Optimising Hospital Building Performance: A Metamodel Approach to Balancing Thermal Comfort and Energy Use

Why It Matters

This research offers a flexible and time-saving approach to complex building design challenges, particularly in sensitive environments like hospitals. By decoupling simulation from optimisation, designers can rapidly assess multiple scenarios and user comfort criteria, leading to more informed decisions that enhance occupant well-being and operational efficiency.

Key Finding

The study found that a flexible optimisation method using metamodels can significantly reduce the time needed to find the best balance between keeping hospital occupants comfortable and minimising energy consumption, with different optimal solutions emerging based on specific locations, times, and comfort needs.

Key Findings

• The metamodel-based approach, particularly the OSMO strategy, offers substantial time savings compared to direct search optimisation.
• Variations in spatial location, time period, and thermal comfort criteria can lead to different optimum design conditions.
• Seasonal variations significantly influence optimum building performance.

Research Evidence

Aim: To develop and test a metamodel-based methodology for optimising building thermal and energy performance, specifically focusing on balancing thermal discomfort and energy use in hospital environments.

Method: Metamodel-based optimisation using Moving Least Squares Regression (MLSR) and Genetic Algorithms (GA).

Procedure: Initial building simulations were used to train MLSR metamodels. A genetic algorithm then optimised for minimal time-averaged thermal discomfort and energy use, presenting the optimum trade-off as a Pareto front. Adaptive coupling of dynamic thermal models (DTM) with computational fluid dynamics (CFD) was used for local thermal comfort evaluation. The 'one sample many optimisations' (OSMO) approach allowed for multiple optimisations from a single set of sample simulations.

Context: Hospital building design and optimisation.

Design Principle

Employ metamodelling techniques to accelerate the exploration of complex design spaces and identify optimal solutions for multi-objective problems, particularly in user-centric applications.

How to Apply

Use simulation software to generate an initial set of building performance data. Train a metamodel (e.g., using MLSR) on this data. Then, use an optimisation algorithm (e.g., a genetic algorithm) with the metamodel to explore numerous design scenarios for thermal comfort and energy use, adapting criteria as needed.

Limitations

The effectiveness of the metamodels is dependent on the quality and representativeness of the initial sample simulations. The selection of design variables and their ranges can influence the optimisation outcomes.

Student Guide (IB Design Technology)

Simple Explanation: This research shows how computer models can help designers find the best ways to make buildings comfortable for people and save energy at the same time, especially in places like hospitals, by testing many options quickly.

Why This Matters: Understanding how to balance user comfort with resource efficiency is crucial for creating sustainable and user-friendly designs. This research provides a method for tackling complex design problems efficiently.

Critical Thinking: How might the 'one sample many optimisations' approach be adapted for optimising the aesthetic qualities of a product, rather than its functional performance?

IA-Ready Paragraph: The development of metamodel-based optimisation techniques, as demonstrated in the study of hospital building performance, offers a powerful approach to efficiently exploring complex design spaces. By training predictive models on initial simulation data, designers can rapidly assess numerous design variations and identify optimal trade-offs between conflicting objectives, such as enhancing user thermal comfort while minimising energy consumption. This methodology allows for a flexible 'one sample many optimisations' strategy, significantly reducing the computational effort required compared to traditional direct search methods.

Project Tips

• When simulating building performance, ensure your initial sample data covers a wide and relevant range of design variables.
• Consider how you will define and measure 'thermal comfort' and 'energy use' for your specific design project.

How to Use in IA

• Reference this research when discussing the optimisation of design solutions, particularly for balancing multiple, potentially conflicting, design objectives like user comfort and energy efficiency.

Examiner Tips

• Demonstrate an understanding of how metamodels can accelerate design exploration and decision-making.
• Clearly articulate the trade-offs identified between different design objectives.

Independent Variable: ["Design variables (e.g., insulation levels, window size, HVAC settings)","Location, time period, thermal comfort criteria"]

Dependent Variable: ["Time-averaged thermal discomfort","Energy use"]

Controlled Variables: ["Building simulation software (ESP-r)","Metamodelling technique (MLSR)","Optimisation algorithm (GA)"]

Strengths

• Introduces a novel 'one sample many optimisations' (OSMO) approach for increased flexibility and efficiency.
• Integrates CFD with DTM for detailed local thermal comfort analysis.

Critical Questions

• What are the potential inaccuracies introduced by using a metamodel instead of direct simulation for optimisation?
• How sensitive are the optimisation results to the initial selection and range of design variables?

Extended Essay Application

• An Extended Essay could investigate the application of metamodelling to optimise the user experience of a digital interface, focusing on factors like navigation efficiency and visual appeal.

Source

Numerical Optimisation of Building Thermal and Energy Performance in Hospitals · White Rose eTheses Online (University of Leeds, The University of Sheffield, University of York) · 2017