We are only now beginning to fully understand exactly how water mist systems work. Here, Glenn Horton and David Yirrell draw on new fire tests to examine how water mist may be used as an alternative to prescriptive compartmentation in apartments.
The design of apartments in england and wales is subject to control under Building Regulations. While the regulations themselves are functional, there is guidance issued in support of them in the form of Approved Document B. The guidance makes a number of assumptions about the potential for a fire to occur in an apartment block and states that it is likely that any fire will originate in the apartment itself.
It considers escape from apartment buildings in two stages: inside the apartment and then in the common areas leading to a place of ultimate safety outside. The focus of this article is on the safety of occupants inside their own dwelling, and the potential benefits that a water mist system could provide in terms of freedom of design over the approaches suggested in Approved Document B.
Currently, the guidance sets out several approaches, depending on the size and layout of the apartment. These can be summarised as follows:
– Apartments can be restricted in size with a maximum travel distance of 9m, with the cooking facilities remote from the entrance to the apartment.
– Apartments can be unrestricted in size but provided with an internal protected hallway where the travel distance is no more than 9m.
– Where duplex units are proposed, as well as control over the protection of the internal staircore, exits from both levels is normally recommended.
While these provisions concentrate on the physical measures necessary to provide a reasonable means of escape for the occupants, provision is also made for early warning in the event of fire.
Currently the typical approaches used for the design of apartments rely on the provision of physical separation, combined with early warning. Although some benefits may be gained by increasing levels of fire detection, by and large open plan apartments present designers with difficulties which are costly to overcome and arguably are not the ideal solution in architectural terms.
From looking at work carried out by BRE1, it would appear that if active fire suppression systems are installed, there may be significant benefits to the occupants in terms of allowing them to escape under a wider range of conditions than would otherwise be the case. It is worth noting that the current guidance in Approved Document B does not consider the use of active systems within dwellings in high rise buildings, when it comes to varying from the designs set out in the guidance document.
This research, and the work of others, has indicated that the use of active systems can enhance the survivability of the occupants in the compartment of origin, to such an extent that it would seem that the use of high pressure water mist systems could be considered as part of a package of measures to compensate for variations from the current guidance, while still achieving the functional requirements of Part B1 to Schedule 1 of Building Regulations 2000.
It is worth pointing out that a holistic approach should be taken of the life safety of occupants. The use of high pressure water mist systems alone may not be considered as the ‘whole solution’, but rather as part of a package of measures which should include automatic fire detection and consideration of the physical measures incorporated.
Exploding a myth
One of the primary concerns with the use of active systems is maintenance and reliability – crucial to the means of escape. Although this is an issue, it is no more so than with other elements, such as the automatic fire alarm system, fire doors and physical fire separation. Arguably a water mist system – unlike a standard domestic smoke detector or self closer on the hallway doors – is less likely to be disabled or removed by an occupant due to nuisance activations. So on balance, the installation of a water mist system could increase the level of fire safety for the occupants, while allowing for greater freedom of design in terms of open plan layouts.
But there has been little published documentation in terms of type testing of water mist for these types of Class A fire applications. Most of the theory and practical discussion for water mist has concentrated on Class B, diesel spray and heptane fires, largely based on the International Maritime Organisation test circulars MSC/Circ 668/728 and MSC/Circ 913.
Sprinklers have been traditionally used in the protection of assets and in life safety, particularly in hotels and, more recently, in residential premises. A recent British Standard, BS9251:20052, provides guidance on the use of sprinklers in these premises.
Sprinklers deliver large droplets that are typically 1,000 microns (1mm) across. Water mist, on the other hand, uses water in fine droplets, ranging from 300 microns to just 20 microns in diameter. The droplet size is dependent on the type of water mist nozzle and the pressure that is behind the system.
Small droplets of water present a large effective area for heat absorption, and a larger density per cubic metre can be contained within any given volume. This cloud of droplets is often referred to as ‘mist’ or ‘fog’. The mist generated tends to fill an entire volume, as opposed to a sprinkler that produces a defined plume. The momentum and lightness of the mist causes the water to be drawn into fires. This process – known as entrainment – is an effective means for the mist to reach fires obscured by the line of sight. The mist readily turns to steam on hot fires, and further cooling is achieved through the latent heat of vaporisation. The mist/steam around the flame front removes oxygen and prevents heat attenuation.
Test conditions
In our tests, we used a high-pressure system, where 100 bars is delivered to the nozzle and the pressure is dissipated within the nozzle during mist formation. Average droplet size, determined through particle size measurement, is 50 microns.
The practical advantage of using water mist over sprinklers is based on the fact that the water quantity used is nearer 6 litres per minute, i.e. 1/10th that of sprinklers. Flow rates – and as a consequence pipe sizes – are therefore reduced and water damage is minimised.
The tests were undertaken in a former show home. The aim of the tests was to measure temperature, carbon monoxide, NOx and oxygen. Temperature thermal couple probes were installed in two locations – in the kitchen and in the lounge/living area – and at 3 heights: ceiling height; head height (1.75m); and bed head height (1m). A total of six temperature probes were used. Unfortunately, the supplied probe to measure gas levels had a faulty O2 cell discovered on start-up. This voided any measurement not only on O2, but also on CO and NOx, and so no measurements could be taken on the toxic gas levels.
