How to Conduct a Life Cycle Assessment?

Conducting an LCA follows the four-phase structure defined by ISO 14040. Each phase builds on the previous one, with iterations refining your assessment as you learn more about the system.

Phase 1: Goal and Scope Definition#

Start by defining why you’re conducting the study and what you need to learn. The goal statement should specify:

• Intended application (product comparison, hotspot identification, EPD creation)
• Reasons for conducting the study
• Target audience
• Whether results will be publicly disclosed

Define the Functional Unit#

The functional unit creates your reference point. Instead of comparing “one product” you compare “the function that product provides”.

For packaging, the functional unit might be “containing 1 litre of beverage for 30 days”. For transport, “moving one person 10 kilometres”. For buildings, “providing 100 m² of usable floor space for 50 years”.

The functional unit must be measurable and relevant to how products actually get used.

Set System Boundaries#

Decide what’s included in your assessment. Will you conduct cradle-to-grave, cradle-to-gate, or gate-to-gate? Define geographical scope (global average, specific region, site-specific). Specify temporal boundaries (current, historical, or future conditions).

Document any exclusions. If you’re cutting off processes contributing less than 1% of mass or energy, state this criterion clearly.

Specify Data Requirements#

Determine what data quality you need. Primary data comes from direct measurement. Secondary data comes from databases. Specify which processes need primary data and where secondary data suffices.

Set data quality requirements for temporal, geographical, and technological representativeness. Data from ten years ago might not represent current conditions. European data might not suit Asian production.

Phase 2: Life Cycle Inventory (LCI)#

LCI collects input and output data for every process in your system. This phase typically consumes most project time and resources.

Map the Process Flow#

Create a diagram showing all processes from raw material extraction through disposal. Each process box needs inputs (materials, energy) and outputs (products, emissions, waste).

Start with foreground processes you control directly. Add background processes for materials, energy, and transport. Connect processes through product and waste flows.

Collect Process Data#

For foreground processes, collect primary data:

• Material quantities (kg, m³)
• Energy consumption (kWh, MJ)
• Transport distances and modes
• Manufacturing yields and scrap rates
• Packaging materials
• Direct emissions and waste generation

Measurement beats estimation where possible. Production records, energy bills, and material purchase orders provide reliable data.

For background processes, use LCA databases:

• ecoinvent for global coverage
• GaBi for industrial processes
• ELCD for European data
• USLCI for United States data

Handle Multifunctional Processes#

When processes generate multiple products, you need allocation rules. Follow ISO 14044’s hierarchy:

1. Avoid allocation through subdivision or system expansion where possible
2. If unavoidable, allocate based on physical relationships (mass, energy content)
3. Use economic allocation when physical relationships don’t apply

Document your allocation method and justify the choice.

Check Data Quality#

Assess data representativeness, completeness, and consistency. Are your data sources reliable? Do you have data for all significant processes? Are units and system boundaries consistent across data sources?

Identify data gaps early. Missing data for significant processes weakens results. Plan how you’ll address gaps through additional collection or justified estimates.

Phase 3: Life Cycle Impact Assessment (LCIA)#

LCIA converts your inventory of emissions and resources into environmental impacts.

Select Impact Categories#

Choose categories relevant to your goal. Standard sets include:

• Climate change
• Ozone depletion
• Acidification
• Eutrophication (freshwater and marine)
• Photochemical ozone formation
• Particulate matter formation
• Human toxicity
• Ecotoxicity (freshwater, marine, terrestrial)
• Water scarcity
• Resource depletion
• Land use

Product Category Rules or regulatory requirements might specify which categories to assess.

Choose Impact Assessment Method#

Select a characterisation method appropriate for your geographical scope:

• ReCiPe for comprehensive global coverage
• CML for established methodology
• TRACI for North American focus
• Environmental Footprint for EU alignment

The choice affects results, particularly for toxicity categories. Document your selection and justify it.

Calculate Impact Scores#

LCA software applies characterisation factors to your inventory data. Each emission or resource use gets multiplied by its impact factor and summed within categories.

Climate change example: CO₂ gets factor 1, methane gets factor 29.8 (100-year GWP), nitrous oxide gets 273. Sum these to get total climate change impact in kg CO₂-eq.

Optional Normalisation and Weighting#

Normalisation compares your results to reference values (e.g., average per capita impacts). This shows relative magnitude across categories.

Weighting assigns importance factors to combine categories into a single score. This involves value judgements and shouldn’t be used for publicly disclosed comparative assessments without clear justification.

Phase 4: Interpretation#

Interpretation analyses results, checks reliability, and draws conclusions.

Identify Hotspots#

Determine which life cycle stages and which impact categories matter most. Does raw material production dominate? Is the use phase critical? Focus improvement efforts where they matter.

Break results down by contribution analysis. Which processes contribute most to each impact category? This reveals improvement priorities.

Conduct Sensitivity Analysis#

Test how methodological choices affect results:

• Change allocation methods
• Use different impact assessment methods
• Vary uncertain parameters
• Modify boundary assumptions

Robust conclusions survive these variations. Sensitive results need careful interpretation and clear communication about uncertainty.

Check Completeness and Consistency#

Review data coverage. Have you included all significant processes? Are boundaries consistent? Does your functional unit actually represent the comparison you intended?

Check calculation consistency. Do units match? Are conversion factors correct? Do process links carry the right quantities?

Draw Conclusions#

State findings clearly. Which product performs better? Where are improvement opportunities? What uncertainties affect conclusions?

Recommendations should follow from your analysis. If the use phase dominates impacts, focus on energy efficiency. If raw materials matter most, consider material substitution.

Documentation#

ISO 14040 requires thorough documentation:

• Goal and scope definition
• System boundaries and exclusions
• Data sources and quality
• Allocation procedures
• Impact assessment methods
• Results and interpretation
• Limitations and assumptions

Transparent documentation enables critical review and allows others to understand and potentially reproduce your work.

Verification#

Third-party verification adds credibility. A qualified reviewer checks methodology, data quality, and conclusions. This matters for publicly disclosed comparative assertions or EPD publication.

Critical review follows ISO 14040 requirements. The reviewer verifies that methods follow standards, data quality is appropriate, and interpretation is reasonable.

Common Mistakes#

Watch for these pitfalls:

• Poorly defined functional units that don’t enable fair comparison
• Inconsistent system boundaries across alternatives
• Missing significant processes due to incomplete mapping
• Using inappropriate allocation methods without justification
• Cherry-picking impact categories to favour one option
• Overinterpreting results given data uncertainty
• Insufficient documentation preventing reproducibility

Software Tools#

LCA software handles the calculation complexity:

• openLCA: open-source, supports major databases
• SimaPro: widely used commercial tool
• GaBi: comprehensive industrial database
• Umberto: material flow analysis integration

Software doesn’t eliminate the need for expertise. You still make judgements about boundaries, allocation, and interpretation. The tool executes your methodology, not the other way around.

When to Seek Help#

LCA requires specialised knowledge. Consider consulting support when:

• Conducting your first LCA
• Results will be publicly disclosed or verified
• Allocation problems are complex
• Your product system is unusual
• Regulatory compliance requires certification

Even experienced practitioners benefit from peer review. Complex allocation decisions, boundary choices, and interpretation all benefit from independent challenge.