Accurate weather data is essential for building energy modeling (BEM), yet the actual urban local weather condition is often overlooked. This study developed an approach to generate local weather data using ERA5, a global atmospheric reanalysis dataset as input for two urban land surface models, Urban Tethys-Chloris (UT&C) and Urban Weather Generator (UWG). The generated datasets (UT&C-ERA5 and UWG-ERA5) are compared to locally measured weather data for a university campus in Singapore. Results show that the original ERA5 underestimates the diurnal temperature range. UT&C-ERA5 significantly improves hourly dry bulb temperature, reducing Mean Absolute Error (MAE) from 1.73 to 1.32 and Root Mean Square Error (RMSE) from 2.31 to 1.67, while UWG-ERA5 shows modest improvements (MAE from 1.73 to 1.70, RMSE from 2.31 to 2.22). UT&C-ERA5 also improves wind speed, lowering MAE from 2.85 to 1.54 and RMSE from 3.23 to 1.79. Subsequently, these weather datasets are employed as inputs for a calibrated BEM. Compared to the original ERA5, UT&C-ERA5 reduces CV (RMSE) of building cooling load from 17.13 % to 15.45 %. By leveraging the global availability of atmospheric reanalysis datasets, this approach can support building energy design and improve energy efficiency in tropical cities.

Many buildings suffer from operational inefficiencies, leading to uncomfortable indoor environments, poor air quality, and significant energy waste. Developing automatic fault detection and diagnosis (FDD) tools in building energy systems is essential to mitigate these issues, reducing both energy waste and maintenance costs. Diagnostic Bayesian networks (DBNs), as probabilistic graphical models, offer a promising solution due to their interpretability, robustness to uncertainty, scalability, and flexibility. In this paper, the practical applications of DBNs for FDD in building energy systems are comprehensively reviewed. The generic modeling procedure is systematically examined and summarized, covering problem formulation, structure modeling, parameter modeling, and fault isolation and evaluation. Then, the paper provides insights into DBN modeling objectives, modeling types, diagnostic samples, and modeling software based on the 43 key relevant papers. Furthermore, the paper discusses practical challenges such as sensor configuration, baseline estimation, threshold determination, and expert knowledge integration. Finally, the recommendations are provided to guide further research, aiming to enhance DBN implementation for building energy systems in real-world scenarios, thereby supporting the transformation of the building service industry into a smart sector and ultimately improving building energy performance.

Urban Building Energy Modeling (UBEM) has become a critical tool for developing local heating and cooling plans, as required by the European Union. Despite growing interest, the reproducibility and reliability of UBEM studies remain limited due to data scarcity and workflow complexity. This paper presents a comprehensive framework to evaluate the data pipeline in UBEM, with a particular focus on data acquisition, enrichment, simulation, calibration, and information application. The approach applies three distinct UBEM workflows (CityEnergyAnalyst, DistrictGenerator, and SimStadt) to the Mierendorffinsel district in Berlin, Germany. We compare the based on quantitative performance metrics and qualitative framework criteria. The results highlight the influence of data sources, archetype definitions, and geometric preprocessing on simulation outcomes. The CEA models consider between 365 and 646 buildings, depending on the scenario. The study provides guidelines for practitioners to enhance model transparency, reproducibility, and accuracy in urban energy modeling. Although official data provide more accurate building functions, the geometries need extensive preprocessing. The residential archetypes are far more refined, e.g., in the status of renovation, which reflects the amount of residential buildings compared to nonresidential buildings. We show that evaluation threshold criteria for the district level are scarce and evaluate multiple metrics. Results depend on the selected evaluation method, but observed differences are generally higher in case of nonresidential buildings, with differences of more than 300 kWh/m2 for several nonresidential building types. The heated area considered differs up to a factor of 1.9, because of different buildings, metadata, and calculation approaches.

