Introduction

Urbanisation presents a dual challenge of supporting population growth and addressing the climate crisis. Cities are at the core of this problem, generating over 70% of global carbon dioxide emissions and consuming more than two-thirds of the world’s energy1. Improving urban energy efficiency and reducing emissions is essential to achieving climate goals. Simultaneously, a significant proportion of the world’s built heritage is concentrated within urban environments, where historic buildings and districts constitute integral components of cultural identity and urban form2. These heritage assets are increasingly seen not as static relics, but as catalysts for sustainable growth and innovation3. Beyond material form, they hold intangible values, such as identity, memory, and knowledge, that strengthen urban resilience4. Their preservation supports multiple United Nations Sustainable Development Goals (SDGs), including poverty reduction (SDG 1), sustainable cities (SDG 11), climate action (SDG 13), and strong institutions (SDG 16). Global frameworks like UNESCO’s Historic Urban Landscape (HUL) approach further advocate for integrating heritage into urban planning and sustainability policy5,6. International conservation bodies like ICOMOS echo this perspective: for example, the 2011 Paris Declaration frames heritage as a driver of sustainable development, highlighting that conserving cultural assets can foster social cohesion, environmental stewardship, and economic prosperity within communities7. This recognition situates heritage conservation at the intersection of urban environmental policy and sustainability agendas8.

However, a critical tension arises from this intersection in seeking to balance reduced energy consumption in cities with the protection of cultural heritage. Historic buildings are integral to urban identity and cultural continuity, yet many face degradation due to age, environmental exposure, and outdated systems9. Renovation is essential to ensure their continued use and resilience. These buildings often contain passive design features, such as thermal mass and natural ventilation, offering lessons for low-carbon design. Moreover, adaptive reuse avoids the embodied emissions of new construction10. Nonetheless, such interventions must prioritise the conservation of architectural authenticity and material integrity, especially in buildings with high cultural significance. While energy efficiency and carbon reduction are increasingly critical in the built environment, these goals should be carefully balanced against the preservation of historic character11. In practice, retrofit decisions that maximise energy performance can sometimes conflict with heritage values: beyond physical fabric, insensitive alterations may erode the intangible “spirit” or ambience of historic places that contribute to their cultural significance. For instance, Atmaca et al.12 found that restoring a heritage building resulted in increased embodied energy and carbon emissions, emphasising the need for careful planning to minimise environmental impact. Similarly, other researchers discuss the importance of integrating energy efficiency measures in a way that respects the cultural significance of historic buildings13,14,15. Therefore, renovation strategies should adopt a holistic approach, considering both the preservation of heritage and the implementation of sustainable practices to achieve long-term environmental and cultural benefits. Developing such strategies requires tools capable of balancing preservation objectives and energy goals in a transparent, evidence-driven manner.

Urban Building Energy Modelling (UBEM) is particularly suitable for addressing this challenge. UBEM is an analytical methodology rooted in bottom-up engineering principles, designed specifically to assess operational performance across large groups of buildings at an urban or district scale16. It integrates detailed data on building characteristics, energy systems, occupancy patterns, and environmental interactions, thus providing comprehensive insights into energy usage and potential efficiency improvements17. However, applying UBEM within historic urban contexts encounters several challenges, including limited data availability, complex heritage typologies, and constraints imposed by existing preservation policies18. A fundamental disconnect further complicates matters, which is that the qualitative, subjective language prevalent in heritage conservation contrasts significantly with the quantitative, data-driven methodologies utilised by UBEM19. This also limits integration of local values and community perspectives into energy decision-making20. This disparity hampers effective integration of conservation objectives and energy strategies, potentially resulting in either suboptimal heritage outcomes or inefficient energy interventions. Although UBEM is widely employed in energy analysis of building clusters21, its comprehensive integration with heritage buildings at the neighbourhood scale, especially within historic districts, remains largely underexplored. These areas concentrate culturally significant buildings, are subject to complex conservation regulations, and often reflect heightened community values, making energy intervention particularly sensitive and impactful. Integrating heritage considerations into UBEM aligns with the HUL approach of inclusive, context-aware urban management, and is increasingly imperative as cities strive to reconcile climate action with cultural continuity.

To address this complexity, this study develops a comprehensive retrofit framework that systematically integrates heritage value considerations into UBEM processes at the historic neighbourhood scale. The framework establishes a structured methodology that bridges the qualitative language of heritage conservation with the quantitative requirements of energy modelling, creating a transparent decision-making tool for retrofit interventions. Using the Wukang Road area, a heritage-rich urban district in Shanghai, as the case study, this framework operationalises the integration of cultural significance assessments, regulatory constraints, and building typological characteristics within the energy modelling workflow. It is important to note that this study adopts a technical, simulation-based approach and does not incorporate stakeholder engagement, community-based value elicitation, or empirical validation. As such, the findings are intended to demonstrate methodological feasibility and scenario-based insight rather than to provide verified behavioural or social outcomes. Within this defined scope, the framework provides a replicable methodology that enables context-sensitive energy analysis whilst respecting statutory conservation controls. The primary contribution lies in transforming the theoretical alignment between heritage management and urban energy planning into a computational framework suitable for policy testing and sustainable urban development. In doing so, it advances UBEM practice by embedding formal heritage constraints as core parameters in retrofit evaluation and by supporting more nuanced, conservation-aware strategies for historic districts.

