Project: Preventing Passenger Ejection from Buses, Coaches and Minibuses

Reference: S0320/VE

Last update: 14/03/2008 12:36:37


The objectives of this project are to:

* determine the frequency, cause and level of injuries in major bus and coach crashes in UK and Europe;
* determine the effectiveness of various restraint systems in holding occupants within the residual space (defined by UN ECE Reg 66);
* identify occupant restraint systems (compartmentalisation, seat-belts, airbags, etc) and quantify the associated risk of ejection or partial ejection taking into account the emergency exit requirements and the fitting of side windows with breakable glass often carried in easily displaced rubber surrounds;
* propose practical methods for increasing the protection of occupants and to cost those methods;
* assess the impact of proposals on other safety considerations existing in the bus and coach operating environment; and
* produce an initial and, subsequently, a partial Regulatory Impact Assessment (RIA) of the effect of introducing each of the safety improvements developed above.


Major accidents involving large passenger vehicles can often result in a roll over. In order to reduce injuries, initial investigation and analysis centred on the vehicle structure and its ability to resist severe deformation. This gave rise to a requirement (UN ECE R66) that the structure should be sufficiently strong to resist crushing and provide a survival space within the vehicle for passengers. This regulation was incorporated in the EC Directive 2001/85, which deals with vehicles of EC categories M2 and M3.

However, the problem remains of passengers being ejected or partially ejected through side windows and subsequently crushed by the overturning vehicle.

Accident statistics and crash testing of cars has demonstrated the improved occupant survival rate associated with the wearing of seatbelts. As a result, legislation for the installation of seat belts in coaches was put in place and legislation governing the wearing of those belts follows.

Buses, with a requirement for frequent passenger movement, are not required to have seatbelts, and in coaches there remains the risk of serious injury when the vehicle rolls onto its side and slides. This continuing risk was illustrated in research undertaken by the department as part of a full review of UN ECE R66 by the UN ECE members. Windows break or are ejected completely and passengers occupying the adjacent seats, whether belted in or not, come into contact with the ground sliding past. The object of this research project is to consider ways in which passengers can be contained within large passenger vehicles subjected to lateral rotation and forward motion. In particular the current philosophy of fulfilling the emergency exit requirements by fitting side windows of breakable glass will be examined.

The usage of emergency exits and their siting will be analysed, and alternative methods of providing such exits will be studied and costed in conjunction with proposals for preserving the integrity of the vehicle envelope.

The outcome will be recommendations for improved legislation, designed to eliminate a remaining cause of serious injury in large passenger vehicles.


Cranfield Impact Centre Ltd
Wharley End, Cranfield, Bedfordshire, MK43 0JR
+44 (0)1234 750111

Contract details

Cost to the Department: £138,400.00

Actual start date: 01 February 2004

Actual completion date: 31 January 2006


Preventing Passenger Ejection from Buses, Coaches and Minibuses. Final Report
Author: Cranfield Impact Centre
Publication date: 20/08/2006

Summary of results

  1. The findings of the research are as follows:
    . For Great Britain, for the 11-year period 1994 to 2004 inclusive, it can be calculated that bus and coach occupants accounted for:
    o 0.49% of all 'killed';
    o 1.45% of all 'killed or seriously injured'; and
    o 3.07% of 'all severities' (killed, seriously injured and slightly injured)

    . It was estimated that, for Great Britain, a maximum of some 3 fatalities and 184 serious injuries per year are due to full or partial ejection from coaches and minibuses. It is not possible to distinguish the individual contribution of coaches and minibuses to this total of ejection related casualties from the available sources.

    . The literature search indicated various technologies that have been considered for reducing passenger ejection from car side windows. Of these, the use of seat belts and bonded laminated side glazing was of most relevance to this study in that these fulfilled the projects' mandate to identify practical and cost-effective means for limiting passenger ejection.

    . Finite Element analyses indicated that:
    o For the coach model:
     Toughened glass was unable to offer any retentive capability for the unrestrained occupant;

     3-point seat belts were more effective than 2-point belts during the selected rollover condition

     Installation and routing of the shoulder belt on the seats nearest to windows should be considered, as outboard installations may be vulnerable to damage from the ground or other intrusion during rollover;

     The use of a seat side bolster, as a form of compartmentalisation for an unrestrained occupant, restricted the occupant from possible ejection, but redirected the occupant into a severe head impact with the overhead luggage rack;

     Laminated glass bonded to the vehicle rather than toughened glass mounted in a flexible gasket demonstrated that the strength limits of the adhesive bond were always exceeded and this caused the window to be released from the structure. It was concluded that standard laminated glass which was adhesively bonded to the structure was insufficient to retain an unbelted occupant within the vehicle under the accident conditions simulated here.

     A series of modifications to improve the fixing strength of the window to the vehicle proved unsuccessful. This was believed to be due to the deflection of the structure around the window during the impact. Ultimately, the inclusion of a bolting system as well as adhesive bonding preserved the bond, but transferred the failure into the laminated glass around the bolts. On this basis, the use of standard laminated glass as a mandatory recommendation was not put forward.

     However, the results indicated that the voluntary use of laminated glass may be beneficial over toughened glass in rollover (and other) accidents where the impact severity was not as high as that used in the simulations conducted here.

     The parametric study also indicated that the reduced height of current window designs may also have some benefit in retaining an unrestrained occupant although it is likely that they will be redirected to an impact with some other object within the vehicle.

    o For the minibus model:
     The occupants were not ejected but were subject to impacts with the interior of the vehicle.

    . It was concluded that 3-point inertia reel seat belts remain the best option for preventing ejection and partial ejection.

    . With the introduction in 2006 of compulsory seat belt wearing in the UK on coaches and minibuses for all occupants over 3 years, the project concluded with the two following recommendations:

    o That future coach and minibus accident data should be monitored following the introduction of compulsory seat belts wearing to measure the effect that this has on rollover ejection casualty numbers

    o As part of the above, current ECE R-66 technical requirements should be reviewed with regard to the increased energy that passengers wearing seat belts will impart to a vehicle's structure during a rollover accident. If excessive roof crush is observed from accidents involving R-66 compliant vehicles (and where passengers wore their seat belts), it may be necessary to increase the energy and strength requirements of roof structures to resist increased roof loading from belted passengers.