LCA Engineering cheat sheet

LCA is a four-phase method: goal and scope definition, inventory analysis (LCI), impact assessment (LCIA), and interpretation; the phases are iterative, not strictly sequential.

Attributional LCA describes the share of global burdens attributable to a product's life cycle; consequential LCA describes how burdens change in response to a decision, using marginal suppliers.

The life cycle stages follow ISO 14040: raw material acquisition, production, distribution and transport, use and maintenance, and end-of-life treatment, including recycling and energy recovery.

Cradle-to-gate covers raw material extraction through the factory gate; cradle-to-grave adds distribution, use, and disposal; gate-to-gate covers only one process step.

A product system is a set of unit processes connected by flows of intermediate products, each modelled as a process with inputs and outputs measured per reference flow.

Elementary flows cross the system boundary to or from the environment (e.g., kg CO2 to air, kg crude oil in ground); they are not transformed by humans and are not products.

ISO 14040 covers principles and framework; ISO 14044 covers requirements and guidelines. ISO 14040 does not prescribe specific methods; ISO 14044 does.

The three pillars of sustainability (environmental, economic, social) map to E-LCA, LCC, and S-LCA; combining them is Life Cycle Sustainability Assessment (LCSA).

ISO 14044 requires an explicit goal statement including intended application, reasons for the study, intended audience, and whether a comparative assertion disclosed to the public is intended.

A comparative assertion disclosed to the public triggers mandatory critical review by an independent panel (ISO 14044 clause 6); internal studies require only optional review.

ISO 14044 requires allocation to be avoided or reduced where possible, e.g., by dividing the unit process or expanding the system to include co-products' functions.

When allocation is unavoidable, ISO 14044 gives a hierarchy: physical relationships (mass, energy) first, then other relationships such as economic value.

LCIA mandatory elements per ISO 14044: selection of impact categories, classification, characterization; optional elements include normalization, grouping, and weighting.

Transparency is a core requirement: data sources, assumptions, allocation choices, and value judgments must be documented so a third party can reproduce the study.

ISO 14044 requires a sensitivity analysis and, where relevant, uncertainty analysis, and requires that limitations be identified and reported in the interpretation phase.

The standard does not mandate a single LCIA method or a single database; method choice must be justified against the study goal and documented in the report.

The goal defines the intended application, the reasons for carrying out the study, the target audience, and whether the results will be used in a comparative assertion disclosed to the public.

The scope defines the product system, functional unit, system boundary, allocation procedures, impact categories and LCIA method, data requirements, and review type.

Scope must state the reference flow, i.e., the amount of product needed to fulfil the functional unit, and the reference year and geography of the data.

System boundary decisions must state which life cycle stages, capital goods, and infrastructure are included, and which are excluded, with justification.

Cut-off criteria must be quantified (e.g., mass, energy, and environmental significance thresholds) and their effect on results checked; 'negligible' without numbers is not acceptable.

The level of detail and the choice of data (site-specific vs generic) must be consistent with the goal; a screening study and a public comparative study need different rigour.

If the study is intended for an EPD, the scope must follow the relevant product category rules (PCR), which fix the functional unit, boundary, and LCIA indicators.

Iteration is expected: goal and scope may be revised as data become available, provided the change and its rationale are documented.

The functional unit is the quantified performance of a product system for use as a reference unit; it must be measurable and relate to the function, not the product mass.

Example: 'provide 1 m2 of wall with a thermal resistance of R = 3.0 m2K/W for 50 years' is a functional unit; '1 kg of insulation' is not.

The reference flow is the quantity of product (in kg, m3, or pieces) required to deliver one functional unit; all inventory and impact results are scaled to it.

Comparisons are valid only when systems deliver the same functional unit and comparable quality of function; differing service life, performance, or use patterns must be normalized.

System boundary types: cradle-to-gate, cradle-to-grave, gate-to-gate, and cradle-to-cradle; the boundary must be stated explicitly and drawn consistently for all compared systems.

Expanding the system boundary to include additional functions of co-products is a way to avoid allocation; the added functions must be credited to the compared system.

Capital goods (machinery, buildings, infrastructure) are included only if the goal requires it; exclusion must be justified and tested for significance.

Geographic and temporal boundaries (e.g., electricity grid mix year, transport distances) must be stated because they strongly affect results.

LCI compiles all inputs (energy, water, materials) and outputs (products, co-products, emissions, waste) for each unit process, normalized to the reference flow.

Data collection follows the ISO 14044 steps: identify unit processes, collect data per process, validate data, relate data to the functional unit, and refine the system boundary.

Allocation hierarchy: (1) avoid allocation by subdivision or system expansion; (2) allocate on physical relationships (mass, energy, stoichiometry); (3) allocate on other relationships such as economic value.

Mass allocation divides a shared burden by mass share of co-products; economic allocation divides by revenue share; results can differ greatly, so the chosen method must be justified.

For recycling, common approaches are the cut-off (recycled content) method, the avoided-burden (end-of-life recycling) method, and the 50/50 or circular footprint formula used in PEF.

Multi-functionality of processes (e.g., combined heat and power, refineries) must be resolved before results are aggregated; the method must be applied consistently across the study.

Data validation includes mass and energy balance checks per process, plausibility checks against similar processes, and comparison with literature or database values.

Elementary flow data should be linked to consistent background databases (e.g., ecoinvent, GaBi) to avoid double counting and to keep the system model coherent.

Classification assigns each elementary flow to one or more impact categories; characterization converts flows to a common unit via a characterization factor (CF).

