The role of the police at major outdoor concert events in the UK

March 12, 2011 1 comment

The role of the police at major outdoor concert events in the UK

Mick Upton

March 2011


The introduction of a Foundation Degree course in crowd safety management by BuckinghamshireNewUniversity attracted a large number of practitioners from abroad to study crowd management at organised social events. Much discussion on the course has centered on the role of the police at such events, particularly their involvement with major outdoor concert events held at a green field sites. It quickly became clear that in many countries the police take a primary role in the crowd safety decision making process. This paper considers the current attitude of the police in the UK toward responsibility for crowd safety at major concert events and the incidents that have shaped police attitudes.

Conflicting police opinion

The year 1995 was possibly the point at which the UK police service made it clear that, contrary to popular belief, they were not responsible for public safety at organised social events. Clarification was necessary because the police role at such events had become a cause of conflicting opinions within the British police service. The event that brought a public safety debate to a head was the 1995 Reading Festival. The Security Gazette that year reported that the Thames Valley Police were very concerned over a decision taken by the local council and the festival organisers not to include the presence of the police on site. The police apparently had specific concerns about crowd management and the handling of any emergency situations that may arise The view of the festival organisers was that they would happily co-operate with the police in allowing their specialist squads – such as the Drug Squad – full access to all areas of the site but they felt that responsibility for crowd management could be effectively undertaken by a private security company.

Thames Valley Police challenged this claim and they asked the sub committee of the council to impose a condition in the event license that required the chosen security company to arrange for a sufficient number of police officers on site. The condition was not imposed however as the council did not believe that they had the power to do so and a license for the event was granted. Thames Valley Police then applied for an urgent judicial review on the decision not to impose the condition. The matter was further complicated when the Reading Borough Council then decided to hold another sub – committee meeting without inviting the police. This time the sub committee apparently decided that they now realised that they did have the power to impose such a condition on a license but they were still not prepared to do so.

Following a High Court decision to support the local authority, the Leader of the council at that time, Mr. David Sutton, is reported as saying; “I am delighted that the High Court has seen fit to dismiss the case by ThamesValley Police. It has always been our aim to achieve a safe and successful Reading Rock Festival and there has never been any doubt in our minds about the on site security and public safety arrangements for this year’s event. They meet all of the Home Office own `pop` guidelines on public safety and welfare” (Security Gazette 1995). The event subsequently went ahead with a low police presence.

The same year that Thames Valley Police put forward a strong argument in court for a police presence at the Reading Festival other forces had a different policy. For example, the London Metropolitan Police published a policy document that year which stated that, “safety at events such as football matches and pop concerts, where members of the public are invited into private grounds; it is the responsibility of the organiser and/or the owner of the property” (Metropolitan Police 1995). It would appear from this statement that the country’s largest police force clearly did notregard public safety at football matches and concerts to be their responsibility.

The Metropolitan Police statement was not made in a direct attempt to contradict Thames Valley Police or to enter into their dispute with Reading Council. They had in fact made their position clear two years earlier in a seminar paper delivered to the Home Office Emergency Planning College, Easingwold, by Assistant Deputy Chief Constable Speed (Metropolitan Police). ADCC Speed argued that the police service was being unwillingly drawn into the role of crowd managers of events. Speed went on to add that the police had also been drawn into a role of guardians of public safety at unregulated events such as that which takes place each New Years Eve in London’s Trafalgar Square and he further argued that in such circumstances the police risk becoming involved in matters that they were not trained for, and therefore not qualified, to undertake (Speed 1993). A reference to being drawn into matters that they (the police) were not trained for did cause some surprise among people that assumed that as the police regularly dealt with crowds they were perhaps the only organisation that were qualified to effectively ensure crowd safety at organised social events.

To understand the claim made by ADCC Speed it is necessary to appreciate the important differences between crowd management and crowd control. The highly respected pedestrian planner John Fruin has argued that many people use the terms as if they were interchangeable titles for one subject when in fact they are two total separate functions. Fruin explained the two functions as follows:

Crowd management is defined as: “The systematic planning for, and supervision of, the orderly movement and assembly of people. Crowd management involves the assessment of the handling capabilities of a space prior to use. It includes evaluation of projected levels off occupancy, adequacy of means of ingress and egress, processing procedures such as ticket collection, and expected types of activities and grou8p behaviour”.

Crowd control is defined by Fruin to be: “The restriction or limitation of group behaviour”.

(Fruin 1993)

The definitions provided by Fruin are accepted by the United Kingdom Crowd Management Association (UKCMA) and the Centre for Crowd Management and Security Studies at BuckinghamshireNewUniversity. With a clear understanding of what crowd management entails it is possible to better understand the statement made by ADCC Speed. It is highly unlikely that the Thames Valley Police would have been trained to undertake a crowd management role according to Fruin`s definition.

A Metropolitan Police view that the police service is not trained to undertake a crowd management role also appears to have been shared by the West Midlands police in 1995. At a planning meeting that I attended for a free concert event in BirminghamCity centre that year, the police representative made it clear to all parties that they would not take an active part in the planning of the event because the safety of the public attending organised social events was not a police responsibility. The police view was later clarified in a letter written by Chief Superintendent Baker (West Midlands Police) to the local authority, which spelt out the position of the Birmingham City Police at the event as follows;

“In respect of the police role I think you are aware of our stance regarding public safety. This is a matter for the organisers of the event and not the police. There will be a patrolling presence on the day to afford both public reassurance and deal with any crime matters which may arise. Additionally there will be a contingency of officers available to respond to any incidents of disorder around the site”

(Baker D.J. 1995)

Clearly there was a difference of opinion in1995 within the police service with regard to responsibility for crowd safety. On the one hand Thames Valley Police were going to court to argue that they should be present at the Reading Festival to ensure crowd safety standards and deal on the other hand the London and Midlands Police were arguing that the police service should distance itself from crowd safety issues. This apparent difference of opinion led Beckley (1995) to draw attention to the fact that “there is no general consensus within the police service about the legalities of police responsibility in this area”.

One year after the Reading Festival dispute, Superintendent Paul Dumpleton (Hertfordshire Police), presented a seminar paper on the subject of police involvement in a two day concert event at Knebworth by Oasis. The event had attracted a total of 240,000 people to two concerts. This experienced senior officer argued that in his opinion, crowd safety at such events should not be the responsibility of the police because the leisure security industry had a better understanding of the management of such events (Dumpleton 1996).

In an effort to introduce a national approach to policing events the Association of Chief Police Officers (ACPO), comprised of UK Chief Constables, Deputy Chief Constables and other senior officers, issued a policy document in 1999 which stated that police responsibility is limited to core policing responsibilities:

· Protection of life

· Protection of property

· Prevention and detection of crime

· Prosecution of offenders

Crowd safety at organised events was not specifically mentioned in this policy statement however and some officers interpreted the `Protection of Life` to include the safety of a concert crowd and it is possibly a misinterpretation of ACPO guidance that continued to prompt debate among some individual police officers over responsibility for crowd safety for a few more years.

Police Special Services

To understand why it was that some police officers insisted on taking on responsibility for crowd safety it is necessary to understand the system under which the service operates. Chief Constables in the UK are entitled to charge a fee for providing officers to carry out any duty that he/she feels is outside of the accepted police role. Such services are deemed to be special police services. A senior officer determines the number of officers supplied and the Chief Constable determines the subsequent fee for their services. It should be remembered that major concert and sports events are likely to take place at week-ends or at peak holiday periods, therefore officers will be required to work overtime and inevitably charges will be high. The Home Office recommendation is that the cost of providing police officers “should be paid in full by those using the service” (Home Office 1991)

On the issue of the cost of police officers, the Concert Promoters Association (CPA) argued that they were the victims of policies put into place to deal with large-scale public disorder at football matches. The CPA pointed to the fact that, with the exception of isolated incidents, concerts had no history of crowd disorder and public safety incidents have been low in the UK. Quoting the same Home Office document, the CPA reminded the police that “there must be a direct link between deployment and charges”.On the basis of this statement the CPA reasoned that as there is a minimal need for police officers to attend concerts police charges should be far lower. The promoter for the Reading Festival further argued that promoters and organisers of events were being forced to pay very high police charges for a service that is not clearly defined (Ben 1995).

It might be argued however that a willingness by some police officers to accept the role of guardians of public safety possibly owes more to a need to protect police service pension funds than a concern over modern day concert events. The Police Act of 1890 was passed to make provision for pensions and allowances to be paid to police constables in England and Wales. By section 16 the Act provided that every police force should have a pension fund and by sec 16(1) (e) it was further provided that there should be carried to that fund:

“Such proportion of any sum received on account of constables whose services have been lent in consideration of payment as the police authority may consider to be a fair contribution to the pension fund in respect of those constables”.

(Police Act 1890)

It would appear from this statement that the practice of hiring out police officers for duties other than the prevention of crime was a well established practice by the mid nineteenth century. Chief police officers used this Act until it was replaced by the introduction of the Police Act of 1964. The 1964 Act clearly empowered Chief Constables to charge a fee to the organiser of any event for supplying police officers to carry out duties other than their accepted role of anti crime or protecting the Queens peace.

“ The chief officer of police of any police force may provide, at the request of any person, special police services at any premises or in any locality in the police area for which the force is maintained, subject to the payment to the police authority of charges on such scales as may be determined by that authority.” (Police Act 1964: 15 (1))

A definition of special police services was not made clear; it was left to the Chief Constable of a force to determine. A lack of a common definition was not seen as a problem in 1964 but it did become an issue in a court case two decades later when the South Yorkshire Police took legal action against Sheffield United Football Club. South Yorkshire Police sued for the non-payment of a bill for what they described as special police services provided to the club at home matches. The South Yorkshire Police case was contested on the grounds that, “The main concern of the police was to ensure that spectators were able to watch in safety and comfort, and both entry to and egress from the ground were smooth and unobstructed” (Harris v Sheffield United 1985). It would appear from this statement that the South Yorkshire Police were clearly willing to accept responsibility for crowd management in 1985.

The police authority accordingly claimed from the club the sum of £51,699 for providing special police services at the club’s matches between August 1982 and November 1983. The club denied liability for the amount claimed, contending that in attending the matches the police were merely fulfilling their duty to enforce the law, therefore their attendance did not constitute the provision of `special police services`. The club further argued that in regard to the period after October 1983, they (the club) had not made a request for police appearance at any matches.

The judge upheld the police authority claim. The club appealed and the case was finally settled on 26.3.86 when an appeal court judge found for the South Yorkshire Police and Sheffield United Football Club were ordered to pay to the police the sum of £51, 699 plus interest to the total sum of £71,479. 64.

The success of the South Yorkshire Police in the case of Harris v SheffieldUnited appears to have led other forces to the conclusion that they were justified in applying the 1964 Act to a variety of events other than football. A practice of senior police officers stipulating how many officers were required at a concert event and at what cost to the organiser or venue, subsequently became common practice throughout England and Wales.

Police attitudes changed in 1989 however when the South Yorkshire police were again involved in a court appearance. This time it was an inquest into the death of ninety-five, (subsequently ninety-six when a victim died later in hospital), football fans at the Hillsborough disaster. The Taylor Report that followed the incident criticised the police for their failure to control public ingress to the game (Taylor 1989).

At the Hillsborough inquiry the police contended that their role on the day had been that of maintaining public order only. In making this submission the South Yorkshire Police appear to have found themselves in the unenviable position of contradicting their submission of Harris v Sheffield United in 1985, at which they had stated in court that the police “should be in attendance at football matches so as to ensure public safety”

The Hillsborough disaster undoubtedly rang alarm bells for many senior police officers. For many years Chief Constables had been happy to supply officers to football or any other event. The level of charge for a service that was not clearly defined was entirely a matter for the Chief Constable. Hillsborough introduced the question of responsibility for public safety and indicated that those found guilty of negligence in that respect could face a criminal charge. Immediately after the Hillsborough inquiry a number of senior police officers adopted a policy of non-involvement in public safety matters. The UK police view in general today is that of non involvement in crowd management operational planning for any organised social event.


