The theory that underpins traditional prescriptive fire codes and building standards, including the current Fire safety: Approved Document B, has its origins in the early 20th century.
But better understanding of the history will show our evacuation strategies depend on assumptions particular to that context – and may succeed more by luck than judgement.
In the late 19th and early 20th centuries, thousands of people around the world were killed or seriously injured in horrific fires that broke out during theatre performances. The events listed in Table 1, for instance, account for some of the biggest losses of life in the history of building fires.
Table 1: Theatre fires from the late 19th and early 20th century, showing the large number of fatalities. Source: The Sandland-Taylor Consultancy
| Year | Theatre name and location | Number of deaths |
| 1871 | Shanghai Theatre, Shanghai, China | 900 (estimated) |
| 1872 | Tientsin Theatre, Tianjin, China | 600 |
| 1876 | Brooklyn Theatre, New York, USA | 278 |
| 1881 | Municipal Theatre, Nice, France | 150–200 (estimated) |
| 1881 | Ringtheater, Vienna, Austria | 448 |
| 1887 | Salle Favart, Paris, France | 70–100 (estimated) |
| 1887 | Theatre Royal, Exeter, UK | 200 |
| 1893 | Kamli Theatre, Kamli, Japan | 1,195 |
| 1903 | Iroquois Theater, Chicago, USA | 602 |
Table 1: Theatre fires from the late 19th and early 20th century, showing the large number of fatalities. Source: The Sandland-Taylor Consultancy
| Year | Theatre name and location | Number of deaths |
| 1871 | Shanghai Theatre, Shanghai, China | 900 (estimated) |
| 1872 | Tientsin Theatre, Tianjin, China | 600 |
| 1876 | Brooklyn Theatre, New York, USA | 278 |
| 1881 | Municipal Theatre, Nice, France | 150–200 (estimated) |
| 1881 | Ringtheater, Vienna, Austria | 448 |
| 1887 | Salle Favart, Paris, France | 70–100 (estimated) |
| 1887 | Theatre Royal, Exeter, UK | 200 |
| 1893 | Kamli Theatre, Kamli, Japan | 1,195 |
| 1903 | Iroquois Theater, Chicago, USA | 602 |
There was an acceptance that death was an inevitable consequence of a fire breaking out in a theatre where there were large numbers of people in attendance, which was a sad indictment of the times.
In 1876, Captain Sir Eyre Massey Shaw, the first chief officer of what would become the London Fire Brigade, published Fires in Theatres. He raised concerns about the frequency of such fires, the alarming number of deaths, and the presence of hazards including combustible drapes, curtains, scenery and large backdrops decorated with multiple layers of oil-based paints.
He also recorded his despair about the grave danger to which theatregoers exposed themselves, commenting that they went 'for amusement; but … with their eyes open, and … prepared to take the consequences, balancing the pleasure against the risk'.
He also bemoaned the lack of regulation and control of such buildings, but hoped 'that the casualties … may be considered of sufficient importance to attract the attention of authorities who are [e]ntrusted with the duty of looking after the safety of life in theatres, if any such authorities exist'.
A later fire in 1911 at the Edinburgh Empire Palace Theatre resulted in the death of nine actors and a musician, as well as destroying the entire stage area. Nevertheless, the safe evacuation of all 3,000 people from the auditorium was hailed as a success because of the timely intervention of the stage manager who, with his staff, had been carrying out regular fire drills.
In 1934, the Home Office published the Manual of safety requirements in theatres and other places of public entertainment. This referred to the Empire Palace Theatre fire, highlighting the fact that several thousand people had calmly evacuated the building in around two-and-a-half minutes, with no deaths or serious injury among them.
It was subsequently established that this isolated evacuation provided sufficient evidence to recommend a safe egress time of two-and-a-half minutes for places of public entertainment.
Post-War Building Studies No. 29: Fire Grading of Buildings, Parts II–IV, published in 1952 by the Ministry of Works for HM Stationery Office, emphasised the need to determine a suitable clearance time. This would be a basis for the number and width of escape routes and exits required for a given number of occupants in any building designed for public assembly.
The document cited the Empire Palace Theatre fire, noting: 'There was a large audience in this building when fire broke out on the stage, and the available evidence indicates that the time taken to clear the building was about 2½ minutes ...'
'Certain measurements [that] have been made over the time taken to empty theatres under normal conditions, suggest that if the auditorium is cleared in 2½ minutes there will be no serious risk of panic in the event of fire … It is proposed to adopt this figure as a basis for calculation.'
The document also recommended that a discharge rate of 40 persons per minute through a unit of exit width, 525mm, should be used in conjunction with the prescribed evacuation time.
Accordingly, one unit of exit width would allow 100 people to exit in the given time; that is two-and-a-half minutes at 40 persons a minute. Two units of exit width would allow 200 people to evacuate in that period, three would enable 300, and so on.
This figure is based on the average shoulder width of a person, and the narrowest width through which a single file of people could pass. Average shoulder width was determined as 21 inches, or roughly 525–530mm. As this is very narrow, the circumstances where such a small width is used are limited. Most codes have tended to recommend a minimum clear door width of 750–850mm.