The rest of this paper is structured as follows. Section 2 details the methodology, including the case study context, data acquisition processes, and the development of the heritage-aware retrofit framework. Section 3 presents the baseline simulation results and comparative scenario analyses conducted under heritage constraints, examining how the proposed framework influences retrofit decision-making and evaluating the implications for heritage conservation policy and sustainable urban planning practice. Section 4 concludes the study by summarising the key findings and contributions.

Methods

In order to reconcile the dual objectives of energy retrofitting and heritage conservation within historic urban districts, a methodological framework is required that is both quantitatively robust and contextually adaptable. This study employs the Wukang Road area as a case study to develop an UBEM-based approach that systematically incorporates conservation constraints. This section outlines the contextual background, data acquisition and processing techniques, and the development of a tiered intervention framework based on heritage protection classifications.

Introduction of case study area

The Wukang Road area is a designated historic neighbourhood located in Shanghai’s former French Concession, now spanning parts of Xuhui and Huangpu Districts. This 1.18 km tree-lined avenue and its surrounding blocks are celebrated for their cosmopolitan heritage character. Dozens of early 20th-century buildings, at least 37 of which are formally protected as cultural heritage, line Wukang Road, showcasing diverse architectural styles ranging from French Renaissance and Mediterranean to Art Deco influences. The district’s rich ensemble of small-scale residences of various countries reflects Shanghai’s modern history as an international metropolis. Many structures in the area carry historical significance beyond their architecture. For example, several villas and apartments were once home to famous writers, politicians, and scientists, embedding the neighbourhood with layers of social memory and cultural nostalgia. In recognition of its outstanding heritage value, Wukang Road was listed in 2011 as one of China’s National Historic and Cultural Streets, and municipal authorities have delineated it as a protected historical and cultural zone.

Preservation regulations in this area are notably strict, governing not only the buildings themselves but also the streetscape, including the iconic plane trees that contribute to the historic ambience. At the same time, Wukang Road remains a lively urban quarter with contemporary functions. Historic residences have been adaptively reused as cafés, boutiques, galleries, and other community amenities, demonstrating a living heritage that continuously serves local needs. This mix of uses means that buildings originally designed for one purpose, such as colonial-era dwellings or mansion apartments, may now host different functions, such as shops or cultural venues, leading to diverse occupancy patterns and energy usage profiles. Related Shanghai studies also note discontinuities in place identity under rapid change, which are pertinent to conservation trade-offs in such adaptive reuse contexts22. In summary, the Wukang Road area offers a distinctive context of high heritage significance, a clear conservation zoning framework, and varied building typologies, making it an ideal testbed for exploring how targeted energy retrofits can be balanced with heritage conservation imperatives in a real-world urban setting. Fig. 1 situates the case study area within Shanghai and provides an overview of its urban fabric and historic streetscape.

Fig. 1
Fig. 1
Full size image

Geographic location and presentation of the case study area.

Building data acquisition

This study acquires urban building data from open-source geographic platforms including BaiduMap and AMap, supported by Geographic Information System (GIS) technologies. Key morphological attributes, including building footprint and height, are extracted from high-resolution satellite imagery using image segmentation for footprint delineation and geometric inversion methods based on shadow length and solar angle for height estimation, following approaches validated in prior remote-sensing-based urban studies conducted in China23. A two-level classification system, adapted from the Chinese Land Use Classification Standard, identifies building functions and types by integrating geospatial big data, crowdsourced inputs and parcel division based on road networks24. The functions fall into three main categories, namely domestic, public and industrial. Further types of building could be developed based on these three categories. Public buildings encompass commercial premises such as shopping centres, workplaces including office buildings, and administrative facilities such as government or public service institutions. Industrial buildings include warehouses, factories and transport-related structures such as stations and associated buildings. Together, these types represent the predominant urban building forms found across China. To ensure data quality, the study applies thresholds to projected area and perimeter-to-area ratio to exclude temporary or irregular structures. Manual verification using satellite imagery removes buildings with unsuitable roofs such as non-concrete surfaces. A site visit is also conducted to validate the accuracy of the building data. The final dataset is managed using QGIS with the WGS 1984 Albers projection, and Python tools including OWSLib, Shapely and GeoPandas are used for geometric processing and attribute classification.

