Sainsbury (client) plans to build up a warehouse along with offices of two storeyes. The building will be constructed of pre-engineered steel structure. For example, Fig 1 shows a typical structure of the pre-engineered steel building. Fig 2 displays the inside of the building under construction
Since the 1960's there has been a rapid increase in the construction of large single storey storage warehouse facilities in the United Kingdom (UK). One of the most notable of these was the 40,000 m2 unsprinklered military storage facility constructed by the Ministry of Defence (MOD) at Donnington. In 1983 this storage facility was completely destroyed by fire despite the presence of an on-site fire service. The fire at Donnington and other unsprinklered warehouses circa 1980 raised concerns in the UK and United States (US). The major concern in unsprinklered warehouses is rapid conflagration, especially if the following features are present:
• 12-30m internal building heights
• High racked storage
• Highly combustible and flammable stored commodities
• Combustible wall and ceiling linings,
• Poor fire safety management
Regardless of the above concerns, recent research undertaken by the Building Research Establishment (BRE) into 'sprinkler installation trends and fire statistics for warehouse buildings' has shown that occupants generally make their escape from fires in unsprinklered warehouses un-hurt.
The likely reasons for this are:
1. The large internal volume can act as a reservoir for combustible products and hence will increase the available safe escape time (ASET)
2. An occupant's awake and familiarity state within the building reduces the required safe escape time (RSET) on this basis, BRE's research stated that 'it could be difficult to justify the provision of sprinklers solely on safety grounds, indicating that the provision of sprinklers would have a negligible impact for life safety purposes.
Since the 1980s there have been several full scale sprinklered fire tests on warehouse commodities (stored to various heights). These tests have been performed by BRE, Underwriters Laboratory (UL), Factory Mutual (FMGlobal) and the National Institute for Standards and Technology (NIST). Such research has assisted in the development of BS5306-2, BS EN12845:2004 and bespoke fire safety engineering guidance such as 'The Design of Smoke Ventilation for Single Storey Industrial Buildings, including those with mezzanine floors, and high racked storage warehouses' by the Smoke Ventilation Association (SVA). Where property protection sprinkler systems are installed with additional enhanced features, significant improvements upon the typical 80% reliability can be achieved. Enhancements such as locking valves in the open position to prevent manual shut-off, fast response sprinkler heads and secondary pumps can all ensure that an enhanced property protection sprinkler system offers an operational reliability of in excess of 90%, which is the equivalent reliability for a system designed to a life safety standard.
The purpose of this Outline Fire Strategy Report is to facilitate preliminary discussions with the Approving Authorities and to outline the approach that will be taken to demonstrate compliance with Part B (Fire Safety) of Schedule 1 to the Building Regulations (2000-2006). This report will provide a brief summary on the all aspects of fire safety which correlate with the corresponding Building Regulations. However it will focus primarily issues pertaining to the Fire Detection and Alarm system and linking factors, such as evacuation times and fire growth. The overarching areas within the project incorporate:
Determine metal building frames and column profiles;
Choose the materials for the walls and roof;
Lay the floors plan out;
Fire sprinkler system and fire-fighting water pipe;
Fire detection and alarm system;
Smoke control system;
Architectural design for egress means;
Structural fire resistance calculations;
Although the provisions for the fire safety aspects noted above are covered under separate regulations, they are interlinked. Therefore, these fire safety aspects will be reviewed holistically in this strategy in order to ensure the requirements of the Building Regulations are achieved in the proposed works. The facility will incorporate approximately 3,500 sq m of warehousing space and a two storey office block. The facility will be typical portal fame construction clad in composite panels with the portal haunch being at approximately 16m above adjacent ground level and the apex of portals being at approximately 18m above adjacent ground level. The two storey office block is compartmented from the remainder of the building in fire resisting construction effectively creating two separated parts. The warehouse compartment is single storey.
Automated Garment Racking System
The automated garment racking system will be spread over five levels (i.e. ground and four upper access gantries/walkway levels). A common feature of automated hanging garment storage systems is that there are no discernible floors. Access to the automated garment racking system will be via elevated walkways and open stairs. A typical arrangement is shown in figure 3
The existing facility is fitted out with high racking bays incorporating in-rack sprinklers. It is intended to alter the existing high racking bay arrangement to incorporate a more efficient system to meet with specific operational requirements. The existing sprinkler system will be amended/extended to ensure that all high rack storage bays are adequately protected with in-rack sprinklers.