In a global context of increasing flexibility in the way workplaces and the districts in which they are located are used, there is a need for occupant-driven approaches to plan urban energy systems. Several authors have suggested the use of agent-based models (ABM) of building occupants in urban building energy simulations due to their ability to reproduce emergent behaviors from individual agents’ actions. However, few works in the literature take full advantage of the ABM paradigm, accounting for both occupant presence and energy-relevant behaviors at the district scale. In this work, we propose a methodology to develop a data-driven, agent-based model of building occupants’ activities and thermal comfort in an urban district. Our methodology combines the use of campus-scale Wi-Fi data to derive feasible occupant activity and location plans, along with thermal preference profiles derived from a longitudinal field study where off-the-shelf, non-intrusive sensors were used to capture the right-here-right-now thermal preference of 35 participants in the same case study district. Our model is then used to explore how different district operation strategies could affect building energy performance in the context of increased workspace flexibility. Our results show that by providing a diversity of building operation conditions, with different buildings having different set point temperatures, occupants’ thermal comfort hours could be improved by an average of about 10% with little effect on district energy performance. Meanwhile, a 6%–15% average decrease in space cooling energy use intensity was observed when implementing occupant-driven ventilation and setpoint controls, regardless of location choice scenario.

Building energy modeling (BEM) is essential for predicting energy use and improving thermal performance in buildings. Traditionally, weather data for BEM comes from built-in tool datasets. Additionally, global atmospheric reanalysis datasets like ERA5, have been used in recent years for BEM. However, the spatial resolution of global atmospheric reanalysis datasets is generally coarse relative to cities, limiting their accuracy in capturing local urban climate effects. Adopting ERA5 as the forcing data, this study examines the use of two urban land surface models, Urban Tethys-Chloris (UT&C) and Urban Weather Generator (UWG), to generate localized weather data for Singapore. The generated local weather data are compared with the data from an on-campus weather station and other weather datasets. Subsequently, these weather datasets are employed as input for an educational building’s energy model that has been validated with energy meter data. The results demonstrate a better agreement between the generated local weather data and locally measured data, compared to the original ERA5 data and typical meteorological year weather data. This leads to an improved accuracy in building energy prediction. By leveraging the global availability of atmospheric reanalysis datasets, this framework for generating local weather data can serve as a universally applicable approach to support building energy design in tropical cities.

The widespread availability of open datasets in cities is transforming the way urban energy systems are planned, simulated, and visualized. In this paper, a cross-scale approach is pursued to better understand the reciprocal effects between building energy performance, the urban climate, and urban dwellers’ indoor and outdoor thermal comfort. On the one hand, monthly building electricity and gas demand data at the parcel level was collected, along with hourly weather station data at the urban scale. On the other hand, a longitudinal experiment was carried out in which 22 participants wore smartwatches for 4–6 weeks and filled out hourly micro surveys on their activities, location, and thermal comfort. In addition to survey responses, the smartwatches collected participants’ physiological data and location throughout the period of the study. The project was conducted in Seoul, South Korea, the highest-ranked Asian country in open data readiness, implementation, and impact. This paper reports on the data collection effort and provides some preliminary analysis of the results. The work carried out is expected to help develop methodologies for the convergence of district-scale and occupant-scale data in urban areas. A number of expected applications are proposed, including urban-scale studies on the impact of urban form on the local climate and building energy performance, district-to-building-scale building energy simulations accounting for occupant thermal comfort-related behaviors, and district-scale analyses of occupants’ outdoor thermal comfort and its relationship with location and wayfinding in urban areas.

Energy waste in buildings can range from 5% to 30% due to faults and inadequate controls. To effectively mitigate energy waste and reduce maintenance costs, the development of Fault Detection and Diagnosis (FDD) algorithms for building energy systems is crucial. Diagnostic Bayesian Networks (DBNs), as graphical probability models, are particularly useful in scenarios where high-quality data is not always available. While many studies have focused on single fault detection using DBNs, the occurrence of multiple simultaneous faults is common, yet the versatility of DBNs in handling such cases is rarely explored. This study adapts a DBN, initially designed for single fault diagnosis, to perform simultaneous fault diagnosis Experiments were conducted on an air handling unit (AHU) in the Netherlands, using implemented simultaneous faults to test the model. The results suggest that the DBN can detect both single and multiple faults effectively.

