Demand
Summary Description
Electricity demand is a vital component in the planning, operation, and resilience of power systems. Climate change is expected to alter electricity demand patterns by affecting not only temperature characteristics, such as mean temperature, frequency of temperature extremes, and seasonal variability, but also consumption behaviours across various sectors.
Numerous studies project increased cooling demand during hotter summers and reduced heating demand during milder winters as this pattern has been consistently observed across multiple global regions (Amonkar et al. (2023), Auffhammer and Aroonruengsawat (2011), Pilli-Sihvola et al. (2010), Trotter et al. (2016), Wood et al. (2015)). However, the magnitude and direction of these changes are strongly dependent on region, latitude, and socioeconomic conditions. For example, countries in warmer climates are expected to experience higher overall demand growth due to increased cooling loads (Bonkaney et al. (2023), Emodi et al. (2018), Zachariadis and Hadjinicolaou (2014)). Climate change also influences electricity consumption by sector and alters temporal demand profiles, including daily and seasonal variations. The residential, commercial, and industrial sectors respond differently to climate drivers (Dirks et al. (2015), Shaik (2024), Taseska et al. (2012)). Furthermore, the timing of electricity demand is shifting, with studies projecting changes in peak load hours and seasonal peaks due to rising temperatures and more frequent extreme heat events (Amonkar et al. (2023), Fonseca et al. (2019), Romitti and Sue Wing (2022)).
To assess the impacts of climate change on electricity demand, various approaches have been employed. These include the use of climate projections from General Circulation Models (or Global Climate Models, GCMs) and Regional Climate Models (RCMs), often combined with downscaling techniques to capture local climatic variability (Auffhammer et al. (2017), Auffhammer and Aroonruengsawat (2011), Fonseca et al. (2019), Lipson et al. (2019), Trotter et al. (2016)). Empirical models, such as regression-based analyses, have also been used to estimate historical relationships between temperature and electricity use (Bonkaney et al. (2023), Emodi et al. (2018), Pilli-Sihvola et al. (2010)). Additionally, simulation tools and integrated energy-climate models are applied to explore future scenarios under varying socio-economic and climatic conditions (Dirks et al. (2015), Lipson et al. (2019), Taseska et al. (2012)).
Role in the Electricity System
Methods and Models
Climate data such as GCMs and RCMs prepared for the appropriate spatial and temporal resolution are integrated into models that simulate or estimate electricity demand responses to climatic variables. The literature broadly categorizes these models into two groups: empirical models, which are typically data-driven, and simulation-based models, which are process-oriented or scenario-based. These approaches differ in their methodological structure and are selected based on the availability of data, the spatial and temporal scale of analysis, and the objective of the study or project.
Empirical models typically use regression-based methods to quantify the historical relationship between electricity demand and weather variables, mostly temperature. Auffhammer and Aroonruengsawat (2011) used a simple log-linear regression models to estimate residential demand sensitivity and climate change impacts estimation in California. Fonseca et al. (2019) developed empirical load models based on hourly utility data and downscaled weather variables to simulate load variability under future climate scenarios. Shaik (2024) extended this approach across multiple sectors and climate variables to evaluate sectoral demand elasticity and response behavior in the United States. While empirical models are statistically robust and adequate for retrospective analysis, they are inherently constrained by the historical range of observed conditions and may not fully capture adaptive or nonlinear responses to future extremes. Ontario practitioners use this method for medium term (11 days to 2 years) with 31 years of historical data incorporating many climate inputs such as dry bulb temperature, cloud cover, wind speed, dewpoint and irradiance.
Simulation-based models, on the other hand, allow to explore prospective scenarios by embedding climate variables within energy system or physics models. These include tools like EnergyPlus for building-level simulations (as used by Berardi and Jafarpur (2020)) and energy system optimization models such as MARKAL (as applied in Taseska et al. (2012)). Simulation models offer the advantage of flexibility in exploring technological, behavioral and policy adaptations under different emissions and socioeconomic pathways. Lipson et al. (2019) incorporated building-urban-atmosphere interaction models to simulate energy demand responses under urban heat island effects, demonstrating how climate-urban feedbacks can influence spatial energy use patterns. Parkinson and Djilali (2015) modeled electricity system planning under hydro-climatic uncertainty using robust optimization techniques, highlighting the importance of integrating demand and supply-side climate risks. Ontario practitioners have developed simulation-based models to address long term horizon of 2 to 20 years; at the moment it does not incorporate climate inputs, but regression of heating and cooling degree-days will soon be added in models (historical only).
