Not Just GWR Then.......

[whole diary]

[SandTEngineer]

From the Greater Anglia website (its rather a long excuse):

Latest information on Greater Anglia rail disruption in Norfolk and Suffolk

We are extremely sorry for the continued disruption to branch line services in Norfolk and Suffolk.

We fully appreciate that this situation is extremely frustrating and it is an absolute priority for us to get these problems resolved as soon as possible.

In the meantime, we are making sure that customers can still complete their journeys.

Please see below, the latest information on the problems as we currently understand them:

What’s going on?

Last Friday (6 December), we were asked to significantly reduce our rural branch line services by Network Rail, who were investigating signalling problems on some of our routes – looking at the impact of extreme autumn weather and leaf fall on track conditions, the components of the signalling system and the interaction of passenger trains with the signalling system.

There is also a separate investigation into an incident at a level crossing on the Norwich-Sheringham line which has led to speed restrictions at crossings on that route. This incident is currently being investigated by the Rail Accident Investigation Branch.

This has led to delays and cancellations for our customers. We are very sorry about this and we’re working round the clock with Network Rail to return services to normal.

Why are you saying its signalling issues when it seems to be just affecting your trains?

Our whole railway relies on the signalling system, which tells trains when to stop and go and controls levels crossings. Put very simply, there are sensors called “track circuits” which detect where trains are. It is essential that these track circuits are functioning correctly – picking up trains’ locations – so that signals and level crossings operate correctly.

The investigation into signalling issues is looking at the operation of some particular track circuits and the different factors involved – including track conditions, the signalling system itself and the interaction of our trains with it.

We do not know the full picture yet and it is not appropriate to rely on rumours (which are being spread by people who are not directly involved in the investigation and in many cases not even in the rail industry).

Are the new trains to blame?

You may have seen some media headlines claiming that our new Stadler trains are the sole cause of all these problems. This is categorically not true.

While our new trains had some software problems initially, these were easily fixed with an upgrade and they are returned to service.

Software problems are not the cause of the disruption that you have been experiencing since 6 December 2019.

The issue is the result of a combination of factors: track conditions, track circuits and the interaction of trains with the track circuits.

We have been working with Network Rail round the clock to find out why, examining every factor including components of the signalling system, the impact of extreme weather conditions and leaf fall, and the interaction between the signalling system and passenger trains, old and new.

The Rail Accident Investigation Branch is also investigating.

What are you doing to sort out the problems?

We are very sorry that this is taking a long time, but safety is paramount and while the investigation continues we are being extremely careful to run services more slowly, and this is leading to delays and cancellations on some of our rural branch lines in Norfolk and Suffolk.

A raft of short-term measures has been already implemented, including running extra rail head treatment trains to blast away fallen leaves, amended signalling operation on some routes and a 20mph speed restriction over level-crossings.

A team of Stadler employees has been set up and is responsible for checking train wheels on the bi-modes when they are in stations and as required, cleaning them. This is done to reduce the build-up of leaves, as well as contamination on the rails. When necessary, they clean the wheels with a pressure washer at the depot,and give them a manual scrub.

The new trains’ flange lubricators have also been temporarily deactivated. However, they do not deposit lubricant directly onto the track and this is purely a precautionary measure to help the investigation.

As soon as we, Network Rail and Stadler know more about the cause of the track circuit failures, we will recommend further long-term solutions, enabling the signalling system on these routes to detect trains consistently and accurately so that trains can run safely and on time.

Are the new trains safe and were they tested enough?

Yes – the new trains are safe. The safety performance of our new trains is in line with our existing high standard of our other train types as we operate them to industry recognised standards. They have been in operation for some time now and over many routes.

The manufacturer, Stadler, has a strong track record in safety, which is at the heart of its business. To date there are 1,900 Stadler FLIRT trains in operation in 19 countries.

Before these new trains for Greater Anglia were introduced, Stadler and Greater Anglia technicians carried out rigorous performance and safety tests, including an electromagnetic compatibility test on every single route, to see how they interact with the signalling system on the route.

Network Rail will only allow Greater Anglia to run trains on specific routes if they are satisfied that they operate safely and correctly. Each type of train also has to be signed off by the ORR^» (Office of Rail and Road formerly Office of Rail Regulation - about) before they can be put it into passenger service.

Stadler engineers conduct regular routine exams of the Greater Anglia new trains to ensure that the required safety standards are maintained.