The claim by ADCC Speed in 1995 that the police service did not train to take an active part in crowd management strategies for organised social events was in my opinion an enlightened view. This very senior officer understood the different values of crowd management and crowd control and although individual officers have enrolled on training courses there is little indication today that training for the police service has moved on to consider the specific issues of crowd dynamics, environmental psychology, cultural behaviour and risk analysis, all of which are fundamental to understanding crowd safety management planning. It is acknowledged here that some individual officers do study crowd psychology but by and large these are academic studies that focus on traditional v contemporary theories of crowd single mind. This is very different from environmental psychology study. In short, police training still focuses on crowd control.

The statement by Superintendent Dumpleton that in his opinion the leisure security industry better understood the management of a huge peaceful concert crowd is interesting because at the time that he made this statement (1996) training for crowd safety management was in its infancy. Today it is possible for individuals that do not hold a higher education award to study for a Foundation Degree in crowd safety management at BuckinghamshireNewUniversity and then go on to take Masters and even a PhD in the subject and this has strengthened the concert promoters claim that there is very little need for the police at concert events. This does not imply support for a total non police involvement in concert events. Provision should be made for a police presence based on an event risk assessment.


Baker J.D. 1995: Birmingham Live 95: Contents of a letter from the West Midlands Police to Andrew Hudson, Producer 22.5.95.: Provided by A. Hudson, Unique Special Projects Ltd.

Beckley A.1995; Chief Inspector: Game Plan: A review of public safety legislation: Police Review p16: November 1995

Benn M. 1995: Synopsis of a dispute between Thames Valley Police and the Reading Festival: Private correspondence Ref MTJ/MD provided by the Mean Fiddler Organisation 17th October 1995

Dumpleton P. 1996: Unpublished seminar paper `Knebworth 1995`, delivered to seminar Mass Crowd Gatherings: Home office EmergencyPlanningCollege, Easingwold 10.10.1996

Fruin J. 1993: The Causes and Prevention of Crowd Disasters: Published seminar paper in Engineering for Crowd Safety, Ed Smith R.A. & Dickie J.F.: Elsevier Science publishers B.V.

Harris v Sheffield United 1985: All England Law Report: Harris v Sheffield United F.C. p838: All England Law Report 1985.

Home Office 1991: Audit Commission Report, Extent of Police Charges, 15 sec 5: Home Office Circular No 36/1991

Metropolitan Police 1995: Public Safety at Non – Regulated Events: Code of Good Practice, Introduction para 1.1.:safety/code/draft4/21/08/95513:05 Metropolitan Police Public Order Branch.

Police Act 1890: Sec 16, (1) (e), Cited in, All England Law Report 24.7.87 p840.

Police Act 1964: Sec 15 (1) Cited in All England Law Report 24.7.87 p840.

Security Gazette 1995: Reading Festival controversy raises public event policing questions. Page 5, 25.8 95

Speed T. 1993: Paper presented to HomeOfficeEmergencyPlanningCollege, Easingwold; Lessons Learned From Crowd Related Emergencies: ADCC Tony Speed (Metropolitan Police) 20.5.1993.

Taylor Rt. Hon. Lord Justice: Final Report Hillsborough 1989: Part 1 – Present And Future, Chapter 1, Three Somber Lessons From Hillsborough, Previous Reports Unheeded p4, para 19. : HMSO 1989.

Extreme Value Analysis

March 6, 2011 1 comment
Extreme Value Analysis: its potential use in crowd safety planning

Edited version of a BNU workshop paper

Delivered January 2011

Mick Upton


This paper considers Extreme Value Analysis methodology as a first step in the process of assessing risk when planning for crowd safety at organised social events. The origins of a two level system of analysis are discussed and the way that the system can be adapted for crowd safety planning use is explained. The paper then moves on to provide practical examples of how the system can improve crowd management planning. .

A scientific explanation

Extreme Value Analysis (EVA) is simply a method of estimating the probability of a rare event. It is perhaps best known for its application in meteorology where it is used to predict the likelihood of extreme weather conditions such as heavy rain maximum frost penetration or extreme wind conditions. The method is also used in engineering to predict the possible failure rate of a mechanical system. A wider application to crowd safety planning was used by Brian Toft (1990) in his thesis, A failure of hindsight, to support his argument that crowd related accidents should not be treated as unavoidable acts of god but more likely a failure of risk analysis.

Toft explained that the EVA method requires the crowd manager to conduct a program of research on two levels in order to complete a comprehensive risk analysis prior to documenting a risk assessment for his/her future event. This two level study of previous crowd related accidents was explained by the researcher to be:

Ø A Low frequency – High intensity accident or incident. Explained as being a crowd-related accident that occurs rarely but when it does happen it results in a high casualty rate

Ø A High frequency – Low intensity accident or incident. Explained as being a crowd-related incident that might occur at regular intervals but result in a low casualty rate.

In support of his thesis argument Toft provided two examples of fires at entertainment venues that were the cause of loss of life. The two fires occurred at the Iroquois Theatre, Chicago, 1903, and the Coconut Grove Nightclub, Boston in 1942. Both fires appear to have occurred in remarkably similar circumstances.

Toft explains that in both cases the decorative fabric of the interiors was highly inflammable, exits were locked or had not been provided, both venues were overcrowded and neither establishment had trained their staff to deal with an emergency such as fire.

It would be a dangerous assumption to dismiss Toft`s comparative study as being simply two fires that occurred coincidentally in similar circumstances almost forty years apart. He simply chose oneexample to illustrate a low frequency – high intensity incident. The rationale applied by Toft was that had the operators of the Coconut Grove nightclub researched the cause and effect of fires at entertainment venues they would have found the basis for improved crowd safety planning at the Coconut Grove nightclub years later.

The time period between the two fires is irrelevant. For as the philosopher George Santayana (1905) famously warned, “Those who cannot remember the past are condemned to repeat it”. Sadly it would appear from the evidence available today that the contributory factors to a high loss of life in the two fires cited by Toft still appear to be contributing factors to a high loss of life in fires at clubs and discos today.

Fires at indoor venues are obviously not the only cause of loss of life at organised crowd gatherings. There have been fatal accidents due to high density crush situations, cultural behavior and falls at crowd gatherings that range from concerts to religious meetings and retail outlets (shops). Wherever crowds are encouraged to form the possibility of an accident exists and it might be argued that the larger the crowd the greater the potential for an accident simply because more people increase the likelihood of an incident. This does not imply that all organized crowd gatherings are dangerous, rather that it is essential to plan for the needs of a crowd.

The event organiser has a mandatory obligation to ensure maximum safety standards for the public attending their event and common practice in the UK is to delegate responsibility for crowd safety to a crowd management consultant. The delegation of responsibility to a consultant does not mean that the event organiser is absolved from all responsibility however because their choice of consultant would be a prime consideration at an inquiry into an accident. On that basis it is reasoned here that it is good practice for all individuals and organisations involved with crowd safety planning to have an understanding of the two level approach to risk analysis that the EVA method recommends in order to make an informed choice of consultant.

Low frequency – High intensity

An obvious start point for a better understanding of the EVA method is low frequencyhigh intensity. Simply because a high intensity indicates a fatal accident and therefore information regarding the possible cause and effect are likely to be available from press accounts, online reports and public inquiry reports.

Great care needs to be taken when considering statements made by members of the crowd however. Regardless of how tall a person is, when they are standing within a dense crowd they cannot see what is taking place even when an incident is unfolding close by. People in a crowd situation tend to make uninformed comments to the media. This is a particular problem with accidents at concert events where fans rarely consider the artiste on stage to have contributed to the cause of an accident, even if the artiste has encouraged reckless or irresponsible behaviour. Fans are far more likely to accuse a front of stage pit team of failure to act or the promoter for failing to take action to stop the show. The crowd manager therefore has to critically appraise the information obtained on crowd-related accidents and a reasonably accurate cause and effect pattern documented for possible reference later when planning an event.

An example of the value of data collection and storage can be seen if we turn the clock back to 1989. Imagine that we are responsible for planning for crowd safety at an important FA Cup football match at Hillsborough between NottinghamForest and Liverpool. Online research would not be easy in 1989 but crowd safety issues have been documented since the first FA Cup Final at Wembley Stadium in 1926 when literally thousands of people without tickets climbed over the stadium walls. Information is available if we look hard enough.

If we had studied the documented history of accidents and incidents at football grounds we would easily have discovered that thirty-three people died and an estimated five hundred others were injured during a FA Cup match between Bolton Wanderers and StokeCity on the 9th March 1946 at BurndenPark, Bolton.

The Molwyne Hughes Report into the BurndenPark disaster indicates that possible contributing factors to the loss of life in the accident were:

Ø Late arrival of a large crowd

Ø Game started while a large number of fans were still outside trying to get into the ground

Ø Failure of crowd control outside of the ground

Ø Ingress system at one section of the ground unable to cope with a crowd rush

Ø This resulted in high crowd density in one section of the ground

Ø A perimeter fence prevented people in distress form moving onto the pitch

Ø The police assumed that they were dealing with a public order problem

Ø Venue staff were untrained volunteers

Ø A perimeter fence then collapsed causing injury

Ø A police officer had to go onto the pitch to ask the referee to stop the game once it was realised that there was a serious safety problem

Ø Crowd members in other parts of the ground then assumed that the police had stopped the game because of disorder by a section of the crowd

Now fast forward forty-three years to our match at Hillsborough and our pre match planning has focused on the need to maintain public order because soccer hooliganism has been an on going problem for years now. On mach day our crowd segregation plan is working well and the match starts on time. Then a large crowd arrives late to one section of the ground where the ingress system cannot cope. The police cannot control the crowd outside and hundreds of people rush into a pen that is already full. We have not trained our stewards to deal with the situation and the police immediately assume that they are dealing with a public order problem. Sadly ninety-six people die and hundreds more are injured as a result of being trapped behind a pitch perimeter fence.

Given the fact that our 1989 disaster has occurred in remarkably similar circumstances to that which occurred at BurndenPark forty-three years earlier, what does that say about our approach to match day planning?

High frequency – Low intensity
Turning to high frequency – low intensity assessment, possibly the best example can be seen regularly at pop and rock concerts where many hysterical or distressed members of an audience are extracted from the crowd at the front of the stage and taken to first aid. Experienced practitioners will know that it is often the case that literally hundreds of young people will be treated on site and then allowed back into the audience to continue watching the show. Responsible concert promoters put front of stage barrier systems into place to deal with these situations and they continually upgrade these systems to cope with new cultural attitudes. Consequently the press rarely reports such incidents.

This does not imply that information cannot be found; rather that it takes more effort. Take for example a casual BBC TV News (2008) report that a crowd of 2000 mainly young women turned up to see a free concert at Fairfields Hall, Croydon, on the 30th October 2008 by the UK group JLS. The venue only holds 1500 people and not surprisingly there were reports of crowd hysteria from young women that could not get in. This report does not however seem to have flagged up a warning for the management of a shopping centre in Manchester because in 2009 the Daily Mail (2009) reported that “Thousands of screaming fans turned out to see the group JLS as they turned on the Christmas lights at the Trafford Centre”. The paper further stated that “The gates were firmly closed by 4pm, as the giant shopping centre was almost overwhelmed”.

On the 14th November 2009 local authority organisers of a free concert by JLS in Birmingham were surprised when a larger than expected crowd of young females turned up to see a free concert. The venue was full, but people appear to have broken through a barrier system and the concert was cancelled for public safety reasons. In each of the incidents cited here the event was free, more people arrived than the organisers expected, the crowd profile was predominantly young female and crowd hysteria was a key factor. Fortunately there were no serious injuries recorded and therefore might be regarded by safety planners as being three `near miss` incidents. It also raises the possibility that at least two previous warnings were overlooked by the organisers of the Birmingham event.

Similarly, the organisers of an appearance by the boy band One Direction in Wolverhampton on the 8th December 2010 seem to have been taken by surprise by the actions of the crowd attending the open air event. A press report claimed that 35 young female fans required medical treatment at the free promotional concert (Daily Mail 2010). It appears that fans turned up very early to the free event, possibly not having eaten and waited for a long period in temperatures below zero. This scenario is remarkably similar to that of the 1979 Cincinnati disaster referred to at the beginning of this paper that cost the lives of eleven young rock fans.

It is not always as easy to predict a potential incident, but the cause and effect of so-called near miss incidents that occur are crucial to a risk analysis process. For as Toft (1990b) quite rightly argued, “near misses should not be shrugged off but instead be treated as fortunately benign experiences, since if the same events were to repeat themselves in less forgiving circumstances then disaster might ensue”.

