‘September 1991 resi WOES SACRE | 3 rch in’ © sOvSIB (87(F47)(K22) . ss : i es Fire spread between caravans MP Shipp, 8Sc, CPhys, MinstP ‘This paper describes work carried out to examine the spacing distance required to prevent the spread of fire between park homes (mobile homes) ‘and holiday caravans. Ignitability tests were conducted on samples of caravan material and two complete caravans were fire tested. Factors taken into consideration were the construction materials, combustible items kept near the caravans and fire screens. The paper will be of interest to the caravan industry, site owners and local authorities. INTRODUCTION Park homes are mobile homes that are permanently of wall materials, curtains, etc were tested for ignitability occupied, whereas holiday caravans may be occupied only using the ISO ignitability’ apparatus at FRS, The ignition of for the holiday season. Both are controlled on sites by the __vertcaly orientated samples of the same materials was also Caravan Sites and Control of Development Act 1960' and investigated. In addition existing ignitabilty data were the Caravan Sites Act 1968. The former calls upon Model sought. Standards issued by the Secretary of State which required that every earavan should be not less than 6 m [rom any ‘Two complete caravans — one a holiday caravan, the other a other caravan in a separate occupation park home — were instrumented and fie tested, Both were representative of the models currently manufactured. Results In recent years there have been requests from the caravan industry and site owners for this separation requirement to 100. —$ aT be reviewed, their case being that the structure and contents of caravans have changed significantly since 1977 when the 90) standard was last revised. With the Department of the Environment undertaking a further revision of the Model aol Standards and the British Standards Institution planning & standard for the spacing of caravans, the Fire Research 20 - Station was asked to re-examine the spacing requirements in the light of changes in caravan design. Only park homes (mobile homes) and holiday caravans were studied, touring caravans being excluded. Details ofthe research results are given in Fire Spread Between Park Homes and Caravans! and the findings have been incorporated into revised Model Standards 60 Ireadiance (kW/m?) A desk study was carried out to seek and examine vackground and statistical information, review current aaravan design and consider other aspects of the spacing sroblem, including the use of fire barriers and the effects of adiation on escaping occupants. In addition the FRS fire investigation team visited a site where a park home had been badly damaged by fire, 1o examine the radiation damage to adjoining properties bre press at aasssesese —_} 6 10 15 20 2 30 Time (minutes) Holiday caravan TESTS + Park home The National Caravan Council (NCC) arranged for materials, ‘nd caravans to be supplied for testing. A number of samples Figure 1 Heat radiation (irradiance) recorded 3m fom each unit - Printed copies of this document are available from BRE Bookshop Building Research Establishment Garston, Watford, WD2 7JRfrom the fire tests provided both input and validation data for a computer model which caleulated the radiation field from a burning caravan, RESULTS AND DISCUSSION Fullscale tests A holiday caravan and a park home were instrumented by Warrington Fire Research Centre and burnt at FRS's Cardington Laboratory. The heat radiation (irradiance in Wim) was measured at a number of locations around each unit. These tests were believed 10 be representative of the range of caravan model designs. ignition conditions, age of| caravans, ambient conditions etc so that the results could be used to propose a ‘design fie” for each type of unit. The tess were carried out instill ait The two units were seen to burnin different ways (Figure 1) The roof of the holiday caravan gave way very quickly so that the fire burnt upwards with the air being drawn in through doors and windows. and later the walls (Figure 2). Because of the more robust construction of the park home its roof remained intact for some time and flames jetted from the doors and windows, resulting in a more severe fire than that inthe holiday caravan (Figure 3), The results ofthe ignitablity tests on component materials hhad suggested that the aluminium of the holiday caravan could withstand thermal radiation for longer than