Water-flow Requirements For Firefighting Purposes In The Uk

Paul GRIMWOOD Firetactics.com _______________________________________________________________________________

Water-flow requirements for firefighting purposes in the UK Paul GRIMWOOD _______________________________________________________________________________ There have been several international research projects over the past 50 years that have attempted to produce an engineered solution to water flow-rate requirements for structural fire-fighting purposes. These studies have often been based upon scientific data associated with heat release rates from compartment fires along with empirical research investigating actual flow-rates used by fire brigades when tackling fires in a wide range of occupancy types. This information is most useful for grading fire-fighting water flow requirements in-line with building codes. It is also of use to the operational fire officer who must assess the resources required at a particular incident to suppress any structure fire of a known or estimated size. This paper is based upon a 100 fire research study in London in 1989-90 that was the first to propose firefighting flow-rates that were controversially lower than those calculated in similar studies by Thomas (UK) 1950; and National Fire Academy (USA) 1983. Fire engineers therefore questioned the resulting flow-rate formula produced and published at that time. However, even some years later, this research remains highly accurate and correlates closely with more recent scientific research completed in London, Sweden and New Zealand.

_______________________________________________________________________________ Before my own research study [1] in 1990, the most established research to date had been completed in the USA although there had been several small-scale laboratory studies investigating theoretical flow-rates to suppress minor compartment fires. The conclusion of my research covering 100 major fires in London 1989-90 demonstrated a recommended flow-rate that appeared controversially low in comparison to those used in the USA and caused a debate that prompted further research. This research occurred between 1994-97 when Lund University Sweden supported London Fire Brigade in a 307-fire study that culminated in the Sardqvist report [2] in 1998. The flow-rates reportedly used by London firefighters in this study were substantially higher than those I had calculated in 1990, but why did this happen? – Were my findings somehow under-estimated or did the Lund 7003 report produce an over-estimate? It is beyond any doubt that the nozzle flow estimates provided by London Fire Brigade upon which Mr Sardqvist based his calculations were not representative of actual flows achieved on the fire-ground. In fact, I have calculated that these theoretical and unrealistic nozzle flows actually resulted in the Lund 7003 flow-rate curve demonstrating a 40 percent overestimate. As a serving operational firefighter in London during part of this research period I am able to attest that flow-rates detailed in SRDB Codes at that time were rarely, if ever, achieved on the fire-ground due to a number of factors including hydrant flow capabilities, frictional losses and nozzle reaction forces. There is no mention in the codes of nozzle/hose sizes, or nozzle reaction forces, that would have direct impact on the amount of water an interior attack hose-line could effectively flow. As an example, any attempt to flow an attack hand-line at 870 LPM would produce a nozzle reaction force that could not possibly be handled safely by an interior attack team [3]. Further still, to suggest that pressures of 5 bars are regularly achieved at the nozzle is generally unrealistic and often impractical as it is well established that UK firefighters (esp. of that time) traditionally under-pumped their attack hose-lines with pump pressures of 4-5 bars being common. My practical experience at that time would suggest that maximum flows of 200 LPM from a 12.5mm nozzle; 450 LPM from a 20mm nozzle and 700 LPM from a 25mm nozzle on interior attack hose-lines used during the research period were far more realistic than those suggested by the theoretical _______________________________________________________________________________ Water-flow requirements for firefighting purposes in the UK

[email protected]

Paul GRIMWOOD Firetactics.com _______________________________________________________________________________ SRDB Codes, as used in the Lund research. At large incidents the flow-rates may even have fallen below these estimates due to hydrant capability at the grid. Sardqvist acknowledged some of these points in his report. In 1994 a further study completed by Barnett [4] in New Zealand produced scientific data, supported by much empirical research, which provided a foundation for the MacBar Fire Design Code in 1997. This research produced a flow-graph that is closely correlated to my own earlier work and interestingly, where the Lund 7003 flow graph is amended (figure one) as demonstrating a 40 percent over-estimate, this too falls much more inline with both the Barnett and Grimwood research findings.