In practice, demand forecasting is often conducted using end-use approaches (IESO (2025)), which can be implemented within either an empirical or a simulation-based framework. Empirical end-use models estimate electricity demand by disaggregated categories such as end-use type, sector, or region based on historical consumption patterns and weather variables. In contrast, simulation-based end-use models, which are more common, simulate the behaviour of appliances, buildings, or sectors using physical parameters.
The figure below briefly resumes decision-making challenges faced by professionals in the demand sector of the electricity system in Canada. These challenges were shared during the workshops held in 2025. These examples linked together the methods and models as well as characteristics presented in the following section such as key climate inputs.
The workshop examples highlighted how demand forecasting is shaped by the North American Electric Reliability Corporation (NERC) requirement to maintain a five-year planning reserve margin. As a result, two forecasting horizons are considered: medium-term (11 days to 2 years) and long-term (2 to 20 years). The medium-term is statistically based on historical data (31 years) while the long-term planning is physically based, as described above. The load forecasting model incorporates historical hourly normals for dry bulb temperature, cloud cover, wind speed, dew point and global horizontal irradiance across ten geographic zones in the province. There is a plan to incorporate relative humidity and temperature (heating degree days and cooling degree days) in the forecast. The main sources for these datasets are Environment Canada and external consultants (wind speed and global horizontal irradiance). Climate data is validated by comparing the model’s outputs with historical load data, ensuring that the forecasts align with observed patterns.
Detailed Discussion
- Seasonal variations:
Rising temperatures due to climate change have led to a significant shift in seasonal electricity demand, particularly by increasing cooling demand during summer and reducing heating needs during winter. This shifting trend has been widely documented across regional and global studies. For example, Auffhammer et al. (2017) showed that electricity demand in the United States exhibits a strong nonlinear response to high temperatures, especially above 25-30 degC, driven mainly by demand for air conditioning. Similarly, De Cian and Sue Wing (2019) also confirmed this global trend that there are steep increases in demand where air conditioning adoption is rising due to climate change. Regional case studies, such as Garrido-Perez et al. (2021) and Bonkaney et al. (2023), have demonstrated that peak demand coincides more frequently with extreme heat days, stressing both power generation and distribution infrastructure. The anticipated rise in the frequency, intensity, and duration of heatwaves further amplifies these concerns, requiring investments in peak capacity, demand-side management, and adaptation strategies. Taseska et al. (2012) projected increased summer cooling demand across European regions due to higher temperatures, while Zachariadis and Hadjinicolaou (2014) confirmed a similar rise in cooling demand in Cyprus. Dirks et al. (2015) also found that seasonal peaks are shifting in the United States, with summer peaks becoming more dominant due to climate-induced changes in building energy use.
However, while winters are projected to be warmer, the demand responses are more variable. Romitti and Sue Wing (2022) introduced a classification of demand response profiles: a “V”-shaped response which is common in mid-latitude temperate cities such as those in North America and Europe, where demand increases at both high and low temperatures; an increasing response, where demand rises steadily with temperature, which can be typically shown in tropical cities; and an unresponsive profile, where minimal or no correlation exists between temperature and electricity demand. Pilli-Sihvola et al. (2010) used multivariate regression in Europe to estimate electricity consumption changes and found that Northern and Central Europe will experience reduced heating demand in winter, while Southern Europe will face increased cooling demand and higher costs in summer. This results in increased annual electricity demand in Spain, while it will decrease in Finland, Germany, and France. Although warmer winters are generally expected to reduce heating loads, the widespread adoption of heat pumps and other climate mitigation technologies may lead to increased winter electricity demand due to fuel switching from fossil-based systems to electric heating. Wood et al. (2015) analyzed how climate change could flatten winter peaks while enhancing summer peaks in the United Kingdom, suggesting that infrastructure planning should shift from winter-dominant to summer-dominant strategies.
