Reproducibility benchmark · Case study
Brightway deterministic LCIA reference
A closed-form two-process product system used to verify Brightway inventory linking, LCIA arithmetic, linear scaling, contribution reconciliation, and deterministic repeated solves.
Claims on this page apply only to the stated system boundary, methods, source revision, and evidence status.
- Source revision
- Brightway MCP 0.3.0 · source 96136cf163b13c17d163bf70b02041513e7a3fca
- Last verified
- 2026-09-15
Decision question
Can Brightway reproduce a complete LCIA result whose correct answer is known analytically, preserve linear scaling, reconcile process and emission contributions, and repeat deterministically?
System boundary
A synthetic two-process product system with one upstream material process, one reference-product process, one CO₂ elementary flow, and a single characterization factor of 1.0 kg CO₂e per kg emitted.
Functional or reference basis
One kilogram of Golden Product, plus a two-kilogram linear-scaling check.
Methodology
The analytical reference is 0.5 kg direct CO₂e plus 3 kg of upstream material multiplied by 2.0 kg CO₂ per kg material, giving 6.5 kg CO₂e per kilogram of product. Brightway builds and solves the same chain, verifies 2× scaling, repeats the calculation five times, and checks that process and emission contributions reconcile to the solved total. A separately implemented openLCA golden fixture is retained as an independent cross-engine corroboration.
Assumptions
The intentionally small foreground and characterization factor isolate calculation and linking semantics from background-database and full-method differences. There is no allocation, recycling loop, external background database, or uncertain input in the core reference case.
Results
Brightway returns 6.5 kg CO₂e for 1 kg Golden Product and 13.0 kg CO₂e for 2 kg to the declared floating-point tolerance. Five repeated deterministic calculations have zero spread, and the process/emission contribution totals reconcile with the solved score. The independently implemented openLCA fixture reproduces the same 6.5 and 13.0 kg CO₂e analytical values.
Interpretation
The closed-form benchmark isolates core inventory linking, scaling, characterization, and contribution accounting from changing background data. The independent openLCA check strengthens confidence in the reference arithmetic without turning this into a claim of general engine equivalence.
Limitations
The benchmark is deliberately synthetic. It does not validate real background databases, allocation, data quality, uncertainty, or broader methodological equivalence between LCA engines.
Reproduction
Run the Brightway golden-system tests in the reviewed Brightway MCP repository and compare the 1 kg and 2 kg scores, repeated-run spread, and contribution totals with the analytical 6.5/13.0 kg CO₂e reference. The openLCA golden fixture may be run separately as a cross-engine check.
Review the underlying methods.
Use the documentation to inspect the tools and workflows behind this study.