In addition to acquiring the building morphology dataset, it is also necessary to define simulation parameters for each building category. These parameters are subsequently used in EnergyPlus software to perform dynamic energy modelling at an hourly resolution over the course of an entire year, thereby capturing seasonal variations and time-dependent performance across the building stock. Building construction periods are estimated using a large-language-model approach that incorporates multi-source geospatial data, including historical land-use layers, remote sensing imagery, and urban development boundaries, and follows established methods from previous research25 to provide a basis for assigning appropriate construction periods. To determine the thermal properties of envelope materials with greater accuracy, the selection of material parameters follows a range of relevant building design standards, issued in 1980, 1995, 2005, and 2015, and also draws on relevant studies to complement and validate the assumptions26,27,28,29,30. Given the potential mismatch between the actual construction year and the corresponding regulatory framework, the thermal performance values are assigned by conservatively rounding down to the nearest earlier standard to ensure consistency, following a commonly used approach in UBEM studies to represent typical envelope performance over time31. This approach allows the model to reflect typical material performance while avoiding overestimation of energy savings. For parameters where official values are not available in specific years, including the building internal heat gains, occupancy rates and Heating, Ventilation and Air-conditioning (HVAC) operation schedules, the study adopts representative values from previous research conducted in the same geographic context, ensuring consistency with locally observed construction and retrofit practices30,32,33,34. Specific HVAC schedules are defined for each building category in this study. For industrial buildings, HVAC systems operate continuously throughout the year. In contrast, domestic and public buildings follow defined operational schedules during daytime hours. Office and administrative buildings operate from 07:00 to 18:00, transport buildings from 06:00 to 23:00, and other public buildings from 08:00 to 21:00. Domestic buildings operate HVAC systems from 19:00 to 08:00 on weekdays and throughout the day on non-working days. Cooling and heating setpoints are set at 25 °C and 16 °C respectively. For all building types except industrial, HVAC systems remain off during the transitional seasons, specifically from 1 March to 30 April and from 16 October to 30 November. Simulations are performed upon the local weather conditions in Shanghai. The detailed modelling parameters applied to each building category are summarised in Table 1. The years indicated in the table correspond to the publication dates of the respective building design standards.

Table 1 Detailed modelling input parameters for building stock simulation
Full size table

Establishing retrofit framework for heritage buildings at neighbourhood scale

This section outlines a structured framework that guides the retrofitting of historic buildings within urban neighbourhoods. The framework aims to reconcile the goals of energy conservation and heritage protection through a step-by-step process that supports evidence-based and context-sensitive decision-making, as shown in Fig. 2. It is composed of three main stages, including data preparation, categorisation and intervention assignment, and performance evaluation. The framework is intended to be adaptable and applicable across diverse historic contexts, providing a systematic way to integrate heritage considerations into UBEM analyses.

Fig. 2
Fig. 2
Full size image

Schematic diagram of heritage-sensitive energy retrofit framework for historic urban areas.

The first stage begins with the preparation of detailed datasets that reflect the geometric and operational characteristics of the existing building stock. These include footprints, building heights, construction periods, thermal properties, and typical HVAC schedules. Such data could be obtained through a combination of open-source platforms, remote sensing techniques, and standard design references, as exampled in Building data acquisition Section. By consolidating this information, the framework ensures that the model reflects both morphological accuracy and regulatory relevance. The quality and completeness of this foundational dataset directly influences the reliability of subsequent energy assessments and retrofit recommendations. Moreover, data standardisation processes ensure consistency across different building types and vintage periods, facilitating meaningful comparisons within the urban context. In a heritage context, the data preparation stage must also compile information on each building’s historical status, including age, and conservation reports. This enriches the energy model with a cultural dimension, establishing the foundation for heritage-informed analysis. The integration of regulatory constraints and protection levels at this stage enables the framework to anticipate potential conflicts between energy objectives and heritage requirements from the outset. Ensuring that the dataset reflects these specific attributes is essential to maintaining both the accuracy of energy simulations and the integrity of heritage-sensitive decision-making.

In the second stage, buildings are categorised according to their level of protection under heritage conservation policies. This step is essential, as it determines the extent to which each structure can be altered. The classification process distinguishes between buildings that must remain entirely unchanged, those eligible for partial intervention, and those with minimal or no restrictions. Once this regulatory typology is established, retrofit strategies are assigned accordingly. Measures range from minor upgrades, such as improving lighting efficiency, to more extensive interventions that address the building envelope and mechanical systems. Importantly, each retrofit scenario respects the limits imposed by conservation requirements, thereby maintaining the integrity of the built heritage. This categorisation formalises the link between qualitative heritage values and technical retrofit actions by embedding policy-defined protection levels directly within the modelling logic. In practice, the assigned level of protection reflects expert appraisal of individual buildings’ historical, architectural, and cultural significance, translating qualitative assessments from municipal heritage frameworks into structured intervention constraints within the simulation environment. Structures of outstanding historical or architectural merit are subject to stricter controls, while those of lesser or more recent significance are afforded greater flexibility. Many international heritage regimes adopt similar multi-tiered approaches. For example, in the United Kingdom, the statutory listing system classifies buildings into three grades: Grade I for buildings of exceptional interest, Grade II* for particularly important buildings of more than special interest, and Grade II for buildings of special interest35. These categories are designed to ensure that any modification is proportionate to the building’s cultural value and preserves its significance. Integrating retrofit strategies with such graded conservation typologies enables the operationalisation of a values-based conservation approach, which has become central in contemporary heritage discourse36. Under this paradigm, acceptable interventions are guided by the cultural, historical, and aesthetic importance attributed to each site, rather than by technical criteria alone. The proposed framework embeds this principle within its modelling logic. Highly protected buildings are constrained to non-invasive strategies that preserve authenticity, whereas structures of lower significance may receive more substantial upgrades, provided they remain visually and materially coherent with the historic context.

