Indoor Environmental Quality in Buildings - Methods for Assessment, Simulation, and Control of Daylight, Artificial Lighting, and Air Quality

Bearbeiter: Daniel Plörer

Betreuer: Rainer Pfluger

Kurzfassung

Complex fenestration systems govern daylight, solar gain and - where ventilation is provided through openings - air exchange at one and the same interface, and they do so through a state that is not fixed at design time but has to be chosen repeatedly during operation. That turns a material property into a control problem. This cumulative thesis frames four peer-reviewed publications and examines the chain on which such a choice depends: characterising the component, representing it in a model, deciding from the model, and verifying by measurement in the occupied building.

On characterisation, comparison of an ISO 15099 model against long-term measurements on a full-scale facade mock-up shows the equivalent-layer approach to reproduce a diffusely reflecting shading system more closely than a specular one, with accuracy degrading towards layers remote from the imposed boundary conditions and depending on the toolchain employed. On the optical side, angle-resolved data are indispensable and neither route to obtaining them is standardised.

On control, a simulation-based logic is shown to evaluate a luminance-based glare criterion at every time step of an annual simulation. What this demonstrates is a decision structure rather than its energetic performance. A requirement follows that reaches beyond the case at hand: a control logic acting on a simplified criterion cannot be evaluated by that criterion.

On measurement, single-point sensing of carbon dioxide proves robust for long-term assessment in the settings examined and not robust for transient assessment - a failure bounded to identifiable ventilation and occupancy events and to a time scale of tens of minutes.

The three strands were not designed to converge. That they nevertheless arrive at the same structural answer is the finding of the frame text itself: in each case the model or the instrument had performed and the deviation had been quantified, and what then stood between that result and a verdict on it was the absence of an agreed acceptance criterion. Criteria of this kind are normative; measurement can inform them but cannot settle them. Several of the most useful improvements available in this field are correspondingly not instrumental at all.