A Comparative Analysis of the Use of Different External Fire Spread Calculation Methods

Transcription

1 A Comparative Analysis of the Use of Different External Fire Spread Calculation Methods (1) Declan Thomson, (2) Paul Currie, and (2) Jianqiang Mai (1) Hoare Lea and Partners (6 th Floor Royal Exchange, Manchester, UK, M2 7FL) (2) School of Engineering, University of Central Lancashire (Preston, Lancashire, UK, PR1 2HE) ABSTRACT Ensuring adequate protection to resist the spread of fire to adjacent buildings has been compulsory in UK building design since the 1212 Great Fire of London, after which King John ordered that all roofs should be constructed from (what was thoughts of at the time as) non-combustible materials. A variety of assessment methods are available to demonstrate that the risk of fire spread to adjacent buildings is controlled as far as reasonably practicable. This study was undertaken to review the background literature and compare the results of the UK assessment methods (i.e. the simple methods and the more complex methods) with other international methods (i.e. North American) and research methods. The results of this study have demonstrated that there are substantial differences between the results of the BR 187 Enclosing Rectangle method (which is generally considered to be the go to method for UK fire engineers) and the other UK methods (e.g. the Simple Geometry Method (Scotland)). However, despite the national differences in the assessments, the BR 187 Enclosing Rectangle assessment is considered to correlate well with the results of the NFPA 80A method, although the NFPA 80A method is considered to be more onerous and may not be suitable for use across the site boundary. The study has also demonstrated that neither of the assessment methods provided in NFPA 5000 or IBC should be used in the UK to determine safe separation distances. Keywords: radiation, fire spread, enclosing rectangle, boundary, separation, unprotected area INTRODUCTION It is recognised both in the UK and internationally that buildings should be designed and constructed such that the risk of fire spread to adjacent buildings is adequately restricted. (1-4) This, like most regulatory requirements, was born from disasters. In 1212 one of the original Great Fires of London destroyed a vast portion of the city and claimed thousands of lives, the result of which lead to King John ordering that all roofs should be constructed from (what was thought of at the time as) non-combustible materials. (5) In 1666 a fire ripped through London, destroying 80% of the city and lead to the London Building Act 1667 wherein buildings were required to be fire resisting with a degree of space separation. (5) However both in the UK and internationally this same fate has befallen other cities; Edinburgh 1824, (6) Chicago 1871 (7) and Vancouver (8) Although UK based Building Regulations are generally only intended for life safety purposes, (1-3) it is clear that fire spread from building to building can have catastrophic results and, as such, the risk of occurrence should be controlled as far as is reasonably practicable. However, due to the performance-based nature of the British and American Building Regulations/Codes, there is no specific method that should be adopted when attempting to achieve compliance, provided that a minimum standard of safety is demonstrated. (9) LITERATURE REVIEW Performance-Based Design UK Guidance In England the Building Regulations require that fire spread to adjoining buildings is controlled as far as reasonably practicable.

2 The external walls of the building shall adequately resist the spread of fire over the walls and for one building to another, having regarded to the height, use and position of the building. (1) Despite the requirement to reduce the risk as far as reasonably practicable, it is recognised that the risk cannot be eliminated entirely. In the UK, guidance on how to achieve this functional requirement can be found in the principal guidance documents for each country (e.g. Approved Document B Volume 2 (AD-B) in England (1) and The Technical Handbook Non-domestic in Scotland (2) etc.). These are the principal documents which will be referred to when reference is made to UK guidance. The aforementioned guidance documents have their own simple space separation guidance which, if met, will demonstrate compliance with the Building Regulations. (1-2) However, these simple methods are limited to particular building types and size and, as such, users are directed towards BR187, (10) which is widely used for assessing space separation in the UK. International Guidance In North America NFPA80A (4) is the principal design guide for assessing space separation. This document follows the same principle as the UK guidance, whereby a standard of protection is recommended as complete protection is impossible. Sprinklers Sprinklers can offer several benefits when it comes to external fire spread, however, the benefit of sprinklers fall outside the scope of this study. Fire Resistance of External Walls and Unprotected Areas In the UK unprotected areas are those which do not achieve the minimum period of fire resistance required for the external walls of the building. (1-3) In accordance with Table A1 of AD-B, (1) buildings more than 1m from the relevant boundary will require integrity fire resistance matching that of the building as a whole from the inside only. However, where the building is less than 1m from the boundary or where the space separation requirements cannot be met, the external walls are required to be provided with fire resistance from both sides, which match the fire resistance of the building as a (1, 3) whole (excluding the fame in some instance). In Scotland this requirement is very different. The external walls are only required to be provided with either a 60 minute or 30 minute fire resistance period. (2) Furthermore, unlike the rest of the UK guidance, where the building is less than 1m from the boundary the fire resistance is only required from the inside-out. (2) Global Guidance The International Building Code (IBC) (11) requires external walls to be fire rated from both sides, however, once the building is more than 3.05m from the assessment boundary this fire rating is only required from the insideout. (11) Relevant Boundaries Nationally space separation is measured to a relevant boundary, this is usually the site boundary or a notional boundary. In terms of space separation the site boundary is the legal site boundary around the site, however, where a building overlooks a public access way or public right of way, river, stream, canal, lake, pond, common land or a public open space, the space separation should be measured to the centre line of this feature, and this is used as the relevant boundary. (2)

