WWF Mamize Firewood-Saving Cook Stove Project III

Audit Analysis

The WWF Mamize cook-stove project delivers roughly 97% of its pro-rata ex-ante expectation for the 2019–2021 monitoring window, with a confirmed investment-test additionality and a quantified 5% leakage deduction. However, the project-specific baseline, absence of a stated buffer pool, a low extraction-confidence rating across 15 source documents, and twenty corrective actions (mostly data-quality and typo fixes) temper confidence in the underlying data. No material integrity failures were found, but documentation quality and data consistency remain the project's weakest links.

Project Description

Red Flags

• ⚠Twenty corrective actions were required in the 2024 verification report, including typos in N and BTotal,i values, cell-reference errors in the ER spreadsheet, and sample-number inconsistencies (304 vs 419 vs 442), indicating weak internal data-quality controls.
• ⚠Minimum extraction confidence across the 15 source documents is rated low, meaning at least one key document was poorly readable and the structured record may under-represent the true state of the project.
• ⚠Usage-rate figures conflict between the 2024 verification report (90%) and the 2024 monitoring report (96.7%), a 6.7-percentage-point gap that is not explained in the extracted record.
• ⚠Buffer pool percentage and reversal-event history are not stated in any available document, leaving permanence risk unquantified for a project whose stoves have a finite service life.

Credit Vintages