CVEN 5019 · Integrated Core · Fall 2026 · Week 4
Process-Level
Life Cycle Assessment
From GHG accounting to full environmental impact — ISO 14040/14044 methodology, ecoinvent, OpenLCA, and comparative analysis of real technology systems.
Instructor Carlo Salvinelli
LCA Assignment distributed this week — due end of Week 5
Software OpenLCA — bring a laptop
Session Overview
Topics
- What is LCA? Process-level vs. economic input-output LCA
- ISO 14040/14044 — the four-phase framework
- Phase 1: Goal & scope, functional unit, system boundaries
- Phase 2: Life cycle inventory (LCI) — data collection and databases
- Phase 3: Life cycle impact assessment (LCIA) — methods and categories
- Phase 4: Interpretation — hotspots, sensitivity, uncertainty
- OpenLCA + ecoinvent demo
- Comparative LCA: EV vs. ICE vehicle with real data
- LCA limitations and relationship to GHG accounting
Learning Objectives
- Explain the four phases of ISO 14040 LCA and the role of each
- Define and defend a functional unit and system boundary for a given technology
- Distinguish process-level LCA from economic input-output LCA and explain when each applies
- Apply ReCiPe or CML impact characterization to a simple LCI
- Identify emission hotspots and design implications from a comparative LCA
- Use OpenLCA with ecoinvent to calculate a basic product carbon footprint
What Is Life Cycle Assessment — and Why Process-Level?
- LCA: a systematic method to evaluate the environmental impacts associated with all stages of a product's or technology's life — from raw material extraction through end-of-life (ISO 14040)
- Why not just GHG accounting? Climate change is one impact category among many. A switch from gas heating to electric heat pumps may reduce CO₂ but increase water depletion (power plant cooling) and mineral extraction impacts (copper, rare earths)
- Process-level LCA builds the inventory from the bottom up — unit process by unit process — using primary data or process databases. More accurate but more data-intensive.
- Economic input-output (EIO) LCA: uses national economic tables to estimate impacts from spending. Much broader coverage, rapid, but coarser — appropriate for Scope 3 estimates, not product design
What LCA Can and Cannot Do
LCA Can
Compare alternative technologies on multiple environmental dimensions simultaneously. Identify where in the lifecycle impacts are concentrated. Guide design decisions early in development.
LCA Cannot
Determine whether a product is "sustainable" in absolute terms. Account for biodiversity, land use change dynamics, or all social impacts without extensions. Replace engineering judgment.
LCA is a comparative tool. Its value comes from consistent application to competing alternatives — not from absolute numbers in isolation.
ISO 14040/14044 — The Four-Phase Framework
Phase 1
Goal & Scope Definition
What question are we answering? Function, functional unit, system boundary, cut-off criteria, data quality requirements
Phase 2
Life Cycle Inventory (LCI)
Data collection: inputs (energy, materials, water) and outputs (emissions, waste) for every unit process in the system boundary
Phase 3
Life Cycle Impact Assessment (LCIA)
Translate inventory flows into impact scores using characterization factors — climate change, water use, toxicity, land use, etc.
Phase 4
Interpretation
Hotspot analysis, sensitivity and uncertainty analysis, conclusions, limitations, and recommendations for decision-makers
ISO 14040 defines requirements; ISO 14044 provides detailed guidance on methodology. Both are required reading for any LCA practitioner. ISO 14067 extends the framework specifically for product carbon footprints.
Phase 1 — Goal & Scope Definition
Goal Statement Must Define
- Intended application: comparative assertion, internal improvement, eco-labeling, investor disclosure? Each has different requirements.
- Target audience: engineers, procurement managers, regulators, public? Determines level of detail and communication format.
