By Darren Crossman



Attending further education as a mature student whilst working full time in the events industry and having aspects of study interrupted by a global pandemic has been difficult and challenging. It is only with the support of my family, my colleagues and my employers that I have been able to complete the course, and gain experience in an emerging field of study.

To my loving wife Donna, thank you for all of the support that you have given and the time that you have sacrificed while I have been studying.

I would like to thank my Employer; The Welsh Rugby Union for their financial support, and the patience that they have shown as I have progressed through the course and to my colleagues who have had to listen to me vocally exploring theories and practices on crowd safety management. I hope that I now provide better advice, support and planning though the competition of this course.

To my fellow students, without whom the group discussion both on campus, off campus and electronically have helped to challenge accepted theory and practice and have enabled context to be added to the tutorials, bringing the subject matter to life.

I’d like to thank the interview participants, without whom, this work would not have been possible, your input has been invaluable and gratefully received at a particularly challenging and difficult period in time.

Finally, I would like to thank all those involved in the design and delivery of the Crowded Places and Public Spaces course at Coventry University for their hard work and enthusiasm. The course design based on emergency planning and disaster management has been an eminently sensible approach.


In 1973 the first Guide to Safety at Sports Grounds (SGSA 2018), known as the Green Guide was published with the subsequent revisions to the current sixth edition providing the basis of flow rates for calculating crowd movement. In addition the Guide To Fire Precautions In Existing Places Of Entertainment And Like Premises (HMSO 1990) – known as the Primrose or Yellow Guide provides further guidance but limited to a single flow rate.

The event safety guide, known as the Purple Guide (EIF 2015) refers users to the flow rates contained within the Green Guide. The guide only provides two rates to use, the first for level surface and the second for a stepped surface. Given that all aspects of the crowd related industry make use of this information, it is extremely limited and does not consider any of the variable that exists in the different locations.

This study aims to challenge the existing levels with the need to establish a broader framework that will be more representative of the flow rates within crowds, based upon the influencing factors. It has looked at what other factors need to be considered and included in the framework for a more realistic calculation of flow rates in crowd related environments and lays the groundwork for building the framework.

Further research and development will be required to populate the numerical data into the framework to allow it to become a tool to be used for accurate calculations and guidance.


  1. Introduction
    • Background
    • Research Aims and Objectives
  1. Literature Review
  2. Methodology

          3.1 Overview

          3.2 Research Methods

          3.3 Research Ethics

          3.4 Limitations and Bias

          3.5 Interview Process

  1. Results and Analysis

          4.1 Overview

          4.2 Results

  1. Discussion
  2. Summary, Conclusions and Recommendations

          6.1 Summary

          6.2 Conclusion

          6.3 Recommendations

  1. References
  2. Notes
  1. Introduction


1.1 Background

When it comes to calculating the flow rates of crowds there has not been much in the way of advancement in terms of usable information for those working in managed crowd related events. Crowds are complex entities that require detailed planning and management, and it’s often perceived that they are simply controlled, when in fact the management takes more detailed planning and consideration, and the control is a small fragment of a crowd safety management plan (Martella et al. 2017). The United Kingdom has developed a series of working documents and legislation over a number of years that provides guidance on planning for crowds and what flow rates should be used in calculations for movement.

However, these documents do provide some form of guidance on crowd flow rates, e.g., the eighty-two persons per metre per minute as identified by the Guide to Safety at Sports grounds (SGSA 2018), but other than a single variation there are no accounting for factors that may affect this figure, hence it has become widely used as a default calculation.  Much of the data comes from Fruins (1971) level of service concept, however given that Fruins work is fifty years old it is reasonable to assume that more analysis and interpretation of the aspects and impacts are required,  but the fundamental basis of the work that was carried out is still valid.

Whilst Fruin looked at the components of a crowd disaster (Still, 2015) to understand crowd movement and flow, there was limited appreciation of the external and internal influences which have a positive or negative effect on the crowd, but Fruin went further by looking and incorporating the anthropomorphic characteristics of human into his work. Even if this work had accounted for such influences at the time, there has most likely been a significant number of changes to crowd influences since it was undertaken.

Fundamentally this means that there a potential a gap in the available published data and the usable information for planning crowd movement.

Hypothesis: that there is no single characteristic that can be used to create a crowd flow model, rather each case needs to be treated on its own merits and that ultimately a framework or a table of factors needs to be created with a formula that will allow the crowd safety manager/ planner to safely plan the egress/ evacuation of a given place using the variables that will be present in that place. It is surmised that practitioners need a framework or a tool that can be used without the need to undertake academic research, but with the knowledge that the elements of the framework have been researched and have peer review support and acceptance.

1.2 Research Aim and Objective

Before the aims and the objectives are stated, it is important to highlight that this paper may be seen as self-contradictory, as it is an academic piece of research aimed at being used by practitioners.  The aim of this study is to analyse the needs of the sector, currently available information on crowd flow calculations used to create a framework that may be used in any given scenario to understand how a crowd is likely to move, and to provide information on potential danger points within a given route or venue. The Hypothesis is that there are a range of factors which will influence the flow rate of pedestrians moving at different phases of an event, which should all be considered to predict the overall movement pace.

