A metal bottle and detached cap sit beside packaging, material samples, drawings and a caliper.
An inventory starts with the product and its components, then follows the activities that deliver its function.
Diagram text description

A brushed metal bottle, separate cap and cylindrical paper package sit on a workbench beside material samples, drawings and a caliper.

To do a product life cycle assessment, decide what question the study should answer and what service the product provides. Then map the activities that deliver that service, collect their inputs and outputs, model the relevant environmental impacts and interpret the results. A useful study makes clear which design or purchasing decision the evidence can support.

The four phases in the ISO life cycle assessment framework are goal and scope definition, life cycle inventory, life cycle impact assessment and interpretation. Goal and scope set the question and boundary. The inventory records physical flows. Impact assessment translates relevant flows into indicators of potential effects. Interpretation tests completeness, consistency and sensitivity before the team draws a conclusion. The phases are iterative. If the recommendation depends on a poorly supported assumption about lifetime, revisit the scope or inventory.

Goal and scope leads to inventory, then impact assessment, then interpretation. Interpretation can send the team back to revisit assumptions.
Interpretation can reopen earlier phases when a conclusion depends on weak evidence.
Diagram text description

Goal and scope define the question and boundary. Inventory records the physical flows. Impact assessment models applicable indicators. Interpretation tests completeness, consistency and sensitivity and can send the team back to revise scope or collect better inventory evidence.

A product carbon footprint addresses the climate question. An LCA designed to consider multiple impacts also examines the other categories relevant to its goal. The reusable bottle example below shows how to turn one inventory flow into one climate contribution, then hand the unfinished study to a reviewer with its open questions visible.

Decide the study question before collecting data

Start by writing the decision in one sentence. “We want to choose which bottle component to investigate for redesign” points towards a different study from “We want to publish a claim comparing this bottle with a competitor’s.” Identify who will use the result, what they need to decide, the product revision and what evidence they will require.

Intended outputWhat to settle before collecting data
Internal hotspot studyProduct revision, relevant stages, impact categories and acceptable uncertainty for prioritising work
Product carbon footprintClimate accounting method, defined unit and boundary, and what review the intended use requires
Environmental product declarationApplicable programme, product category rules and route to verification and registration
Public comparisonEquivalent service, common methods and data quality, and the applicable critical review and communication requirements

Table 1, shown one row at a time. Every field is retained.

Row 1

Intended output
Internal hotspot study
What to settle before collecting data
Product revision, relevant stages, impact categories and acceptable uncertainty for prioritising work

Row 2

Intended output
Product carbon footprint
What to settle before collecting data
Climate accounting method, defined unit and boundary, and what review the intended use requires

Row 3

Intended output
Environmental product declaration
What to settle before collecting data
Applicable programme, product category rules and route to verification and registration

Row 4

Intended output
Public comparison
What to settle before collecting data
Equivalent service, common methods and data quality, and the applicable critical review and communication requirements

Choose among an internal hotspot, product carbon footprint, EPD or public comparison based on the purpose and review needs.
Set the intended output and audience before selecting a method or evidence package.
Diagram text description

An internal hotspot uses the relevant revision and stages. A climate footprint needs a defined unit, boundary and accounting method. An EPD follows programme and product category rules. A public comparison needs equivalent service and applicable critical review.

Describe the audience as well as the output. A design team can use a provisional result to choose where to measure next. A customer reading a public comparison needs to know what the products were assumed to do, for how long and under which conditions. Determine how the team will deal with confidential supplier inputs while still giving the reviewer enough information to assess them.

Agree the review plan early. Where a conclusion may be shared publicly, decide which method and rules govern it before collecting a dataset that could prove unsuitable.

Set the functional unit and boundary

Define the functional unit in terms of service. For a reusable bottle, a meaningful service might be carrying 300 litres of drinking water in 600 half-litre fills, with the same leak resistance and comparable cleaning conditions. The reference flows are the quantities of products and activities needed to deliver that service, such as the number of bottles, replacement lids and washes.

Boundary choiceWhat the team models
Manufacturing studyMaterials, manufacture and other specified activities up to the stated gate, with later stages disclosed as outside scope
Study of the delivered serviceMaterials, manufacture, distribution, use, replacement and end-of-life processes required to provide the defined service
Common functions or inputsProcesses shared across scenarios, with consistent treatment and a reason for any exclusion

Table 2, shown one row at a time. Every field is retained.

