What is predictable is preventable.Graham, 2016
When developing risk assessments and contingency plans, have you accounted for the threats that may harm your attendees? I looked at crowd disasters over the last thirty years and identified the ones that
Crowd disasters occur when people are given insufficient information and space (Sime, 1999). It is rarely the case that only one factor causes a disaster (Helbing and Mukerji, 2012). I aim to identify the most common causes of disasters so they can be mitigated in future events. Within crowd science, well-known models of classification include Fruin’s (2002) development of the FIST model of classification, Force, Information, Space and Time. Still (2014) developed the DIM-ICE Model, which categorises elements of an event plan into Design, Information, Management – Ingress, Circulation, and Egress in order to adequately assess and mitigate risk to the crowd. Using these models gives a structure to the elements of an event or crowd management plan and allows the plan to be broken down into stages of planning, licensing and operations.
Why did it really happen?
Understanding causality is crucial in the prevention of crowd disasters. Without identifying the reasons why disasters happen, prevention will never be effective. The HSE (2004) defines causality as follows;
- Immediate (proximate) causes: the agent of damage, injury or ill health (stadium stand collapses)
- Underlying (distal) causes: unsafe acts and unsafe conditions (untested stadium design)
- Root (distal) causes: the failure from which all failures grow (lack of respect to safety by management, prioritising finance over safety etc.)
Classifying causality into proximate and distal causes helps to understand the sequence of actions or inactions that allowed the situation to develop and disaster to occur. It helps crowd managers identify where risk can be mitigated within distal causes in order to prevent a proximate cause ever developing. It gives evidence that the development of positive attitudes towards safety; in design, planning, training and operations can only improve the health and wellbeing of staff and crowds which in turn improves business and experience for attendees.
A century of crowd disasters
Crowd experts have compiled and shared databases of disasters over the last century in the form of excel sheets, online lists (Still, 2018) and interactive maps (Asgary, 2018). Tables listed in this paper are extracted from the database compiled of disasters spanning from 1988 to 2018. In the last thirty years, this database recorded a total number of 148 crowd disasters resulting in the death of 10,976 people and 13,381 injuries.
From the database analysis, the below list identifies types of events that would motivate people to attend or be the reason a crowd formed. These include Religious, Sport, Political, School, Entertainment, Transport and Shopping.
What external threats affect a crowd event?
Threats identified below refer to any external factor that is out of control of or unbeknownst to the event organiser or those responsible for the environment that which the crowd gathered. These are listed in the table below with a description for each threat.
So what is the most common threat to my event?
According to the table above, the most common threat to a crowd is ‘accidental fire’, followed by ‘external communications’. Identifying the most common external factors likely to impact a crowd allows a crowd manager to pay particular attention to them in the event risk assessment, developing robust contingency plans.
The responsibility of the event manager
It is evident, that although external threat is out of the control of the organiser, the ability to prepare and respond is not. For example, where fire is an external threat that is out of the control of the organisers, it is also a risk that should be addressed, as it is likely to happen if protective measures are not put in place. Considering distal and proximate causes of fire related disasters, investigations show that although the proximate cause was external, the distal cause included ignorance by management to ensure a venue or space was up to building code standards and escape routes fit for purpose and capacity.
For example, if pyrotechnics were used in a venue for a music event, then the surrounding area would need to be fireproof, have adequate space to allow for any fallout or excessive heat and safety measures put in place to protect attendees from potential harm. In the case of the Porto Alegre in Brazil in 2013 (Flor, 2013) the proximate cause were the pyrotechnics setting the surrounding area on the stage alight. However, exploring the distal causes highlighted the fact that the venue was overcrowded, exceeding licence condition limits, and the emergency exits were not fit for purpose. In 1990, when an arson attack set the Happy Land venue on fire in New York (Blumenthal, 1990), 87 people died as a result of the inability to escape. Prior to the event, the venue had been condemned due to failure in building code standards. There was only one entrance and exit and it was where the fire had started.