To achieve the aim of this study the following objectives have been established:

  • In the first instance, review what guidance already exists and challenge any potential shortcomings.
  • Secondly research from primary and secondary sources, what they considered influences on crowd movement are.
  • Third, to review how the modern world has changed and incorporate any influences on crowd movement.
  • Fourth, to create an initial framework that may be used to construct bespoke models that will outline how a crowd is likely to move, assessed against all influencing factors.

2. Literature review


The available guidance (mainly for United Kingdom consumption) comes in the form of the Guide to Safety at Sports Grounds (SGSA 2018) and the Fire risk assessment series of publications (Fire Safety Risk Assessment 2006). However, the Guide to Safety at Sports Grounds is severely limited in its advice and guidance and only provides information for two flow rates but it differs greatly (being much lower) to the flow rates in the fire safety risk assessment guidance creating confusion between the documents. This literature review seeks to examine the work that is available and may be linked used to assist in the calculation of crowd flow rates.

The Built Environment

consideration is required of  how the built environment has an impact on pedestrian movement (Martella et Al 2016) which may present obstacles, pinch points, corners etc a point further illustrated by by  the examination of how flow rates are affected by interaction of people with spaces and buildings (Shahoseini et al 2019) and how flow rates can be significantly reduced by restrictions.

General forward motion

Designing for Pedestrians: A level of service concept (Fruin 1971) has widely been used as the basis for crowd movement calculations but has not been challenged but reviewed and assessed that the crowd behaviour, size and environment could all have an impact (Haghani et al. 2019) . In a challenge these concepts, the “movement adaption model” (Thompson et al 2019) takes account of the space occupied by each person whilst walking, and the space left between potential points of contact with others in the crowd. However, this study was limited to direct travel and did not take account of any lateral movement or sway and subsequent work to create a holistic model was required. 

Surfaces and the impact on Gait

The identification of walking surfaces being widely different in terms of size, composition, and environment and when combined, the factors influence how humans walk with some being more challenging than others to maintain stability or gait (Thomas et al 2020).  The authors explore the theory of complexity of surfaces and the impact that it has on the ability to walk across single, multiple, and changing surfaces. Their findings indicated that where surfaces need to be observed to safely walk, there is less observation ahead of the person and will thus result in a change of pace. The study confirmed that complex surfaces are seen to be more challenging to stability whist walking across them and when it comes to steps or stairs, additional care would need to be taken, which would again have an effect on speed of pace, gait, and thus, flow rate. 

The impact of surfaces should be examined, alongside suitable preparations made for changes in weather that may affect the surface (Martella et al 2016). Further to this, the safe movement of the crowd should be considered not only at the ingress and egress phases but throughout the event, and that the designed infrastructure and crowd behaviour must be factored into the planning process for safe crowd movement (Martella et al 2016).

Directional change impact within Crowds

Observing pedestrian footprints in the snow is like the streamlines of fluid (Helbing 1998), who analysed the movement of pedestrians in a crowd and found other synergies, e.g., to the dancers on a dance floor who take more space than those surrounding the edge of the dancefloor, due to the rapid direction changes and the less than single directional choice. Further analysis showed when the elements of a crowd all move in a single direction and in a formation like group the directional changes are reduced, which results in a steady level of velocity or flow.  Helbing (1998) presented that the reaction between pedestrians in a crowd is very important as when the pedestrians are in a high-density crowd and there is a slight change of direction or pace at any given point “pedestrian jams”, or increased densities and slower flow rates are likely to occur behind the direction of travel. This highlights how, even when a calculation has been made based upon standard information, without considering the effects within a crowd, the flow rates can change quickly having a significant ripple effect downstream. Helbing (1998) identified that pedestrian flow will demonstrate the characteristics of waves throughout its composition, where the peaks will be the denser, more slow-moving parts of the crowd and the troughs will be the faster, less dense areas of the crowd, with the overall goal to achieve a uniform appearance and thus flow rate.  Fu et al (2016) examined the effects of ‘speed matching’ in pedestrian flow, with their hypothesis being that a pedestrian changes their speed to maintain pace with the surrounding pedestrians which impacts the density of the crowd. This work goes some way to explaining how the waves that Helbing (1998) described, due to changes in pace and increases/ decreases in density and flow rate. Smith (1995) contributed to the theory of higher speeds resulting in increased density and increased density results in lower speeds, but this falls short of looking at any other influences on a crowd.  Supporting work on the wave theory (Ivancevic and Reid 2011) does recognise the additional effects of cognitive, motivational, sociological and emotional effects that may be influencing movement.

Crowd Psychology and Subgroups

Sime (1995) briefly touches on crowd psychology and mentions the psychology of a crowd should be factored in when designing for any movement as their historical evidence from disasters that assumes that the crowd is acting as a single entity and not as individuals or subgroups.  Taking subgroups (Drury and Reicher 1999) as an element that needs further investigation and separating from the now debunked theories of those such as Le Bon (2002) and the early theories of the crowd being a single mindless entity.