- Reason for carrying out the study: be explicit — comparative LCAs for public disclosure require third-party critical review (ISO 14044 §6.1)
Scope Decisions
- Cut-off rules: processes contributing <1% of mass/energy/environmental relevance may be excluded — but must be justified and documented
- Cradle-to-gate: raw material extraction through manufacturing only (common for product LCA components)
- Cradle-to-grave: includes use phase and end-of-life — required for comparative assertions about use-phase-dominant technologies (EVs, appliances)
The Functional Unit — Most Critical Decision
Definition
A quantified description of the function that the product or service delivers. All inputs and outputs are normalized to this unit.
Example: Transport
"Transport of 1 passenger over 1 km for 10 years with 90% reliability" — not just "one car." Includes the function, quantity, quality, and duration.
Example: Electricity
"Delivery of 1 kWh of electricity to the consumer at grid reliability" — not "1 kWh generated," which excludes transmission losses.
Why it matters
Changing the functional unit changes the result. A CFL vs. LED comparison changes drastically depending on whether you compare per lumen-hour or per unit sold.
System Boundaries — What's In, What's Out
System Boundary — Example: Lithium-Ion Battery Pack
Lithium mining
Chile/Australia
→
Lithium carbonate
processing
→
Cathode material
manufacturing
→
Cell assembly
(China/Korea/US)
→
Pack assembly
& BMS
→
Use phase
charging cycles
→
End of life:
recycling / landfill
- Upstream boundary: How far back in the supply chain? For lithium batteries, cobalt from DRC artisanal mining has very different impacts than cobalt from large-scale Chilean operations — the choice matters enormously
- Capital equipment: typically excluded (<0.1% of lifecycle impacts for most systems) unless exceptionally long-lived or energy-intensive manufacturing (e.g., semiconductor fab)
- Infrastructure: road networks, electricity grid typically excluded — consistent with allocation convention
- Multi-output processes require allocation: when a process produces multiple outputs (crude oil refining → gasoline, diesel, jet fuel, plastics), how do you allocate the inputs/emissions?
- Allocation approaches: mass allocation, economic allocation, energy allocation, system expansion (preferred by ISO for most cases)
- End-of-life modeling: 50:50 cut-off rule, recycled content method, or end-of-life recycling rate — each gives a different result for high-recyclability materials like aluminum
- Consistency in boundary decisions across compared systems is more important than perfection in any single choice
Phase 2 — Life Cycle Inventory (LCI)
- LCI = the data collection phase: for every unit process within the system boundary, document all inputs (materials, energy, water) and outputs (products, emissions to air/water/soil, waste)
- Primary data: measured from the actual process (plant metering, supplier EPDs, direct sampling) — highest quality, most expensive to collect
- Secondary data: from LCA databases representing industry averages or literature values — necessary for most upstream supply chain processes
- Data quality criteria (ISO 14044): time-related coverage (within 5–10 years), geographical coverage (region-specific), technology coverage (representative of actual technology mix), precision, completeness, consistency
Key LCA Databases
ecoinvent v3.9
Industry standard — 20,000+ background processes, global and regional datasets, multiple allocation system models (Cutoff, Consequential, APOS). Swiss Centre for LCA. Academic license available.
US LCI Database
NREL/DOE database for US-specific processes. Free. Covers energy, metals, plastics, agriculture. Less comprehensive than ecoinvent.
GaBi / Sphera
Industry LCA database with strong coverage of chemicals, polymers, metals. Commercial license. Widely used in automotive and consumer goods sectors.
OpenLCA
Free, open-source LCA software (GreenDelta). Compatible with ecoinvent, US LCI, and others. Used in this course — installation guide on Canvas.
Phase 3 — Life Cycle Impact Assessment (LCIA)
LCIA translates raw LCI flows (kg of CO₂, liters of water, m² of land, etc.) into meaningful impact scores using characterization factors.
| Impact Category | Unit | Examples of Contributing Flows |
|---|
| Climate Change | kg CO₂-eq | CO₂, CH₄, N₂O, HFCs |
| Water Consumption | m³ world-eq | Freshwater withdrawal, evapotranspiration |
| Terrestrial Acidification | kg SO₂-eq | SO₂, NOₓ, NH₃ deposition |
| Freshwater Eutrophication | kg P-eq | Phosphorus to freshwater |
| Human Toxicity (cancer) | CTUh | Carcinogenic substances to air/water/soil |
| Fine Particle Formation | kg PM₂.₅-eq | PM₂.₅, SO₂, NOₓ precursors |
| Land Use | m² · yr crop-eq | Agricultural, urban, and forest land occupation |
| Mineral Resource Scarcity | kg Cu-eq | Ore grade depletion for metals |
LCIA Methods — Which to Choose?
