Reference case · Case study
Municipal waste LF/WTE uncertainty
A qualified SwolfPy municipal-solid-waste reference workflow covering deterministic LCA, seeded routing uncertainty, repeatability, and lifecycle recovery.
Claims on this page apply only to the stated system boundary, methods, source revision, and evidence status.
- Source revision
- SwolfPy qualification S1 20260915T200335Z · U1 20260915T214654Z
- Last verified
- 2026-09-15
Decision question
Can the SwolfPy workflow reproduce a frozen landfill-plus-waste-to-energy reference result, propagate non-degenerate routing uncertainty reproducibly, and recover cleanly across project lifecycle operations?
System boundary
The lf_wte_basic municipal-solid-waste fixture and a separately frozen synthetic routing-uncertainty case. The benchmark tests SwolfPy/Brightway adapter behavior and stochastic plumbing; it is not presented as an empirical waste-management study for a specific municipality.
Functional or reference basis
Annual system basis 28,421.83 Mg/year for the deterministic reference, with two seed-42 N=200 stochastic runs plus a different-seed check for the synthetic routing fixture.
Methodology
The qualification first locked the deterministic project and LCIA method, then corrected adapter-contract issues around CommonData access, LCIA method shape, and Monte Carlo result retrieval. The authoritative stochastic run used the same production-user execution context as the qualified deterministic project. A separate U1 run introduced frozen synthetic routing variability and checked repeatability, non-degeneracy, state restoration, and cleanup.
Assumptions
Routing distributions in U1 are synthetic qualification inputs rather than measured municipal uncertainty distributions. The deterministic lf_wte_basic fixture is a reference workflow and should not be interpreted as a location-specific waste policy recommendation.
Results
The deterministic reference repeated exactly at -586.5298675913574 kg CO₂-eq/Mg on the stated annual basis. In the qualified U1 uncertainty run, two seed-42 N=200 runs were identical at full stored numeric-array/API precision, seed 43 diverged, each routing variable produced 200 unique samples, LCIA scores were non-degenerate, deterministic state was restored, and cleanup passed. Interrupted-create/retry recovery was also qualified separately in an isolated lifecycle run.
Interpretation
The study supports the tested deterministic and stochastic execution paths and shows that lifecycle context matters for long-running LCA software. The uncertainty result is evidence about the plumbing and repeatability of the workflow, not evidence that the synthetic distributions describe a real waste system.
Limitations
The negative deterministic score depends on the frozen model boundary, credits, LCIA method, and reference data. The study does not validate empirical waste composition, process data, policy feasibility, uncertainty magnitudes, or optimization constraints for a real municipality.
Reproduction
Run the lf_wte_basic project under the qualified SwolfPy runtime, verify the deterministic score and method fixation, then execute the frozen U1 synthetic routing case with N=200 at seed 42 twice and seed 43 once. Compare full stored arrays, restore deterministic state, and verify cleanup.
Review the underlying methods.
Use the documentation to inspect the tools and workflows behind this study.