Guards in Name Only? Dwell Times And The SWR Guards Dispute

In December 2019 the RMT union called a strike by guards on South Western Railway (SWR). It is entirely possible that it escaped your notice. The reasons for that, and for its calling, are worth exploring in detail.

The strike started on Monday 2nd December and the RMT union effectively called on the guards not to work for the entire month of December 2019 and the start of the new year. In contrast to a superficially similar strike on Southern a couple of years ago, this did not become a near-daily news item. Even on the first day of the strike reporting was limited – as if the event was not something particularly out of the ordinary.

In general, the strike seems not to have caused the same level of enraged anger amongst commuters that the Southern strike caused. This is despite the fact that the number of trains run was nominally cut to around 50% of the normal level of service although in fact quite a few extra trains were run. Furthermore, trains tended to be formed of the maximum number of carriages possible, meaning that overall capacity was reduced by considerably less than 50%.

A silver lining, not obvious to passengers, was that SWR were able to take advantage of the spare drivers and spare train paths to carry out interference testing of their class 442 stock, which had been banned in service due to the train interfering with signals. This has now been rectified and proved to be working. If it hadn’t been for the strike, we probably would have been many months away from seeing these trains back in service.

There may have been many factors that affected the lack of news coverage of this dispute but a crucial one may have been that the service that had been run was generally running pretty reliably and on time. So no new news to report and at least passengers knew where they stood – although they probably were standing rather than seated. Another factor limiting press coverage could have been the air of resignation about the whole thing. A third could be that the the strike didn’t seem likely to (and didn’t) drag on for months unlike the Southern dispute.

Tactical mistake?

Whatever the rights or wrongs of the dispute, tactically this strike does seem to have been a bad move by the RMT. It failed to attract attention beyond the area affected and it did not get much publicity (or attention from MPs in the middle of a general election). Furthermore, either the RMT union either had to drain its coffers to pay sufficient compensating strike pay or it expected members to forego an income at the very time of year when people are at their most reluctant to have their income cut. More to the point, the strike had a limited effect as the level of service run throughout the duration of the dispute demonstrates.

Not the same as the Southern Dispute

Superficially, this strike shared some similarity with the Southern dispute, but there were some major differences beneath the surface. The exact reasons for that strike were never clear, but a claimed reduction of safety standards, a fear that guards (recategorised as onboard supervisors) would in future be made redundant and deep unhappiness about the principle of any increase in the number of trains run without a guard being present regardless of the circumstances, were all key issues.

As with the Southern dispute, on SWR, here the RMT union argued that having the doors opened and closed by the driver on SWR is less safe than having the guard in charge of this. Also, as with the Southern dispute, management in this latest dispute vehemently denied that safety would be compromised. At heart, the RMT are opposed to any operation of trains on SWR without guards under any circumstances, but the SWR appear to have withdrawn any such proposals quite a while ago. So this dispute now seems to have largely revolved on who opens and closes the doors. Or, as we shall see later, just on who will close the doors in future.

Reasons left unsaid

Never stated in either dispute but at the top of many people’s thoughts is the idea that these disputes are really about industrial muscle. If the trains doesn’t need a guard then any future threat of industrial action by the RMT union would be largely ineffective. Effective action would be limited to the withdrawal of labour of the few train drivers who are members of the RMT union rather than ASLEF.

In the Southern dispute, the RMT insisted that there was a hidden agenda to get rid of the guard’s job, but in the SWR dispute the management have assured the union that this isn’t the case. Nor are the union claiming this is the intention of the current management. Of course the current management can’t answer for the next operator of the franchise, who would potentially be in a position to get rid of guards if they wanted to– or (more importantly) were mandated to do so by the DfT.

Past reasons left unresolved

In the earlier dispute, one could well ask why Southern was so determined to have the drivers close the doors, if there was no intention of getting rid of the guard (opening themselves up to the accusations of a hidden agenda). The answer is that Southern Railway were severely hampered operationally by having to have a guard on board. They wanted to be in a situation where if a guard was unavailable for any reason, they wouldn’t have to cancel a train. As well as having a shortage of guards on occasions, Southern relied on trains being joined and divided to maximise capacity on their network which brings an added level of roster complexity . A train waiting at a critical platform for want of a guard for a portion of the train could – and did sometimes did – bring an entire route to a halt. Southern were anxious to eliminate this unnecessary risk.

As noted earlier, SWR (unlike Southern) have agreed to have a guard on every train. Officially, the outstanding issue in the RMT’s dispute was (and remains) that having the guards no longer closing the doors compromises safety. The RMT union clearly regards this as a fundamental part of the guards’ job and say that without this the guards are merely “glorified porters”.

THE SWR CAPACITY PROBLEM

Over the past 30 years there has been a massive expansion of commuting into London, which has required more capacity on just about every route into the capital. On almost all of the major routes into the city there have been major schemes aimed at resolve this situation. So for example, Southeastern has been augmented by HS1 services. Great Western and Great Eastern lines will get some relief from Crossrail and the Great Western line also benefits from electrification. Southern and the Midland Main Line have benefited from Thameslink. A scheme is also in preparation to enhance the Brighton Main Line and remove the critical bottleneck for Southern there. Euston station is being expanded for HS2. Meanwhile King’s Cross is having two extra tracks just north of the station reinstated, which will further increase capacity when combined with the benefits of Thameslink.

Unfortunately, the South West Main Line has also experienced considerable growth (especially long-distance growth) but there has been little done to accommodate it. Options are limited. There is only one terminus served and this means there is little opportunity to ‘mix and match’ different solutions. There really has to be one solution to fit all circumstances. The only alternative option that was available, re-opening the former Waterloo International Terminal for domestic trains, was implemented in 2019.

The only real hope for significantly increasing capacity by increasing the number of trains coming into London on the South West Main Line is Crossrail 2. Unfortunately, it is hard to see how Crossrail 2 could be open before 2035 at the absolute earliest and a solution is needed for South Western Railway now.

SWR – definitely sub-optimal

SWR has a number of inherent problems with its current set-up and these problems have dogged and them and their predecessors for years. A lot of it is down to the rolling stock on the line. A variety of different classes of train are used, which hinders many aspects of operation but, in particular, it hinders performance optimisation. Incredibly, some of the line’s new class 707 trains were even under-specified in terms of future performance required, meaning they will be withdrawn in the next year or so.

When the older trains were conceived in the 1980s, minimising dwell time was not a fundamental design consideration. They have insufficiently-wide doorways and have vestibule areas not planned around actual passenger behaviour in doorway areas. The problem is that the design of doorways and the vestibule area becomes more critical as trains get more crowded and have a greater number of standing passengers.

Worse still, there have been issues with the designed length of the trains themselves. Suburban stock into Waterloo used to generally be only eight coaches long when many train operating companies were already operating twelve car suburban trains. This has been partially rectified by the introduction of some 10-car trains.

Another problem with SWR is the timetable. This was recast a few years ago and a key objective was to increase dwell time (stationary time at a station stop) so that it was more reflective of reality. Unfortunately, as trains get more crowded, dwell time increases disproportionately. One cannot keep increasing dwell time though, as that then eats into overall capacity as one cannot then run as many trains.

Seconds really do matter

We can’t emphasise enough at LR Towers how critical dwell time really is on the modern railway. We generally despair when people suggest that a few odd seconds here and there won’t make much difference.

Just to give a few examples from the recent past, we have reported on how a vision was created for a Victoria line with trains every 90 seconds and where “every second counts”. More recently we have seen the reality of how this laudable objective hasn’t quite been achieved and a train every 100 seconds is the best that can be attained. This is largely because it is currently impossible to reduce the dwell time at King’s Cross southbound platform.

We have also criticised the former junior transport minister, Steven Hammond, who himself criticised Network Rail’s proposed plans for Thameslink with the comment “This would cause considerable inconvenience and disruption all because that would allow Network Rail some marginal seconds in efficiency gain”. In fact the marginal seconds were critical.

Just to show how critical marginal seconds are, in a different scenario, the plans for Thameslink were altered at a late stage to send trains to Gillingham rather than to Tattenham Corner when it was realised that a set of particular junctions north of East Croydon could not be made to work with the proposed timetable. This late change of plan turned out to be extremely disruptive. In this case “a few [extra] marginal seconds” would have made all the difference between a working timetable and a non-working one.

We will look at why seconds really do matter in an SWR context later.

The SWR capacity solution

Fortunately for the beleaguered commuter, the future capacity problem on what is now SWR was recognised years ago and the problem was comprehensively addressed. It was recognised that a large part of any solution that did not involve a multi-billion pound spending and major reconstruction plan lay with having the appropriate trains.

Arguably, the biggest problem on SWR was that the trains were too short. If you run 8-car trains then you can increase capacity by 25% just by adding an extra two coaches. This is the obvious big-hitter solution that does a lot to solve your problems in one go.

The obvious first problem with having longer trains is that you need extra coaches. The next problem, not quite as obvious, is that you need more power because the trains are longer. You also need extra storage capacity in the depots which means longer sidings in existing depots – or new depots entirely.

The issues do not stop with the trains. Signalling is generally optimised for a particular length of train. Whilst a lot of trains (longer distance ones) on SWR are 12-car there will be portions of the network optimised for 8-car trains. Without a lot of careful planning, there could be locations where longer trains foul junctions if brought to a stand.

If all that wasn’t bad enough, to lengthen the trains you need to lengthen a lot of platforms. Platform lengthening is generally a time-consuming and expensive process, as some of it can only be done when trains aren’t running. To cap it all, some of the critical platforms for suburban services at Waterloo were only long enough for 8-car trains.

New trains

It might seem from what has already been suggested that lengthening trains is a potentially good idea even if it is expensive. It turns out though, that it actually more cost-effective not to lengthen the existing trains but to buy new longer ones instead.

There are a number of reasons for preferring to buy new trains. You can reduce the number of driving cabs now that trains aren’t lengthened and shortened daily to the extent they used to be. Fewer driving cabs not only creates more capacity and reduces maintenance costs but also reduces dwell time as passenger doors are more evenly spread along the train. Those passenger doors can also be wider and the vestibule areas can be designed to discourage people from blocking doors. Fewer cabs also save a lot of money as they are a significant portion of the train-build cost.

Another measure can increase capacity to an even greater extent. You can design the longer train so it has more standing capacity in proportion to seating capacity whilst keeping or increasing that absolute number of seats.