Cross organisational learning

Where there is a lack of credible data on accidents at a particular category of event the researcher must extend their area of search beyond their particular event or venue in order to take into account the importance of what is often referred today as environmental psychology. In his discussion on the concept of isomorphism (the similarity of organisms of different ancestry), Beishon (1980) argued that:

`It did not matter whether a particular system was biological, sociological or mechanical in origin, it could display the same (or essentially similar) properties, if it was in fact the same basic kind of system`.

Consequently, even when two systems might appear to be completely different, if they posses the same or similar component parts or procedures then they will both be open to a common mode of failure. It can be assumed from this that a queue system requires the same level of attention regardless of what people are queuing for. For if seemingly different systems can display the same or similar properties it follows that they can also be subject to the same failure modes. Toft drew on this hypothesis to argue that the similar features of accidents were that any failure that occurs in one system would have a propensity to recur in another `like` system for similar reasons.

Concert planners for example can learn lessons from queue systems failures at special offer sales at retail outlets. The common factor is that pre event marketing shapes the psychology of the crowd prior to their arrival. At concert events crowds will arrive early regardless of the fact that they have a ticket because they are determined to get as close to the front of stage as possible while at sales people arrive early to obtain a bargain. In this scenario specific attention needs to be paid to queue and ingress management, staff training and communications

Some retail managers appear to have been slow to take advantage of inter organisational and cross organisational learning. Reports indicate that there has been a number of `near miss` incidents at retail outlets. Stores that have failed to learn include: IKEA, (one serious incident in London and a fatal incident in Jeddah); Primark (two serious incidents in London); Gucci’s flagship store in Knightsbridge, West London (public disorder during queuing) and Curry’s store Birmingham (one serious incident). In each case it was reported that thousands more people turned up than was expected and staff were not able to control queuing and ingress and the police were called to restore public order. Far more serious was a queue failure at a Wal-Mart store in America where one person died. From TV and press reports on these incidents it would appear that store managers and staff in all cases did not have the knowledge and training to understand crowd behaviour therefore they failed to understand the powerful forces that can be generated by crowds in spite of the fact that there were numerous warnings.


At the key stage of event risk analysis there is much to be gained by conducting a wide-ranging program of research into crowd related accidents and incidents. It is a fatal mistake to restrict the search to your own particular event or venue type. For example, football Safety Officers should not focus solely on football incidents or sports grounds, equally the concert promoter should not ignore incidents that appear to have nothing to do with concert events. Crowd safety risk analysis needs to focus on the possible environmental psychology of crowds during four key phases of;

Ø Crowd arrival and queuing
Ø Ingress and/or processing
Ø Attendance
Ø Egress to include possible evacuation
The overall aim throughout each of these phases is to control crowd density. The nature of the event or venue type is irrelevant as even the best architecturally designed venues can experience an accident if arrangements are not made to take into account the attitude of an attending crowd.
Ref used

BBC TV 2008: BBC Television News report: Hysteria at Croydon Theatre 9th December 2008

Beishon 1980: The concept of isomorphism: quoted by Toft B. 1990 in A failure of hindsight: Published thesis

Daily Mail 2009: Its JLS mania as thousands of fans turn out to see X Factor group turn on lights: 5th November 2009

Daily Mail 2010: Girls go wild for One Direction: Page 3 Thursday December 9th 2010

Santayana G.1905: The Life of Reason, Volume 1: BiblioBazaar 2009

Toft B.1990: Thesis: A failure of hindsight: University of Exeter 1990

Practical use of command and control by crowd managers

March 13, 2010 1 comment
Practical use of a command and control concept by crowd managers
Mick Upton

This paper considers how the leisure security industry has adapted a police concept of command and control. The origins of the concept are first reviewed and then practical application by crowd managers as a planning aid is discussed. Finally the problems associated with implementing a command and control concept within the leisure security industry are considered.


The military would no doubt argue that a command and control concept has always existed within military units, they call it a `chain of command`. A chain of command has two primary objectives, a) to ensure that the strategic aims of commanding officers are met, and b) to ensure a disciplined approach to meeting tactical objectives. It is generally accepted now however that the police service restructured a chain of command system to suit their particular needs.

What is now commonly referred to as command and control was developed by the police following serious public disorder that took place at the Broadwater Farm council estate at Tottenham, North London, 1985, in what is commonly referred to now as the Broadwater Farm riots. In the aftermath of the very serious riots in which a police officer was murdered the Metropolitan Police considered it necessary to ensure that they had in place a system of planning that would ensure that could respond to any future large-scale chaotic crowd situations in a disciplined manner. The system developed was command and control and the Home Office explains these terms as follows:

“Command means the authority for any agency to direct the actions of its own resources (both personal and equipment”).

. “Control means the authority to direct strategic and tactical operations in order to complete an assigned function an includes the ability to direct the activities of other agencies engaged in the completion of a function. The control of the assignment functions also carries with it responsibility for the health and safety of those involved”. (Easingwold KH/SH 95720)

The aim of the concept was to establish a chain of command within the police service of England and Wales that would support public order policing. Subsequently the system was adopted by a number of other agencies as good practice when dealing with crowds.

In a workshop presentation paper Gareth Owen (2009) explained how the police use the terms Gold (strategic), Silver (tactical) and Bronze (operational) to identify specific roles. It was also explained that a police gold commander is likely to be stationed some distance away from the incident in order to have the benefit of CCTV that can provide a wide view of the situation. It is the Silver, or tactical, commander that will direct operations at the scene and Bronze, or operational, commanders that will supervise response teams. It can be seen that this structure follows the military officer and non-commissioned officer rank structure system.

Unlike some countries where the police insist on taking an active part in planning for public safety at events, the UK police service distance themselves from that responsibility. In London, the Metropolitan Police have gone so far as to publish a policy document which states that:

Safety at events such as football matches and pop concerts, where members of the public are invited into private grounds, public safety is the responsibility of the organiser and/or the owner of the property” (Metropolitan Police 1995).

The above statement makes it clear that the use of a police command and control system is to ensure a disciplined response to a major incident or accident. In police terms it is therefore a reactive concept that is designed to assist the restoration of normality to what is otherwise a chaotic situation.

The primary aim of the leisure security industry however is to channel all of their efforts and resources into the safety management of peaceful crowd activity at an organised social event. Their approach to an event is therefore proactive. Should there be any large-scale public disorder control of the site would immediately be handed over to the police silver commander who would co-ordinate, or direct, a combined agency response as necessary.

The leisure security industry does not have the training or resources to respond to a major incident and they certainly do not have the authority to `direct the activities of other agencies engaged in the completion of a function“. The best that a leisure security company can do is to respond to a localised incident, secure an accident scene or assist the emergency services by the provision of crowd control at the scene.

If then the police system of command and control is reactive and the leisure security industry approach to an event is proactive it naturally leads to two questions:

Should the leisure security industry bother with a reactive concept that was designed to deal width public disorder?

If so, do the necessary organisational structural frameworks exist within the industry operate such a concept?

In my opinion the answer to the first question is yes, but it follows that the concept must be adapted to a proactive system. The use of command and control style and language in a plan presented to a local authority for approval demonstrates that a company is taking a structured approach to safety management. Private companies however are unlikely to use the term’s Gold, Silver and Bronze, they will refer to strategic, tactical and operational in reference to levels of responsibility. In other words, all parties involved in discussions for the event are now talking the same language. This use of terminology also demonstrates that the leisure security industry has moved on from an outdated image of being a bunch of cowboys to become a professional industry in its own right.

To answer the second question: do the necessary organisational frameworks exist within the industry operate such a concept? It is first necessary to consider how the leisure security industry has attempted to adopt a command and control concept.

Strategic Level

At strategic level the primary aim of the crowd management planner is to agree the level of responsibility that the event organiser or promoter wishes them to accept. Having reached agreement for the provision of either a crowd control or crowd management service and any pre/or post event security cover required, the strategic planner can move on to conduct a site survey. A site survey is intended to aid the estimation of the resources necessary to complete the agreed task. The next step is to negotiate a financial agreement with the event organiser that will cover cost and provide a profit.

It should remembered however that leisure security companies exists in a hash commercial climate where the bottom line figure on an estimate can be the deciding factor on gaining a contract over a competitor. The fee therefore has to be competitive. This means that any unexpected costs will have to be absorbed by the company and will naturally affect the bottom line profit. The possibility of obtaining additional funds from the organiser to pay for unplanned expenses is extremely unlikely; this is a business agreement in which the promoter also has to protect his/her bottom line.

A key factor at this stage is to make sure that the company is fully insured for the level(s) of responsibility that it has accepted, particularly when the company has agreed to provide a crowd management service. The fact that a company pays a premium to an insurance company at the start of a year might perhaps lead that company to assume that all the services offered by them is covered by their policy. Unfortunately this is not necessarily the case. Insurance companies have very strict rules that govern polices for insuring crowd services and security, any variation of their terms could mean a refusal to pay out on a claim. Two scenarios perhaps serve to illustrate the importance of getting insurance cover clarified prior to the event; both these scenarios were actual insurance claims.

Scenario 1

A leisure security company that is providing a crowd management service to a green field site concert event has also agreed to provide a security service for the build up and de-rig. The event production manager has stipulated that he wants two-security staff overnight, one to safeguard office equipment on site the other to safeguard forklift trucks and vehicles parked overnight.

The perimeter fence for a very large site has not been completely erected, there is no lighting, no transport provided for a mobile patrol and there is no access to a telephone.

Two nights before the show, a criminal gang steels ten valuable fence panels that are stacked up at the far end of the site ready to be installed. As each panel weighs a quarter of a ton the gang must have had a large truck and there would have been a considerable amount of noise involved in the lifting operation. The event organiser therefore claim that security staff was negligent as they did nothing to prevent the theft of valuable equipment and a claim was made against the security company for the loss

The terms of an insurance policy covering theft from an event site or venue, particularly during the night, are likely to be subject to a security survey having been conducted of the site prior to accepting responsibility.

In the case of a casual concert event, an insurance survey is very difficult. A production manager is unlikely to be able to specify the total value of all the kit and equipment left on site overnight by a number of contractors and performers and if they did they would very likely have to employ a large security team. The production manager also works to a strict budget and he/she will therefore look to have a security company provide cover for the minimum price.


In this particular case the company was able to argue successfully that the two guards provided were assigned to specific guarding duties by the production manager and that these duties did not include the protection of fence equipment. Furthermore it was impossible to see what was happening in the dark and had they left their positions they would have risked leaving their positions unguarded. As the production manager did not provide access to a telephone the guards were not able to alert the police. The claim was finally settled by the fence company policy.

The lesson learned here is that when a security service is requested by a client, that service is provided in order to validate their (the clients) insurance policy. What this means is that if an item is stolen or damaged the organiser has to prove negligence by the security staff if they wish to claim against the service provider. When agreeing to provide a security service it is therefore important that the conditions attached to the provision of service are stated in an agreement.

Scenario 2

A supervisor is in charge of a team of twelve stewards operating a five-lane ingress system at a green field site. Prior to gate opening time it rains heavily. When doors open there are approximately 5,000 people queuing to get in. Pedestrian flow is slow and difficult because the ground becomes very muddy due to continual heavy rain.

One hour after doors open, the supervisor becomes concerned over the flooded state of the ground in his area and he takes the decision to stop ingress. He then opens a nearby exit gate and directs all the people queuing to move to the converted 5m exit where the ground is better.

Most of the crowd move across without problem but a woman slips on wet grass and injures her ankle. The supervisor calls a medical team and the woman is taken away for treatment.

One month after the event the security company receive a letter from a solicitor representing the accident victim claiming substantial damages for loss of earnings for their client due to injury and compensation for missing the event. The solicitor claims that the supervisor was responsible for the accident because he directed the crowd across wet grass to a new entrance.

The promoter has already rejected the claim on the grounds that the security company was responsible for crowd safety management therefore they should deal with the claim. The security company subsequently send the claim on to their insurers.

In this scenario it is possible that a court would find a leisure security company liable even if the decision to move the crowd were taken in the interest of crowd safety. In this particular case the insurance company immediately asked to see evidence of training and competence for the supervisor. As he had only attended a company-training program the insurance company refused to pay out on the grounds that there was no evidence to prove that the supervisor had been trained to direct crowd activity. The supervisor had therefore taken an instant decision based on his personal (common sense) assessment of the situation, he did not seek advice or approval from a senior manager as he considered the situation required urgent attention.