3 A notional boundary is assumed where buildings exist on the same site which are under different ownership/management or if there is considered to be sufficient risk that additional protection is required, for example, residential buildings or assembly and recreation buildings. (1) In Scotland, the idea of a notional boundary is taken one step further, as it is a requirement to consider all buildings under the same occupation on this basis and, opposing compartments of the same building. (2) It is noted that fire spread between compartments is only considered elsewhere in the UK in health care buildings. (12) Buildings Less than 1m from the Relevant Boundary Buildings erected less than 1m from the relevant boundary are required to be provided with fire resisting construction from both sides (in Scotland from inside out only), however, small openings in the façade are permitted provided they comply with Figure 1. (1) Figure 1 Unprotected areas which may be disregarded (1) The Technical Handbook (2) also permits the external walls of a protected corridor to be discounted and, goes one step further, restricting the provision of the above between 1m to 0.5m from the boundary. Below 0.5m from the boundary, only such elements as solum vents or cavity vents, wallhead fascia, soffit or barge boards are permitted. (2) BR187 The relevant boundaries referred to in BR187 (10) were the site boundary and the notional boundary and users are directed towards the appropriate guidance document for definition of terms. (10) The guidance later infers that the separation distance requirement calculated using the documents enclosing rectangle method will be measured to the relevant boundary as defined in the original guidance documents. (10) This is taken from the original research which stipulated that the assessment is based on a mirror image building on the opposing side of the boundary, (1) the results of the assessment would therefore provide half of the overall required separation distance. (13) This has been done for many reasons, one of which is to reduce land waste. (13) However, in 2014 BR187 (5) was updated and, as part of the revision, the document provides a critical review of some of the assumptions in the original report. In this case, the document explicitly describes the requirements of the calculation in terms of the relevant boundary as demonstrated in Figure 2.

4 Figure 2 BR187 space separation requirements (5) As shown above where a large building overlooks a small building the separation distance should be increased such that the large building is not at risk of igniting the smaller building. This is contradictory to the guidance provided in the principal UK guidance documents. (1, 2) However, in accordance with Chitty (14) this should be considered but there is no requirement to deviate from the original application of the methods. UK Radiation Analysis The fundamentals of all space separation assessments within the UK have their roots in Fire Technical Paper No. 5 Heat radiation from fires and building separation. (13) This paper sets out the assumptions which are required to be made in order to carry out any space separation assessment. Stefan-Boltzmann Law The radiation calculation derived separately by Joseph Stephan and Ludwig Boltzmann forms the basis of the more complex external fire spread assessments. (15) 4 I = εσt Where: I = Intensity of radiation (kwm -2 ) ɛ = Emissivity of the radiator σ = Stefan-Boltzmann constant (kwm -2 K -4 ) T = Absolute temperature (K) This law applies when measuring the radiation from a blackbody (i.e. a perfect radiator). However, in the built environment, generally there is no such thing as a perfect radiator and, as such, emissivity has been introduced to the formula. Notwithstanding, for conservative reasons emissivity is generally taken to have a value of (or close to) unity (1). (13) Ignition The external fire spread assessment is generally based on the pilot ignition (i.e. ignition with an accompanying ignition source) of oven-dried wood (which in the UK is considered to occur around 12.6kWm -2(7, 10, 13, 16) ), unprotected by paint. (13) It is noted within Fire Technical Paper No. 5, (13) that this is an onerous assumption, particularly in the UK where the building façade can be considered to retain moisture for most of the year, which in turn would raise the minimum radiation intensity required for ignition. Fire Intensity One of the fundamental assumptions of the radiation assessment is that the fire has reached flashover. (13) However, this, and the resulting radiation intensity of a fire is influenced by many factors. In an effort to simplify the assessment, these factors have been restricted to fire load density and ventilation.