Regarding fire safety, the proposed automated hanging garment racking system and associated works must comply with The Building Regulations 2000 (England and Wales) but in existing areas only to the extent where no new or greater contravention of any regulation is created. As a general principle the recommendations of Approved Document B (ADB) Volume 2 - Buildings Other Than Dwelling Houses can to be applied to most buildings. Although, in respect of certain buildings and structures, in particular those purpose-designed for automated storage plant, the recommendations set out in ADB might be either inappropriate or unreasonably restrictive.
A common feature of automated hanging garment storage systems is that there are no discernible floors and no fire separation provided between deck levels. Automated hanging garment delivery, picking or transportation systems are generally accessed via elevated walkways, stairs and ramps. The Building Regulations are written in a functional manner and rather than follow the prescriptive guidance of a particular code of practice, a fire engineered approach can be taken to demonstrate compliance with Statutory requirements where deemed appropriate. In this regard ADB states the following: 'Fire safety engineering can provide an alternative approach to fire safety. It may be the only practical way to achieve a satisfactory standard of fire safety in some large and complex buildings and in buildings containing different uses'
ADB: Fire Safety Engineering (0.3) Therefore, where the proposed development does not fully comply with the recommendations of ADB m it is intended to incorporate the latest guidance available as part of an alternative fire safety engineering approach. This approach will be based on the recommendations of BS7974 with the overarching aim of achieving the optimum design solution and Building Regulations approvals.
THE REGULATORY REFORM (FIRE SAFETY) ORDER 2005
When the premises become a workplace, the 'responsible person', i.e. employer, must ensure that any duty imposed by The Regulatory Reform (Fire Safety) Order 2005, England Wales, specifically articles 8 to 22 or by regulations made under article 24 is complied with. Generally, compliance can be achieved with robust maintenance, staff training and housekeeping regimes and by undertaking regular fire risk assessments (FRA). The local Fire and Rescue Service will be responsible for enforcing these Regulations.
Purpose groups can apply to the whole building or to a compartment in the building. ADB recommends that different uses within a building should be assessed as separate purpose groups except when certain criteria are met (i.e. classified as ancillary if less than1/5 of the compartment floor area). The design of the existing building incorporates a significant amount of single storey open spatial planning, with the warehousing areas being completely separated from the main office block by fire rated construction. Therefore, when account is taken of proposed material alterations and the definitions noted in Appendix D of ADB the proposed warehouse compartment is classified as Purpose Group 7(a) Storage and Other Non-Residential. The existing multi-storey office compartment is classified as Purpose Group 3, Office.
In tandem with the application of recommendations in ADB, a fire safety engineered approach has been used where considered appropriate for aspects of the building's design. This approach has used recommendations from the following guidance documents to achieve a robust fire strategy:
• BS7974 Application of fire safety engineering principles to the design of buildings (BSI 2003)
• BS9999 Code of practice for fire safety in the design, management and use of buildings (BSI 2008)
• BS5588 Fire Precautions in the design, construction and use of buildings; Part 5 Access and facilities for fire-fighting (BSI 2004)
• BS5588 Fire Precautions in the design, construction and use of buildings; Part 8 Code of practice for means of escape for disabled people (BSI 1991)
BS7974 is specifically noted in standard fire safety guidance material as providing further guidance on fire safety engineering principles. Figure 4 presents the basic fire engineering design process suggested in BS7974.
BS7974 states the following regarding the use of fire safety engineering in buildings that are otherwise compliant with standard practice:
'In many projects it is likely that the provisions of existing codes of practice and other guidance will be largely followed and that fire engineering techniques will not be necessary (or may be used only to justify limited departures from the codes). At its simplest the QDR team may therefore, define the acceptance criteria in terms of compliance with existing code recommendations. The acceptability of a particular design may be evaluated by means of a comparison. The level of safety provided by alternative fire safety strategies can be compared with that achieved by the well-established codes. This approach involves deterministic and/or probabilistic techniques and requires less extensive analysis than a full study. The objective of a comparative study is to demonstrate that the building, as designed, presents no greater risk to the occupants than a similar type of building designed in accordance with a well-established code. Most existing codes and guides allow elements of trade-off and/or alternative measures. Examples can be found for fire resistance, compartment sizes and building separation In a comparative study these examples may be applicable without the need for detailed analysis.'
The evacuation strategy for the proposed buildings will involve simultaneously evacuating all areas on activation of the fire alarm system. As part of the Management and Health and Safety procedures the management team shall have prepared and implement procedures for the safe evacuation of occupants from the building.
