Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Back-filling Missing Data When Predicting Domestic Electricity Consumption From Smart Meter Data

Xianjuan Chen, Shuxiang Cai, Alan F. Smeaton

TL;DR

This work tackles the challenge of predicting annual electricity consumption when smart meter data are incomplete by developing a profile-based back-filling method. It performs K-means clustering on a 36-feature representation derived from 12 months across three ToU slots to identify $k=5$ consumption profiles, then assigns missing months for new users using their closest profile to reconstruct monthly usage. The approach is evaluated with SMAPE-derived accuracy across day, night, and peak periods, showing reliable back-filling for up to six missing months and enabling personalized tariff recommendations. The findings highlight significant potential savings from ToU tariffs for nighttime-heavy profiles and demonstrate the practical impact of tailoring tariff choices to individual usage patterns in Ireland's evolving smart-meter landscape.

Abstract

This study uses data from domestic electricity smart meters to estimate annual electricity bills for a whole year. We develop a method for back-filling data smart meter for up to six missing months for users who have less than one year of smart meter data, ensuring reliable estimates of annual consumption. We identify five distinct electricity consumption user profiles for homes based on day, night, and peak usage patterns, highlighting the economic advantages of Time-of-Use (ToU) tariffs over fixed tariffs for most users, especially those with higher nighttime consumption. Ultimately, the results of this study empowers consumers to manage their energy use effectively and to make informed choices regarding electricity tariff plans.

Back-filling Missing Data When Predicting Domestic Electricity Consumption From Smart Meter Data

TL;DR

This work tackles the challenge of predicting annual electricity consumption when smart meter data are incomplete by developing a profile-based back-filling method. It performs K-means clustering on a 36-feature representation derived from 12 months across three ToU slots to identify consumption profiles, then assigns missing months for new users using their closest profile to reconstruct monthly usage. The approach is evaluated with SMAPE-derived accuracy across day, night, and peak periods, showing reliable back-filling for up to six missing months and enabling personalized tariff recommendations. The findings highlight significant potential savings from ToU tariffs for nighttime-heavy profiles and demonstrate the practical impact of tailoring tariff choices to individual usage patterns in Ireland's evolving smart-meter landscape.

Abstract

This study uses data from domestic electricity smart meters to estimate annual electricity bills for a whole year. We develop a method for back-filling data smart meter for up to six missing months for users who have less than one year of smart meter data, ensuring reliable estimates of annual consumption. We identify five distinct electricity consumption user profiles for homes based on day, night, and peak usage patterns, highlighting the economic advantages of Time-of-Use (ToU) tariffs over fixed tariffs for most users, especially those with higher nighttime consumption. Ultimately, the results of this study empowers consumers to manage their energy use effectively and to make informed choices regarding electricity tariff plans.

Paper Structure

This paper contains 17 sections, 2 equations, 7 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Background and Context
  3. Smart Meters
  4. The Irish Energy Market
  5. Electricity Consumption Profiles
  6. Experimental Setup
  7. Data Collection
  8. Data Cleaning
  9. Data Overview
  10. Methodology
  11. Profile Extraction
  12. Back-filling Energy Consumption Data
  13. Evaluation of Clustering Accuracy
  14. Results
  15. Creation of User Profiles
  16. ...and 2 more sections

Figures (7)

  • Figure 1: Installation rate for smart meters in the EU up to 2022 ACER.2023.Report6.
  • Figure 2: Average electricity consumption for each calendar month from 107 consumers.
  • Figure 3: Consumption ratios for day, night and peak timeslots across users
  • Figure 4: Workflow for estimating HDF data for missing months
  • Figure 5: Elbow plot for K-means clustering on users.
  • ...and 2 more figures