Row 1

Boundary choice
Manufacturing study
What the team models
Materials, manufacture and other specified activities up to the stated gate, with later stages disclosed as outside scope

Row 2

Boundary choice
Study of the delivered service
What the team models
Materials, manufacture, distribution, use, replacement and end-of-life processes required to provide the defined service

Row 3

Boundary choice
Common functions or inputs
What the team models
Processes shared across scenarios, with consistent treatment and a reason for any exclusion

The delivered-service scope contains the manufacturing boundary and the later distribution, use, replacement and end-of-life stages needed for the service.
Set the shared service before comparing designs across the larger boundary.
Diagram text description

A manufacturing study can cover materials and manufacture to the specified gate. The delivered-service scope adds relevant distribution, use, replacements and end of life. Match the functional unit and document common inputs and exclusions.

A gate-based study of two bottle bodies can inform a manufacturing decision. If the bodies have different expected lifetimes, that result alone cannot tell you which design has lower impacts over the same service. Add the relevant replacement and cleaning activities to answer that question. For common inputs such as the drinking water carried, state whether they are modelled or excluded and why their treatment fits the decision.

Record geography, study period and any cut-offs, which are rules for leaving certain inputs outside the model. A low-mass seal can matter if it drives early failure, so assess the relevance of an exclusion against the study goal. Include the reason for each exclusion and revisit it if the result becomes sensitive to it.

Map the product system and build the inventory

Draw the product system as connected processes. Begin with the versioned bill of materials, which lists the components used in the product. Add process steps that the bill does not describe, including factory energy, water, waste and direct releases. Extend the map to supplier processes, freight, use and disposal according to the boundary.

Process or stageStarting recordData owner
Bottle body, lid, seal and packagingComponent mass, grade, supplier site and versioned bill of materialsEngineering and procurement
Forming, assembly and packingBatch output, energy, water, scrap and direct release recordsProduction and site environmental team
DistributionShipment mass, distance, mode and packing configurationLogistics
Use and replacementCleaning scenario, wear tests, replacement parts and expected service lifeProduct team
End of lifeTreatment routes, material composition and geographic evidenceStudy practitioner

Table 3, shown one row at a time. Every field is retained.

Row 1

Process or stage
Bottle body, lid, seal and packaging
Starting record
Component mass, grade, supplier site and versioned bill of materials
Data owner
Engineering and procurement

Row 2

Process or stage
Forming, assembly and packing
Starting record
Batch output, energy, water, scrap and direct release records
Data owner
Production and site environmental team

Row 3

Process or stage
Distribution
Starting record
Shipment mass, distance, mode and packing configuration
Data owner
Logistics

Row 4

Process or stage
Use and replacement
Starting record
Cleaning scenario, wear tests, replacement parts and expected service life
Data owner
Product team

Row 5

Process or stage
End of life
Starting record
Treatment routes, material composition and geographic evidence
Data owner
Study practitioner
  1. Match each record to the product revision, period and facility. Record the original units before converting or scaling it to the functional unit.
  2. Separate intermediate inputs such as purchased steel or electricity from exchanges with the environment such as emissions to air. Link intermediate inputs to suitable upstream datasets so their relevant exchanges enter the model.
  3. Reconcile inputs and outputs where possible. For a metal forming step, compare stock issued, finished parts and offcuts. Record measurement gaps, yield and treatment of waste.
  4. Give every input a source reference and a status. Document whether the number was measured, provided by a supplier, estimated from a secondary dataset or still missing.

Version the process map and the inventory together. A new lid revision may change mass, lifetime and use conditions. The inventory should show which flows change and which earlier evidence remains valid.

Handle missing supplier data and allocation openly

Request the missing evidence in terms the supplier can answer. If a lid supplier cannot provide a product inventory, ask for the component material, mass, plant and process route. Ask which activities its figures already include. That gives the practitioner a basis for choosing an upstream dataset and checking where boundaries might overlap.

  1. Request product and site-specific quantities for the important inputs, with units, periods and the evidence supporting them.
  2. If a value remains missing, select a documented proxy that matches the material, process and geography as closely as possible. Identify the mismatch and retain the source and version.
  3. Where a process serves several products, try to measure the studied product’s share directly. If a split is necessary, explain the basis and whether it reflects how the process consumes resources.
  4. Run an alternative plausible assumption and show which indicator or decision it changes. Assign the remaining gap to an owner and record the review decision.

