The cost of failure

As you may have noticed, The Rail Engineer likes to feature upbeat articles that focus on rail engineering successes and the deployment of clever new technology. Fortunately such stories are not hard to find. Triumphant conclusions to highly complex projects are the norm rather than the exception. Everyone expects excellence, and indeed why shouldn’t they?

Sometimes, though, things can go badly awry and this article takes a look at the heavily delayed and problematic introduction of new high-speed diesel multiple units (DMUs) in Denmark – the infamous IC4 trains. An independent report published in December 2014 attempts to set out a way forward for this beleaguered project.

No one sets out to create an exhibit for the rail engineering cock-up hall of fame. It’s tragic, therefore, that a brave engineering endeavour like IC4 should find itself a contender for that dubious honour. Serious delays and a host of technical shortcomings have caused the IC4 project to become a contentious issue within Danish politics. For those of us who are not involved, the technical issues behind a failed project can be interesting and we can gain insight. As ever, though, when things don’t work out, there is a price to pay. Failure always comes at a cost!

IC4

As part of its ‘Good trains for everyone’ plan (‘Gode tog til alle’), the Danish national railway operator Danske Statsbaner (DSB) ordered 82 IC4 4-coach trains (and 23 IC2 2-coach trains) from Italian train builder AnsaldoBreda in 2000, to be delivered from 2003 onwards. Each IC4 train comprises four articulated coaches of lightweight aluminium alloy construction. The trains were designed for a top speed of 200km/h (120mph). Braking is achieved by means of a blended mix of hydrodynamic braking and pneumatic braking. There is also a magnetic track brake.

Delayed

The IC4 fleet was intended to replace the existing IC3 trains used on the country’s trans- Great Belt routes. Unfortunately, completion was delayed by ten years, with the last of the IC4 trains only arriving in Denmark in October 2013. Because their intended routes are set to be electrified, this has reduced the life cycle of the fleet from 25 years to 15.

Numerous technical problems have further delayed the introduction of the trains and only 67 sets are currently approved for passenger use, of which just 12 are in regular passenger service. Indeed, five of the new trains have been cannibalised for spare parts. By October 2014, the entire fleet had covered just 10.8 million km, 30% of which had been completed during 2014.

DSB’s choice of a heavily-customised train has been criticised as being a major reason for the delays. The Nordic minimalist design philosophy of DSB’s own designers was paired with Italian chic from the design studios of Pininfarina. Only four trains were running by the end of 2007 (in regional service) but even they were suspended from service in February 2008 because of exhaust fume problems and many minor faults.

Ultimatum

In June 2008, DSB gave AnsaldoBreda an ultimatum whereby at least 14 trains had to be approved and ready for regular service before  May 2009 as otherwise the contract would be cancelled. This ultimatum was met, following which DSB announced it had reached an agreement with AnsaldoBreda on the delivery of the remaining trains. The final delivery date was to be extended to 2012, AnsaldoBreda would drop all further development, and all subsequent train sets would be identical to the then-current test train. Any future upgrades would be performed by DSB and AnsaldoBreda was to pay DSB compensation of 2 billion kroner. Together with previously paid compensation fees, this meant that half the original value of the contract was refunded.

Also, for what seems to be a bizarre reason, the Danish IC4 fleet remains one train short. The missing set does exist, however. It can even be seen on Google Earth at its current resting place in Tripoli, Libya (at 32.828791°N 13.111563°E). AnsaldoBreda and the then Italian Prime Minister Silvio Berlusconi gifted the train in 2009 to Libyan dictator Muammar Gaddafi to mark the fortieth anniversary of his rise to power! An ambitious scheme to construct standard gauge lines in Libya had been started in 2008, but the project stalled because of the 2011 civil war leaving the IC4 train with nowhere to go.

PROSE report

Now an independent report produced by Swiss rolling stock engineering company PROSE Ltd, in association with Swiss Federal Railways (SBB) and Enotrac, has assessed whether it is worthwhile for DSB to continue spending resources on IC4, given the expected 15-year operating life of the trains. It concludes that DSB can hit its target of having 74 of the IC4 trains in service by 2019, with a reliability figure of 20,000km between failures.