Benefits of new trains aren’t limited to what passengers can see. You can have a homogeneous fleet with the same acceleration characteristics and you can take advantage of state-of-the-art traction control (which have advanced spectacularly over the past 30 years). Modern trains also have a lot of self-diagnosis features and these can be used so that maintenance is done when needed rather than after a given number of hours. This both reduces the time needed for maintenance and also the likelihood of a breakdown.

Not just new trains – a new fleet

It is not enough to simply replace the trains that ‘need’ replacing. A homogeneous fleet means consistent dwell times, acceleration and retardation curves. All this further increases capacity but it means prematurely retiring any recently-introduced stock. Fortunately for those set upon this course, leasing, rather than outright purchase, of the existing fleet makes this proposition more feasible.

Having just one variety of train also reduces driver training costs and the potential for accidents due to different driving techniques and cab layouts on different stock. On the maintenance side too it reduces costs as fitters only have to service one type of train and there is a smaller pool of spares required.

With new features

What you can also do with new trains is have new features that aren’t practical to retrofit in old trains. These will not be limited to features for the passenger such as better customer information displays or USB sockets. The big plan for future trains, for which passive provision is being made in trains built recently, is something called Automatic Braking and Door Opening (ABDO).

Automatic Braking and Door Opening

ABDO is one of the big new ideas for the future on suburban lines. This would only apply for stopping at stations and is something that is designed to produce a lot of the benefits of automatic train operation at a fraction of the cost. The idea is that fixed transponders on the approach to a station would enable a train to stop there accurately and in the correct position, without requiring any action from the driver. Since you are only stopping a train, the safety challenges are considerably reduced over automating an entire line.

The automatic braking is done optimally in conjunction with the wheel-slip protection system (ABS for trains) which is already fitted. The big benefit of this is it ensures not only that any potential sliding is eliminated but also that optimal and consistent braking performance is achieved. Once automatic braking is implemented, trains should automatically stop in the correct place on the platform. As soon as it is detected that the train has stopped in the correct place the appropriate doors can be opened.

Apart from saving time, there are enormous safety potential benefits with automatic braking. Instead of concentrating on bringing the train to a halt, the driver can instead concentrate on making sure that if someone falls or jumps onto the track in front of the train they can apply the emergency brake promptly.

Automatic Door Opening

The DO of ABDO stands for door opening and takes the concept of automatic braking one stage further. It (unsurprisingly) enables the doors to be automatically opened. Here too there are distinct advantages. The doors will be opened on the correct side of the train and only the doors that are in the platform will be opened. Whilst one can argue as to whether it is safer for the driver or the guard to release the doors for opening, it is difficult to argue against it being safer still for them to be opened automatically with all the different additional safeguards that this brings.

There is a further operating advantage with automatic door opening in that the doors will be released for opening at the earliest opportunity. This already happens on Thameslink in the central area. One can also, a bit cynically, state that, by having the doors open automatically, you effectively kill any argument as to whether the doors should be opened by the driver or the guard.

Implementing the plan

The good news is that the capacity plan developed for SWR is coming together with practically all aspects in place. The main item missing is the new trains, which are behind schedule but are now due to be phased in from summer 2020.

The order for the trains consists of 60 10-car sets and 30 5-car sets. They are expected to have ABDO capability, although this may not be initially present. What they do not have is guard’s panels for guards to operate the doors, other than a facility to prevent the doors being closed. This is in case the guard is on the platform assisting mobility-impaired passengers, for example. Clearly, if the driver cannot close the doors, the train cannot depart.

Other aspects of the plan have been completed or are underway. The 8-car suburban platforms at Waterloo were extended to 10 cars a couple of years ago, a major task. Other platforms that needed extending have already been extended. Some of these, such as at Feltham station, have been a major challenge.

The power supply has been upgraded. A new stabling depot is being built at Feltham on part of the site of the old Feltham marshalling yard. Waterloo has been resignalled (admittedly not without long-term issues) and Feltham signalling area is in the process of being resignalled too.

Seconds really count on SWR

For many years now SWR and its predecessors have been plagued by needlessly long dwell times. One way of reducing dwell time is to introduce driver only operation (DOO) so that the guard does not have to confirm to the driver that the doors are closed before sending the start signal to them. DOO schemes on the lines into Waterloo have come and gone leaving DOO mirrors and monitors unused for many years, before finally being removed.

SWR and the DfT have really focused on reducing dwell times in order to increase capacity in terms of numbers of trains that can be run. Hence one of the reasons to almost completely replace the rolling stock with something more appropriate. One factor that was tackled was to minimise the time taken for the door opening and door closing sequence. This will eventually be improved further by ABDO. The closing sequence time will be tackled by transferring the task of closing the doors to the driver. The difference between the latest attempt to do this and all such exertions in the past is that the proposals currently on the table make no attempt to remove the guard from the train.

The RMT case

As management have given ground it is inevitable that the RMT union have shifted their case. They seemed to recognise that it is going to be hard to argue on safety grounds against the doors being automatically opened, if all the appropriate safeguards are in place. If the guard is present on the train then any arguments about the plight of mobility-impaired passengers are also hard to make. This is especially true as the guard will be able to ensure the train physically cannot depart whilst assisting a passenger on the platform – something that is not actually true today.

The RMT state that:

It is our view that adding three or four seconds dwell time at each station to ensure the safest method of working and despatch is surely a small price to pay to guarantee the safety and accessibility for all, which is the issue at the heart of the dispute.’

From this statement we can see that the RMT:

  • Claim only three or four seconds are added
  • Have stuck to their line that having the guard closing and opening the doors is the safest method of working
  • Refer to ‘accessibility to all’, which is a red herring if the guard is present
  • Agree that this is the heart of this dispute.

Is it worth those extra seconds?

We have covered the issue of whether it is actually safer for the guard to close the doors in our previous report on the Southern guards dispute. Once this argument is removed, we are left with the issues of how long extra the procedure takes and whether this is, in fact, ‘a small price to pay’.

We are also left slightly unclear as to whether the RMT are happy for the doors to be opened automatically. If not then it is hard to see how they think that the extra time taken for the guard to open and close the doors can be anything like three or four seconds.

Even if it only applies to dispatch, this low figure would doubtless be disputed and it is hard to see how it can be as little as this given the actions the guard must take:

  • See that it is safe to close the doors
  • Checks the signal (if present) has a proceed aspect
  • Close all the doors except their own
  • Check again now the doors are closed
  • Close their own door
  • ‘Ding-ding’ the driver to confirm it is safe to depart.

After all this, the driver then has to ding-ding the guard to acknowledge the signal. Only then can the train depart.

Compare the above procedure with the situation where the driver closes the doors which is that the driver:

  • Checks that the signal (if present) displays a proceed aspect
  • Checks it is safe to close doors
  • Closes doors
  • Checks it is safe to depart (no-one stuck in doors or leaning against train for example)
  • Departs

Industry insiders suggest that the second approach results in a minimum saving of six seconds and potentially a typical saving of ten. Note that once the guard has closed their own doors, he or she cannot see if it is safe to depart. So this check can only be done by the driver anyway.

Why those seconds really matter

A major issue in this dispute is that SWR are desperate for any increase in capacity until Crossrail 2 comes along. If running 20tph on a suburban stopping route, then an additional six seconds saved per train amounts to two minutes per hour. Ten seconds per train amounts to over three minutes per hour. This is roughly the equivalent of being able to provide an extra high-density 10-car train per hour.

As an example, consider Earlsfield, the only station between Clapham Junction and Wimbledon and a station where passengers physically cannot board the train in the morning. Sources suggest that SWR believe that if guards are not involved with closing the doors, they can provide an extra four trains in a three hour peak period on top of all the benefits already planned.

Earlsfield might be the best example of a station to benefit from such a change, but it is almost certainly not the only one. Even if, elsewhere, it does not lead to an increase in capacity, it will lead to an increase in reliability and punctuality.

It’s always about the money

Apart from any wider issue about the desirability to increase capacity for the health of the general economy, it makes a lot of financial sense to run a more frequent service. This is because the suppressed demand on SWR extra trains can be filled. The conventional argument is that commuter services don’t make money because of the infrastructure needed to provide for them for only a few hours of the day. However this argument is only valid if you have to provide new infrastructure. If you already have the infrastructure and the rolling stock then squeezing additional trains into the peak when they are full of season-ticket holders can be very profitable indeed. And it is almost all pure profit.

It is also probably easier to achieve than trying to save an equivalent amount of money by eliminating guards.

What does the RMT union hope to achieve?

With the Southern dispute it was hard to see what it was really about. With the current SWR dispute it appears to be quite clear exactly where the sticking point lies. Ignoring for the moment any rights or wrongs of the dispute, it is hard to see what the RMT have been trying to achieve by their action.

The problem for the RMT is that historical precedent suggests that they will always lose a battle of attrition if the other side is determined enough. This is especially true when the other side is financially compensated for any losses by the DfT – as was almost certainly true in December.

Whilst some commuters have suffered enormously, there is little evidence to suggest that they are blaming SWR, perhaps down to SWR being able to get their message across. Furthermore, on most lines SWR demonstrated it was able to run its limited-capacity service using ‘contingency guards’ – management trained up in the guard’s role.

If A further strike is called…

If the RMT were to call a further strike of a month’s duration it would seem almost certain that SWR would keep going and provide a contingency service – as they have done on a number of occasions in the recent past.

On Monday 6th January 2020 a new managing director started at South Western Railway. He has made it clear from the outset that he will not give in to union demands. The Times reports him as saying that caving in would “sacrifice our ability” to improve punctuality. The new managing director, Mark Hopwood, comes from GWR where he has presided over changing in working practices as a result of replacing HSTs with state-of-the-art Hitachi Inter City Express trains. Mark is a well-respected, practical railwayman who knows how to run a punctual railway. He knows he cannot give away a few seconds per station in order to placate the RMT.

Once the new trains start coming into service, there is the further danger that ASLEF (the main train drivers union) and SWR management might come to some agreement to run trains without guards during the dispute. An agreement has been in place for years that gives ASLEF drivers on the SWR (and predecessors) an increased pay packet in return for agreeing to driver only operation. This was originally introduced to cover an agreement to operate trains without a guard and, based on their past experience with Southern Railway and the law courts, ASLEF might not be keen to breach their agreed terms of that contract.