The water mist pipe work was installed according to the manufacturer’s design guidelines. All installation was carried out using 12mm x 1.5mm stainless steel grade 316L tube and compression fittings. Pressure was supplied using a single phase, high pressure piston pump at 100 bars and the maximum flow available was 12 litres per minute. Nozzles installed were Fogex FA20 nozzles c/w temperature bulb of 57 degrees C with a k-factor of 0.64 litres/minute. At 100 bars upon activation the flow per nozzle is 6.4 litres per minute.
Flow calculations were performed in the system allowing for worst case total discharge. The flow calculations were based on the Darcy-Weisbach formula for high-pressure systems.
The 3-bedroom apartment contained a kitchen and living area. All access was from an entrance/escape corridor that terminated in the living area. The plan in the figure shows the configuration of pipe work, the positions of the sensors and the test fire locations.
The test was to represent a ‘Class A’ fire. A makeshift sofa was fabricated using foam sandwiched between two pallets and the fire was ignited using paraffin accelerant. The ‘sofa’ was located approximately 1.5m away from a single nozzle. Although rapid cooling by water mist suggests that only a single nozzle would discharge, we made sure this would happen by limiting the pump capacity to a single nozzle at design pressure. The system activated after only 40 seconds and extinguished the fire in approximately three minutes.
There was a rapid drop in temperature post-discharge. Maximum temperature reached in the kitchen was 63oC and in the living area 70oC. A cold discharge test (non-fire) was also carried out to demonstrate the mist characteristics.
Visibility
During the water mist discharge the room was filled with mist and smoke, although there was visibility across the room. The exit corridor was clear of mist/smoke from 1.5m downwards (i.e. the smoke was above 1.5m or head height). As soon as the discharge was stopped (bearing in mind that the fire was extinguished), the smoke dropped to about 0.5m above the floor. So it would appear that the mist kept the smoke at high level and permitted for a safe escape.
It is unfortunate that gas measurements could not be taken. However, from a similar set of unpublished tests, the carbon monoxide reduces on discharge, and it is reasonable to assume that this would have occurred here. People in the vicinity of the test were not adversely affected by carbon monoxide levels reached. Temperature reduction was rapid, as measured by the thermocouples, and it was observed that no other nozzle bulb discharged.
Standard development
British Standards Institute (BSI) plans to develop a code of practice for water mist in domestic premises. Such a document should provide a level of protection at least equivalent to that given by the sprinkler systems standard, BS9251: 2005. In drawing parallels and suggesting proposals based on the information above, the following comments should be considered.
BS9251 for sprinkler systems is restricted to residential occupancies not exceeding 20m in height. This restriction need not apply to high-pressure water mist systems that have the ability to supply the necessary pressure and flow to heights many times greater than this, from a single pump set. The same applies to the maximum individual room size.
BS9251 details several configurations of installations that use either mains fed pressure, or with a supplementary booster pump, or with a pump and stored water; on both metered and un-metered systems. The tests detailed here have been based on a high-pressure water mist system and mains fed pressure would not be adequate. The system uses a pump, so this would be a standard requirement. Stored water could be an option with low flow rates, but the type of pump used requires a feedback to the storage tank, and so a storage tank is recommended. For domestic requirement, cylinders could be used, as the required water demand is minimal.
The criteria for sprinkler systems is to provide flow for a single sprinkler head of 60 litre/min, or up to four sprinkler heads of 42 l/min. Maximum flow is therefore for four sprinkler heads. There is no problem translating this to the number of water mist nozzles as long as the coverage is the same. The FOGEX FA-20 nozzles tested are at least equivalent to sprinkler coverage. Water mist nozzles, as tested, do not work on reduced flow and the tests showed that once one nozzle discharged, the likelihood of any subsequent nozzle activation is reduced considerably. It would be reasonable to assume that a maximum flow for only two nozzles at full flow rate should be applicable. Maximum run time is for 10 minutes domestic and 30 minutes residential. Based on the BS9251 standard, the minimum water quantity for an equivalent FA20 nozzle would be:
Domestic
1 off @ 6.4 l/min for 10 minutes = 64 litres
2 off @ 6.4 l/min for 10 minutes = 128 litres
Residential
1 off @ 6.4 l/min for 30 minutes = 192 litres
2 off @ 6.4 l/min for 30 minutes = 384 litres
Minimum operating pressure has been type-tested to 70 bar.
Annex A details hydraulic flow calculations and pressure loss based on the Hazen-Williams formula. This can be used as an approximation for high pressure water mist systems, although the Darcy-Weisbach formula will give more accurate values. The pipe table can be replaced by those relating to the smaller bore water mist pipe for smaller flow rates (normally stainless steel).
1 Effectiveness of Sprinklers in Residential Premises, BRE
2 Executive summary available from www.odpm.gov.uk and a full version from www.bre.co.uk
This article is based on a study by Glenn Horton of Locke Carey Consulting and David Yirrell of Water Mist (Fire) Technologies. The authors would like thank London Fire and Emergency Planning Authority and St George Ltd for helping to make the tests possible.
