Windows in new buildings in Korea are equipped with double or triple glazing, a low-E coating, and gas injection between the panes of glass, in accordance with the regulations for the reinforced insulation of windows. However, these windows have certain issues, such as thermal transmittance variations and injected gas leakage. In this study, we investigated the current status of double-glazed windows filled with argon gas and analyzed the actual deterioration degree in terms of insulation performance via field experiments and simulations. Accordingly, we manufactured a 26-millimeter low-E double-glazed window test specimen and conducted tests in accordance with the KS F 2278 standard. The test results indicated that the thermal performance decreased by 10.9% when the argon gas filling rate was reduced from 95% to 0%. The simulation results showed that the thermal performance of windows that were insulated using only glazing decreased by 22.6% with the decrease in the argon gas filling rate; the thermal performance of the double-glazed windows also decreased by 13.6%. A comprehensive analysis of the field surveys and simulations predicted a 92% probability that the argon gas filling rate of double-glazed windows would be below 65% by two years after completion, in addition to a ~4.3% decrease in thermal performance.

With global efforts to strengthen various energy-saving policies for buildings to reduce greenhouse gas emissions, in South Korea, new laws and regulations have been in force since May 2015 to install shading devices in public buildings and to include the solar heat gain coefficient (SHGC) reduction performance of shading devices in the evaluation of building performance. By making a ventilated air layer outer glass and inner glass to lower the temperatures of the air layer and glass surface, it is possible to reduce the amount of heat flowing into the building while maintaining the same level of light transmission as plain window systems. This study proposes a double-skin façade window with a 20 mm ventilated air cavity, and assumes that insolation inflow indoors would be reduced through ventilation in the air cavity. The artificial solar lab test results show that the SHGC can be lowered through ventilation by 28% to 52.9%. Additionally, in an outdoor test cell experiment, the results show that the mean temperature was 0.6 K and the peak temperature was 0.9 K lower with ventilation in the air cavity than that without ventilation in the air cavity.

This study presents a modular diagnostic framework for evaluating thermal degradation in aging building envelopes by integrating infrared thermography, panoramic reconstruction, and deep learning-based semantic segmentation into a unified workflow. The methodology combines image registration, panoramic synthesis, façade component segmentation, and quantitative surface temperature analysis to provide scalable and reproducible diagnostics. By excluding fenestration zones—where infrared measurements are physically unreliable—the framework focuses on opaque wall regions and window surroundings to ensure physically meaningful evaluation. Field validation was conducted on a multi-story office building constructed in 1996. The diagnostic indicators revealed a mean wall surface temperature of 14.3 °C with a standard deviation of 5.6 °C, and a temperature factor ranging from 0.67 to 0.78 under measured conditions. The vulnerable area ratio reached 9.1% for walls, while window areas showed greater vulnerability at 12.74%, with anomalies concentrated at frame–glass interfaces and perimeter seals. These quantitative results confirmed the framework’s ability to detect thermal irregularities and visualize localized anomalies. More importantly, the contribution of this study lies in establishing a systematic and extensible diagnostic pipeline that advances building envelope analysis, supporting large-scale energy audits, retrofit prioritization, and sustainable building management.

Existing relative benchmarking of building energy use relies on the registered ”principal use” recorded in the building register to define peer groups and to derive energy-use intensity (EUI, kWh·m⁻²·yr⁻¹). This assumption breaks down for mixed-use buildings, where multiple activities with very different operating schedules, occupant densities and equipment loads coexist in a single building. Type 1 and Type 2 neighborhood living facilities in the Republic of Korea, which account for the largest share of the non-residential building stock, are the most extreme example: a single registered use masks restaurants, retail, offices, clinics and lodging, producing an unusually wide EUI distribution that a single-profile evaluation cannot represent fairly. This study develops a reproducible, non-intrusive benchmarking model that reconstructs the actual activity mix of a building from open public data and evaluates each activity within a like-for-like peer group. Building-register records, monthly metered energy use and business-license (operating-permit) records were linked through map-based spatial join, and a two-step cluster analysis was used to consolidate 53 raw activity types into 15 energy-use peer groups. Within each peer group the EUI was standardized to a z-score and converted to a 1–9 stanine grade; the building-level grade was obtained as the floor-area-weighted sum of activity grades. The model was applied to a district-scale dataset comprising 3,001 neighborhood-facility buildings in Mapo-gu, Seoul (17,413 linked business-license records). Compared with the conventional EUI grade, buildings previously assigned to a single grade were redistributed across grades 2–9, and a change-point disaggregation of metered data showed that, for the most energy-intensive group, the coefficient of variation of cooling and heating energy fell by 33.4% and 49.1%, respectively. The proposed model therefore groups buildings with genuinely similar end-use profiles, providing a low-cost screening approach for retrofit prioritization and building decarbonization strategies.