- Regional variations:
The impacts of climate change on electricity demand show considerable regional heterogeneity, influenced by geographic location, climatic baseline, economic development, and infrastructure characteristics. A clear latitudinal pattern exists: countries at lower latitudes are projected to have higher increases in electricity demand for cooling, while higher latitude regions may observe mixed outcomes due to the opposing effects of warmer winters and hotter summers. Similar to the Romitti and Sue Wing (2022)’s classification, Hu et al. (2024) analyzed Temperature Response Functions (TRFs) across Europe by fitting the relationship between electricity demand and temperature to three types of curves: a linear decreasing curve for Northern European countries, a linear curve with a horizontal segment for cold and intermediate climate countries, and a V-shaped curve with a comfort zone for intermediate or warm countries, where both heating and cooling demands are affected.
Regional variations in electricity demand due to climate change across Japan show a clear dependency on latitude (Hiruta et al. (2022)) Northern regions, such as Hokkaido and Tohoku, will experience decreased annual electricity demand owing to reduced heating needs, whereas southern regions from Tokyo to Okinawa will see increased consumption due to extended cooling requirements. Transition zones near the boundary of the northern and southern regions present a balance between reduced heading and increased cooling demands, while these transition zones are shifting northward as climate warming continues. McFarland et al. (2015) also confirmed a decrease in HDD in the northern US, while the increase of CDD is significant in the southern US.
In tropical regions, where average temperatures are already high, even modest increases in temperature can trigger substantial rises in electricity use. Bonkaney et al. (2023) found that rising temperatures will affect electricity demand across all warming levels due to increasing air conditioning needs, with low adaptive capacity exacerbating system stress in Niger. De Cian and Sue Wing (2019) projected larger increases in energy consumption, primarily driven by higher frequency of extreme temperatures and the resulting increased cooling demand. However, in temperate regions, impacts on demand are mixed, varying based on local climate and socioeconomic conditions, as demand can either increase or decrease depending on local climate and geographic incidence of climate change (De Cian and Sue Wing (2019)).
- Sectoral variations:
Climate change impacts electricity demand across sectors in different ways due to sector-specific end-use patterns, varying weather sensitivities, and adaptive capacities. Therefore, understanding differences in electricity demand patterns across sectors is essential for ensuring grid resilience under changing climate conditions.
In the residential sector, electricity demand is particularly sensitive to temperature fluctuations, primarily due to the use of heating and cooling appliances to meet residents’ comfort expectations, which increase faster than the climate is changing (Wood et al. (2015)). De Cian and Sue Wing (2019) found that temperate regions, warming reduces heating needs, resulting in lower electricity consumption, especially in households using electricity-based heating systems, such as heat pumps. However, in tropical and warmer temperate zones, the demand for space cooling increases substantially with rising temperatures and humidity levels, leading to a net increase in electricity use in residential buildings. Berardi and Jafarpur (2020) emphasized that while heating energy use intensity may decline by 18-33%, cooling energy user intensity could rise by as much as 126% by 2070 in Canadian buildings.
The commercial sector is also highly climate-sensitive, primarily due to its dependence on heating, ventilating, and air conditioning (HVAC) systems. Taseska et al. (2012) projected that under a warmer climate scenario, commercial electricity demand for cooling would increase, especially during peak summer periods, while the reduction in heating demand would be relatively minor. Véliz et al. (2017) showed that commercial electricity demands would increase, driven not only by higher consumption but also by rising electricity prices. This suggests that the commercial sector faces both quantity- and price-based vulnerabilities under climate change.
The industrial sector shows more heterogeneous responses. Its demand is generally less directly related to air temperature, but certain processes, particularly those requiring climate-controlled environments, are affected by heat. Shaik (2024) showed that electricity demand in the industrial sector remains relatively price-inelastic and shows modest temperature sensitivity compared to other sectors. However, long-term changes in the demand may emerge through indirect pathways, such as workforce comfort requirements, advances in technology, or regulatory standards.
In the transportation sector, electricity demand is currently limited as the sector remains largely dependent on petroleum. However, with increasing electrification of transport systems, future electricity demand is expected to rise significantly (De Cian and Sue Wing (2019)). Shaik (2024) also noted that while the sector’s electricity use is still low, it is highly responsive to economic and price signals, suggesting future variability under combined climate and market pressures.