The security company was only able to avoid payment of a very heavy claim by demonstrating that at the time that they undertook a site survey the venue operator had stated that the site drained very well after heavy rain. This statement by the venue operator had been recorded in the minutes of a combined services planning meeting. It was on the basis of this wrong information that ingress gates had been located at the original position, therefore the site operators insurance was responsible for settling the claim.

There are two lessons that are learned from this scenario. The first is that it is important to undertake a comprehensive site survey, this includes questioning the site operator. The results of the survey and subsequent discussions at planning meetings should be recorded and the minutes circulated to all parties. Most importantly, members of staff must be trained to a credible standard and copies of qualifications obtained by individuals made available to an insurance company prior to an event.

The two scenarios provided above are intended to illustrate that strategic planning for pro-active crowd safety management is a team effort. A wide range of variables needs to be considered. These variables include: insurance; health and safety issues; risk analysis and assessment; staff level and shift patterns; staff training and logistical issues that are equivalent to moving, feeding and accommodating a battalion of troops.

Practical issues will include planning for queue management; calculating pedestrian flow during ingress and egress; density calculation; effective event monitoring; communications and emergency action drills.

The planning measures taken for all of these issues need to be reviewed at team meetings and amended where necessary. Once agreement is reached all issues need to be documented in a crowd management plan that will be submitted to a local authority safety-planning meeting at which all interested parties have the opportunity to comment on the plan.

Tactical Level

On the day that the planned event takes place the strategic planner is likely to revert to a role of tactical manager, often under the title of Head of Crowd Safety and Security. Other members of the strategic planning team are also likely to take on tactical roles. These roles will include:

Responsibility for localised risk assessment and management of teams stationed in their specific area

The provision of situation reports (sit reps) on necessary queue diversions and pedestrian flow rates to the control room

Dealing with customer complaints

Securing an accident scene as a scene of crime

Incident reports and witness statements

Tactical managers may also:

Manage an event control room

Act as a crowd safety advisor on stage (show stop)

Manage a front of stage pit

Operational Level

Operational level generally refers to team management. For example, a team leader might be responsible for:

The supervision of ten stewards that control an ingress gate

Supervision of a security response team

Supervision of a section of a large front of stage pit team.

The operational supervisor, or team leader, would normally be responsible to an area manager but would be expected to be able to manage queue systems, assess density, calculate pedestrian flow and deal with customer complaints in the absence of a tactical manager.

Do the necessary organisational frameworks exist within the industry operate such a concept?

To answer my second question it is necessary to consider training qualifications that are currently available for the private security industry. My assessment refers to standards for UK training only and I would be interested to hear of how other countries approach training and qualification.

Unlike the police service, which has a formal rank structure and a systematic approach to training, the UK private security industry has lacked formal training for crowd safety management for many years.

The introduction of what are called National Occupational Standards (NOS) in the UK was intended to create a career path from steward to crowd manager as follows:

Level 2 NOS Event steward

Level 3 NOS Supervisor

Level 4 NOS Manager

Licensing security staff

Academic awards offered by Buckinghamshire New University:

Foundation Degree in crowd safety management

BA (Hon)


Theoretically these levels create a career path from steward (L2 NOS) to crowd manager (FdA) but unfortunately the system is weak at NOS level in my opinion. The problem is caused by the fact that two government-sponsored bodies were involved in the introduction of standards on the NOS. On the one hand the Skills for Security organisation introduced frameworks for levels of training for the guarding industry. On the other hand the Skills Active organisation introduced frameworks for what they called spectator control.

While the spectator control Level 3 and 4 awards would seem to be appropriate for crowd safety management closer examination reveals that they were written primarily to satisfy the needs of football grounds and in my opinion these standards are little more than upgraded steward training programs. There is nothing in them that would improve our understanding of important issues such as crowd dynamics or crowd psychology. These issues are not considered to be necessary because the programs are aimed at people working in architecturally designed stadiums.

The answer to my second question therefore has to be NOT ENTIRELYbecause existing NOS frameworks do not effectively address necessary crowd management training at levels 3 and 4 and no provision is made to address the key issues for planning for green field sites or street events. Current practice by leisure security companies is to for their staff to gain a Level 3 or 4 to satisfy an insurance company. The real crowd expertise necessary at Level 3 and 4 is still provided by those forward thinking companies that bolt on their own version of training for key issues not covered by current frameworks. In this sense it might reasonably be argued that practitioners are teaching trainers but I would argue that in terms of proactive crowd safety management this has been the case since the emergence of rock `n` roll in 1955.

Ref used

Easingwold KHSH 950720: Command and control: Advice paper published by The Home Office Emergency Planning College, Easingwold.

Metropolitan Police 1995: Code of Good Practice: Issued paper safety/code/draft4/21/08/95513/05:Metropolitan Police Public Order Branch.

Owen G. 2009: Theory of Command and Control: Unpublished workshop paper delivered to Foundation Degree module LS235 7th Oct 2009

BDO Sydney Australia

March 6, 2010 Leave a comment
Report to the Coroner’s Inquiry – Sydney Australia 2001


In February 2001, I was contacted by the legal team representing Mr. Vivian Lees the promoter of the Big Day Out (BDO) concert event in Australia. The lawyers informed me that there had been a fatal accident at the Sydney BDO in January 2001and I was asked to submit a review report of the actions taken by the front of stage pit team at BDO. My review should also cover European front of stage safety systems in general use.

I was provided with videotape evidence from six camera positions shot by Channel V, the company filming the concert in order for me to form an opinion of the actions of the pit team. On viewing the tapes it appeared that the headline act (Limp Bizkit), had claimed that the promoter had failed to implement their request for a centre thrust barrier system to be installed for the concert. It was also obvious from comments made from the stage that Limp Bizkit had been very critical of the actions of the front of stage pit team at the time of the accident.

Following the submission of my report I was invited to attend the coroner’s inquest in Sydney to expand on it and answer questions on European front of stage safety systems. A coroners inquiry system in Australia operates in much the same way as it does in the UK. That is to say that the aim of the inquiry is to establish the cause of death not to decide blame. I accepted the invitation to attend on condition that I could state my opinions freely regardless of whom I might offend. The lawyers representing both the promoter and the pit team agreed to this in April 2001.

I spent a total of four days answering questions from lawyers at the inquiry. With regard to the specific issue of the installation of a centre thrust barrier, Mr. Lees had stated previously that the system was not called for in the artiste contract and that there were not enough barriers available in Australia to construct such a system. Questions to me regarding the barrier therefore focused on my opinion of the effect that the parallel system that had been installed might have had on crowd density and the actions of the front of stage pit team.

In my opinion the video evidence indicated that the standing crowd density did not exceed 0.5m2, a level that would be acceptable throughout Europe. With regard to the actions of the pit team, it was my opinion that they had done an excellent job and possibly saved many lives by their response to high-energy (moshing) cultural activity.

Following publication of my evidence to the inquiry by the press, I was criticised by one American web site for failing to blame the promoter for the accident. In my defense, I could only repeat that I was not there to point the finger of blame. I subsequently received a personal apology form the operator of the American web site in a telephone call but no apology was published on his web site.

The paper provided here is an edited version of a much longer document provided to the lawyers.

Review of the actions of the front of stage pit team at

The Big Day Out Concert, Sydney 21.1.01.


European front of stage barrier systems




This review considers the crisis management response by security staff during a fatal incident at the Sydney Royal Agricultural Show Ground on the night of 26.1.01 during a concert performance by Limp Bizkit and front of stage barrier systems in common use in Europe. The review has been conducted at the request of Mr. Vivian Lees, the terms of reference are set out in a letter signed by Mr. Lees, dated 7th April 2001, as follows:

To comment as to the conduct of security staff`, this is interpreted to mean the crisis management response by front of stage security staff immediately prior to and during the fatal incident.

To review front of stage barrier configurations in current use in the UK and Europe at rock festivals and to provide any comments or suggestions that I may have to offer.

It is emphasized that any comments or suggestions that are made in this review are personal opinion of operating standards. This review does not seek to establish how or what triggered the incident at Sydney. That is entirely a matter for the appropriate Australian authorities.

The opinions expressed in this review regarding the circumstances at the Big Day Out concert are based on video footage of the incident of the front of stage and communication with Mr. Lees (Event producer) and Mr. Jeff Grey (Security Director), and information provided by Charlesworth Josem Partners PTY Ltd.

The video footage provided to me is from six (6) different angles of the incident. It is therefore possible to form firm conclusions regarding the actions of the security team and the effectiveness of the design of the barrier system.

Glossary of terms used in this report

The following terms used throughout this review are explained for the benefit of the reader who might not be entirely familiar with the language used in concert promotion and rock cultural behaviour.

Body surfing: When people are lifted above a crowd they appear to be swimming as they roll their bodies and flail their arms and legs. Generally the surfer is attempting to reach the front of stage area.

Front of stage pit: A sterile area in front of the stage to allow security staff to protect the stage and mount a crisis management response to a crowd incident.

Kilo Newton (Kn): Pressure rating – approximately 240lbs

Mixer: A temporary structure constructed in front of a stage within the pubic area for the
purposes of monitoring and controlling sound and lighting systems.

Moshing: The act of body slamming into people within an area in front of the stage. Moshing is an extension of punk rock’s “slam dancing” period.

Mosh pit: The area that moshing takes place, not to be confused with the front of stage Pit, moshing can occur anywhere.

Pit team: Security staff stationed in the pit to prevent a stage invasion and to extract persons in distress from a crowd.

Primary barrier: Atemporary structure, generally of demountable type, erected in front of a stage in order to create a sterile zone (the front of stage pit) that will enable a pit team to operate effectively.

Stage diving: The act of a fan or performer diving or jumping into the crowd from the stage.

Secondary barrier: A temporarybarrier similar to a primary barrier, erected to control a crowd. There may be as many secondary barriers as the crowd manager feels necessary within a crowd.

The concept of a front of stage pit

The origins of a sterile zone, or pit, in front of a concert stage can be traced to sixties youth culture. The sixties decade brought about a change in rock `n` roll style by introducing `Beat Groups`, typified by the Beatles and the Rolling Stones. Beat music introduced a change in cultural behaviour from dancing to spectator hero worship. The wish to touch, or hug, a `pop star` became an uncontrollable urge for many young fans and accessible dance hall stages only served to encourage fans to invade the stage, often resulting in shows ending in chaos.

To overcome the problem of over enthusiastic fans, promoters moved their shows back into theatres where it was common to have an `orchestra pit`. These pits had been introduced at the front of stage at music halls in the nineteenth century for the specific purpose of allowing an orchestra to accompany an artiste musically without intruding on the audience view of the stage. Orchestra pits were therefore constructed at a lower level than the stage itself, a concept that remains today in modern theatre design.

Initially when orchestra pits were adapted for a security purpose they were unmanned. This was due to the fact that they were often very deep and this in itself was considered at the time to be enough to deter stage invasions. Deep pits however soon presented an unacceptable hazard when determined fans attempted to climb over them and fell into the pit! Promoters overcame this problem by placing security staff in a line in front of the pit to act as an added deterrent. By the end of the sixties decade a concept of a front of stage pit protected by a strong security team was established for the sole purpose of preventing stage invasions.

During the mid seventies; rock, beat and pop concerts moved outdoor to play to bigger audiences. Early pop festivals at the Isle of Wight played to over a quarter of a million people. This number of people obviously had the potential to create problems at the front of the stage. Festival promoters theorised that by building the stage very high the audience would be forced to stand back in order to see, therefore pits were considered unnecessary. Unfortunately this theory was fundamentally flawed in that it failed to take into account the fact that natural laws might cause a spontaneous dynamic surge which could create an intolerable pressure load on those persons at the front. A major incident was only avoided by the fact that cultural behaviour of the period was peace and love. A great number of people were happy to simply sit on the ground and enjoy the music. Nevertheless, by the end of the seventies it was generally accepted that security teams should be stationed within a pit barrier for two reasons; first to protect the stage and performers, second to assist persons in distress within the crowd.

