View the PDF here
By ANDREAS SÆTER BØE
What would you guess the mixing temperature to be if you mix hot air (1,000°C) with cold air (20°C) in a 1:1 volume ratio? Many people would intuitively guess a temperature of about 500°C. However, this is far from true; the correct answer would be 230°C as warm air has much lower density than cold air. In fire safety, a temperature of 230°C compared to 500°C represents two completely different risks; the first case is less likely to cause a fire caused by radiation. This short example is the reason it may be safe to reduce fire insulation on ventilation ducts after a mixing point.
RISE Fire Research in Norway has conducted a series of full-scale laboratory tests to study the need and effect of fire-insulation of HVAC ventilation ducts. The results suggest that fire-safe and cost-efficient ventilation concepts can be developed without excessive use of fire insulation.
Background
All buildings of a certain size have a heating, ventilation, and air-conditioning (HVAC) system, and it must be installed in a way that does not contribute to fire and smoke spread between fire cells. There are three ways to install a ventilation system: the separate, the seal-up, or the extraction principle (Figure 1).
- The separate principle is based on each fire cell having a separate ventilation system. This is often used in small buildings such as row housing due to costs and practical concerns.
- With the seal-up-principle, fire dampers are installed in ventilation ducts wherever the ducts pass through a fire partition, and the ventilation fans are turned off during a fire. This physical barrier of fire dampers prevents smoke transport in the ducts, and insulation of the ducts is not needed. The seal-up-principle has been the preferred strategy in many countries; in the Nordic countries, the extraction principle is more common.
- When ventilation systems are installed in accordance with the extraction principle, ventilation fans are active during the fire, and smoke is transported through the ducts and out of the building. This causes warm duct surfaces, as the ducts will be heated by hot smoke transported through them. This problem is normally solved by fire insulating the ducts to prevent fire spread due to radiation or direct contact with combustible materials. After the introduction of EN-1366-1:2014, Fire resistance tests for service installation - Part 1,[1] the practice for ventilation in Norway has developed to include insulation of ventilation ducts in their full length. This is costly, and there is no documentation that verifies this is necessary.
Figure 1. A simplified sketch shows the main principle of three ventilation systems: separate-principle (left), seal-up-principle (middle), and extraction principle (right).
Effect of Mixing with Ambient Air Temperature
All buildings where the ventilation system passes through several fire cells have mixing points where air from the different fire cells is mixed. In case of a fire, hot smoke from the fire compartment (assuming the fire is located in one fire cell) will rapidly be cooled down by room temperature air from neighbouring fire cells (where there is no fire) at the mixing point and beyond.
Intuitively, one might think that mixing hot smoke (1,000°C) with room-temperature air (20°C) in a 1:1 volume ratio would yield a mixture with temperature about 500°C. However, this is not the case. The reason for this is that warm smoke has a much lower density than air at room temperature. Room-temperature air has a density of about 1.2 kg/m3, while 1,000 °C air has a density of about 0.27 kg/m3. A simple calculation then yields a mixing temperature of 230°C[2].
RISE Fire Research arranged a laboratory test where hot smoke (1,000°C) was mixed with room temperature air (20°C) and tests where no air was added (Figure 2).
Figure 2. Test Set-Up With and Without Mixing
Temperatures were measured both inside the duct (smoke-gas temperatures) and on the outside surface of the duct (duct-surface temperature). The effect of mixing hot gas from the furnace with 20°C ambient air temperature in a 1:1 volume ratio was examined. This scenario was conducted to simulate a case where hot smoke from a fire room was mixed with cold air from a neighbouring room that was not affected by the fire.
The results show that both gas temperature and duct-surface temperature were effectively reduced at the mixing point and beyond. The smoke-gas temperature was reduced to 229°C and 254°C for the uninsulated and the insulated duct, respectively.[3]
Critical Surface Temperature Due to Radiation
Ventilation ducts are covered with insulation to prevent a warm duct surface from igniting combustible materials in its vicinity, thereby spreading the fire out of its fire cell. As the radiation increases with temperature in the fourth power (~T4), it is the most critical heat transfer path from a hot ventilation duct to its surroundings.
The lowest critical heat flux to get ignition of a combustible material is about 10 kW/m2, although it will be higher for many materials. A rough estimate is that the duct surface must have a temperature of about 500°C to give a heat flux of 10 kW/m2 at 0.1 m (0.33 ft) from the duct. A surface temperature of 228°C, on the other hand, will give a heat flux of about 2 kW/m2.[2, 3]
Experimental Surface Radiation Test
A test was carried out to see whether samples of 3mm-thick (0.12 in) plywood samples would ignite due to radiation from an uninsulated hot duct surface during a 60-minute ISO-curve test. The samples were positioned 100 mm (3.94 in) below and 150 mm (5.91 in) to the side of the duct. No mixing was present in this test.