ReCiPe 2016
Most widely used globally. Midpoint (18 categories) + endpoint (damage to human health, ecosystem quality, resource availability). Default in most studies.
CML-IA Baseline
European standard, midpoint only, conservative characterization. Common in EU regulatory contexts and EPDs.
TRACI 2.1
US EPA method, US-specific characterization factors. Recommended for North American product systems by EPA.
EF 3.1
EU Environmental Footprint — mandated for EU Product Environmental Footprints (PEF) and Category Rules. 16 impact categories.
Phase 4 — Interpretation
Hotspot Identification
- Which life cycle stage contributes >80% of a given impact category? Often called a "hotspot" — highest priority for design improvement
- Example (battery EV): manufacturing stage dominates mineral resource depletion and human toxicity; use phase dominates climate change (depends on grid); end-of-life dominates water quality (if batteries landfilled)
- Contribution analysis: decompose results by process or material — e.g., how much of climate change comes from steel vs. aluminum vs. battery cells?
Sensitivity Analysis
- How much do results change when a key assumption changes? E.g., vary the electricity grid carbon intensity from 100 g CO₂e/kWh (Norway hydro) to 800 g CO₂e/kWh (coal grid)
- Identifies which inputs most influence the conclusion — and which conclusions are robust regardless of uncertainty
Uncertainty Analysis
- Parameter uncertainty: range of possible values for emission factors, process yields, transportation distances
- Model uncertainty: choice of allocation method, system boundary, characterization method
- Scenario uncertainty: which future grid mix? Which end-of-life pathway? Which use pattern?
- Monte Carlo simulation: OpenLCA supports stochastic analysis — run 1,000+ iterations varying all uncertain parameters within defined distributions to generate result probability distributions
A good LCA interpretation section is honest about what the study cannot tell you — just as important as what it can.
Comparative LCA — Electric vs. ICE Vehicle
Functional unit: Transport of 1 person over 200,000 km vehicle lifetime. Data: ecoinvent v3.9, US LCI, NREL GREET. Based on representative 2023 mid-size sedan (Tesla Model 3 vs. Toyota Camry).
Lifecycle CO₂e
(kg CO₂e / 200k km)
–52%
Manufacturing CO₂e
(kg CO₂e / vehicle)
EV higher: ~8.8t vs. ~6.4t
+38%
Water Consumption
(m³ / 200k km)
context-dep.
Mineral Resource
(kg Cu-eq)
EV ~4× ICE (Li, Co, Ni, Cu)
+300%
Human Toxicity
(CTUh)
+50–100%
Key insight: EVs have significantly lower lifecycle climate impact on the US grid — but the trade-offs on minerals, water, and toxicity require full LCA to understand. The breakeven point for climate is ~20,000–50,000 km depending on grid intensity.
🔬
Week 4 · OpenLCA Workshop (30 min)
Calculate the carbon footprint of 1 kWh of electricity from three different sources using OpenLCA and ecoinvent v3.9.
Pre-requisite: OpenLCA installed (free at openlca.org) and ecoinvent 3.9 database imported (academic license via CU Boulder — instructions on Canvas).
