Normalization overview

Last updated January 23, 2026

Normalization adjusts your energy use data by applying the weather conditions from a selected comparison year to all bills. This removes weather as a variable, making it easier to compare energy use across years.

Why normalize data?

Make accurate year-over-year comparisons of energy use.
Measure energy savings from operational improvements—not weather changes.
Improve long-term energy analysis, especially in Energy Use Intensity (EUI) charts.

Limitations of normalization

You cannot tune weather regressions by meter (like in Cost Avoidance) so you cannot adjust balance point temperature by meter or delete outliers.
No adjustments are made, or can be made, for any other significant variables over time (occupancy, new loads, etc., although EUI charts do account for floor area).
Non-weather sensitive meters return normalized data that matches calendarized data.

Weather-sensitive meters use the degree days from the user-defined normalization comparison year. Use Weather Data Depot to determine a typical degree day year as the comparison year. Avoid using an extreme weather year.

Normalization shows estimated (hypothetical) data. "How much would I have used if the weather for all years were equal to 2017?"

Normalization diagram

Diagram of how normalization works: actual data entered into EnergyCAP is split into not weather-sensitive meter data, which is prorated by average daily use and cost, and weather-sensitive meter data, which gets weather from a selected normalization year allocated by degree days, producing a Monthly Use Trend chart.