Are these new trains too sophisticated for old British infrastructure?

The new trains made by Stadler have been subject to industry standard safety checks and authorisation before they were permitted to enter passenger service. They are also inspected as part of planned routine exams to ensure the required safety standards are maintained

Why is it that other operators on the same lines, old trains and freight trains are still allowed to operate normally unaffected by the signalling problems?

Network Rail is responsible for prioritising which trains run on our network. Some of our services have been cancelled to allow extra rail head treatment trains to run on our routes. Others have been affected as the investigation is looking at how our passenger trains, both new and old are interacting with the tracks.

When is this going to end and the train service get back to normal?

But we are working closely with Network Rail to get the service back to normal as quickly as possible – we are planning for a full service on Monday on all routes except Norwich to Sheringham and Ipswich to Peterborough.

Can I get compensation?

Yes you may be eligible for delay repay compensation if your journey has been delayed by 15 minutes or more.

Why have you sent your old trains back before the new ones are ready?

We haven’t. We still have nine of our old diesel trains in service.

Replacing every single old train on our network with brand new trains is a very complex project, which has involved careful planning over several years and many different contingencies.

It is not possible to entirely predict the results of safety and performance tests, when we check how the trains interact with the infrastructure, such as signals and points - which is why we drew up these different contingencies, which included worst case scenarios.

We have more new trains in our fleet than the number we have returned.

As we continue to increase the number of new trains in service, we will have a wider pool of back-up trains should faults occur.

Ultimately, our fleet of new trains will be larger than our old fleet was. 

[stuving]

So, the 755s are not activating track circuits - hence the link with the RAIB^▸ (Rail Accident Investigation Branch) investigating how Norwich Road level crossing "lost" a 755/4 and opened when it shouldn't.

That is rather surprising, as that problem (déshuntage) is expected with short light trains - few axles and lightly loaded. But 755s are not that short and while lightly built use shared bogies; indeed there was talk earlier on about their axle loads being too high.

But the design is an odd one, with very unequal loadings on the bogies. I was puzzled by the plated weights reported for these trains, and apparently I'm not the only one - basically adding the vehicle weights gives 51.2 t less than the weight for the whole of a 755/4.

However, the average axle load is know to be about 13t, and for ten or twelve of those (with some unevenness) to not make contact with the rails does seem odd. No doubt there's there's a lot of poking about in Stadler's design going on, looking at their idiosyncratic engineering practices to see what's new and might not suit the local variety of track circuit.

[FarWestJohn]

I am surprised this was not noticed earlier. I was staying in Norwich July and August and the 755 units were out all the time on driver training.

Are they fitted with track circuit assisters as many other light weight units are?

Good write up here when the 166/165 went through the process:

https://www.sparkrail.org/Lists/Records/DispForm.aspx?ID=9690

[SandTEngineer]

I think the problem here might be that the level crossings utilse track circuit predictors that are supposed to adjust the strike-in time based on the approaching trains speed.  So losing train detection due to rail contamination, even for a short period, affects that.  Thats my understanding at present.

Of course in the old days (am I allowed to mention those?) the track circuits were fitted with treadles to backup the track circuit operation, and thus avoid loss of detection.  Well, that problem was known about in the mid-1960s, so certainly 'Lessons were Learnt' at that time but of course NR^» (Network Rail - home page) doesn't have many people now who experienced those days......

[ellendune]

I am surprised this was not noticed earlier. I was staying in Norwich July and August and the 755 units were out all the time on driver training.

If IUIC the problem relates to the interaction of track circuits, leaves on the line, and new trains. 

This recent period is the first time all these things have come together so I am not surprised it was not noticed earlier. 

[stuving]

I am surprised this was not noticed earlier. I was staying in Norwich July and August and the 755 units were out all the time on driver training.

If IUIC the problem relates to the interaction of track circuits, leaves on the line, and new trains. 

This recent period is the first time all these things have come together so I am not surprised it was not noticed earlier. 

Except that it's a known problem, with known solutions - TCAs^▸ (Track Circuit Actuator), and something to scrub leaf residue off at least some of the wheels. So I presume Stadler had to justify those areas of their design to NR^» (Network Rail - home page) and ORR^» (Office of Rail and Road formerly Office of Rail Regulation - about) in a safety case, and the testing would have covered exactly the kind of leaf-infested worst cases that cause this problem.

