Flow method

THE CURRENT national guidance¹ for the provision of water for firefighting in the UK, issued by the Local Government Association in 2007, offers only limited information to fire authorities in estimating the optimum water flow-rates needed for fire control or final suppression in building fires. However, this guidance does provide an engineering-based methodology for grading hydrant flows in line with the risks they serve, although the source of this information is not clear. Both of these topics are now subject to review through a national consultation process.

In the UK, some guidance for estimating needed flow-rate exists in the Fire Service Manual², although the ‘Iowa methodology’ referred to therein is inappropriate for UK fire and rescue services because the associated firefighting tactics are very limited in scope and are rarely used nowadays.

Accurate methodology

Based on proposals by Dorset Fire and Rescue Service, there are currently recommendations that a particular New Zealand methodology³ for estimating firefighting water flow rates be implemented nationally.
This methodology is based on the combination of two research projects into firefighting flow-rate requirements: this author’s 1989 study of tactical flow-rates in London Fire Brigade (TFR 1989) and the TP 2004/1 study by the late Cliff Barnett on behalf of the Society for Fire Protection Engineers (SFPE) (New Zealand).

This author believes that the methodology provided in the New Zealand approach, TP 2004/1, for estimating minimum firefighting water flow-rates offers the most accurate of all known methods. There have been many other similar engineered solutions around the world based solely on scientific theory, while very few have used empirical evidence of actual flow-rates used at real building fires upon which to base their final formulae. The solutions that are based solely on the physical attributes of water as a cooling agent generically provide gross overestimates in needed flow-rates.

In the UK, there have been only two research studies into firefighting flow-rates used at building fires in the past four decades: this author’s TFR 1989 study of 100 serious working fires in London; and Sardqvist’s 1993 study of 307 fires in London. (The latter research was based mainly on a large number of very small fires and the data inputs used for flow-rates were inappropriate).

TP 2004/1 research

Cliff Barnett was a renowned fire engineer in New Zealand and around the world until he passed away in 2008. He was a past president of the SFPE (New Zealand) and his professional contribution to fire engineering is held in high esteem. He sourced over 700 pages of calculations he had used to develop a simple computer programme⁴ for evaluating fire growth and flow-rate requirements, to produce the draft document, TP 2004/1. At this stage, his work was solely limited to the physics associated with the mechanisms of fire suppression, coupled with the cooling attributes of water.

The flow factor originally used in the draft paper was 0.58, which was commonly referred to by fire engineers, based on established efficiency factors used at the time (0.33) for water as applied to structural fires. This suggested that 0.58 litres/second was required to suppress each MW of heat output from a building or enclosure fire. Based on theory alone, and using the 0.58 flow factor, the flow charts produced in the draft document demonstrated a range of flow curves based around a flow-rate of 8.7 litres/m² for 250kW/m² of involved fire load.

In the early part of 2004, Cliff Barnett approached this author and requested a coordinated approach in combining the theoretical research from TP 2004/1 with the empirical research recorded through TFR 1989 (100 serious fires in London), in order to provide a more rounded methodology for estimating flow-rate needs with a far greater level of accuracy.

TFR 1989 study

In 1989, as part of the London Fire Brigade Fire Investigation Team (North and Central Area), this author undertook practical research into the water flow-rates used to suppress 100 serious working fires in London. The fires were all ‘make-up’ fires requiring the assistance of four pumps or more, involving a wide range of occupancies, mostly residential flats and houses but also including offices, restaurants, schools, factories, warehouse and industrial units where fire damaged areas up to 1,720m². The research was validated by comparing needed flow-rates as used at a further 20 serious fires in offices in the USA.

The objectives of the research were to record the actual firefighting water flow-rates used by firefighters to control and suppress structural fires. This entailed a generic range of fire stream applications that included 87% interior offensive and 13% exterior defensive operations, utilising a mix of solid-bore straight stream attacks and some constant flow fog patterns used mainly for operator protection, cooling hot surfaces or extinguishing flaming gaseous combustion.