Templeton et al (2018) investigated the psychological impact of groups moving together within a crowd. One of their first moves was to distinguish the different types of crowds; for instance, a crowd of shoppers in a mall have a similar purpose but are not unified, whereas a crowd attending a concert have the same purpose and are unified by their aim. They theorised that social identity of a group within a crowd could have an effect on the crowd density and therefore, on the flow rates of the overall crowd. In the observations Templeton et al (2018) found that those outside of the ‘in group’ had to manoeuvre around the group creating faster walking speeds on the outside of the group, whilst the group moved at a much slower pace. This has the initial impact of highlighting that there are different flow rates within a crowd, on a flat level surface – purely driven by the psychological aspect of the crowd. Further discussion highlights how computer models and modellers have not yet accounted for the psychological aspect in their models, which would indicate that a computer model is not an accurate model. This work highlights the subgroups that are likely to be active within a crowd, and will have an effect on the flow rates, particularly at the ingress phase. Supporting the sub-group theory Drury and Reicher (1999) evaluated the social identity within a crowd and how similar goals and aims would unite individuals with the crowd space and how they are likely to act together but reject external influences from outside of their social group, the ‘ingroup and the outgroups’ theory (Reicher 2001).

Crowd Behaviour and Movement

Crowd disasters which result in death or serious injuries resulting from crushing or trampling are not the preserve of emergency situations but may occur at different types of events (Sports, Music, religious gatherings, etc) during the main phases; ingress, circulation, egress (Al-Gadhi 1996).  Such incidents may be down to poor design where the limits of crowd flow exceed the space available for the given crowd to move beyond a specific point or series of points in a given time period, otherwise defined as “the rapid coming together of a group of persons in a restricted space with sufficient force to cause human injury or death (Fruin, cited by Al-Ghadi 1996). Therefore, the understanding of crowd dynamics and the typographies (Berlonghi 1995)  behaviour of crowds is essential in planning for movement.

Al-Gadhi (1996) stated that there are three component parts; crowd characteristics, facility design and layout, and management practices that influence crowd behaviour echoing the sentiments of Sime (1995). Crowd behaviour can be attributed and influenced by the nature of the event, which leads to the need to understand the characteristics of a crowd for the planning of a safe event. Al-Gadhi (1996) suggested that further research to develop useful means for predicting crowd behaviour would be required. Such a recommendation could be interpreted as the need for a universal framework or modelling tool to understand the relationship between crowd behaviour and crowd flow, and in particular what design features any crowd modeller should be aware of when planning for the movement of their crowd.  For instance , pedestrians have a sense of task that enables them to reach their common destination and the speed at which they travel is determined by their surroundings, and that they try to reduce their travel time (Hughes 2003). While he focusses on the movement of the crowds, he considers the behavioural aspects of crowds in high density and low-density scenarios. In both cases he references that there is little exploration of this, with most work being conducted at more common densities. He concludes that this field of research is still in its infancy (in 2003) and again references that more needs to be learned about crowd behaviour as it has a direct effect on flow. Helbing et al (2014) consider that dense crowds can lead to involuntary bodily contact, and the transfer of force between bodies, emulating the domino effect. They refer to this as crowd turbulence, which if left unchecked or without management can be the beginning of crowd disasters, linking to the work of Al Ghadi (1996) allowing for sufficient space in the infrastructure. 

The Effect of mobile phones on walking

Society has changed significantly since the original flow rates that many work to were introduced, and arguably one of the biggest influences in this change has been the introduction of the smart phone. These devices have been embraced by large areas of the population and provide easy access to social media applications. Through clever design the applications have enticed people to become almost addicted to them, generating a real fear of missing out (Servidio 2021) if they do not have ready access to instant updates on what is happening in their social circles. As most people now carry these devices, there is likely to be an impact on how they move as an individual and as a group and thus how they impact on crowd movement. It has been identified that this distraction physically impacts gait and speed (Crowley et al 2019) as well as leading to detrimental impacts on awareness of the surroundings. The investigations found that there was a marked effect on speed when the user was texting or talking on the phone compared to just walking with the biggest decrease in speed being during texting. Crowley et al (2019) observed that walking whilst being distracted, could be connected with slips, trips and falls even when there is no impairment to the cognitive function (e.g., though alcohol or drugs).

The Effect of Intoxication on pedestrian Movement.

This area of the literature review is surprisingly light on research. There is plenty of information on the effect of intoxication on pedestrians involved in accidents but limited work on the actual movement of intoxicated pedestrians.


As shown, there is a mass of supporting information but no central framework to join all of the theories together, supporting the need to research the field potentially leading to the design of such a framework.

3. Methodology

3.1 Overview

Consideration was given to which methodology to select for research purposes. At this stage of the research the initial thought of selecting a quantitative approach was disregarded as there was insufficient insight (Barnham 2015) into the potential influences on crowd flow rates that the use of qualitative data may provide. The use of a qualitative methodology allowed the research to examine the observations of professionals that are involved in the crowd safety sector.

3.2 Research Methods

The research was undertaken via series of interviews conducted with crowd safety managers who have experience across the industry. The interviews were conducted via video conferencing with transcripts being made of the conversations to enable coding analysis (Gibbs 2007) to be undertaken post interview. The coded data was then analysed to evaluate the data for discussion and outcomes.

