LCA of Sachet Water Production in West Africa — Full Documentation

Project: LCA of Sachet Water (Lagos, Nigeria). Files and data are in project folder.

1. Project overview and scope

Goal: Compare cradle-to-grave environmental impacts of sachet water and PET bottled water for Lagos, Nigeria. Functional unit: 1 L potable drinking water consumed at point of use. System boundary: raw water abstraction → (treatment) → packaging manufacture & filling → distribution → consumption → end-of-life of packaging (recycling, landfill, open burning/mismanaged).

2. Process followed

1. Defined scope, functional unit, and system boundaries (see protocol.md).
2. Compiled open-access data sources and literature (see bibliography.md).
3. Created an inventory template (inventory_template.csv) and populated Lagos-default numeric parameters (packaging mass, filling energy, transport distance, end-of-life shares).
4. Prepared OpenLCA import templates and mapping guidance (files: openlca_import_template.csv, openlca_premap.csv, openlca_mapping.md).
5. Developed reproducible Brightway2 script templates (in scripts/) to automate mapping and LCIA when run locally.
6. Computed deterministic LCIA results (GWP, water depletion, cumulative energy) using literature-default emission factors and the inventory defaults; saved in results/.
+
7. Performed sensitivity runs and a Monte Carlo (approximate) uncertainty propagation; summary outputs saved in results/mc_summary.csv and sensitivity files.

3. Data used and sources

Primary inventory data (defaults) are in inventory_template.csv. Key numeric defaults used for deterministic calculations:

• Sachet packaging mass: 6 g / L (two 500 mL sachets at ~3 g each)
• PET packaging mass: 24 g / L (two 500 mL bottles at ~12 g each)
• Filling water: 1.0 L per FU (explicit)
• Filling energy: 0.02 kWh / L
• Transport distance: 40 km (average roundtrip equivalent)
• End-of-life fate (sachet baseline): recycling 5%, landfill 20%, open burning/mismanaged 75%
• End-of-life fate (PET baseline): recycling 15%, landfill 35%, open burning/mismanaged 50%

Open-access source list is in bibliography.md (ELCD, USLCI, World Bank "What a Waste", LAWMA, UNEP reports, FAO AQUASTAT).

4. Emission factors & characterization

Default emission and intensity factors used (literature/open-dataset proxies):

5. Calculation approach (deterministic)

For each product (Sachet and PET), total GWP was calculated as the sum of:

• Packaging production: packaging_mass_kg * polymer_GWP
• Filling energy: filling_energy_kWh_per_L * electricity_GWP
• Transport: packaging_mass_t * distance_km * EF_transport (tkm calculation)
• End-of-life: emissions from open burning (mass burned * open_burn_EF) minus recycling credit (recycled_mass * virgin_production_GWP) for closed-loop recycling approximation

Example: packaging production (sachet)

packaging_mass = 6 g = 0.006 kg. LDPE GWP = 2.7 kg CO2-eq/kg. packaging GWP = 0.006 * 2.7 = 0.0162 kg CO2-eq per 1 L.

6. Deterministic results

See detailed numeric breakdown: results/gwp_results.csv, and hotspot table: results/hotspot_table.csv.

Key deterministic outcomes (per 1 L)

7. Monte Carlo uncertainty & approach

I used the uncertainty bounds recorded in inventory_template.csv as independent multiplicative factors (uniform between uncertainty_low and uncertainty_high), propagated them to obtain means and standard deviations. The MC summary (approximate) is in results/mc_summary.csv.

MC summary (means ± std)

8. Sensitivity analysis highlights

• Packaging mass reduction (e.g., 20% lighter sachet) produces ~20% reduction in packaging contributions — strong lever for water and energy.
• Increasing PET recycling to 50% could halve PET packaging water and energy footprints, reducing overall PET water depletion by roughly 50% and GWP by ~15–40% depending on credits.
• Filling energy scales linearly; doubling filling energy significantly affects sachet relative impact due to its smaller packaging burden.
• Transport impacts are small under baseline 40 km; large distances or inefficient fleets can increase energy share materially.

9. Files generated

• README.md — project overview and next steps.
• protocol.md — LCA protocol and assumptions.
• inventory_template.csv — filled inventory defaults and uncertainty bounds.
• openlca_import_template.csv, openlca_premap.csv, openlca_mapping.md — OpenLCA import helpers.
• scripts/ — Brightway2 script templates to run mapping and LCIA locally.
• results/ — CSVs and figures created (GWP, hotspot, MC summary, driver_shares.svg, sensitivity_figures.svg).

10. Assumptions and limitations

Key limitations: use of open-default proxies rather than fully localized LCI processes; simplified modelling of end-of-life (open burning approximated; landfill methane neglected); MC approximations use independent multiplicative uncertainty and normal approximation — for publication-grade results use full sample-based Monte Carlo with correlated parameters where appropriate and use detailed LCI datasets (Ecoinvent) if available.

11. Next steps & deliverables

1. Refine foreground inventory with primary data (sachet mass sampling, plant energy intensity, transport/fleet specifics, waste fate shares from LAWMA/local surveys).
2. Run full random-sampling Monte Carlo (10k draws) and compute distributions/histograms and statistical comparisons.
3. Expand impact categories (ReCiPe): human toxicity, ecotoxicity, particulate matter, freshwater eutrophication.
4. Draft manuscript and policy brief (I can prepare these — see policy memo).

12. How to export to Word

Open this HTML file (documentation_full.html) in a web browser, then in Microsoft Word: File → Open → choose the HTML file, then File → Save As → Word Document (.docx). Alternatively, open this file in the browser, select Print → Microsoft Print to PDF (or Save as PDF) and import into Word.

Generated by the project assistant — repository path: lca-sachet/.