The third stage shifts focus to evaluation, introducing two metrics to quantify the performance of retrofit options. Energy use intensity (EUI), which expresses consumption per square metre of floor area, allows for cross-comparison among buildings of varying size and function. Total energy use (EU), by contrast, aggregates demand across the entire neighbourhood, providing a broader view of urban energy performance. To ensure reliability, model outputs are compared with figures reported in previous studies and official publications. This cross-referencing process strengthens the validity of the simulation results and builds confidence in subsequent scenario assessments. The energy performance of the existing building stock in its current state is treated as the baseline. Retrofit scenarios are then simulated and compared against this baseline, enabling a clear assessment of intervention effectiveness under different levels of heritage protection. Notably, evaluating results in light of the heritage categories allows stakeholders to see the trade-offs between efficiency improvements and conservation constraints. By examining EUI and total EU changes for each scenario, decision-makers can identify where energy gains are achieved with minimal heritage impact and where stringent protections might limit improvements. This evaluative stage effectively mirrors a heritage impact assessment logic applied to energy measures. Positive effects, such as energy savings, improved comfort, are weighed against any negative implications for heritage, including alterations to fabric or appearance. Through such an approach, this framework quantified the benefits of retrofit interventions while explicitly acknowledging the conservation boundaries, thereby facilitating a balanced interpretation of results.

The abovementioned indicators are defined through the following Eqs. (1) and (2), where the subscript \(i\) denotes individual buildings, \(j\) refers to building types within the stock, and \(h\) indicates each hourly time interval. The superscripts \({Cool}\), \({Heat}\) and \({Elec}\) correspond to energy associated with cooling demand, heating demand and general electricity use respectively. These notations allow the simulation to distinguish between different thermal and electrical energy demands. \({Floorage}\) represents the total floor area of each building. The subscript Build refers to the full set of buildings included in the analysis, and any summation over this index reflects aggregation of energy metrics across the entire stock being evaluated.

$${{EUI}}^{{Cool}/{Heat}/{Elec}}=\frac{\sum {{Load}}_{i,j,h}^{{Cool}/{Heat}/{Elec}}}{{{Floorage}}_{i}}$$
(1)
$${{EU}}^{{Cool}/{Heat}/{Elec}}={\sum }_{i,\,j}\sum {{Load}}_{h}^{{Cool}/{Heat}/{Elec}}$$
(2)

The Wukang Road area in Shanghai serves as an illustrative case for applying this framework. As a designated historic district, the area features a diverse array of building types and a clearly defined heritage zoning plan. The heritage protection levels applied in the Wukang Road district are defined by local regulatory guidelines, which differentiate the degree of permissible physical intervention based on cultural, architectural, and historical significance37. As illustrated in Fig. 3, protected buildings (red) are subject to strict conservation controls prohibiting alterations to both exterior and interior elements, thereby excluding any envelope upgrades or HVAC modifications. Retained historical buildings (orange) must preserve façade and spatial layout, but may allow selective internal updates such as lighting replacements. General historical buildings (green) permit more flexibility, including controlled insulation and system replacements, provided the external character is preserved. Unprotected buildings (blue) face no formal conservation constraints. These restrictions directly affect the thermal performance outcomes: buildings under higher protection are limited to lower-impact interventions, leading to more modest improvements in heating and cooling loads. In contrast, less restricted buildings benefit from comprehensive upgrades, resulting in greater reductions in energy use intensity.

Fig. 3
Fig. 3
Full size image

Schematic plan for the classification of historic buildings and corresponding 3D models in the Wukang Road area.

Retrofit strategies are customised to align with each category. No operations are carried out on the protected building. For retained historical buildings, only lighting systems are upgraded. In line with best practice, lighting loads are reduced to 5 watts per square38,39. General historic buildings receive both lighting and envelope upgrades, calibrated to comply with the most recent national efficiency standards shown in Table 1 without compromising visual consistency. For all other non-protected buildings, further improvements are implemented, including the replacement of HVAC systems with more efficient models. Specifically, for domestic buildings, cooling and heating systems are assigned coefficients of performance of 2.90 and 2.20 respectively, and other building types adopt a coefficient of 5.6029. These interventions are defined so as to respect each category’s limitations, for example, envelop upgrades on general historic buildings are designed to be internally applied or otherwise invisible externally, thus preserving the facades. In this way, each scenario adheres to the spirit of the conservation guidelines while exploring the maximum feasible energy improvements. Once defined, the interventions are simulated to assess changes in EUI and EU. The results not only quantify the benefits of each strategy but also illustrate the framework’s capacity to support decision-making in complex heritage contexts. Through this application, the Wukang Road case demonstrates how sensitive energy interventions can be effectively coordinated with long-term conservation goals, providing a template for other historic districts facing similar sustainable challenges.

Results

Following the establishment of the modelling framework that incorporates heritage classification and intervention strategies, this study conducts energy efficiency simulations based on the Wukang Road area, assessing the energy performance of various historic buildings under both current conditions and retrofit scenarios. This section not only presents quantitative results of energy consumption variations, but also reveals the tensions and balance between energy-saving potential and conservation constraints through comparative analysis with heritage protection levels, thereby further validating the adaptability and effectiveness of the proposed framework in practical applications.

