Start 7-day trial

ISO 50006:2014

ISO 50006

How to choose energy performance indicators, establish baselines, and normalize for weather and production — the measurement detail ISO 50001 requires but does not spell out.

What the standard covers

ISO 50006:2014 (Measuring energy performance using energy baselines and energy performance indicators — General principles and guidance) is the technical specification behind ISO 50001’s measurement requirements.

ISO 50001 says you must track energy performance. ISO 50006 explains how: which indicator to choose, how to set a baseline, and how to adjust for variables that change between periods.


Three types of EnPI

An Energy Performance Indicator (EnPI) is a quantitative measure of energy performance. ISO 50006 defines three types.

Value-based

A direct consumption total for a period.

Example: 412,000 kWh of electricity in Q1.

Works when conditions are stable. Fails when weather or production shifts between comparison periods — a mild winter and a harsh winter are not comparable on raw kWh alone.

Ratio-based

Energy divided by an activity variable.

Example: 85 kWh per tonne of product; 120 kWh per occupied bed-day.

Suited to production-driven facilities with one dominant driver. Assumes a relationship through zero — which is wrong for most buildings that carry a weather-independent base load.

Model-based

Energy modelled as a function of one or more variables using regression.

Example: Monthly electricity = 28,000 kWh + 420 × HDD + 310 × CDD

Best for weather-sensitive buildings and facilities with multiple simultaneous drivers. The model separates base load from weather-driven load, making fair comparison possible across different climatic conditions.

This is the approach ISO 50001 auditors most often expect for building-level verification — and it requires reliable degree day data matched to the same periods as your utility bills.


The energy baseline (EnB)

The baseline is the reference against which EnPI values are compared. A weak baseline makes it impossible to tell whether performance improved or the weather simply changed.

Requirements for a sound baseline:

  • At least 12 months for weather-sensitive facilities
  • Represents typical operating conditions — not an unusual year
  • Established before improvements wherever possible
  • Documents all relevant variables and static factors in effect during the reference period

Adjustments:

  • Routine — systematic normalization for changing relevant variables (weather, production). Defined in advance.
  • Non-routine — one-off corrections when something structural changes (extension, outsourced production, equipment replacement). Must be pre-defined and documented.

Relevant variables vs. static factors

TypeDefinitionExamplesTreatment
Relevant variableChanges routinely and significantly affects energyOutdoor temperature, production volume, occupancyInclude in the EnPI model
Static factorAffects energy but does not routinely changeBuilding size, equipment type, number of shiftsDocument; non-routine adjustment if it changes

Degree days are the canonical relevant variable for buildings. They quantify the temperature difference that drives heating and cooling demand. A model-based EnPI that includes HDD and CDD separates weather-driven variation from genuine performance change.


What auditors ask

An ISO 50001 auditor will typically want to know:

  1. Which EnPI you use and how it is calculated
  2. How the baseline is defined and what period it covers
  3. How relevant variable changes are accounted for
  4. Whether non-routine adjustments are documented

Answers backed by a citable weather data source and reproducible methodology hold up. “We adjusted for weather informally” does not.

Ed runs model-based EnPI regressions against billing-period degree days, projects baselines forward to reporting-period conditions, and displays model quality metrics — based on the principles of ISO 50006, not as a formal EnPI certification package.


Further reading

M&V workflow

Prove savings with documented degree days. Regression, normalization, and model quality in Ed.

Upload billing data, match degree days to billing periods, and generate weather-normalized savings evidence auditors can review.