Had the venues been fit for purpose, appropriate risk assessments been produced for these events and action taken, this disaster may not have occurred, or the severity of impact could have been limited. Reflecting on distal causes, if management do not have a respectful attitude towards safety, this will have a knock on effect on all operations within an event or venue. For example, in the case of Happy Land, the building had already been condemned and the venue owners should have not allowed the event to take place.
Yet not all distal causes are due to ignorance of management. Perhaps a risk assessment was produced for an event and necessary measures put in place, however, the site is safe in good weather but with the addition of heavy rain and pressure of crowds on grass causes a new situation to develop, causing significant risk for crowds. In the case of the Guns n’ Roses concert in Donnington, England 1988 (Upton, 1995); the stage was at the bottom of a hill and after heavy rain, deep mud formed at the front of the stage. This environment paired with a dense crowd attending a concert resulted in the death of two people. Had the land been dry, this cause, removed from the situation, could have narrowly avoided the disaster.
It could be suggested that if a dynamic risk assessment been carried out on the site, management could have identified the dangerous patch of mud developing and put in measures to secure it from crowds or make the ground safe to stand on. This is another example of how an external threat can negativity impact an event, but if management are not responsive in mitigating risk or putting contingency plans in action, then the risk continues to grow before it reaches a critical point and results in catastrophe.
Preventing future disasters
As of yet, there is no legislation or law that enforces those who are responsible for the management of a crowd to have appropriate credentials or proven experience. This has naturally resulted in unregulated and inexperienced organisers in the industry. However, those responsible for crowds can continue their training and development to improve planning.
With the correct attitude to crowd safety, attention paid to regulating density and the producing of robust operational and contingency plans, can event managers reduce risk and protect lives. Venue owners and those responsible for crowds are the only parties that can prevent crowd crushes from occurring (Pearl, 2015).
Develop contingency plans
Using a simple structure, such as the DIM-ICE model when planning an event or space where a crowd will congregate supports the development of contingency plans arising from risk assessments. Developing contingency plans for each phase of a crowd – ingress, circulation and egress
The most common threats to a crowd have been identified. If these are addressed in all crowd risk assessments, and contingency plans created, it increases the understanding and responsibility of management and greatly reduced risk to the crowd. Robust contingency plans need to be produced for every threat to a crowd and these must be agreed with agencies, communicated clearly and tested with all those involved in executing the plan.
Graham (2016) categorises risk into ten families including external, legal and regulatory, strategy, organisational, operational, information, human resources, technology, political, financial and administrative risks. These families of risks offer the crowd manager a useful structure when assessing risk to a crowd.
Predict and Prevent!
Simple measures can be put in place now to prevent further unnecessary pain and loss to human lives around the world. The right attitude of management and training of those responsible for crowds is essential. Only those with experience and appropriate training should be permitted to manage the safety of crowds. The tools available are inexpensive and easy to use, the suggestions and structure in how to approach a crowd risk assessment are being shared. There exist too many disasters in order to learn from hindsight and not allow a repeat.
Crowds in their very nature will always carry risk, however, what is predictable is manageable. With experience, training, information and most of all, the attitude to ensure the safety of all who attend, can crowd disasters be avoided and attendees kept safe.
Asgary, D. A. (n.d.) World Crowd Disasters Web V1. yorku.maps.arcgis.com. [Online] [Accessed on 7 December 2018] http://yorku.maps.arcgis.com/apps/webappviewer/index.html?id=e7c52856187642e19bd227865393432c.
BBC News (2000) Zimbabwe Stadium Deaths Inquiry. news.bbc.co.uk. 10 July. [Online] [Accessed on 10 December 2018] http://news.bbc.co.uk/1/hi/world/africa/826951.stm.
Ben Doherty (2010) Cambodian stampede: Phnom Penh counts the cost of water festival disaster. The Guardian. 23rd November. [Online] [Accessed on 11 October 2018] http://www.theguardian.com/world/2010/nov/23/cambodia-stampede-rainbow-bridge-water-festival.
