Archive for the ‘Royal and Sun Alliance (RSA) Insurance Group’ Category

Research testing with water mist systems for commercial office buildings

Friday, March 26th, 2010

24 March 2010

 Louise Jackman and Kelvin Annable report on the findings of research into the use of water mist systems in large open-plan office spaces.

In the UK water mist systems are increasingly being considered and used for the fire protection of buildings, including commercial premises and as an alternative to sprinkler systems. However, the impact of design variables in the application of such systems is often not well understood.

Water mist systems employ a spray of fine water droplets that can suppress a fire by cooling, wetting and displacing oxygen (by droplet conversion to steam). In a small compartment, such as a prison cell (3m by 4m and 3m high) with a closed door, water mist has been shown to be very effective, both at suppressing the fire and improving tenable conditions [1]. However, in larger spaces water mist may not be as effective, as small water droplets are not contained in the vicinity of the burning fuel and air/fire dynamics can deflect droplets away from combustion gases. Hence, water mist system designs for larger spaces will often require greater water delivery rates and closer nozzle spacings.

For a water mist system to be accepted for use in a building, it is necessary to undertake a full review of a particular system in the context that it will be used [2]. One key part of this review is a requirement to demonstrate the system’s effectiveness against fire performance tests that are appropriate to the real life application, because each water mist system is a bespoke system. Currently, there are only a very limited number of fire performance tests for different end use applications in the draft British Standard for commercial and industrial water mist systems [3].

BRE Global has recently completed a three year experimental research programme to investigate the parameters that influence the performance of a water mist system. This work was commissioned by the BRE Trust and was supported by industry partners. The aim of the work was to investigate water mist system design parameters, investigate building/room parameters that influence suppression effectiveness, and develop a fire performance test that could be applied to large open office areas.

BRE Trust experimental research programme
BRE Global conducted 48 fire tests, with low pressure and high pressure water mist systems. The commercial systems were provided by industry partners.

To assess the performance of water mist systems, three stages of experimental work were completed:

• Parameter testing with crib fire tests;
• Development of a full scale fire test protocol for open plan office spaces;
• Testing with the full scale fire test protocol for open plan office spaces.

Tests were carried out under an open ceiling, i.e. a freestanding ceiling supported on columns, but with no walls at the edge of the ceiling. Several compartment tests were also conducted.

Parameter testing with crib fire tests
A series of wood crib fire tests in open conditions was conducted, where the fire source was a single wood crib. The wood crib was developed for the project, so that the fire had the potential for progressive horizontal spread. The fire size was relatively small (approximately 500kW), operating an automatic water mist nozzle with a quick response glass bulb rated at 68°C at approximately three minutes under a 2.8m open ceiling.

The crib was ignited at one end and allowed to burn freely until the fire had spread to involve approximately half the crib. A low pressure water mist system (a single operating nozzle) was activated and water was discharged for a 10 minute period. Any remaining fire was manually extinguished at the end of the 10 minutes. During the tests the flames were observed, the temperatures were measured and at the end of the test the damage was assessed. The arrangements assessed included the following ‘water mist system design parameters’: nozzle type; nozzle spacing; and water flowrate; and the following ‘building/room design parameters’: obstructions; ventilation; and compartmentation.