Current energy use patterns in selected urban heritage area

This section presents the baseline energy performance of the historic building stock within the Wukang Road area. The analysis focuses on five primary building use categories in this area, namely administrative, commercial, industrial, office, and domestic. These functional categories not only reflect diverse energy demand profiles but also show distinct patterns of distribution across heritage protection levels. As shown in Table 2, commercial and industrial buildings make up the majority of protected and retained historical structures, suggesting their strong historical and architectural significance. Domestic buildings, though fewer, are evenly spread across all protection levels, while administrative and office buildings appear rarely and are mostly unprotected. This distribution indicates a clear link between building function and heritage designation, which in turn influences retrofit feasibility. By disaggregating EUI into cooling, heating, and electricity end uses, the study provides a nuanced understanding of functional energy behaviour across the heritage district. Figure 4 visualises the distribution of EUI values through violin plots, while Table 3 summarises average and total consumption metrics for each building type in the baseline scenario.

Fig. 4
Fig. 4
Full size image

Violin plot of baseline EUI by end use for case historic building stock.

Table 2 Distribution of heritage protection levels by building function in the Wukang road historic district
Full size table
Table 3 Energy performance statistics for baseline models
Full size table

The simulation results align closely with previous stock level modelling studies conducted in Shanghai29,40, thereby reinforcing the validity of the modelling approach and supporting its application to historic contexts. Cooling EUI remains substantial in non-domestic buildings, particularly in industrial buildings (50.75 kWh/m²) and commercial buildings (36.49 kWh/m²). These values reflect the typical characteristics of larger historic structures with extensive internal spaces and high equipment or occupancy loads. Administrative and office buildings show comparatively lower cooling EUI values, at 21.60 kWh/m² and 21.07 kWh/m² respectively, likely due to shorter operating hours or smaller floor areas. Domestic buildings present relatively high cooling intensity (37.31 kWh/m²), highlighting great cooling energy demands even in unmodified historic residences, and underlining the importance of targeted retrofit strategies. Heating demands show a distinct pattern, with non-domestic buildings reporting very low heating EUI, such as administrative at 0.29 kWh/m² and industrial at 0.12 kWh/m², consistent with Shanghai’s mild winter climate and limited traditional heating provision in these building types. In contrast, domestic buildings demonstrate higher heating EUI at 25.78 kWh/m², reflecting generally lower efficiency HVAC systems in residential buildings and clearly indicating substantial potential for retrofit improvements. Electricity consumption, primarily for lighting and equipment, is highest in industrial buildings which is 256.98 kWh/m², reflecting equipment-intensive operations. Commercial and administrative buildings exhibit moderate electricity EUI values (148.20 kWh/m² and 114.09 kWh/m²), while domestic buildings have the lowest electricity intensity (62.15 kWh/m²). In aggregate terms, commercial buildings dominate total cooling (33.97 GWh) and electricity (137.98 GWh) consumption, reflecting their considerable floor area and operational intensity. Conversely, domestic buildings significantly contribute to total heating consumption (2.16 GWh), despite lower individual cooling and electricity intensities. Overall, the findings clearly demonstrate the linkage between building functional categories and their energy performance characteristics. The outcomes further confirm that the proposed methodological framework effectively captures energy use patterns within heritage conservation constraints, providing a solid data-driven foundation for precisely tailored retrofit planning and implementation in historic urban contexts.

Crucially, these baseline findings demonstrate that heritage buildings exhibit diverse energy performance profiles depending on the type of use. The modelling outcomes therefore validate the framework’s capacity to capture differentiated energy needs within a culturally and architecturally heterogeneous urban fabric. By aligning with empirical patterns observed in previous heritage-related energy studies, the results lend credibility to the simulation approach and establish a robust reference point for subsequent retrofit scenario analysis. Notably, several of the performance patterns observed in this study correspond with findings from international research on historic building energy behaviour. Traditional buildings are frequently characterised by passive environmental strategies such as thermal inertia, natural ventilation, and solar shading, which allow for relatively efficient thermal performance when occupied and operated in accordance with their original design intent41. However, when these buildings are adapted to contemporary comfort expectations, such as the provision of continuous indoor heating during winter months, their energy performance often deteriorates significantly due to a lack of thermal insulation and outdated building systems42. The exceptionally low heating demand recorded in the non-domestic buildings of the Wukang Road area, in combination with high cooling and electrical loads, is consistent with the idea that historic public architecture was designed to optimise summer ventilation and shading, with minimal provision for winter conditioning. In contrast, the relatively high heating consumption observed in historic domestic buildings suggests that the introduction of modern thermal comfort standards, especially during colder periods, can lead to substantial increases in energy demand unless appropriate retrofitting measures are implemented. Collectively, these findings reinforce the necessity of implementing targeted and function-specific energy interventions. It is now essential to assess how such interventions, particularly when governed by varying degrees of heritage protection, may enhance building performance while maintaining the integrity of cultural values.

Scenario analysis of retrofit strategies under heritage constraints

As mentioned in the preceding Methods section, the heritage protection levels in Shanghai determine the scope of permitted retrofit interventions and are thus a key structuring variable in the scenario design. These categories directly shape the technical feasibility of energy interventions and, in particular, the potential for reducing heating and cooling loads, which depend heavily on mechanical and envelope upgrades. There is also a clear correlation between building function and protection status as shown in the Table 2. Therefore, this section assesses the impact of differentiated retrofit strategies across the historic building stock, as informed by typological function and heritage protection level. Figure 5 visualises the EUI distributions of retrofit models, and Table 4 presents the corresponding performance metrics of the retrofit models and their comparisons to the baseline scenario. Figure 6 shows building-level annual total energy savings for cooling, heating, and electricity from retrofit measures. Cooling savings are concentrated in the south-central area, heating savings vary across the site, and electricity savings are more prominent in areas with higher building density.

