[
There is a growing trend to specify a combination of toughened glass and sprinkler protection in place of fire resistant glazing. But by doing so, you are not replacing like with like, warns Mike Wood.
Functional fire safety design is undoubtedly an important development for today’s complex commercial buildings. But this non-prescriptive approach is dependent on expert opinion and, therefore, has to be kept informed if risks are to be evaluated and balanced as much as possible on a sound technical basis. Nevertheless, the risk-based approach can lead to some challenging judgments.
One such question, for example, follows from the specification of standard ‘toughened’ glass (thermally toughened, or tempered, soda-lime-silica glass) together with sprinklers as a fire-resistant barrier. The approach is not an issue in itself. What is questionable is the presumption that the level of performance of the hybrid toughened glass/sprinkler combination is equivalent to that provided by specifically designed fire-resistant glass. This not only raises the question of how ‘level of performance’ and ‘equivalence’ are interpreted. It also focuses attention on what is meant by fitness for purpose and prompts consideration of whether equal weight of information is actually available to allow an adequate comparison of different systems, when their reactions and sensitivities to fire are entirely different. Solving the equation of value, risk and overall cost may not, in practice, turn out to be a straightforward one.
Impact resistant
What are the features of standard toughened glass that lead to the faith placed in its use with sprinklers as a fire-resistant barrier? Firstly, toughened glass is stronger than standard annealed (i.e. ‘window’) glass, typically by a factor of five. It is, therefore, more likely to resist the stresses generated by fire. But the level of those stresses in practice is uncertain, difficult to calculate and dependent on events in any case. Defining a single characteristic strength for glass is also not possible because of the atomic structure of glass and the weakening effect of surface flaws (both micro and macro). Toughened glass is no exception, and it is therefore normal to refer to strength as a range of likely values. Secondly, toughened glass is effectively a commodity product. It is available from numerous processors who are accustomed to routinely churn out thousands of square metres with only minimal pre-conditioning of the glass. As a result, commodity toughened glass is relatively – and seductively – cheap when set against fire-resistant glass.
So why should the designer pause and carefully consider the risks of using toughened glass as part of a fire-resistant barrier? Commodity toughened glass is produced for impact safety – it is not specifically designed and produced to be resistant against fire. Impact safety requires that if breakage should occur then it is fail safe, i.e. it breaks in a given area into a defined minimum number of small pieces without sharp edges, to reduce the possibility of injury. Paradoxically, the strength of toughened glass is its fundamental weakness under fire conditions: it is susceptible to unpredictable thermal shock failure and breaks catastrophically, forming a pile of pieces. Clearly, what is a decided advantage under human impact turns into a potentially fatal disadvantage under fire conditions, when a fire barrier may effectively disappear in an instance without warning.
Because of its nature, toughened glass needs special care and consideration if it is to be exposed to fire. Failure in fire is triggered by stress generated at the edge, by the tension between hot exposed glass and the cold edge shielded by the frame section. And the stress levels and components are very different in fire situations, compared with human impact. Factors which can weaken toughened glass under fire conditions are nowhere near as critically significant if impact safety is the requirement. But processors of the standard toughened product will be unaware of the fire requirement, and are not set up to do anything special or different from the norm. In particular, the edge working of standard toughened glass is basic (merely to remove the cut edge features) and may actually randomly introduce flaws which serve to heighten the risk of fire stress failure.
Probability of failure
The risk of the sudden and complete failure of toughened glass when exposed to fire is a possibility well recognised by the fire-resistant glass and testing industry. The difficulty of avoiding such failure is documented and test failures do not come as a surprise to those who have seen a number of tests. Modified toughened glass for fire resistance applications is a special product requiring special attention and processing. But even with this extra level of care and consideration, the probability of stress failure in a fire may not be entirely discounted, and a successful level of repeat performance is conditional upon certain framing requirements.
Determining the probability of failure is not straightforward: it requires a detailed technical programme using expertise in the fracture mechanics of glass. Such programmes show that several interacting factors are important and that widely different failure probability envelopes apply – depending on circumstances and the particular history of any individual glass plate. Factors such as edge condition, quality of finish, and the risk of damage during transport, handling and installation, are uncontrollable, which makes calculation of the individual probability of failure impossible.
It is very risky to assume that two pieces of superficially similar toughened glass will turn out to have the same failure probability characteristic in fire. Success in one case, such as in a controlled fire test, is no guarantee that success will apply in each and every case. This is even less so in the final installed condition, when the necessary specialist supervision is not always available. Probability of failure is only one aspect. Functional considerations of fitness for purpose, and a comparison of different systems, also need to evaluate the type of failure, their consequences, and the limits on the application that these characteristics may lead to.
All materials deteriorate in fire but the mode of failure, and its implications, may be dramatically different. Implications for the building and its occupants can also be dramatically different as a result. Toughened glass, for example, fails catastrophically under thermal fire stress; in this context, “catastrophically” means suddenly, unpredictably and without warning, resulting in complete disintegration of the fire barrier. It is especially vulnerable to uneven heating, local hot spots, or sudden cooling when hot. Failure may therefore be initiated by relatively small or local fires – for example at floor level – which may not even have the time or energy to activate the associated sprinklers.
