Experimental evaluation of a naturally ventilated PV double-skin building envelope in real operating conditions

Abstract

Building integrated photovoltaic systems are fast becoming a feature of urban landscapes in France and other countries tackling similar pressures to improve the energy footprint of residential and commercial building sectors. As active components of building envelopes, the technology represents a promising solution to the local electrical and thermal demand. However, despite significant recent interest and investment into the technology, few studies have been undertaken to study full-scale installations operating under real conditions.In this paper we present an experimental evaluation of a prototype naturally-ventilated photovoltaic double-skin facade, designed to maintain favourable operating conditions for electrical performance by utilising the stack effect to cool photovoltaic components whilst improving the thermal performance of the building to which it is attached. Developed to meet both technical and aesthetic specifications, the prototype comprises a two-storey, vertically pleated, tinted glazed facade with a heterogeneous arrangement of photovoltaic cells electrically grouped into a vertical stack of three arrays. Installed on the West-North-West facade of an occupied office building in Toulouse, France, the component was tailored to the specific requirements of building into which it was incorporated. The result is a complex system in terms of geometry and environment. The objectives of the study were to test the assumption that the behaviour of simplified double-skin components can be generalised to real multifunction systems, and to propose analysis techniques suitable for this context. The prototype was instrumented to monitor the minute-wise thermal, electrical and air flow behaviour during the first year of operation. We present an analysis approach adapted to both the complexity of the system and the size of the dataset, wherein data were first classed according to daily indicators of environmental conditions using aggregated database queries. Periods of normal and anomalous behaviour were distinguished using data visualisation techniques. As well as assessing the daily and seasonal variation in performance, the predictability of system was also tested as a function of environmental conditions by means of stationary empirical models for heat extraction and array temperature. Despite the inherent complexity and the limitations of approximate or incomplete instrumentation, the system was found to behave in a predictable manner and could be described by simple relationships between environmental conditions and system state.