Fig. 5
Fig. 5
Full size image

Violin plot of retrofit model EUI by end use for case historic building stock.

Fig. 6
Fig. 6
Full size image

Spatial distribution of end-use energy savings from retrofit compared to baseline scenario.

Table 4 Energy performance statistics of retrofit models and their comparisons with baseline models
Full size table

Retrofit interventions lead to significant reductions in EUI across all building types. On average, cooling EUI decreases by approximately 31 percent, with the most substantial reductions occurring in industrial buildings, where cooling intensity drops by nearly 40 percent. These improvements result from upgraded HVAC systems and refined envelope conditions, particularly in building types with fewer regulatory restrictions. Administrative and office buildings also exhibit marked decreases, suggesting that even moderate upgrades can achieve measurable energy performance gains. Electricity EUI also decreases across all non-domestic categories. Industrial buildings show the largest reduction, at over 40 percent, followed by commercial and administrative types. These reductions reflect the impact of improved lighting and equipment efficiency. By contrast, domestic buildings exhibit no change in electricity intensity. This outcome is expected, as domestic buildings already have relatively low lighting and equipment loads, and the retrofits applied to this category do not alter system-level efficiencies. Heating EUI decreases across all building types. Industrial and domestic buildings show the largest proportional reductions, with heating intensity falling by more than 50 percent. These outcomes reflect both the efficiency improvements achieved through the replacement of underperforming systems and the reduced reliance on electricity for space heating. In domestic buildings in particular, the sharp decline in heating EUI underscores the relatively high baseline demand and the potential for substantial savings when HVAC performance is improved. These results confirm that targeted, function-specific interventions can reduce energy demand even under heritage constraints. The observed variations across building types highlight the importance of aligning retrofit strategies with both building function and conservation requirements, allowing for energy efficiency improvements that remain sensitive to the historical context.

To enhance interpretability, the protection status, which is shown in Fig. 3 and retrofit performance layers in Fig. 6 can be analysed in spatial conjunction. A comparative reading reveals that buildings with high protection levels, which are primarily concentrated in the central and western parts of the district, exhibit relatively limited retrofit gains, particularly in heating and envelope-driven interventions. This reflects the restricted range of physical modifications permitted under heritage regulations. Conversely, the most significant energy savings tend to occur in areas with lower levels of heritage constraint, such as general historic buildings or other buildings, where full-scale envelope and HVAC upgrades are feasible. These figures together suggest a spatial trade-off between conservation value and energy retrofitting potential. This insight is critical for formulating place-specific retrofit policies, while highly protected areas may benefit more from low-intrusion or passive measures, less constrained zones offer opportunities for more aggressive decarbonisation strategies.

The abovementioned findings underscore the value of building-level stock modelling in heritage contexts. By capturing the distinct energy behaviour of each structure while aggregating results across the district, the approach enables both detailed technical insight and broader strategic planning. The observed divergence between EUI and total energy use, as well as the varied heating responses across building types, illustrates how individual characteristics and heritage constraints shape retrofit outcomes in non-uniform ways. Stock-level analysis provides a means of identifying such patterns systematically for neighbourhood-level historical building, while still allowing for granular interpretation at the building scale. This dual capability is particularly important in historic districts, where policy decisions must balance area-wide energy goals with the specific architectural and regulatory conditions of each structure. As such, modelling existing stock on a per-building basis offers a robust foundation for formulating energy interventions that are both effective and conservation-sensitive. In the case of Wukang Road, the model results clearly identify the building types, and by implication their corresponding heritage zoning categories, that offer the greatest potential for energy efficiency improvements. This insight provides valuable guidance for conservation authorities and urban planners, enabling them to prioritise interventions where the impact is maximised. Moreover, the findings can inform a re-evaluation of specific regulatory constraints in situations where the projected energy benefits may warrant a carefully controlled relaxation of conservation rules. The scenario analysis thus confirms that substantial gains in energy performance are feasible within historic districts, provided that retrofit strategies are applied with due sensitivity to heritage values and regulatory frameworks. The following discussion places these results within the broader context of conservation theory and urban sustainability planning, building on contemporary scholarship that advocates for value-based, adaptive, and interdisciplinary approaches to heritage-led urban transformation.

Bridging quantitative modelling and qualitative heritage values

A core achievement of this study is addressing the conceptual divide between the quantitative methodologies of UBEM and qualitative frameworks of heritage conservation. Heritage conservation traditionally relies on qualitative evaluations, assessing authenticity, historical significance, and socio-cultural values. By systematically categorising retrofit strategies according to heritage protection levels and translating these into quantifiable energy scenarios, the research operationalises heritage value assessment within a quantitative modelling process. In practice, this bridging is accomplished through the direct translation of heritage protection levels into decision rules that govern permissible retrofit interventions within the UBEM workflow. Each conservation category is linked to a specific set of allowable measures, reflecting the regulatory limits placed on physical change according to the building’s attributed significance. This rule-based differentiation ensures that the energy model does not apply technically optimal solutions uniformly, but instead adjusts its recommendations in line with heritage value hierarchies derived from planning policy.