Tasks:
1. Open the ecoinvent 3.9 Cutoff database in OpenLCA
2. Search for: "electricity, high voltage, production mix, US" → run with ReCiPe 2016 Midpoint (H)
3. Repeat for: "electricity production, wind, 1-3MW turbine, onshore, US" and "electricity production, photovoltaic, 3kWp slanted-roof installation, multi-Si, US"
4. Compare results across at least 4 impact categories (climate change, water consumption, mineral resource scarcity, land use)
5. Identify which electricity source wins on each impact category
6. What does this tell you about the risk of optimizing for a single impact category?
Work in pairs. LCA Assignment distributed at end of class — due end of Week 5.
LCA Limitations — What to Watch For
- Data quality and age: ecoinvent background datasets for some regions and materials may be 10+ years old — coal power in China in ecoinvent v3.5 is significantly different from current reality
- Geographic aggregation: national averages miss regional variation — a manufacturing plant in Wyoming (coal-heavy grid) is very different from one in Washington state (hydro-dominant)
- Dynamic LCA: standard LCA is static — it doesn't capture how impacts change over time (e.g., grid decarbonization during an EV's 15-year lifetime). Prospective LCA methods are emerging but not standardized.
- Biodiversity: characterization factors for biodiversity loss are highly uncertain — species sensitivity distributions (SSDs) exist but are incomplete for many chemicals and regions
- Social impacts excluded: standard LCA covers environmental impacts only — worker health and safety, community disruption, and supply chain labor conditions require separate Social LCA methodology (covered in Week 6)
- Use phase behavior: ICE vehicle MPG varies widely with driving patterns — LCA results depend on assumed usage. Sensitivity analysis on use-phase assumptions is critical.
- "LCA-washing": selective reporting of favorable impact categories while hiding trade-offs. Comparative assertions for public disclosure require ISO 14044 §6.1 third-party critical review to prevent this.
Despite limitations, LCA remains the most comprehensive, standardized method available for multi-dimensional environmental comparison — no substitute exists for product design decisions.
📋
LCA Assignment — Distributed This Week · Due End of Week 5
Conduct a process-level comparative LCA of two competing technology options for the same function — and present design recommendations based on your findings.
Technology pairs to choose from (or propose your own):
A) Natural gas boiler vs. air-source heat pump for residential space heating in Colorado
B) Lithium-ion vs. vanadium flow battery for 4-hour grid storage application
C) Conventional concrete vs. low-carbon concrete (GGBS blend) for a structural column
D) Cotton vs. recycled polyester T-shirt (per wearing occasion over 2-year garment life)
Deliverables:
• Defined functional unit, system boundary (with justification), and data quality assessment
• OpenLCA model with ecoinvent background data — screenshot evidence required
• Results for at least 5 impact categories using ReCiPe 2016 Midpoint (H)
• Hotspot analysis and sensitivity analysis on 2 key assumptions
• Design recommendation: which technology is preferable and why — with clear caveats
• 6–8 page report + OpenLCA model file submitted to Canvas
Engineering students: include process-level analysis of the highest-impact manufacturing stage. Business students: include supply chain sourcing recommendations based on LCA hotspots.
Key Takeaways — Week 4
- LCA is multi-dimensional: climate change is just one of 15–18 impact categories. Technology decisions that optimize for CO₂ alone may create serious trade-offs in water, minerals, toxicity, or land use.
- The functional unit is the most consequential methodological choice — it defines what you're comparing and therefore what the results mean. Always scrutinize it when reading others' LCA studies.
- Process-level LCA built from ecoinvent provides the highest quality background data — but is only as good as the foreground (primary) data you supply for the specific technology under study.
- EV vs. ICE LCA: EVs win on lifecycle climate change on most grids, but lose on minerals, water, and manufacturing toxicity. The full picture requires all categories.
- Comparative LCAs for public disclosure must be critically reviewed by an independent panel (ISO 14044 §6.1) — this is a quality safeguard, not bureaucracy.
- OpenLCA + ecoinvent is your primary tool for this course — invest time learning the software this week. Tutorial videos on Canvas.
Next week: Net Zero Pathways & Sustainable Procurement — how organizations translate LCA insights and GHG accounting into actionable net zero strategies and supply chain commitments. LCA Assignment due end of Week 5.