Characterization result = sum over flows of (inventory amount x CF); e.g., GWP = sum of kg of each GHG x its 100-year global warming potential.

Midpoint indicators are at the cause-effect chain level (e.g., kg CO2-eq, kg SO2-eq, kg N-eq); endpoint indicators model damage to human health, ecosystems, and resources (e.g., DALY, species-yr).

Common methods: CML (midpoint, problem-oriented), ReCiPe (midpoint and endpoint), TRACI (US EPA), IMPACT World+, and EF/PEF (Environmental Footprint).

GWP values come from IPCC assessment reports; AR5 uses 28 for fossil methane and AR6 uses 27.9 (fossil methane, 100-yr), so the chosen IPCC version must be stated.

Normalization divides characterized results by reference values (e.g., per-capita emissions in a region); weighting aggregates categories using value-based factors and must be transparent.

Optional LCIA elements per ISO 14044: normalization, grouping, weighting, and data quality analysis; weighting is not permitted in comparative assertions disclosed to the public in some schemes.

LCIA results are relative expressions of potential impacts, not predictions of actual damage; they do not model spatial or temporal exposure unless the method explicitly does.

Interpretation per ISO 14044 has three elements: identification of significant issues, evaluation (completeness, sensitivity, consistency checks), and conclusions, limitations, and recommendations.

Sensitivity analysis varies a key parameter (e.g., allocation method, electricity mix, transport distance) and reports how results change; it identifies which assumptions drive conclusions.

Uncertainty analysis quantifies variability in inputs and propagates it (e.g., Monte Carlo simulation) to give a range or distribution around results.

Scenario analysis compares alternative plausible futures or system configurations rather than varying a single parameter.

Pedigree matrix data quality scoring rates data on reliability, completeness, temporal, geographic, and technological correlation; scores feed uncertainty factors in ecoinvent.

Consistency check verifies that the same method, boundary, and data quality were applied to all compared systems; a difference in method invalidates a comparison.

Completeness check confirms that all relevant flows and processes within the boundary were included, and documents any gaps and their likely effect.

Critical review types: internal expert review, external expert review, and review by interested parties (panel); the choice depends on the goal and on whether a public comparative assertion is made.

Common LCA software: SimaPro, GaBi (now Sphera), openLCA (free, open source), Umberto, and Brightway2 (Python-based, open source).

Common databases: ecoinvent (Switzerland, broad coverage), GaBi Databases (Sphera), Agribalyse (agriculture, France), US LCI (US EPA), ELCD, and EF datasets.

ecoinvent system models include cut-off (recycled content), APOS (allocation at the point of substitution), and consequential; the choice changes recycling and co-product burdens.

Data quality dimensions: reliability, completeness, temporal correlation, geographic correlation, technological correlation, and sample size; each is scored and documented.

Foreground data describe the specific product system under study (collected by the practitioner); background data describe upstream processes and are taken from databases.

Database version and release year must be cited because results change between versions (e.g., ecoinvent v3.8 vs v3.10) and comparability requires the same version.

Data gaps should be filled with proxy data, estimated values, or excluded with justification; each choice must be documented and its influence assessed.

OpenLCA uses the ILCD/ecospold formats and can import ecoinvent; it supports LCIA methods such as ReCiPe, CML, and EF via method packages.

ISO 14067 specifies principles, requirements, and guidelines for quantifying and reporting the carbon footprint of a product (CFP), based on ISO 14040/14044.

A carbon footprint covers only greenhouse gases, reported as kg CO2-eq using IPCC GWP values; the chosen IPCC version and time horizon (usually 100 years) must be stated.

ISO 14067 requires the CFP to be reported per functional unit or per declared unit, with system boundary, data quality, and any offsets reported separately, not netted.

ISO 14025 defines Type III environmental declarations (EPDs): quantified environmental data for a product, based on PCR, verified by an independent third party, and intended for business-to-business communication.

An EPD is developed from an LCA, follows the relevant PCR for the product category, and is verified by an approved verifier before registration and publication.

Type I labels (ISO 14024) are third-party certified, multi-criteria, and based on life cycle considerations (e.g., Nordic Swan, Blue Angel).

Type II labels (ISO 14021) are self-declared environmental claims by the manufacturer; they must be specific, verifiable, and not misleading.

PCR define the product category rules: functional or declared unit, system boundary, cut-off rules, LCIA indicators, and required reporting modules for an EPD.

The circular economy aims to keep materials in use through reuse, repair, remanufacturing, and recycling, and to design out waste and pollution; LCA quantifies whether a circular strategy actually reduces impacts.

The waste hierarchy (prevention, reuse, recycling, recovery, disposal) is a qualitative guide; LCA can show cases where recycling has higher burdens than energy recovery due to transport or processing.

Eco-design integrates environmental considerations into product design; LCA identifies the life cycle stage with the dominant impact (hotspot) to target design changes.

Design for disassembly, material selection, and design for recycling reduce end-of-life burdens; the benefit must be verified over the full life cycle, not assumed.

The circular footprint formula (PEF) accounts for recycled content, recycling at end-of-life, and the quality of secondary materials via a quality factor.

LCSA combines E-LCA (environmental), LCC (life cycle costing, economic), and S-LCA (social) using the same functional unit and system boundary for consistency.

S-LCA uses social impact subcategories (e.g., worker health and safety, child labour, local community) assessed via activity variables and performance reference points.

Rebound effects and market-mediated effects (consequential modelling) can offset circular economy benefits; they should be considered when the decision changes consumption patterns.