If the new trains have no guard’s panels, as is the overwhelming likelihood now, then the RMT can be accused of preventing them from being operated. Even if the RMT refuse to work on the new trains, their introduction using contingency guards would mean a significant increase in capacity on suburban services during strike days, due to the trains being longer and having more overall capacity.

The crux of the matter

It does seem that, in the long term, it is next to impossible for the RMT to win this dispute. The recent month-long strike and its failure to bring SWR to its knees shows the struggle the union faces.

In order for the RMT to ‘win’ the dispute, the 90 new trains would have to be retrofitted with guard’s panels and, if that were done, some of the advantages of modernising SWR would be lost. We also need to emphasise that, as with the Southern dispute, the safety case argued by the RMT is disputed within the rail industry.

Just to add to the RMT’s woes, there is a government in power with a large, safe majority and it has no need to hold a general election for almost five years. This government is likely to be unsympathetic to the RMT’s cause and has already talked about legislation to ban total strikes on the railways. We might not be close to witnessing the end of a second crew member on board suburban trains in London, but it looks like the traditional role of the suburban guard is disappearing – perhaps preserved on SWR in name only.

One thing this dispute has highlighted is that modern technology means rethinking the role of the second person on the train (if retained) is now inevitable. What purpose (if any) are they there to fulfil? Until that question is answered we risk having the people with the wrong skills appointed to the job.

The current occupants of SWR guard’s jobs and their union may complain that their job is being devalued, but technology has already done that. So long as the role of guard is seen as the 21st century progression of a 19th century role, rather than one which meets 21st century requirements, we risk an era where strikes are endemic or, at best, a fragile and uneasy peace maintains.

Thanks to ngh, in particular, for providing much and verifying much of the background information. London Reconnections is advert-free thanks to your support. If you enjoy our work, you can buy us a coffee or support us on Patreon.

Cover photo by Andi North.

101 comments

  1. Good article, but I think it misses a few things.

    It reads like the December action was the first strike by the RMT, but the RMT has been striking regularly ever since SWR took over in August 2017. The change of focus to being just about who closes the doors, is the most recent version of this dispute. The strikes started off as pre-emptive action, just in case SWR were thinking of going to DOO, but every time SWR thinks they have an agreement, the RMT change their argument a little. At one time, the RMT were demanding that SWR guarantee the role of the guard even if SWR lose the franchise, which is obviously nonsense. December 2019 was the first time the RMT had striked for an entire month, but they have done nearly entire weeks before.

    Research by SWR has also shown that passengers want a customer-focussed staff member on board, rather than an operations-focussed person, so this is what SWR are aiming to implement, and so have guaranteed a second member of staff on board, but that isn’t good enough, mainly because RMT see guards as a key operational responsibility.

    Otherwise, a good overview of the dispute.

  2. There is also, as hinted, “other” politics involved.
    RMT certainly seem to be involved in a “political” strike & statements from their Scretary seem to support that hypothesis. Whereas ALSEF seem to be taking a much more pragmatic stance, as is usual for them, as well as having a pre-existing agreement, which I was not aware of.
    A second-person “not-a-guard” on densely-populated suburban trains is a very good idea, but SWR do not seem to have a problem with this, which would suggest that RMT are “on to a loser” here.

  3. A very useful update, thanks

    Their claim of 3 or 4 seconds extra is simply not true. It’s often takes that long to wait for the doors to open, and the closing sequence is most commonly 8 seconds. I really notice all the hanging around compared to Southern, TL and Overground, so have got into the habit of counting. Anything under 8 seconds between the doors having closed and the train moving off is a bonus in my experience. I’m not aware of any safety failures since the North London Line went to driver operation of the doors.

  4. PoP…..excellent article as always.

    Pointing out a typo….managing directory! And……

    Something got slightly mangled in the description of automatic braking. The train version of ABS works regardless of whether the control is manual or automatic. The big benefit is again on time: a consistent time to brake into a station and, if a higher braking rate than drivers are trained to use is employed, then a few seconds might be saved. On pure tunnel ATO systems, the run time benefits from high braking rates are as important as improved acceleration, possible more so.

    Also, I’d argue, though that a response to someone jumping in front of a train would be quicker if the driver had his/her hand on the brake controller, than if the driver is merely supervising the automation. That said, train’s stopping distances are such that, sadly, neither approach is likely to allow the train to stop short.

  5. Thanks for another well researched article.

    @NICKBXN “I’m not aware of any safety failures since the North London Line went to driver operation of the doors”
    Not in respect of door opening/closing. But look at the wider picture – the absence of a second staff member on board may have been a factor in the “driver overload” issues that occurred in the Peckham Rye incident ( see https://www.londonreconnections.com/2018/thirty-centimetres-of-safety-an-incident-at-peckham-rye/ and possibly also the stranding incident north of Shepherds Bush.

    I think the RMT are running a poorly directed campaign here. As the article points out, there doesn’t seem to be any safety deterioration if the driver operates the door. But there are wider safety issues about operating trains with only one staff member on duty. That is something I have become less comfortable about, and I think the RMT could win “hearts and minds” if they concentrated on that rather than arcane points about door operation.
    Do we have any statistics of how many times per annum a driver becomes incapacitated “at the wheel”? I’m aware that in the airline industry there are multiple events per year (sadly including a few who have died in the cockpit) which is why two person cockpit crew remains the minimum for virtually all commercial flights. While it is true that a train will “fail safe” if a driver becomes incapacitated (unlike an aircraft) but it’s clear from the Peckham Rye incident that a safety issue will rapidly develop if there is no trained staff member on board to manage the safety aspects of such a ‘stranding’ situation.

    An additional point about second staff members. Expectations about the ability of persons of reduced mobility (PRM) to use trains have moved on greatly in the last few years – but it’s clear that many with mobility impairment still encounter difficulties when they try to use trains. (For those who follow twitter, Captain Deltic has highlighted many incidents where a passenger with PRM needs has been abandoned or stranded whilst attempting to use the rail system). If the railway industry wants to both increase capacity and simultaneously enable PRM to freely use the system, then there has to be a second staff member on board. Every time a driver has to get out to operate a wheelchair ramp then the schedule has gone to hell in a handcart right there.

  6. 100andthirty,

    Typo corrected and paragraph on automatic braking clarified so as not to suggest that ABS is specifically a feature of ABDO.

    On the issue of emergency braking when the train is entering the platform, I based my comment on a TV clip with a Northern line train operator describing their job as they were entering the platform. She made the point that with ATO her concentration was entirely focused on being able to stop in an emergency and she had her hand poised over the appropriate control ready to do so. She clearly implied she thought she could react faster with ATO taking over the normal driving aspect.

  7. As frequent traveller who is virtually tied to SWR for rail transport, I happily confirm that the door-opening delay is a regular and undefendable nuisance that drives passengers mad. As a prominent contributor here would say – the only justification seems to be for “the guard to check that the platform is still there”

  8. Island Dweller,

    A few pertinent points raised on which I would like to comment further.

    The general approach for train evacuations is that the train crew (whether one or two or more) should not be doing this on their own. Normally any evacuation route has to risk assessed at the time taking into account the present conditions (think of the evacuation at Lewisham a couple of years ago). Practically the only exception would be in the event of fire but I suspect then that passengers wouldn’t wait to be told anyway.

    Normally evacuation requires quite a few members of staff. Having one extra crew member aboard won’t make much difference.

    The problem at Peckham Rye entirely stems from the simple fact that the driver broke the most basic of rules and did not contact the signaller first. Had he have done so then the problem could have been escalated in the proper way and a proper safe evacuation carried out.

    I agree with you entirely about the RMTs campaign being poorly directed. But first they need to come to terms with themselves. There is no point in having a strike once they accept that they do have an important role but not one that involves an arcane procedure to open and close doors.

    Drivers do die at the front of the train. Unlike an aircraft (but see below) the train can stop safely automatically, the signaller realises that something is wrong and there is a standard procedure to deal with it. The passengers are not at any risk. In such circumstances the public address system on the train can be accessed remotely to inform passengers on the train. However, drivers dying or becoming totally incapacitated is extremely rare although it did happen a few years ago on c2c.

    A bit off topic but the ‘two pilots’ mentality is something that will be challenged in the forthcoming decades. Remember most planes are small and have a single pilot. Already there is a system available to safely automatically land the plane at the nearest airfield. See YouTube videos on Garmin Autoland such as this one. There is no reason that this cannot be scaled up.

    Furthermore the aircraft industry is already talking about having a single pilot on-board. For take off, landings and emergencies there will be a second remote pilot. It is rumoured that the future Boeing 797 will optionally come with this feature though I suspect the airlines will wait for someone else to adopt it first before slowly introducing it.

    When it comes to disabled access, yes you do need a second person. But they don’t have to be train based! So long as the station is staffed that should be adequate or even better. Using a driver can be appropriate for exceptional circumstances such as boarding at a quiet station. What you also need is flexibility with station staff based at larger stations so that they can accompany or meet the passenger if necessary at the quiet station. Note this is irrelevant to the current dispute as there is no plan to take the guards away – only redeploy them on the train.

  9. Is the label “Guard” a misleading historical artefact?

    I’m not sure what could replace it, some of my what-if thoughts reek of Orwellian Newspeak, but would some sort of passenger service label fit better? And maybe some sort of tech solution for covering the passenger mobility problem, some device carried to signal a hold to the train systems.

  10. Re automatic braking: Most drivers use “braking zones” rather than rigidly fixed braking points. This of course takes into account factors such as anticipated low rail adhesion as well as the braking characteristics of the train (that can vary somewhat even within the same class of traction running the same formation). Additionally however drivers may vary their braking style based on the busyness of the platform and the sighting on approach of the full length of the platform. Certainly some drivers use a more cautious approach than others but at busy platforms most of us have adopted a braking style where we bring the train down to a “reaction friendly speed” roughly at the start of the platform, and in some situations even coasting so that we can apply appropriate brake as we observe passenger behaviour. This is not just about crowded platforms, it’s also about passengers standing beyond the yellow line, protruding umbrellas, young children not held by the hand, drunken revellers and (the biggest bane of my working life) – people transfixed by texting or listening to music on mobile phones.