I read that it's wheel contamination that's the likely cause, hence why it's 755-specific. Of course the whole subject of wheel/rail/leaf interaction it notoriously resistant to theory, hence the possibility that a new train (from a new supplier) provides the first example of new behaviour.

[Trowres]

There are 1900 FLIRT trains in operation in 19 countries (according to the the GA^▸ (Greater Anglia) website quote above). Have the problems been observed elsewhere or is this unique to NR^» (Network Rail - home page) for some reason? ... type of track circuit for example.

[SandTEngineer]

There are 1900 FLIRT trains in operation in 19 countries (according to the the GA^▸ (Greater Anglia) website quote above). Have the problems been observed elsewhere or is this unique to NR^» (Network Rail - home page) for some reason? ... type of track circuit for example.

Think you might need to read my earlier post above.

[stuving]

There are 1900 FLIRT trains in operation in 19 countries (according to the the GA^▸ (Greater Anglia) website quote above). Have the problems been observed elsewhere or is this unique to NR^» (Network Rail - home page) for some reason? ... type of track circuit for example.

I wonder what kind of answer that question could ever get. Taken literally, the answer (if one exists) would go a long way to solving the current problem, so if anyone knows it they should go and talk to the team investigating that. However, since the answer depends on knowing precisely what has caused the GA 755 issue, finding it first is likely to be impossible.

This (seasonal WSTCF) is a generic issue for track circuits and all modern trains:

An unwanted side effect of some forms of rail contamination, and indeed possible from some forms of low adhesion remedial treatments such as those involving sand, is a failure of the train to activate a track circuit. These incidents occur when the conductivity of the wheel / rail interface is insufficient for the track circuit to detect the presence of a train. Such failures, known as Wrong Side Track Circuit Failures (WSTCF), have the potential for safety related incidents with severe consequences such as train collisions, or failure of other signalling related equipment to operate level crossing barriers for example.

Much research into this phenomenon was undertaken by BR^▸ (British Rail(ways)) in the late 1980s / early 1990s following the introduction of new diesel multiple units which were causing more WSTCFs than the units they replaced. It was noted that low voltage DC^▸ (Direct Current) track circuits were particularly vulnerable to problems where modern diesel multiple units with good riding bogies and disc brakes operated. A range of solutions were found for the most common cause of WSTCF; rust film on the rail. However, such solutions are not always effective for other causes of WSTCF such as dry leaf film and excessive applications of friction improvers such as TG60, Traction Gel or sand. As a result, the railway suffers a number of WSTCFs every year.

The first priority is to identify vulnerable locations so that mitigation measures may be implemented. Such locations do not necessarily coincide with low adhesion locations identified for traction or braking problems, as track circuit failures can occur at locations where trains normally do not need to accelerate or brake. It is also rare to have both adhesion and track circuit problems at the same time since poor adhesion requires damp conditions, while poor track circuit performance requires dry conditions.

That's from section 6 "Train detection measures" of the Adhesion Working Group's manual "Managing Low Adhesion" - WSTCF is not itself an aspect of adhesion, but the two are closely linked and countermeasures to both need to be managed together.

So the related questions of "why just 755s" and "what do we do about it" may lead to something about Stadler's trains that's new to this network, but it could be something else - e.g. something new for Stadler they did in adapting their FLIRT design. One obvious thing is they are smaller and so no doubt lighter, but what other (or secondary) changes did that lead to? Then again, it maybe there's a known higher risk (due to the size and weight of the trains) but something about the measures expected to prevent WSTCFs hasn't worked as planned - perhaps due to an unrelated change in operational methods at GA or NR.

I think we'll just have to let the investigating team, with all the data they have available, get on with working this one out.

[stuving]

I think we'll just have to let the investigating team, with all the data they have available, get on with working this one out.

I was thinking about the investigation by the railway - NR^» (Network Rail - home page)/GA^▸ (Greater Anglia)/Stadler - when I wrote that, but the RAIB^▸ (Rail Accident Investigation Branch) are also investigating, specifically, the incident at Norwich Road crossing. They have said:

Since the incident, Network Rail has modified the settings of this and similar level crossings on the line to reduce the chance of a repeat of this incident.

Our investigation will identify the sequence of events which led to the incident. It will also consider:

•   the design, implementation and operation of the predictor system, including any effects of rail head contamination due to fallen leaves
•   the design of relevant elements of the class 755 train and the process for accepting it for use on this route
•   any underlying factors.

