Sierra Leone Safe Water Project

Integrity

The verification report (2024-05-29) states that additionality was confirmed by the VVB (SustainCERT) and the baseline was reassessed in 2024, which supports integrity. However, reversal events are described as not addressed in the monitoring record and no buffer pool percentage is provided in available documents; leakage treatment is inconsistent (0% deduction in the 2024 monitoring report versus a 5% figure in an earlier/unknown document). These inconsistencies and the lack of explicit reversal safeguards reduce the integrity score.

Transparency

SustainCERT is named and a clear monitoring period for 2024 is provided, and multiple evidence documents are listed (validation, monitoring, verification, PDD, stakeholder consultation). Yet key numeric outputs are missing from the extracted record (total emissions claimed/verified and grid emission factor are not found) and the minimum extraction confidence is low, which undermines transparency and traceability.

Claim Safety

There is elevated risk of over‑crediting: the latest monitoring report uses a national default FN/RB factor (0.95) and an annual survey for usage monitoring, but the merged facts show multiple contradictory FN/RB values and usage rates across earlier documents (PDD and earlier monitoring reports). CORSIA/CCP status is not stated, leaving double‑claim risk unresolved. These contradictions and gaps lower claim safety.

Documentation

A substantial set of documents is listed and 23 documents were used, including a recent verification report (2024-05-29). However, extraction confidence is low and some sources are listed as 'unknown' or missing, and there are unresolved contradictions flagged in the record. The low extraction confidence and unknown/contradictory inputs reduce the documentation score.

Overall

I privileged the most recent, higher‑priority documents: the 2024 monitoring report (dated 2024-05-18) and the 2024 verification report (2024-05-29) over older PDD entries and unspecified/unknown documents because they are more recent and are formal verification outputs. Specific contradictions addressed: (1) usage rate assumed — monitoring report (2021/2024) shows 100% assumed and verified usage ~95% (monitoring report 2024); the older PDD (2019) reported 85% assumed so I privileged the more recent monitoring values. (2) leakage deduction — I privileged the 2024 monitoring report value of 0% with a 'deemed negligible' justification over an earlier 5% figure from an unknown 2023 source because the monitoring report is the latest verification document; nevertheless the contradiction reduces confidence. (3) leakage justification — recent monitoring report states 'deemed negligible' while a 2021 monitoring report had quantified leakage; I privileged the 2024 monitoring report but penalized scores for the inconsistency. (4) FN/RB — I used the 2024 monitoring reported FN/RB value (0.95, national default) over an anomalous 63 value from an unknown 2023 source; the huge discrepancy flag reduces claim safety. (5) usage_rate_verified — recent monitoring (2024) reports 95% which I privilege over earlier lower verified values (2019/2020). (6) FN/RB method — monitoring report uses national default (2024) over older/local methods. (7) usage monitoring method — annual survey in recent monitoring is privileged over earlier self‑report. (8) reversal_events — recent monitoring lists reversal events as 'not addressed' while a later unknown source claimed none reported; I treated the verification report's lack of reversal treatment as a concern. (9) crediting period — I relied on the verification report (2019-05-19 — 2024-05-18). Overall score is lowered because multiple contradictions and low extraction confidence indicate data reliability issues despite positive elements (VVB confirmation, recent reassessment).