The main functions that need to be improved are as follows:

• The IC4 sets can’t work in tandem to provide the required train capacity, because the coupling system doesn’t work properly.
• The brake blending and wheels slide protection systems are deficient, meaning that a speed restriction has been imposed on the entire fleet.
• Cracking has been found in axle box housings.
• The power packs have several integration problems and component defects which have reduced the availability of the train sets.

Coupler

The Prose report highlights several problems associated with the Dellner inter-unit coupler system. There is no suggestion within the report that Dellner Couplers itself has been at fault; rather it was the way in which the train manufacturer incorporated the Dellner coupler system into the overall design concept, and especially the train computer and monitoring system. Deployment

of the IC4 trains on their intended long distance routes, e.g. Copenhagen-Aalborg, has been hampered because these services require 8-car units at peak times. The 4-coach IC4 trains have therefore tended to be used on regional services, for which they are unsuited.

When not in use, the coupler is hydraulically retracted and protected by sliding covers. The system is supervised by sensors and is controlled by the train computer. It has been decided that the sliding covers will be replaced by a fixed gaiter solution to reduce the complexity of the system. Three prototype gaiters from different suppliers have been trialled. The version selected is actually from Dellner itself and uses the same high-end silicone as the company’s gangways. Dellner has successfully supplied similar gaiters in Norway and Sweden.

According to the Prose report, the electrical connections will not separate in about 20-30% of uncoupling operations. An investigation by Dellner, working in conjunction with DSB, found that there was no problem with the movement of the electrical coupler itself. However, it was the solid attachment of the cables to the car body that was restricting the couplers’ movement. This has been overcome by using a sliding clamp which allows the cables to move in and out when the electrical couplers are moving.

Incorrect assembly of the guiding arms that open and close the cover for the electrical couplers contributed to the problem. Replacement of these moving covers and adoption of the new Dellner gaiter will provide better protection in the future.

Brakes

There are two issues concerned with the IC4 braking system. Poor braking performance on slippery rail has meant that a reduced maximum speed limit has had to be imposed, which limits the ways in which the trains can be deployed. There is also a reliability issue associated with some of the braking system components.

As built, each axle has independent Wheel Slide Protection (WSP) control, except axles 5 and 6 (on a trailer bogie) which are both controlled by one Brake Control Unit (BCU). During November 2011, two IC4 sets failed to stop at signals. As a result, the maximum brake percentage was reduced from 170% to 130% and the maximum operating speed was lowered. In addition, axle 5 was no longer braked so that it could always provide an accurate position reference for the ATC system.

These measures have led to the currently imposed 169 km/h speed restriction. During the autumn period the maximum speed is reduced further to 140km/h. The restrictions have an important impact on the operation of the IC4 fleet and the fulfilment of its deployment plan.

In January 2014, a task force was set up to eliminate the braking problems. It has identified incorrect piping of the parking brake anti- compound device, which could lead to locked wheels during braking. The parking brake was being automatically applied when the air pressure within the brake pipe and brake cylinder fell below thresholds of 2.7 bar and 2.0 bar respectively (as might occur during emergency braking). The anti-compound device then received the cylinder pressure from the WSP valve outlet instead of from the integrated relay valve outlet. During braking on slippery rail, the parking brake could therefore be applied, resulting in locked wheels.

The task force also found problems with the brake blending. IC4 has conventional pneumatic brakes and hydrodynamic brakes (HDB). Powered axles have both types, but the trailing axles only have pneumatic brakes. The HDB is prioritised in order to reduce heating and wear of the brake pads, but it is blended with the pneumatic braking on the powered axles. There is also a cross-blending function for the pneumatic braking on the trailer axles. When wheel sliding is detected, the HDB is switched off and the pneumatic brake force is increased to compensate.

The task force found that, during wheel slide conditions, the additional pneumatic braking was not always applied. When it was applied, it occurred mainly on the trailing axles. The brake system will now be simplified by removing the cross blending function and an equal brake force will be applied to all axles. There will, however, be a reduced brake force on trailer axles 5 and 6 in order to ensure a reliable speed reference for the ATC.

