026 is the last reasonable window for Taiwan’s manufacturing sector to begin.

The previous article painted a panorama of the entire industry—the system formed by stacking four layers (method, standards, data, certification) on top of one another, and the historical forces that drove this system from 1969 to today. But for a company about to enter this field, the real question isn’t “what does this industry look like,” but two more direct questions:

1. Why is every company now being pushed to do this?
2. Where are we now? What should we do next?

This article answers these two questions. The first half discusses the historical evolution of compliance pressure, explaining why EPDs and carbon footprints shifted from “voluntary CSR” to “mandatory trade requirements” in the 2020s; the second half translates the four-layer framework into an actionable roadmap.

Why Is 2026 the Point of No Return? Fifty Years of Accumulated Compliance Pressure

Three Waves from “Voluntary Disclosure” to “Mandatory Verification”

When LCA tools were born in the 1990s, their primary applications were internal corporate analysis and marketing communication. But over the past three decades, the compliance pressure surrounding LCA has undergone three distinct waves of evolution.

First Wave (1990–2010): The Voluntary Framework Era

• Main drivers: rising consumer awareness, ISO standardization, the Nordic-pioneered EPD system
• Corporate motivation: brand differentiation, self-regulation, CSR reporting
• Compliance pressure intensity: low, primarily “peer pressure”

Second Wave (2010–2020): The Quasi-Mandatory Era

• Main drivers: policies derived from the Kyoto Protocol, the EU’s Integrated Product Policy, the Paris Agreement
• Corporate motivation: procurement tender requirements, supply chain demands from major clients, green building certification needs
• Compliance pressure intensity: medium—specific regulations began requiring LCA results, but enforcement was lax

Third Wave (2020–present): The Mandatory Regulation Era

• Main drivers: the EU Green Deal, CBAM, the Battery Regulation, ESPR, CSRD
• Corporate motivation: market access, avoiding tariffs, avoiding penalties
• Compliance pressure intensity: high, with trade-blocking effect

We are now in the middle of the third wave. 2023–2026 is the critical window in which this system leaps from “quasi-mandatory” to “truly mandatory.” Any company that fails to enter the field during these years may face being forced to scramble to catch up later.

The Concrete Pressure of Three Key Regulations

For any manufacturer looking to export to the EU, the following three regulations directly determine the priority of EPDs and carbon footprints.

CBAM (Carbon Border Adjustment Mechanism): Already in Effect

The EU launched the CBAM transitional period in October 2023, and it entered the definitive phase on January 1, 2026. It covers steel, aluminum, cement, fertilizers, electricity, and hydrogen, and the EU has proposed extending it to downstream products (chemicals, plastics, glass, textiles, etc.) before 2028.

The real pressure on manufacturers:

• Exports to the EU must declare a product’s “embedded emissions”
• For emissions exceeding the EU industry benchmark, manufacturers must purchase CBAM certificates (sold from February 2027, equivalent to the EU ETS carbon price)
• No credible data → the “default value” applies → and the default value is usually significantly higher than the actual figure, effectively penalizing manufacturers who haven’t done LCA

CBAM doesn’t directly require an EPD, but it makes “product-level carbon footprint calculation” mandatory—and once a company has built carbon footprint calculation capability, it is only a PCR and third-party verification away from an EPD.

The New Battery Regulation (EU Regulation 2023/1542): Tightening in Stages

In effect since 2023. From February 2025, EV batteries sold into the EU must provide a carbon footprint declaration; from February 2026, industrial batteries (>2 kWh) follow; from February 2027, all batteries must have a Battery Passport; with further tightening thereafter.

The real pressure on manufacturers:

• Not just battery makers, but the entire battery supply chain—from mining, cathode and anode materials, and electrolytes to assembly plants
• Upstream and downstream must pass along “company-specific datasets,” with DQR ≤ 2
• Once carbon footprint grading kicks in, low-scoring products effectively lose the market
• A “maximum carbon footprint cap” will be set later, with non-compliant products banned from sale

The Battery Regulation is the first EU regulation to directly bind LCA results to market access, and other industries may follow suit.

ESPR (Ecodesign for Sustainable Products Regulation): In Effect 2024

In effect since 2024. The broadest-reaching product regulation in EU history, it will require most physical goods to have a Digital Product Passport (DPP) before 2030.