5 The original data for compartment fire intensity was based on experimental assessments of large (i.e. 3m by 3m) and small (i.e. 0.3m by 1m) scale brick structures, with a fire loading bench mark of 25kgm -2 (wood equivalent). (5) It was established that buildings with a high fire load density over 25kgm -2 and a lack of adequate ventilation would reach a maximum temperature of 1040 o C, and those with a low fire load density (i.e. below 25kgm -2 ) provided with sufficient ventilation to sustain the fire, would reach a maximum temperature of 830 o C. (5) This formed the basis upon which the UK guidance is based and, correlates to radiation intensities of 168kWm -2 and 84kWm -2 (1, 3, 5) respectively. Law (13) adopted the same principles for her work and divided compartment fires into two distinct groups; (i) fires with restricted ventilation, and (ii) fully-ventilated fires. However, unlike the other guidance documents which direct users towards the resulting fire intensity, Law s (13) approach favoured the fire temperatures, which she rounded to 1100 o C for ventilation restricted fires and 800 o C for fully-ventilated fires, based on the experimental results shown in Figure 3. (5) Flame Projection Figure 3 Maximum fire temperature verses the enclosure size (13) Generally in the UK the temperatures generated by flames projecting out from an opening can be disregarded from the space separation assessment. (13) It is recognised that the radiation intensity from a single surface is greater than that of the external flames and, furthermore, when considering small windows (i.e. 1m to 2m in height) the calculation process provided in BR187 (5) would suggest an almost insignificant value for the emissivity of a projected flame such that the radiation from the flame could be discounted from the space separation assessment. METHODS OF ANALYSIS Introduction In the UK there are a variety of approaches which can be adopted to determine the risk of fire spread to adjacent buildings. Generally users are directed either towards the methods (1-3) provided in the guidance documents or BR187 (1991). (10) Internationally different approaches are adopted. In North America, in particular the United States of America (USA) and Canada, individual states/provinces have their own approaches and Building Codes. However, the National Fire Protection Association (NFPA) provides an external fire spread assessment which is generally accepted in the USA and Canada. (4) The approach provided in NFPA80A has its roots in the St. Lawrence Burns experiments and work undertaken at the Fire Research Station (4) fire tests which also form the basis of BR187. As such, it is anticipated that this assessment will provide comparable results to those obtained from the BR187 enclosing rectangle approach. AD-B Two methods are proposed within AD-B, however, they have limited applicability. Due to the limited scope of this study only the method known as Method 1 has been reviewed. Method 1 (as shown in Figure 4 below) applies only to residential buildings other than institutional buildings which are more than 1m from the relevant boundary. Furthermore, the building should be not more than 24m in length and not more than three storeys above ground level.

6 Figure 4 AD-B Method 1 (1) There is no recorded reason why at 1m distance the maximum total area of unprotected area is 5.6m 2 when the rest of the values increase by increments of 6 times the separation distance. (16) Chitty (16) presumes that this value was born from experimental data and from the Fire Research Station in 1968 which included a test with a window of that size. Technical Handbook The Technical Handbook provides a variation of AD-B s Method 1. (2) However, in 2005 when the first Technical Handbook (17) was introduced this method was modified such that it applied to low risk factories, residential, assembly and office buildings which are not more than 9m in height and 24m in length. (17) Furthermore, the Technical Handbook has removed reference to the 5.6m 2 opening and now states that the unprotected area is permitted be six times the separation distance and, if the separation distance is 6m or more, (2, 17) then the amount of unprotected façade is unlimited. BR187 Enclosing Rectangle Method The enclosing rectangle method is one of the most user friendly methods of calculating the risk of external fire spread. As such, it is widely used throughout the UK and is generally the go to method for most UK fire engineers. The fundamental calculation method is based on the radiation calculation described above, and assesses radiating panels which increase vertically by multiples of 3m up to 27m (30m in the 2014 edition) and horizontally by multiples of 3 up to 30m before the rectangles increase by 10m then 20m at a time. The amount of unprotected façade is given as a percentage, however, the assessment is only valid for unprotected areas over 20% non-fire rated. As the background calculations have already been completed, the engineer is required to read values off the relevant tables and interpolate between the results were required. BRE s Alternative Method During work undertaken with the Forestry Commission, Scotland, BRE concluded that the simple method from the Technical Handbook, did not compare well with other calculation methods and proposed an alternative calculation. (16) U = 2.5A Where: U = The maximum unprotected area (m 2 ) A = The minimum boundary distance (m) NFPA 2

