Efficient and Quiet: Optimisation of Ventilation Systems by Coupling Airflow with Acoustics in a Multipole Approach

Ventilation systems are used to ensure a good indoor climate even in increasingly airtight buildings. It is predicted that ventilation systems will be installed in 90% of all new buildings in Germany by 2035 in order to achieve the EU’s climate targets. To design energy-efficient ventilation systems, topology decisions have to be made about the number, size, location and interconnection of fans and volume flow controllers, and operation under different load scenarios has to be anticipated. This leads to enormous degrees of freedom and methods from discrete optimisation become helpful. The resulting optimisation problem is a 2-stage stochastic MINLP. Preliminary work has confirmed the enormous potential of the approach to reduce life-cycle costs by up to 23%. However, so far consideration of acoustic phenomena has been lacking. The noise from a ventilation system must not exceed certain sound pressure levels in rooms. To integrate the acoustics into the optimisation, this paper models sound propagation from the duct network to the rooms as a network flow problem. Then, airflow and acoustics are coupled in a multipole approach. To overcome the strong nonlinearities and to reduce the model size, the model’s structure is exploited. The addition of acoustics is an important step that takes mathematically supported designing of ventilation systems to a new level.