Required Safe Egress Time and Available Safe Egress Time
In performance based fire safety engineering approaches it is fundamental that the evaluation of the complex interactions between fire, occupants and buildings for means of escape considers the following:
• Required Safe Egress Time (RSET)
time to detection
warning time (alarm)
response time (occupant pre-movement)
• Available Safe Egress Time (ASET)
Time to untenable conditions governed by fire growth, smoke production and size of room. A simple diagrammatic representation of the relationship between ASET and RSET is shown in figure 5
Method of Analysis
To ensure that the proposed works meet with the functional requirements of the Building Regulations regarding means of escape a comparative analysis as discussed in Section 4.3 has been carried out.
The recommendations of BS7974-6 have been used to determine the Required Safe Egress Time (RSET) for the proposed and code compliant layouts.
RSET can be calculated using the following formula:
trset Δ = t det Δ + t alarm Δ + t pre Δ +
t travel Δ
t rset Δ = Total time required for escape (secs)
t det Δ
= Time to detection (secs)
t alarm Δ
= Time to alarm (secs)
t pre Δ
= Pre-movement time (secs)
t travel Δ = Travel time (secs)
The occupancy noted in Table 1 illustrates that for the worst case shift, the warehouse will be sparsely occupied in regard to floor area and final exit provision.
Research on Evacuation Times
Standard fire safety recommendations for means of escape relate to travel time, i.e. the time taken to travel to and through exits to a place of relative safety. However, it is widely acknowledged that the time taken for occupants to start to evacuate can be considerably longer than the time taken to actually move out of the building. Significant research has been carried out in recent decades on occupant behaviour during building evacuations. Fires and human behaviour by Canter in 1990 and Safety and the built environment by Sime in 1988 were significant contributions. Through such research it is now widely acknowledged that occupants' behaviour varies appreciably depending on a wide range of factors such as social affiliation, role and responsibility, commitment to a present task, the presence of a focal point, etc. BS7974-6 Human Factors: Life safety strategies - occupant evacuation, behaviour and condition is the latest guidance available in relation to human behaviour in evacuations.
Recognition and Response Times
How occupants respond to a fire will depend initially on the cue they receive. The components of occupants' evacuation time are recognition time, response time and travel time. Both recognition and response times are known as pre-movement time. The former represents the time taken to appreciate that the alarm is a fire alarm relating to the occupant (as opposed to a security alarm, car alarm, etc).
During this time occupants will continue with activities as they were prior to the activation of the alarm, e.g. staff members undertaking duties, etc. The latter represents the duration taken by occupants to begin to move towards an escape route, e.g. investigative behaviour, shutting down machinery, etc.
In warehouse buildings occupants are generally familiar with their surroundings and are trained in fire emergency evacuation protocols. In the proposed storage facility and associated offices, occupants should be familiar with the sounding of the fire alarm and as such would commence evacuation reasonably promptly. Where a reasonable standard of fire safety management is provided and a high standard of automatic fire detection is provided, BS7974-6 recommends that familiar occupants of a building can be expected to take between 1mins and 2mins before they start to evacuate. By comparison, occupants of an unfamiliar building with a complex internal layout and a more basic standard of fire alarm system could take up to 15mins before commencing an evacuation. Recent research would indicate that even in unfamiliar buildings, occupants prompted by familiar staff can commence their evacuation within the first 30secs.
Impact of Evacuation Management
To comply with ADB and other fire safety legislation a reasonable standard of management will need to be provided in the building. This will include regular fire drills and the appointment of fire wardens for various areas to encourage an efficient evacuation of the building in the event of a fire. Such protocol should assist in limiting the recognition and response times of occupants. However, the proposed strategic approach to fire safety adopted in this fire strategy will not place any greater onus on management of this building than would be considered reasonable for any other storage facility of this size. An extract of Table C.1 from BS7974-6 is partially replicated in Figure 6. This demonstrates the potential variation in pre-movement times of evacuees which is dependent upon on the level of management and fire alarm provision. The management of this building will adopt a proactive fire safety management approach to comply with standard fire safety regulations for the workplace.