Blumenthal, R. (1990) Fire in the Bronx. nytimes.com. 26 March. [Online] [Accessed on 11 October 2018] https://www.nytimes.com/1990/03/26/nyregion/fire-bronx-87-die-blaze-illegal-club-police-arrest-ejected-patron-worst-new-york.html.
Flor, A. (2013) Nightclub fire kills 233 in Brazil. Reuters. 27 January. [Online] [Accessed on 11 October 2018] https://www.reuters.com/article/us-brazil-nightclub-idUSBRE90Q04T20130127.
Fruin, J. J. (2002) The Causes and Prevention of Crowd Disasters. www.crowdsafe.com. 17 February pp. 1–10. [Online] [Accessed on 18 February 2016] http://www.crowdsafe.com/fruincauses.pdf.
Graham, G. (2016) REAL RISK MANAGEMENT. 4 May pp. 1–10. [Online] [Accessed on 3rd January 2018] http://www.lexipol.com/wp-content/uploads/2016/05/Lexipol_Real_Risk_Management_Part_1.pdf.
Hillsborough Independent Panel (2012) Hillsborough. House of Commons, pp. 1–389.
Helbing, D. and Mukerji, P. (2012) ‘Crowd disasters as systemic failures: analysis of the Love Parade disaster’. EPJ Data Science, 1(1) pp. 1–40.
HSE (2004) Investigating Accidents and Incidents. Liverpool: HSE Books, pp. 1–88.
Pearl, T. H. (2015) ‘Crowd Crush: How the Law Leaves American Crowds Unprotected’. SSRN Electronic Journal pp. 1–43.
Lee, R. S. C. and Hughes, R. L. (2006) ‘Prediction of human crowd pressures’. Accident Analysis and Prevention, 38(4) pp. 712–722.
Fricke, D. (2000) Nine Dead At Pearl Jam Concert. Rolling Stone. 17 August. [Online] [Accessed on 10 December 2018] https://www.rollingstone.com/music/music-news/nine-dead-at-pearl-jam-concert-235167/.
Pearse, D. and Weaver, M. (2009) Fire in Bangkok nightclub kills scores of New Year’s Eve revellers. The Guardian. 1st January. [Online] [Accessed on 11 October 2018] http://www.theguardian.com/world/2009/jan/01/bangkok-nightclub-fire-deaths.
Seabrook, J. (2011) Crush Point. The New Yorker. 7 February. [Online] [Accessed on 11 October 2018] https://www.newyorker.com/magazine/2011/02/07/crush-point.
Sime, J. D. (1999) ‘Crowd facilities, management and communications in disasters’. Facilities, 17(9/10) pp. 313–324.
Still, G. K. (2014) Introduction to Crowd Science. CRC Press.
Still, P. D. G. K. (2016) ‘3.3 Catastrophe Theory’. MSc Crowd Safety and Risk Analysis pp. 1–7.
Still, P. D. G. K. (n.d.) Crowd Disasters. [Online] [Accessed on 2nd December 2018] http://www.gkstill.com/ExpertWitness/CrowdDisasters.html.
Smith, R. A. (1995) ‘Density, velocity and flow relationships for closely packed crowds’. Safety Science, (18) February, pp. 321–327.
The Guardian (2014) 16 dead at South Korean pop concert after ventilation grate collapses. The Guardian. 17 October. [Online] [Accessed on 11 October 2018] http://www.theguardian.com/world/2014/oct/17/14-feared-dead-south-korea-pop-concert-4minute.
Upton, M. (1995) ‘INCIDENT AT DONINGTON MONSTERS OF ROCK 1988 ’, pp. 1–11.
Watson, J. (2000) Guards Blocked Exit in Mexico City Disco Fire. ABC News. 22nd October. [Online] [Accessed on 10 December 2018] https://abcnews.go.com/International/story?id=82313&page=1.
Zheng, X., Sun, J. and Zhong, T. (2010) ‘Study on mechanics of crowd jam based on the cusp-catastrophe model’. Safety Science. Elsevier Ltd, 48(10) pp. 1236–1241.