Axle box housings

IC4 has also suffered from failures within the castings that carry the axle boxes. Cracking and fracturing has occurred where the primary damper supports meet the lower part of the housings. The primary dampers have also suffered from sweating and leakage.

From February 2010 to October 2014, three broken damper supports and six cracked damper supports were discovered. The Root Cause Analysis (RCA) revealed the following issues:

• Poor wheel maintenance due to incorrect instructions and the measurement of wheels at 180,000km intervals;
• Over-braking of trailer axles due to incorrect cross-blending of the braking system;
• Incorrect brake piping, resulting in locked wheels and wheel flats;
• High compression speeds within the dampers due to wheel faults and the striking of objects on the rails.

Cracking and fracturing has occurred both above and below the point where the damper support bracket meets the wheel bearing housing. Mitigations put in place by DSB have included increased wheel inspection frequency (visual inspection at 30,000km intervals) and it has been decided that ‘comfort turnings’ of the wheels will be undertaken at 80,000km on front bogies and 120,000km on the intermediate bogies.

As a high priority, new dampers will be fitted, which provide one third of the reaction force at high compression. Changes at the damper support of the axle box itself are not considered appropriate since the above measures (including improvements to the braking system) should eliminate the over-stressing of the damper supports. It is thought that some initial cracking might be caused by obstacles on the track, so the guard irons will be lowered by 75mm to prevent the leading wheels striking larger objects.

Power packs

An IC4 train set is equipped with four power- packs mounted underneath the coach body, two per end coach. Each power pack has an IVECO 560 kW low-emission 12-litre V8 common rail diesel engine, turbocharger, alternator, air compressor and cooling system. The alternator, air compressor and cooling pump are driven hydraulically. Each power pack powers one axle through a mechanical transmission.

Reliability of the power packs has been poor and the PROSE report covers 25 different  faults. It discusses four critical problems in detail, starting with broken turbocharger fixing bolts and cracked exhaust manifolds. The root cause was quickly identified as being the weight of the turbochargers, which are mounted eccentrically and fixed only onto the manifolds. DSB has now designed a turbocharger support bracket that is expected to solve the problem.

Broken fuel pipes are another problem, with leaks occurring in a 12mm fine-grain E235 steel piping that links the primary and secondary fuel pumps. The report says that IVECO is responsible for this component, but it has not responded to requests for assistance from DSB. The cause and solution have yet been determined and because leakage of fuel is considered a high risk, weekly inspections are undertaken. Two alternative fuel pipes are currently undergoing trials.

Crankshaft thrust bearing damage (scoring) is identified as another problem with the engines. One instance of a broken crankshaft has been attributed to this problem. It is not yet clear though whether the cause is specific to the affected engines, or whether there is a design problem affecting the entire IC4 fleet. A design change would be costly and could become a major issue for the outcome of the IC4 program and the fulfilment of the deployment plan. PROSE recommends the implementation of a task force to assess the risk of a design failure and to define specific corrective actions.

The alternators (three per IC4 train set) provide three-phase 400V AC power, but problems have arisen with overheating. Instead of an airflow through each alternator of 22.4 m3/min, as specified by the alternator manufacturer, only 7 m3/min was measured. To improve this, the air outlet has been enlarged by replacing the original cover with a grid.

The bottom line

It is estimated that the remedial action required on the power packs, axle box housings and brake systems will cost DSB in the region of 1.5 million kr (€202,000) per train set. The final availability goal to have 74 IC4 train sets in operation in 2019 is nevertheless regarded as realistic. In order to reach this level of operation, the report recommends that DSB should employ more engineers and workshop staff. This includes a taskforce to deal with the power pack issues, at least two more train computer and monitoring systems engineers and at least 40 additional workshop employees. The total cost of the programme will be more than 111.2 million kr (€15 million), although reduced operating costs will partially offset this.

Even as the PROSE team was at work in September 2014, the Danish department of transport launched an investigation to determine whether it would be more economical to have the IC4 trains scrapped, or have them rebuilt for slower regional traffic. Implementation of the PROSE recommendations therefore looks set to go ahead.

Regardless of the number of train sets used and their duration of operation, they only have a value if the functionality and the reliability goals are reached.