3.3 Research Ethics

The research has been conducted ethically to ensure that the set criteria are maintained during the study to avoid any harm, physical or otherwise being inflicted though physical or mental means (Wisker 2018). All of the research participants were asked for their consent to be interviewed and for their responses to be used in the analysis (CU ETHICS 2021). All participants were free to withdraw from the research programme at any time and have their submissions withdrawn.

3.4 Limitation and bias

It is recognised that there will be some limitations to this work and there is the possibility of some bias based upon pre-existing working knowledge, experience, and practices, however a conscious effort to eliminate this bias has been made to allow for an open interview and research process.

3.5 Interview Process

A number of potential participants were approached for interview; however the responses were limited. This being the case and with the potential pressures on operational matters, the research proceeded with the limited numbers. A semi-structured interview process was selected to allow both interviewer and interviewee to expand on the baseline questions and to explore the subject matter in some detail. The pre-determined questions provided a framework for discussion but did not use the traditional question and answer format, rather it allowed for the interviewee to elaborate on their observations and experiences and allowed the questions to fall into the conversation at the right moment to maintain the flow and thought processes. In the classification of questions (Booth 2005) these would fall into the exploratory type. The interviews were conducted via video conferencing in place of face-to-face conversations and consisted of an ‘ice breaking’ introduction session before the main body of the interviews began. The interviews were recorded to allow the maximum focus on the conversation with the transcripts (annexes) being created post interview.  Post interview, the transcripts were used for the qualitative analysis. Table 1 (below) gives a brief outline of the participants operational experience.

Table 1 Showing participants experience in the field and qualification for interview.

4. Results and Analysis

4.1 Overview

The analysis has been conducted using software that looks for similar themes within the transcripts of the interviews. This allowed the interviews to be examined and highlight where there were repeated experiences on the same subject.  The analysis was conducted by coding the transcripts. The choice of coding fell to selecting the inducting methodology (Yi 2018) as the subject matter is known, but in this case exploratory research is being conducted, therefore a codebook could not be created and used to evaluate the data for recurring themes and potentially conflicting themes although, the transcripts were manually coded to ascertain commonalities and differences in the statements. The participants did on the whole echo each other’s sentiments, however there was a notable difference between the practitioners who plan for and then deliver crowd safety management on the ground and the crowd modellers who plan for the movement of crowds using knowledge, simulation, and computer simulation. 

4.2 Results

The first and perhaps most expected finding was that there is very little in the way of guidance that is in circulation or used outside of the Guide to safety at Sports Grounds (SGSA 2018) otherwise known as the Green Guide, the Purple Guide (EIF 2015) which tends to be used more on green field sites and generally in conjunction with the Green Guide, and the fire safety guidance (HMSO 1990). Taking account that the participants operate primarily in the UK but do operate and have experience operating in a global capacity, the same guidance tends to be used.

There is an acknowledgement that aspects of the guidance that has been created is perhaps on the back of flawed science but has found its way into circulation and therefore requires crowd managers to follow it, irrespective of whether it has a sound background or not. All of the participants agree that the current guidance is not necessarily fit for purpose.

It’s identified that the current guidance provides figures on flow rates, which crowd safety professionals are aware are the maximum rates that can be achieved, but potential clients are not. This may cause issues when a crowd movement plan is put together as the clients’ expectations may not be met by the plan, which is most likely to take a safer and perhaps more realistic view of how the crowd is likely to move.

Demographics is where the participants disagree, most agree that the demographic of a crowd, based on what the event is, and the likely attendees will have an effect on how it will move, but there was disagreement from one participant who would be looking at more generic behaviours from the crowd. For crowds with a younger demographic the movement is likely to be relatively fast, but for crowds with an older demographic the reverse can be said.

Behaviour featured in all participant answers, and it was expressed that behaviour could be linked to a number of other factors which would have an impact on the crowd and the behaviour may change in response to external stimulus or influence. Most cited sporting events, where the crowd is made up of regular spectators who know where they are going and how to get to their destination, need little in the way of supervision and behave the same way event after event and enter the venue quite quickly and without support but, when it comes to different events such as concerts, the demographic changes and as there is little likelihood of the spectators having been regular attendees to the venue, then flow rates are likely to be lower than a sporting event. Culture is also likely to play a part in how crowds move, as a crowd located in one country may have different ethics and routed behaviour and be more compliant than a crowd in a different country or part of the world.

Turning to the external influences, most respondents agreed that the surface that the crowd would move over would have a big impact on the speed that it moves. A hard flat and level surface is likely to produce a constant flow rate, whereas a surface that has a gradient or changes type is likely to cause the crowd to slow down. The examples that were given included beaches, where the crowd may move along easily navigated hard sand and then come across soft sand which will slow down their movement. This effect on the rate of arrival and the rate of departure from a given point will cause the crowd to increase in density and may have an upstream effect in a similar manner to phantom traffic jams on highways.

Again, there was agreement that the guidance is flawed as it does not take into account scenarios such as greenfield sites, where the surface can change rapidly. A flat, firm and level surface can become a sticky muddy surface or a puddle very quickly if it rains, which changes the way that the crowd interacts with it and the speed at which they move.

