LCIA Method Selection: ReCiPe, CML, TRACI, or EF?

"Which LCIA method should I use?" is a question that confuses even experienced practitioners. This guide explains the differences between major methods and provides practical selection criteria.

Understanding LCIA Methods

What is an LCIA method? An LCIA method is a package containing:

• Impact category definitions
• Characterization factors for elementary flows
• (Optional) Normalization references
• (Optional) Weighting factors

Why so many methods? Different methods reflect:

• Different scientific models
• Different geographic contexts
• Different policy priorities
• Evolution of scientific understanding

Major LCIA Methods Compared

FAQ: LCIA Method Selection

"Which LCIA method should I use: ReCiPe, CML, TRACI, or EF?"

Quick decision framework:

Method profiles:

ReCiPe 2016 - The versatile choice

• Pros: Comprehensive, both midpoint and endpoint, globally applicable
• Cons: More complex, three perspective variants
• Use when: You need flexibility and thorough analysis

CML-IA - The academic standard

• Pros: Well-established, widely cited, conservative approach
• Cons: Older (though updated), fewer categories than modern methods
• Use when: Academic publication, when tradition matters

TRACI 2.1 - The US choice

• Pros: US-specific characterization, EPA-developed
• Cons: Only applicable to US context
• Use when: US-based products, US regulatory context

EF 3.0 - The EU regulatory method

• Pros: Official EU method, harmonized with policy
• Cons: Mandatory for PEF, may not suit non-EU studies
• Use when: EU Product Environmental Footprint, EU compliance

"Why do midpoint and endpoint results differ so much?"

Understanding the difference:

Midpoint = Impact at the mechanism level

• "How much acidification potential?"
• Lower uncertainty, less intuitive
• Example: 15 kg SO₂-eq

Endpoint = Damage at the protection level

• "How much harm to ecosystems?"
• Higher uncertainty, more intuitive
• Example: 0.0003 species × year lost

Why results "differ":

They measure different things along the same cause-effect chain:

Emission → [Midpoint] → Damage mechanism → [Endpoint] → Actual harm
   │           │                                │
   │      "Acidification                  "Ecosystem
   │       Potential"                       damage"
   │                                            │
   └────────── Lower uncertainty ───────────────┘
                        │
              Higher interpretation needed

Example: Climate change

The endpoint modeling chain adds uncertainty:

• Climate models (temperature change per kg CO₂)
• Impact models (health effects per degree)
• Valuation models (DALYs per health effect)

Each step adds assumptions and uncertainty.

When to use each:

"How do I interpret normalization and weighting results?"

Normalization puts results in context:

Normalized value = Impact score / Reference value

Reference values are typically annual per-capita or regional totals.

Example:

• Your product's GWP: 10 kg CO₂-eq
• EU per-capita annual GWP: 8,000 kg CO₂-eq
• Normalized: 10/8,000 = 0.00125 person-equivalents

What normalization tells you:

• Which categories are relatively significant
• How your product compares to "average" activities
• NOT which impacts are more important

Weighting applies value judgments:

Weighted value = Normalized value × Weight factor

Weights reflect importance rankings, typically derived from:

• Expert panels
• Policy priorities
• Damage costs
• Distance-to-target

Critical notes on weighting:

Appropriate weighting use:

• Internal decision-making
• Screening and prioritization
• Sensitivity analysis ("how would ranking change?")
• Communication with clear caveats

"What impact categories are most relevant for my product type?"

Category relevance by product sector:

Standard category sets:

Minimum set (most studies):

1. Climate change (GWP)
2. Ozone depletion
3. Acidification
4. Eutrophication (freshwater + marine)
5. Photochemical ozone formation
6. Resource depletion (fossil)

Extended set (comprehensive): Add:

• Human toxicity (cancer + non-cancer)
• Ecotoxicity (freshwater)
• Particulate matter
• Water use/scarcity
• Land use
• Ionizing radiation

Sector-specific additions:

• Mining: Resource depletion (minerals)
• Agriculture: Biodiversity, soil quality
• Marine products: Marine eutrophication
• Electronics: Rare earth depletion

Method Deep Dives

ReCiPe 2016

Structure:

• 18 midpoint categories
• 3 endpoint damage categories (Human Health, Ecosystems, Resources)
• 3 cultural perspectives (Individualist, Hierarchist, Egalitarian)

Perspectives explained:

Default choice: ReCiPe 2016 Midpoint (H) - Hierarchist

CML-IA

Structure:

• 11 midpoint categories (baseline set)
• Additional categories in extended set
• No endpoint aggregation

Key characteristics:

• Conservative scientific approach
• Widely cited in literature
• Good for comparability with older studies

Categories:

• Abiotic depletion (elements + fossil)
• Global warming
• Ozone depletion
• Human toxicity
• Freshwater aquatic ecotoxicity
• Marine aquatic ecotoxicity
• Terrestrial ecotoxicity
• Photochemical oxidation
• Acidification
• Eutrophication

TRACI 2.1

Structure:

• 10 impact categories
• US-specific characterization factors
• Midpoint only

US-specific aspects:

• Ozone depletion uses US background
• Smog based on US conditions
• Toxicity using US population data
• Eutrophication for US water bodies

Categories:

• Ozone depletion
• Global warming
• Smog
• Acidification
• Eutrophication
• Human health (cancer + non-cancer)
• Respiratory effects
• Ecotoxicity
• Fossil fuel depletion

EF 3.0 (Environmental Footprint)

Structure:

• 16 impact categories (mandatory set)
• Normalization references (EU + global)
• Weighting factors (for single score, internal use only)

Regulatory context:

• Required for EU PEF studies
• Specified in EU Green Claims Directive proposals
• Linked to EU policy objectives

Key features:

• Climate change split: fossil, biogenic, land use change
• Water scarcity uses AWARE method
• Consistent with EU policy definitions

Practical Method Selection

Decision Flowchart

Start
   │
   ▼
Is this for EU PEF compliance?
   │          │
  Yes         No
   │          │
   ▼          ▼
Use EF 3.0   Is this US-focused?
                │          │
               Yes         No
                │          │
                ▼          ▼
           Use TRACI   Academic publication?
                          │          │
                         Yes         No
                          │          │
                          ▼          ▼
                    Use CML-IA   Use ReCiPe 2016
                    or ReCiPe      (H) as default

Running Multiple Methods

Best practice: Run at least two methods and compare.

Comparison checklist: ☐ Are rankings consistent between methods? ☐ Do hotspots appear in same life cycle stages? ☐ Are conclusions robust to method choice?

If results differ significantly:

• Report the range
• Discuss which method is more appropriate
• Consider category-specific method selection

Software Implementation

openLCA: All major methods available via Nexus

SimaPro: Most methods built-in, some via updates

GaBi: Proprietary method variants + standards

Key Takeaways

1. No single "best" method—choose based on context
2. EF 3.0 is mandatory for EU PEF studies
3. TRACI is required for US-specific characterization
4. ReCiPe is versatile and widely applicable
5. Midpoint has lower uncertainty; endpoint is more intuitive
6. Weighting is prohibited for public comparative claims
7. Test multiple methods to check robustness

Method Selection Checklist

Before finalizing your LCIA method:

☐ Check if regulation/standard mandates a specific method ☐ Consider geographic scope of your study ☐ Confirm method is available in your software ☐ Verify compatibility with your database version ☐ Decide if midpoint, endpoint, or both are needed ☐ Plan to test at least one alternative method ☐ Document rationale for method selection

---

Next Steps

With LCIA methods understood, the next lesson is our Process Modeling Cookbook—practical recipes for modeling common processes like electricity, transportation, and manufacturing.