Figure One: Comparison of research studies for fire-fighting flow-rate requirements based upon low-category risk.

When converted to an area formula my original calculation, based on a mean average for comparison to Lund 7003, for minimum (and realistic) fire-ground flow-rate requirements (based on office compartments with 2.5 metre high ceilings) suggests that A x 2 = LPM (where A = area in Sq. m). If high-risk occupancies are involved then my own 1990 calculations converted for area flow-rate (A x 4 = LPM) appear far less controversial, especially when applied to fires involving up to 100 Sq. m of floor space. Interestingly, the Barnett 1994 and Grimwood 1990 studies demonstrated a flow-curve directly proportional to the area of the fire and not roughly proportional to the square root of the area of the fire as suggested by Sardqvist in 1998. In 1984 we were introduced to the innovative 'new-wave' Swedish concepts of 'offensive' water-fog applications by Mats Rosander and Krister Giselsson. Their approach revolutionized compartmental firefighting techniques and several fire authorities around the world have since approved this form of attack. In 1994 the UK Fire Research Station finally confirmed the principles involved were sound and the techniques were adopted a year later as national policy. That same year the U.S Navy adopted this style of fire attack on board ships and by 1998 firefighters in Australia were developing the techniques further still. Currently fire authorities in Holland, France, Germany, Spain, Italy and USA are seen to be exploring the most recent developments concerning 'new-wave' water-fog techniques. However, there are current and emerging trends throughout the UK fire service, in terms of compartment firefighting flow-rates, that may appear somewhat disturbing and should prompt some cause for concern. _______________________________________________________________________________ Water-flow requirements for firefighting purposes in the UK

[email protected]

Paul GRIMWOOD Firetactics.com _______________________________________________________________________________ Minimum Flow-Rate When Rosander & Giselsson originally reported on the Swedish concepts they recommended a 'minimum' flow-rate of 100 litres per minute (LPM) for gaseous-phase cooling applications. The TA Fogfighter jet/spray branch that is commonly used by Swedish firefighters incorporates a flow-range of 100-300 LPM or 175-450 LPM. In the UK we have adapted the Swedish techniques in-line with our initial attack strategy which is based around 19mm (some brigades use 25mm) hose-reel tubing supplied by high-pressure pumps. Where operated within the previously accepted range of 15-40 bars (pump pressure) the expected flows from the nozzle would be somewhere within the range of 85-230 LPM for 19mm bore and 250-400 for 25mm bore hose-lines. These flows were considered generally acceptable for compartmental gas-cooling applications and routine fire-ground work including most one-room fires, car fires, rubbish fires etc although some Swedish FBT specialists, at the time, did voice their concerns over the lower flow-range being utilized by UK firefighters. However, as fire behavior training has developed over the past five years in the UK and firefighters have become accustomed to the 20 Cubic metres of 'container' fire where they learn, practice and perfect the techniques of 'gaseous-phase-cooling', a realization has dawned that the 'ideal' flow-rate for tackling a container 'burn' is around 40 LPM. It has been suggested that where flow-rates are utilized in excess of this 40 LPM the danger of steam-burns exists to nozzle operators. It is certain that flow-rates much above 40 LPM in such facilities serve to hinder the training effect and appear ‘too effective’ in suppressing the fire. CFBT instructors therefore prefer the lower flows to enable continuity in repeating ‘flashover’ evolutions in the simulators. This has resulted in manufacturers offering 'lowflow' options for the hose-reel branches and brigades are now seen to be equipping themselves with an initial 'attack' flow-range of 45-90 LPM on the hose-reel system fitted to front-line fire engines - This is dangerous and throws up major Health & Safety implications. It is not difficult to recognize that the average one room fire may be 3-4 times the size of a standard container training burn. Additionally, the fire loading of a furnished 'room' is far higher than will be found in the standard container, be it carbonaceous chip-board or gas fuelled. It is essential to remember that container systems are only simulators and they simulate ignitions of fire gas layers, not class 'A' furnishings. As an extinguishing medium, water has a theoretical cooling capability of 2.6 Meg-Watts per litre per second although in the practical application of a 'real' fire it's capability is more likely to be around 0.84 MW per litre per second. To the firefighter this simply means that the nozzle in use has a maximum 'practical' cooling capability and reliable estimates may be derived (table one). This is a simplistic approach but is generally accurate in practical terms. 50 lpm

