This article summarises the latest issue of Beyond Fast and Easy Answers, the personal LinkedIn newsletter by NEPSA Director Chresten Heide-Anderson. The newsletter explores how cognitive biases and decision-making affect complex technical and policy questions. In FIRE-04, rooftop PV fire safety is used as an illustration of a broader question: can we choose a fire-safe solution before we have defined the fire behaviour the complete system is expected to deliver?
When a new technology changes the way a building system behaves in a fire, comparing familiar materials and constructions may not be enough. The decision-making lesson is to define the outcome before allowing the available solutions to define the problem.
The installation of photovoltaic systems on flat roofs provides a useful example.
PV panels do not simply add another product to an otherwise unchanged roof. They change the geometry above the roof surface and can affect airflow, heat transfer and flame behaviour. This raises an important question for fire-safety assessment: does evidence developed for the original roof construction still represent the conditions under which the roof must perform after PV has been installed?
Research by Jens Steemann Kristensen, Benjamin Jacobs and Grunde Jomaas has demonstrated the importance of this changed fire scenario. Their experimental work examined PV installations above a B_ROOF(t4)-compliant flat-roof construction and varied factors including panel inclination, the gap beneath the panel and insulation material.
In the tested set-up, the authors found that introducing PV above a compliant roof construction could enable flame spread and concluded that the standards considered in their study were inadequate for the scenario investigated. They also emphasised that the critical gap heights observed in the experiments are not universal design values and that the limited-width mock-up does not represent an infinite flat roof.
Kristensen, Jacobs & Jomaas, Fire Technology, 2022
The UK review was commissioned by the Building Safety Regulator and prepared by OFR Consultants. Its findings and recommendations are those of the authors and do not necessarily represent government policy.
Better than does not necessarily mean good enough
When a new technology changes the behaviour of an existing system, it is natural to test the constructions already in use and compare their performance. That provides valuable evidence.
But comparative performance answers only one question. It tells us which solution performed better among those tested. It does not, by itself, establish whether the best-performing solution achieved an acceptable level of fire safety.
The best of the existing options may be the appropriate solution. It could also be only the least inadequate of the solutions examined.
To distinguish between those possibilities, the required fire behaviour has to be defined independently of the solutions being compared.
That means asking how far fire may spread beneath the array, how far heat or fire may penetrate into the roof construction, how continued fire development should be controlled and which consequences for occupants, firefighters, property and neighbouring activities need to be considered.
Classification remains important, but it answers a defined question
This is not an argument against classifications, standard tests or prescriptive solutions. They allow accumulated knowledge to be applied consistently and efficiently.
Material properties remain relevant. Where EPS is used, its combustible nature and any protection relied upon as part of the roof build-up should be considered openly as part of the complete system rather than treated as a separate component question.
A classification records performance according to defined test conditions and criteria. If a new technology substantially changes the conditions under which the system operates, it is reasonable to ask whether the existing evidence covers the new scenario or whether additional system-level evidence is required.
The same reasoning applies beyond PV. New technologies, new combinations of materials and changed building uses can all create situations in which established solutions face conditions different from those for which the original evidence was developed.
Start with the outcome
A useful distinction is between inputs, outputs and outcomes. A material specification or mitigation measure is an input. A classification or successful test can be an output. The behaviour of the complete construction in a fire is an outcome.
Confusion becomes possible when the measurable output is treated as though it automatically defines the desired outcome.
The question should not only be which available construction performs best. It should also be whether it performs well enough.
The full Beyond Fast and Easy Answers article explores the wider decision-making question, including Daniel Kahneman’s work on substitution and WYSIATI and Stephen Covey’s principle of beginning with the end in mind.
Read the original BFEA article.