Weather was further cited due to its effect. In permanent fixed locations such as stadia, it is likely that rain will drive spectators to move more quickly to get into the shelter of the building, whereas the greenfield sites or open-air sites may see people seeking shelter or holding an acceptance that they are going to get wet and continue to move a unified pace. The time of the event was also identified as having an effect, daytime events where vision is good often present better flow rates than night-time events, where vision is impaired.

All agreed that the social identity of a crowd does play a significant part in the way in which it moves. Generally, in the events sector, crowds have a unified purpose (whether it is watching a game of football or whether it’s watching their favourite artist) they have a synergy with all of those around them.

When it came to discussing technology there was a difference of opinion. Some felt a crowd could be encouraged to follow direction which would result in better movement, whilst others suggested that people may be distracted by the applications on the phone and the need to switch between the applications to get the result they were looking for. The power requirement was also highlighted, that if an event was relying on technology for items such as e-tickets, should a mobile phone lose power it would slow entry as the owner would need more attention from the staff.

The changing social landscape and the desire of people to show what they are doing to the world at the time that they are doing it has created some differences in the way that people watch events. For example, in a greenfield site, they are far more likely to move from stage to stage, taking images at each one to show where they have been. In addition, at events where tickets are limited, not enabling large groups to attend, groups are forming through the use of social media who have similar intentions and interests.

Examining how other factors may affect the abilities of spectators the use of alcohol and drugs needs to be evaluated, as they are likely to be widely used. The use of alcohol or drugs is likely to impair their movement and their cognitive abilities to respond to surroundings and their physical abilities to move, which again will result in variation in their predicted movement.  Further impacting their movement will be the duration that they are in attendance, for instance if they are entering a venue early in the day, they are more likely to be able to move quickly but if they have been on their feet all day, in a queue for example then their movement is likely to be more sedentary. 

5. Discussion

Having examined the available guidance on flow rates, it can be said that it is limited at best and is considerably lacking with regards to review and currency with the modern word and levels of understanding that those managing crowds possess.

This field has had little research conducted in preceding years but recently the importance of maintaining the safety of humans in crowded places has been seen as becoming more important and research has been increasing (Hagani and Sarvi 2017). In fact, much of the guidance and legislation, certainly in the United Kingdom is created from so called ‘tombstone’ legislation emerging from disasters where people have lost their lives e.g., the Hillsborough disaster in 1989 and the subsequent report (Taylor 1990) or the Bradford fire that led to the Fire Safety and Safety of Places of Sports Act 1987.

With the changes in technology and a growing interest and general knowledge in the field of crowd safety, research has gathered pace and more literature is being produced, however there is still little or no research and guidance that links all of the aspects of a crowd together and connects it to how crowds move. The participants of the research have identified from their observations and experiences that they are relying on their own knowledge gained over a number of years with the guidance holding little influence.

Taking this into account It is clear that a better methodology of calculating a flow rate needs to be established for safer crowd movement planning and management. The following breakdown of items is intended to form the basis of a suggested framework for calculating the overall movement capability of a crowd. It is noteworthy that all aspects are inter-related and interdependent and should not be treated in isolation from each other.

Figure 1: showing the interconnected relationship between aspects affecting flow rates.

Suggested Framework Composition:

Type of event

The type of event is likely to provide an indicator to the planner of what each spectator is going to bring with them, for example a person attending a festival that spans several days will need some form of shelter and somewhere to sleep which means that they will be carrying a reasonable sized bag and depending on the weight of it and the distance that they have walked, may cause fatigue and slow their pace. Conversely a summer concert in a stadium is likely to attract lightly dressed individuals who only need to be prepared for a few hours in a protected environment.


This is perhaps the most well researched aspect of the discussion and features heavily in crowd dynamics (Still 2015).  All available spaces, routes obstacles and access points must be accounted for, for a measured space to allow a safe calculation of how many people can pass through the space. The information presented, signage, direction, stewards on the ground providing direction, toilets, doors, gates, turnstiles etc all need to be considered as individual points within the whole scope of any assessment.  The rate at which they can pass will then rely on the actual flow rate, but again this section links intrinsically to all of the other sections and cannot be treated in isolation.


We know from the existing guidance that limited accountability has be shown to the surface that spectators are moving over, and as the participants have identified, the surface becomes extremely important. Detailed attention is required to the type of surface that the crowd is to move over and how it will impact their gait, their speed, and their lateral movement. Take the example cited by participant eight, where overnight a puddle had formed, the perceived width available was still present from a planning perspective but the crowd were using a much smaller space due to the presence of the water, which links the surface to the weather forecasts. In other locations the surface may have an impact on the speed of movement based on what footwear the pedestrians are wearing, which links across to the demographic aspect of planning. Further to this, the current guidance accounts for steps but does not account for gradient, where an uphill gradient will undoubtedly slow movement, whereas a downhill gradient is likely to increase speed for certain demographics but may actually cause a slower movement for the more infirm demographic.