It should be noted that Table 2.3 of Approved Document B – Volume 2: Buildings other than dwellings requires that anything less than two units of exit width – 1,050mm, two persons side by side – has to be taken as a single unit width, as it will only allow single-file evacuation.
Hence, the maximum number of people for 850mm is restricted to that of one unit width, and erring on the side of caution 750mm opening is restricted further. However, you will see from note 4 to Table 1 that a figure of 530mm is used.
These recommendations have been broadly adopted, with the odd modification, in many global fire codes and building standards over the past 70 years. Although the current version of Approved Document B – Volume 2 does not refer to discharge rates, units of exit width or a predetermined egress time, they are nonetheless embedded in the design philosophy that determined the document's figures for widths of escape routes and exits.
In Table 2.3 of the Approved Document B Volume 2, two units of exit width – 1,050mm – permit a discharge capacity of 220 persons, and for three units – at 1,500mm – the capacity is 310.
This approach is based on the following assumptions:
evacuation starts as soon as an alarm is raised
the rate of at which people move through each exit is constant throughout the two-and-a-half-minute evacuation period
evacuees make use of all available exits.
However, such an effective and efficient evacuation is rarely achieved when an actual fire breaks out. Normally, time is lost for various reasons before people begin to evacuate; delays in evacuation were for instance observed in all the fatal theatre fires listed in Table 1.
Most commonly, this is because even when people see, smell or hear fire they do not believe that they are in any danger, and tend to wait for further information or additional cues before they begin to evacuate. They are not usually aware of the number and location of all fire exits, either.
Consequently, they are far more likely to evacuate through an exit they know – in most cases, the route by which they entered the building.
This raises important questions. If the assumptions do not represent typical evacuation behaviour, was the analysis of the Edinburgh evacuation flawed? Is it possible that key information was either overlooked, or just missing?
Colin S. Todd's book A Comprehensive Guide to Fire Safety, published by BSI in 2008, reviews the investigation carried out by the then British Fire Prevention Committee.
The stage manager's prompt response significantly influenced the audience's behaviour and undoubtedly ensured the safe evacuation of the auditorium. As he was familiar with the performance of The Lion's Bride, during which the fire broke out on stage, he knew it was not part of the script and took decisive action.
He instructed the orchestra to play 'God Save the King', insisted the fire curtain be lowered, and asked the ushers to guide the audience safely to the fire exits. As was customary in those days, the national anthem marked the end of a performance, at which point people would stand in readiness to leave the auditorium.
The lowering of the recently installed fire curtain was crucial to the audience's survival, because it ensured that most of the heat and toxic smoke was contained in the stage area. While fire curtains were not common, there being no legal requirement for them, theatres had begun to install them on an ad-hoc basis following the Exeter blaze of 1887.
The curtain's effectiveness was evident given that, although the stage side was completely destroyed, there was very little damage in the auditorium. Good housekeeping ensured that all exits were accessible and free of obstructions. Ushers and staff had together run monthly fire drills, which meant that all employees were prepared in the event of a fire breaking out.
Failure to identify this context and understand how vital it was to the swift evacuation of the auditorium has resulted in incorrect assumptions being made subsequently. The analysis of the Empire Palace Theatre fire in Post-War Building Studies and its recommendations are flawed – and, given that the approach was then adopted in many prescriptive codes and standards, including Approved Document B, these too are put in doubt.
'Even when people see, smell or hear fire they do not believe that they are in any danger, and tend to wait for further information or additional cues before they begin to evacuate'
Fortunately, such guidance has tended to work in most cases. This is due to several factors that, unintended by those who prepared fire safety codes, allow significant margins of error.
For example, in many building fires the actual time available for safe egress is much greater than two-and-a-half-minutes, and the flow through exits is generally faster than the conservative rate used for all building types.
It should also be noted, though, that in certain fires the available time for safe egress may be considerably less. The devasting fire in 2003 at the Station nightclub in West Warwick, Rhode Island, USA – where 100 people were killed and a further 220 injured – is one example that I will cover in a future article.
More to the point, the adoption of a two-and-a-half-minute evacuation time after the Empire Palace Theatre fire has overlooked what influenced the behaviour of the audience in that case. Without the stage manager's intervention, the lowering of the fire safety curtain and the well-trained theatre staff, it is more than likely that this would have gone down in history as yet another disaster with considerable loss of life.
Altogether, this demonstrates how using information without context or consideration can lead to unsafe assumptions being adopted as best practice.
Simon Sandland-Taylor FRICS is director, founder and owner of the Sandland-Taylor Consultancy
Contact Simon: Email
Related competencies include: Fire safety, Legal/regulatory compliance, Property management
BUILT ENVIRONMENT JOURNAL
Paul Mallion FRICS and Gavril Postnikov 23 July 2024
BUILT ENVIRONMENT JOURNAL
Dr Richard Atkins 16 July 2024
BUILT ENVIRONMENT JOURNAL
Simon Davis MRICS 09 July 2024
