Fire Resistance of Historic Fabric
Peter Jackman
A room with a fairly restricted fire load and hence moderate fire growth potential |
The implementation of fire safety regulations has taken its toll on historic building fabric over the last half-century as enforcers have tried to impose the prescriptive recommendations given in the guidance published in support of national regulations such as Approved Document B (Fire Safety). These recommendations are frequently imposed without due consideration of the need for them.
Elements of construction have to be ‘certificated’ as being capable of providing fire resistance for the pre-determined periods stated in prescriptive guidance documents. These periods are generally stated as 20, 30, 60 or 120 minutes when evaluated against the fire exposure conditions given in the appropriate standards, normally BS 476: Part 20: 1987 or its predecessors. The standards are based on fixed pass and fail criteria relating to structural load-bearing capacity, integrity (flaming) and temperature rise on the protected side (insulation) of the element.
The objectives behind these predetermined durations are never questioned, nor do we challenge the assessment criteria used, such as the fire severity replicated in the furnace, or the failure criteria used to establish integrity (the resistance to flames passing through the element) and insulation. In truth, these questions do need to be asked. Unlike new construction, where materials and fittings are supplied with no knowledge of the conditions in which they will ultimately be used, we can define exactly what we want a particular element of a particular building to achieve in the context of reducing the impact of fire. Once this has been defined, we are in a position to ascertain whether the element in question is capable of fulfilling that role either with or without the application of upgrading measures.
While ultimately this may require the assistance of a professional, initially an analysis can be made by the interested parties following the principles outlined here, which will lead to an informed approach to the specification of works.
The first myth that must be addressed is the time element. Practitioners are often shocked to learn that fire resistance (or FR) minutes as referred to in the guidance, have no correlation with real time minutes. The actual correlation between FR minutes and real minutes is not only an unknown, it varies building by building, structure by structure. This statement is confirmed in ISO/PD TR 834-3, the International Standard Organisation’s fire resistance test method commentary. ISO 834 forms the basis of BS 476: Part 20, which explicitly identifies this lack of correlation between FR and real time minutes. In a situation with a plentiful supply of combustible material and a good draught an element with a 30-minute fire resistance might provide just 10 minutes fire separation, while in another situation it might provide 45 minutes. Nothing worries fire engineers more than seeing less experienced designers halving the fire resistance period on the grounds that ‘everybody will escape in under five minutes’. This is acceptable if they are halving an application where the 30 minutes FR provides a real 45 minutes, but where the actual duration is only 15 minutes, this could reduce the protection afforded to 7½ minutes.
What features of the building determine whether 30 FR minutes result in an under or over-provision in practice? There is no easy answer to this, but a major influence is the likely severity of the ‘real’ fire and the way the fabric responds to these conditions. A new fire standard, BS 9999, gives a clue as to how this is determined. This standard generates fire safety measures that relate to an alpha-numeric risk profile which is the product of the building’s use and contents. A major component in this risk profile is the severity/rate of fire growth, which has four categories: slow, medium, fast and ultra-fast.
A slow-growing fire will take much longer to damage the fabric, or generate untenable conditions on one or both sides of a separating element, than a fire that is growing ultra-fast. The furnace used to measure fire resistance displays a fixed fire growth rate, probably ‘fast’ for the first 10-15 minutes and ‘medium’ thereafter. A construction which achieves 30 FR minutes against the furnace curve will obviously last longer when subjected to a slow growing fire and vice versa.
A corridor lined with drapes with a low integrity tolerance |
What factors control fire growth? The fire load is important; a fire in an art gallery generally has significantly less fuel than one in a library. However, this simple analysis is also dependent upon the amount of air available for combustion. An unfenestrated basement full of books would probably have a slower growth rate than a modestly furnished large dining hall, despite the disparity in the fire load. The fabric of the construction will then influence the fire conditions. Unplastered stone walls will produce a lower rate of heating than a hollow lath and plaster wall. Clearly, this analysis is not ‘rocket science’, and a consultant with a certain empathy for the likely fire dynamics will be able to predict the appropriate rate of fire growth and likely applicability of the furnace-generated fire conditions with some confidence.
What then is the validity of the standard failure criteria, particularly integrity (passage of flame)? The obvious failure criterion is the presence of flaming on the unexposed face. The presence of any flame on the protected face is irrefutable evidence of failure against this criterion, but what is the real hazard? In a tapestry-hung drawing room the hazard is very real but in an unfurnished plastered corridor, is this hazard significant before flames reach a certain size? The test operated in the UK for over 20 years utilises an oven dry cotton pad placed in the vicinity of any through-gap, even when no flames are being generated, which indicates a loss of ‘integrity’ when it starts to smoulder. Even a tapestry-hung room could tolerate that level of leakage for some time and a stone-walled armoury could withstand it almost indefinitely. Does it always constitute failure?
The failure criteria for insulation are deemed to have been exceeded when the average temperature of the unexposed face exceeds around 160°C (a rise of 140°C). What hazard does such a temperature represent in terms of life safety and fire spread?
In respect of life-safety, touching a wall at 160°C would result in serious burns: this is far from being a safe temperature. However, in respect of fire spread, this temperature is not going to ignite most combustibles in the adjacent space, even in our proposed tapestry-hung room. Ignition is only likely if temperatures of around 400°C are reached on the protected face and the combustibles are in close contact. Therefore, if the standard criterion of 160°C is used, the safety margin is very large.