This integration responds to calls in recent literature for methodologies that reconcile the subjective evaluation inherent in heritage studies with more empirical and data-driven disciplines such as energy modelling. It demonstrates the feasibility of balancing tangible, quantifiable outcomes like energy savings against intangible heritage values, fostering interdisciplinary communication between conservationists, urban planners, and sustainability practitioners. While this framework primarily employs formal protection status as an indicator for qualitative values, it is acknowledged that statutory designations do not fully reflect the breadth and depth of heritage significance, particularly those values informally ascribed by communities, users, or other non-institutional stakeholders. Although this approach introduces a degree of simplification, it offers a structured and transparent means of embedding conservation constraints within quantitative modelling. The reliance on regulatory classifications reflects Shanghai’s operative heritage management system and enables alignment with enforceable planning mechanisms. At the same time, this foundation offers a clear pathway for future enhancement. The integration of participatory evaluation methods, such as value mapping or stakeholder consultation, could allow for the inclusion of community-defined and intangible heritage dimensions43. Such extensions would enhance the framework’s cultural sensitivity and support more inclusive and socially responsive approaches to sustainable heritage retrofit planning.

Furthermore, the stratified retrofit logic employed in this framework reflects a growing international consensus that conservation decisions should be calibrated according to the cultural significance of individual buildings. This aligns with values-based heritage management approaches found in charters such as the Burra Charter44, and with academic models that advocate for multi-dimensional heritage evaluation45,46. The approach also resonates with emerging discourse advocating for methodologies where measurable sustainability goals complement rather than compete with heritage preservation imperatives47. While it currently reflects the formal structure of Shanghai’s conservation system, its underlying logic is transferrable and compatible with international frameworks that emphasise minimal intervention, reversibility, and the preservation of authenticity. Crucially, the methodology positions UBEM not simply as a technical tool, but as a platform capable of accommodating diverse heritage priorities. As heritage science continues to advocate for integrated planning across environmental, cultural, and social domains, this study contributes to advancing methodological coherence between conservation principles and data-driven retrofit decision-making. It sets the foundation for future iterations of the framework to incorporate diverse heritage values, thus moving closer to a genuinely inclusive and context-sensitive model of sustainable heritage management.

Implications for heritage policy and sustainable urban governance

The study provides a clear pathway for incorporating quantitative energy analysis into heritage conservation policy and urban development strategies. By modelling retrofit scenarios under different heritage protection levels, the framework generates a robust evidence base to support more informed and transparent decision-making. This is particularly valuable in regulatory processes, where interventions in protected buildings are often evaluated through qualitative heritage impact assessments. The quantitative outputs of UBEM can complement these assessments by projecting energy and carbon performance under various intervention strategies, thus promoting context-sensitive energy retrofits that align with international heritage standards and climate commitments48.

The results demonstrate that tailoring retrofit interventions according to heritage zoning categories allows planners to balance cultural significance with environmental performance. Such alignment reflects contemporary best practices in conservation, which emphasise proportionality between the significance of heritage assets and the scale of permitted interventions. The analytical structure presented here helps to operationalise this principle, enabling the calibration of retrofit policies that respect the constraints of conservation while maximising achievable energy gains. Planning authorities could thus leverage the modelling outcomes to refine conservation area regulations. For example, the identification of retrofit measures that yield significant energy savings with minimal visual impacts could lead to tailored policy adaptations that balance heritage conservation with energy efficiency objectives. Furthermore, aligning heritage zoning with UBEM outputs supports broader urban sustainability goals, aligning local conservation practices with global frameworks such as UNESCO’s Historic Urban Landscape approach, which emphasises integrating heritage into urban sustainability agendas49. Spatially explicit energy performance analyses can be integrated into municipal planning tools, aiding planners in identifying and prioritising heritage retrofit projects within comprehensive urban climate action strategies.

Limitations of the study and directions for future research

This study’s chief limitations are its reliance on secondary data sources, the use of standardised retrofit strategies, the absence of empirical (post-retrofit) validation, and a limited treatment of social and cultural dimensions. The accuracy of UBEM simulations hinges on the granularity and precision of input data50. The current analysis depends on secondary geographic and historical datasets, inherently introducing uncertainty into simulations, particularly regarding older, less documented buildings. Moreover, while the model adopts several standardised intervention strategies, including HVAC upgrades, envelope insulation, and lighting improvements, for their compatibility with large-scale simulation and data availability, these are necessarily generic in nature. Recent research increasingly advocates for more specialised, context-sensitive retrofitting approaches that are particularly pertinent to heritage buildings. For instance, research has investigated the application of low-visual-impact solutions like internal wall insulation using advanced materials such as phenolic foam, mineral wool, and vacuum-insulated panels to achieve significant thermal improvements while preserving historic facades51. Other work has demonstrated the effectiveness of innovative internal finishing materials composed of diatomaceous earth and phase change materials for hygrothermal control, offering a trade-off between performance improvement and heritage conservation52. A holistic analysis of ground-coupled heat pumps as a renewable energy solution for historic buildings has been explored, demonstrating their potential for sustainable revitalisation53. These and other distributed, micro-retrofit interventions are highly relevant to heritage conservation but often fall outside the current scope of UBEM due to the scale of analysis and data resolution requirements. For example, the precise application of PV systems in a way that minimises visual impact, such as using tile-like panels54,55, requires a level of detail that is difficult to model at the urban district scale. While promising, these techniques often require high-resolution material data or bespoke modelling, which fall outside the operational scope of urban-scale UBEM platforms at the current stage, and it is also difficult to evaluate the post-retrofit performance of such interventions. Furthermore, structural and material-specific heritage conservation issues are not fully captured by high-level energy simulations, necessitating detailed building-level assessments for comprehensive decision-making56. Another limitation lies in the absence of direct local community input and behavioural factors in the current modelling framework. Effective conservation, especially within inhabited urban areas like Wukang Road, depends heavily on resident acceptance and active participation. Although the UBEM framework developed here is primarily technical, future applications could significantly benefit from integrating participatory methodologies to capture local cultural practices, user preferences, and community heritage values, thereby ensuring retrofits resonate with local identity and social dynamics57. This limitation potentially overlooks critical socio-cultural dimensions influencing retrofit acceptance and effectiveness. Thus, while the model provides strategic insights at the district level, it requires augmentation with micro-level socio-cultural analyses and empirical validation through community engagement.