Findings from parameter tests
The findings from the parameter tests with wood crib fires included:
• Wetting of the wood crib fuel load was demonstrated to be an effective means of either reducing fire spread or preventing any further fire spread, depending on the ‘wetting’ flux density.
• Suppression was observed when the crib fire was exposed within the water mist spray envelope. Test results demonstrated that it is critical to install water mist nozzles at spacings that deliver sufficient water over the area of the fire. At small increases in nozzle spacing it has been shown in the open scenarios in this project that fire suppression may not be achieved.
• A reduction in the water flow rate of the tested water mist system resulted in reduced fire suppression effectiveness.
• With the crib fully shielded from the water discharge, only minimal fire control was demonstrated.
• With the crib fire partially shielded, the water mist system was effective at preventing spread to exposed fuel protected by a high water flux discharge density. A level of fire control was also demonstrated within the part-shielded portion of the fire.
• Ventilation, in open conditions, had a highly significant influence on the water distribution pattern of the mist discharge. A significant amount of the water mist droplets could be seen being ‘blown’ away from the fire. The water coverage flux density at the fire location was much reduced. The suppression outcome was dependent on the ventilation rate and location. Ventilation affects both the fire and the water mist system. With ventilation across a floor, the development of the crib fire was affected. Initially, the flaming struggled to become established but then spread quickly. The performance of the water mist system was detrimentally affected by the ventilation. The fire was not effectively suppressed, particularly in comparison to testing in ‘still air’ conditions, and fire continued to spread along the crib.
• Within a compartment, water mist performed well in a fully sealed enclosure, but its performance was reduced with ventilation, such as an open door with or without mechanical room ventilation.

In summary, installation parameters, e.g. nozzle spacings, were found to be critical in achieving effective fire suppression. Building parameters, e.g. fuel shielding, ceiling height, ventilation and compartmentation were all found to potentially have a highly significant influence on the effectiveness of the tested water mist systems. Hence, it is important that for proving and acceptance testing, a fire test protocol accurately addresses the potential impact of these parameters.

Development of a full scale fire test protocol for open plan office spaces
Prior to developing a fire test protocol, an assessment was made of typical open office areas. Information was gathered and reviewed from an office survey, office fire load surveys, office fire test data and standard test fires. In the development of the draft British Standard [3] the UK committee did not adopt the European office fire performance tests [4], preferring to wait for the outcome of this BRE Trust project.

BRE Global sought to develop an office test fire that met the following criteria:

• A ‘stylised’ scenario that represents a typical office fire, in terms of the fire load distribution, fire growth rate and heat release rate.
• A scenario that challenges a water mist system, by including a shielded fire source and an open ceiling.
• A readily repeatable scenario.
• An ignition scenario with the potential to spread both within the initially ignited fuel and to other fuels in the vicinity, so that a reliable progression of fire spread within the fire load is achieved.
• A fuel arrangement that was not susceptible to collapse (in the early stages of the fire).
• A fuel arrangement that allows a clear means for determining pass/fail criteria in terms of limiting the fire spread and the extent of fire damage, by means of temperature and assessment.
• Fuel and materials that can be closely specified easily sourced and repeatedly obtained.
• A scenario that is simple and relatively cost effective.

The BRE Global developed arrangement consisted of two combustible 12mm thick plywood walls at right angles to each other and two 22mm thick chipboard tables. The tables were positioned with a 10mm gap to the walls. Additional fuel loading consisted of two cribs of wood and polypropylene. These cribs were beneath the table. All of the wood-based products used in these tests were non fire-retardant treated. An arrangement of cardboard box files (four of which are filled with paper) and polyurethane foam sheets were placed on the table.

The ‘corner’ wood crib was ignited at the end closest to the corner of the wall arrangement. The fire was allowed to develop freely, involving the first crib, plywood wall and table. The gap between the chipboard table and the plywood walls allows flames to penetrate easily and involve the ‘target’ box files and foam sheets above. For the development work and to establish the repeatability of test, the fuel arrangement was located below the BRE Global calorimeter hood which allowed for the measurement of heat release from the test fire. Additional thermocouples were located above the wood cribs, folders and foam sheets and at a height of 2.5m. You can see pictures of the test below:

Shortly after ignition Flames spreading above table
Fully involved Post test

The heat release from the fire is shown below:

Gas phase temperatures during the test are shown below:

