Engineering Analysis of Half-Hourly Hydropower Plant Records: Operating-Regime Reconstruction, Energy Estimation, and State-Aware Data-Quality Assessment

Half-hourly hydropower plant records support the engineering analysis of operating states, energy, and measurement consistency; however, legacy spreadsheet structures complicate their integration within a traceable workflow. We analyzed authorized records from an anonymized plant using source-preserving data processing. The workbook comprised 31 sheets and included 1519 rows. Of these, 1488 complete intervals constituted the analytical dataset, with no imputation applied. A common unit-state model was used to reconstruct the operating regimes, allocate the interval energy, and condition the data quality diagnostics. The estimated monthly energy was 3689.75 MWh (119.02 MWh/day). The both-online configuration represented 291.0 allocated hours and accounted for 54.04% of the estimated energy; single-unit configurations represented 435.5 allocated hours; and the both-offline configuration represented 17.5 allocated hours. Daily energy was strongly associated with the mean turbine flow (Pearson’s r = 0.967; Spearman’s rho = 0.950; n = 31). Unit power-factor formula errors coincided with inactive unit states, all 35 negative line-power records occurred while both units were offline, and one interval had a 0.584 MW aggregation discrepancy. By using the reconstructed unit state as a shared analytical layer, the framework links energy allocation and measurement diagnostics while preserving record-level provenance. Energy and time allocations follow an explicit representative-interval convention.