The current approach, while offering strategic insights at the district scale, risks overlooking critical socio-cultural and heritage-specific dimensions that influence retrofit acceptance, effectiveness, and long-term sustainability. Future research could significantly enhance the robustness, applicability, and societal relevance of the framework through several avenues. First, longitudinal empirical studies validating model predictions via detailed monitoring of retrofit implementation could refine future simulations and improve predictive accuracy. Second, incorporating qualitative, socio-cultural research methodologies would enable the capture of resident behaviours, cultural perceptions, and context-specific heritage values, thereby transforming the UBEM into a socio-technical decision-support tool. Furthermore, future studies should prioritise the inclusion of heritage-specific criteria, such as historical significance, conservation status, material authenticity, and aesthetic sensitivity, within the modelling parameters and performance evaluation metrics. This would allow for a more nuanced understanding of the trade-offs between energy efficiency and heritage preservation. Enhanced interdisciplinary collaboration with fields such as heritage studies, conservation science, anthropology, sociology, urban informatics, and environmental psychology would deepen the analytical capacity to interpret heritage environments as complex, socially embedded systems58,59. Finally, extending the current methodology to incorporate advanced analytical tools, such as digital twins, artificial intelligence, and machine learning, could further enhance the predictive and adaptive capabilities of UBEM. Alongside this, future frameworks should also explore how emerging retrofit technologies can be integrated into urban-scale simulations. Doing so would help bridge the current gap between strategic energy modelling and conservation-led implementation, enabling more contextually appropriate and culturally sensitive heritage policy and governance at multiple scales.

Discussion

While the ambition to decarbonise urban environments grows increasingly urgent, historic districts face a unique tension, which is the need to reduce energy use without compromising cultural integrity. This study demonstrates that heritage buildings need not represent obstacles to energy transitions but can instead become active participants when retrofit interventions are informed by conservation principles and supported by rigorous simulation frameworks. By integrating protection status, typological diversity, and regulatory constraints within the UBEM methodology, this analysis transcends conventional one-size-fits-all approaches to retrofitting, establishing a pathway for nuanced, heritage-sensitive energy planning. Key findings are as follows.

  • Under heritage protection constraints, the framework at selected case study area achieved average reductions of 31.28% in cooling EUI, 53.15% in heating EUI, and 24.66% in electricity EUI, demonstrating that significant energy savings are still attainable within regulatory boundaries.

  • Across the targeted building stock, retrofitting under heritage constraints led to a 29.73% reduction in total cooling energy use, a 44.86% reduction in total heating energy use, and a 23.66% reduction in total electricity use, indicating that heritage districts can still deliver meaningful energy savings at scale.

  • The framework maintains regulatory fidelity by respecting intervention limits based on conservation zoning, while enabling scenario-based exploration of energy benefits within these limits.

  • It operationalises a heritage-informed energy modelling logic, linking building-level significance with appropriate retrofit scope, thereby aligning quantitative performance outcomes with qualitative cultural values.

  • As a decision-support tool, it allows urban planners and conservation authorities to simulate trade-offs spatially and prioritise interventions that maximise energy performance without compromising cultural integrity.

This research regards historic buildings as active contributors to sustainable urban transformation. The framework functions as both a methodological tool and a strategic platform, enabling heritage-sensitive retrofit planning within formal policy contexts. In Shanghai, where heritage protection operates within a tiered regulatory system, the model provides a means to calibrate conservation regulations based on energy objectives and building typologies. Simulation results indicate that general historical buildings, which are subject to moderate controls, present significant opportunities for low-visibility upgrades such as internal insulation or decentralised HVAC, all while preserving cultural significance. In highly protected districts, the model supports incremental, component-based interventions that align with façade retention and contribute to overall energy reduction targets. At the policy level, this framework is expected to offer institutional value as a decision-support tool for conservation planning, impact assessment, and funding prioritisation. It helps to underpin the development of retrofit standards and adaptive reuse guidelines that connect permissible actions with demonstrated performance benefits. The model also offers broader applicability for historic cities with complex heritage governance, supporting the development of digital planning tools that unite energy and heritage management. Future developments should aim to incorporate participatory value assessments and post-retrofit monitoring to enable more culturally resilient and performance-driven heritage retrofit strategies.