A1-A2: Automatic fire detection and alarm throughout the building
A3: Automatic fire detection and alarm only in certain areas, or manual call points only throughout
Components of Available Safe Egress Time
The Available Safe Egress Time is the time from ignition until conditions on occupants' escape routes become untenable. The time taken for hazardous conditions to occur on the escape route will depend on three governing factors:
• the growth rate of the fire (and rate of smoke production)
• time to sprinkler activation and containment of fire size
• the compartment size/smoke reservoir
Determination of Design Occupancy in the Warehouse Compartment
To assess the likely required capacity of escape routes in speculative buildings, floor space factors are commonly used. When a designated bespoke occupancy is not known there are varying recommended floor space factors in current guidance documents that can be used. The proposed works are not for a speculative end user and Sainsbury's have a good understanding of necessary staffing levels to carry out their logistics operations based on current arrangements and their other warehousing operations across the UK.
Fire Growth Rate
During the evacuation period the main hazard to occupants is smoke. The amount of smoke produced is directly linked to fire size and fire growth rate. The most common method of determining fire growth rates for unsprinklered fires is through the use of 't-squared' fires. Figure 7 illustrates the heat release rates over time of the four common fire growth rates published by the National Fire Protection Association (NFPA). BS7974 acknowledges that storage warehouse commodities growing unchecked will generally follow an ultra-fast fire growth rate. This attracts a coefficient of g = 0.187kW/s2. Charters et al undertook a recent study to review the probabilistic aspect of fire growth in storage warehouses serving retail facilities. One of the main findings of the research was that fires in high-racked storage cash and carry warehouses were ultra-fast growing. This correlates with BS7974-1. An ultra-fast growth rate will be used for the comparative assessment of both the proposed and code compliant layouts.
Sprinklers are installed in buildings to suppress fire size by the distribution of water so as to decrease the heat release rate and pre-wet adjacent combustibles while controlling gas temperatures to avoid structural damage. The proposed warehouse compartment will be fitted with a comprehensive in-rack sprinkler system to ensure fire size is kept to an absolute minimum. Figure 8 below from BS7974-1 demonstrates sprinkler control and extinguishment.
Given the enhancements to the property protection sprinkler, the proposed model will be based on a sprinkler controlled fire scenario. Although, there is significant potential that the provision of in-rack sprinklers will fully extinguish a fire, for the purposes of this comparative assessment, it will be considered that the fire size on sprinkler activation in the proposed building will be controlled as per (b) in Figure 7 rather than (d) i.e. the fire being extinguished. This provides an additional factor of safety for the analysis.
There has been a number of real fire tests carried out on high rack storage arrangements by insurance companies. Tests carried out by Underwriters Laboratories Inc. concluded that fire size is unlikely to exceed 5MW where in-rack sprinklers have been provided. Smaller tests on single racks of clothing would indicate a typical peak heat release rate of 2.5MW for isolated arrangements. BRE 368 would also suggest a peak heat release rate of 2.5MW for sprinklered retail fires using fast response sprinkler heads.
Travel Distances at Ground Floor Level
The final exits from Ground Floor Level, independent of the Office Block, will be maintained as part of the proposed works. The existing exits are well distributed around the external façade thus ensuring that occupants can turn their back on a fire and move towards an alternative exit. Existing travel distances to final exits will remain are unlikely to exceed 80m.
When minimum fire protection measures are provided (i.e. manual fire alarm, 2m floor to ceiling height, etc), ADB suggests maximum travel distances in warehouse accommodation of 25m in a single direction of travel and 45m where two means of escape are provided. The proposed building contains a number of inherent fire safety features such as high ceilings, automatic fire detection, sprinklers and an open plan environment. Such features assist in prolonging the onset of hazardous conditions and increase the chance of a person becoming aware of a fire in the early stages of its development independent of the alarm being raised by others. Therefore, the basic recommendations set out in ADB are considered to be unduly restrictive for the proposed building.
A detailed assessment of allowable travel distances will be based on the recommendations of BS7974 and the results of this analysis will be presented in the detailed Fire Strategy Report (FSR). A detailed bespoke fire engineering analysis of allowable travel distances based on the recommendations of BS7974 and incorporating Computational Fluid Dynamics (CFD) modelling will be carried out when layouts have been finalised. The results of this analysis will be presented in the detailed FSR in due course. For the purposes of this OFSR and preliminary design we have carried out a preliminary two zone model analysis to gain an understanding of the likely results to be obtained from the detailed CFD modelling.
