RAIB: NR must assess manual track packing limitations
Network Rail should assess the suitability and limitations of manual lifting and packing, after carrying out track renewal and other work likely to result in significant change to track geometry or the supporting ballast.
It is the first recommendation made by the Rail Accident Investigation Branch (RAIB) in its inquiry into the derailment of a sand train at Courthill Loop South Junction, Lewisham, on January 24 2017.
Two wagons were derailed, one of which overturned. Investigations showed that the immediate cause of the accident was that there was insufficient wheel load at the lefthand wheels of the leading bogie of the 16th wagon to prevent the wheel flanges climbing over the rail as the wagon negotiated the newly laid track.
The double junction at Courthill Loop South Junction had been renewed in an engineering possession on January 14/15. After installation, the lead engineer established that the vertical level of the Up slow line was considerably lower than that specified, and that there was a transition step where the newly laid track needed to rise to match the level of the existing track. As no tamping machines were on site at the time, he arranged for manual lifting and packing of the ballast to smooth the transition step.
In a follow-up possession the following weekend, investigations found that the track had settled by about 20mm over the intervening week. A track twist was identified and additional packing carried out under three short bearer sections (connected with ‘U’-shaped bearer ties designed to link sleepers in modular switch and crossing installations), and after re-measuring it was confirmed that the maximum twist was within Network Rail tolerances.
RAIB’s investigation found that the trackbed poorly supported the bearers on the track affected when the renewal site was handed back, and that the configuration of the bearers made one side of the track more susceptible to poor trackbed support than the other. The bearer ties were found to have a degree of mechanical flexibility that allowed differential rotation of the short bearer sections on either side.
Another causal factor was the probable lateral offset of the sand payload in the 16th wagon of the train. The second recommendation the RAIB makes is that NR should assess and define the criterion for which it is expected that the vertical trackbed geometry should be confirmed under load, and define specific mitigation measures that need to be applied when this is not possible. It should then update its process and guidance to include objective limits and mitigation measures.
The third recommendation is that NR should determine the circumstances when cant gradient should be measured before handing back track into service following renewal and other work likely to result in significant change to track geometry or supporting ballast. A review of the design and validation of the standard bearer tie NR uses on modular switch and crossing layouts should take place, and the RSSB (Rail Safety and Standards Board) should “expedite” work to define an acceptable limit for the lateral offset of payloads carried by bulk hopper wagons.
RAIB Chief Inspector of Rail Accidents Simon French said: “The track was poorly supported when it was handed back for traffic on the day before the derailment, because there had not been time for the machines to finish tamping the ballast, and manual consolidation work was ineffective.
“The effect of all these deficiencies was to create a situation in which, when it was handed over for traffic the track had - or very rapidly developed - geometry faults. These created the conditions for the wheels of a freight train with a slightly offset payload to climb over the head of the rail.”