Divyanshu Sood is a PhD student in UCD’s School of Mechanical and Materials Engineering and the UCD Energy Institute. His project, entitled ‘Development of next-generation residential building archetypes to evaluate energy consumption and indoor environmental quality at various spatial and temporal scales,’ is funded by the Sustainable Energy Authority of Ireland (SEAI). Divyanshu recently won the CIBSE Building Simulation Young Modellers Award 2023. We catch up with him below.
What are you working on at the moment?
I’m currently engaged in a project that involves developing residential building archetypes
representative of the Irish residential building stock. My main objective is to predict heating
energy consumption, thermal comfort, and CO2 concentrations in these buildings. This study
is crucial for identifying areas that are energy inefficient, which allows me to recommend
suitable retrofit options. These insights are important for amending policies and focusing on
specific areas to aid the sustainable transition of the building energy sector.
Once developed, I use these archetypes to create machine learning models which can
predict building performance under various retrofit options. The ultimate aim is to provide a
faster and more reliable optimization framework. This framework can help in quickly
identifying an optimized set of retrofit options for achieving maximum energy efficiency and
the highest level of thermal comfort in buildings. In this sense, my work is not just a
contribution to the field of sustainable building practices but also a step forward in the
integration of machine learning for energy conservation and building retrofit strategies.
How did you become interested in this research field?
My interest in this research field was sparked by a keen awareness of the challenges faced
in the building sector, particularly regarding energy efficiency, thermal comfort, and
environmental impact. The realization that residential buildings play a crucial role in energy
consumption and CO2 emissions drove me to explore ways to make them more sustainable.
The turning point was observing the gap between theoretical energy performance and actual
building performance. This discrepancy often results in inefficient energy use and reduced
comfort levels. I became intrigued by the potential of using advanced computational
methods, like machine learning and optimization algorithms, to bridge this gap. The idea of
developing a metamodel to predict building performance and optimize retrofit options
seemed like a promising solution to these widespread issues.
Can you share an interesting finding from your research?
Our study presents a ground-breaking approach in the field of sustainable building design,
utilizing the multi-objective optimization method NSGA-II. The use of the metamodel has
significantly enhanced the efficiency of this process, reducing the optimization time from
several hours to just 10-15 minutes, a notable improvement over traditional simulation-based
optimization methods. A key finding of our research is the metamodel's high prediction
accuracy, nearly 95%, which is critical for the success of the optimization process. A
comparative analysis reveals that using the metamodel-based method results in an 80%
reduction in completion time compared to the simulation-based approach, while maintaining
a high correlation coefficient of 0.98. This significant time efficiency, coupled with the
method’s accuracy, positions our approach as a highly effective tool for architects and
engineers seeking to design energy-efficient and comfortable residential buildings. Our
findings demonstrate the potential of integrating advanced computational methods like
NSGA-II in the building sector, paving the way for more sustainable, efficient, and
comfortable living spaces.
What is the wider relevance of your research to the energy transition?
–Societal benefits: my research contributes to enhancing the living conditions within
residential buildings. By optimizing for thermal comfort and reduced energy
consumption, I hope my research will lead to more comfortable and healthier living
spaces and help in reducing household energy expenses. This is particularly
significant in societies where energy costs are a substantial burden on families.
– Climate change mitigation: residential buildings are significant contributors to
global greenhouse gas emissions, primarily due to heating and cooling requirements.
By focusing on energy-efficient building designs and retrofitting options, my research
aids in reducing these emissions. Lower energy consumption translates to less
reliance on fossil fuels and a consequent reduction in CO2 emissions, thereby
contributing to global efforts to mitigate climate change.
– Policy impact: the findings and methodologies developed in my research should
provide valuable insights for policymakers. They can use this data to formulate more
effective building regulations and energy policies. The research highlights the areas
that require focus and investment, such as retrofitting older buildings or incorporating
energy-efficient designs in new constructions. Policies informed by such research
can lead to more sustainable urban development, aligning with national and
international goals for energy efficiency and environmental protection.
What’s something people may find surprising about you?
I have a keen interest in participating in hackathons. In addition, I can swim 50 meters in a
pool using the backstroke, but interestingly I’m not able to do so with the traditional freestyle