The RAIB have released an update regarding the above incident which can be found here. I've included a couple of extracts below:
13 The RAIB has found no evidence of any malfunction of the traction and rolling stock, signalling, AWS or TPWS equipment involved in the incident.
18 The driver indicated to the fireman that an AWS brake demand had occurred. His expectation was that the fireman would open the AWS isolating cock in order to by-pass the AWS brake demand and release the brakes. The fireman has stated that he believed that he was following the driver’s instructions when he subsequently crossed the cab and opened the AWS isolating cock. Witness evidence suggests that the AWS isolating cock was not sealed before the fireman opened it.
19 The railway rule book permits the drivers of trains that are in service to isolate the AWS system only if it has become defective or if it is inoperable due to the configuration of the infrastructure. In these circumstances, the railway rule book requires drivers to immediately bring their trains to a stand and then contact the signaller. Certain conditions must then be met before the train can proceed any further.
20 However, in this case, the train was not brought to a stand and instead continued on its journey. OTDR data shows that the brake demand made by the AWS system ceased to be effective around 12 seconds after it was initiated. The brief brake application which resulted from the AWS brake demand before it was by- passed reduced the train’s speed by a total of 8 mph (13 km/h). Witness evidence and OTDR data show that the AWS isolating cock remained open during the remainder of the incident; this had the effect of making any subsequent AWS or TPWS brake demands ineffective.
21 During this 12 second period, a second AWS warning occurred. This was created by the fixed AWS magnet located around 275 metres on approach to signal SN43, which was displaying a single yellow caution aspect. This warning occurred around 2.5 seconds after the AWS warning from the TSR had been acknowledged; OTDR data shows that it was acknowledged by the driver within 0.5 seconds. Witness evidence suggests that the driver was unaware that he had received two separate AWS warnings and that he instead believed that he had received a single warning. Because he had seen the TSR warning board, he also believed that the warning he had acknowledged was associated with a TSR. The driver did not see signal SN43 and, therefore, was not aware that it was displaying a caution aspect.
22 Just over a minute after the AWS brake demand had been triggered, train 1Z67 approached signal SN45, which was displaying a red danger aspect, at a speed of around 52 mph (84 km/h). As it did so, it passed over the first TPWS OSS for this signal, located approximately 750 metres on the approach. This OSS is configured so that the TPWS system fitted on any train passing it at a speed greater than 65 mph (105 km/h) will demand a brake application. Because the train was travelling more slowly than this set speed, the TPWS system on Tangmere did not generate a brake demand.
23 Train 1Z67 then passed over the second OSS at a speed of around 53 mph (85 km/h). This OSS is located approximately 360 metres on the approach to the signal and has a set speed of 45 mph (74 km/h). The TPWS system correctly identified that the train was travelling over the set speed and demanded a full brake application. However, because the AWS isolating cock was still open, this demand had no effect on the train’s braking systems.
24 At some point on the approach to SN45, the driver of 1Z67 saw that the signal was at danger and fully applied the train’s brakes. Analysis of the OTDR data available suggests that the earliest point which the driver could have applied the brakes would probably have been around 220 to 230 metres on approach to the signal. By the point where the driver applied the brakes, there remained insufficient distance to bring the train to a stand at signal SN45. The train subsequently came to a stand around 550 metres beyond the signal, standing on both the crossovers and the up and down Badminton lines, at just after 17:26 hrs.
26 The evidence available to the RAIB indicates that the SPAD at SN45 was not the only occasion on which the AWS isolating cock was used by a train crew operating Tangmere in order to by-pass an AWS brake demand. The extent of this practice continues to be the subject of further investigation.