the plywood of the park home. This difference in the standard of construction was evident in the burning behaviour of the two types. though the heating in the full-scale tests was by direct flame impingement, not just radiation, Ignitabil ‘The wall structure of an aluminium skinned holiday caravan (Ge outer skin, insulation and wood:-based inner skin) did not, ignite at irradiances below 50 kWim, although the paint burnt briefly without sustained ignition. The external fixtures and fittings (eg window frames, vents, tc) probably would have ignited at irradiances below this igure, but the actual critical value was not determined from these tests, though published data suggested values of 15-20 kWin?. Howeve curtains exposed directly through an open window did not ignite at izradiances below 17 kWims, Figure2 Holiday caravan after six minutes ‘The plywood skin of a park home (with or without finish) did not ignite at irradiances below 17 kW Other materials such as the tyres of cars, wood:- based products and gas cylinders — all of which are likely to be found close 10 caravans on a site found unlikely to ignite at irradiances below 15 kW/m, and more generally 20 kWim:. although wood may have done so at 12.6 kWim?, Computer pre‘ ‘Accomputer model was developed to enable the irradiance from a burnin; fon an adjoining unit at varying distances and orientations, The results from the full-scale tests were used to validate a esign fire model for each of the two types of caravan. This was used to interpolate or extrapolate irradiances under selected conditions. such as the presence of a fire screen, for example. The model was used to show the distance from a burning caravan at which some defined irradiance would be experienced caravan to be calculated where itis incident Some specific limitations of the model were: © fire conditions 10 minutes after ignition when irradiances ‘were at a maximum, © results ofa single fire (for each case) in a single mode! ‘under specific conditions of ventilation and ignition © Windless conditions, © included some assumptions and were not exact. and © only plane rectangular flames could be modelled. Separation distances The design fire ‘The ignitability criteria proposed, based on results of laboratory tests and available literature, and presu presence ofa pilot flame (eg burning brand), were the 50 kWim? 17 kWir holiday caravan structure holiday caravan and park home generally other surrounding or intermediate items (eg cars, sheds, gas cylinders — including hazard to humans) k Wim Figare 3. Park home after six minutes‘The results of these tes fires were modelled to produce irradiance fields around the two types of caravan 10 minutes after ignition, and were used to create design fire contours for the holiday caravan (Figure 4) and the park home (Figure 5). Manufacturers’ variations in materials were presumed not to affect the design fire. o~ s7 Wie re6kWine Based on 3m high target Figure 4 Holiday caravan desig fire contours 12.8 kWim? 7A W i? om Figure S Park home design fire contours Holiday caravan ‘The Model Standards’ spacing requirement of 6 m might be reduced to 5 m for holiday caravans with aluminium exteriors, The spacing contours at the corners could be cut ‘off so that where the units are positioned ‘corner to corner’ a lear space of 3.5 m should be sufficient to ensure no fire spread (Figures 4 and 6). This presumes no major difference in the proportions of a holiday home caravan from the one that was tested. However, the 6 m spacing requirement should be retained for holiday caravans with a plywood exterior, Park home Figure 5 shows there was no flexibility to reduce the current {6m spacing requirement for park homes. A simplified separation boundary was therefore inappropriate. There is, however, no evidence from reports of real fires to indicate any risk where the spacing is 6 m, so there was no reason to increase this distance, Other intermediate items ‘There appeared to be no statistical or anecdotal evidence to suggest that the presence of sheds, cars, porches, ele between park homes (or indeed between 6 m spaced holiday caravans) was creating a hazard to life. There was therefore litle strong argument to support the use ofthese separation ‘boundaries for intermediate items where holiday caravans or park homes are spaced 6 m apart. ‘On the basis of the fire test results it has to be presumed that any items (other than those classified as non-combustible") within 3 m of a burning caravan could ignite and be an additional source of fuel and radiation, or indeed, direct ‘lame impingement, However, certain items need to be located between units for practical reasons. The limitations that might be applied are summarised in Table 1. RECOMMENDATIONS 1 There should be no change to the Model Standards? ‘6 m spacing requirement for current designs of park homes. 