Home › Energy › EnPI Workbench

📈 EnPI & Baseline Workbench

The ISO 50001 §6.3/§9.1 engine room: build a regression baseline (kWh = a·driver + b) from your baseline-period months, check R² ≥ 0.75, then track the reporting period with predicted-vs-actual and a CUSUM of savings — the IPMVP option C logic, free.

Monthly data

Period: B = baseline (builds the model) · R = reporting (savings measured against the model). Driver = production units, tonnes, DJU/HDD, hours…

MonthPeriodDriver (X)Energy kWh (Y)PredictedDiff

Actual vs baseline model

CUSUM of savings (kWh)

Method: ordinary least squares on baseline months; savings(month) = predicted − actual; CUSUM = running sum of savings over the reporting period (a rising CUSUM = real, persistent savings; a flat/falling one = drift). Validity checks: R² ≥ 0.75 and ≥ 10–12 baseline points recommended (ISO 50006, IPMVP).