Improving Evacuation Safety of Construction Sites

Construction Site Evacuation Safety
evacuation strategies for tall construction sites

May 2016 - May 2019

Funded by Institution of Occupational Safety and Health (IOSH)

Click here to see FSEG YouTube channel playlist featuring presentations and results from the construction site project.

In London alone there are an estimated 250 high-rise building projects planned over the next few years which will contribute a significant proportion of the projected £21 billion per year construction spend. A typical project, such as the £400 million ‘100 Bishopsgate’ building will have a peak workforce of some 1500 workers onsite with a total number of workers onsite through the life of the project estimated to be 12,000. Based on these figures it is estimated that the total number of workers expected to be exposed to construction sites in London during the lifetime of construction for these 250 projects can easily exceed 1 million people.

A significant risk to the health and safety (HS) of workers on construction sites is emergency evacuation associated with fire or some other on-site emergency. Evacuation during the construction phase of a building is one of the most challenging evacuation scenarios particularly given the ever changing nature of the construction site. Nevertheless, over the past 50 years around the world, very little research has focused on this topic. Clearly, during construction the evacuation plans and procedures for the completed building are not applicable. The layout of the building and even its interconnectivity may be changing on a daily basis requiring that evacuation plans are adapted and conveyed to the construction workers on a regular basis.

While construction sites may undertake regular evacuation drills as part of HS practice, these are seldom unannounced evacuation drills. Thus workers are usually aware that a drill will take place on a given day and in some cases at a given time. Thus workers can prepare for the drill and may pre-empt the evacuation. This inadvertently reduces the realism of the drill and it fails to test construction site workers’ knowledge of the evacuation process, the effectiveness of the procedures in place and the effectiveness of the training processes employed.

Construction of a high rise building.

This project will be the first systematic fundamental study into the evacuation behaviour and performance of workers on construction sites and will provide important insight into, and quantification of, the likely behavioural responses of construction workers during evacuation situations. The project will also provide a unique data-set for the validation of evacuation models that can be used in the development and optimisation of evacuation procedures for construction sites. With this knowledge it will be possible to frame more realistic and achievable evacuation procedures improving the safety of construction workers and as a result emergency responders.

Project Objectives

The project’s objectives are to establish a unique evidence base characterising, for the first time, the actual performance and behaviour of construction workers during emergency evacuation. This information combined with computer simulation can be used to inform the development of more reliable evacuation procedures improving the safety of construction workers.

Objective 1
Develop an understanding of how perception of risk may influence evacuation behaviour on construction sites.

Objective 2
Develop an understanding of the level of construction worker knowledge of evacuation procedures on various construction sites and how this is influenced by type of construction, phase of construction and population demographics.

Objective 3
Collect human performance data characterising the evacuation behaviour of construction workers, including response times, movement rates, wayfinding abilities and preparedness to use hoists or lifts for evacuation.

Objective 4
Provide evacuation data that could be used to validate evacuation models specific to construction sites.

Objective 5
Through the use of evacuation modelling utilising data collected in this project, demonstrate how evacuation procedures for construction sites can be optimised

Objective 6
Through the better understanding of construction worker evacuation behaviour and the optimisation of evacuation procedures, provide improved certainty of the outcome of evacuation situations, enabling a more efficient and safer response from emergency services.



Several views of a construction site in Melbourne showing the construction of a high rise building. The floors shown have three types of surface, steel beam lattice, metal decking and metal decking with rebar. Several workers can be seen on the surfaces. Evacuating from these surfaces poses several difficulties that need addressing.

Project Methodology

The project has five distinct tasks, Planning and Preparation, Data Collection, Data Analysis, Model Calibration, and Optimising Evacuation Procedures:

1. Planning and Preparation
A key component of the project will be the planning of four unannounced evacuation trials of construction sites. The unannounced nature of the trials will ensure that the behaviour of the workers is as realistic as possible. FSEG have considerable experience in undertaking these type of evacuation experiments and have safely undertaken unannounced evacuations of many complex buildings. In 2013 FSEG successfully completed a pilot trial in collaboration with Brookfield Multiplex at their Aldgate Tower high-rise construction site to demonstrate the feasibility of such evacuation trials.

The sites will be selected in consultation with the project advisory board, and will be examples of two types of construction sites at two differing phases of development.

The worker’s performance and behaviour during the evacuations will be collected using strategically located video cameras. Questionnaires will be distributed at the end of the evacuation to supplement the video data. The questionnaires will assess risk perception, knowledge of the evacuation procedures and other subjective experiences of the participants.

2. Data Collection
Each of the four construction sites will be instrumented with video cameras that will record the performance and behaviour of the participants. The cameras will be set up at a time when there are very few people on site e.g. very early in the morning. A team of FSEG researchers, with assistance from the site management staff will conduct the surveys of the construction workers after the evacuation. They will remain on-site for the day assisting the workers to complete the surveys during their scheduled work breaks. Furthermore, the surveys will also be provided on-line for completion at a later stage.

