Dynamic Framework to Quantify Thermal Resiliency and Passive Survivability During Prolonged Air Conditioning Outages in Extreme Hot Climate: Arizona Case Study

This dataset supports the evaluation of thermal resiliency and passive survivability of residential buildings during prolonged air-conditioning (AC) outages under extreme hot-arid climate conditions in Arizona, United States. It provides field-validated evidence of residential heat exposure and recovery dynamics during extreme events, supporting strategies to protect public health, strengthen grid resilience, and inform climate-adaptive building design and policy. High-resolution field measurements were collected during controlled AC shutdown experiments lasting approximately 77-78 hours in two occupied apartment units: a two-bedroom apartment in Mesa, Arizona, and a one-bedroom apartment in Chandler, Arizona. Indoor dry-bulb temperature (°F) and relative humidity (%) were recorded using distributed data loggers positioned at multiple locations within each unit, including mid-room zones, near ceilings, near exterior walls, and adjacent to windows. These measurements capture spatial and temporal variations in indoor environmental conditions during both disruption (cooling loss) and recovery (cooling restoration) phases. The dataset includes both raw and processed data. Raw data consist of time-series indoor environmental measurements, while processed data include derived thermal resiliency metrics and analytical outputs. The analysis framework integrates multiple performance indicators to quantify building response under extreme heat stress: Exposure Metrics (E): Thermal Exposure Index (TEI) and Intensity Index (II) Delay Metrics (Del): Delaying Index (DI) Dynamic Metrics (D): Thermal Time Constant (τ) and Thermal Lag (φ) These metrics are normalized and combined to compute a Thermal Resiliency Index (TRI), defined as a weighted function of exposure, delay, and dynamic response (TRI) = w₁E + w₂Del + w₃D, with greater emphasis on exposure to reflect occupant risk conditions. To capture post-outage performance, a Recovery Resiliency Index (RRI) is also calculated based on time required to return to setpoint temperature and cooling recovery rate. A composite thermal resiliency index is developed by combining TRI and RRI, ensuring that both disruption severity and recovery dynamics are represented. In addition, the dataset includes a predictive modeling approach for indoor temperature evolution during extended outages. The model combines an exponential formulation during the transient phase with a sinusoidal representation after thermal stabilization, enabling estimation of indoor conditions over prolonged outage durations. Unlike prior studies that rely predominantly on building energy simulations, this work provides empirically measured, field-validated data capturing both outage and recovery phases. The dataset addresses a critical gap in measurement-based research on building thermal resilience. This resource is intended to support research and applications in: Extreme heat risk assessment Indoor environmental quality during power outages Climate-adaptive and resilient building design Energy system resilience and demand-side planning All analysis steps are reproducible using the provided documentation, with the exception of the multivariate regression used in the predictive model. Additional details are available in the accompanying README and analysis files.