Consider a washer serving several production lines. A share based on number of batches might be misleading if the runs have different durations or temperatures. Check whether meter readings or a physical driver such as cycle energy can represent the product more closely. If the available basis is weak, flag it and test a different share.

Keep the original reading, the split and the resulting product quantity together. The reviewer needs to see how much of the inventory relies on the allocation and whether a better measurement would alter the recommendation.

Translate flows into impact categories

Impact assessment estimates potential effects from the inventory. Start with the categories relevant to the decision, then select a documented impact assessment method and version. Its characterisation models relate exchanges with the environment to category indicators. The relationship depends on the substance, the receiving environment and, for some categories, the location.

The diagram shows how one inventory can feed distinct indicators. A greenhouse gas emitted to air can contribute to climate change. An appropriate water consumption flow may contribute to a water scarcity indicator when the basin is known. Extraction of minerals can contribute to a resource indicator under the selected method. Each indicator uses its own model and unit. Some flows can contribute to several categories.

The inventory can feed separate climate change, water scarcity and resource use indicators.
Select the appropriate characterisation method for each relevant category.
Diagram text description

Applicable emissions to air can contribute to a climate indicator. Water flows can need location and consumption evidence for water scarcity. Resource flows follow a separate indicator method. The categories keep their own units.

An intermediate input such as one kilowatt-hour of purchased electricity needs the exchanges from its supply dataset before those exchanges can be characterised. If you use a pre-calculated factor instead, check its indicator, scope and method to avoid counting the same upstream processes again.

Report the category results separately so the reader can see trade-offs. A more durable bottle body could lower replacement needs while using a resource-intensive material. Look at the relevant indicators before deciding which option to investigate. Summing kilograms of carbon dioxide equivalent and a water scarcity indicator would mix unlike quantities.

Interpret hotspots and test the result

Interpretation asks whether the model supports the decision. Check whether all relevant stages are covered, whether the same methods and rules were used across scenarios and which assumptions dominate the result. A hotspot is a large contribution within a specified indicator and boundary. Its prominence can change when an omitted activity or different use scenario is added.

Fictional scenarioReference flows for 600 fillsWhat could change the design choice
Light body lasts 600 fillsOne light body. Add the lids and washing needed for the same serviceA lower manufacturing burden could matter if the rest of the service is comparable
Light body lasts 300 fillsTwo light bodies. Include additional transport and treatment where relevantAn extra body could outweigh the initial material saving
Reinforced body lasts 600 fillsOne reinforced body. Include its greater material demand and replacement lid if neededDurability could help, but the effect depends on the complete inventory and indicator

Table 4, shown one row at a time. Every field is retained.

Row 1

Fictional scenario
Light body lasts 600 fills
Reference flows for 600 fills
One light body. Add the lids and washing needed for the same service
What could change the design choice
A lower manufacturing burden could matter if the rest of the service is comparable

Row 2

Fictional scenario
Light body lasts 300 fills
Reference flows for 600 fills
Two light bodies. Include additional transport and treatment where relevant
What could change the design choice
An extra body could outweigh the initial material saving

Row 3

Fictional scenario
Reinforced body lasts 600 fills
Reference flows for 600 fills
One reinforced body. Include its greater material demand and replacement lid if needed
What could change the design choice
Durability could help, but the effect depends on the complete inventory and indicator

The 600-fill scenario needs one light body if it lasts all 600 fills, or two light bodies if each lasts 300 fills.
The fictional service stays fixed at 600 fills while the required number of bodies changes.
Diagram text description

The upper track uses one light body for all 600 fills. The lower track uses a body for the first 300 fills and a replacement for the next 300. The table adds the reinforced alternative. No indicator outcome or preferred design follows from these counts alone.

The scenarios show what would have to be tested. They do not establish a winning design. Compare a short and long service-life assumption, and vary lid replacement and washing where evidence is weak. If the preferred option changes, identify the data that would resolve the choice.

Return to scope if the assumed service no longer matches the decision. Return to the inventory if the missing evidence concerns a particular process, such as the lid failure rate or how users clean the bottle. Record what you changed and how it affected the result.

Work through a small hypothetical example

Consider this fictional example. Rillway’s promotional bottle R500 revision B is studied for an internal decision about whether to investigate a replaceable lid. The common service is carrying 300 litres of drinking water in 600 half-litre fills. The scenario requires one steel body, the original lid and one replacement lid, assuming a body life of 600 fills and a lid life of 300 fills. Those lifetimes and all example data are invented.