    Even in my short 10 years as a train driver the issue of passenger behaviour on platforms has become the single greatest challenge I face on a daily basis. I have certainly adapted my drivestyle accordingly. Automatic braking would presumably not take into account any of these factors and without hugely expensive platform edge doors I can imagine drivers blood pressure going the the roof unless we were able to take instantaneous control. I appreciate the likes of the dlr stop trains automatically in platforms without a driver in the seat at all, but if you are there in the seat I foresee a lot of emergency brake applications and very liberal use of the horn.

    One further point (and I’m sorry if this debate has been had before), I’m not entirely opposed to the use of “him” or “he” in a generic way when the flow of an article reads better with it. But there were times when reading this otherwise interesting article, that the role of driver or guard is one that is only undertaken by men.

  11. ‘Industry insiders suggest that the second approach results in a minimum saving of six seconds and potentially a typical saving of ten. Note that once the guard has closed his own doors, he cannot see if it is safe to depart. So this check can only be done by the driver anyway.’

    I have a few problems with this statement. Firstly, my own source, with access to enough data that they should really be in a good position to know, tells me that actually, in practice, guards closing the door is on the order of a couple of seconds quicker. I should imagine this is likely because a quick glance where you can physically see anyone in a dangerous position is much faster than spending several seconds staring at each view on a tiny monitor. In addition, I’m not sure what you mean by your last sentence. When a guard is closing the doors, a driver isn’t involved with the dispatch besides checking that the signal is clear, which will probably already be on their mind by the time they receive two on the bell…

  12. Excellent article, as ever.

    One gripe. If twitter/Facebook is to be believed, the main reason for a comparative lack of noise over the month-long strike is that the service was so dire anyway, it was hard for it be worse (and it may even have been better!!).
    With the best part of 2 weeks’ of the strike being over a period when many don’t commute anyway, and leisure trips to London being achievable by driving to an alternate rail/tube head, it wasn’t the best time for a strike in terms of disrupting daily life, either.
    It seems to me – but of course don’t have comparative figures – that the NR infrastructure is terribly unreliable at the moment. Yesterday, there was a track circuit failure between Earlsfield and Clapham Junction for much of the day. On the up fast – which anyone who knows about the Clapham Junction disaster will realise is auto signalled. All up trains were going on the up slow and, of course, Salisbury and beyond trains were turning at Basingstoke.

  13. Latecomer,

    Thanks for the insight. It will be interesting to see how ABDO gets implemented. Rightly or wrongly such cautious driving does not apply to the Underground which can be extremely crowded. TfL would argue that the national rail network is unduly reticent in applying Underground operating practices.

    It will be interesting to see what kind of braking curve is applied should ABDO get implemented. Of course, proponents of the scheme would argue that what is needed is consistency and it shouldn’t be down to a particular driver. I would presume there would be some override so that, in particular circumstances, the driver could reduce the speed still further but I don’t think such a feature exists on the Underground (other than taking the train out of ATO and using ‘protected manual’).

    It would be better if some of the problems you mention were tackled at source and a properly supervised and monitored platform would assist. So announcement, or having a word with intending passengers, could be used to discourage or prevent issues such as having people over the yellow line or free-range children.

    All that said, I am not unsympathetic to your comments.

    On the issue of implied sexism, I try to be very careful not to use ‘he’ – although I have found a single example that crept in. But I dislike ‘they’ for singular so generally specifically wrote ‘the driver’ or ‘the guard’. It is our policy to write in a gender-neutral style but old habits die hard and examples even get past the editor (who is surprisingly passionate on this issue).

    The main problem was a few uses of ‘his’ or ‘him’ which didn’t raise the same alert level in my brain. I believe I have fixed all the previous references to inappropriate gender-specific pronouns but, if there are more, I am happy to correct them if they are pointed out.

  14. Bear also in mind that larger stations – the ones most critical for dwell time! – are unlikely to be certified for self-dispatch at least in the peak hours, meaning there will be dispatch staff who have to check that it’s safe and then give the signal to the driver. So really, not much different to the case with guards.

  15. Great article and im reminded that Southerns operation was transformed before DOO became widespread across its network on the Electrostar operated services just by granting the driver door release the guard still closed the doors then and buzzed away. This allowed very swift egress of passengers and thus reloading at the busy stations, as well as removing passenger frustration as to why the doors haven’t been opened,
    Whilst there was pain getting to full DOO some of the benefits gained in dwell time have been eroded by sloppy practices at a stations where platform dispatch has been retained. All too often dispatchers aren’t positioned for the train arrival especially when its a 12 car and weather is inclement leading to prolonged dwell times. However, GTR didn’t stop there and on Thameslink they moved to full driver control in the London Bridge to St Pancras section, despite it being considerably busier then most Southern dispatch stations, but even under heavy loading conditions the drivers achieve very good dwell times. This demonstrates what can be achieved and the benefits to operations so SWR (with DofT backing) have to push this through. There are operators who still have this arcane practice of the guard taking a platform stroll before releasing the doors.

  16. Muzer,

    “When a guard is closing the doors, a driver isn’t involved with the dispatch besides checking that the signal is clear”
    If that is true then surely having the driver close the door and check after he has closed the doors is then safer? My recollection, which may be wrong, is that even in the days of EPB stock drivers poked their head out the window and looked backward as they moved off.

    Platforms 4 and 5 (Thameslink) at London Bridge are self-dispatch. I also believe that Farringdon (Thameslink) is self-dispatch and that is a really awkward, busy, curvaceous platform.

    But surely, regardless of whether fully self-dispatch or the signal to close the doors is given by platform staff, the saving in time, compared to now, will be roughly the same in both cases.

  17. Re Mike P

    Yesterday, there was a track circuit failure between Earlsfield and Clapham Junction for much of the day. On the up fast – which anyone who knows about the Clapham Junction disaster will realise is auto signalled. All up trains were going on the up slow and, of course, Salisbury and beyond trains were turning at Basingstoke.

    The track circuit “failure”* occurred as the 2nd pair of modified 442s in the am peak went through!
    It appears the quick set of mods isn’t working as expected. Previously the 442s had just been upsetting track circuits in front of them, this is a first recorded affecting the train behind as well.

    *it didn’t actually “fail” merely just get upset by a pair of 442s…

    The power draw and conditions in a full am peak seem impossible to replicate at other times for testing..

    Thankfully new static converters for the 442 are on the way.

  18. Great and informative article as always.

    As a slight digression, I assume the layout of the unrefurbished class 455s was partially determined, as with the Networkers, by BR’s method for determining train capacity. In this the standing capacity was deemed to be 35% of the number of seats (10% for slam door trains). This meant that the more seats you squeezed in – and the less standing space you provided – the more the nominal standing capacity increased. Great on paper for persuading the Treasury to part with some cash to buy extra trains, but not so good in reality.

    Having a defined capacity for a type of train is useful for a number of reasons, such as determining how effective a new timetable will be, producing economic appraisals of new projects, understanding how bad crowding is across the network, and ensuring that car seating plans are appropriately laid out for the type of traffic they’ll see. The method of calculating these capacities has changed a few times since BR days under OPRAF, the SRA, and DfT, to try and achieve a sensible compromise between these often conflicting requirements.

    The SWT 455s were, of course, refurbished by removing the poles just inside the doors, large stand-back areas so people weren’t obstructing the doors when leaning against the draught screens, 2+2 seating with wide aisles to encourage people along the train, larger open areas for bikes, luggage, push-chairs and standing passengers, and (I think) the doors were made to open to almost the full width of the aperture as well. This helped a lot, but you still had relatively narrow doors, doors at one third and two thirds along the length of the car instead of one quarter and three quarters, and some nasty stepping distances.

    Incidentally, I have seen in the past a rather complex formula that estimated door open to close time taking into account number of passengers boarding and alighting, number of ‘door sentries’ partially blocking the door, door width, step height, platform gap width, and so on. It’ll be interesting to see how dwell times on the level-boarding Flirts work out – how much does the level boarding reduce the times? These units aren’t being used on high intensity commuter routes, and the single door per car would make them unsuitable, but I’m still intrigued to see how the boarding times work out.

    (Side grammar note: Using ‘they’ as a singular dates back to at least the 1300s. it seems that the reason the singular form has been disparaged is because an 19th century grammarian didn’t like it on some spurious ‘logical’ grounds, and determined that it was bad English. ‘You’ was also plural once, but its transformation to a singular was successful.)

  19. I was on a train from Vauxhall to Clapham Junction this morning. There was a guard on board, and yet the on-train station announcements were out of synch. Surely, this should be a primary responsibility of the guard?

  20. What I have found strange about the class 707 trains is that dwell times on those units seem even longer than on the classic 455’s that have been around since the 1980’s. Of particular concern is the extended delay before the doors are even opened.

    I would agree with others that the SWR dispute got less coverage because too many users expect the service to be poor. Most of my friends live in SWR land. They all believe the service is worse than it was under SWT.

  21. @Gareth Gouldstone
    Out of sync announcements are a regular occurrence on SWR, especially in my experience on the 458s. Sometimes the guards make corrective announcements, but most of the time they don’t seem to bother, possibly not appreciating the impact this has on irregular and disadvantaged passengers.
    You’d have hoped they’d at least be able to switch the automated announcements off when they’re not working correctly.

  22. Re; NGH & 442s
    NR state that the track circuit failure was not related to the passage of the 442s. The SEG asked their twitter feed – so the old modified convertors appear to be working.

  23. I was once on a SWT train at Clapham Junction when upon arrival the guard opened the doors on the wrong side of the train onto the adjacent live track. How that was possible I don’t know – it was probably 10 years ago but from what I recall was a fairly new train at that time. He hastily closed the doors but then somewhat sheepishly he told us we couldn’t continue until a replacement guard arrived or someone to accompany the train/him to Waterloo. I would have assumed there were safeguards in place to prevent this but wonder if there is a difference in the ability of the guard or the driver’s systems to allow/prevent this from happening.

  24. Pedantic “But I dislike ‘they’ for singular so generally specifically wrote ‘the driver’ or ‘the guard’.”

    There’s not a man I meet but doth salute me
    As if I were their well-acquainted friend
    (A Comedy of Errors)

    Good enough for Shakespeare, 400 years ago. Should be good enough even for Pedantic of Purley

  25. @MikeP:

    “If twitter/Facebook is to be believed, the main reason for a comparative lack of noise over the month-long strike is that the service was so dire anyway, it was hard for it be worse (and it may even have been better!!)”