A primary purpose of the research, which was published in 1990⁵, was to develop an easy-to-use fire ground method (formula) to enable fire ground commanders to estimate water-flow requirements with a reasonable level
of accuracy.

The research demonstrated flow-rates that clearly correlated with the extent of fire damage in buildings. Where the applied flow-rate was below 2 litres/min/m² (83% of fires) then fire suppression was achieved after a struggle and well into the decay stages (burn-out) of fire development on 53% of occasions.

From this assessment, it was considered impractical and potentially unsafe to extinguish fires during later stages on the fire development curve, as compartment fires might well attack the structural elements during this phase and risk structural collapse. For this reason, flow-rates below 2 litres/minute/m² were seen as ‘critical’ and potentially unsafe for future use. Flow-rates of around 3.75 litres/minute/m² were seen as a far safer and more practical option. By rounding the 3.75 to 4.0 litres/minute/m², a useful target flow-rate fire ground formula was produced as:

F = A x 4

Where

F = Flow required in litres/minute

A = Area of floor fire involvement in m²

4 = Flow factor based on an optimum target flow of 4.0 litres/minute/m²

This formula became known as the Tactical Flow-Rate (TFR 1989)

Although the flow of 4.0 litres/minute/m² provided a safe approach in most situations, it was always proposed by the author that in situations where a high fire load existed; or where compartment boundaries had been breached and structural elements were becoming involved; or where the fire development was influenced by increasing ventilation factors; then the applied flow-rate should be increased to 6.0 litres/minute/m² per 250kW/m² of involved fire load.

Flow-rate formula

The F = A x 6 factor became particularly relevant as lightweight construction forms became common, where structural elements may add a percentage to the involved fire load, or where energy efficient compartments with high levels of insulation may increase the risk of energy rich gases in under-ventilated fires suddenly becoming ventilated.

It was further established from the author’s practical research⁶ that a primary firefighting flow-rate below 200 litres/min may be detrimental to safe firefighting operations, even in small enclosures or rooms, where the potential for involved fire loads exceeded 270MJ/m².

Because of this, several UK fire authorities, including London, are now evaluating the use of 22mm or 25mm high-pressure hose-reels, since 19mm hose-reel cannot flow ≥120-160 litres/min. However, the use of a 200 litres/min hose-reel still has limitations in anything more than a one- or two-room residential fire, and transporting fire loads in heavy smoke will suggest that flow-rates should be higher still.

Combining the research

The data from TFR 1989 was closely analysed by Cliff Barnett and its insertion into TP 2004/1 caused an amendment to the flow-rate factor where a greater efficiency factor had been noted with flow-rates used at actual building fires in London. These 100 fires were all serious in nature with some fire involvements up to 1,720m² in floor area.

A new efficiency factor of 0.50 was used to produce a flow factor of 0.38, which resulted in core flow rates of 5.7 litres/min per 250kW/m² of involved fire load for optimum suppression of enclosure fires. This was consistent with the A x 6 ‘rule of thumb’ formula (6 litres/min) suggested by TFR 1989.

The TP 2004/1 approach compares favourably with flow-rate derivations as follows:

6.0 litres/min/m² – TFR 1989/1

6.3 litres/min/m² – UK Water/Local Government Association National Guidance for a 76m² house

6.5 litres/min/m² – BS 9990: 2006: Code of practice for non-automatic fire-fighting systems in buildings (derived) for a 70m² flat

6.25 litres/min/m² – US National Fire Academy plus back-up line of same

This improvement in fire stream efficiency led to major reductions in the needed flow-rates detailed through the modified flow charts in TP 2004/1.

Conclusions

The following conclusions can be made:

1. The vast majority of fires are confined to the compartment or origin.

2. Most building fires only require a very small amount of water to achieve control and final suppression.

3. The features associated with modern lightweight construction and energy efficient compartments suggest that fire development may involve more of the gas-phase than in the past.