With the increasing stock of ageing infrastructure and resource constraints in Singapore, related risks and carbon emissions can be mitigated through long-term resilience planning, automated building inspection, and effective maintenance. Sustainable actions are needed to maintain Singapore's ageing infrastructure. Hence, a state-of-the-art control and management system is required in the form of smart city digital tools. We introduce an Urban Digital Twin (UDT)—GHG App for decision-makers in Singapore's operational building greenhouse gas (GHG) emission mitigation and decarbonisation initiatives. Based on multiple-criteria decision analysis (MCDA), a Potential for Intervention (PFI) map was created to rejuvenate the building system. Decision-makers can use this map to prioritise the rejuvenation of low-carbon building systems in the built environment. A heat map of the PFI results highlights which buildings need urgent rejuvenation based on critical parameters. The GHG App utilises this method to generate maps and enables users to modify parameter weights based on their priorities, automatically updating the map. Users can plan an intervention for buildings with higher PFI values once the map is generated. The GHG App provides interactive data visualisation of 119,872 features representing Singapore's built environment, including the context size of 6,785 existing residential buildings modelled and used to demonstrate the analysis results. Our research findings can contribute to the development of standards for accounting for operational GHG emissions, setting emission limits, and planning decarbonisation in the built environment sector.

Fault detection and diagnosis (FDD) provides several interrelated benefits, including reducing energy waste, enhanced operational efficiency, and maintaining indoor comfort. The initial step in FDD is to detect deviations from normal or expected operation. However, establishing a reliable baseline can be challenging, especially when there is a lack of sufficient system documents or when complex control strategies are involved. This study investigates three feature selection methods for the baseline estimation: expert knowledge-based, correlation-based, and causality-guided, using heating coil valve control estimation as an example. These methods were tested in an office building in the Netherlands. The results show that while the correlation-based method achieved the best estimation, it may lead to false negatives due to features with reverse causality. This study aims to emphasize the necessity of causal analysis in the baseline estimation to achieve reliable FDD in buildings.

The expansion of solar energy in high density cities highlights the crucial need for optimal capacity planning in building-integrated photovoltaic (BIPV) systems. However, uncertainties present significant challenges for robust planning in these systems. Addressing this challenge, this paper proposes a distributionally robust optimization (DRO) model employing scenario robust ambiguity sets for BIPV system expansion planning to handle uncertainties in both demand and solar irradiance. This model stands out for its comprehensive incorporation of factors including the dynamics of climate change, the properties of building geometry, the intricate non-linear correlation between solar generation potential and global horizontal irradiance and uncertainty handling. The model can be transformed into a tractable formulation using the linear decision rule approach. In evaluating the model’s performance in data uncertainty handling, the study compares it against three alternative approaches – deterministic optimization, stochastic programming, and robust optimization – in a case study for BIPV system expansion planning for a college in Singapore. The findings demonstrate the superiority of the proposed model: it achieves a 5%–12% reduction in grid reliance, saves electricity cost by 3%–10% and reduces carbon dioxide (CO2) emission by 3%–10% compared to the benchmark approaches. These results underscore the capability of the proposed model in effectively handling data uncertainties in demand and irradiance in the BIPV system expansion planning.

The increasing availability of urban-scale, open-access datasets can support decision-making in urban planning, in particular in relation to climate resilience and climate change mitigation. Such data-driven initiatives however often neglect the central role of urban dwellers, whose activities create the demand for energy and mobility in urban areas. This is due in large part due to the difficulty of data collection at this scale, along with privacy concerns arising from any such data collection effort. The use of wearable technologies for self-reported comfort feedback from urban dwellers provides a promising opportunity for citizens to actively participate in the adaptation of urban areas to better support outdoor comfort and climate resilience.

In this work, subjective feedback data from 22 participants in a longitudinal test in Seoul, South Korea was collected through a smartwatch application. Participants were required to wear a smartwatch for 4–6 weeks, during which time their location as well as environmental and physiological data were collected. Participants were also requested to complete hourly micro-surveys, in which they were asked about their activities, location, thermal preference, clothing level, comfort adaptations, and mood. This information was complemented by an urban scale dataset comprising building geometries and data from 1000+ weather stations over the same period.