0.70MW

100 lpm

1.40MW

150 lpm

2.10MW

200 lpm

2.80MW

300 lpm

4.20MW

550 lpm

7.70MW

800 lpm

11.20MW

1000 lpm

14.00MW

Table One - Practical Cooling Capability of Water

The true relevance of these figures is only realized when Heat Release Rates (HRR) common to one-room fires (compartments) are closely examined. (Note: These are conservative _______________________________________________________________________________ Water-flow requirements for firefighting purposes in the UK

[email protected]

Paul GRIMWOOD Firetactics.com _______________________________________________________________________________ figures as they generally only apply to individual items of furniture and not surrounding fuel-loads or wall and ceiling linings etc).

Pine Bunk-Beds

4.5 MW

Sofa (2-seater)

3.0MW

Sofa (3-seater)

3.5MW

Upholstered Chair

2.0MW

Rubbish Bin (small)

300KW

Light Bulb

100W

Xmas Tree

0.7MW

Small Dresser

1.8MW

Single Mattress

1.0MW

Pine Bunk Beds

4.5MW

5 Timber Pallets

1.8MW

2 Panel Work Station

1.8MW

3 Panel Work Station

7.0MW

Kings Cross Fire

15-25MW

(not NIST)

1.5MW

(not NIST)

Container Simulation

NIST

Table Two - Typical Heat Release Rates (HRR) of Common Furniture/Fires – Source NIST USA & others

_______________________________________________________________________________ Water-flow requirements for firefighting purposes in the UK

[email protected]

Paul GRIMWOOD Firetactics.com _______________________________________________________________________________ It is generally accepted, as a 'rule of thumb' guide, that a fire of at least one Mega-Watt HRR is required to achieve flashover. However, it has also been observed that flashover may occasionally be achieved with a HRR as low as 300KW and a recent US report described a backdraft that had been modelled on trash bags emitting just 25KW HRR! It should further be noted that a HRR recorded from a reconstructed hotel bedroom 'furniture' fire peaked from 2MW to 7MW within seconds when the timber wall linings became involved! It now becomes apparent that flow-rates below 100 LPM are dangerously low and would not provide the firefighter with a safe and effective means of tackling anything more than the smallest of fires. I have spoken to firefighters who have occasionally struggled with the 40 LPM flow-rate in a container simulation and then just managed to control the 'burn' when opting for the higher flow-rate of 90 LPM. This would confirm the above data (table one) as practical estimates of flow-rates needed to deal with various sized fires. It is important to remember that container burns are three-dimensional simulations producing burning gas layers where there is no major class 'A' fuel load. The heat absorption capacities associated with the application of water sprays, for cooling fire in the gaseous phase, are increased in comparison to dealing with surface cooling of burning fuels. There have, in the past, been situations where fire authorities have been held vicariously liable for allowing firefighters to utilise low-flow rates in ‘real’ compartment fires where the risk of 'flashover' is existent (isn't that almost every compartment fire?). Maximum Flow-Rate A further 'cause for concern' in relation to flow-rates is emerging in terms of 'maximum' available flow for compartment firefighting. As UK fire brigades progress the modern transition from solid stream 'smooth-bore' branches to the more versatile combination fog nozzles a disturbing trend has seen several brigades (including the largest UK metropolitan brigade) reduce their maximum compartmental flow capability for hand held hose-lines from 700 LPM (25mm nozzle @ 3 bars NP) to 400 LPM (combination nozzle @ 7 bars Pump Pressure). This has been the result of several internal operational reviews of compartment fire-flow requirements and I have discussed this drastic 43% reduction in flow-rate with several fire officers who are of the opinion that 400 LPM is the highest flow needed for a single attack hand-line. It is worthy of note that such flow-rates fall way below the recommended minimums recognised by recent European standards commissions in France and Germany in relation to main-line firefighting nozzles. Let's take a close look at this 'maximum' flow-rate and assess its practical capability on the fireground.