The weather conditions and the time of year associated with the conditions will have positive and negative effects of crowd movement. A summer event will see people in a pleasant environment, where they are warm and comfortable, and are likely to move at a steady but less rushed pace, but if rain is introduced into this scenario then they are likely to move at a much faster pace to seek shelter. Winter conditions and the cold are likely to see the crowd prepared to move at a faster pace to keep warm but form a personal space perspective will have a minor increase on the space each person required due to the change in clothes (larger coats etc).  Weather links back to surface as it can have a direct impact and change the way crowds move, a dry firm surface can quickly become a muddy, boggy surface when it begins to rain as has been seen to great effect at Glastonbury Festival.

An example of surface change after bad weather (Mud at Glastonbury Festival – In Pictures 2021)


The composition of the crowd will have a marked effect on the way in way in which it moves and how the flow rates are affected (Larsson et al. 2021). As many events, whether it be a sporting event, music, or other type of event, they develop a fan base that is likely to support of follow them. This being the case as the years pass, the followers are likely to age which may influence mobility and speed of movement.  Further to this will be the gender orientation of the crowd and whilst some sporting events have a changing gender profile from their spectators, other events such as concerts may attract predominantly gender biased crowds. Take for example the reference made by participant nine, to the horse racing crowd and the spectators wearing high heels. Clearly, something as simple as this type of footwear will have a noticeable effect on the way in which they walk and how their movement will affect those around them.

Social Identity

Accounting for the impact of social identity is something that has largely been ignored to date (Moussaïd et al. 2010) no matter what type of crowd it is related to as the spatial movement of a group will depend on its function and goals. As has been identified by the participants, the social identity of the crowd is likely to have an effect other the movement, whether it be individual attendees, small or large groups or the groups that from through a shared identity and purpose once inside the venue, all will have their own purposes which will have an impact on how others respond and react to them.  This invokes the groups within groups theory (Drury and Reicher 1999) and how understanding and planning to manage the crowd and sub crowds becomes extremely important.

Mental Capabilities

As a generalisation, it can be said that spectators at most crowd related events may develop a cognitive impairment of some form or another. Whether this is through the consumption of alcohol, drugs, euphoria related to the event, fatigue from the length of the day, distractions such as technology or any other form, it will have a detrimental effect in the way in which they move. Within the scope of this research, it is more likely to be levels of euphoria, intoxication and excitement that affect their movement, and this is likely to be linked to the social groups that the spectators are a part of (again reinforcing the links between the factors).

External Influence

It would be remis not to include the external influences that may cause an issue with the crowd. Such things may include responses to social media and as the participants mentioned, monitoring social media to get the ‘atmospherics’ of the crowd can play an important part in safe management. A rogue message can create mayhem as was seen in Oxford Street, London 2017, when there was a false reporting of a shots fired that cause people to flee and respond in the fight or flight manner. This highlights how quickly and simply a wrong piece of information can spread so quickly that it’s difficult to maintain management of the situation.



Overall Picture

Throughout this discussion there is a thread that links all of the elements that are being considered, with none being able to be considered in isolation. These interconnected relationships give support for the need for such a wide-ranging framework to consider flow rates as the failure to consider one aspect will have a negative effect on the overall result (Reason 1990). An example of a suggested way of using the framework model is located at annex A which requires further design and development to create a usable tool.

6. Summary, Conclusions and Recommendations

6.1 Summary

Still (2015) gave us the DIM-ICE meta model and paying due cognisance to crowd dynamics, each aspect of the crowd movement should be examined and calculated. As this work is extremely detailed only the ingress phase of crowd movement is being examined in this study, but the other elements should also undergo the same reviews. Clearly the established flow rates do work in certain scenarios and provide a base to work from, but they were created some time ago and the understanding of crowd movement and crowd behaviour has changed significantly since the inception of this data which means that crowd safety professionals are often working outside of the guidance to establish the correct flow rates for their environment. It also provides inexperienced crowd planners with an expected rate of low that may not be achieved, and cause flow related issues at other points within the crowd route system.

6.2 Conclusions

It is concluded that the flow rates that are generally used, do not accurately represent how a crowd moves and requires detailed review and renewal. A simple one size fits all approach, is not suitable for an industry where there is as much difference in the events and locations as there is in the people who attend them. A detailed process to evaluate the most likely composition of the crowd, their behaviour and influences is lacking from the current guidance and planning processes. All too often, a one size fits all approach is demonstrated and although on most occasions there are no significant occurrences, this may be more by luck than by detailed planning. The information that is available and the knowledge level of crowd professionals is far beyond what is essential breaking every crowd down to two speeds of movement. It would be remis not to plan in much more detail and could eventually put lives at risk. 

6.3 Recommendations

This research challenges the conventional approach to the use of a single set of guidance that provides flow rates to the events sector, which does not account for the detail and movement within a crowd and whilst this research has established an outline framework to use when establishing crowd flow rates, further research is required. The next phase is recommended to be observational with measurement at different types of events with different types and constitutions of crowds to enable population of the suggested framework.

The table below is the next likely step in the development of this framework. Research will be required to ascertain the data to support each aspect of the table in a numerical form and the formula will need to be developed to provide the interactions between the cells. One or more columns may be required for each aspect, depending on the complexity of the route and the time of day that the crowd is to be moving.

In some cases, more than one table will be needed to show different entry profiles, if there is repeated entry at different times, for example where a concert has support acts and the main act, these will most likely require a flow rate to be calculated for each performance time. Festivals may need to consider this as different acts head to different stages and the audience level of interest in the artist.

