Quebec’s winter months are synonymous with biting cold, extended darkness, and a relentless demand for energy. Despite Hydro-Québec’s provision of abundant and affordable electricity, the province faces persistent challenges in energy waste and inefficiency. A groundbreaking study conducted by researchers at Concordia University now offers a cutting-edge analysis of household energy consumption across Quebec’s major urban centers by diving deeper into neighborhood-level usage patterns and socio-demographic variables. This study unveils a nuanced understanding of how human factors—not just infrastructure—drive energy consumption, offering promising avenues for smarter, fairer energy policies in cold-climate cities.
Harnessing the power of detailed smart meter data collected hourly from 2019 to 2023, the researchers mapped out residential electricity usage across Montreal, Trois-Rivières, and Quebec City. They segmented the data according to Canada’s Forward Sortation Areas, geographical subdivisions usually employed for optimizing mail delivery. This granular approach allowed the team to correlate energy consumption not just with building characteristics, but with demographic features drawn from the 2021 Canadian census, such as income levels, employment rates, household sizes, car ownership, and age distributions. By leveraging these interdisciplinary data sets, the team aimed to pinpoint which societal factors most significantly influence energy demand patterns.
Advanced machine learning algorithms played a pivotal role in this research. The team employed sophisticated models to detect the activation points of heating systems in response to changing outdoor temperatures. This change-point analysis illuminated long-term consumption behaviors, particularly the timing and intensity of heating periods. Meanwhile, another machine learning framework clustered daily electricity use into distinct profiles, enabling the researchers to interpret short-term variations in energy use relative to socio-demographic characteristics. Importantly, these AI-driven methods did not merely generate predictions; they revealed the underlying drivers and relative importance of various social variables.
One of the striking revelations was the heterogeneous relationship between income and energy consumption. Higher-income neighborhoods tended to exhibit elevated baseline electricity use coupled with more pronounced spikes as outdoor temperatures dropped. This suggests that wealthier households maintain higher indoor comfort levels and possibly consume energy for a broader array of devices year-round. Conversely, lower-income communities often initiated heating earlier in the cold season, a trend that may reflect older, poorly insulated housing stock requiring prolonged heating despite the relatively affordable power supply.
Age demographics emerged as another key determinant. Neighborhoods populated by older residents showed notably higher electricity use on a per capita basis. This pattern is understandable given that seniors typically spend more time indoors and prioritize thermal comfort, which can intensify heating needs. Conversely, areas characterized by younger residents, high-rise apartment complexes, densely crowded living conditions, or higher proportions of non-Canadian residents tended to have lower energy consumption overall, reflecting different lifestyle patterns, energy efficiency of newer buildings, or economic constraints.
Daily behavioral rhythms also sculpted energy demand curves. Communities with high employment rates and car dependency exhibited sharp morning and evening peaks in electricity use—corresponding with the times residents leave and return from work. In contrast, neighborhoods with higher unemployment or more walkable environments produced flatter, less volatile energy consumption profiles throughout the day. These distinct patterns underscore the profound influence of social structure and urban design on electricity demand, with implications for grid management and peak load forecasting.
Integrating such socio-demographic insights into energy modeling represents a radical shift from conventional approaches that primarily focus on physical building attributes. This holistic perspective acknowledges inhabitants as dynamic agents whose socioeconomic status, lifestyle, and cultural backgrounds directly shape energy consumption. Lead author Masood Shamsaiee, a PhD scholar at Concordia’s Next-Generation Cities Institute, emphasizes the necessity of embedding these human elements into policy frameworks to enhance both efficiency and equity in energy systems.
This pioneering research holds practical value for utility operators like Hydro-Québec seeking to optimize distribution networks and design tailored energy efficiency programs. By identifying which communities contribute most to peak loads and inefficiency, utilities can prioritize investments in retrofits, subsidies, or educational outreach. Furthermore, this data-driven segmentation supports the development of equitable policies ensuring that lower-income and vulnerable populations are not disproportionately burdened by energy costs or systemic inefficiencies, aligning with broader social justice objectives.
The study, published in the journal Energy and Buildings, also opens avenues for future interdisciplinary research at the intersection of urban studies, social science, and electrical engineering. The coupling of machine learning with census data heralds a new era in understanding complex urban energy dynamics, where the granularity of behavioral and demographic factors is no longer lost in aggregate statistics. This framework could be adapted for cities worldwide facing similar challenges of balancing cold-climate energy needs with sustainability and social equity.
As Canadian cities, particularly those in harsh winter zones, grapple with decarbonization imperatives and evolving energy demands, insights gleaned from this research are timely and vital. They reiterate that addressing climate change and energy efficiency requires more than technological upgrades; it demands nuanced appreciation of human diversity and consumption contexts. Data-informed, community-centered strategies, as exemplified by this study, could well transform the future landscape of energy policy and urban resilience.
Ultimately, this investigation illuminates the complex interplay between society and technology, illustrating that energy demand arises not solely from infrastructure but from the people within it—their habits, socioeconomics, and lifestyles. For cities like Montreal, Quebec City, and Trois-Rivières, such knowledge equips city planners, utilities, and policymakers with the tools to craft smarter, fairer, and more effective energy solutions that honor the lived realities of their residents while advancing sustainability goals.
Subject of Research: Urban building energy consumption influenced by socio-demographic factors in Quebec’s major cities.
Article Title: Socio-Demographic insights on urban building energy consumption
News Publication Date: 10-Dec-2025
Web References:
- https://www.sciencedirect.com/science/article/pii/S0378778825015233
- DOI: 10.1016/j.enbuild.2025.116793
Image Credits: Concordia University
Keywords: Urban populations, Demography, Energy consumption, Machine learning, Smart meters, Heating demand, Socio-demographic factors, Energy efficiency, Electricity use patterns, Hydro-Québec, Climate adaptation, Urban studies
