Smart City Digital Twin–Enabled Energy Management: Toward Real-Time Urban Building Energy Benchmarking

TLDR

Cities must assess building energy performance to meet reduction goals, yet current top‑down benchmarking lacks actionable insights, while smart‑meter analytics offer the chance to capture previously undetectable temporal fluctuations. This study develops daily, period‑segmented building energy benchmarks from smart‑meter data to quantify differences from conventional annual benchmarks and assess their utility for near‑real‑time energy management. The authors segment data into occupied/unoccupied periods during the school year, occupied/unoccupied periods during summer, and peak summer demand, using these intervals to generate the benchmarks. Temporally segmented benchmarks differ from overall benchmarks, exposing periods of over‑ or under‑performance, providing a more accurate efficiency measure, and establishing a foundation for digital‑twin‑enabled urban energy management and large‑scale smart optimization.

Abstract

To meet energy-reduction goals, cities are challenged with assessing building energy performance and prioritizing efficiency upgrades across existing buildings. Although current top-down building energy benchmarking approaches are useful for identifying overall efficient and poor performers across a portfolio of buildings at a city scale, they are limited in their ability to provide actionable insights regarding efficiency opportunities. Concurrently, advances in smart metering data analytics combined with new data streams available via smart metering infrastructure present the opportunity to incorporate previously undetectable temporal fluctuations into top-down building benchmarking analyses. This paper leveraged smart meter electricity data to develop daily building energy benchmarks segmented by strategic periods to quantify their variation from conventional, annual energy benchmarking strategies and investigate how such metrics can lead to near real-time energy management. The periods considered include occupied periods during the school year, unoccupied periods during the school year, occupied periods during the summer, unoccupied periods during the summer, and peak summer demand periods. Results showed that temporally segmented building energy benchmarks are distinct from a building’s overall benchmark. This demonstrates that a building’s overall benchmark masks periods in which a building is over- or underperforming during the day, week, or month; thus, temporally segmented energy benchmarks can provide a more specific and accurate measure for building efficiency. We discussed how these findings establish the foundation for digital twin–enabled urban energy management platforms by enabling identification of building retrofit strategies and near-real-time efficiency in the context of the performance of an entire building portfolio. Temporally segmented energy benchmarking measures generated from smart meter data streams are a critical step for integrating smart meter analytics with building energy benchmarking techniques, and for conducting smarter energy management across a large geographic scale of buildings.