Derailment at Knaresborough, 7 November 2015

[whole diary]

[Chris from Nailsea]

The Rail Accident Investigation Branch (RAIB^▸ (Rail Accident Investigation Branch)) have published their report on their website:

Derailment at Knaresborough, 7 November 2015

At 07:22 hrs on 7 November 2015, a Northern Rail passenger service from York to Harrogate derailed on a set of points on the approach to Knaresborough station. The train consisted of two class 150, 2-car multiple units.

The leading five bogies derailed and damage was sustained by both the train and track. None of the train crew or five passengers on board were injured. The line was re-opened at 12:58 hrs on 8 November 2015.

The signaller in Knaresborough signal box had authorised the train to pass a signal at danger (red), without realising that the set of points beyond the signal was in an unsafe condition. The signaller had not checked the associated points position indicator in the signal box and misinterpreted the significance of being able to reverse the signal lever, leading him to believe that the route was correctly set and safe.

The signaller that day was a mobile operations manager. As a mobile operations manager, his core work was to respond to faults and incidents on the railway network; he operated signal boxes infrequently.

The RAIB concluded that the signaller did not have a full understanding of the working of Knaresborough signal box and that this lack of knowledge may have been the result of either poor initial training or the way his knowledge had been maintained.

An underlying factor to this incident was the lack of robustness of Network Rail’s competence management system for non-signallers (the people within Network Rail whose core duty is not to operate signal boxes but who occasionally have to do so).

In March 2016, Network Rail re-issued the operations manual for the staff in charge of operating signalling equipment. In April 2016, Network Rail started an end-to-end review of the way it manages the competence of its signallers.

As a result of this investigation, RAIB has made one recommendation on Network Rail to review whether the changes that it has recently made to the operations manual have resulted in non-signallers maintaining the required level of knowledge and experience.

RAIB has identified five learning points. The first three learning points relate to the actions of signallers in degraded operating conditions. The fourth learning point relates to the importance of investigating and understanding the underlying reasons for repeated asset failures. The final learning point relates to the actions of drivers when authorised to pass a signal at danger and after an incident.

[grahame]

None of the train crew or five passengers on board were injured

One of the things I find interesting (but off topic) is just how few passengers there seem to be in trains in RAIB^▸ (Rail Accident Investigation Branch) reports.    I'm not suggesting that quiet trains are more accident prone - just that it's a reminder to us all who travel in busy trains just how many aren't actually all that busy.

[grahame]

There were about 109 passengers and two train crew on the train

That's the 4 car one / King's Cross buffer collision.  27 per carriage.
http://www.firstgreatwestern.info/coffeeshop/index.php?topic=17334.msg199376#msg199376

[broadgage]

Yes, not only are overcrowded trains more memorable than ones with plenty of room, but also consider the following simplified but illustrative example.

Suppose that a TOC^▸ (Train Operating Company) operates 100 very overcrowded services, each with seats for 700 passengers but actually carrying 1,200 passengers.
And also 900 services with plenty of room, 700 seats and only 100 passengers on each train.

The TOC could say, truthfully, "only 10% of our trains are crowded, the other 90% have plenty of room"

Yet an honest and genuinely random survey of the passengers would show that the majority of those surveyed reported overcrowded  trains.
120,000 passengers experience overcrowded trains and only 90,000 experience plenty of room.

So the apparently contradictory statements that "most trains have plenty of room" and "most passengers find the trains overcrowded" can both be true !

[ellendune]

Yet an honest and genuinely random survey of the passengers would show that the majority of those surveyed reported overcrowded  trains.
120,000 passengers experience overcrowded trains and only 90,000 experience plenty of room.

So the apparently contradictory statements that "most trains have plenty of room" and "most passengers find the trains overcrowded" can both be true !

Of course they can both be true. If all the 120,000 were trying to use the 10% of trains they would be overcrowded, equally if the 90,000 were spread out among the other 90% they would not be!

[grahame]

So the apparently contradictory statements that "most trains have plenty of room" and "most passengers find the trains overcrowded" can both be true !

Exactly what we see and have learned.    Calculations on your hypothetical example:

100 "heavily overcrowded" with seats for 700 but 1200 on board
900 "light loading" with seats for 700 but just 100 on board

1000 services
Total seats 700,000

Passengers on overcrowded journeys 120,000
Passengers on light loaded journeys 90,000

Total passengers 210,000
Total seated passengers 160,000
Total standing passegers 50,000

Passengers reporting "heavy overcrowding" - 57%
Trains heavily overcrowded - 10%
Load ratio on these trains - 171%

Passengers reporting "light loading" - 43%
Trains lightly loaded - 90%
Load ratio on these trains - 14%

Passengers standing when they travel - 24%
Seats occupied - 23%

What you see in the press:
Passengers reporting "heavy overcrowding" - 57%
Passengers standing when they travel - 24%

What you see when you work out the line's economics:
Less than 1 seat in 4 occupied on average (23%)
9 trains out of 10 carrying just 10 people per carriage
Overall loading 30%

It's very instructive to look at passenger numbers based on published ticket sales data, and to work out the number of seats based on timetables and stock used, and come up with an estimate of actual figures; easy enough to do for a branch line on which you can assume that most of the traffic runs to or from the junction station, much harder elsewhere.