This cross-scale dataset was then used to investigate the relationship between urban form and environmental parameters with occupants’ survey responses. The relationship between indoor comfort and environmental parameters in the case study is discussed, with recommendations for further research into this topic. The use of machine learning to leverage the combination of spatial, temporal, and subjective preference data to predict occupants’ outdoor comfort as a function of their urban environment is also explored.

The growing adoption of clean solar energy in urban environments emphasizes the critical importance of efficient capacity planning in the development of building-integrated photovoltaic (BIPV) systems. Nevertheless, uncertainties associated with solar generation and demand pose substantial obstacles to systems planning. In response to these challenges, this paper introduces a novel approach: an optimization model that leverages ambiguity sets based on scenarios and is robust to distributional uncertainty for the expansion planning of building-integrated photovoltaic (BIPV) system. The distinctiveness of the model lies in its consideration of the non-linearity between global horizontal irradiance as well as PV solar power generation potential. A case study conducted in Singapore demonstrates the proposed model's substantial advantages including a 2 % -5 % improvement in self-sufficiency, 3 % -8 % electricity cost savings, and 3 % -8 % reduction in carbon dioxide emissions compared to benchmark models.

Automated fault detection and diagnostics (FDD) can support building energy performance and predictive maintenance by leveraging the vast amounts of data generated by modern building management systems. Diagnostic Bayesian Networks (DBN) offer a particularly promising approach due to their robustness, flexibility and scalability. However, FDD applications in whole building systems are rare, as they require the integration of different building subsystems, with their own potential faults and symptoms, which increases complexity and makes the resulting DBNs system-specific. In order to overcome these limitations, the 4S3F (four symptoms and three faults) method offers a simplified, adaptable framework for FDD implementation across building systems. In this paper, we implement the 4S3F methodology to a whole-building HVAC system in a case study office building located in the Netherlands. Our methodology uses generic, aggregated representations of individual subsystems within the building, such that FDD methods for specific subcomponents can later be incorporated where available. We first define aggregated building system groups (boiler group, chiller group, hydronic groups, ventilation groups, and end user groups) and subsequently define generic faults that can be detected with the existing sensor infrastructure. This simplified system representation is then used to define a DBN to isolate the most probable system-level faults that lead to building-level symptoms. By focusing on the whole building system, this work aims to provide the groundwork to incorporate occupant feedback and behavior in FDD.

The widespread availability of open datasets in urban areas is transforming how urban energy systems are planned, simulated, and visualized. Urban energy models, however, require an understanding of urban dwellers, as their activities create the demands for energy in buildings. In this paper, we explore using campus-scale Wi-Fi data to identify typical occupant activity patterns as an input to an agent-based model of building occupants at the district scale. The data is taken from a Singaporean university's Wi-Fi network at high resolution. Each record comprises a timestamp, a device identifier, the location of the device within the campus, and the access point to which it is connected. The Wi-Fi dataset contains 120 different buildings on campus and 10,300 anonymized individual devices. Activities are then assigned to each location on campus according to the building use type. In order to test the methodology, the activity plans of 27,604 undergraduate students, 8,304 graduate students, and 12,018 employees were simulated over a workweek. The results show the model's ability to produce plausible activity plans but could be improved by implementing sampling rules and expanding the source dataset to include off-peak dates. Nevertheless, using such an agent-based modeling approach at the district scale appears to be a promising methodology to assess the impacts of different planning strategies on occupant behavior and district energy demand.

Smart city initiatives have been a driving force for city-level dataset collection and the development of data-driven applications that benefit effective city management. There is a need to demonstrate use cases for effective city management using the available dataset. Urban Digital Twin (UDT) is a 3D city model that can integrate multi-disciplines and improve systems operability on a digital platform. However, UDTs are developed within organisations, and there is only limited availability of authoritative open 3D datasets to explore the potential of UDT concepts. This paper reports a methodology for creating a UDT platform for visualising and querying city energy data. We demonstrate a bottom-up approach to constructing an integrated 3D city dataset and create a query system for rapid access and navigation of the 3D city dataset through a visualisation platform using Cesium Ion. Various use cases are explored based on the dataset, such as building material stock management, energy demand simulation, electric vehicles (EV) demand and flexibility, and estimation of greenhouse gas (GHG) emissions. These use cases can help decision-makers and stakeholders involved in city planning and management. Furthermore, it provides a guideline for developers willing to create UDT applications for smart city initiatives.