Our investigation is independent of any investigation by the railway industry, the Office of Rail and Road.

We will publish our findings, including any recommendations to improve safety, at the conclusion of our investigation. This report will be available on our website.

So we should get to find out what happened and why, perhaps with a lot of detail.

[SandTEngineer]

I think we'll just have to let the investigating team, with all the data they have available, get on with working this one out.

I was thinking about the investigation by the railway - NR^» (Network Rail - home page)/GA^▸ (Greater Anglia)/Stadler - when I wrote that, but the RAIB^▸ (Rail Accident Investigation Branch) are also investigating, specifically, the incident at Norwich Road crossing. They have said:

Since the incident, Network Rail has modified the settings of this and similar level crossings on the line to reduce the chance of a repeat of this incident.

Our investigation will identify the sequence of events which led to the incident. It will also consider:

•   the design, implementation and operation of the predictor system, including any effects of rail head contamination due to fallen leaves
•   the design of relevant elements of the class 755 train and the process for accepting it for use on this route
•   any underlying factors.

Our investigation is independent of any investigation by the railway industry, the Office of Rail and Road.

We will publish our findings, including any recommendations to improve safety, at the conclusion of our investigation. This report will be available on our website.

So we should get to find out what happened and why, perhaps with a lot of detail.

From information published elsewhere there is a bit missing.  They have placed restrictions on the number and types of train allowed to run.

[Bob_Blakey]

A possibly stupid question - I have a basic understanding of how track circuits work - but could this problem be eliminated by replacing the 'traditional' TCs^‡(resolve) with axle counters? And, if so, what would be the time & cost of doing so on the routes over which the Stadler trains run?

[stuving]

A possibly stupid question - I have a basic understanding of how track circuits work - but could this problem be eliminated by replacing the 'traditional' TCs^‡(resolve) with axle counters? And, if so, what would be the time & cost of doing so on the routes over which the Stadler trains run?

It's true that the main selling point of axle counters is avoiding the things that can go wrong with track circuits, including these rust and leaf ones. But as I understand it, the speed estmation that is used by the predictor for level crossing is specific to track circuits. Axle counters have two sensors per head and use them to sense direction of travel, but they are too close together to estimate speed with any accuracy. You could presumably do that with two axle counters, of course (and a bit of software).

[stuving]

There's a report in the East Anglian Daily Times of further details from GA^▸ (Greater Anglia):

'Not the new trains' fault' says Greater Anglia after rail chaos
PUBLISHED: 10:37 17 December 2019 | UPDATED: 11:22 17 December 2019  Paul Geater

The new trains introduced on Greater Anglia's rural routes were not the cause of the signalling problems that have plunged services into chaos over the last few weeks, the company said today.

Their engineers have been working with Network Rail since signalling problems first hit services at the end of November.

They are now satisfied that the signalling problems only affect the Norwich to Sheringham line in Norfolk - and that all trains that use that line, old diesel units and freight trains travelling to North Walsham - have been affected by it.

There have been no problems with signalling on any other lines in the region, and Greater Anglia is now trying to reintroduce normal services on all of these.

However the problems of the last three weeks have delayed the commissioning of new trains and training of drivers, so it will be a few days before a normal service is introduced everywhere.

And the Ipswich to Peterborough service could remain suspended until early in the new year because it has not been possible to carry out route testing with the new trains.

Greater Anglia officials had to introduce speed restrictions and extra tests on the new trains because they were the one new factor that had been introduced into services before the problems became apparent - officials say if there had been a problem with them it would have been irresponsible to run them normally without making changes for safety reasons.

A spokesman for Greater Anglia said: "We are very sorry for the major disruption to services on our regional lines recently. These delays and cancellations were not due to the performance or design of our new trains. We were unable to run our normal train service due to signalling issues.

"We are now able to operate the normal timetable on our rural branch lines again, except for the Norwich to Sheringham line, so the majority of services have been reintroduced.

Saying it's not a general issue with the 755s, or even that other trains are affected, doesn't rule out 755s being affected worse, of course.

[stuving]

The cab video of the original near miss at Norwich Road crossing (Thorpe End) has leaked and is visible on this BBC^» (British Broadcasting Corporation - home page) report (with no other new news). Scary!