4. While the rapid deployment of a hose-reel may achieve some rapid knockdown of fire in the gas-phase, it may struggle to prevent a fast developing fire situation where the fuel-phase fire continues to grow. It has been noted⁷ that where a fixed fire load exceeds 270MJ/m², the deployment of a single 19mm hose-reel against a fast developing fire must
remain questionable.

5. The flow concept of TP 2004/1 is based upon 50/50% suppression/exposures. However, this was a topic that was debated to some great extent by this author with Cliff Barnett during the final draft, and it was agreed that the term ‘exposure’ included the potential for a large percentage of the fire load to transport in smoke and ignite in surrounding areas of a structure, or to become involved in
gas-phase combustion where exterior winds were forcing additional oxygen into the fire zone.

6. For example, where the area of involvement (fire) is just 20m² but the area of involvement (heavy smoke) is 70m² in a low fire load (office), then the overall potential for sudden energy release might be around 17.5MW.
This situation would demand a minimum safe deployment of two hose-lines (attack and back-up) with a combined minimum flow-rate of around 900 litres/min.

7. This author also undertook research in 2002 into the estimated average flow-rate used by firefighters at UK building fires. A series of flow tests with 58 UK brigades demonstrated that the average 45mm primary attack hose-line was being pumped to just 290 litres/min.

8. There is a clear need to train firefighters in the need for effective flow-rate and also to equip them with branches, nozzles and effective hose sizes that are able to deliver the required flow-rate for any specific situation, within reasonably practical limits. Ideally, all fire appliances should be fitted with flow meters, since pressure gauges provide extremely limited information on their own.

9. The concepts of optimum firefighting flow-rate for enclosure fires and water provisions from the hydrant grid are two separate issues and, while they are inherently linked, they should not be confused. It is suggested that Appendix 5 of the current UK national guidance for water provisions¹ should be updated using TP 2004/1 and a detailed engineering analysis of water provisions be undertaken so that a suitable means of grading the hydrant grid may be developed.

10. It is also recommended that TP 2004/1 should be included in any future revision of the UK national guidance document¹ and that any engineered solution to grading water provisions should be based around this methodology.

11. There are a number of engineered grading systems of firefighting water provisions that also take account of the service delivery of a local fire authority, in their ability to effectively meet their requirements in applying water to fires at predetermined capacities and timescales.

12. It may still be the case that the current national guidance for water provisions¹ will remain effective to form the basis of water provision arrangements, in support of the TP 2004 flow-rate methodology.

13. Great care must be taken in addressing issues associated with the removal of fire hydrant stock; depletion of hydrant grids and mains; and any proposals to place greater reliance on water tank supplies as carried on fire appliances. These are matters that must again be closely analysed from an engineering and operational perspective.

14. Any belief that building fires ≤100m² are not in need of firefighting flow-rate assessment, or that appliance water tank supplies are nearly always enough, is strictly flawed. With health and safety concerns, the trend in the UK fire and rescue service is now for secondary back-up hose-lines and higher flow-rates than previously used, and any planned reduction in hydrant stock must be assessed with a critical eye

Paul Grimwood FIFireE is principal fire safety engineer with Kent Fire and Rescue Service

References

1. The National Guidance Document for the Provision of Water for Firefighting 2007, Water UK, Local Government Association.

2. Fire Service Manual: Hydraulics, Pumps and Water Supplies 2001, HM Inspectorate of Fire Services.

3. Barnett, C., SFPE (NZ) TP 2004/1.

4. FireSys Fire Engineering Tools, MacDonald Barnett Partners, Auckland, New Zealand, 2004.

5. Grimwood, P. T., Fog Attack 1990, FMJ International Publications (Fire Magazine), UK.

6. Grimwood, P. T., EuroFirefighter 2008, Jeremy Mills Publishing Limited, UK.

7. KFRS-TB1-FSE-091201-19mm hose-reel Performance REV 1.