This work proposes the use of a data-driven, agent-based model of building occupants’ activities and thermal comfort in an urban university campus in order to assess how district operation strategies can be leveraged to support the transition to flexible work arrangements. The results show that when users are given the flexibility to pursue more comfortable workspaces, they are still comfortable only 58% of the time.

Before 2020, the way occupants utilized the built environment had been changing slowly towards scenarios in which occupants have more choice and flexibility in where and how they work. The global COVID-19 pandemic accelerated this phenomenon rapidly through lockdowns and hybrid work arrangements. Many occupants and employers are considering keeping some of these flexibility-based strategies due to their benefits and cost impacts.

This paper explores how demand-driven control strategies in the built environment might support the transition to increased workplace flexibility by simulating various scenarios related to the operational technologies and policies of a real-world campus using a district-scale City Energy Analyst (CEA) model that is calibrated with measured energy demand data and occupancy profiles extracted from WiFi data. These scenarios demonstrate the energy impact of ramping building operations up and down more rapidly and effectively to the flex-based work strategies that may solidify. The scenarios show a 5–15% decrease in space cooling demand due to occupant absenteeism of 25–75% if centralized building system operation is in place, but as high as 17–63% if occupancy-driven building controls are implemented. The paper discusses technologies and strategies that are important in this paradigm shift of operations.

District-scale energy demand models can be powerful tools for understanding interactions in complex urban areas and optimising energy systems in new developments. The process of coupling characteristics of urban environments with simulation software to achieve accurate results is nascent. We developed a digital twin through a web map application for a 170ha district-scale university campus as a pilot. The impact on the built environment is simulated with pandemic (COVID-19) and climate change scenarios. The former can be observed through varying occupancy rates and average cooling loads in the buildings during the lockdown period. The digital twin dashboard was built with visualisations of the 3D campus, real-time data from sensors, energy demand simulation results from the City Energy Analyst (CEA) tool, and occupancy rates from WiFi data. The ongoing work focuses on formulating a resilience assessment metric to measure the robustness of buildings to these disruptions. This district-scale digital twin demonstration can help in facilities management and planning applications. The results show that the digital twin approach can support decarbonising initiatives for cities.

This paper presents a digital twin of a university campus in Singapore as a demonstrator for a digital-twin enabled approach to district energy resilience. This paper focuses mainly on the development of the building energy and occupancy models in the digital twin, which are complemented by a user interface for real-time data visualization and scenario assessment. The building energy demand model of the case study area was calibrated using measured hourly cooling and electricity data collected in the case study area. Occupant presence was estimated using WiFi connection counts, and a simple regression model was developed to assign electricity loads as a function of occupant presence and time of day. The digital twin’s scenario assessment capabilities were explored through scenarios on the long-term effects of climate change and of the increase of remote working and studying as a result of the COVID-19 pandemic. Four different “work-from-home” cases were considered, and three different building operation strategies were assumed for each case. The results show that a decrease in building occupancy post-COVID-19 would lead to minimal space cooling savings in the case study area unless building operation was proactively adjusted to adapt to the new needs of the campus.

Input uncertainty is one of the major obstacles urban building energy models (UBEM) must tackle. The aim of this paper was to quantify the effects of two of the main sources of stochastic uncertainty, namely building occupants and urban microclimate, on electrical and thermal supply system sizing at the district scale. In order to analyze the effects of the former, three different methods of occupant modeling were implemented in a UBEM. The effects of the urban heat island on system sizing were studied through the use of measured temperature data from a weather station in the case study district compared to measured data from a national weather station. The methods developed were used to assess the sizing and costs of centralized and decentralized technologies for a case study in central Zurich, Switzerland. The choice of occupant modeling approach was found to affect the district’s total annualized costs for space heating and cooling by ±5%, whereas for the costs of electricity the variation was ±8%. Regarding outdoor temperature, the effects on the heating demands proved be negligible, however the costs of the cooling alternatives were found to vary by about 4% at the district scale due to the effect of urban climate, for individual buildings this deviation was as high as 40%.