IDEAL Household Energy Dataset: 255 UK Homes, Electricity, Gas & Sensors [Nature Scientific Data]

Overview

The IDEAL Household Energy Dataset is a high-quality, multi-modal energy monitoring corpus collected from 255 UK residential homes as part of the IDEAL (Identification of Demand flexibility in rEsidentiAL buildings) project. The study ran for 23 months, ending in June 2018, and represents one of the largest and richest open datasets linking household-level energy consumption to occupant behavior, demographic characteristics, and building properties.

For all 255 homes, the dataset includes high-resolution electricity and gas consumption data alongside individual room temperature and humidity readings from IoT sensors, and boiler temperature readings. For a sub-cohort of 39 homes, more detailed data is available including individual electrical appliance monitoring (plug-level current clamps) and individual radiator heat output data.

Sensor data is augmented by anonymised survey data and metadata covering occupant demographics, self-reported energy awareness and attitudes, building characteristics (construction type, insulation, glazing), room types, and appliance inventories. The dataset was described in a paper published in Nature Scientific Data (2021) and is freely accessible from the University of Edinburgh DataShare repository.

Column Schema

Key Statistics

• Homes: 255 UK residential homes (full dataset); 39 homes with appliance-level detail
• Duration: 23 months (ending June 2018)
• Data Streams: Electricity, gas, room temperature, room humidity, boiler temperature, appliance-level power (sub-cohort)
• Augmentation: Survey data — occupant demographics, energy attitudes, building and appliance metadata
• File Format: CSV (structured directories per home, room, and sensor)
• Published: April 2021, University of Edinburgh DataShare (DOI: 10.7488/ds/2836)

Use Cases

• Residential energy consumption modeling and occupancy inference
• Demand flexibility identification and demand response research
• Heating and cooling load forecasting using temperature and gas data
• Behavioral energy analysis linking occupant demographics to consumption patterns
• Smart home IoT sensor fusion across electricity, gas, temperature, and humidity streams

Source and Attribution

Created by the IDEAL project team at the University of Edinburgh, led by Dr. David Murray-Rust. Published in Nature Scientific Data (2021) and deposited in the University of Edinburgh DataShare repository. DOI: 10.7488/ds/2836. Licensed for research use.