Planning of distributed ventilation systems for energy-efficient buildings by discrete optimisation

The EU's climate targets are increasingly affecting the building sector. Energy-efficient buildings should therefore be built as airtight as possible in order to meet these targets - ventilation systems are necessary to ensure a comfortable indoor climate. In the planning of ventilation systems, the placement, wiring and operation of the fans, among other things, must be considered. If, instead of the conventional planning for the maximum load case, the partial load scenarios are included, oversizing is reduced and energy efficiency is increased. In addition, there are new ventilation approaches that include distributed components in the central duct network and thus offer further opportunities to increase the energy efficiency of the systems. If one now considers multiple load cases and allows distributed components in the design, the number of combinations exceeds any human manageable amount and the human made design decision becomes far from optimal as a result. Therefore, in this paper a method is presented that uses mathematical optimisation techniques to control the complexity and support the design. The planning task is modeled techno-economically as a minimisation problem with respect to life-cycle costs (Mixed-Integer Nonlinear Program) and solved with discrete optimisation methods. The approach is then tested on a case study, with which savings of 22% in life-cycle costs are achieved while reducing energy consumption by 28%. Furthermore, the embedding in the real planning process is considered and shown that the division into several planning phases has a negative impact on the efficiency of the ventilation system. These results show that increasing the complexity of the planning task and modelling and solving it using discrete optimisation methods allows for a huge increase in the efficiency of the ventilation system. •Mastering the complexity of designing ventilation systems to increase efficiency.•Combining distributed ventilation systems with discrete optimisation methods.•Implementation of a priori estimation of the number of distributed fans.•Reducing life cycle costs by up to 22% and energy demand by 27.9%.•Demonstration of the dependencies in the real planning process.