    I had an alternative route, using Southern rather than SWR/LU, during the strike … and it was enlightening. Now that the strike has been resolved, the alternative route has become my primary route. I will not be using SWR again unless I am forced to.

    The alternative route was on paper about five minutes longer but, in practice, took less time because a persistent 7-8 minute delay on my (SWR) journey – of 40 minutes – has crept in over the past few years and SWR is either incapable of removing it or unable to remove it. I suspect that the second is the case, given the infrastructure issues described here.

    There was merely one significant delay, caused by trespass.

    The dwell time problems were also made only too obvious. No more waits for the light to come on as a prompt for the door button press …

  26. Re PHIL R

    I was once on a SWT train at Clapham Junction when upon arrival the guard opened the doors on the wrong side of the train onto the adjacent live track. How that was possible I don’t know – it was probably 10 years ago but from what I recall was a fairly new train at that time.

    Easy: 100% manual with no automated safety systems hence human error – No ASDO or CSDE on SWR at that point in time and only in limited amount for 458s when extended to 5 car units for 10car operation.
    Compare to Southern/Thameslink next door being the national rail home of such innovations…

  27. In the sections regarding station duties you have
    “Even if it only applies to dispatch, this low figure would doubtless be disputed and it is hard to see how it can be as little as this given the actions the guard must take:

    See that it is safe to close the doors
    Check the signal aspect if one is visible
    Close all the doors except their own
    Check again now the doors are closed
    Close their own door
    ‘Ding-ding’ the driver to confirm it is safe to depart.

    After all this, the driver then has to ding-ding the guard to acknowledge the signal. Only then can the train depart.

    Compare the above procedure with the situation where the driver closes the doors which is that the driver:

    Checks it is safe to close doors
    Closes doors
    Checks it is safe to depart (no-one stuck in doors or leaning against train for example)
    Departs”

    It mentions the Guard would “Check the signal aspect if one is visible”
    But it does not mention the driver would do the same so there is only the process of the Guard closing the local door that is different.

  28. What it comes down to is
    A. Drivers are in ASLEF, guards are in RMT.
    B. Mick Cash is not Bob Crow.

    But we used to have much shorter dwell times. I recall doing some timings at Blackfriars. A train of EPB stock bound for Holbon Viaduct could arrive, unload, load and be on its way in 20 seconds. A following Thameslink class 319 took around two minutes to do the same. Admittedly the 319s have about the worst designed doorways and vestibules on the network.

  29. PoP…..noted your point about the Northern line driver/ATO. However, when commissioning the new Victoria line, we noticed that operating the emergency brake in and around the station didn’t usually reduce the stopping distance compared with letting the ATO do its job. Sometimes the stopping distance increased. What was/is this? On a modern train the overwhelming majority of service (ie ATO) braking effort is delivered by the dynamic brake. The emergency brake is delivered by the electro-pneumatic friction brake. If a train is already braking and the emergency brake is applied, the following happens. 1) simultaneously the dynamic brake is released, and air is admitted to the brake cylinders. 2) the air pressure takes up to 2 seconds fully to apply. 3) whilst the air pressure builds up, the train slows at a lower rate and 4) once the air pressure has fully applied braking at the emergency brake rate resumes). Underground the emergency brake rate is very close to the ATO service brake rate. In the open brake rates are lower. But the point remains that stopping short of an obstruction that falls in front of a train in a platform is very much in the lap of the gods; the driver may or may not be lucky.

    Your article and the various commentators lead to a much wider discussion about how the railway should be staffed and whatever training they are given. The points about dealing with an incident on the train and helping people who need it are well made. Equally I am convinced that people who work on stations should have safety training so that they can help with evacuation too. Moreover, the whole situation for people with mobility assistance needs requires review. The compliance that is just about achieved for the rolling stock reflects societal expectations from 25 years ago. Expectations have moved on, not least the reasonable expectation that once can turn up and go!

  30. @Joe Grey As an slightly larger digression on your digression, I read your use of the term ‘door sentries’ and had a good chuckle.

    For those wondering who they are, people who stand (at attention) at the edge of doors facing each other, but that block entrance and egress.

  31. @PoP under “The RMT Case” – This is especially true as the guard will be able to ensure the train physically cannot depart whilst HE is assisting a passenger on the platform”, and last paragraph under “Is It Worth Those Extra Seconds” – “Note that once the guard has closed HIS own doors, HE cannot see if it is safe to depart. So this check can only be done by the driver anyway.”

    In relation to your other comments. Passenger behaviour on platforms is much better managed on those platforms where there is a well trained and capable member of staff whose specific role is to focus on the safety of waiting passengers and intervene or offer advice as necessary. Second best is live “human” announcements especially where specific advice is being issued. Automated announcements have a role but these get disregarded through familiarity and are not heard by those wearing headphones.

    Regarding the Peckham Rye incident, you are right in that the primary cause was because the driver broke the rule book, however, subsequent events led to driver overload. The lack of clarity by the signaller and the fact that a controller was distracted by another event. A new technician who took over from the previous person they relieved got the driver to repeat all the previous fault rectification checks which involved the driver walking through a train for a second time with increasingly hostile passengers. All these issues compounded the drivers mistakes. If another member of staff had been on board (like a guard or onboard supervisor) they would have deflected a lot of the stress from passengers, reduced the stress overload and would have likely intervened to prevent an uncontrolled evacuation onto a live line.

  32. Is it not possible to ask the RAIB what proportions of PTI/dispatch incidents involved guard, platform and driver(self) dispatch? My impression is that there isn’t much difference.

  33. Latecomer,

    Further generic-specific statements corrected.

    Clearly a second person on board would have helped in the Peckham Rye incident and ought to have taken some of the stress off the driver. I note that you say guard or onboard supervisor so the safety role is not fundamental (as in ‘safety-trained’). It is why I think there needs to be a debate as to exactly what the second person is for. For example, this may be justifiable on a day-time service through Peckham Rye but it doesn’t follow that it makes sense on a Sunday morning service to Tattenham Corner. And should the service be cancelled if this second person is not present?

    Whilst not personally of this opinion I heard a former SWT chief executive, Christian Roth, support the idea of a second person on (then) South West Trains on the basis that a second person can reduce delays due to on-train incidents (e.g. medical situations). He felt that they were potentially justified on that basis alone and went further and stated that, if the train was not walk-through, he would like a member of staff on each part of the train.

  34. Richard,

    You are probably correct. Anecdotally there are many RAIB reports where one thinks ‘if only there was a guard present’ but there are also reports where the presence of the guard let to confusion as to what was going on and one thought ‘if only this had been left to the driver’.

    TfL have produced evidence to show that Gospel Oak – Barking is actually safer without a guard and one reason against using trains with a driver and a guard during the problems with the class 710 introduction was that it would compromise safety – a general principle being that a variation in procedures must be safer than the current situation.

  35. A good article. However, I think one angle has been missed.

    Most of SWR’s suburban guards do not carry out any revenue duties. There is a separate grade of Commercial Guard for that. I think this situation is unique on our railways and is a hang-over from the failure to introduce DOO out of Waterloo.

    The guards are merely there to operate the doors and carry out any required safety-critical tasks. Although they’re supposed to patrol the train and answer enquiries etc, far too many sit in the cab for the whole journey and don’t even make announcements when the automatic system isn’t working.

    The leaked SWR proposal from ACAS shows that SWR wants the guards to be patrolling the train regularly, using the door close inhibit key when required as described in the article. My guess is that some guards don’t want to sacrifice being able to sit in the cab all day long.

    Om dwell times, it’s ridiculous on SWR. You can get good guards who carry out the procedure very quickly and others where it seems to take a lifetime. Such inconsistent dwell times are so obviously wholly inappropriate a for a high frequency metro service.

  36. @Steve “far too many sit in the cab for the whole journey”
    As a regular SWR traveller I have also noticed this. Most galling are the ones who cheerfully read the scripted announcement that says “You can find me in the middle of this 10 coach train” when they’re actually in one of the middle cabs behind a locked door, so not very findable at all.

  37. Pedantic of Purley. A great and insightful article as always. With regards to your comment about not needing a second person on board for PRM if all stations are staffed. Not all stations are staffed. Some are only staffed during certain hours and others are not staffed at all. Would SWR be happy to employ more staff to ensure all stations are covered from first train to last, including relief staff to cover illness/ annual leave? Certainly from a security stance all passengers would be happy to know all stations are fully staffed.

  38. Some very thoughtful comments here – what this site excels in.
    @POP. Re automatic systems taking over from pilots. I am clearly far more skeptical than you about that. Well programmed systems can fail (think Air France from Rio to Paris – automatic systems and software protection dropped out due to unreliable airspeed indicators, though it’s ironic the reason everyone then died was appalling lack of basic flying skills from the AF crew). Alternatively, automated systems can be very badly programmed (Boeing 737 max I’m looking at you).
    I think getting rid of human pilots on commercial aircraft is a long way ahead

  39. Brialliant article, as ever.

    Two key differences with the Southern strike:

    1) the RMT and ASLEF came to an agreement that there would be no further introduction of DOO, and thence when RMT were in dispute with Southern on the matter, ASLEF swiftly followed. Whilst Southern could operate a service with contingency guards etc with RMT on strike, they couldn’t do it when ASLEF came out. ASLEF were then offered the king’s shilling, did a u-turn (much to the RMTs displeasure) and signed up. (The safety argument mysteriously being resolved)

    2) the majority of the Southern fleet was already DOO capable. Thus when ASLEF withdrew their dispute, DOO (or more strictly, Driver Controlled Operation, DCO) could be swiftly implemented without delay. This is not the case on SWR where none of the fleet is DOO/DCO capable, at least not without modification. Hence RMT striking early (in both senses), as once the new trains are in, short of a dispute with ASLEF the strike becomes irrelevant.

    As I understand it, the striking guards have received strike pay from the RMT (funded partly by contributions from other unions) that has effectively compensated them in full for basic pay over the month. Strike pay from a union is not taxable or subject to NI, so the actual amount is less than normal pay. However it must be a significant 7 figure sum.

