# Studying Buildings Outlines to Assess and Predict Energy Performance in Buildings: A Probabilistic Approach

**Authors:** Zohreh Shaghaghian, Fatemeh Shahsavari, Elham Delzendeh  
**arXiv:** [2111.08844](https://arxiv.org/abs/2111.08844)  
**Date:** 2021-11-17  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2111.08844v1/

## Key points

- The paper explores the use of predictive models to inform architects and designers about the impact of design decisions on energy consumption in buildings during early design stages.
- Building performance is typically calculated during the final phases of design, so incorporating predictive models earlier can provide valuable insights.
- The study focuses on evaluating the impact of a building's outline on energy consumption using synthetic data.
- Buildings with square shapes tend to be more energy-efficient compared to T, U, and L shapes of the same volume.
- Principal Component Analysis (PCA) is used to assess correlations between variables in the data distribution.
- Wall material specifications explain approximately 40% of data variation.
- Degree 2 polynomial regression models achieve high accuracy levels with over 98% R squared (coefficient of determination).
- Considering a building's outline is important in assessing and predicting its energy performance during early design stages.
- Synthetic data and probabilistic approaches, along with advanced modeling techniques like polynomial regression and PCA analysis, can help architects and designers make informed decisions regarding energy efficiency.

## Summary

The paper titled "Studying Buildings Outlines to Assess and Predict Energy Performance in Buildings: A Probabilistic Approach" explores the use of predictive models to inform architects and designers about the impact of their design decisions on energy consumption in buildings during early design stages. The authors highlight that building performance is typically calculated during the final phases of design when most specifications are fixed and unlikely to be modified based on design programs. Therefore, incorporating predictive models earlier in the design process can provide valuable insights. The study focuses on evaluating the impact of a building's outline on energy consumption using synthetic data. Four office outlines, including square, T, U, and L shapes, are selected for analysis. In addition to the shape parameter, other building features commonly used in literature such as Window to Wall Ratio (WWR), external wall material properties, glazing U value, windows' shading depth, and building orientation are considered. The results indicate that buildings with square shapes tend to be more energy-efficient compared to buildings with T, U, and L shapes of the same volume. Furthermore, T, U, and L shape samples exhibit similar behavior in terms of energy consumption. Principal Component Analysis (PCA) is applied to assess the correlations between variables in the data distribution. The analysis reveals that wall material specifications explain approximately 40% of data variation. To predict a building's energy consumption based on its outline, polynomial regression models with different degrees of complexity are employed. The findings demonstrate that degree 2 polynomial models achieve a high accuracy level with over 98% R squared (coefficient of determination). This suggests that these models can effectively predict new samples. In conclusion, this paper emphasizes the importance of considering a building's outline in assessing and predicting its energy performance during early design stages. By utilizing synthetic data and applying probabilistic approaches along with advanced modeling techniques like polynomial regression and PCA analysis , architects and designers can make informed decisions regarding energy efficiency . These findings contribute to the development of sustainable building practices and can guide future design processes .