Having recognised how the exposure conditions will vary, and how appropriate the test failure criteria are to the hazard, the adequacy of the existing construction can then be established in the context of the objective. There is a significant difference in the acceptability of the unexposed face conditions depending upon whether the barrier is to provide life safety, to restrict the degree of damage to the structure, or to preserve as many of the artefacts as possible. The design of the fire barrier will vary significantly against these three objectives.
If the objective is solely life safety, (which, as the supporting guidance indicates, is the sole concern of the Building Regulations), it will be necessary to provide a short period that is primarily smoke-free and not too hot, after which the fire can be allowed to spread as the structure and the artefacts will be seen to be sacrificial; the unwritten result of adopting national Building Regulations. In order to prevent the main structure from being ravaged by fire, we would not only wish to keep the occupants alive, we also try to contain the fire for as long as possible, but not being overly-concerned about allowing more smoke/steam or elevated temperatures to develop. When the artefacts have historical or monetary value, then our protection levels need to be higher, controlling smoke and temperature rise for much longer and so allowing snatch teams to do their job safely.
Once an understanding of the fire growth, the objectives and the relevant criteria has been achieved, it becomes possible to establish the adequacy of the existing construction by bringing in professional help. For many applications, little physical upgrading may be necessary: it may simply be a case of making good the existing fabric as required. If inadequate, the upgrading measures can be minimised by focussing solely on any shortfall in performance.
LATH AND PLASTER CEILINGS
Lath and plaster ceilings have a major role in preventing fire spread. They are critical to the protection of horizontal elements such as timber joisted floors, including the flooring on top, which in terms of fire performance is often in a poor condition due to the presence of gaps. Following failure of the ceiling, fire may spread either as a result of integrity losses through gaps in floorboards or as a result of collapse due to the joists burning away. The longer the ceiling remains in place, the greater will be the fire separation between floor levels.
Published data on the fire resistance of lath and plaster ceilings is limited. Not surprisingly, it has turned out to be impractical to remove a fully aged, installed and ‘abused’ ceiling to a fire test laboratory for testing without causing unknown degrees of damage. Ceilings built in place in test laboratories are rarely adequately aged, although age is known to have a major influence on the fire resistance of lime plasters. This is because lime cannot achieve its maximum strength and hence full fire potential until it has fully carbonated, and this can take many months.
The testing that has been undertaken as part of post-war building studies on lath and plaster ceilings has revealed wide variations in performance when tested under the standardised furnace conditions, with some exhibiting collapse quite quickly. The poor performance can, it is thought, be due to the very rapid heating rate (thermal shock) and the high temperatures reached in a standard furnace test. The furnace test is artificial because it ignores any preheating which would be experienced in the period before ‘flashover’, the point of sudden escalation. Furthermore, in a less well furnished room the rate of heating will be slower, with restricted fenestration slowing it further. When the ceiling is heated at this much slower rate, the ceiling may be adequate for some applications without any need to upgrade the ceiling at all. Early testing of an American intumescent coating system is already indicating that the rate of plaster heating can be reduced by a brush/ roller applied coating, significantly increasing the duration of its integrity.
Whether the extended protection is needed will depend upon the objectives set for the protected space. An expert analysis of any floor/ceiling assembly, measured against the set objectives may conclude that, once any repairs required have been carried out, it is:
- adequate without treatment
- adequate with, for example, a brush/roller applied coating to the ceiling
- not adequate without additional linings or floor cavity insulation being incorporated, although reducing the fire load and/or reducing the ventilation may permit a positive re-analysis without such measures.
FIRE DOORS
A similar approach can be applied to fire doors. The performance of a door that is required to provide fire separation will also be a product of the fire severity and the rate of heating, which are both related to the quantity of combustible material (the fire load) and its vulnerability. Whether the door is adequate for the required purpose will depend on whether the objective is life-safety only, property protection or protection of the contents.
Typical door assembly that could easily be upgraded to provide adequate fire resistance in a low to medium rate of fire growth application |
To achieve life-safety, one of the most important measures will be the fitting of a non-invasive smoke seal and a face-fixed high quality intumescent seal.
When the objective is extended to containing the fire to protect the structure and/or the artefacts, the door must prevent burn-through for more prolonged periods. In a panelled door, this will invariably mean that the panel must be upgraded but, above all, the method of retaining the panel must be improved. This will probably require a contribution from a fire-door expert. Painting the whole of the door with intumescent paint is highly questionable, as the main framework of the door will rarely burn through. Upgrading the panel is most important, but any treatment of the panel should utilise proven applied membranes such as thin sheets of intumescent mill-board and not just rely on a coat of intumescent paint or a layer of intumescent paper. Coatings may, however, slow down the rate of heating.
Fire doors almost exclusively fail as a result of the leaf distorting out of the frame, rather than by burning-through; the result of asymmetric heating. Again, this distortion can also be the product of rapid heating, so if the potential fire development is slowed down by reducing the contents and controlling the air supply, then the door will remain adequate for longer without the need for major work on the leaf. While this process may sound complicated, it is possible for a specialist fire consultant with experience in dealing with historic fabric to apply an expert system that utilises these principles and generates the minimum number of upgrading measures.
MINIMUM INTERVENTION, MAXIMUM EFFECT
These examples demonstrate that, with a review of the fire resistance objectives and an analysis of the environment in which the building exists, it is possible to dispense with a great deal of unwanted ‘surgery’ without compromising the objectives. These principles can be used in relation to any element of a building, allowing a scientific common sense approach to the upgrading process.
Ultimately, as with modern hospital surgery, the process of upgrading our historic buildings for fire safety should never be unnecessarily invasive. Rather, it should be a case of making the minimum of intervention for the maximum effect.