The real pressure on manufacturers:

• The DPP must include product environmental footprint information
• Textiles, furniture, building materials, electronics, and the like are expected to be prioritized for inclusion
• Not just a one-off carbon footprint, but data continuously updated across the full life cycle

ESPR advances LCA/EPD from “a compliance tool for specific industries” to “a universal requirement for nearly all physical goods.”

The Echo in Asian Markets

EU regulations have prompted responses from major Asian economies, creating multiple pressures manufacturers must face simultaneously.

For Asian manufacturers, “do it first for the EU, then reuse it for Asian compliance” is a pragmatic dual-track strategy—EU EPDs are usually recognized by Asian authorities, but the reverse is not necessarily true.

Six Questions to Ask Yourself Before Launching an EPD Project

Before launching any LCA/EPD program, honestly answer six questions. The answers to these six will determine every subsequent choice in your roadmap.

Q1. What is our motivation for doing this?

• Market pressure: client requirements, tender needs, supply chains demanding disclosure
• Regulatory requirements: exporting to the EU, CBAM-covered products, mandatory labeling in specific countries
• Brand positioning: green marketing, ESG reporting, sustainability claims
• Internal decarbonization: finding emission hotspots, setting reduction targets

Different motivations determine the rigor you need. Market pressure and regulatory requirements call for a “formal certification level”; brand and internal uses can start with a lightweight version.

Q2. Does our product have an existing PCR (Product Category Rule)?

• An existing PCR → drastically shortens preparation time; apply it directly
• None, but a similar category exists → can serve as reference but requires discussion with the program operator
• None at all → you may need to draft one yourself or apply for development, extending the timeline by 6–12 months

Q3. Which markets are we targeting for export?

Choosing the wrong system means wasted work, since the subsequent choices of database, PCR, and verifier are all tied to it.

Q4. What is the current state of our internal data?

• Already have an ISO 14064 organizational carbon inventory → some data is transferable
• Already have ISO 50001 energy management → energy consumption data is more complete
• Already have ISO 14001 environmental management → a foundation on the environmental side
• None at all → must collect from scratch

Q5. What is our budget and timeline tolerance?

📊 Key Figures

• First EPD, from zero to publication: 8–18 months
• Budget range: NT$500,000–5,000,000 (database licensing, consulting fees, verification fees, and internal labor combined)
• Marginal cost from the second EPD onward: 6–9 months, roughly half the budget

Q6. Will we maintain it on an ongoing basis?

An EPD is typically valid for 5 years. It must be updated if a major process change occurs in the interim; it must be re-verified at expiration. This affects whether you build internal capability or treat each instance as an outsourced project.

A Five-Stage Maturity Model: Which Stage Is Your Company At?

Roughly dividing a company’s progress in this field into five stages helps clarify where you stand.

When pushed into this field by the market, most manufacturers fall between Level 0 and Level 1. Crossing from Level 1 to Level 3 typically takes 12–24 months and one full project cycle.

Working backward from the compliance timeline: If your product falls within the scope of CBAM, the Battery Regulation, or ESPR, and 2027–2028 is a compliance deadline, you should be at least at Level 1 right now (mid-2026) and launch a project toward Level 3—because going from Level 1 to Level 3 takes about 12–18 months.

The Method Layer: The Foundational Homework of LCA Methodology

Why does this layer exist? This layer exists because of the methodological chaos of the 1990s—back then, the same product would yield wildly different results from different consultants, forcing ISO TC 207 to be established in 1993 and to release ISO 14040 in 1997. Even with ISO standards today, there remains substantial “discretionary room” in method choices, requiring someone internal who knows the field to judge whether a consultant’s recommendation is reasonable.

This layer usually doesn’t get stuck, because the standards documents are public and course resources are abundant. Problems will arise in the later layers, but how solidly this layer is done affects the quality of subsequent decisions.

The typical current situation

• A vague awareness of the LCA concept, but no complete training
• May have taken a short external course
• No dedicated staff, often handled part-time by the ESG / QA / R&D unit

Recommended next steps

1. Assign a dedicated point of contact: 1–2 internal liaisons. Even with full outsourcing, you need someone who knows the field to interface with consultants—otherwise what the consultant delivers will be unintelligible and impossible to judge as right or wrong.