    For what it’s worth, I’m a little surprised that an independent ‘authority’ hasn’t been commissioned to demonstrate the difference in safety of DOO services, and / or to seek passenger views. Having commuted on DOO services for the best part of two decades I have never once felt that the absence of a guard was in any way less safe than a train with one. Speaking to passengers (as I do) what most actually want is some form of visible customer service, assistance and security, particularly on trains that are delayed or run in the late evening. If you ask a passenger ‘do you want a guard on the train’ you’ll usually get an answer in the affirmative (albeit less so on lines that are already DOO); if you ask them ‘what do you want them for?’ you’ll be hard pressed to find anyone who mentions doors. A happy solution for me would be DOO, with very visible, friendly and authoritative customer service staff appearing on trains where and when they are most likely to be needed, which would not include peak services.

    Finally, given that well over half of all rail system passengers are travelling on DOO services today, and over three quarters of passengers on suburban systems, it seems odd that the RMT are sticking to the safety argument. Perhaps we should have guards on all the other services that operate from Wimbledon.

  40. Re SFD,

    If you ask a passenger ‘do you want a guard on the train’ you’ll usually get an answer in the affirmative (albeit less so on lines that are already DOO); if you ask them ‘what do you want them for?’ you’ll be hard pressed to find anyone who mentions doors.

    Quite – when MerseyTravel did ask about the role of second member of staff before tendering and ordering new trains as DCO only – the main response was for “bouncers” in the evenings which MerseyTravel agreed with as it would save them ~£4m a year. Now their Mayor folded to RMT pressure and guard panels are having to be retrofitted and they have a large hole in their operating costs as they had put the £4m annually towards the new trains maintenance contract (inc. depot rebuild).

  41. Unfortunately there are a lot of mistakes in this article.

    The 442s are still causing problems

    The 707s are being dumped purely because SWR managed to buy other trains cheaper

    Those replacement trains are now over 2 years behind schedule and none have yet been introduced

    SWRs latest timetable had significantly *increased* dwell times in an attempt to reduce their delay stats

    SWR got £25M compensation for the recent strike, *plus* paid less staff wages, *plus* paid significantly less Delay Repay compensation to customers, so would have loved the strike to continue

    Ultimately no one noticed the strike because the normal service is so dire that effectively the strike has been played out continuously since August 2017

  42. Strikes ballots are only valid for 6 months after voting closes, the result was announced on 24th July 2019 so RMT had to call a strike in either December or January and any strikes after this month will require another ballot. As Jimbo points out the dispute has been dragging on for over two years so obviously this was not the first ballot, the previous one was back in February 2019.

  43. Unhappy of Hampton: “SWR got £25M compensation for the recent strike, *plus* paid less staff wages, *plus* paid significantly less Delay Repay compensation to customers, so would have loved the strike to continue”

    Interesting that in news reports yesterday (yes I now this is not a “news” site but the timing and content is IMO pertinent to this thread) :
    “The strike-hit South Western Railway (SWR) franchise faces the prospect of nationalisation after its accounts revealed “significant doubt” that it could see out the year without going bust.”

    And what do they claim in mitigation? “The firm has been in talks with the Department for Transport (DfT) to claw back money for losses partly ascribed to infrastructure issues overseen by the government-backed rail system owner Network Rail, the delays in implementing new timetables, as well as industrial action.”

    So while they, as many employers in their position would, play down the impact of strikes while they are on, they claim that they are a factor in their financial plight.

    I note the comment above about deliberately lengthening dwell time to play the game of avoiding penalties for late running. If the technology and staffing developments outlined in the article come to fruition the heat will be on, as one imagines expectations will be for tighter adherence to schedules.

  44. Sad Fat Dad – Southern ASLEF drivers initially rejected the “King’s Shilling” when it was offered in June 2017, called strikes in July (later called off) and reinstated the overtime ban, They finally voted to accept a deal in November after the safety issues had been addressed (not “mysteriously being resolved”) along with strict conditions for operating without a second member of staff on board.

  45. Re UNHAPPY OF HAMPTON

    Unfortunately there are a lot of mistakes in this article.

    And plenty in some of the comments

    The 442s are still causing problems

    The article was written before Monday’s “problem” hence unknown whether problem solved at that point…

    The 707s are being dumped purely because SWR managed to buy other trains cheaper

    Also key specification issues and the benefit of a common fleet so not purely.

    Those replacement trains are now over 2 years behind schedule and none have yet been introduced

    They aren’t 2 years late, try 1 year, expect the first to be in service in late spring with the first unit being delivered soon.

    SWRs latest timetable had significantly *increased* dwell times in an attempt to reduce their delay stats

    No it is called being realistic about dwell times (suggest you read the franchise ITT). The railway uses 30s timetabling increments so, if the typical dwell time further out is 45s in the peak, you need to recognise that and stop believing that it is the magic 30s in the timetable when it was operated by a 4VEP/EPB and be realistic – often this means 60s, 60s, 30s or similar. What you won’t like is that now means 60s mostly or 90s in places.
    Passenger numbers have grown (doubles during SWR’s tenure) so thing have to change to enable on time running.

    Southern did the same in May 2018 and has managed to run to time better than than at some point late in the slam door era…

  46. Joe Grey
    Step height On ex- LSWR routes, this, yes.
    The LSWR seemd to have had a much lower “standard platform level” than many railways & this legacy is still with us, so that, even on straight or almost-straight track, it can be a long hike either up-into or down-out-of a train.
    Never mind places with curvature like P8 @ Vauxhall or P8 @ Clapham Jn.

    AP
    They all believe the service is worse than it was under SWT.
    A professional railwayman friend commutes Woking-Clapham/Croydon, daily. His judgement is that SWR are consistently worse, & less “on the ball” than SWT were.
    See also perstreperous, & Unhappy of Hampton – made me smile that did. ( Ultimately no one noticed the strike because the normal service is so dire that effectively the strike has been played out continuously since August 2017 )

    Andrew Bridle
    The course of action you urge has already been taken by TfL for their “Overgound” services.
    DCO + platform staff. And it seems to work very well, especially compared with my fairly regular trips to SWR-land.

  47. To gain a few more seconds, could ASDO be programmed to start opening the doors before the train comes to a complete halt? I’ve only ever seen it on the Paris Métro, where the manual door levers are unlocked once the speed falls below a certain level. It certainly appears to improve dwell times. (See what I mean here: https://youtu.be/RGcfC9xijGM )

  48. “could ASDO be programmed to start opening the doors before the train comes to a complete halt?”

    There are serious safety issues with opening doors before the train has come to a halt so whether it could or not is irrelevant,

  49. I have one comment and a slightly cheeky request.

    As a passenger on the Hounslow Loop, I think there’s a remarkable difference in how quickly the doors on the latest trains (class 707) open compared to the slightly older (but still newish) trains still sometimes used (class 458/5 I think). If I’ve got these right the 458/5s must add seconds to each station stop. The guard presses the button and you hear a ‘beep, beep, beep’ before the doors slide slowly open. On the 707s they seem to ping open immediately.

    My cheeky request is to ask if anyone can clarify whether the 455s will all be withdrawn in 2020? Or more specifically on the lines to Shepperton and the stopping lines to Guildford and Epsom? Have considered moving house to some areas on these lines but would prefer my daily commute to involve a little air-con in the warmer months…

  50. So, if signalling staff were traditionally called “Bobby”, should we start calling the guards “Gino” (Guards In Name Only)?

  51. Re Andrew M and ASLEF Shrugged:
    This is already the case on the Victoria Line – the doors begin opening while the train is still moving, with the accelleration/jerk causing the two leafs to open at visibly different rates. It is less extreme, however, than the Paris Métro example, as the doors are only half open when the train stops, which prevents passengers disembarking while the train is still moving.

  52. Of course in the not too distant past it was the passenger who opened and closed the doors on the slam door stock. I like many others developed the skill of opening the door before the train had halted (at Waterloo) and jumping off without falling.

    Over the strike period I and many fellow travellers were impressed at how well the services ran and, in particular, how good the contingency guards did their at job.

  53. “There are serious safety issues with …”. This could be rephrased as “Some people believe that … is seriously unsafe”. (May well be some whole groups of people). The appropriate response is for discussion and research to take place, to evaluate (and ideally agree) the likelihood and potential severity of any bad consequences – multiply these together, set against any benefits from … , and conclude whether or not … should occur. Shutting down discussion is rarely the best approach.

  54. Re door opening times: it is instructive to compare the time between wheels stop and door opening commencing on a Class 700 at St Pancras to that of, say, a Class 350 at Euston. There’s a good 10 second difference. And when the ATO is in (fuller) use on Thameslink, the doors will be opening instantly on wheels stop, rather than after a second or two.

  55. And highly relevant that there are no plans to introduce 24tph working on Thameslink without full ATO in operation on the central section. The only effective difference (as far as dwell times go) between full ATO and ASDO is that the acceleration curve on starting from the platform is under computer-control rather than human-control.

  56. Changing trains at Wimbledon or Clapham Junction, it’s instructive how passengers on Thameslink or Southern are normally ready to get off the minute the train stops. On SWR regular travellers often don’t even bother to get up until the train has come to a stand. They know they will have plenty of time before the guard eventually releases the doors.
    On mechanical speed of door operation, I seem to recall reading that one reason Thameslink went for pocketed doors was that they are inherently faster in operation than plug doors.

  57. JIMS ….The decision on pocketed doors was made based on perceived reliability at the time, not directly speed. Both Electrostar and Desiro introduced sliding plug doors in quantity in the 2000s. The former respond well and open and close with what feels to be appropriate speed. Conversely, the latter are positively lethargic, albeit they close with a satisfying “clunk”. When the decision came to be made for Crossrail, there was a predisposition to rule out sliding plug doors, again, on grounds of reliability, but common sense prevailed, and I think a comparison between the class 700 and the class 345 will show that sliding plug doors were the way to go. Sliding pocketed doors give lots of disadvantages in terms of space used, clarity of view from the DOO CCTV equipment and the space taken up by the door pockets.

    Apropos comparisons between class 350 and class 700, not only is there the lethargic response of the doors to the open command, there is also the challenge that the train must stop (even a complete stop!!!!!), the guard must open his/her door, the guard must step onto the platform and check that the train is correctly berthed, and only then can he/she reach into the train and release the doors! No wonder it’s slow!

