Actually not ALARP at all.
How do you divert freight flows down a dead-ended line, without making things massively more expensive for everybody? Rust on the railhead shouldn't be a problem with the passenger service on the Cambrian line, it's not like it gets 1 train a week or something.
I'm no expert but would have thought that a rusty surface would offer greater adhesion than a polished one?
It had in run both directions the night before and was due to follow the last service that night.
It would, but the issue with rust is that train detection from track circuits will go to pot.
Although less issues with Axle Counters, which I believe this line has.
Axle counters are the primary means of train detection for the ETCS, but the line still has many track circuits at level crossings.
Needs to be a fiendishly clever piece of kit to guarantee it won't WSF, actuating in the right direction and unlocking the FPL underneath a moving train in the facing direction.
I understand that when the weather is wet the adhesion is worse but there's no major problem with track circuit detection, but when it's dry it's the other way round. Alerting for marginal track circuit detection may have benefit to warn of impending non-detection but it probably isn't a reliable predictor of poor adhesion.
A run-through of points isn't a primary safety issue as it's highly unlikely to directly injure anybody (although there are downstream safety risks if it blocks the line and leads to things like self-evacuation from stranded trains). So it would be very difficult on safety grounds to introduce something that introduces several potential significant hazards just to mitigate it.
As an example of faulty sanders causing a serious accident, look at the Glasgow Queen Street accident of October 1928. The sanders on the loco of an outgoing service were faulty. Tho loco started slipping badly on the 1 in 41 gradient and, as it was in a tunnel, the driver was unaware that the train was actually sliding backwards! It slid back into a shunting movement and 3 people died in the collision.
We don't know, because we don't have the detailed analysis. The RAIB's statements suggest sanding is likely to be a major factor, but we need to wait for the report.
I think we can be certain that sanding (or lack of) is a major factor otherwise RAIB wouldn’t have made mention in their initial update. It has also been stated elsewhere that TfW issued a NIR (National Incident Report) about blocking of 158 sanders within days of the accident.
The devil will be in the detail of when and how the sanders became blocked.
I've commented several times above regarding the effectiveness of sanders. There was also a RAIB report a few years back about an overrun of several miles somewhere on Southern/Southeastern when the sand had run out. Sorry can't find and link it right now.
The question is did the train leave it's overnight stabling location with sander pipe(s) blocked already, or did they become blocked during the day's operations. If the latter that's worrying and the mechanism of that needs to be understood and mitigated.
Doesn't that depend on whether the sanders are mandatory or just a nice to have? All our vehicles have sanders, but these are only checked during maintenance and a failure is not relevant from an operational point of view. However, the majority of them are mostly intended for accelerating and not for braking.
Stonegate in 2010 with a 375; it ran away for three miles. A summary, and the link to the full report, can be found here.
Apologies for messing up the quote tags, my phone doesn't like them!
If the unit is fitted with sanding equipment for braking (as 158s are) then it's a rule book requirement that the sanding equipment that will be at the leading end of the unit at any point during the day's diagram must be operational or the unit must not leave the depot. If the driver notices that the sanding equipment has become defective in service and they believe they will have difficulty stopping without it (i.e. because there are low railhead adhesion conditions) then they must stop and report to the signaller immediately.
Or one could ask - is the infrastructure / signalling ready for trains that slide past a red signal (or limit of ETCS movement authority) and don't slow down noticeably in the next 900 metres?
From https://www.rssb.co.uk/research-catalogue/CatalogueItem/S189 (from 2014 - registered only).
Note I'm only advocating the MTB type for classic heavy rail emergency use in the UK. The IRJ electrical bridging by metallic debris can only lead to 'right-side' failures to occupied in traditional track circuits, so that is not directly a safety issue, but a reliability one. Clearly, any IRJs passed over could be inspected and cleaned as neccesary following a slide incident where MTBs were used. I'd consider axle counter sensors passed over during a slide possibly vulnerable to some types of malfunctions as they use sensitive inductive techniques directly in the flange area. Again the most likely result is a miscount showing occupied. Much knowledge might be gained in this area from German experience, as both MTBs and axle counters have been in common use there for decades.
The metal mass of magnetic track brakes can have negative effects on existing signalling equipment even when not in active use. And Not every bogie is designed to absorb the additional forces of a magnetic track brake.