2. Build foundational understanding: Read ISO 14040 and 14044 (Taiwan has the CNS 14040 Chinese version) and the General Guide chapter of the ILCD Handbook. These two documents total over 200 pages, but 80% of the key concepts are in them.

3. Take practical courses: Compared with general-overview courses, prioritize case-study-based training. Seeing how someone calculates a real case is far more useful than hearing 50 slides on methodology.

Why Is Choosing a PCR the Most Critical Early Decision?

Why does this layer exist? When Sweden created the world’s first EPD system in 1998, it found that “within the same product category, different companies use inconsistent methods, making EPDs incomparable.” So it introduced the concept of “Product Category Rules (PCR)“—the industry jointly sets the rules of the game for each product category. This design was later formally institutionalized by ISO 14025 (2006), and the EU’s PEFCR today is a strengthened version of the same logic.

Decisions in this layer affect all subsequent work. Choosing the wrong PCR means redoing the entire LCA.

Common blind spots in the current situation

• Not knowing what a PCR is
• Knowing, but unable to find a version suitable for your own product
• Finding one, but using an expired version
• Finding multiple versions and not knowing which to use

Priority Order for Searching for a PCR

1. EPD International’s PCR database — the broadest coverage, in English, used globally
2. Industry-specific operators — IBU (focused on building materials), EPD Norge (Nordic building materials), PEP ecopassport (electrical and electronic)
3. The EU PEFCR list — on the European Commission website, regulation-oriented
4. Regional EPD programs — Japan’s SuMPO, South Korea’s KEITI, etc.

What to Confirm After Finding a PCR

The Impact of the PEFCR Wave

The EU is systematically absorbing the PCRs of various industries into PEFCRs (PEF Category Rules). The implications for manufacturers:

• Short term (2024–2028): existing PCRs coexist with newly released PEFCRs; you need to confirm which applies to your target market
• Medium term (2028–2030): PEFCR coverage increases, and EU regulatory use will progressively point toward PEFCRs
• Long term (post-2030): PEFCRs are expected to become mainstream within the EU, and existing PCRs may need to be realigned

If your EPD program is primarily for the EU market, we recommend first checking whether your target category has a PEFCR draft or final version—lest you finish a round under a PCR version only to redo it shortly after.

The Two Battlefields of the Data Layer: Foreground Data vs. Background Databases

Why does this layer exist? The construction of the data layer reflects the evolution of LCA from “an internal corporate tool” to “cross-organizational collaboration.” In the early days, companies did LCA themselves and cared only about their own factory’s numbers; starting in the 1990s, LCA needed to incorporate upstream (raw material production) and downstream (waste disposal) data, which a single company could not produce alone—hence the emergence of centralized foundational databases (the big four: ecoinvent, GaBi, IDEA, CLCD). The distributed architecture of the LCDN today is the latest stage of this evolution.

The work of the data layer can be split into two battlefields with entirely different types of challenges.

Battlefield One: Foreground Data (Your Own Factory’s Data)

The typical current situation

• Energy bills are scattered across different units—accounting, facilities, production control
• Raw material procurement records aren’t divided by batch, making them hard to map to a single product
• Waste and emissions aren’t metered separately by production line
• In factories where multiple products share a line, energy consumption cannot be precisely allocated

Recommended next steps

1. Begin data collection centered on the “functional unit” Determine whether you’re calculating “1 kilogram of product,” “1 square meter of panel,” “1 packaging unit,” or something else, and organize all data around this unit.

2. Build a Data Inventory Sheet Categorize by inputs and outputs:

• Inputs: electricity, fuel, water, primary raw materials, auxiliary materials, transportation
• Outputs: target products, by-products, waste, air emissions, wastewater

3. Typical data gaps

• Auxiliary materials (packaging, cleaners, lubricants)
• Indirect energy consumption (air conditioning, lighting, the allocation of shared facility infrastructure)
• Internal transport and distribution
• Employee commuting (some PCRs don’t require it, but some do)

4. Data representativeness period Usually take the most recent complete year(12 consecutive months); you cannot cherry-pick especially good or especially bad months.

Battlefield Two: Background Databases (Upstream Data)

The choice is constrained by budget, target market, and PCR rules

Recommended next steps

1. For small companies starting out, consider: openLCA software (free) + a short-term ecoinvent license (a single-project license is available). This combination can push the entry cost to the minimum.