AUTOMATIC FIRE DETECTION AND ALARM
Time to Detection and Alarm for the Proposed Arrangement
It is proposed to provide a high standard of automatic fire detection and alarm system throughout the Warehousing compartment. The standard of detection and alarm in the building will affect the general time to warning for occupants and provide an enhanced level of fire safety through early warning and shorter evacuation times. Activation of the sprinkler system will also lead to the fire alarm sounding throughout the building. To ensure a conservative estimation of fire alarm activation our analysis has been based on an anticipated sprinkler activation time of 123sec
Time to Detection and Alarm for a Typical Code Compliant Arrangement
Many code compliant warehouse facilities with code compliant travel distances only require manual fire detection to meet with code recommendations. Therefore the time to detection would depend on the time taken for occupants in the room of fire origin to raise the alarm. However, should the room of fire origin be unoccupied then the time to alarm for people who may be in rooms remote from the fire would depend on them or other occupants becoming aware of cues of the fire scenario, e.g. sounds of burning material, smell of smoke or seeing smoke through a vision panel. Although BS7974-6 recommends a time to detection of at least 15mins for such scenarios, a reasonable worst case for a typical code compliant open plan warehouse could be 3 - 5mins.
The elevated walkways will be accessed by a relatively small number of transient staff who should be familiar with their surroundings and location of their escape routes. In compliance with health and safety legislation, fire drills will be carried out on a regular basis. Therefore, staff should be familiar with the fire alarm signal and the evacuation procedure. Current fire research would suggest that occupant pre-movement times in a managed facility such as this Sainsbury's Distribution Centre is between 30secs and 60secs after alarm. However, BS7974-6 Table C1 (see Figure 5) conservatively presents a 120secs occupant premovement time for this type of occupancy. In line with BS7974-6 recommendations the code compliant layout could be 900secs. However, the comparative analysis will consider a pre-movement time of 120secs for both layouts. This is an onerous consideration and as such adds a significant degree of conservatism to the analysis.
Extensive studies have been carried out on the travel speeds of people of various abilities. CIBSE Guide E Fire Engineering and BR459 Fire Safety Engineering 'A Reference Guide' both state that a travel speed of 1.2m/s is considered reasonable for able bodied people in low occupant density areas. This is based primarily on work by Pauls and Fruin. Personnel who will be accessing the proposed elevated walkways will be doing so to maintain/ensure the smooth operation of the automated system and will need to be reasonably mobile/physically fit to carry out such operations. Therefore, a travel speed of 1.2m/s is considered reasonable for this assessment. As previously noted travel distances on elevated walkways of the Automated Hanging Garment Racking System could be up to 100m when racking layouts are finalised. Based on a 1.2m/s travel speed it could take these occupants 84secs to reach a storey exit. For a typical compliant layout with a 45m travel distance it would take 37.5secs to reach a storey exit. For the above RSET components, Table 2 below presents the corresponding trset Δ .
The above analysis would indicate that occupants located on one of the elevated walkways could Conservatively take up to 327secs to evacuate from the upper walkway levels of the automated hanging garment racking system. The analysis also demonstrates the benefits of providing a good standard of automatic detection and an in-rack sprinkler system that is linked to the fire alarm system as such systems should ensure that the required safe egress time from the proposed building is kept
to a minimum. As can be seen from Table 2 above such systems should also ensure evacuation times from the proposed building would be comparable with many typical code compliant arrangements, if not better.
To meet with code recommendations a Type 'M' manual fire alarm system would be necessary in a standard warehouse building. However, the provision of an automatic fire detection and alarm system is an integral part of the overall fire safety package from a fire engineering perspective. The current fire alarm system will be extended to ensure the earliest possible detection of fire in the vicinity of the Automated Hanging Garment System areas. The system will provide high standard smoke detection coverage above and to the underside of elevated levels of the Automated Hanging Garment System level. The sprinkler system will also be linked to the fire alarm system so that in the event of a sprinkler head activating the alarm will sound throughout the building thus ensuring the earliest possible alert of the need to commence evacuation. The additional provision of smoke detection and linking the sprinkler system to the fire alarm is considered a significant additional life safety benefit which should substantially reduce the time to alarm and pre-movement time of occupants in the building. The Type 'L5' fire alarm system will be designed in compliance with BS5839-1.
Office Block Compartment
No changes are proposed for the Office Block Compartment. Means of escape from the Warehouse Compartment is independent of the Office Block. Therefore, no further analysis of this area is necessary as part of the proposed works.
Comprehensive fire safety management procedures should be developed in consultation with the Enforcing Authorities. A detailed Fire Safety Management Plan should be developed for the management of the proposed scheme to ensure the safety of all occupants is not comprised. This section only serves as a brief introduction to the evacuation procedures which will need to be written into the Fire Safety Management Plan. Further guidance can be found in BS5588-12 'Managing fire safety'.