Contemporary concert audiences are very active in terms of producing a crowd energy release. In this respect modern youth culture is very different from that of previous decades and predicting the energy released by cultural behaviour and natural laws of dynamics are an undeveloped science. Spontaneous surges, moshing and crowd surfing can result in literally hundreds of people being extracted from the crowd for their own safety. The primary purpose of a modern pit at outdoor concert events is therefore not to prevent stage invasions; they are rescue zones that allow a trained pit team to carry out the vital function of extracting persons in distress from a dense crowd. Unfortunately however there is no common standard for pit team training in the UK or Europe at this time. Currently, team competence levels are based purely on team experience and the dedication of those individuals that choose to undertake the often-difficult role of ensuring public safety.

Actions of the pit team during the Sydney incident

It is understood from information provided that the Sydney pit team was comprised of eighty-four (84) persons. This number would be considered in Europe to be above the normal level for a primary pit system, which would average fifty (50) persons. The manning level is however subject to the width of the pit. In this case there was a duel stage concept in place and this could account for the higher number of staff.

Identifying the chain of command within the pit team was easy due to the fact that supervisory staff was wearing white shirts, this distinguished them from team members who wore yellow.

After observing the actions of the team on video I reached the following conclusions:

The team was managed by supervisors that showed a keen sense of awareness of the seriousness of the situation in front of them. At least two (unsuccessful) attempts were made to halt the show by appealing directly to the artiste on stage and further appeals were made to the crowd to co-operate with their efforts to help people.

The fact that team discipline was maintained throughout the incident was evident by the fact that a second line was maintained in order to pass extracted persons out of the pit. This second line also prevented an attempt by one determined individual to climb onto the stage.

Team members, led by two supervisors, went into the crowd specifically to form a cordon so as to enable others to extract a casualty immediately in front of the primary barrier. By doing so they placed themselves at risk of injury from a pressure load created by cultural activity that included crowd surfing and moshing.

My overall conclusion with regard to the actions of the pit team is that they were well led, disciplined and they responded very well to a very serious situation. The extraction of the casualty was done as quickly as possible in the circumstances. While the loss of a life is obviously of deep regret, it is my firm opinion that the actions of the pit team actually saved lives on the night.

Barrier systems design

At European festivals it is generally the responsibility of the promoter to decide if a front of stage barrier system is to be installed. While it is also the responsibility of the promoter to decide on system design, the artiste’s contract rider can often stipulate a particular system.

A design system is therefore generally a matter of agreement between the promoter and the artiste based on an analysis of crowd size, crowd profile, density, anticipated cultural behaviour, topography and natural laws of crowd dynamics. A well-informed local authority officer can however influence a barrier system decision. A good example of local authority input can be seen at the Wembley complex where the local licensing officer will have the final say in barrier system design. At the Limp Bizkit 2001 Wembley Arena concert the local authority officer would not allow the band to have a system of their choice.

The common denominator in the UK, and most European countries, for front of stage barrier systems is the fact that a barrier must be able to withstand an applied horizontal pressure load of 3kn per metre run at indoor events and 5kn per metre run at outdoor events. Pressure loads are generally measured at a height of 1.2m.

It is acknowledged here that a static horizontal pressure load of 1kn applied over a time period of 3/4minutes can cause serious injury or even death to a person. However, the rationale applied to barrier system loading specification is that a weak barrier system would collapse and therefore cause a high casualty rate. Whereas increasing the tolerance of a barrier allows vital minutes for the pit team to extract people in distress. In such circumstances the speed of reaction by the pit team and the time taken to halt the show are crucial factors.

It is important to emphasise at this point that in the case of the fatal accident at Sydney, the video appears to indicate that a vertical load might have been imposed on the victim due to a crowd collapse rather than dynamic load imposed against the pit barrier. Research indicates that in such circumstances death can occur in less than 3 minutes. If there are periods when the stage is in darkness, as there were in Sydney, it is possible that a fatality could occur in front of the pit without the knowledge of the team. A pit team relies totally on light from the stage to enable them to see the crowd, prolonged periods of darkness can therefore create high risk situations.

When using a single primary barrier system there are two options, a parallel system or a curved one. There is a valid opinion among practitioners that a curved barrier improves audience viewing and has the advantage of dissipating a pressure load if there were to be a dynamic crowd surge. A curved system is reliant on the fact that there are major emergency exit points stage left and right. Where it is not possible to install these exit points, a curved system has the potential to trap people in the corners during surge activity.

There is no data available that I am aware of at this time to support an argument that a parallel barrier system is less safe than a curved one. Where a parallel system is installed in a wide arena the crowd can extend both sides of the stage. Naturally crowd spread is governed by an ability of individuals to have a clear view of the stage and the location of emergency exits in that area becomes a key issue. The key factor in all crowd conditions is density control.

The video footage of BDO Sydney appears to show a sanding crowd density of 0.5, with ample room for crowd spread. Mosh pits are clearly visible within the crowd. A density factor of 0.3 only appeared to be evident for approximately 6/8 rows in front of the primary barrier. This density level is considered to be normal for rock concerts and therefore acceptable under UK current guidance.

Centre thrust systems

Following a fatal incident at the Donington Monsters of Rock event in 1988 in which two people died, the concept of a Centre Trust barrier (often referred to in America as a T barrier) was developed. The purpose of the T barrier is to minimise the effect of lateral crowd surges. The concept is currently used by traditional rock acts such as AC/DC.

One advantage of this barrier system design is that it allows the pit team to operate approximately sixty (60) feet into a crowd. In my opinion the thrust should not be extended to reach the mixer or used when high-energy crowd moshing is anticipated as it splits the crowd into two halves.

A split crowd situation requires a pit team to work in two directions at the same time therefore a T system should only be installed where the event warrants it. An extended centre thrust can also encourage the artiste to use it to get closer to a crowd during a performance. This can cause a two-way pressure load within the crowd half way back from the stage that is extremely difficult to deal with. The temptation for the artiste to stage dive into the crowd from within the thrust might also cause a crowd collapse.

When the T system was first introduced there were a number of casualties due to trapping at the base of the T. Design has now improved but the system is now only used where a lateral surge is highly likely. It has been used in America to control cultural behaviour but the system was coupled with a drastic reduction of crowd capacity in front of the stage. An example of a T barrier and a reduced crowd capacity combination is the concert by Metalica at the Los Angeles Coliseum last year (2000) when crowd capacity on the pitch area in front of the stage was reduced by approximately 50%. This was possible due to the venue design, which incorporates a high level of seating together with field perimeter fencing.

Pen systems

In Denmark, the system in use at the 2000 Roskilde Festival was a complex design that incorporated crush barriers in front of the primary barrier. The crush barriers at Roskilde were permanent fixtures that had been in place for over thirty years. Prior to the 2000 incident it had always been theorized that the system prevented lateral and dynamic crowd surges. Unfortunately it was found that they also created trapping points and they were removed immediately after a fatal incident to be replaced by a new system.

The new system introduced was based on a theory that a high-energy release within a crowd mass is activated by approximately 2% of its capacity, therefore by splitting the front of stage crowd into manageable groups of 500 people energy can be controlled. Entry into the system is via two controlled queues. Pens are constructed with de-mountable sections rated at 5kn and ingress and egress is calculated as separate units, similar to a permanent building. The crowd is directed by marshals via an integral crowd barrier network through gates into a pen until that pen reaches a density of 0.5 at which point that pen is closed. Crowd migration is prevented within the system by using the same principle of marshals and crowd barriers. Gangways within the system are kept clear for emergency evacuation requirements and provision has been made for the extraction of people from the secondary pit. A separate pit team staffs the primary and each pen has a team of marshals to supervise it. Safety zones are maintained as a contingency plan to control density in the event that the crowd invades the pens.

It is intended that each pen will be cleared during a one-hour interval between acts. Diverse acts are presented on the main stage in order to encourage people to leave the pens after each act. Emergency evacuation has been calculated into the system. A senior crowd manager, based on the stage with an overall view, is in control of the system and he/she is authorised to stop the show immediately if necessary. It is a condition of the licence that neither the artiste nor their representatives have control of the system and they must stop the show immediately if they are instructed to do so.

It remains to be seen how effective this system will be. It is however an adaptation of a system piloted in Belgium where indications are that it did improve crowd safety. At this early stage of development there are two obvious problems, both financial. The system is very costly in terms of the amount of barrier needed plus the cost of installation. Secondly, there is increased staff cost due to the fact that two pit teams are needed plus security and marshals. Present estimates for Roskilde this year indicate that one hundred and seventy staff will be required to make the system fully effective.

At the 2001 conference of the International Live Music Conference (ILMC), an annual meeting of international concert promoters and other interested parties, a resolution was passed to form a safety group that has a mandate to review concert safety standards in Europe and report back at the 2002 conference. The first indication of change as a result was the 2002 festivals at Lowlands (Holland), Hultsfred (Sweden) and Roskilde (Denmark) which banned crowd surfing. In the U.K. promoter Stuart Galbraith (SFX), has stated that he would like to see a common policy on the issue and he intends to meet with other UK promoters with a view to following the European lead. A major difficulty with a ban however is implementing it within a crowd mass. Roskilde organisers believe that they can overcome this problem by splitting the crowd in front of the stage into smaller manageable groups of 500 persons.

Multi barrier systems

In the wake of the Roskilde tragedy last year (2000) there is now a popular support for multi barrier systems when a mass high-energy crowd activity is anticipated. A good example of a multi-barrier system is the triple system that was used by Midland Concert Promotions Ltd for a crowd attendance of 125,000 people at an Oasis concert held at Knebworth Park during August 1996.

A multi -barrier system should provide controlled access and egress into the first two areas. Each person entering the system at Knebworth was issued with a colour wristband relevant to the area they were directed into. Emergency evacuation was catered for by the provision of exits each side of the arena.

Promoter opinion

In order to present a balanced picture of the of the Concert/festival promoters attitude to the use of front of stage barrier systems, this review considered it relevant to examine European systems used during the year 2000 and 2001. This course of action was taken in view of public statements that have been made in Australia to the effect that T barrier systems are now standard at European events.

The organisers of major events held in six European countries were contacted during the research for this review. These countries were Belgium, Denmark, Holland, Republic of Ireland, Sweden and the U.K. The following criterion was established when selecting European events: they were outdoor concerts, crowd attendance was in the region of 50,000 persons and cultural behaviour was `high energy`. It was found that none of the events contacted used a T barrier system for the main stage. In Holland, Sweden, UK and Ireland the most common design was a primary parallel 5kn system. It is perhaps worth noting that Limp Bizkit performed without serious incident at a UK concert in Leeds with a parallel system in 2000. Two further UK concerts were then researched during 2001. At the Ozfest event at Milton Keynes Bowl in June 2001 it was found that the event operated a duel stage set-up with a parallel primary barrier system. At the Sterophonics event at Donington in July 2001 the promoter installed a single stage set-up incorporating a primary and secondary barrier system.


It has been stated here that in my opinion the actions of the pit team during the Big Day Out event were highly commendable. I therefore believe that comments made from the stage with regard to ineffective security are entirely unjustified. Furthermore these remarks demonstrate that whoever made them had very little comprehension of what was actually taking place in front of the stage.

The presence of persons in the pit that were not identifiable was confusing; it is possible that they were crew members or band security. If this was the case it might be reasoned that they should have taken action to halt the show and bring stage lights up to full potential rather than become actively involved in rescue attempts in semi-darkness. As it was, pit team members were forced to rely on the use of torches during the early stages of the incident and their pleas to halt the show were ignored.

With regard to front of stage barrier systems generally, this review has attempted to present an overview of systems common in Europe. It is not claimed that these are the only systems in use. The issue of system design is very much personal preference therefore it is possible that there are others in use that I am not aware of.

The introduction of a person at Roskilde to be solely responsible for the effectiveness of a crowd safety system and crisis management response is an interesting concept as it removes the promoter one step from responsibility. However, while the recognition of crowd management as a social science is long overdue in my opinion it should be remembered that it could only be effective if coupled with a comprehensive training program that will lead to a formal qualification. Current practice throughout Europe generally is that an individual or a security company with an established reputation for working with crowds is contracted to take on the role. Often a person will combine two roles, for example it is common for production managers to assume responsibility for safety but as they have other tasks to perform on the day I believe this to be a high-risk strategy.

