Session 2: Urban microclimate; Building embodied energy; Building energy demand
Circular Design in Vertical Extensions: Reclaimed Steel in Structure
Zahra Ghenaat
Based on the Global Status of Buildings and Construction report, approximately 37% of global CO2 emissions are from materials that are used in building construction. On the other hand, with the Government’s requirements to achieve Net Zero emissions in the UK by 2050, rapid solutions are required to decrease carbon and CO2 emissions from the building sector and minimise its environmental impact. Moreover, 80% of today’s buildings are still expected to be in use by 2050, which shows retrofit strategies are more sustainable solutions than new construction.
This study explores the potential of circular design strategies with a focus on reusing reclaimed steel and aluminium in both structural and façade systems in vertical extensions by using the Crowndale Centre in London as a live case study. The research explores how parts of the existing structure can be retained or repurposed for a three-storey rooftop extension and uses environmental simulations to assess operational and embodied impacts.
Preliminary findings suggest that reclaimed materials can substantially reduce embodied emissions, but wider adoption requires overcoming technical and regulatory barriers. The study concludes with practical recommendations for integrating reclaimed elements into façades and structural systems, contributing to more circular and climate-positive design strategies.
Keywords: Circular Design, Retrofit, Vertical Extension, Embodied carbon.
– Theme: Building embodied energy –
Environmental Design Approaches for Maximising Outdoor Comfort Using Microclimate-Based Strategies in Hyderabad
Julia Brunning
This study presents microclimate responsive design strategies to improve outdoor comfort and community engagement utilising an existing building (Dr. Kallam Anji Reddy Memorial in Hyderabad, India) as a case study in a Composite Climate type climate.
The methodology utilised a diversity of analysis and strategy techniques such as EPW climate data, climate trends, bioclimatic charts, and thermal comfort models (PMV, PET, ASHRAE/IMAC) to assess outdoor human comfort, reveal daily extremes and suggest potential solutions to maximise comfort. Computational Fluid Dynamics (CFD) simulations of wind flow, together with radiation and shading analyses assess the performance of various design interventions.
Results show that rammed earth vertical communal gardens and the use of a butterfly-shaped ferrocement shade with integrated rainwater collection and cross-ventilation panels, provide effective evaporative cooling, reduce solar radiation, and enhance airflow as confirmed through the adjusted bioclimatic chart analysis. Additional benefits such as air pollution filtration, noise buffering, biodiversity enhancement, rainwater collection and support for community food growing were also presented.
This study demonstrate how integrating microclimate-informed design approach with innovative materials can create healthier, ecologically rich outdoor spaces with enhanced comfort and meaningful social and environmental benefits.
Keywords: Bioclimatic charts, comfort models, Environmental Simulation, Composite Climate, Passive Strategies.
– Theme: Urban microclimate –
Indoor Temperature Prediction for Residential Heat Pumps: A Physics-Informed Machine Learning Approach
Alexander Gledhill
Heating in buildings represents a major share of UK energy demand and associated greenhouse gas emissions, making the transition to decarbonised heating technology crucial. Air-source heat pumps offer a promising solution; however, successful integration depends on maintaining thermal comfort while enabling energy flexibility. This research builds upon a large field trial exploring third-party controlled flexibility in residential building clusters across Southern England. Expanding on findings that demonstrate the viability of demand side interventions, this study adopts a building-physics-informed framework to model the relationship between indoor ambient temperature and heat pump flow temperature. By coupling machine learning with the interpretability of physics-informed models, the approach aims to deliver indoor temperature predictions to varying building typologies and household characteristics. Preliminary analysis formulated a random forest machine learning model, trained on heating season data, including supply water temperature, outdoor air temperature, solar radiation and occupancy. At a 96-hour prediction horizon, the model achieved high accuracy (R² = 0.9788, RMSE = 0.1143). Ongoing work seeks to extend the model by incorporating ventilation heat loss and thermal capacitance, enhancing model interpretability.
Keywords: Air-Source Heat Pumps, Physics-Informed Machine Learning, Building Thermal Modelling, Energy Flexibility.
– Theme: Building energy demand (operational carbon vs. energy) –