Similarity assessment of using helium to predict smoke movement through buildings with double skin fa?ades during building integrated photovoltaics fires

In comparison to the building attached photovoltaics (BAPVs), Building Integrated Photovoltaics (BIPVs) drive more changes in building envelope. Therefore, tackling the fire safety of building integrated photovoltaic is more complicated due to these changes and the interaction between photovoltaic panels and the building. The photovoltaic fire hazards, including smoke generation of solar glazing, closely affect the fire safety of the building and occupants. This effect is further exemplified in the case of solar double skin facade systems where the stack effect influences smoke spread from ignited integrated photovoltaic panels via the vertical space of plenum in double skin facade. In scientific research field, the sub-scale model or laboratory-scale experiments is often applied to study the building fire smoke spread, which is much less costly, time and effort consuming, compared with full-scale test. However, the major concern of the sub-scale fire tests is providing the efficient but safe ignition source. There are methods such as cold smoke and hot gas (hot smoke test), each of which has their specific limitations on the cost and safety. This paper proposes an alternative method of helium smoke to substitute the real fire test for simulating a real integrated solar facade fire. First the fire dynamic simulator model with helium source is validated by experimental helium test along with particle image velocimetry in terms of helium concentration and flow velocity. The differences between the experiments and simulations are in the acceptable range (23 and 24%). Helium smoke result similarity is then compared to small scale real fire which showed close agreement according to the small coefficient of variation (between 1.6 and 11%).