Testing with the full scale fire test protocol for open plan office spaces
A series of tests was undertaken to assess the performance of industry provided water mist systems (both low and high pressure) and a sprinkler system against the BRE Global developed office fire test protocol. The test work was used to develop criteria for the determination of effective fire suppression.
Tests included:
• Sprinkler system – an array of four sprinkler heads on a 3m by 3.5m spacing at a 5mm/min coverage (to aid the development of criteria for effective fire suppression);
• Low pressure water mist system – an array of four nozzles with water supply pressure of approximately 12 bar, on a 2.5m by 2.5m spacing with water supply pressure of approximately 12 bar and nominal coverage of 5mm/min; and on a 3m by 3m spacing and nominal coverage of 3.5mm/min;
• High pressure water mist system – an array of four nozzles with water supply pressure of approximately 100 bar, with both a 3m by 3m spacing and 4m by 4m spacing, and nominal coverage of 1.6 and 2.8mm/min respectively.

The fuel arrangement for the described tests was located centrally within the array of four nozzles. A 6m by 6m open ceiling was used for the installation of the water mist systems. The floor to ceiling height for all tests was 5 m. Nozzles were sealed and the suppression systems automatically activated on operation of heat sensitive elements; at approximately 5 minutes in normal operation.

Success criteria
The criteria determined for a successful test was as follows:
• The water mist system, operating without manual intervention, shall successfully suppress the test fire.
• The temperature, measured 75mm below the centre of the ceiling, after operation of the water mist system, shall not exceed 80oC for a period longer than two minutes for the entire 30 minute duration of system operation.
• There shall be evidence of unburnt foam and box files remaining after the completion of the test.
• Fire damage to the plywood shall not extend to the ends of the walls.

Findings with the full scale fire test protocol for open plan office spaces
All the systems, as a minimum, demonstrated temperature reduction at ceiling level and reduced fire damage, compared with the unsuppressed fire test. However, not all arrangements demonstrated effective fire suppression meeting the criteria for a successful test:
• The sprinkler and low pressure water mist system (at 2.5m by 2.5m spacing) were successful.
• The low pressure water mist system (3m by 3m spacing) was not successful.
• The high pressure water mist system (installed with various arrangements) was not successful.

Fire performance tests are necessary to demonstrate the effectiveness of a particular water mist system for specific end use applications. In this work, the primary focus was on open plan office areas.

BRE Global has developed a fire test protocol that can be employed for testing the effectiveness of water mist systems in this scenario. This stylised office fire test will be submitted to the relevant British Standards committee for their consideration to include as a new Part containing a fire test protocol for open plan office areas in draft British Standard DD 8489 [3].

Overall, the full scale test results were of concern. Most water mist system arrangements were not able to provide expected levels of fire protection for the tested scenario (open plan office areas with a high ceiling). Or, in terms of the design of the tested systems, the spacing between nozzles was too great and the quantity of water discharged too low, to provide effective fire suppression.

Critical for the successful operation of a water mist system are the system design details, in particular, nozzle type, nozzle spacing, water flowrate and building/room design details, in particular, obstructions, ventilation, ceiling height, compartmentation and openings.

The general findings from this work are likely to be equally valid to other types of application and occupancy types. It is therefore advisable to always carry out fire performance tests to support the use of water mist systems in different applications.

This study was conducted as part of a BRE Trust project. The authors would like to thank the following for their contribution to the work, BRE Global colleagues, Tyco Fire and Integrated Solutions (TFIS), Ultra Suppression Systems Ltd, and Royal and Sun Alliance (RSA) Insurance Group.

1. K Annable and P Reading, IWMA Conference 2008 Proceedings, Fire safety in prison cells – effectiveness of water mist suppression systems,
2. CLG Guide 2006, C Williams and L Jackman, An independent guide on water mist systems for residential buildings, CI 71/5/24 (BD2502), BRE,
3. British Standard Draft for Development 8489 Part 1 to Part 7 – Commercial and industrial water mist systems (committee draft only).
4. ‘Fixed firefighting systems – Watermist systems – Design and installation’, European Committee for Standardization, CEN Technical Specification, CEN/TS 14972: 20