Sporting events may need to consider that with ‘double header’ games or long duration events with aspects that pique interest from the spectators, different entry profiles will be required thought the day.

However, as has been the theme throughout this research, such a framework must be used with caution and as guidance only although it may produce more accurate guidance for planning purposes than the current level of guidance that is available.

7. References

(Using the Harvard Referencing Style)

Greenwood T (1990) A Guide To Fire Precautions In Existing Places Of Entertainment And Like Premises  London: HMSO Books

Al-Gadhi, S. (1996) ‘A Review Study Of Crowd Behavior And Movement’. Journal Of King Saud University – Engineering Sciences 8 (1), 77-107

Armstrong, K. (2021) Keeping Up With The Crowd [online] available from <> [15 May 2021]

Au, S., Gilroy, J. and Haslam, R. (2011) ‘Assessing Crowd Dynamics And Spectator Safety In Seated Area At A Football Stadium’. Pedestrian And Evacuation Dynamics, 663-674

Shmidt F L & Hunter J E (2015) Methods of Meta-Analysis: Correcting Error and Bias in Research Findings Third Edition Sage Publications , 513-551 DOI:

Berlonghi, A. (1995) ‘Understanding And Planning For Different Spectator Crowds’. Safety Science 18 (4), 239-247

Bøhling, F. (2014) ‘Crowded Contexts: On The Affective Dynamics Of Alcohol And Other Drug Use In Nightlife Spaces’. Contemporary Drug Problems 41 (3), 361-392

Booth, A. (2006) ‘Clear And Present Questions: Formulating Questions For Evidence Based Practice’. Library Hi Tech 24 (3), 355-368

Crowley, P., Madeleine, P. and Vuillerme, N. (2019) ‘The Effects Of Mobile Phone Use On Walking: A Dual Task Study’. BMC Research Notes 12 (1)

Coventry University (2021) Ethics [online] available from <> [20 June 2021]

Drury, J. and Reicher, S. (1999) ‘The Intergroup Dynamics Of Collective Empowerment: Substantiating The Social Identity Model Of Crowd Behavior’. Group Processes & Intergroup Relations 2 (4), 381-402

Drury, J. and Reicher, S. (2021) Collective Action and Psychological Change: The emergence of new social identities [online] available from <> [16 May 2021]

Filingeri, V., Eason, K., Waterson, P. and Haslam, R. (2017) ‘Factors Influencing Experience In Crowds – The Participant Perspective’. Applied Ergonomics 59, 431-441

Fire Safety and Safety Of Places Of Sports Act  (1987) London: HMSO

Fire Safety Risk Assessment (2006) London: Department for Communities and Local Government Publications

Fruin, J. (1971) Pedestrian Planning And Design; Metropolitan Association Of Urban Designers And Environmental Planners Inc. New York

Fruin, J. and Strakosch, G. (1987) Pedestrian Planning And Design. Mobile, Alabama: Elevator World

Fu, Z., Luo, L., Yang, Y., Zhuang, Y., Zhang, P., Yang, L., Yang, H., Ma, J., Zhu, K. and Li, Y. (2016) ‘Effect Of Speed Matching On Fundamental Diagram Of Pedestrian Flow’. Physica A: Statistical Mechanics And Its Applications 458, 31-42

Gayathri, H., Aparna, P. and Verma, A. (2017) ‘A Review Of Studies On Understanding Crowd Dynamics In The Context Of Crowd Safety In Mass Religious Gatherings’. International Journal of Disaster Risk Reduction 25, 82-91

Gibbs, G. (2007) Analyzing Qualitative Data 38-55. California:Sage Publications

Guide To Safety At Sports Grounds (2018) Sports Ground Safety Authority, 6th edn

Haghani, M. and Sarvi, M. (2018) ‘Crowd Behaviour And Motion: Empirical Methods’. Transportation Research Part B: Methodological 107, 253-294

Haghani, M., Sarvi, M. and Shahhoseini, Z. (2019) “When ‘Push’ Does Not Come To ‘Shove’: Revisiting ‘Faster Is Slower’ In Collective Egress Of Human Crowds”. Transportation Research Part A: Policy And Practice 122, 51-69

Helbing, D. (2021) A Fluid Dynamic Model For The Movement Of Pedestrians [online] available from <> [10 March 2021]

Helbing, D. and Johansson, A. (2021) Pedestrian, Crowd, And Evacuation Dynamics [online] available from <> [10 March 2021]

Helbing, D., Brockmann, D., Chadefaux, T., Donnay, K., Blanke, U., Woolley-Meza, O., Moussaid, M., Johansson, A., Krause, J., Schutte, S. and Perc, M. (2014) ‘Saving Human Lives: What Complexity Science And Information Systems Can Contribute’. Journal Of Statistical Physics 158 (3), 735-781 doi:10.1007/s10955-014-1024-9

Helbing, D., Johansson, A. and Al-Abideen, H. (2007) ‘Dynamics Of Crowd Disasters: An Empirical Study’. Physical Review E 75 (4) doi: 10.1103/physreve.75.046109

Hogg, M. and Tindale, R. (2001) Group Processes. Malden, MA: Blackwell Publishers

