Beyond Comfort: Understanding How We React to Buildings

by

Author: Denis Avetisyan

New research reveals occupant behavior during demand response events is driven by factors extending beyond simple thermal preferences.

A study utilizing ecological momentary assessment demonstrates the significant role of non-thermal motivations in human-building interactions and demand flexibility.

Despite growing efforts to enhance grid flexibility through demand response programs, occupant overrides often undermine their effectiveness. This limitation prompted an investigation, detailed in ‘Motivations and Actions of Human-Building Interactions from Environmental Momentary Assessments’, into the underlying drivers of these interactions within the home. Findings from a three-month study utilizing ecological momentary assessment revealed that nearly half of occupant actions related to thermal comfort were motivated by non-thermal factors, challenging models focused solely on temperature preference. How can a more nuanced understanding of occupant motivations inform the design of demand response strategies that better align with real-world behaviors and improve grid resilience?

The Escalating Strain on Modern Energy Systems

Contemporary energy infrastructure, designed for historically stable climates, now faces unprecedented strain from escalating extreme weather events. Rising global temperatures fuel more frequent and intense hurricanes, wildfires, and winter storms, all of which directly threaten power grids and fuel supply chains. Coastal power plants are increasingly vulnerable to storm surges and flooding, while inland facilities grapple with heat-related equipment failures and disruptions to cooling water sources. Transmission lines are susceptible to damage from high winds, ice accumulation, and falling trees, leading to widespread outages. These events not only compromise the immediate delivery of energy but also necessitate costly repairs and upgrades, placing a significant economic burden on utilities and consumers, and highlighting the urgent need for resilient infrastructure planning.

Conventional energy management systems, often relying on predictable load forecasting and centralized control, are increasingly challenged by the escalating frequency of extreme weather and shifting consumption patterns. These systems typically struggle with the inherent unpredictability of renewable energy sources, like solar and wind, and fail to adequately address the localized impacts of events such as heat waves or severe storms. The rigidity of older infrastructure, combined with a lack of real-time data integration and dynamic response capabilities, results in inefficiencies and increased risk of grid instability. Consequently, power outages become more frequent, and the overall resilience of energy networks is compromised, necessitating a shift towards more adaptive and intelligent management strategies capable of handling volatile conditions and rising demand.

The ongoing shift towards electrification of both the heating and transportation sectors, while essential for decarbonization, presents a significant challenge to existing energy grids. As more households and businesses adopt electric heat pumps and electric vehicles, electricity demand is becoming increasingly concentrated during peak hours – typically late afternoons and evenings. This surge in demand can overwhelm grid infrastructure, leading to increased strain on power plants and distribution networks, and potentially triggering outages. Unlike traditional fuel-based systems with more distributed load, electrification amplifies these peak demands, necessitating substantial investments in grid modernization, energy storage solutions, and innovative demand response programs to maintain reliability and prevent cascading failures. Effectively managing this evolving load profile is now paramount to realizing the full benefits of an electrified future.

Successfully managing energy demand requires more than just technological upgrades; it fundamentally hinges on comprehending how and why people use electricity. Demand Side Management (DSM) programs, designed to incentivize altered consumption patterns, often falter when they fail to account for the complex interplay of occupant behavior, comfort preferences, and ingrained habits. Studies reveal that simply offering financial incentives isn’t enough; effective DSM necessitates detailed insights into daily routines, appliance usage, and the psychological factors influencing energy choices. Sophisticated modeling, incorporating behavioral economics and data analytics, is increasingly employed to predict responses to DSM initiatives and tailor programs to specific demographic groups. Ultimately, a shift towards a truly responsive energy grid relies on recognizing that energy demand isn’t solely a technical problem, but a human one.

Decoding the Human-Building Interaction

Human-building interaction is not solely determined by physiological requirements; instead, occupant behavior arises from the interplay of multiple motivational factors. Thermal comfort, encompassing perceptions of temperature and humidity, constitutes a primary influence, but is consistently accompanied by routine motivations – habitual actions performed regardless of immediate physical need – and non-thermal motivations. These latter factors include considerations of social context, aesthetic preferences, and perceived control over the environment. Research indicates that while thermal discomfort prompts a measurable proportion of occupant actions, nearly half are attributable to these non-thermal drivers, demonstrating a complex relationship between building systems and human behavior that extends beyond simple cause and effect.