I am currently negotiating with UK government approved training organisations, academic bodies and promoters to introduce a European standard of training that will include; basic steward/marshal, supervisor, pit management and crowd management planning and safety officer, this project is still in its infancy. There has been a favorable response from all sides to pilot programs that I have implemented however and it is possible that a full syllabus will be available for accreditation by next year.

This concludes my review of the issues outlined in the instruction letter. If there are issues that require clarification or expanded on please do not hesitate to get in touch with me.

______________ ___________

Mick Upton
Contact: e-mail


  • Since his report was written the Roskilde system has proved to be very successful and the pen capacities have been increased from 500 to 1000 persons.
  • The introduction of crowd safety training has been successfully introduced at Buckinghamshire New University (BNU), first in the form of the Foundation Degree and now with higher degree courses.
  • In the UK it is now possible to gain a certificate in front of stage pit training from BNU
  • In the UK there are now occupational standards of training for supervisors (Level 3) and managers (Level 4) on the National Occupational Standards
  • My e-mail address has changed to mick.upton

Riverfront Arena disaster 1979

February 26, 2010 Leave a comment
Case study paper based on an account of the disaster written by John Fuller in his book Are the kids alright? Part of my research into crowd safety planning in 1985

Ingress failure at the Riverfront Arena December 1979

Mick Upton


On the evening of the 3rd December 1979 eleven young people died during an ingress failure to a concert by the `Who` at the Riverfront

Coliseum Arena, Cincinnati, USA, in what still is rocks worst ever tragedy. A study of the Cincinnati disaster is recommended for all students of crowd safety. In his book, `Are the kids all right? John Fuller (1981) explained that the queue area for the Riverfront Coliseum Arena was a plaza located one level above street level. The pedestrian route to the plaza was via five entry points from street level, two entry points were designated as entry points on the night of the concert. In the United States, off duty law enforcement officers are allowed to work for private companies and all five entry points were controlled by off duty police officers paid by the venue to ensure that only ticket holders were allowed up onto the plaza.

The Incident

Fuller’s chronological account of the disaster claims that during the early afternoon of the concert the venue became concerned over the fact that early arrivals might affect safety. A venue manager apparently contacted the police in an effort to have a start time of 1600hrs for twenty-five off duty police officers brought forward. It appears that the police were not able to contact all officers but fourteen officers plus one sergeant were able to report for duty at 1330hrs and they took up positions at the five entry ramps to the plaza. At this time it is estimated that there were approximately 200 – 300 people already on the plaza but there was no sign of disorder. The weather was extremely cold, estimated at 34f, and with a strong wind coming off of the nearby Ohio River the chill factor could have made conditions worse. Later, high crowd density would have increased the body temperature of individuals caught up in a crowd mass, there were reports of people witnessing steam raising above the crowd and this is common with high crowd density. If individuals were suddenly removed from the crowd mass, as some apparently were a sudden change from very high to very low temperature conditions might have had an adverse affect on their physical condition as their body temperature fluctuated. Particularly if these individuals had removed or lost outer clothing during the period of their exposure to high-density conditions. As already been stated, loss of outer clothing is a common factor if crowd density reaches a high level.

By 1700hrs it is estimated that approximately one thousand people were on the plaza, clustered around the main entrance in a fan shape. Half an hour later, at 1730hrs, crowd density appears to have increased to a level where several witnesses reported that lateral and dynamic surge activity was taking place spontaneously. Witnesses stated that by 1815hrs surge activity was becoming serious with people losing their shoes as they struggled to stay upright. Another witness stated that at this time the pressure load imposed on the crowd was frightening and people were being lifted up and passed over the head of the crowd to get to safety.

At 1815hrs the Who begin their sound check and some sections of the crowd, possibly thinking that the show had started early, pressed forward believing that they may gain entry, but the entry doors were firmly closed. Jake Pauls (1985) has referred to this common problem as a `front-to-back communications failure`, a condition where people at the back of a crowd or queue may contribute unknowingly to the forces which can build in a crowd or queue. People at the back of a crowd are encouraged to press forward in what appears to be a forward flow when in fact all that has really happened is that density has increased at the front subsequently creating a false illusion that there is more space at the back.

As the situation became serious police officers met with venue security staff to discuss what could be done. It was found that door-opening protocols were obscured by the maze of contracts that existed between the venue and the promoter, artiste management, stewards, security and production crews. All of which had a greater or lesser degree of input into opening the venue doors. It was not possible to hold an all party discussion but an agreement was reached to open a set of doors on the North side of the arena in an effort to divert a part of the crowd away from the main entrance in order to relieve pressure. This decision was announced to the crowd out front over a public address system. The crowd was also asked to take a step back to relieve pressure on those at the front. Making a request to move back is common in such situations but unfortunately it rarely achieves the desired effect all crowd members caught up in a dense crowd consider themselves to be a victim of the action of others, in other words there are no perpetrators only victims

On hearing an announcement that other doors were open, a section of the crowd apparently ran round to that location. Unfortunately however these doors were closed. It appears that communications had failed and the message to open the doors never reached staff in that location. Finding the North doors closed the crowd appears to have turned around to rush back to the main entrance, only to be caught up in a reverse flow situation with more people heading for the North doors. The casualty rate for this incident is not recorded but it is reasonable to assume that there would have least been minor injuries.

At sometime between 1905hrs and 1930hrs security staff attempted to open the front doors to allow ingress forward to sixteen turnstiles. The pressure load on the outside of the doors however prevented security staff from opening doors outward. By sheer brute strength some doors were opened by security staff and a mass surge into the foyer area took place, causing several turnstiles to jam, people then climbed over them. In the belief that the foyer area should be cleared so as to restore order, security staff closed the entry doors. Once again causing a pressure load on those outside pressed up against the doors. Security staff then began to open doors as best as they could and drag people into the foyer. Once the pressure load eased doors were open fully. Unfortunately by this time eleven people were either dead or dying on the Plaza concourse.

Ref used

Fuller G.J. 1981: Are The Kids All Right ?: Times Books

Pauls J 1985: in Crowd Psychology and Engineering p4: Published Paper by Sime J. D.: Jonathan Sime Associates, Godalming Surrey, June 1995.

Disaster at the Atrock Festival

February 25, 2010 Leave a comment


Mick Upton

Summary of a Report to the Israeli Concert Promoters Association


On the 18th July 1995 a sixteen year old young woman and a seventeen year old young man died and many others were injured as a result of a crowd crush during ingress to the Atrock Rock Festival in Israel. Two days later a fifteen-year-old young woman also died as a result of injuries she received during the disaster. The immediate response from some Israeli politicians was to call for a ban on all similar concert events.

Shortly after the Arad disaster concert promoter Yahuda Talit (Talit Productions) contacted me. Mr. Talit was not involved with the Arad concert; he called on behalf of the Israeli Concert Promoters Association (ICPA). The ICPA invited me to go to Israel to make an independent assessment of the disaster, the primary purpose of the visit being to identify the root cause(s) and suggest ways of improving public safety at future concert events. I accepted the invitation on condition that no fee should be paid to me, as my report must represent a truly independent opinion. I had no interest in finding excuses for badly promoted shows. I also insisted that my findings should be made freely available to the official government inquiry team if they wished to read it and that I would be allowed to speak freely to the press if interviewed. These terms were accepted unconditionally by the ICPA and I arrived in Israel on Sunday 6th August 1995, just over one month after the fatal accident. The following account of my visit to Israel is taken from notes written at the time, unfortunately the original report was lost due to a computer failure. To the best of my knowledge however this account reflects an accurate summary of my visit and subsequent conclusions. The findings presented here should not be misinterpreted to be simply an Israeli problem. My international study of concert safety standards leads me to believe that the failures found in Israel can still be found today in other countries, including the UK.

Research method

On my arrival in Israel I was advised that a meeting had been arranged with the ICPA membership that evening. At this meeting concert promoters explained their approach to safety planning, unfortunately however the Atrock Festival promoter was not a member of the ICPA and not available for interview. Press reports of the tragedy had been translated into English for me and I was able to have access to press photographs. I was also advised that two official inquiries had been conducted into the incident, one by the police and a second by the Education and Culture Committee.

I requested a copy of their report from the police but I was advised that it was classified as secret and therefore not open for inspection by anyone not even relatives of the victims. The reason for secrecy was not explained. Interviews with individual police officers involved with the event were however permitted and given freely. The conclusions of the Education and Culture Committee had not been finally reached therefore it was not possible to have a copy of their report either. I did manage to interview security staff that worked at the event and members of the public who claimed to have attended the concert. These interviews combined with accounts from police officers and press reports and a site visit did enable me to build up a picture of the role of the police, security staff training level, marketing strategy for the event, the ingress system design, and most importantly, what actually happened on the 18th July.

The Concert

Arad is a small town in southern Israel approximately 15mins drive from the Dead Sea. For twelve years prior to the 1995 event the town council, under the direction of the mayor promoted the Arad Festival. Each year the complete town was given over to performers and artistes to stage drama, poetry, mime, street theatre and music. The festival lasts for a full four days and ran continuously twenty four-hours a day. As a performance of one type concludes so an alternative performance began at a different venue or location. A trip to Arad for the festival had become traditional for young Israeli’s and foreign visitors to the country.

After the 1991 festival it was decided to allow a promoter to take on the specific task of presenting concerts for a youth culture. Accordingly, the council brought in the FORUM company to stage contemporary rock and/or pop events. Forum was chosen on the basis of their reputation as operators of popular discos. They had also been involved in the promotion of smaller live music events. The added attraction of rock to the festival appears to have worked well during 1992/3/4. During this period there seems to have been little cause for concern to the council, although the major rock promoters in Israel stated at my meeting with the ICPA that they had been concerned over safety standards following the 1994 event. Exactly why the promoters were concerned was not made clear by them, it was simply stated that they felt that crowd safety management was lacking. The ICPA did not voice their concern to any official body, it appeared to be confined to informal association meeting discussion.

By 1995 Forum had graduated from small events to promoting major concerts when they announced their intention to promote the American rock band R.E.M. at the sports stadium in Tel Aviv, although this concert was not actually scheduled until three weeks after the Arad festival. Nevertheless the company appears to have felt sufficiently confident to promote the Arad concert. The 1995 show would be bigger and better than anything that they had done previously. To ensure working capital they took on two brothers by the name of Schwartz and together they secured an open field site that was approved for a capacity of 18,000 people. Capacity attendance appeared to be based on the UK principle of 0.5m2 (two persons per square metre).

As a temporary site the promoters needed to fence off the required area and this they did by erecting an 8ft high temporary wire mesh fence. The fence was supported at the base by standing each panel in concrete blocks. Each panel is then linked to the next by inserting horseshoe type clips into the top of the hollow tubes of each panel. This type of fence is in common use at building sites and events held at temporary sites throughout Europe but it is not designed to take a pressure load without any form of bracing. No bracing appears to have been used at Arad.

Entry to the site for the public was via one entrance only. Staff supplied by a private security company controlled this entry gate. Police officers were also in attendance at the gate in some considerable number and they had complete authority to take on the spot decisions in all matters both inside and outside of the site. It was interesting to note during interviews with police officers that every one of them insisted that they had final responsibility for public safety matters not the private security company. The attitude of the police was perhaps shaped by the volatile political situation that existed in Israel at that time. Terrorist attacks were common then, and still are today in Israel.

To ensure that people without tickets did not attempt to gain entry by climbing over the fence the promoters appear to have took the decision to fix barbed wire at the top of each panel. The decision to install the wire could perhaps have been regarded by the organisers as a visible deterrent because, as stated previously, this type of temporary fence will not take the pressure load imposed by a single person let alone a crowd intent on seeing a free show. Although the local police claimed responsibility for public safety they did not appear to be aware of the weaknesses of this particular type of fence installation and they did not object to it at the planning stage.

The festival was billed as a presentation of three days of rock music to be held on the evenings of 18th, 19th and 20th of July 1995. Customers could purchase a one-event ticket for sixty-shekels (approximately £20 then) or by purchasing a three-day ticket for which customers would be entitled to a discount. The system for retailing tickets in Israel requires that a licensed ticket agency handle the distribution of all tickets sold. This rule is to ensure that events are not over sold. On this occasion ticket sales were handled by the HADRAN AGENCY based in Tel Aviv. Tickets for the three-day event sold well and there was a particularly heavy demand for the first concert on the 18th. This show was billed as the farewell performance by Israel’s top rock group MACHINA. Although unknown to European audiences, Machina had dominated rock sales charts in Israel for the past twelve years and the demand to see their last performance was such that this particular show sold out quickly.