The fictional R500 needs one steel body and two lids for 600 fills. Obtain lid and seal masses and verify durability.
Hold the service constant while testing the original and replacement lid assumptions.
Diagram text description

The fictional body lasts all 600 fills. The original lid and one replacement each serve 300 fills. Confirm lid and seal masses, material routes, failure rates and cleaning conditions before comparing design options.

The inventory file identifies the material mass of each part, the number of replacements, the cleaning activities and the fate of the components. The table isolates three entries allocated to the production of the one bottle body needed for the service. One entry is an elementary flow to the environment. The other two need different evidence before they can support an impact result.

Inventory entry for the exampleHow it enters impact assessment
0.40 kg of fossil CO₂ emitted to air from a modelled forming operationApply the CO₂ reference factor of 1 kg CO₂e per kg CO₂. This gives a climate contribution of 0.40 kg CO₂e from that flow
0.20 kWh of purchased electricity for the same activityLink a suitable grid dataset and characterise its relevant exchanges. No electricity contribution is calculated here
2 litres of process water withdrawnDetermine source, location and how much is consumed or returned before modelling a water scarcity contribution

Table 5, shown one row at a time. Every field is retained.

Row 1

Inventory entry for the example
0.40 kg of fossil CO₂ emitted to air from a modelled forming operation
How it enters impact assessment
Apply the CO₂ reference factor of 1 kg CO₂e per kg CO₂. This gives a climate contribution of 0.40 kg CO₂e from that flow

Row 2

Inventory entry for the example
0.20 kWh of purchased electricity for the same activity
How it enters impact assessment
Link a suitable grid dataset and characterise its relevant exchanges. No electricity contribution is calculated here

Row 3

Inventory entry for the example
2 litres of process water withdrawn
How it enters impact assessment
Determine source, location and how much is consumed or returned before modelling a water scarcity contribution

The 0.40 kg CO₂e is an illustrative characterised contribution from one modelled emission, with no measured or supplier-verified input. The other stages, flows and categories remain unquantified. Record the climate factor and the allocation to the reference flow beside the value so it can be reproduced without treating it as a bottle result.

Before using the model to choose a lid, obtain the lid and seal masses, material routes and durability evidence. Vary the number of lids over the same 600 fills. Then test cleaning and disposal assumptions alongside the material change. The design decision stays open until the relevant categories have enough evidence to assess the trade-offs.

Prepare a reviewable study handoff

Hand off the model and the decision record together. Give the next practitioner a starting point that identifies the exact files, what can be rerun and the assumptions that still need agreement. The following procedure uses the same fictional bottle study as an example of what each handoff step should contain.

StepWhat to sendWho checks it
1. Confirm the scopeR500 revision B, internal lid redesign question, 600 fills, included stages, geography and exclusionsStudy sponsor and practitioner
2. Reconcile the inventoryVersioned process map, reference flows, BOM, original units and links to source recordsEngineering, procurement and data owner
3. Record the impact methodSelected categories and method version, mapping of exchanges to indicators and missing models or datasetsLCA practitioner
4. Test the assumptionsShort and long lifetime cases, alternate cleaning scenario, allocation choices and unresolved data gapsProduct team and practitioner
5. Decide the permitted useReview log, open issues, evidence request, deadline and the decision that must stay provisionalStudy reviewer and sponsor

Table 6, shown one row at a time. Every field is retained.

Row 1

Step
1. Confirm the scope
What to send
R500 revision B, internal lid redesign question, 600 fills, included stages, geography and exclusions
Who checks it
Study sponsor and practitioner

Row 2

Step
2. Reconcile the inventory
What to send
Versioned process map, reference flows, BOM, original units and links to source records
Who checks it
Engineering, procurement and data owner

Row 3

Step
3. Record the impact method
What to send
Selected categories and method version, mapping of exchanges to indicators and missing models or datasets
Who checks it
LCA practitioner

Row 4

Step
4. Test the assumptions
What to send
Short and long lifetime cases, alternate cleaning scenario, allocation choices and unresolved data gaps
Who checks it
Product team and practitioner

Row 5

Step
5. Decide the permitted use
What to send
Review log, open issues, evidence request, deadline and the decision that must stay provisional
Who checks it
Study reviewer and sponsor

A review package links source data and units to the inventory, models, sensitivity tests, and review limits. Missing data need a request, documented proxy or explicit gap.
Keep the inventory sources, assumptions and permitted use in the same handoff.
Diagram text description

Start with versioned bill of materials, supplier, factory, logistics, use and disposal evidence. Preserve original units and reference flows. Request missing data, qualify any proxy and test material assumptions. Deliver scope, inventory, methods, uncertainty, review limits and an owner for the next evidence request.