2. The spacing requirement for holiday caravans with aluminium or other metallic exterior skin could be reduced to 5 m clear space, with 3.5 m space at the corners (Figures 4 and 6), 3 ‘The spacing requirements for holiday caravans with an exterior skin of plywood or similar material should remain at6m. 4 Where there is a mixture of either park homes and holiday caravans, or holiday caravans made of aluminium (or other metal with similar ignitability characteristics) and plywood (or other material with similar ignitabilty ‘characteristis), the separation distance should remain at 6m, 5. Combustible intermediate items such as cars, gas bottles, cle associated with a separate holiday caravan occupancy with the 5 m spacing and within the separation boundary of an adjoining occupancy should be restricted (Table 1). 6 No combustible intermediate item should be within 3m of an adjoining unit (ie there should be 3 m clear ‘space between separate occupancies) except as specified in Table 1. Space for C ‘Space for B ay Areas where ce for A combustible ae items should be restected sace for \ . Space fore \ |-Space for A Space for 8 Figure 6 Holiday caravan layouts‘Table 1 Restrictions on the placement of items within the separation boundary of an adjoining occupancy Notes: 1 The bie aplieso both park homes end holiday caravans, execpt where sated chews. 2 Except forcovered maka? thee weed be ao estrkton onthe or lation or ems outside the separation Pounder ‘Now-combastbie NO) tm comes) Recommend ts cow Pa Wain apa a ang 1 Atpowecony carping - E Guunoneonbote sae * 5 Niwtotaenesuce 4 Oatonee per ae ape ee xe xe Vers c Dera gen rt er ta Sit om nt 1 Nemo ths ie og Porches and bey windows Ne 2 easiest ator oce oe ° 3 Seeds te aro ng Ens ipacione) = ‘Ceara neon tr ing eS Eaves (holiday earavas) = ‘Clearspace between eaves of adoning unis abe | = we Dos oi fin) aa c Thin thermoplastic sheet only “ ae Jo not permit Cond oka s co x on em ae 1 Enna sony ening © {2 setae 3 Dave et = Ne ‘ein < {etm pec ny oa = Se Cm = Teed 7. 3 mbigh fire sereens may be used to reduce the spacing 4-_ Department of the Environment and the Wels Office. for holiday caravans to 3.5 m clear space. They would need to be imperforate, non-combustible and robust. 8 Consideration should be given to producing holiday ‘caravans with an imperforate wall and root, Such units, with blank walls facing, could be separated by only 3 melear space, NB Care must be taken that implementation of 7 and & ‘would not compromise existing means of escape provisions. REFERENCES 1 Caravan Sites and Control of Development Act 1960, London, HMSO, 1960. 2 Caravan Sites Act 1968, Chapter $2. London, HMSO. 1968. 3. Shipp M P. Fire Spread Between Park Homes and Caravans — Recommendations on the spacing prescribed in the Model Standards to prevent fire spread. London, HMSO, 1989, Caravan Sites and Control of Development Act 1960, Section 5, Model Standards 1989: Holiday Caravan Sites, London, HMSO, 1989. Department of the Environment and the Welsh Office. Caravan Sites and Control of Development Act 1960, Section 5, Model Standards 1989: Permanent Residential Mobile Home Sites. London, HMSO, 1989, International Organization for Standardization, Fite tests — Reaction (o fire — ignitabilty of building products. Inernational Standard ISO 5657, Geneva, 1SO, 1986. Department of the Environment and the Welsh Office. 3ravan Sites and Control of Development Act 1960, ‘Model Standards (Revised 1977). London, HMSO, 1977, British Standards Institution. Fire tests on building ‘materials and structures. Part 4: Non-combustibilty test for materials. Brcish Standard BS 476;Part 4:1970, London, BSI, 1970. Price group 2. Also avalabe on subscpon For current ries poaso contact BRE Bookshap, Buling Reseach Fstalstet, Garston, Wao 02 7A (lephone 0823 654448), Ful cas lat roent ames of BAL puoieabons are given in BRE News sen ee subscoers © Crow cxpright 1891, Pubishod by he Bulking Researen Establishment Apolicasons te reproduce extracts should be mado the BRE Puente Manage.