Event marketing

The marketing policy for the Atrock festival appears to have played a major part in the incident that cost three lives. The promoters of the three day event undertook a private agreement with local banks to offer any person that opened a bank account with a minimum deposit of thirty shekels (£10), would receive a voucher that could be exchanged for a reduced price single day ticket. This offer was subject to availability of tickets on any particular day at the box office and did not a guarantee a ticket. Many young Israeli’s however mistakenly took this marketing campaign to be an offer of a single day ticket at a reduced price. In short, they thought that they had already got a ticket for any day. Not surprisingly many young people rushed to open bank accounts in order to get a voucher. Very few people appear to have read the small print on the back of the voucher. Literally hundreds of young people now held what they firmly believed to be a ticket for a show that was already widely known to be sold out when in fact what they held was a conditional voucher that did not guarantee entry to the event. Subsequently a great many voucher holders turned up on the 18th expecting to get in.

The incident

Doors opened to the public at 2045hrs the opening of the doors was dictated by the length of time taken for the sound check of those groups that were appearing. Sound checks are necessary to establish a correct balance for the artiste and to enable the sound engineer to set his/her control panels for light and sound for each act. This particular sound check appears to have taken a great deal longer than most as it was not completed until 2030hrs. The key factor in this case is that the first act was due to appear on stage at 2100hrs. This time frame indicates that only fifteen minutes was allowed from the gates opening for the public to enter and the first act appearing on stage. Witnesses stated to me that at the time gates was opened there was a large crowd, estimated in thousands, waiting to get in. With only one entry gate in use it was highly unlikely that more than a few hundred could have been admitted in time for the opening of the concert. Although the first act was not the headline act, the sound of live music being heard by those on the outside was bound to cause crowd control problems for the security staff. A normal admission time for this size attendance would be expected to take in the region of two hours, however this is largely dependent on the number and location of ingress gates.

Interviewees stated that there were approximately twenty private security guards at the entry gate backed up by ten/ twelve police officers and they do not appear to have coped effectively with the initial opening rush. Reports suggest that there was little attempt at establishing a linear queue pattern and bulk queuing quickly formed at the gate. A bulk queue situation was allowed to grow out of control for the next forty-five minutes as security and police made frantic efforts to get as many people as possible through the gate.

The security guards stated that they had not been made aware of the likelihood of people turning up with vouchers and expecting to gain entry. When this did happen security staff attempted to turn them away fearing that had found forged tickets, or at best referred them to the box office which had sold out of tickets for that show. With so many people arriving with vouchers the gate soon became totally blocked. People refused to move until they had exchanged their voucher for a ticket, and those people with tickets became very frustrated by the fact that they could hear that the show had began but could not get in.

At 2145hrs a police officer became so concerned over the situation at the gate that he took it upon himself to go to the stage and ask the support group to stop playing. The group stopped playing and the police officer left the stage. After waiting a short while and with no further information given to them however the group began to play again. After two songs it was made known to the group that there was a serious problem on site and they stopped for the second time, this time they left the stage. At 2200hrs the police officer in charge at the site gave instructions to close the gate because he was afraid that the site would become overcrowded. A combined team of police officers and security staff did this.

Once the gate were closed the pressure from the crowd on the outside built up to a point where a lateral surge was spontaneously triggered onto the perimeter fence which then collapsed taking with it the closed gate. The fence fell inward onto a section of the crowd on the inside, trapping them underneath it. This sudden pressure release carried people from outside to inside like a tidal wave. As people were carried forward they trampled over the ones trapped underneath. Those underneath were not able to lift the fence due to the weight on top of them, others close to what had been the top of the fence were trapped by the barbed wire fixed to each panel and of course, the panels were linked together.

Once the combined security and police team was able to establish control it was quickly realised that two people were dead and many injured. The show was then cancelled that night. Press reports of the incident claim that it took three hours to clear casualties from the site to hospital. Two days later the news was announced that a third person had died as a result of her injuries. Attempts were made to stage the remaining two shows but the artistes refused to take part and these shows were then cancelled.

The promoter(s)

The decision on ingress facility design and fence type for an event that was known to have sold out 18,000 tickets in advance appears to indicate a lack of knowledge of pedestrian flow planning. The fixing of barbed wire to prevent non-ticket holders gaining entry actually indicates that those in charge of the production possibly suspected those problems could occur that night. In which case measures should have been taken to brace the whole fence line and install at least two more ingress gates.

No ingress flow calculations were available from the promoter, the police or private security at Arad therefore it is not known how the organisers considered design for crowd safety in terms of pedestrian flow.

The police

According to information provided by the Israeli Concert Promoters Association the police officer in charge of the event was immediately dismissed and the area police chief posted to another area, consequently it was not possible to interview these two officers. The disciplinary action taken by the police however appears to have been taken before either of the two subsequent independent inquiries had even began. If accurate, the actions of the police are puzzling, as they (the police) appear to have acknowledged their lack of crowd safety management planning.

Interviews with individual officers led me to understand that the Israeli police do not train officers of any rank in proactive crowd management. This is by no means rare, I have never discovered a police force in any country that I have worked in to have provided crowd management training, they all train for crowd control. By setting themselves up as the authority on site however an assumption might be made that the Israeli police did approve site design and safety standards. The role of the police prior to the concert and during the incident is, by their own speedy disciplinary actions, to say the least questionable.

Private security

Interviews with security staff in Israel revealed that formal training in proactive crowd management strategies was not a condition of their licence to work. All of the persons interviewed (8) claimed that their experience of working with concert crowds gave them an advantage over police officers but nevertheless they were not invited, or permitted, to advise the police on such issues.


It is common for the media to report that a crowd related fatal incident occurred as a result of panic or irrational crowd behaviour. The Arad tragedy was no exception, it was widely regarded to have been caused by people trying to gain free entry to a sold out show. In my opinion however the tragedy was caused by a combination of serious errors by separate organisations involved with the event. These errors occurred for a number of reasons, listed here as follows:

A failure to understand natural laws of crowd dynamics by the event planners

No contingency planning for ingress

A failure to understand crowd cultural behaviour (crowd psychology) by event planners

Poor systems design (perimeter fence & ingress system)

A marketing strategy that invited chaos

Lack of police training for proactive crowd management

Poor communications on site

Total absence of command and control

No risk analysis – risk assessment – event plan

In view of what I considered to be the above failures, it was my opinion that the crowd tragedy at Arad was not caused by irresponsible crowd behaviour or disorder. It was caused by a combination of lack of understanding of planning for crowd safety management by organisers and a failure by the police to cope with a role that they had insisted that they be responsible for. As stated previously, these are common problems today at some events at international level. It is a constant source of amazement to me that the safety management of peaceful crowds is still not recognised as a social science anywhere in the world to my knowledge.



February 23, 2010 3 comments
Presentation Paper


Mick Upton

Delivered to

The Home Office Emergency Planning College


Mass Crowd Events

8th December 1995

Published by Buckinghamshire New University in:

Case Studies in Crowd Safety Management

By Kemp, Hill, Upton & Hamilton

Entertainment Technology Press Safety Series 2007


On the 28.8.1988. two young men died and a third was seriously injured in a fatal crowd related incident during an open air rock concert billed as the `Monsters of Rock` at the Donington Park motor race circuit North West Leicestershire. This paper is a personal account of my involvement in what is now commonly referred to as the Donington Disaster.


My involvement at the Monsters of Rock event was as head of security and crowd management planning. As such, I was employed by Aimcarve, the company responsible for promoting the event. I attended all pre meetings and I had a good working knowledge of the site having worked on all events since 1980. At the 1988 event I was responsible for all security issues, crowd management planning for the event and the implementation of crisis management should the need arise.

At the time of the event (1988) there was no established practice to set up an Emergency Liaison Team (E.L.T.) therefore the event did not have one. The police attended all Safety Advisory Group (SAG) planning meetings for the event but no police officers were present within the arena during the concert. The police were fully aware of what was taking place during the event via radio communication but they did not take an active part in containing any of what turned out to be three separate incidents that took place or the subsequent rescue operation of the injured. These actions were entirely the responsibility of ShowSec International Limited.

The following account is not claimed to be a factual minute by minute account of what happened on that day, it is simply the recollections of one who was actually there. As with all accidents or incidents there will doubtless be other witness accounts from people who were present that day and will recall things differently. This is natural when people see things from different perspectives. My version of the disaster is however based on a broad view of the sequence of events and the benefit of radio reports to me from trained supervisors. From this wide perspective I believe my version to be an accurate account of the events that lead to the loss of two lives, the serious injury to one person and a traumatic experience for many other people that were involved either as victims or rescuers.

The incident at the 1988 Monsters of Rock concert has been the subject of much discussion by those that have an interest in crowd safety matters. At the time of the incident the media showed a natural interest, as would be expected when two people die in such high profile circumstances. Unfortunately much of what was reported by the press as the cause and effect of the incident was inaccurate. For example, the name of the act quoted in some reports, as being on stage at the time was the singer David Le Roth, it was in fact Guns `n Roses. Other reports indicated that a high crowd density build triggered the incident directly at the front of the stage subsequently causing crushing, this was also inaccurate. Several other papers claimed that the group on stage refused to stop playing, this was also inaccurate. Yet another report claimed that police were directly involved within the crowd at the time of the incident when in fact the police were not in the arena at all.

Confusion over what actually happened was possibly due to the fact that there had in fact been three separate incidents that day. Reporters were unaware of this when they questioned people. Perhaps, confused by the different versions given by some witnesses, they (the press) simply cobbled together a version of a single incident. While this version may have satisfied newspaper editors it did little to help students of crowd safety to discover the true course of events. This account might therefore help those that have an interest in crowd safety to understand what happened on the 28th of August 1988 at the Monsters of Rock.

Previous history of the event

Promoter Maurice Jones, Chairman of Aimcarve Ltd, the organisers, first staged the Monsters of Rock in 1980. Jones enjoyed a well-earned reputation for promoting good rock shows. The 1980 show was an immediate success with rock fans. Many of whom traveled from different parts of Europe to be there. The demand for an annual rock show was clearly evident and the show has continued to be held annually.

Heavy rock groups came to regard the Monsters of Rock show as being a major international showcase. To play at Donington was seen to be performing at the rock event in the heavy metal calendar. It was covered by virtually all the trade press, and in addition to being broadcast live on radio bands often filmed their performances for future video promotion to enhance a sales campaign of their latest recording.

Many fans will travel great distances to be there regardless of was performing as the headline act. At the first concert in 1980 the demand for tickets was such that people arrived and set up camp on the Monday prior to Saturday’s concert. Over the years the promoter had actively discouraged camping and tried to present the event as a one-day concert not a festival. Nevertheless some fans still arrived with tents on the Friday evening. The promoter did provide facilities for these people. These facilities included; toilets, lighting, large marquees (for those without tents), food vendors and vast amounts of firewood.

The campsite itself was an amazing prelude to a rock concert. Nobody appeared to want to sleep. There was perhaps what could only be described kindly as a carnival atmosphere. Rock fans have a well deserved reputation for a love of alcohol and giving them the opportunity to spend all night in a field with a ready supply of beer while listening to ear splitting rock music appears to be their vision of heaven. In fact some of them have been known to enjoy the night on the campsite so much that they have slept through the entire day of the concert and gone home on the Sunday having had a wonderful time!

The venue

Donington Park is a motor race circuit situated approximately midway between Nottingham and Derby at the village of Castle Donington. The site itself is next to the East Midlands International airport. The circuit has a long history of motor sport and the track has been upgraded to handle F1. However it is perhaps best known as the home of the British Motor Cycle Grand Prix. Motor sport attracts large crowds and the venue has developed good working relationships with the emergency services. As one would expect, there is a major incident plan in place, which takes into account virtually any scenario that could occur at a motor sport venue situated next door to major airport.

