Normalized Consumption

Weather normalization in energy data analysis involves adjusting measured energy consumption to account for weather variations, providing a clearer view of energy efficiency. This process uses weather data, such as heating and cooling degree days, to make energy consumption comparable across different times and locations. We will explain how to achieve this.

Normalize all heating consumption (gas, oil, etc.) directly linked to the outside temperature, excluding consumption for domestic hot water. Electricity consumption should not be standardized.

Degree-days (DD) represent the temperature difference between the outside and the inside of a building. This concept helps compare heating requirements across different buildings, regardless of the outside climate. Typically, 15/15 degree-days are used, with a comfort temperature set at 15°C.

Use the following formula:

Normalized Consumption = Actual Consumption (kWh) * Normal DD / Recording Year DD

1. Normal DD: In Belgium, use 1869.13 (average of 15/15 degree-days from 1988 to 1997).
2. Recording Year DD: Refer to the table for degree-days corresponding to your consumption period.

To normalise your consumption you must:

• record the degrees day using a ‘manual entry’ type device
• display your normalised consumption compared with your reference consumption via a comparator widget

Record monthly degree-days from a reference year using official data for your location. Update this list monthly with the past month's DD value.

We use the following data source for Brussels - https://www.synergrid.be/fr/centre-de-documentation/statistiques-et-donnees/degres-jours

Display reference and normalized consumption in the comparator widget

Steps:

1. Define Monthly Consumption Time Series of the reference year:
   • Select a data source.
   • Set it as the reference time serie of the comparator
   • Group by month.
   • Define a static time frame (e.g., 01/01/2023 - 01/01/2024).
2. Define Monthly Normalized Consumption of the current year:
   • Select the DD static list as the data source.
   • Group by month.
   • Define the time frame as inherited from the data selector (by year).
   • Apply a MathJS transformation:
     • x: DD time series of the selected period.
     • g: Monthly consumption of the selected period to be normalized.
     • ref: Monthly DD of the reference year (e.g., 01/01/2023 - 01/01/2024) with a static time frame forwarded to the selected period. The “clock icon” let you define a specific time frame.
     • Formula: g * ref / x (see mathJS transformation below)

Normalized consumption - time frame definition to forward the reference year to the selected one

Normalized consumption - mathJS transformation

By following these steps, you can effectively normalize your heating consumption data on IoT Factory.