Figure Two - Maximum flow capability from one Line of 2 x 45mm Delivery lengths with modern

combination jet/spray type nozzle.

_______________________________________________________________________________ Water-flow requirements for firefighting purposes in the UK

[email protected]

Paul GRIMWOOD Firetactics.com _______________________________________________________________________________

Examples of ineffective, low-flow, 500 lpm nozzles attempting to control or reach a fire beyond the capability of the stream in use - Photos London and Paris Fire Brigades (note 'break-up' of stream patterns)

The 450 lpm combination nozzle is a versatile tool that will provide optimum flows normally at a factory pre-set of 7 bars nozzle inlet pressure although this can be lowered as an option. It can be seen, from table one, the 450 lpm nozzle may appear to handle most residential room and contents fires although it will certainly be on its limit if a set of pine bunk beds become fully involved! However, what happens if these 'furniture' fires spread to involve additional items? What if they achieve flashover in the room and involve everything including flammable wall linings? What if the fire spreads into the hallway or even involves two rooms? What if the fire's in a high-risk occupancy with an increased fire load? What if it's involving the 3-panel work station in a modern office complex? Suddenly things look different and the critical flow rate (CFR) we normally try and achieve to counter such propagating fires is beyond the capability of the hose-line in our hands. Things can only get worse! What if the hose-line we are using is the 45mm line consisting of three 25m lengths a standard run into an average to large property. The pump pressure required to deliver 450 lpm at the nozzle is 11 bars (on the level)! That's above the UK fire service routine test pressure for hose so let's reduce it to a workable 9 bars. You will now see our original 450 lpm at the nozzle has reduced drastically! If it's the bunk-bed fire we will either have to lay a second hose-line in support of the initial attack line or the crew on the nozzle are in for a hard time and will take unnecessary punishment. The fire may even propagate in its transition from compartmental to structural as the firefighters attempt to play 'catch-up' inline with it's development. _______________________________________________________________________________ Water-flow requirements for firefighting purposes in the UK

[email protected]

Paul GRIMWOOD Firetactics.com _______________________________________________________________________________ As we adapt to new techniques and progress the gradual transition towards technologically advanced firefighting branches it is essential we recognize the true capability of our firefighting streams. If we fail to respect the demands placed on firefighters advancing decreasing interior hose-line capabilities against increasing fire loads then we may undermine their own capable limits also. Iowa Gallonage Formula. In 1954 Keith Royer and Bill Nelson of Iowa State University created the Iowa gallonage formula based on scientific research and supported by experiments completed in a series of test fires. The formula demonstrated that Gal = V/200 where V = compartmental volume in cubic feet. This formula referred specifically to indirect extinguishing methods. When converted for comparative purposes it can be seen (Fig 3) how closely correlated the various flow formulas are.

Fig 3 Comparison of research studies for Fire-fighting flow-rate requirements based upon low-category risk.