7 NFPA 5000 NFPA 5000 (18) gives no reasoning or background justifications for its assessment method although, it would appear to adopt the same notional boundary (or imaginary line ) principle as BR187, (5, 10) whereby the assessment is measured to a notional boundary. Similar to the BR187 enclosing rectangle method, NFPA 5000 s method is table based. Users can work out the required separation distance based on the percentage of unprotected façade for selected areas of exposure ranging from a 9.29m 2 area of exposure to 1,858m 2. NFPA 80A The method is based on the same research as BR187 but, the decision has been made to adopt a pilot ignition of cellulosic materials at 12.5kWm -2. (4) This is a more conservative figure however, it is noted that the original figure used was 0.3calcm -2 s -1. (13) The discrepancy in rounding is considered to be marginal and, as such, both values are considered to be acceptable for use in practice. Again the method is table based, Table is used for buildings of equal or greater height and Table is applied to buildings of lesser height. However, before users can apply Table the separation of the taller building should be determined using Table Due to the limited scope of this study, only the assessment for buildings of equal or greater height has been assessed. The NFPA method takes into account more aspects of fire severity than the BR187 (5, 10) methods, users are required to establish the severity of the fire based on fire load and the interior walls and ceiling finishes. The factor which result in the most onerous assessment should be applied. Fire severity group Table 1 Fire severities adapted from NFPA80A (4) Fire loading of floor area (kgm -2 ) Average flame spread index of interior wall and ceiling finishes Light Moderate Severe Similar to the UK guidance (1-3) methods, NFPA80A (4) considers all openings that may contribute to the spread of fire within the enclosing rectangle. However, additional consideration is given to building façades which have a fire resistance period less than 20 minutes or are more than 20 minutes but less than the estimated fire duration. Considering the former, the building façade should be considered as 100% open and, considering the latter, the façade should be considered as 75% open. Using this information, a guide number is selected form Table Table 2 Adapted from Table of NFPA80A (4) Severity Width/Height or Height/Width (Whichever is greater) Percentage of Openings Guide Number (Multiply by lesser dimension, and add 1.5m to obtain building to building separation Light Moderate Severe As noted in line two of the Table 2 the equation for separation distance is:

8 D = GS +1.5 Where: D = The separation distance (m) G = The guide number S = The lesser dimension of W or H (m) The 1.5m that is added to the equation accounts for flame projection out of the openings. (4) IBC The IBC s (11) method is another tabulated method which determines the required separation distance based on the unprotected area per storey giving consideration to the level of protection afforded to the buildings and the openings. Similar to NFPA5000, (18) no reasoning or justification is provided to account for these values and no other space separation assessment is provided. COMPARISON OF RESULTS A series of comparative assessments have been undertaken using some of the assessment methods discussed in above. The intention of these tests is to compare the results of the BR187 (5, 10) enclosing rectangle method to other methods. Low Fire Load Buildings In UK guidance buildings with a low fire load (i.e. less than 25kgm -2 ) are considered to be offices, residential and assembly and recreation buildings. (13) Methods marked with a star (*) assess the full separation distance between buildings. To provide a comparative assessment the results of these methods have been halved. Test 1 The building façade is considered to be fire rated and there is no provision of compartment walls or floors. Each window is 2.25m 2 (i.e. 1.5m x 1.5m) and the door is 5m 2 (i.e. 2m x 2.5m) and, therefore, the total unprotected area is 34.25m 2. NFPA 80A* NFPA5000 IBC BR187 (ENCLOSING RECTANGLE) BRE ALTERNATIVE METHOD SIMPLE GEOMETRY METHOD (SCOTLAND) Figure 5 Test 1 geometry Figure 6 Separation distance to the notional boundary or mid-point