2. Primary principles for choosing a database

• First principle: Which one does the verification body in your target market recognize?
• Second principle: What does the PCR restrict?
• Third principle: Can your budget and timeline bear it?

3. Cross-checking with multiple databases (an advanced approach) For important products, we recommend calculating once each with two databases and seeing the gap in results. If the difference is too large, investigate why—it could be different model assumptions, differences in geographic representativeness, or outdated data in one of the databases. This is very helpful for later addressing the verifier’s questions.

Battlefield Three: Modeling and Calculation

After data collection, you move into modeling and calculation. On the surface this stage is technical work; in substance it is a series of “method choice” decisions.

Main tools

• openLCA: free and open-source, suitable for academia and SMEs
• SimaPro: the European industry standard, paid, full-featured
• GaBi: under Sphera, commonly used in automotive and industry
• One Click LCA: building-material-oriented, user-friendly interface, suitable for BIM integration

Key decisions at this stage

1. System boundary: strictly follow the PCR rules, but where the PCR leaves gaps, decide and explain on your own
2. Allocation method: when multiple products share a line, do you use mass allocation, economic allocation, or substitution? The choice can affect results by more than 20%
3. Cut-off criteria: are trace materials included? ILCD defaults to allowing omission when cumulative contribution is < 1%, but this must be declared in the report
4. Choice of reference flows: for the same electricity, choosing ecoinvent’s “Taiwan, market for electricity, low voltage” versus “TW, electricity production, hard coal” yields very different results

Why Has the EPD Become a “Legal Qualification Certificate for Marketing”?

Why does this layer exist? After the Volkswagen Dieselgate scandal broke in 2015, EU regulators realized that voluntary environmental disclosure could no longer earn the market’s trust—if even a multinational automaker would cheat on emissions certification, the credibility of voluntary declarations collapsed. After Dieselgate, the EU pivoted entirely toward the path of “mandatory third-party verification”: EPDs must be reviewed by independent verifiers, PCRs must be developed by industry consensus, and results must be traceable. Every institutional design in this layer is, in essence, “trust repair.”

The choices in this layer appear to come last, but in fact they shape the specifications of all the work that comes before.

Considerations in Selecting a Program Operator

Choosing a Verifier

• Usually chosen from the program operator’s approved list
• The type of experience must match: prioritize those who have done your product category
• Geographic considerations: a local verifier means lower communication costs and fewer time-zone issues
• Some PCRs mandate a multi-person mode (panel review), which needs to be confirmed in advance

Typical Questions You’ll Be Asked During the Verification Stage

Preparing answers to these categories of questions in advance can substantially shorten the verification cycle:

• Why is the system boundary drawn this way? Has anything been omitted?
• What is the rationale for the choice of allocation method? Have you tested the sensitivity of other methods?
• What is the basis for the DQR scoring? Can the low-scoring items be improved?
• What is the logic for handling by-products?
• Why was this background database chosen? Why this particular dataset rather than that one?
• Are the data sources traceable? Are there original invoices and measurement records?

The Impact of the Green Claims Directive

The Green Claims Directive proposed by the EU in 2023 (expected to pass in 2026 and take effect in 2028) will set a new threshold for “environmental claims” beyond EPDs.

Main provisions:

• Any “green,” “low-carbon,” or “eco-friendly” claim must be based on an LCA
• Claims must be third-party verified
• Comparative claims (e.g., “X% lower carbon emissions than a competitor’s product”) must follow strict comparison rules
• Violators face fines of up to 4% of annual turnover (similar to the GDPR level)

The implication for manufacturers: Even if you don’t do an EPD, the moment you mention “eco-friendly” or “sustainable” in marketing, you fall under this regulation. The EPD thus shifts from a “marketing bonus” to a “legal qualification certificate for marketing.”

Five Cross-Layer Decisions You Must Think Through in Advance

In actual execution, several cross-layer decisions recur and are worth thinking through before launch.

Decision Point 1: Cradle-to-gate or Cradle-to-grave?

