Ask what the comparison is for
Life-cycle assessment (LCA) is a structured method for evaluating potential environmental impacts associated with a product system. It does not produce a context-free score. The result depends on the question, the functional unit, the system boundary, data quality, allocation choices, impact assessment method and the scenarios used for end of life. ISO 14040 and ISO 14044 set out principles and requirements for LCA; an individual study should state how its choices follow those standards.
For glass manufacturing, comparing one tonne of cullet with one tonne of virgin raw material may not answer whether two products provide the same service. A better functional unit might describe a defined quantity of packaging delivered with specified capacity and protection, or a defined construction function over a stated period. The right unit depends on the decision. It should be clear enough that readers can see what is held constant.
Boundaries decide what is counted
A study might begin at resource extraction and end at disposal, or focus on a narrower gate-to-gate process. Glass recycling comparisons can include collection, sorting, cullet transport, processing residues, furnace operation, product manufacture, packaging, distribution and end-of-life treatment. If important stages are omitted, the result should say so. Recovered material may carry burdens from collection or sorting, while allocation methods determine how burdens and benefits are assigned between product systems.
Transport distance matters because recovered glass can be heavy and value per tonne varies by grade and route. Reprocessing also uses equipment and energy. For remelting, electricity and fuel mix, furnace efficiency, cullet quality, batch composition and plant conditions influence results. For alternatives such as foam glass or powder, crushing, grinding, additives, heating, yields and product service life are additional factors. Avoid transferring a result from one technology or country to another without checking those conditions.
Look for a real substitution
Recycled content is an input description, not a complete impact metric. A product can contain recovered glass and still require other materials, energy, processing and transport. To claim an avoided impact, an analysis needs a defensible account of what the recovered glass displaces. Does cullet replace a portion of virgin batch material? Does glass aggregate replace a specified aggregate in an engineered application? Does fine powder displace a cementitious constituent at a demonstrated rate? The substitution must be technically feasible and linked to actual product performance.
Quality loss or contamination may redirect material to a lower-value application, which can influence both the substitution scenario and the next stage of its life. Recycling routes can also compete for the same feedstock. A scenario should make its assumed allocation, market response and recycling loop explicit. Avoid counting the same recovered material benefit twice in connected systems.
Read a declaration carefully
An environmental product declaration (EPD) reports life-cycle information under defined product category rules and programme procedures. Check the declared unit, modules reported, geography, reference period, data sources, allocation method and whether the declaration is product-specific or an average. An EPD is not by itself a ranking or a guarantee that one product is preferable. Two declarations are comparable only when their scope and methods support that comparison.
Useful frameworks include ISO 14025 for type III environmental declarations, EN 15804 for construction-product EPDs in its applicable market, and programme instructions from providers such as EPD International. The European Commission's Product Environmental Footprint work and the EU's environmental policy initiatives provide additional context. Verify the document's current version and rules before using it in a specification or public claim.
State uncertainty and avoid false precision
Data for industrial processes can vary by facility, equipment, energy source, product mix and year. A single point estimate may conceal that variation. A transparent study explains where it uses measured primary data, where it relies on secondary databases such as ecoinvent, and where assumptions fill gaps. Sensitivity analysis can show whether the conclusion changes under different transport distances, energy mixes, recovery rates or service-life scenarios.
Keep impact categories distinct. Greenhouse-gas results do not automatically describe water use, resource depletion, toxicity or local pollution. A claim about carbon should state the metric, time horizon if relevant, unit, boundary and comparison basis. Avoid broad phrases such as “zero impact,” “infinitely recyclable” or “always better” unless the precise statement can be supported and qualified.
A concise evidence checklist
- Name the product, destination and intended service.
- Define the functional unit and explain why it is equivalent.
- Show system boundaries, allocation and end-of-life assumptions.
- Report collection, sorting, transport and processing losses.
- Identify data sources, geography, date and energy mix.
- Show actual substitution, product performance and uncertainty.
- Link to the study, EPD or methodology behind the statement.
The European Environment Agency, International Energy Agency, European Commission, ISO, EPD International, ecoinvent and sector bodies can provide background, but a broad sector statistic cannot replace a product-specific analysis. The most useful environmental claim is often modest: it tells readers what was measured, where the evidence applies and what remains uncertain.