9 The BR187 (5, 10) enclosing rectangle method requires 2.7m to the notional boundary, this is the shortest separation distance required. The IBC method and the simple geometry method, all required over double the distance of the BR187 (5, 10) enclosing rectangle method. Test 2 The same test conditions as Test 1 have been used for Test 2, with an alternative opening arrangement. As such, the building façade is considered to be fire rated and there is no provision of compartment walls or floors. Each window is 2.25m 2 (i.e. 1.5m x 1.5m) and the door is 5m 2 (i.e. 2m x 2.5m) and, as such, the total unprotected area is 34.25m 2. Figure 7 Test 2 geometry Figure 8 Separation distance (m) to the notional boundary or mid-point Due to the more compact distribution of openings (i.e. square as opposed to rectangular) the required separation distance has generally been increased using the more complex methods. These results are considered to conform well with the background physics behind the methods. The UK assessments which do not require additional distance (i.e. the simple geometry method (2) and BRE s alternative method (16) ) are more simplistic in nature and could be more readily applied in where a rough assessment is required. As the results of these methods are more onerous than the complex methods, their use for a rough assessment is considered reasonable, provided a detailed assessment is carried out at a later date. The NFPA5000 (18) and IBC (11) assessment do not require any additional space separation with the altered opening arrangement. This is considered to be due to the simplistic nature of the assessment method. The NFPA5000 (18) method requires the shortest separation distance to the notional boundary, however, the method has been taken to its maximum distance and, therefore, is not considered to provide an accurate reflection of the space separation required. Of the complex methods, the BR187 (5, 10) enclosing rectangle assessment requires the shortest separation to the notional boundary (i.e. 3.1m) and overall, discounting the mirror image principle. Test 3 NFPA 80A* 4.15 NFPA IBC 6.1 BR187 (ENCLOSING RECTANGLE) 3.1 BRE ALTERNATIVE METHOD 3.7 SIMPLE GEOMETRY METHOD (SCOTLAND) The purpose of this test is to assess the space separation requirements when considering a large non-fire rated construction. The building is 27m in height and 50m in length and, therefore, the unprotected area 1350m 2.

10 Figure 9 Test 3 geometry Figure 10 Separation distance (m) to the notional boundary or mid-point The NFPA5000 (18) and IBC (11) methods requires the shortest separation to the notional boundary, however, these methods has been taken to their maximum limits and, therefore, are not considered to provide an accurate reflection of the space separation required. Excluding these methods, the BR187 (5, 10) enclosing rectangle method requires 24m to the notional boundary. Only the BRE alternative method (16) requires less separation (i.e m). In previous test this method has been more onerous and, therefore, its use for a rough assessment was considered reasonable, but, despite the shorter (5, 10) distance the results still bare a good correlation to the BR187 enclosing rectangle assessment. High Fire Load Buildings In UK guidance, buildings with a high fire load (i.e. over than 25kgm -2 ) are considered to be shops and commercial units, most industrial and storage facilities. (13) Note that the BRE alternative method (16) has been modified to account for a high fire load assessment. The proposed new formula is provided below: U A =. 1.1 Test 1 NFPA 80A* NFPA5000 IBC BR187 (ENCLOSING RECTANGLE) BRE ALTERNATIVE METHOD The building façade is considered to be fire rated and there is no provision of compartment walls or floors and, as such, the windows are considered to be the only unprotected areas. On this basis, the total unprotected area is 96m Figure 11 Test 1 geometry

