Develop one or more energy performance indicators and an energy baseline for each EnPI, and compare EnPI values with their respective EnBs on a regular basis [S2].
Define the energy management system scope and record and regularly review the method used to determine and update EnPIs and associated energy baselines [S2][S4].
Define the boundary before choosing the reference period
An energy baseline is a quantitative reference for comparing current EnPI values against past performance, not an arbitrary average of historical consumption. It is built for a defined organizational boundary, energy type, process scope and set of operating conditions [S2][S3].
The boundary document should identify the site, included electricity, gas and steam systems, main meters, submeters, production processes and reporting interval. Electricity, gas and steam have different conversion factors and measurement uncertainties, so combining them into one unexplained total removes information an auditor or energy manager needs to challenge [S3].
Clauses 6.4 and 6.5 address energy performance indicators and energy baselines [S4]. Significant energy uses should be ranked using energy magnitude, variability and improvement potential, and both data ownership and indicator ownership should be assigned [S3].
Use a representative baseline period
The reference period must contain reliable, consistent data for the selected EnPIs and represent normal operation. The baseline period should represent normal operation and sufficient seasonal or production variation, because a period that is too short may miss seasonality and an old period may no longer represent the process [S3].
Major shutdowns, meter failures and unusual order mixes should not be silently deleted; explain their effects and define exclusion criteria in advance [S3]. Define exclusion criteria in advance and explain the effects of major shutdowns, meter failures and unusual order mixes [S3].
Data completeness, the range of operating variables and process stability should be assessed together. Once selected, the baseline data must not change without a version record and rationale, while corrections to meter or operating records should remain traceable [S3].
Select an EnPI matched to significant energy use
An EnPI is a measured value, ratio or model accepted as a meaningful representation of energy performance. It must be evaluated against a logical baseline with the same form and units, such as an intensity indicator in Btu/lb compared with a baseline in Btu/lb [S1][S2].
Simple indicators include kWh per unit produced, energy per tonne and energy per square metre.
Record and regularly review the method used to determine and update EnPIs and establish the conditions under which baseline adjustments will be made [S2].
Normalize relevant variables without overfitting
Relevant variables are quantifiable factors that routinely change and materially affect energy performance, including operating performance. Production volume, product mix, operating hours, degree days, building occupancy and data-centre IT load are examples, while floor area, installed capacity and process design are static factors [S2][S5].
Normalize EnPIs where relevant variables significantly affect energy consumption, and assess data completeness, variable range and process stability [S2][S3]. A period that is too short may miss seasonality, while an old period may no longer represent the process, so variable range and process stability should be assessed together [S3].
Implement a process for ongoing monitoring, measurement and analysis of EnPIs, EnBs and energy performance improvements [S2].
Compare the main EnPI approaches
Simple energy intensity is inexpensive to calculate and easy to communicate, but it can obscure fixed baseload when production changes.
For a manufacturing KPI structure, the EnPI should connect operational output to energy performance without becoming a separate KPI hierarchy. The ISO 22400 manufacturing KPI guide provides a complementary reference for organizing manufacturing indicators around agreed definitions.
Use the baseline where managers can act
The EnPI is suitable when the responsible team can influence the measured result.
Quarterly or monthly EnPI comparisons are useful for management review, but the interval must match the operating cycle and decision speed of the process. The measurement, analysis and improvement process should run continuously rather than wait for the annual management review, with EnPIs, baselines and identified improvements reviewed regularly [S2].
Utility invoices can support organization-level totals, but they usually cannot identify the systems responsible for consumption. Sub-metering of significant energy uses provides the granularity needed to distinguish baseload, shift patterns and changes in daily consumption, with production and weather measurements taken over the same periods [S4][S5].
Control baseline changes through formal triggers
Capacity expansion, process redesign, a new product, a changed meter boundary or a durable operating-regime shift can make a baseline structurally invalid. Routine production, weather and occupancy variation belongs in normalization; structural change requires a formal baseline revision rather than an undocumented adjustment [S3].
Adjustment triggers should be defined before they occur. Each revision should retain the old model, revised model, transition date, rationale and effect on earlier performance reports, with management review testing both the result and whether the method remains suitable [S3].
Changing a baseline solely because a target was missed removes the reference needed to judge performance. The AMO implementation sequence separates establishment, comparison, record maintenance and adjustment conditions, while the LBNL workflow requires regular comparison, documented update methods and ongoing monitoring, measurement and analysis [S1][S2].
Maintain an auditable evidence chain
ISO 50006 provides the companion methodology for defining EnPI boundaries, normalizing baselines where relevant variables significantly affect consumption and defining adjustment triggers [S4]. An auditable record should include the energy review, significant-energy-use register, baseline-period justification, raw energy data, variable data, formulas, units, model validation, comparison records, revision history and assigned ownership [S1][S2][S3].
The evidence chain must preserve scope consistency. Each EnPI needs a corresponding baseline, and the baseline period, boundary and calculation method must remain suitable for the performance claim. If a 50001 Ready project is planned, the selected EnPI, EnB and improvement method must fit that program’s recognized scope and boundaries [S2].
The next control point is a documented EnPI register containing at least one metric and baseline for each significant energy use, with matching units, a named owner and a scheduled review. A baseline should be approved only after the reference-period data, exclusion rules, relevant variables and adjustment triggers have been reviewed against the intended management decision [S1][S2][S3].