Hughes, R. (2003) ‘THEFLOW OFHUMANCROWDS’. Annual Review Of Fluid Mechanics 35 (1), 169-182 doi:10.1146/ANNUREV.FLUID.35.101101.161136

Ivancevic, V. and Reid, D. (2011) ‘Turbulence And Shock-Waves In Crowd Dynamics’. Nonlinear Dynamics 68 (1-2), 285-304

Kelly, C., Tight, M., Hodgson, F. and Page, M. (2011) ‘A Comparison Of Three Methods For Assessing The Walkability Of The Pedestrian Environment’. Journal Of Transport Geography 19 (6), 1500-1508

Larsson, A., Ranudd, E., Ronchi, E., Hunt, A. and Gwynne, S. (2021) ‘The Impact Of Crowd Composition On Egress Performance’. Fire Safety Journal 120, 103040

Le Bon, G. (2002) The Crowd. A Study Of The Popular Mind. New York, United States: Dover

Martella, C., Li, J., Conrado, C. and Vermeeren, A. (2017) ‘On Current Crowd Management Practices And The Need For Increased Situation Awareness, Prediction, And Intervention’. Safety Science 91, 381-393

Minegishi, Y. (2021) ‘Experimental Study Of The Walking Behavior Of Crowds Mixed With Slow-Speed Pedestrians As An Introductory Study Of Elderly-Mixed Evacuation Crowds’. Fire Safety Journal 120, 103041

Moussaid, M., Helbing, D. and Theraulaz, G. (2011) ‘How Simple Rules Determine Pedestrian Behavior And Crowd Disasters’. Proceedings Of The National Academy Of Sciences 108 (17), 6884-6888

Moussaïd, M., Helbing, D., Garnier, S., Johansson, A., Combe, M. and Theraulaz, G. (2009) ‘Experimental Study Of The Behavioural Mechanisms Underlying Self-Organization In Human Crowds’. Proceedings Of The Royal Society B: Biological Sciences 276 (1668), 2755-2762

Moussaïd, M., Perozo, N., Garnier, S., Helbing, D. and Theraulaz, G. (2010) ‘The Walking Behaviour Of Pedestrian Social Groups And Its Impact On Crowd Dynamics’. Plos ONE 5 (4), e10047

The Guardian (2021) Mud At Glastonbury Festival – In Pictures [online] available from <> [12 July 2021]

Timmermans H (ed.) (2009) Pedestrian Behavior Models, Data Collection and Applications Emerald  doi:10.1108/9781848557512

Polit, D. and Beck, C. (2010) ‘Generalization In Quantitative And Qualitative Research: Myths And Strategies’. International Journal Of Nursing Studies 47 (11), 1451-1458

Servidio, R. (2021) ‘Fear Of Missing Out And Self-Esteem As Mediators Of The Relationship Between Maximization And Problematic Smartphone Use’. Current Psychology

Shaaban, K. and Muley, D. (2016) ‘Investigation Of Weather Impacts On Pedestrian Volumes’. Transportation Research Procedia 14, 115-122

Sime, J. (1995) ‘Crowd Psychology And Engineering’. Safety Science 21 (1), 1-14

Smith, R. (1995) ‘Density, Velocity And Flow Relationships For Closely Packed Crowds’. Safety Science 18 (4), 321-327

Still, G. (2015) Introduction To Crowd Science. Boca Raton, Fla: CRC Press, Taylor & Francis Group

Taylor, P. (1990) Lord Justice Taylor: The Hillsborough Stadium Disaster Inquiry Final Report. London: HMSO

Taylor, P. (1990) The Hillsborough Stadium Disaster: 15 April 1989 : Inquiry By The Rt Hon Lord Justice Taylor : Final Report : Presented To Parliament By The Secretary Of State For The Home Department By Command Of Her Majesty January. London: HMSO

Templeton, A., Drury, J. and Philippides, A. (2018) ‘Walking Together: Behavioural Signatures Of Psychological Crowds’. Royal Society Open Science 5 (7), 180172

Reason, J. (1990) ‘The Contribution Of Latent Human Failures To The Breakdown Of Complex Systems’. Philosophical Transactions of the Royal Society B Biological Sciences 327 (1241), 475-484

Thomas, N., Gardiner, J., Crompton, R. and Lawson, R. (2020) ‘Physical And Perceptual Measures Of Walking Surface Complexity Strongly Predict Gait And Gaze Behaviour’. Human Movement Science 71,

Thompson, P., Nilsson, D., Boyce, K., Molloy, M. and McGrath, D. (2020) ‘Exploring The Biomechanics Of Walking And Crowd “Flow’. Fire And Materials 44 (6), 879-893

Wisker, G. (2018) The Undergraduate Research Handbook. London: Macmillan Education UK

Yi, E. (2018) Themes Don’t Just Emerge — Coding The Qualitative Data [online] available from <> [4 July 2021]

Zhang, X., Weng, W., Yuan, H. and Chen, J. (2013) ‘Empirical Study Of A Unidirectional Dense Crowd During A Real Mass Event’. Physica A: Statistical Mechanics And Its Applications 392 (12), 2781-2791

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.

Translate »