Crashworthiness standards may block locomotive approvals

Next month, the German standards authorities are due to consider the adoption of a local edition of the provisional EuroNorm prEN 15227 covering the crashworthiness of rail vehicles. Already being applied to many builds of locomotives and multiple-units, this norm is due to be adopted as a full standard in the second half of this year, and each country is currently preparing a 'mirror' specification embodying the key rules in its own language.

The meeting on April 10 is a critical deadline for both train operators and manufacturers. The standard as currently adopted would prohibit the future development of several successful locomotive types used predominantly for freight traffic by open-access operators amongst others.

Efforts to get the standard amended last year failed narrowly, so attention is now focused on getting an exemption clause included in the preface to the German version. This would have an immediate impact for the market as many of the locos are used in Germany.

Protecting passengers

The crashworthiness specifications outlined in prEN 15227 are intended to enhance the safety of passengers and train crew, particularly in the event of a collision. Thus the theoretical survival area, for example part of the driving compartment, must remain intact following a collision. It is also highly desirable that the axles should not leave the rails, or that the amount by which they do so is strictly limited, so as to minimise the risk of over-riding.

Development of the crashworthiness standard began with accident data collected in 1991-95 by a number of European railways (BR, CFL, DB, DSB, NSB, SJ, SNCB and SNCF). This data was summarised - albeit in an anonymous form - in a report by the former European Railway Research Institute1. Using this data, ERRI defined four reference accident scenarios2. But the study only looked at passenger trains, as the equivalent data on freight train accidents had not been collected.

From 2003 standardisation comm-?ittees began looking at issues of structural integrity and load assumptions, using provisional EN 15227. This in turn led to the publication in April 2005 of a first draft for prEN 15227. The initial draft contained many data errors and was widely interpreted as a 'ranging shot' from which the final standard would evolve. A first round of consultation elicited a great deal of comment which was fed back to the committee.

When a Comments Resolution Meeting was held in Brussels in February 2007, it was decided that a second round of consultation was not required, despite the substantial differences between the first draft of prEN 15227 and the version under consideration at the time.

Short locomotive problem

However, Vossloh Locomotives GmbH had been undertaking its own crash simulations with various locomotive types, and had discovered that the proposed norms posed significant problems for shorter locomotives with a fixed wheelbase or short bogie centres. These findings were published in May 2007, and notified to the standards committee. But when the adoption of prEN 15227 was discussed on September 5 last year, the manufacturers of small locomotives were outvoted. Only Vossloh and Voith lodged dissenting votes, as the third supplier Gmeinder participated as a guest.

As it currently stands, prEN 15227 considers the crashworthiness of rail vehicles under four possible collision scenarios:

• a head-on collision between two identical train formations;
• a head-on collision between the train and a different type of rail-mounted vehicle;
• a collision between the train and a large road vehicle on a level crossing;
• a collision between the train and a smaller obstacle, such as a car on a level crossing, an animal or debris.

These four scenarios were chosen to replicate the most common accidents which had led to personal injuries on passenger trains.

When applied to locomotives, the standard train formation consists of the locomotive hauling an 80 tonne wagon, and the 'different rail-mounted vehicle' for Scenario 2 is also represented by an 80 tonne wagon.

The specified limit values were clearly designed for relatively long high-performance locomotives used on passenger trains, but they do not take account of the normal operating conditions for a short locomotive with a centre cab. Such locos are not normally used in passenger service, and thus the only person likely to be affected in the event of a collision is the driver, who already benefits from the protection afforded by the central cab.

However, many of the locomotives affected are used on the main line for local freight movements, or to deliver and collect wagons at local sidings. This requires formal authorisation, which would no longer be available if the German Federal Railway Office (EBA) insists on full compliance with EN 15227.

For example, under Scenario 1, Chapter 6.2 on over-riding sets close tolerances on the maximum distance that wheelsets can be permitted to lift off the rails during a collision. According to Vossloh's simulations, the limiting values cannot be met by a short locomotive. Taking a three-axle locomotive with a mass of 50 tonnes and a length of 11 m, the back-and-forth motion set up by a Scenario 1 collision could see one set of wheels lift off the rails by as much as 800 mm, compared to the 21 to 100 mm limit in the standard. This is a physical effect based on the design of the loco (Fig 1).

Since the risk of overriding increases dramatically with the momentum to be exchanged, short, hard collisions are less risky. The energy exchange in collisions as foreseen by EN 15227 leads to violent vertical movements of the locomotive. Thus the danger of over-riding increases. The use of A and C-buffers in accordance with UIC 526-1 has provided sufficient momentum exchange capability, and the manufacturers argue that safety would not be enhanced significantly by the application of EN 15227 to this type of locomotive.

Analysing the risk

To support its argument for an exemption, Vossloh has analysed EBA accident data from 1997-2006, looking in detail at the number and type of incidents, the type of locomotive, and the outcome of each accident. During this period there were 64 144 accidents covering all types of traction, resulting in 9 607 injuries and fatalities (Table I).

Of this total, only 624, or 0·87%, involved centre-cab locomotives. And little more than half of these - 351 - involved collisions with other rail-mounted vehicles. Limiting the analysis to the safety of train crews or locomotive drivers showed that there were only six serious injuries, 24 minor injuries and no fatalities over an entire decade that could be attributed to collisions involving centre-cab locomotives (Table II).

Further assessment of the individual accidents found that there were only two involving locomotives colliding with a rail-mounted vehicle under scenarios 1 and 2 of prEN 15227, which resulted in two serious injuries and no fatalities.

Despite all efforts on the part of the manufacturers, short locomotives cannot fulfil the current requirements of EN 15227 due to their design. But at the same time the accident data shows that for locomotives of this design the problem is relatively inconsequential.

Given that these locomotives are rarely involved in the type of accidents that the standard was designed to protect against, and have a very good safety record in terms of both incidents and personal injury, Vossloh Locomotives has therefore proposed that a preface should be added to exempt such locomotives from EN 15227 (panel). This is due to be presented to the German standards committee meeting on April 10.

Whilst the objective of improving safety for rail users and staff is clearly welcome, this example shows that prEN 15227 and the test scenarios are not strictly applicable to all locomotives or vehicle types. The areas of application and the limit values have to be considered carefully in their application to ?normal usage.

Proposed exemption

'Locomotives with a central cab which are also to be used on main lines (but not at the head of loaded passenger trains) are excluded from the demands of this standard due to their technical and operational characteristics. The design categories and collision scenarios defined in the standard are not valid for these vehicles.'

Table I. EBA accident data 1997-2006*

* Accident data includes incidents of trespass and covers the German federal network. Private and industrial railways are excluded. Reporting methods changed from 2000 onwards to include data on 'near-miss' incidents where safety standards were breached but no accident occurred.

Table II. Accidents involving centre-cab locomotives

• CAPTION: Vossloh Locomotives' G12 design will replace the existing G400B and G765 families. Photo: Vossloh Locomotives
• Fig 1a. Distribution of static forces
• Fig 1b. Energy absorption and reaction forces
• Fig 1c. The locomotive leaves the rails as a consequence of the collision forces

References

1. ERRI report B 205.1/DT 357.
2. ERRI report B 205.1/RP 1.