Start the review by tracing one input from the source record through the inventory and the impact model. Reproduce the selected calculation, then check that it is presented within the study’s boundary. If a key dataset is confidential, arrange access for the person reviewing its fitness and document what others can see.

Deliver the scope note, process map, inventory, model, results and sensitivity file with their versions. Record the next action and what change would trigger a new review. For Rillway, a verified lid lifetime or a different cleaning assumption could change the reference flows and therefore the recommendation.

When to use a climate-only footprint

A product carbon footprint is appropriate when the decision concerns greenhouse gas emissions over a stated life cycle boundary. For example, a factory investigating the climate effect of lower electricity consumption can use a bounded greenhouse gas study, provided the scope and allocation are clear. The GHG Protocol’s Product Standard overview describes how life cycle greenhouse gas accounting can reveal hotspots.

Choose a broader assessment when the decision could move an impact into another category. A cleaning change could alter energy demand, water use and wastewater characteristics. A climate result would address one of those questions. Decide which other potential effects warrant modelling given the study’s goal and evidence.

The same inventory may support both kinds of study, but check the completeness of the exchanges and the model used for each indicator. A footprint factor for a material might give only a climate value and omit the underlying exchanges needed for other categories. Record that gap and get the relevant inventory data before extending the analysis.

When an EPD or public comparison needs more

An environmental product declaration, or EPD, reports product information in a programme’s specified format. A completed LCA can be part of the evidence package. If an EPD is the goal, choose the programme and applicable product category rules early, because those rules can affect the unit, data, modelling and reporting requirements.

Intended useAdditional work to check
Internal design studyDocument goal, assumptions, model limits and the review needed for the decision
EPDConfirm programme and product category rules, prepare the declaration and follow verification and registration requirements
Public product comparisonAlign function, boundary, method, data quality and applicable rules. Establish the critical review and communication requirements

Table 7, shown one row at a time. Every field is retained.

Row 1

Intended use
Internal design study
Additional work to check
Document goal, assumptions, model limits and the review needed for the decision

Row 2

Intended use
EPD
Additional work to check
Confirm programme and product category rules, prepare the declaration and follow verification and registration requirements

Row 3

Intended use
Public product comparison
Additional work to check
Align function, boundary, method, data quality and applicable rules. Establish the critical review and communication requirements

For example, the International EPD System asks the producer to choose the applicable product category rules, complete the study and declaration, obtain independent verification and submit the declaration for registration and publication. Confirm the requirements of the programme you will actually use. A finished internal report does not carry a registration number or demonstrate that those checks were done.

Even two declarations can cover different modules, product versions or assumptions. Read the underlying scope and the applicable category rules before using them in a comparison. Agree the required review before finalising the model, and make sure any public wording describes the service and conditions the evidence actually covers.

Prepare product data for the practitioner

Prepare the product records before the practitioner starts modelling. Pick the SKU and revision. Use the bill of materials to identify which components and packaging belong to it, then map the supplier’s part numbers and production sites. Keep the source reference with each quantity so the practitioner can see why it belongs to this version.

If two suppliers sell parts under the same purchasing code, record which batches use which part and whether the materials or process routes differ. If a spreadsheet merges the two without a defined production mix, resolve the mapping before treating its quantities as one dataset. Put changing versions in separate records and explain which model each one feeds.

The product and supplier files support inventory preparation. The practitioner still needs to define the functional unit and boundary, select datasets and impact methods, and assess uncertainty. Assign a review owner to resolve gaps and record when the evidence is ready for the intended use. Preserve the earlier values when new information changes the study.

Record the decision and the next evidence request

End the study with the decision the evidence supports and the next question that could change it. For Rillway, that might be a decision to test lid durability and collect cleaning data before choosing a design. Record the functional unit, scope and most influential assumptions alongside the decision so another person can tell when it needs reopening.

If a lower result depends on a different lifetime, data source or boundary, align the studies before presenting it as a comparative finding. Where the analysis remains incomplete, state what is missing and which use remains open. Send the study sponsor the current evidence file, the named owner for the next measurement and a review date. That lets the team return to the right phase when the new evidence arrives.