• Cradle-to-gate (from the cradle to the factory gate): data is easy to control, suitable for B2B and intermediate materials
• Cradle-to-grave (to the grave): requires estimating the use phase and waste disposal, suitable for B2C and green building use

Some PCRs allow a Modular EPD, slicing the life cycle into modules such as A1–A5 (production and construction), B1–B7 (use), C1–C4 (disposal), and D (recycling benefits), disclosing module by module. This A1–D modular architecture originates from EN 15804 (the European building-material EPD standard) and has since spread to other industries.

Decision Point 2: Internal Team or Outsourced Consultant?

A pragmatic compromise: outsource the first EPD, but explicitly stipulate a knowledge-transfer obligation in the contract, requiring the consultant to leave behind complete modeling files and the data inventory sheet.

Compliance consideration: If your product falls within mandatory scopes like CBAM or the Battery Regulation, building internal capability is recommended in the long run—because declarations must be updated every year, and the accumulated cost of pure outsourcing will exceed that of an internal team.

Decision Point 3: Single-Product EPD or Platform-Type EPD?

• Single-product EPD: precise, one per SKU, but costly
• Platform-type / representative-product EPD: uses one representative product plus a description of the variation range—lower cost but limited representativeness of results. Some PCRs set an upper limit on the degree of variation.

SMEs usually start with the platform type; only large manufacturers or those with many differentiated products take the single-product route.

Decision Point 4: What Do You Communicate Externally After Completing an EPD?

• Pure technical disclosure: post it on the official website, provide it for clients to review, supply it for tenders and bids
• Marketing use: packaging, advertising, ESG report citations

Marketing use requires special attention to “greenwashing” risk. The term “greenwashing” was first coined in 1986 by environmental activist Jay Westerveld, but for a long time it carried no legal consequences. Starting in the 2020s, the EU, the UK, and the US FTC all strengthened scrutiny of green claims—the EU’s Green Claims Directive raises penalties for violations to the GDPR level.

EPD results are “objective numbers,” but “comparative claims” (e.g., “30% lower carbon emissions than a competitor’s product”) carry additional rules; usually both EPDs must be based on the same PCR and verified by the same body to be comparable.

Decision Point 5: Do It for the EU, for Asia, or Dual-Track?

In practice, an “EU-version EPD” can almost always be accepted in Asia; the reverse is not necessarily true. So for multi-market manufacturers, the EU-version EPD is the “greatest common denominator” choice.

Common Pitfalls Across the Three Phases of an EPD Project

Compiling the places where things easily go wrong during execution, they can be divided into three phases.

Early Phase (Preparation to Data Collection)

• Starting data collection before selecting a PCR, only to find the scope doesn’t match
• Failing to define the system boundary clearly first, only to discover missing items after data collection is done
• Underestimating hidden costs like database licensing, employee time, and verification fees
• Directly repurposing data from the organizational carbon inventory (ISO 14064) without allocating it by product
• Failing to check whether the target category has a PEFCR draft, and discovering shortly after finishing the PCR version that it must be redone

Mid Phase (Modeling to Internal Review)

• Sloppy energy-consumption allocation methods in multi-product factories, undermining overall credibility
• Upstream suppliers unwilling to provide detailed data, forcing the use of secondary data and dragging down the DQR score
• Inconsistent terminology between the Chinese and English versions, which fails to align during later translation
• Failing to do result sensitivity testing, leaving you speechless when the verifier asks
• Over-relying on ecoinvent’s “global average” data without adjusting for the Taiwan grid

Late Phase (External Verification to Publication)

• Discovering only at the verification stage that a method choice is contentious, requiring large sections of calculation to be redone
• No update plan after publication, only realizing at the 5-year expiration that everything must be redone
• Misleading claims when EPD results are used in marketing, triggering external scrutiny or Green Claims legal risk
• No trigger mechanism for updates when processes or suppliers change
• Discovering during CBAM declaration that the EPD scope doesn’t align with CBAM requirements, requiring separate calculation

8–18 Months: The Timeline Rhythm from Zero to EPD Publication

From the decision to do it to EPD publication, the typical rhythm is as follows:

The first EPD usually falls in the 12–18-month range; with experience, the second onward can be shortened to 6–9 months.

Timeline reminder under compliance pressure: If your goal is to be ready in 2027 for CBAM’s product expansion or the launch of battery grading, you must launch the project before mid-2026 to have a reasonable chance.