  58. Thank you – a good analysis as far as it goes but you are right and wrong about the desirability of a uniform rolling stock inventory. For stations as far out as, say, Surbiton the need is for ‘metro’ ‘ type ehicles (as recently introduced on London Undergound Met line) which provide adequate seating for off-peak travel and lots of standee room for peak movements. The doorways appear to function very well in ‘crush’ situations. Further out, where the housing growth is taking place, the need is for more seats as journey times are in excess of30 – 40 minutes. These trains also call at stategic stations to allow interchange, but the movements are predominantly ‘on’ or ‘off’, so while the doorway/ vestibule design is important, it is not as crucial as on ‘metro’ services. SWT talked about double-cecking for these ‘outer services’ but nothing seems to have come of that idea. It should be looked ata gain as line and terminal capacity inwards of Wimbledon / Richmond is now fully used. The majority of the SW route is ‘overground’ and the cost of lifting bridges or lowering track should be balanced against provision of extra ‘right-of-way’ facilities.

  59. @VideoP
    The desire for a uniform fleet of new trains is for [inner] suburban services, replacing the 455s/456s/458s, with the 450s being retained for medium distance and 444/442s for longer distance. Matching the internal layout and number of seats/standing space to the range of peak and off-peak journey types made on each stock type has to involve a number of difficult compromises, which are never going to please everyone.

    As for double decker trains, it’s been discussed extensively elsewhere that they lose more capacity in dwell time than they gain in floor space. In other words, because it would take so much longer for passengers to get on and off a double decker train, trains would have to spend longer stopped at stations meaning that fewer trains overall would be able to run.

  60. Please clarify the operation of the doors in ABDO mode. With the present DOO metro fleets, doors become available to open when the ‘Open’ button becomes illuminated. The implication in the article is that with ABDO all doors will open, whether or not someone wishes to enter or exit (rather like a tube train). Since the metro fleets spend their time outside, rather than mostly in warm tunnels, a lot of heat will be lost in winter, particularly outside the peaks (and cool air from the aircon on hot summer days) if all the doors open at all stops.

    Surely ABDO will simply offer the ‘Open’ illumination a little earlier? All doors open at every stop would be tremendously wasteful of energy.

  61. @RAYL. Isn’t this what already happens in the Thameslink core. All doors open automatically.

  62. @RayL

    If opening all doors can significantly reduce the dwell time, then the extra heat lost (if any) is likely to be offset at least partially by the fact the doors will be open for less time.

    At busier stations this will be especially significant since most doors would likely be opened there anyway.

    Or, as I’ve written about on here before, using “door ordering* (requesting a door be opened using the button before the train has stopped, with it then opening automatically when this has happened) would allow the ABDO system to be used without opening all the doors where this is not needed, whilst still reducing dwell time.

  63. @RAYL

    To save the dwell time at a station is necessary for the doors to open at all stations. This is what the tube trains do – even when “out in the sticks” at Epping, Grange Hill.

    Of course, the system without drivers, the DLR, requires someone to press the door open button.

    The Met trains auto-open, but seem to have a timer to close after a few minutes. You can see this at Amersham/Cheshunt/Watford and the Edgware Road Circle Line trains.

    Remember the famous fact that only 45 percent of the London Underground is underground…

  64. Re 100ANDTHIRTY

    JIMS ….The decision on pocketed doors was made based on perceived reliability at the time, not directly speed. Both Electrostar and Desiro introduced sliding plug doors in quantity in the 2000s. The former respond well and open and close with what feels to be appropriate speed. Conversely, the latter are positively lethargic, albeit they close with a satisfying “clunk”. ….
    Apropos comparisons between class 350 and class 700, not only is there the lethargic response of the doors to the open command, there is also the challenge that the train must stop (even a complete stop!!!!!), the guard must open his/her door, the guard must step onto the platform and check that the train is correctly berthed, and only then can he/she reach into the train and release the doors! No wonder it’s slow!

    Bombarider + Desiro City (700, 707, 717) doors are electrical operated, the original Desiro doors (185, 350 ,444, 450) are pneumatically operated (like most older sliding door installations,

    One suspects that there will be a demand for new door mechanisms for the 350s ,444s, 450s at some point…

  65. NGH………….yes, indeed, Desiro sliding plug doors are pneumatic, but it would be wrong to say that they are slow because they are pneumatic. I have travelled on many train (Tyne and Wear metro is an example) where they have pneumatic sliding plug doors that operate quite smartly. I don’t understand why the design on the Desiros was ever accepted. I wouldn’t have accepted them!

  66. A very carefully researched article – as ever! May I make on or two additional comments:
    – whatever RMT may say, requiring the guard to carry out a visual check before initiating door closure, is not either a safety panacea, or indeed, often feasible. I am thinking here of very sharply curved platforms (Godalming in SWland a very good example), where there are no positions on which anyone could see the entire length of train; also, those very crowded stations, where only some of the punters are leaving on the departing train, the guard cannot get any sort of overview from the platform and level boarding (not available at Godalming…) means that he will not get a better view from leaning out of the train.
    – SWT in particular has, along with many other operators, extended journey times to maintain performance targets, and this has the side effect that a substantial number of trains arrive several minutes before departure times (again in SWland, Woking and

  67. ,,, (and here is the rest of my comment which was lost when, Aksim, presumably, decided to run a loopy/looping scan on what I had written so far) and Guildford are especially bad. This destroys any dispatch discipline or,, in the minds of platform staff, any need to worry about time. I have yet to see any modelling on the interaction between timetable slackness and despatch speed but I would expect it to be complex and related to size of station
    – I recall that when advising one of the bidders for the TLK fleet, for which I was the bankers’ nark), they concluded that there were at least 17 processes involved between train stopping and doors opening. Many of these appeared to be systems crosschecking between themselves, which is why they are invisible to even the better informed punter, and whilst they may have been no more than a second each, cumulatively, they took up several precious seconds. Alas, I failed to take away (or steal as we might say to use a technical term) a copy of that list…
    – Chris Green always referred to those who stand back by the doors as “sentinels”.

  68. @Brian
    Chesham not Cheshunt. Doesn’t help that today they are both terminal stations on TfL services.

  69. DM1
    “Door ordering” works v well on Croydon Trams …
    And Brian B – disagree
    No ( I think) .. that it is empahtically not necessary for all doors to open, in fact especially at places like Loughton / S Ruislip / Debden in the winter – though this means having a reliably working button on the outside.
    Which, oops, SE/Southern/SWR/TfL_Overground/etc seem to manage to do quite well ….

  70. @Greg

    Croydon trams (or at least the Bombadier trams – I haven’t worked out the door options on Staedler trams) appear to have two options for the driver – unlock doors and open doors. Unlocked doors will only open if the passenger presses the button – including if you’ve pressed the button in advance. In practice most of the time the drivers use the open doors option which opens all the doors – and if they don’t there’s usually a delay until the passengers realise they need to press the button for once.

  71. You are purely looking at things from a suburban point of view. SWR do operate long distance “Inter City” services. Many stations are destaffed or unstaffed after lunch times. The stations are often of a design where assistance is needed, are we to offload this onto drivers too? Or are we saying to mobility impaired passengers “naff off, you are not welcome on the railways”. A patronising and selfish response there and under the equality act, something you seem to be ignoring in your ramblings, this is not possible, unless you are intent on getting your friend Boris to repeal this law in line with anything else regarding human rights.
    What if someone gets attacked on a train, do we leave them to bleed to death, police down in these rural parts are few and far between and the likelihood of any pepetrator being aprehended is minimal. Also respect the fact that who is there to look after or attend to a passenger should they be taken ill or call the paramedics should medical attention be needed.
    The role of the guard is many fold, not just opening and closing doors. This measure is purely for profit motives only (not as if they are not making large profits from the taxpayer and passenger for a shoddy service already). Technology cannot be allowed to be our master, it should be our servant. What would happen if there were a failure of the on train technology? Passengers inconvenienced? Delays as you sort out the ensuing shambles?
    I speak as a lifelong rail enthusiast, rail traveller and former employee.

  72. Paul, VideoP,

    As for double decker trains, it’s been discussed extensively elsewhere that they lose more capacity in dwell time than they gain in floor space.

    Yet on the Paris RER, they’ve recently expanded their use of double-decker trains. Line RER-A used to have a mix of single- and double-deckers, but the latter proved so popular that in 2011 they switched to an all-double-decker fleet. I don’t have any figures for dwell times, but the published timetable shows 22tph between 7am and 8am (when they aren’t on strike). Not too shabby for heavy rail.

  73. Martin Mynard,

    I think you are getting a bit carried away emotionally and reading things into the article and comments that simply aren’t there.

    For a start, the main thrust of the article is about the role of the second person on a SWR train and any comments about driver-only trains have been only peripheral. The article hasn’t suggested getting rid of the second person on SWR – only having a rethinking as to what their purpose is and plan accordingly.

    South Eastern javelins already operate without a ‘guard’ but do have a second person. On the GWR the new Inter-city trains are designed so that doors can be open and closed by the driver but it hasn’t been suggested, to my knowledge, that the idea is not to have a second person aboard.

    In any case you are overlooking that on longer distance trains there is the very practical need for revenue protection. Either you have properly staffed stations to discourage fair evasion (in which case disabled assistance can be given by station staff) or you need an on-board capability to issue tickets and check that people have paid (in which case you have a second person on-board).

    You then talk about someone being attacked on a train and bleeding to death – a bit dramatic as trains are actually very safe (and carriages are usually covered by CCTV). But isn’t that the whole reason a discussion is needed? If the real role for ‘guards’ is public order then surely you should be employing people more suited to it – in a security and medically-trained role. And I fail to see how a SWR guard who locks himself away can even be aware such a thing is going on let alone assist. See previous comments on Merseyrail.

    I suspect the way you actually get help is for passengers to call the police and for the emergency services to arrange for the train to stop in an appropriate place so that the incident can be addressed. And could I suggest that sometimes calling for assistance is all that the guard can do.

  74. To enlighten the debate (and indeed myself) could someone confirm which of the activities identified by Martin Mynard – boarding assistance, passenger information in disrupted service, calling paramedics in the event of injury, etc. – are in fact included in the formal role of a guard, in particular a South Western suburban guard?

  75. BALTHAZAR @21.09, yes I’ve experienced non-commercial guards on SWT/SWR doing all those things. Under SWR’s proposals they will have more time to do all these things because they won’t be tied up with operating the doors every few minutes. Surely a win-win situation.