The Whole Energy Homes Project employed Ecological Momentary Assessment (EMA) to gather detailed data regarding occupant behavior and subjective experiences within a residential setting. This methodology involved prompting participants, via smartphone, to report their thermal comfort, current activities, and thermostat settings at random intervals throughout the day. EMA captured data points in situ, minimizing recall bias and providing a high-resolution temporal record of occupant perceptions and actions. The resulting dataset comprised thousands of responses, detailing the relationship between environmental conditions, occupant behavior, and stated preferences, allowing researchers to analyze patterns beyond those revealed by traditional post-occupancy surveys or building energy monitoring alone.

Data from the Whole Energy Homes Project indicates a consistent pattern of occupants adjusting automated thermostat settings, despite the presence of control systems designed to maintain pre-defined comfort levels. This behavior suggests a misalignment between the algorithmic logic of the building’s energy management system and individual occupant preferences or perceived needs. Specifically, occupants frequently intervened to alter temperature settings, even when conditions were within the system’s established comfort parameters, demonstrating that automated control does not consistently align with, or adequately address, occupant expectations regarding thermal comfort and overall environmental control.

Analysis of occupant behavior within the Whole Energy Homes Project indicates that while approximately 54% of actions related to building control are responses to thermal discomfort, a substantial 46% are driven by non-thermal factors. These non-thermal motivators include established daily routines, such as adjusting settings upon waking or leaving for work, and social influences, where occupants modify the environment based on the presence or anticipated arrival of others. This data demonstrates that occupant interaction with building systems is not solely predicated on achieving thermal comfort, but is significantly influenced by behavioral patterns and social contexts.

Harnessing Demand Flexibility Through Behavioral Insight

Demand Response (DR) programs demonstrate the capability to modulate electricity consumption; however, program efficacy is directly correlated with minimizing instances of occupant override of automated control strategies. Occupant overrides—manual adjustments to thermostats or appliance settings that counteract DR signals—negate the intended load shift and introduce unpredictability into grid management. High rates of override indicate a disconnect between program implementation and occupant behavior, potentially stemming from discomfort, inconvenience, or a perceived lack of control. Therefore, successful DR implementation requires a focus on user acceptance and integration with established occupant patterns to ensure automated adjustments align with comfort preferences and daily routines.

Demand Response programs benefit from personalization based on occupant behavior as demonstrated by the Whole Energy Homes Project. This project identified correlations between thermostat adjustments and occupant activities, indicating that successful demand flexibility strategies require understanding pre-existing routines and preferences. Programs designed without considering these factors are more likely to experience occupant overrides, reducing their effectiveness. Tailoring control strategies to align with established patterns – such as recognizing when occupants are typically home, asleep, or engaged in specific activities – improves program participation and minimizes disruption to comfort levels, thereby maximizing demand reduction potential.

The integration of Building Performance Modeling (BPM) with Demand Flexibility principles enables the development of predictive control strategies for optimized energy management. BPM utilizes data on building characteristics, occupancy patterns, and environmental conditions to simulate energy consumption. When combined with an understanding of how and when occupants are willing to adjust their energy usage – Demand Flexibility – these models can forecast future energy needs and proactively adjust building systems. This allows for pre-emptive control actions, such as pre-cooling or adjusting HVAC setpoints, to shift demand away from peak periods while minimizing disruption to occupant comfort. The predictive capabilities derived from this combined approach move beyond reactive demand response to a more sophisticated and efficient system of energy management.

Analysis of occupant behavior within the Whole Energy Homes Project demonstrated statistically significant correlations between thermostat adjustments and concurrent changes in occupant actions. Specifically, thermostat interactions were significantly associated with changes in activity level (p=5.21e-05), clothing adjustments (p=0.00), and fan usage (p=2.50e-09). These findings indicate that occupants respond to temperature changes by actively modifying their behavior, suggesting that effective demand response strategies require integrated control systems capable of considering these interconnected actions to avoid occupant overrides and maximize energy savings.

