Algorithmic Planning of Ventilation Systems: Optimising for Life-Cycle Costs and Acoustic Comfort

TL;DR

The paper tackles the challenge of balancing life-cycle costs and acoustic comfort in building ventilation by formulating a holistic optimization that couples airflow and acoustics within a 2-stage stochastic MINLP. It introduces reduced-nonlinearity models for fans, fan stations, VFCs, and silencers and uses a 0D acoustic framework with octave-band level additions to enforce noise constraints. A key contribution is the integration of acoustics directly into the design phase, enabling simultaneous optimization of component purchases and per-load-case operation, and the development of a solving algorithm that yields a Pareto front of lifecycle costs versus noise limits. Case study results show that, under certain noise constraints, the holistic approach can reduce investment and overall life-cycle costs and transparently reveal trade-offs, while also highlighting that acoustics often do not dramatically increase energy costs. The work lays a foundation for scalable, acoustics-aware planning and suggests directions for applying the method to more diverse building types and more complex duct networks.

Abstract

The European Union's climate targets challenge the building sector to reduce energy use while ensuring comfort. Ventilation systems play an important role in achieving these goals. During system planning, the primary focus tends to lie on reducing life-cycle costs, including energy and investment expenses. Acoustic considerations which contribute significantly to occupant comfort, are either addressed as an afterthought or overlooked. This can result in suboptimal designs, where silencers are added indiscriminately without properly assessing their necessity. This paper introduces a novel method for optimising life-cycle costs through mathematical optimisation while adhering to predefined noise limits. We propose new model equations with reduce non-linearity better suited for integration into the optimisation framework. Further, they present a comprehensive approach to optimising ventilation systems under multiple load scenarios. Our method surpasses the traditional sequential approach by enabling simultaneous consideration of airflow and acoustics in a single, holistic optimisation step. A case study demonstrates the method's practical application, showing that optimal solutions can be computed efficiently. The results reveal that, with appropriate fan selection, many silencers can be eliminated. Additionally, the method supports decision-making by transparently illustrating the trade-offs between life-cycle costs and noise limits. Notably, while optimal solutions from the sequential and holistic approaches align for most noise limits, the holistic method achieves a 12 % reduction in costs under specific noise constraints. These results demonstrate the benefits of integrating airflow and acoustic design while underscoring the need for further application on more diverse building types and more complex ventilation system configurations.

Figures (17)

• Figure 1: procedure of the planning approach and organisation of the manuscript.
• Figure 2: example configuration as a graph representation of a simple ventilation system with different components on the edges.
• Figure 3: two different acoustic 0D model approaches
• Figure 4: generic 9-pole element. Figure adapted from Breuer et al. Breuer2024
• Figure 5: sources of noise in a room. (1) supply air system, (2) exhaust air system, (3) radiation noise of in-room ducts, (4) radiation noise of VFC/fans, (5) noise sources within the rooms, (6) noise transmission between rooms through walls and finally (7) cross-talk transmission between rooms