11 Figure 12 Separation distance (m) to the notional boundary or mid-point The BR187 (5, 10) enclosing rectangle method requires 9.4m to the notional boundary. The modified BRE alternative method requires slightly less separation (i.e. 9.34m). Given this good correlation, it is considered that the use of the modified alternative method for a rough assessment is reasonable. All three NFPA80A (4) fire severity levels have been assessed as the fire loading in the NFPA is considered light below 34kgm -2, however, the moderate fire load provided the best correlation to that of the UK guidance. Excluding that the assessment should be measured to the opposing building and not a notional boundary, the NFPA80A (4) and radiation assessment require a larger space separation than that of the UK guidance, this is considered to be due to the lower pilot ignition point. Test 2 NFPA 80A* SEVERE NFPA 80A* MODERATE NFPA 80A* LIGHT NFPA5000 IBC BR187 (ENCLOSING RECTANGLE) BRE ALTERNATIVE METHOD This test is considered to represent a warehouse/industrial process facility with a 100% unprotected façade Figure 13 Test 2 geometry Figure 14 Separation distance (m) to the notional boundary or mid-point The separation distance required by the NFPA5000 (18) and IBC (11) assessments is required to be taken to the maximum distance recommended in each document (i.e. over 9.14m). Due to the simplistic nature of the methods, the results are not considered to accurately reflect the space separation requirements of the building. CONCLUSIONS UK Methods NFPA 80A* SEVERE NFPA 80A* MODERATE NFPA 80A* LIGHT NFPA5000 IBC BR187 (ENCLOSING RECTANGLE) BRE ALTERNATIVE METHOD The BR187 (10) enclosing rectangle method generally requires a shorter separation to the notional boundary than any of the other UK methods. On this basis, the less complex methods are still considered to be fit for purpose, however, as these methods required additional separation, it is believed that these methods will be applied less in the future. The BRE s alternative method (16) and the modified version proposed in this study provide a good rough assessment. This method could be readily used in meetings and in the absence of adequate information. The results of this assessment are generally more onerous than that of the BR187 (5, 10) enclosing rectangle method but as the compartment size increases the results become less onerous

12 Global Methods Both the IBC (11) and NFPA5000, (18) use a mirror image approach when assessing the external fire spread requirements. However, the results of these assessments bare no correlation to any of the UK or research methods. Furthermore, it is noted that the results of the NFPA80A (4) method do not correlate with the IBC (11) or NFPA5000. (18) Given that there is no clear evidence provided as to how these methods were derived, it is recommended they are not used in the UK. The NFPA80A (4) method would satisfy the requirements of the UK Building Regulations as the method is more onerous than the UK methods, given the lower radiant intensity assumed for pilot ignition and the addition of flame projection. However, this method assesses the full separation distance between buildings, something which is not undertaken in the UK assessments. If this method is proposed for use in the UK, it should be used as intended (i.e. separation should be measured from building façade to building façade). To avoid legal issues, it is recommended that this method is only used for buildings on the same site. REFERENCES 1) Communities and Local Government Approved Document B (Fire Safety) Volume 2 Buildings other than dwellinghouses, NBS (2013), London, EC2P 2EA 2) The Scottish Government Technical Handbook Non-Domestic Fire, BSD (2013) Available online from: 3) British Standards Institute BS 9999: 2008 Code of practice for fire safety in the design, management and use of buildings, BSI (2008), London, ISBN ) National Fire Protection Association NFPA80A 2012, Recommended Practice for Protection of Buildings from Exterior Fire Exposure, NFPA (2011), 1 Batterymarch Park, PO Box 9101, Quincy, MA ) Building Research Establishment BR187 External fire Spread: building separation and boundary distances, Ed. 2, ISBN , BRE (2014), Herts, WD24 9XX 6) McLean, D. The Scotsman, Lost Edinburgh: The Great Fire of Edinburgh, (2013) Available from: 7) Chicago History Society The Great Chicago Fire, Chicago History Society, (1999) Available from: 8) Past Tense Vancouver History The Great Vancouver Fire, (2009) Available from: 9) Bukowski, R. W. The Role of Standards in a Performance-based Building Regulatory System, NIST Building Fire Research Laboratory (no date given), Available from: 10) Building Research Establishment BR187 External fire Spread: building separation and boundary distances, ISBN , BRE (1991), Herts, WD6 2BL 11) International Code Council 2015 International Codes International Building Code, ISBN: , International Code Council Inc. (2014), 4051 West Flossmoor Road, Country Club Hills, IL 12) Department of Health Health Technical Memorandum 05-02: Firecode, Guidance in support of functional provisions (Fire safety in the design of healthcare premises), The Department of Health (2014), Available from: 13) Law M. Heat Radiation from Fires and Building Separation, Fire Research Technical Paper 5, HMSO (1963) 14) Chitty, R. (2014), Personal Communication (1/06/2015) 15) Encyclopaedia Britannica Stefan-Boltzmann Law, (2014) Available from: 16) Chitty, R. External fire spread: Building separation and cladding, BRE Global (2009), BRE Fire and Security, Bucknalls Lane Watford Herts WD25 9XX 17) The Scottish Government Technical Handbook Non-Domestic Fire, BSD (2005) Available online from: 18) National Fire Protection Association NFPA , Building Construction and Safety Code, ISBN: , NFPA (2014), 1 Batterymarch Park, PO Box 9101, Quincy, MA 02169