The result can be seen in Figure 3. None of the samples ignited during the test. However, on the samples at 100 mm (3.94 in) distance, signs of charring were observed within 0.45 m (1.48 ft) from the furnace, and a slight colour change occurred for the samples positioned between 0.45 m–1.15 m (1.48 ft–3.77 ft) from the furnace. The samples positioned at a greater distance than 1.15 m (3.77 ft) from the furnace were undamaged. Correspondingly, for the samples at 150 mm (5.92 in) distance from the duct, the samples positioned within 0.45 m (1.48 ft) from the furnace had a slight colour change, and the remaining samples were undamaged.[3]
Figure 3. Plywood samples 3 mm (0.12 in) thick were exposed to radiation from an uninsulated duct carrying hot smoke from a furnace with temperature profile following the ISO-curve in 60 minutes. The flat plywood samples were positioned at a distance of 100 mm (3.93 in) below the duct; the standing samples were positioned 150 mm (5.90 in) to the side of the duct. The first sample (No. 1) was located 250 mm (9.84 in) from the furnace.
The Effect of Insulation
To study the effect of the insulation, similar tests were carried out with and without insulation. The insulation used was 80 mm (3.15 in) thick and corresponded to EI 60 classification. For the uninsulated duct, the temperature decreased rapidly with increasing distance from the furnace. With a furnace temperature of 1,000˚C, the smoke-gas temperatures 5.5 m (18.0 ft) from the furnace were reduced to 567°C and 478°C for ducts with diameter 250 mm (9.84 in) and 125 mm (4.92 in), respectively. The duct surface temperatures at the corresponding locations were reduced to 328°C and 317°C for ducts with a diameter of 250 mm (9.84 in) and 125 mm (4.92 in), respectively.[3]
For the insulated duct, the smoke-gas temperatures barely decreased with increasing distance from the furnace, and the duct-surface temperatures (underneath the insulation) were approximately similar as the smoke temperatures. At the point where the insulation ended, smoke-gas temperatures rapidly decreased and demonstrated a similar response as the uninsulated case.
Deformation of Ducts
The uninsulated ducts did not experience any damage during the tests, only a change of colour. The insulated ducts were clearly deformed and damaged due to the high heat exposure during the test. The deformations were largest in the vicinity of the suspensions. Ducts supported by legs from beneath had more severe damage and even holes. The ducts suspended from above also had large deformations, but no visible holes ( Figure 4). The cross-section was reduced by 25% at the most due to the deformation.
Figure 4. The duct was clearly deformed around the supports. The image was taken after the insulation was removed after the test.
Implication of results on Need for Fire Insulation
As both the surface temperature and the smoke-gas temperature were effectively reduced after the mixing point to a temperature that may be defined as a non-critical temperature with regard to radiation, one might argue that the need for insulation after the mixing point is effectively reduced. In addition, these results suggest that the risk of fire spread through radiation from an uninsulated duct is mainly a problem before the first mixing point and is reduced further after every new mixing point as more room temperature air is mixed in.
The insulation effect of 80 mm (3.15 in) of insulation is so good that almost all the heat is stored inside the duct. The outer insulation surface will then have a temperature far below what gives a critical heat flux. However, two additional problems appear:
- The heat front is simply shifted to where the insulation stops, and the problem with a potentially hot duct surface must be dealt with after the insulation.
- The storage of heat inside the insulation may cause deformations of the duct.
We do not know the full consequences of the large deformations seen on the insulated ducts. One thing is clear: Such deformations occur only if the smoke gas is really hot (~800–1,000°C) and will likely not happen if an automatic extinguishing system activates.
Given that the deformation does take place, we do not know exactly what consequences this may cause. If the deformations do not cause any hole or leakage of the duct, the deformation would perhaps limit the flow through the ventilation duct, but not contribute directly to the spread of fire or smoke.
On the other hand, if the deformations cause leakages or a collapse of the duct, this may contribute more directly to the spread of fire and smoke and should be avoided.
A potential way of reducing this problem may be to reduce insulation thickness. This would give a more controlled heat reduction of the smoke gas in a way that keeps the surface temperature of the insulation below a critical temperature and also reduces the thermal stress on the duct inside the insulation.
Note: These tests were carried out in a large test hall; results may deviate to some extent if the duct is placed in a cavity.
References
[1] EN 1366-1:2014, Fire resistance tests for service installation - Part 1: Ventilation ducts 2015.
[2]Bøe, A. S., Stensaas, J. P. & Sesseng, C. BRAVENT - Teori- og kunnskapssammenstilling. RISE Fire Research, 2019.
[3]Bøe, A. S., Hox, K. & Sesseng, C. BRAVENT - Varmespredning i ventilasjonskanaler. RISE Fire Research, 2019.
Andreas Sæter Bøe is with RISE Fire Research AS, Norway.