However, it is generally recognized that US firefighting flow-rates (as used) are much higher than that suggested by the Iowa rate-of-flow formula and actual flow-rates nearer the NFA flow formula (some three times higher) are probably more realistic. There are several reasons why US firefighting flow-rates may appear higher than elsewhere 1. Residential building construction in the USA is mainly timber-framed whilst many of the older inner-city tenements still provide timber staircases and flooring with little fire resistance. Fire spread in such structures may be greatly affected by wind and weather conditions. 2. US firefighters commonly practice VES (Vent- Entry- Search) tactics and are more likely to make openings in a fire-involved structure to release smoke, heat and gases. This causes structural fires to burn with greater rates of heat-release, demanding an equally high water flow to counter the rise in temperature. 3. The US water supply system for firefighting is generally conditioned to flowing larger amounts at higher pressures. The water mains, the hydrants, the pumpers and the hoses in USA are all capable of flowing more water than their European equivalents. 4. Traditionally, US firefighters attempt to pitch their fire attack at the base of the initial gradient on the critical flow rate curve in an attempt to suppress the fire quickly and effectively, often with an element of 'over-kill'. This is seen through the increasing popularity of high-flow nozzles such as Vindicator. However, US firefighters have occasionally been forced into using lower flow-rates on a par with European flows to _______________________________________________________________________________ Water-flow requirements for firefighting purposes in the UK

[email protected]

Paul GRIMWOOD Firetactics.com _______________________________________________________________________________ suppress fires, and have done so with success, for example, at major high-rise fires such as the Interstate Bank (LA) and the Empire State building (NY), as reported in Fog Attack (1). 5. US Firefighters have gradually moved away from the widespread use of 'indirect' waterfog applications due to apparent misunderstandings of application technique. _______________________________________________________________________________ REFERENCES (1) (2) (3) (4)

Fog Attack (DMG Publications) http://www.firetactics.com/FOG-ATTACK.htm Sardqvist Stefan (Lund University Sweden) Research Note 7003 (1998) Appendix 1 - http://www.firetactics.com/NOZZLE-REACTION.htm Barnett CR (Macbar Fire Design Code) 1997

Further reading – A Review of Firefighting Water Requirements (Simon Davis) University of Canterbury NZ 2003 - http://www.firetactics.com/sdavis.pdf ______________________________________________________________________________ Paul Grimwood served 26 years as a professional firefighter, mostly within the busy inner-city area of London's west-end. He has also served in the West Midlands and Merseyside Brigades as well as lengthy detachments to the fire departments of New York City, Boston, Chicago, Los Angeles, San Francisco, Las Vegas, Phoenix, Miami, Dallas, Metro Dade Florida, Seattle, Paris, Valencia, Stockholm and Amsterdam. During the mid 1970s he served as a Long Island volunteer firefighter in New York State USA. It was in his book Fog Attack (1992) where the conclusions drawn from his ground breaking empirical research study into ‘firefighting flow-rates’ first appeared. This research, based upon a 100 fire study in London 1989-90, proposed firefighting flow-rates that were controversially lower than those calculated in similar studies by Thomas (UK) 1950; and National Fire Academy (USA) 1983 and fire engineers therefore questioned the resulting flow-rate formula he produced at that time. However, even some years later, his work in this field remains highly accurate and correlates closely with more recent scientific research completed in London, Sweden and New Zealand. _______________________________________________________________________________________

APPENDIX 1 In 1990 I completed a research project (Fire Magazine UK - November 1992) that evaluated the operational capability of fire fighting hand-line streams as used by London Fire Brigade. At that time we had main-line options of 45mm (1 3/4") hose-lines with 12.5mm (1/2") nozzles and 70mm (2 3/4") hose-lines with either 20mm (3/4") or 25mm (1") nozzle options. One of the basic laws of physics - Newton's third law - states that for every action there is an equal and opposite reaction. Quite simply, to the firefighter this means that as water is projected from a nozzle to form a 'jet' or firefighting stream, the nozzle tends to recoil in the opposite direction. This effect, termed nozzle or jet reaction (or kick-back) requires the firefighters at the nozzle to exert sufficient effort into over-coming this reaction force. The entire force of this reaction takes place as the water leaves the nozzle and whether or not the fire stream strikes a nearby object has no effect on the reaction. Thus, whether or not a hose-line's stream is allowed to strike a wall whilst a firefighter is working it from the top of a ladder is immaterial to his stability on the ladder, which is governed solely by the reaction at the nozzle. By evaluating maximum flow capability for a hose-line that could be effectively directed and safely handled whilst advancing and working inside a fire-involved structure It was observed that there was a maximum nozzle reaction force that could be handled by one, two and three firefighters as follows _______________________________________________________________________________ Water-flow requirements for firefighting purposes in the UK