A Decision Checklist for Management’s Kickoff Meeting

If you are the manager deciding whether to launch an EPD project, we recommend confirming the following checklist before the kickoff meeting.

Strategy

• The motivation is clear and put in writing (market, regulation, brand, decarbonization—pick the primary one)
• The target market and corresponding EPD system are selected
• The budget ceiling and timeline targets are approved
• The long-term strategy on whether to build internal capability is decided
• The compliance-pressure timeline has been mapped (the timeline of CBAM, the Battery Regulation, ESPR, etc., and their impact on your own products)

Standards

• The corresponding PCR is found and its version verified
• The functional unit, system boundary, and impact categories specified by the PCR are clarified
• If there is no PCR, an alternative plan is decided
• The PEFCR draft status has been confirmed (if primarily targeting the EU market)

Execution

• A dedicated internal point of contact is assigned (name, title, hours committed)
• A consultant is selected (if outsourcing) or an internal team is established
• LCA software and the background database are procured
• The data inventory sheet is drafted and the data-source units are notified

Certification

• The program operator is selected
• A shortlist of verifier candidates is prepared
• An ongoing maintenance mechanism is planned (update trigger conditions, arrangements for the 5-year expiration)

Communication

• The scope of external use for the EPD results is defined
• Greenwashing risk is assessed (in preparation for the Green Claims Directive)
• Internal stakeholders (marketing, sales, R&D) are aligned on expectations

The 2024–2030 Compliance Map: Key Milestones for the Next Six Years

For any manufacturer running a long-term business, the following are the compliance milestones worth tracking over the next six years.

2024–2025: CBAM Transitional Period, Battery Regulation Phased Launch

• CBAM transitional period: importers must declare, but CBAM fees are not yet levied
• Battery Regulation initial phase: basic due diligence, recycling obligations, and labeling requirements take effect
• ESPR takes effect: the EU begins announcing priority product groups and launches DPP pilots

2026 (now): CBAM Enters the Definitive Phase, Green Claims Expected to Pass

• CBAM entered its definitive phase on 2026/1/1—importers must prepare to purchase CBAM certificates starting this year
• CBAM certificate sales: scheduled to begin on February 1, 2027
• The Green Claims Directive is expected to pass and enter a transposition period
• The EU is expected to complete several key-category PEFCRs

2027–2028: Battery Grading, CSRD Expansion

• Battery carbon footprint grading kicks in (grades A–G), and low-scoring products are blocked from sale
• The Battery Passport becomes mandatory from February 2027
• The scope of CSRD expands to mid-sized enterprises (>250 employees)
• The Green Claims enforcement period

2029–2030: ESPR Becomes Mandatory Category by Category, CBAM Product Expansion

• ESPR covers multiple categories (textiles, furniture, building materials, electronics)
• CBAM is expected to extend to chemicals, plastics, and glass
• The Battery Regulation sets a “maximum carbon footprint cap”

Post-2030

• The EU goes fully DPP, Asian systems mature, and a global mutual-recognition attempt for an “environmental passport” may emerge
• Most physical goods exported to the EU must have an LCA/EPD foundation

Two Failure Modes and Launch Recommendations

The most common failure mode is pouring too much effort into the data layer, only to find that the decisions in the standards layer were wrong—for example, the wrong PCR was chosen or the system boundary was incorrectly defined—causing earlier work to be redone. The way to prevent this is, in the month before launching the project, to first nail down the choices in the standards layer and certification layer (which PCR, which EPD system, which target market) before moving into the execution of the data layer.

The second most common failure mode is treating the EPD as a one-off project. An EPD is valid for 5 years, during which changes to the supply chain, processes, or material formulations may require an interim update. Viewing the EPD as an ongoing data-management mechanism rather than a one-off report can substantially reduce subsequent costs.

The barrier to entering this field is not low, but the system itself is open and learnable. The first time you go through the entire process, the capability built internally will keep accumulating, and the marginal cost of EPDs for subsequent products will drop quickly.

More importantly, this system will not disappear—it will only expand. The investment in entering this field is usually an irreversible capability build, not a one-off compliance expenditure.

Previous in the series: LCA Is Not Science, It Is a Compliance System: Fifty-Six Years of Evolution from Coca-Cola 1969 to CBAM 2026