Events following the accident
At 02:14 hrs, the Vale of Glamorgan signaller spoke to driver 2 on the Global System for Mobile communications – Railway (GSM-R) radio system to inform him that two track circuits continued to show occupied after the passage of his train through the Llanharan area.
In this conversation, driver 2 confirmed to the signaller that his train was travelling through the Pontyclun area and that it was complete as his train still had a continuous brake pipe.
One of the possible reasons for an occupied track circuit is the presence of one or more wagons left behind by the train, which would manifest itself to a driver by a drop in brake pipe pressure resulting in an automatic brake application.
At 02:14 hrs, the Vale of Glamorgan signaller called the Port Talbot signaller back to say that he had spoken to the driver who had confirmed that the train was complete.
In this conversation, the Port Talbot signaller queried whether anything was being dragged or hanging off the train.
At 02:21 hrs, train 6A11 passed over the HABD site at Pontsarn, and soon after, St George’s CCTV level crossing.
The Vale of Glamorgan signaller observed the passage of the train using the level crossing CCTV monitor.
He did not notice anything amiss with the train.
The Vale of Glamorgan signaller called the Port Talbot signaller at 02:24 hrs to confirm that the train was complete and that there did not appear to be anything being dragged or hanging off the train.
At 02:30 hrs, train 6A11 passed in front of Cardiff ROC, located to the west of Cardiff Central station.
By that point, the train had left the area of control of the Vale of Glamorgan signaller to enter the area of control of the South Wales main line signaller at Cardiff ROC.
As the train passed the ROC, staff inside the building knew that something was amiss because of the loud repetitive banging noise made by the train.
The main line signaller immediately started making arrangements to bring the train to a controlled stop and sent a message on the GSM-R system for driver 2 to contact him.
At 02:32 hrs, driver 2 contacted the main line signaller who asked him to bring the train to a stand and wait for further instructions.
Train 6A11 came to a stand at Cardiff East Junction, just outside Cardiff Central station at 02:33 hrs.
The shift signalling manager and incident controller at Cardiff ROC had heard the banging noise made by the train as it passed them.
They agreed that the noise was most likely to have been caused by the presence of wheel flats on a wagon and decided that the most suitable course of action was to allow the train to continue towards the wheel impact load detector (WILD) site at Marshfield, approximately 6.5 miles (10.4 km) away.
Equipment at WILD sites measures the impact loads on the rails to detect abnormalities with wheels, including wheel flats.
The expectation of the shift signalling manager and incident controller was that the WILD site would confirm the presence of any wheel flats.
In this case, their intention was then to route the train to Alexandra Dock sidings for examination.
This course of action was conveyed to the South Wales main line signaller.
At 02:43 hrs, the main line signaller spoke to driver 2 to instruct him to continue towards Marshfield and to then expect a further call.
Train 6A11 moved off a few minutes later and passed the Marshfield WILD site at 02:56 hrs.
The incident controller was remotely monitoring the passage of the train over the WILD site expecting to receive an alarm.
However, no alarm came.
As a result of this, the train was allowed to continue on its journey.
Train 6A11 passed Newport station at 03:05 hrs and Bishton HABD at 03:12 hrs.
It arrived at Severn Tunnel Junction at 03:19 hrs.
Having emerged from the tunnel, it then passed Pilning station at 03:32 hrs.
As the train passed the train maintenance depot at Stoke Gifford, staff heard loud noises and saw sparks coming from one wagon on the train.
They reported this to the signaller at the Thames Valley Signalling Centre (TVSC).
The TVSC signaller contacted driver 2 and asked him to stop and examine his train.
Train 6A11 came to a stop at signal BL1587 on the down Filton main line beyond the disused Horfield station at 03:48 hrs.
Upon examining his train, driver 2 noticed evidence of ‘scaling’ on the wheels of three wagons (7th, 17th and 23rd) and that the trailing bogie of GERS89016 (11th) was feeling warmer than its leading bogie.
He did not immediately identify the collapsed brake rigging.