[email protected]

Paul GRIMWOOD Firetactics.com _______________________________________________________________________________ One firefighter - 266N (60 lbf) Two firefighters - 333N (75 lbf) Three firefighters - 422N (95 lbf)

These were interesting findings and from these I was able to establish baseline flows for interior firefighting operations. To achieve this it became necessary to take acceptable pumping practice into consideration without contravening the limitations placed upon european pumps, hoses and equipment available at that time. This resulted in safe and effective maximum baseline flows of 277 lpm (73 gpm) on 45mm hose-lines with 12.5mm nozzles, 650 lpm (172 gpm) on 70mm hose-lines with 20mm nozzles and 750 lpm (200 gpm) on 70mm hose-lines with 25mm nozzles, as advanced by two-man crews. However, these 'base-line' flows were rarely, if ever, achieved in practice as tradition had established a base-line pumping pressure of 3-4 bars (45-60 lbs psi) to which a small adjustment may have sometimes been made for frictional loss and pressure head. Actual flow-rates were, in fact, far lower than had previously been thought - Ref: SRDB Codes of the period. Interestingly, similar research has been carried out by other fire departments, notably San Francisco, Los Angeles and Chicago, who proposed that a safe and practical baseline flow for a workable firefighting hand-line would be around 550 lpm (150 gpm). More recently (1996), the City of St. Petersburg in Florida USA have established that, for their purposes, the ideal baseline flow is around 600 lpm (160 gpm) using a 7/8" (22mm) nozzle with a 50 lbs psi nozzle pressure on a 45mm (1 3/4") hose-line. This set-up will create an acceptable reaction force of 266N (60 lbf) and offers a hose-line that is easily advanced and maneuvered for interior position. However, the change to combination fog/straight-stream or automatic nozzles brings a demand for higher nozzle pressures to achieve similar flows and with that comes an increased reaction force. A baseline flow of 600 lpm (160 gpm) being discharged from a combination/automatic type nozzle operating at 7 bars (100 psi) NP will produce a reaction force of 356N (80 lbs lbf) which would cause a two-man team to struggle with any workable advance of the line. There are combination/automatic nozzles available that have been adjusted to provide rated flows at lower nozzle pressures but be sure to test these yourself as manufacturer's 'rated' flows are sometimes unachievable! Top US branded nozzles must meet the stringent demands of NFPA standards and Low-Pressure combination nozzles are able to achieve their rated flow-rates at factory-set nozzle pressures of just 5 bars. This would enable a flow of 600 lpm (160 gpm) to be achieved with a reaction force of just 303N (68 lbs lbf) which is more easily handled and advanced by a two-man team. The firefighter is able to calculate the amount of nozzle reaction (NR) by resorting to various formulae NR (Newtons) = 1.57 x NP x d squared/10 (European Smooth-bore), or; NR (Newtons) = 0.22563 x lpm x Sq.root of NP (European Combination fog/jet or automatic Nozzles)

These are metric formulae where P = Nozzle Pressure; d = Nozzle Diameter; lpm = Flow in Litres Per Minute and NR is in Newtons. In the USA different formulae are used as follows NR (lbf) = 1.57 x d squared x NP (US Smooth-bore), or; NR (lbf) = 0.0505 x gpm x square root of NP (US Combination fog/straight or automatic Nozzles)

_______________________________________________________________________________ Water-flow requirements for firefighting purposes in the UK

[email protected]