Kenya Solar Water Farm Project

Integrity

Additionality is supported — the validation report confirms an investment test and the VVB confirmed additionality (validation_report 2021). The baseline is project‑level but was reassessed in 2024 (monitoring_record), and leakage is quantified with a 5% deduction and a quantified justification (monitoring_report 2022‑07‑29). However, permanence is weakened because no buffer pool percentage is stated in the extracted records and there are many corrective action requests and material findings reported in the monitoring record (multiple CAR/CL entries and CAR01, FAR01 listed), which reduce confidence in implementation.

Transparency

Transparency is mixed: the VVB is named (4K Earth Science) and a recent monitoring period is provided (2024‑04‑01 — 2025‑03‑31), but critical emission figures (total ERR claimed and verified, grid emission factor) are not found in the extracted record. The low extraction confidence flag indicates at least one key document was poorly readable, which lowers transparency and the ability to cross‑check numbers.

Claim Safety

There is measurable mitigation of over‑claim risk: the project uses the national default FNRB (0.92) and applies a 5% leakage deduction with justification. But the project is CCP‑approved (status: approved), CORSIA eligibility is not stated, and there are contradictions in monitored usage rates (verified 79 vs earlier 90), which raise dual‑claim and over‑crediting concerns.

Documentation

Documentation coverage appears broad (validation report, monitoring report, PDD, stakeholder consultation; n_docs_used = 30) but the minimum extraction confidence is low and the latest extracted monitoring report is dated 2022‑07‑29. Corrective actions remain recorded (CAR01, FAR01), and some important fields (buffer pool pct, grid EF, ERR totals) are missing from the extracted record. These gaps lower the documentation score.

Overall

I privileged the more recent and higher‑priority verification/monitoring documents where contradictions exist. Specifically: fPIC — the validation report (2021) indicating no FPIC is preferred over the older PDD claim; usage assumptions — the more recent 2024 extraction value (69) is preferred over a 2021 90 figure; leakage justification — the monitoring report (2022) showing a quantified justification is preferred over a PDD that did not address leakage; additionality verification — the validation report (2021) confirming additionality is privileged over an unknown 2024 source; usage rate verified — the more recent reported 79 (2024 extraction) is preferred to the 90 figure in the 2022 monitoring report; benefit sharing — the validation report (2021) description is preferred over a 2017 stakeholder note; usage monitoring method — the monitoring report's annual survey is preferred where 'assumed' appeared elsewhere; additionality test type — the validation report's investment test is preferred to the PDD's combined test; reversal events — recent records reporting no reversals are preferred; crediting period — the monitoring report (2022) crediting period was used rather than an unknown future record. These contradictions and the low extraction confidence reduce the overall score despite some positive evidence. If documentation quality were higher and remaining CARs closed with emission figures reported, scores would improve.