Because of the risk of extreme thermal stress, water coverage from sprinkler systems has to be complete and without diversion by, for example, transoms, mullions, blinds, or other materials. For the effective operation of such systems, they need to be designed so that the fire activates the sprinklers before attacking the toughened glass. The sprinklers must also continue to operate effectively without interference during the whole time of the fire. Larger plate sizes resulting from no transoms or fewer mullions will also increase the vulnerability to unpredictable failure.
In comparison, fire-resistant glass, such as special intumescent laminated insulation or integrity glass, fails in a totally different manner from toughened glass. Failure is designed to be progressive, slow and gradual as the laminate erodes layer by layer – the whole laminated structure bonds together under fire conditions as a robust and resilient barrier. The functioning of the interlayer is reliable and repeatable. This means that deterioration is predictable and controlled, providing plenty of time for those in danger to react and escape in safety.
A fire-resistant glazed system based on intumescent fire-resistant glass is self contained and does not require additional associated systems. Maintenance is minimal and, once built into the structure, it can be relied on to function effectively as intended. Activation of the intumescent interlayer even signals the location of the fire.
Because of its controlled mode of deterioration, glass of this type can also maintain a level of reliable performance for hours – depending on the product formulation – to provide both a high level of property protection and an extended safe working environment for firefighters. Large scale, non-standard tests and real fires conclusively demonstrate this capability.
Apples and pears
But the level of performance has more than one dimension: it is not just a question of resilience for a rated time of exposure, as in a prescriptive furnace test. Fitness for purpose, when considered on functional design grounds,, should also take into account the likelihood of consistently delivering that performance under real fire conditions. The characteristic level of reliability therefore also needs to be considered in any comparison of different systems.
Performance repeatability and reproducibility need to be evaluated, in conjunction with a sensitivity analysis related to fire conditions. Performance time after time is an essential parameter, to be evaluated on realistic commercially available systems (and not just ideal laboratory samples in a controlled research environment). This should also be reinforced by non-standard, practical large scale tests and real fire cases.
Such evidence is all available for intumescent laminated fire-resistant glazed systems based on, for example, Pilkington Pyrostop (integrity and insulation) and Pilkington Pyrodur (integrity). There is an extensive track record over 30 years of successful testing and application, which now includes hundreds of successful third party independent tests of commercial framed systems. The grounds for confidence in the performance of such systems are therefore far better than can be the case for standard toughened glass with sprinkler combinations.
Such aspects are rarely evaluated for representative commercial toughened glass-sprinkler installations. The manufacturers of commodity toughened glass are not set up to do so, as the normal use of this glass does not demand fire testing. Justification for the hybrid toughened system, therefore, tends to be founded more on individual successful but nominal “type tests” and individual research papers, which risk being non-representative of commercial product installations.
In conclusion
Toughened glass/sprinkler hybrid systems should not be placed on the same level of performance as specifically designed fire-resistant glass types. If water flow is restricted, gives only partial cover, or does not start at all on the toughened hybrid system, its vulnerability under fire conditions is likely to dictate the course of events. Reliable and consistent resilience against fire cannot, therefore, be taken as a generic characteristic of plain toughened glass. Performance in practice of such standard toughened glass is subject to circumstance, probability and the individual condition of each plate.
Intumescent laminated fire-resistant glass, in marked contrast, has a reliable, predictable and repeatable function with a track record of extensive testing with successful outcomes. Differences in the probability of failure, the mechanism of failure, and the consequences of failure translate into different levels of risk. In these respects it offers a safer choice.
A much deeper and wider consideration is really necessary than one simply made on initial capital outlay and nominal fire-resistant performance times, based on individual prescriptive tests. The ultimate target should be reliable fitness for purpose in fire, given that the consequences of fire can be disastrous for people and property, and that there is always the potential for surprise.
Mike Wood is global consultant for fire protection at Pilkington Group Limited.
CLEAR ABOUT GLASS
The extensive use of glass in modern building design means that fire-resistant glazing can be a key element in any fire protection strategy. The following should therefore be strictly adhered to:
– System thinking is critical. A fire-resistant glass can only function as intended if it is installed as part of an approved fire-resistant glazing system, where all components are compatible. These include the glass, the glazing seal, the beads, the frame material, frame design, and the fixings for both the beads and the frame.
– Standard frames and glazing systems are not adequate against fire.
– Test evidence of performance is essential. The installed fire-resistant glazing system must be tested and approved as a complete system and test data must be relevant to the installation.
– Assessments are an expressed opinion and should only be carried out by appropriately accredited individuals or organisations, and should be based on relevant and applicable test evidence.
– Don’t mix and match between approved fire-resistant systems.
– No two types of glass are exactly the same. What is achieved with one fire-resistant glass may not be achievable with another.
– The installed fire-resistant glazing system must be the same as the system that was approved. It is essential that there are no last minute changes or short cuts. Above all, manufacturers’ instructions must be followed.
Important distinctions must be made between integrity only and integrity and insulation performance, as the differences in levels of protection are vary dramatically, and could be significant where both life safety and property protection against fire are concerned.