Structural security in terms of crowd management is good. The track area is surrounded by a brick wall approximately twelve feet high topped with barbed wire. Entry and exit gates are designed to cope with ingress and egress crowds in the region of 200,000 people. The arena area is situated in the inner circle of the track with a standing room area that could accommodate 150/200,000 people. In practical terms however this number of people could pose problems with sight lines for a concert therefore attendance at the Monsters event would normally range from 50/100,000. Attendance for the 1988 event was given by the promoter as 85,000 people.

The stage was built at the north east corner of the arena, facing approximately south east so as to direct PA sound away from the village of Castle Donington and toward the open area of the airport. In 1988 the grass area immediately in front of the stage sloped down toward the stage at a gradient of approximately 1/20. At the time this was considered advantageous to viewing, however after the incident this area was laid flat.

The Police

As stated previously, I had been present at every Donington concert from 1980 and I have never seen a serious confrontation between the police and the fans. The police have always taken what I would describe as a realistic approach to the event. This is not to suggest that they ignored crime. Local officers have come to understand the rock culture and realised that a torn denim or leather look of the average rock fan did not signify that they were a rioting mob. All police control vehicles, temporary police station and facilities were therefore positioned near the main entrance to the site not in the arena.

Officers patrol the campsite during the evening and night of Friday but the event itself is left in the hands of a private security company employed by the promoter. The only officers in the arena in 1988 were two that were assigned as liaison officers and they were in the back stage area. In my opinion the working relationship between private security and all the emergency services prior to 1988 were good and since the incident they could be classed as excellent.

Pre Planning

The event was licensed by the Northwest Leicestershire District Council who chaired regular meetings for some six months prior to the event. For me planning for the 1988 concert began in March of that year with informal meetings with the promoter. At this point I was advised that Iron Maiden would be the headline act. This particular act had a long history of successful record sales and the indications were that they would attract a large crowd but their stage act was unlikely to cause serious problems.

During the months leading up to the concert the local authority arranged a number of meetings that were attended by the emergency services, the promoter, St.John Ambulance, myself and Tony Ball, who at that time was my assistant and would act as the site controller on the show. At these meetings it became clear that an attendance of 85,000 people was likely. Accordingly, the promoter agreed to provide daylight vision screens each side of the stage to avoid crushing at the center of the crowd and a front of stage barrier. A single primary barrier constructed of scaffold and ply board was installed for the event. This design was common at the time. The event used only one stage.

On completion of a site survey I proposed that a total of 520 steward staff be deployed on the day and all parties agreed this number. In July we were advised that the line up of major artistes for the concert was to be;





I had worked previously with three of the four named acts and they were considered to be low risk in terms of crowd behavior. However the fourth act, Guns `n Roses, were a new young American act that we knew little of. Our risk assessment indicated that their stage act was aggressive in style and that they had strong record sales that would possibly attract a large number of their own fans to the show. Our audience profile indicated that these fans would be younger and more volatile than we would normally expect. The term used to describe this new trend in heavy rock was Thrash Metal. This title was given to indicate the way that these fans thrashed about wildly to the music to a point where they appeared to be hitting each other. As a result of our risk analysis we increased the number of security staff in the front of stage pit to forty.

The Incident(s)

The weather conditions prior to the concert were bad. It rained continuously from the Monday to the Thursday. Nevertheless the crew completed construction of the site on time. By Friday the weather had improved but conditions were still very wet and it was necessary to impose restrictions on vehicle movement around the site. Saturday, the day of the concert, the weather remained overcast with occasional showers.

The doors opened to the public at 0800hrs and ingress was completed without serious incident. Those fans that had arrived early made their way directly to the front of stage where security were not able to sit them down due to the wet ground conditions. However the crowd was cheerful and friendly and appeared to be happy to wait until the supporting acts would start to appear. During the period of their wait recorded music provided by a disc jockey from B.B.C radio,

Incident # 1

At 1300hrs the control room received a message from the supervisor back stage advising that high winds were causing a serious problem to the daylight screen position at stage right. A rigging crew was alerted and I arranged to meet with Tony Ball at that location to assess the situation. By the time that we reached the location however the complete screen assembly had collapsed and was only prevented from falling onto a section of the crowd by a steel fence that surrounded the backstage area.

The ShowSec supervisor had moved the crowd from the danger area and a rigging crew was attempting to retrieve the damaged screen rig. Efforts by the rigging crew were continuously frustrated at this time as a small section of the crowd insisted on attempting to stand underneath the suspended rig because the screen was still relaying live pictures of the act on stage. Their actions however put them in great danger. My actions at this point were to deploy more staff from the pit to secure the danger area, cut the power to the screen and secure the damaged screen so that it would not fall. It was not necessary to stop the show as the band on stage shortly finished their act. The team operation to retrieve the screen took approximately one hour but before it was finally completed Guns `n Roses went on stage. Almost immediately I received a report from the pit supervisor that there was an unusual high level of crowd activity taking place in front of the stage. At this point I detailed a security team to return to the pit. I remained at the screen location to ensure that it was safe and I detailed my assistant, Tony Ball, to go to the pit and send me a situation report.


At approximately 1415 hrs Tony called me on the radio and asked for me to come to the pit urgently. I immediately went to the area which was approximately one hundred yards from where I was. On arrival at the pit I saw that there was a great deal of activity within the crowd. Density was clearly at 0.3 immediately in front of the barrier, however this was not a serious problem. At approximately ten rows back however a density of approximately 0.5 allowed room for lateral surges, which were becoming a problem. I witnessed crowd surges that ran across the complete front of the stage. One particular surge started at stage right and stopped suddenly with a crowd collapse at the centre of the audience approximately fifteen yards out from the front of stage. It was immediately obvious that this was a serious situation as approximately fifty people were involved in a crowd collapse. I sent a four-man team into the crowd to assist and assess the problem. At the same time I sent a message to the stage to ask the singer with Guns `n Roses to stop the show as we had a problem. The singer immediately stopped the show and he then used the stage PA to calm the crowd and advise them of the problem.

The advance team reached the spot and attempted to send back a radio message but unfortunately their communications failed due to the fact that victims of the incident grabbed at their radios and pulled the microphone lead out. At this point the team leader signalled to me to go to the spot. I instructed Tony Ball to remain in charge of the pit and I advised the control room, which was managed by my business partner Gerry Slater, that I was going out into the crowd with another team. The show was still stopped at this point.

On reaching the spot I found that the advance team was dealing with approximately 10/15 people that had obviously been at the bottom of a crowd collapse. Ground conditions were bad and the people involved were covered in mud, I decided that they should all be extracted from the crowd for their own safety. We managed to lift most people up and were passing them toward the pit when unfortunately the band on stage assumed that the incident had been fully contained and they resumed playing. Suddenly, the whole crowd around us erupted. A large section of the crowd, and two of our own security team, collapsed in front of us. I witnessed approximately 30/35 bodies that suddenly piled up in front of me covered in mud. Our efforts to pull people off of the pile were hampered by the fact that people behind us climbed onto our backs in an attempt to what we now know to be crowd surfing. In some cases these people dived over our heads onto the pile of bodies.

At this point I lost communications, my radio and earpiece was ripped from me. I was not therefore able to advise Tony of the situation. He realised however that we were in serious trouble and he quickly dispatched another ten-man team to assist us. At the same time he stopped the show for the second time. Once assistance reached us I was able to establish a cordon around the scene and retrieve the bodies. Tony had managed to establish a line of security people that extended from the pit to us and this enabled us to pass people back to St. John Ambulance staff, who were stationed at stage right. We managed to retrieve over thirty bodies, all were covered in mud. It was noticeable that some were also bleeding and others had obviously vomited. When we reached the bottom of the pile we discovered one person unconscious. This person was immediately passed to the pit where he was resuscitated by Steve Johnson, a ShowSec pit team member.

Unfortunately, as we removed what we took to be the last casualties we discovered another two bodies underneath them. These two were both laying face down in about four inches of mud and they were almost covered over. At first I did not realise that they were people. The pressure load on these two victims was such that we had to dig under them with our bare hands to turn them over. We managed to extract both of them from the mud but they appeared to us to be lifeless. Suddenly the band started to play again and crowd conditions made it impossible for us to examine them in any detail. We immediately evacuated the victims to St. John where they were removed to hospital but were found to be dead on arrival.

Two senior police officers that had been advised of the seriousness of the incident arrived at the scene. The first indications were that the police wanted to stop the concert. After further discussions with the promoter however the show was allowed to continue as the police then decided that no purpose would be served by stopping the event. It was agreed in fact there might be a risk to public order by doing so as the majority of the audience were not aware of the seriousness of the incident.

Incident # 3

At 1700hrs David Le Roth was on stage. Both Tony Ball and myself had remained in the pit area to monitor the crowd. There had been no further incidents since Guns `n Roses had finished their set but we had decided to remain in the area. At approximately 1715hrs I noticed a young woman stage right about ten yards out who appeared to feint. She disappeared from my view and she did not reappear. As I was the one that knew exactly where she was I advised the team that I was going in to find her. I went into the crowd and on arriving at the spot I found a young woman aged approximately fifteen, laying face down on the ground. The crowd parted and I was able to check her condition, she was still breathing, there was a pulse and she there were no obvious signs of injury. As she would have been in danger where she was I lifted her up to take her back to the pit where medics could attended to her. At this point the pit team signaled to the crowd to part to allow me room to bring her in and I started to do so.

As I approached the pit I noticed one member of the pit team begin to climb the front of the stage. I can only assume that he was in a state of shock from the earlier incident because he climbed onto the stage and pleaded with David Le Roth to halt the show as there had been too many injuries and the show should be abandoned. The singer, not realising the fact that he was a member of the pit team, assumed he was about to be attacked and called his two personal bodyguards onto the stage. They promptly grabbed hold of our man and literally threw him off the stage into the pit. Some sections of the crowd that had realised what was happening took offence at this and a dynamic surge toward the stage occurred. It was possibly this incident that the press later mistakenly reported as being the point where Le Roth refused to stop the show.

At this point I was still in the crowd with the young woman in my arms. A member of the crowd, presumably seeing my security shirt, then decided that all security staff should be attacked and he hit me in the face with a one-litre plastic beer bottle even though I was obviously carrying a casualty. Although the bottle was only half full he hit me with enough force to knock me sideways onto the ground. This happened near to the barrier and fortunately the pit team was able to grab the young woman and I suffered the indignity of being rescued by my own pit team. The team also managed to recover our shell-shocked security man, who had actually missed the pit and landed in the crowd. He was also taken to St. John medics where he made a full recovery.

There were no further incidents, Iron Maiden completed their set and the show closed on time.

The inquest

The inquest into the two deaths at Donington was held at Loughborough Town Hall during February 1989. After four days of listening to testimony from more than thirty people that were involved in the incident, the verdict reached was accidental death. The coroner praised the efforts by the pit team by saying “their efforts undoubtedly saved lives”. It was recommended that the following measures should be taken to prevent a similar disaster:

Muddy conditions should be made less hazardous A person should be positioned on stage to have overall control of safety

The giant stage should be moved so that it is not at the foot of a slope

It was also recommended that a working party be set up to discuss safety at concerts.

The aftermath

The sense of shock resulting from the Donington disaster was the catalyst for a review of the Greater London Council (G.L.C.) guidance for pop concerts, which was current at the time. A review body chaired by Richard Limb from the Northwest Leicestershire District Council invited members of the concert promoters association, private security companies and volunteer groups to work with local authority officers and the emergency services throughout 1989/90 to formulate a new guidance document for concert events. The result was the Guide to Health, Safety and Welfare at Pop Concerts and Similar Events. Perhaps more commonly referred to by the colour of its cover as, `The Purple Guide`. The front cover of which had a photograph taken of a crowd at a later Monster of Rock concert. I was pleased to have had input into the document in the form of the chapter on crowd management.

The Purple Guide was also significant in that it advocated that an Emergency Liaison Team (ELT) should be established for each major concert event. The ELT should be made up from members of the emergency services, the local authority and the security team. The role of the ELT is to immediately take command if a serious incident occurred, the police officer present would then take charge of co-ordinating command and control, and evacuation if necessary. The system is now standard practice in the UK.


On a personal level I learnt a number of lessons from that day.

The importance of a good control room

Good communications are essential

Staff need firm leadership – casual staff might not cope with disaster

All senior staff must have a very clear understanding of their role

Someone must take action to stop the show quickly

Staff at the scene can be affected by what they witness, they to might need care and attention

7. Training is fundamental to safety management