Towards a Resilient and Adaptive Energy Future

Optimizing energy consumption hinges on a deeper understanding of how people interact with their buildings, a relationship far more nuanced than simple efficiency metrics suggest. Demand Side Management, traditionally focused on technological solutions, now recognizes that occupant behavior significantly impacts overall energy use. Studies reveal that factors like thermal comfort preferences, daily routines, and even psychological responses to building environments dramatically influence heating, cooling, and lighting demands. Recognizing these complexities allows for the development of more responsive and personalized energy strategies, moving beyond blanket approaches to tailor systems to actual human needs. This responsiveness not only enhances comfort and productivity for occupants but also unlocks substantial potential for reducing peak loads and integrating intermittent renewable energy sources, ultimately fostering a more sustainable and resilient energy future.

The successful integration of intermittent renewable energy sources – such as solar and wind – hinges on a power grid capable of adapting to fluctuating supply and demand. Demand flexibility, the ability to adjust energy consumption patterns in real-time, emerges as a critical component of this adaptive infrastructure. Increasingly, utilities are leveraging behavioral data – insights into how and why people use energy – to incentivize and enable this flexibility. By understanding occupant behavior, grid operators can predict energy needs with greater accuracy and proactively manage demand through dynamic pricing, direct load control, or personalized energy-saving recommendations. This data-driven approach moves beyond traditional, static energy management, fostering a more responsive and resilient energy system capable of accommodating a future powered by renewables and minimizing energy waste.

Buildings traditionally operate on fixed schedules, often disregarding the actual patterns of human activity within them. However, a shift towards occupant-centric energy management promises a future where buildings dynamically respond to the needs and preferences of those who inhabit them. This approach recognizes that comfort and well-being are paramount; by tailoring heating, cooling, and lighting to align with individual schedules and desired environmental conditions, energy waste can be substantially reduced. The integration of smart sensors, data analytics, and automated control systems enables this responsiveness, creating environments that are not only ecologically sound but also enhance the productivity and satisfaction of occupants. Ultimately, this alignment of building systems with human behavior represents a pathway towards genuinely sustainable and resilient infrastructure.

The consistency with which researchers identified underlying motivations for energy-related behaviors significantly strengthens the study’s conclusions. A Jaccard Index of 0.8, a measure of agreement between raters, demonstrates a high degree of reliability in tagging these motivations within the collected data. This robust inter-rater reliability minimizes the risk that observed patterns are due to subjective interpretation, bolstering confidence in the validity of the findings. Consequently, the identified motivational factors can be applied with greater assurance to the development of effective and broadly applicable energy management strategies, informing interventions designed to encourage sustainable practices across diverse populations and building types.

The study meticulously pares away extraneous variables to reveal the core drivers of occupant behavior during demand response events. It demonstrates that motivations extending beyond simple thermal comfort—psychological factors, perceived control, and even social influences—are paramount. This aligns with Andrey Kolmogorov’s assertion: “The greatest discoveries often occur when one abandons preconceived notions and looks at things in a completely new way.” The research effectively strips away the assumption that thermal considerations alone dictate responses, illuminating a more nuanced and accurate understanding of human-building interaction. It isn’t merely about adding layers of complexity to existing models, but rather removing the false simplicity that obscured the true dynamics at play.

What Remains to Be Seen

The pursuit of demand flexibility, predictably, has not yielded to simple thermal equations. This work demonstrates, with a commendable austerity, that human response is burdened by factors beyond mere temperature. To insist on thermal comfort as the primary driver is to mistake a symptom for the disease. It reveals a foundational problem: the field attempts to predict behavior, when it should be striving to understand it. The accumulation of data, however voluminous, will not reveal intent.

Future efforts must resist the urge to add complexity. More sensors, more variables—these are the tools of those who fear admitting they do not grasp the underlying principles. Instead, the focus should shift toward discerning the core motivations—the genuinely irreducible needs—that dictate occupant action. This necessitates a move beyond passive observation and toward methods that can probe the why, not merely the what.

The ultimate test will not be the accuracy of a predictive model, but the elegance of its explanation. If the behavior cannot be articulated simply, it is not understood. The goal, therefore, is not to build a more intricate simulation of human fallibility, but to expose the surprisingly few, fundamental truths that govern it.

Original article: https://arxiv.org/pdf/2511.10467.pdf

Contact the author: https://www.linkedin.com/in/avetisyan/

See also:

2025-11-15 21:13