Different types of flooring exists during the various stages of development. Each poses its own difficulties in people's movement and evacuation.

3. Data Analysis
The video footage from the trials will be analysed frame by frame to extract human performance and data appropriate for the validation of evacuation models.

Data from the questionnaires will be statistically analysed to determine how perception of risk influences evacuation behaviour on construction sites. The human performance data will be invaluable in defining how construction workers actually behave during on-site evacuation, as opposed to how site managers think they behave, or would like them to behave. As such the proposed dataset will be unique and represents the only data of its type available anywhere in the world. This information will be extremely useful in its own right and will enable site managers to make more informed decisions when it comes to planning onsite evacuation procedures.

All the collected data will be made publicly available via the FSEG web site and through the publication of academic and professional papers and reports.

Part of the FSEG and Multiplex Team

4. Model Calibration and Validation
The human performance data will be used to calibrate the buildingEXODUS evacuation modelling software. For this software to be able to make reliable predictions of human behaviour during evacuation it requires reliable human performance data describing parameters such as; response times, travel speeds, wayfinding choices, etc. Other software calibration may be required to ensure that features that currently function for completed buildings, such as lifts can accommodate on site variations such as the use of hoists for evacuation. The validation dataset will then be used to assess the performance of the calibrated buildingEXODUS evacuation model in performing evacuation simulations of construction sites by comparing its predictions to the observed evacuations.

5. Optimising Evacuation Procedures
The validated model will then be used to assess the evacuation procedures of a Multiplex construction site and the model used to suggest procedural improvements to the evacuation process. The project team will work closely with the construction site team to ensure that the proposed optimised procedures are practical and viable. The successful completion of this project will demonstrate the viability of applying evacuation modelling tools to construction sites, thereby improving the planning and reliability of evacuation procedures and ultimately, the safety of construction workers.

Project Findings

1. Project Results
The project has now been completed and successfully met all of its six objectives. In total four full-scale evacuation trials and the five walking speed experiments were undertaken. The data-set generated from these nine trials involving 1,072 participants incorporates around 2,200 data points, extracted from 3 GB of video data, and information from 61 worker questionnaires.

Inside the formworks

The data-set involves:

the evacuation of 920 participants
the measurement of 920 exit times
the measurement of 275 response times
walking experiments involving 152 participants
the measurement of 545 walking speeds over four different types of surfaces
the measurement of 126 stair walking speeds on two different types of stairs
the measurement of 59 ladder speeds
the measurement of 203 interpersonal distances on temporary stairs.

Equipment used in data collection

The analysis of this data has produced a unique evidence base characterising, for the first time, the actual performance and behaviour of construction workers during emergency evacuation. The evidence base consists of:

response times for workers in the main building and the formworks, as measured from the sounding of the alarm in the main building,
worker walking speeds on different types of surfaces, such as concrete, decking and decking with rebar, and
worker ascent and descent speeds on temporary dog-leg and parallel scaffold stairs and ladders.

Analysis of Response Times


Walking speed and stair speed trials

Measurement of ladder ascent/descent speeds

The data have also been incorporated in the building evacuation simulation tool buildingEXODUS, providing it with a unique capability to simulate evacuation from high-rise construction sites. The performance of the software has been validated using measured data collected from the trials. The validated software has been used to explore how evacuation procedures for high-rise construction sites can be improved, including the impact of reducing worker response times, replacing ladders with temporary scaffold stairs within the formworks, and using hoists to assist in evacuation.

Validation Case Geometry

Validated buildingEXODUS used to explore how hoists can be used for construction site evacuation

2. Conclusions
The use of the evidence base and the modelling software will inform the development of more reliable evacuation procedures, improving the work environment, through better preparation for, and management of, on-site emergency evacuation, and advancing the safety of construction workers. Potential uses of the evidence base and modelling approach include:

addressing limitations, assumptions and omissions in guidelines and regulations, including those produced by the Health and Safety Executive, through the incorporation of the evidence base.
use of the evidence base by construction site health and safety managers to inform training of workers and the formulation of best practice.
use of suitably validated modelling tools by construction site managers to define enhanced evacuation procedures.

3. Useful Links

The full report can be downloaded from: www.iosh.com/constructionevacuation
Construction site evacuation simulation videos and this video can be found on the FSEG YOUTUBE Channel at: https://www.youtube.com/user/FSEGresearch
The High-Rise Construction Site Evacuation Modelling Validation Data-Set can be downloaded from: http://fseg.gre.ac.uk/validation/building_evacuation/

Project Partners

Project Funding

Advisory Board

Multiplex
London Fire Brigade
Build UK
Loughborough University (representing IOSH)
Fire Safety Engineering Group (University of Greenwich)

Further Information

Prof. Ed Galea
Fire Safety Engineering Group
University of Greenwich
Greenwich Maritime Campus
Old Royal Naval College
Queen Mary Building
Greenwich SE10 9LS
UK
Tel: +44 (020) 8331 8730
e-mail: E.R.Galea@gre.ac.uk