Information note

14/04-2003
File.: D76333933-3
Case: RE02255-3
ST/GRA/af

 

Introduction

Thatched roofs have been protected against fire hazards for many years in Denmark. Methods used have often been one of the three methods de-scribed in DIFT’s Pyrotechnical Information no. 29, “Brandsikring af straatage” (Fire-protection of Thatched Roofs) (BtI 29).

However, the last few years have seen an increasing number of cases of thatched roofs that have been plagued by seriously reduced durability, some only lasting 5 to 10 years. This has resulted in an increasing scepticism regarding the methods of fire-protection usually used, on the grounds that those thatched houses that were protected against fire hazards were then no longer properly ventilated to an extent that permitted the thorough drying out of the thatch itself. This suspicion has been thought to be the cause of the very rapid deteriation of the thatch.

Preliminary experiments, carried out by thatcher Erling Bach Petersen in Kolind, Denmark, where various mineral wool batts and other sheet materials have been replaced by glass matting as a protective covering against fire, have shown that this attempt apparently does work. These preliminary experiments have shown that the thatched materials only burn at reduced speed and without the usual content of airborne debris.

Use of glass matting as an alternative to the far more dense fire pro-tection methods described in BtI 29, indicates that the passage of air through the thatch is not affected to any great degree at the most usual wind speeds.

Aims

The aim of this experiment has been to show that fire protection of thatched roofs with glass matting results in the same advantages as with other types of fire-protective measures, as it would be advanta-geous to use the far more openly-woven glass matting instead of the presently accepted methods, thus ensuring adequate ventilation of the thatch itself.

Description

Fire Experiment no. 2 was carried out 20th March 2003 at Carlo F. Chris-tensen’s firm, Kastbjergvej 15, DK-8585 Glesborg, Denmark.

The same house was used for the experiment no. 2, as was used for ex-periment no. 1, namely a 4,5 x 5,0 m house, with a wooden bearing con-struction, and with a 13 mm plasterboard as a covering, see enclosed drawing (encl. 1). After completing work on rafters and laths, the roof was fire-protected as follows:

The fire protection was carried out by applying a layer of glass mat-ting sold as Sepatec 9679/M300, which is a combination matting with an area weight of 600 g per m2.. The glass matting was stretched taught, so that it was still immediately close to the underside of the thatch, after the thatching process was completed.

The matting was laid out in horizontal lengths along the roof, with overlaps being stapled along the length of a lath, overlap min. 100 mm. Along the length of the eaves and at the barges a layer of 50 mm Rock-wool A-batts 600 mm wide was laid out and fixed into place before thatching. Along the ridge, a similar layer, this time 700 mm wide, was laid out and fixed on top of the glass matting.

The roof was then thatched, using the Danish rod-method, i.e. the thatching materials were fastened by pressure between the laths and a 6 mm galvanised iron rod (replacing materials used historically, such as hazel rods), which was tightened down to the underlying laths with a stainless steel wire, fastened to the underlag by means of a screw. This method is shown in BtI 29 on page 8.

The thatching materials were well-stored, dry reeds that were rela-tively irregular and uneven in shape. The thatching was somewhat less loose and more solid than the first experiment, but still a little loose compared to standard commercial thatching quality. The experimen-tal roof was therefore also on this occasion a very easily combustible thatch, where a fire would under normal circumstances develop very quickly indeed.

The ridge was made of heather, which is the most usual material used in this area, and fastened by means of chicken netting.

Weather conditions for the experiment

On the day of the experiment, the temperature was approx. 4C, wind speed 4–5 m/sec, blowing predominantly at right angles in toward the roof surface that was set on fire. No precipitation.

Forsøgsbetingelser

På forsøgsdagen var temperaturen ca. 4°C, vindhastighed 4–5 m/s, fortrinsvis vinkelret ind på den tagside, som blev antændt. Ingen nedbør.

Method

The thatch was set on fire with 4 torches of straw that were planted in the eaves thatch, in a way corresponding to a deliberate act of arson or a rapid outbreak of fire from a neighbouring house.

After 38 minutes, the fire brigade from Grenå town began dousing the fire with the help of a couple of thatchers that cut the thatch loose, so it could be pulled down onto the ground.

The following observations were made while the roof was alight:

1. amount of airborne debris
2. heat radiation expressed as a measure of distance necessary between the roof and spectators
3. complete combustion, possible breakthrough into the loft room
4. fire burning in the loft room
5. percolation of water during fire-fighting.

Observations

When lit, the fire spread very rapidly to the whole of the surface of the roof where the eaves had been set on fire. For the first few min-utes, the fire was strong and burnt with high flames. Soon after this, however, the fire became a bed of embers, and stayed that way until the fire-fighting started.

Re 1) At no point was any airborne material observed that would have been able to start another fire.

Re 2) During the whole of the period while the experiment was carried out, it was possible to stand about 2.5m from the burning roof, without any inconvenience due to heat radiation.

Re 3) There was no breakthrough into the loft room at any time.

Re 4) The end of one lath was seen to be charred at the corner between the eaves and the barge.

Re 5) No water was seen to percolate through the glass matting during fire-fighting. The altered procedure of laying out the glass matting so that it was drawn around the eaves edge and out, had the effect that water did not collect here and run down the in-side wall, as was seen during the first test.

Supplementary observations

• After 15 minutes of fire, the glass matting was still quite cold and unaffected by the fire overall inside the house.
• There was a great deal of smoke in the loft room during the fire, which underlines the openness of the glass matting.
• After 35 minutes, a smoke diver examined the underside of the burn-ing roof, and found the temperature to be so high that one could only hold one’s hand against the surface for a few seconds at a time.
• It was possible to stand under the roof, and observe the glowing thatch through the glass matting. Subsequently it was found that the laths had not been adversely affected by fire at these places.

Conclusion

Fire protection in thatched roofs using glass matting of the kind known as Sepatec 9679/M300 fulfills the conditions laid out in BtI 29 as re-gards reduction of rapid outbreak of fire from neighbouring properties and radiation.

Fire protection with Sepatec 9679/M300 must be laid out with extra in-sulation consisting of approx. 50 mm Rockwool A-batts in a 600 mm wide belt along the eaves and barges, as well as 50 mm Rockwool A-batts approx. 750 mm wide at the ridge of the roof, so that the ridge of the roof is positioned above this extra insulation.

Final Comment

Fire protection in thatched roofs in accordance with the instructions in this note as well as in the enclosed instructions can be considered equivalent to the methods already mentioned in Brandteknisk Information nr. 29, “Brandsikring af straatage”, and will be accepted as a new method in future revisions of this information pamphlet.

 

Søren Tofte
Dansk Brandteknisk Institut