  76. “…on most lines SWR demonstrated it was able to run its limited-capacity service using ‘contingency guards’ – management trained up in the guard’s role”.

    Does anyone have a feel for how many ‘contingency guards’ have been utilised during these strikes? It does lead one to wonder what was happening to their ‘normal jobs’ while they were out acting as guards. If their normal roles were left uncovered, you might ask whether they are really essential to the running of the railway.

    Scope for financial savings, anyone ?

  77. @PoP, Martin Mynard

    I witnessed a physical fight on a train a few months ago. The guard was nowhere to be seen, even after I pulled the emergency handle to speak to the driver. The train was halted for a good while at the next station while local police arrived. I only saw the guard in person after the offenders had been escorted off the train. (I offered to make a witness statement but no one took me up on it. There was CCTV.)

    As a woman (I hate that phrase), the use of a generic “he” doesn’t bother me in the slightest. What does bother me is sitting though belligerent and antisocial behaviour from other passengers, especially in the evenings, because the guard is sitting in a cab between stations. Physical violence is rare. Drunken harassment isn’t. In the evenings at least, a second person is needed as a bouncer.

  78. Re doors opening before train fully stops…happens regularly on Bakerloo, doors just starting to open in last second or so of braking, so you can’t get out.

    Re support for passengers needing boarding ramps or other assistance. I think it is already a challenge to handle this in peak (space and dwell time) and I wonder when the railways will give up trying to serve passengers who need assistance in peak at every station and set up dial and ride minibus service (from home) instead to hub stations offering dedicated provision. Maybe even to final destination station or door in some cases.

    Some shocking videos on YouTube which shows how unreliable assistance is generally, especially multioperator long distance.

    It is sad that assistance is not working and may need to change…not saying that change us desireable but may be inevitable in light of other constraints.

  79. @Andrew M
    Re: double decker trains
    The Paris RER uses 4 platform stations to allow for longer dwell times with 22tph. Trains enter busier stations alternately on each side of a single Island for each direction.

  80. Re: double decker trains

    Of course you could always double-deck the platform too.
    Although this only solves the dwell time issue, not the space one, since staircases would still be needed for emergencies..

  81. Regarding a comment by Paul: “The desire for a uniform fleet of new trains is for [inner] suburban services, replacing the 455s/456s/458s, with the 450s being retained for medium distance …”

    This doesn’t seem quite right. The 455/456 units are to be replaced by the new uniform Bombardier train fleet. But the 455/456s do some quite long journeys that are certainly not “inner suburban”. For example from London Road Guildford, or Clandon, journeys that are initially rural and take about an hour to Waterloo.

  82. @Peter
    I’m not sure I understand the argument here. The services you mention may start at Guildford but they are all stations stopping services which serve the inner suburban area and need to be suitable for that purpose.

    As I said in my original comment: “Matching the internal layout and number of seats/standing space to the range of peak and off-peak journey types made on each stock type has to involve a number of difficult compromises, which are never going to please everyone. ”

    The new Class 701s will nevertheless represent a significant improvement for longer distance passengers on these trains as, unlike the 455/456 fleet they will come with toilets, wifi and power points, along with walk through carriage ends and 2+2 seating. I think they sound quite appropriate for the range of routes and journey times due to be covered.

  83. Anecdata from yesterday. I travelled from Vauhall to Kew Bridge & timed the interval from door-closing to start of movement. I noted that the interval from stopping to door-opening was longer than on the overground, but that was all.
    However, the shortest timed interval was about 6 seconds, the longest over 18 seconds & most were at the 9-11 seconds mark.
    Add one of those for each stop around, say either the Hounslow or Kingston Loops, plus the shorter delay at opening, and you have 23*( 5+10 ) seconds = 5.75 minutes added ( unnecessary? ) delay for each individual train journey,

  84. An interesting and well researched article. As a long term inner surburban SWR traveller the current service is as bad as I’ve ever known it. There are multiple reasons for this, most of which are covered above far more informatively than I could hope to achieve.

    However one that I haven’t seen mentioned is the removal of the door interlock control from the guard. I am no expert but it seemed the additional delay caused by driver having to berth the train and then release the doors to control of the guard resulted in a spike in dwell time. Prior to that the guard had control of their local door which could be opened on approach to the station the door release process could begin as soon as (or on occasion just before) the train came to a halt and closed during departure, so that they could actually see any issues along the length of the train.

    I can think of at least two reasons why this practice was removed, possibly a precursor to an earlier attempt at DOO which was abandoned or more likely (and not entirely unreasonably) for Health and Safety reasons.

    I’m not exactly sure when this change was made, no doubt there will be more qualified people on here who will. However I make similar regular inner suburban journeys on Southern and I remember this as the point where their process which up until then had seemed to take forever suddenly seemed to be a lot quicker than SWR/T. Once that change had been made, in effect removing control of the doors from the guard, their days were numbered.

    Thanks for the article.

  85. Slightly ironic that while dwell times will be improved by the new 701s and working practices, at the same time they are reintroducing the 442s on Pompey fasts which will INCREASE dwell times due to their narrow doors and high step…

  86. Regarding my earlier comment and Pauls’s response “I’m not sure I understand the argument here. The services you mention may start at Guildford but they are all stations stopping services which serve the inner suburban area and need to be suitable for that purpose”

    The reason for my earlier comment was simply to point out the inaccuracy of describing SWR’s new stock as “inner” suburban trains.

    The peak hour trains from Guildford do not do a good job of serving the inner suburban area – they are already overcrowded when they reach it. Meanwhile, their slowness is a frustration for passengers from the outer suburban area.

    I agree that as far as can be judged at the moment, the design of the new SWR stock actually seems to be a compromise between inner- and outer- suburban needs (it thus perpetuates the present unsatisfactory services). The inclusion of toilets is indeed welcome for the out-of-London routes (if, that is, the toilets don’t get vandalised and locked out of use). For inner suburban trains though, toilets are both an unnecessary expense and a waste of space.

    It may be a pity that the new stock could not have been split into inner and outer suburban fleets, with designs tweaked for the needs of each. This would however have required a total recasting of the timetable. So the real issue is likely to be the compromises forced on the timetabling of SWR services by the insufficient track capacity in the inner suburban area. Looking at the Guildford via Cobham service for example, this was notably speedier in Southern Railway and BR days when many trains ran fast to Waterloo from Surbiton. That has not been possible for many years (apart from a token handlful of trains in the peaks), following the growth of long-distance commuting which has eaten up the capacity of the fast lines.
    .

  87. Surely, apart from the doors’ issue, the task of the guard on the mainline and outer surburban services is to check passengers’ tickets. I have made innumerable journeys on the Waterloo – Reading services, and from memory, have had ONE on-train ticket check. It would be no great hardship on the least heavily loaded section – Ascot to Reading – for the guard to go down the train between stations checking tickets and making him/herself known to travellers. But then, in the round, ticket inspection since privatisation of BR has largely collapsed.

  88. Unless things have changed dramatically over the years, union strike pay was generally a pittance. It’s certainly not a replacement for any pay lost when on strike, and usually only paid after x days of being on strike. As far as a person going on strike was concerned, they would be losing all the money they would have earned for the duration of the strike. That’s manageable if you only lose a day per month, but would be a severe hardship for many with mortgages etc. if the strike was more drawn out.
    According to the RMT site, strike pay (Allowance payable) is £1 a day, maximum £5 per week, and only then at the discretion of the NEC.

    I was surprised that the RMT would call for a one month strike, and even more surprised that so many members supported it. If the strikers were being supported by other unions, as has been mentioned in the comments, then that helps, but I can’t see that the extra money would amount to much.

  89. @Peter
    OK I understand that your frustration is with the service pattern rather than the stock.

    I referred to these services as “[inner]” suburban, not because they only serve inner suburbs but because they’re the innermost services on that route, running the whole distance on the slow lines and stopping at all stations. I did include the square brackets as an acknowledgement that this might be controversial for some, and I do now regret using the word inner at all! Shall we just agree on suburban?

    I can definitely see, from a distance, the appeal for more Guildford via Effingham trains to run fast into Waterloo from Surbiton, but outside the peaks, where the constraints are probably operational, these trains aren’t in my experience especially heavily loaded through the inner area? And the journey time saving is limited – the peak time 0807 ex Guildford, which is fast from Surbiton, takes 54 mins. The daytime stopping service (via Cobham) takes an hour.

    OK I get it, adding that 6 minute saving up across multiple journeys is significant, and it is frustrating to sit for over an hour on a train that seems to crawl along and stop everywhere… but isn’t this just one of those compromises that we all have to live with because the world can’t be perfect?

  90. I see that Paul Clifton , the BBC South Transport Correspondent has news on more strikes…
    Thiis from Twitter @PaulCliftonBBC

    Guards in @RMTunion on @SW_Help have voted for more strikes over future role. Eligible voters: 851. Votes cast: 505. Vote for strike: 416. Against: 88. So fewer than half of RMT guards voted to strike, but most of those who voted chose action.
    2:40 PM – 23 Jan 2020

  91. on the 7th January Phil R

    I was once on a SWT train at Clapham Junction when upon arrival the guard opened the doors on the wrong side of the train onto the adjacent live track.
    As far as I am aware there was no mechanism to prevent opening the doors on the wrong side or even when a train was not in a platform. This has been implemented using GPS on trains such as the Elecrostars but relies on a database of which side the platform is on. I am not sure how they get over the problems of GPS position uncertainties if a station has platforms where require different door side opening depending on which platform is used. This is most common at Terminals and GPS did (does ?) not work at Victoria due to the building above. There used to be a delay opening the doors at Victoria as the GPS signal was lost and the driver had to override the system to open the doors. Electrostars are also programmed with the stopping pattern that prevents the doors being opened at stations where the train is not scheduled to stop.

    LU had a different method – a transmitter was fitted under the platform nose and a detectors was installed on the train. The doors would only open when the detector recognised the signal from the platform. In some cases where doors could be opened on both sides a transmitter is fitted under both platform nosings.

    With reference to the comments about the doors opening before the train has come to a complete stop it is interesting to note that most lifts in the world commence opening the car & landing doors before the lift has come to a complete stop and has levelled with the floor. If the lift fails to stop it it is quite possible for the landing doors to open as those on the car open (they are mechanically linked) and most of the landing doors will then remain open even though the lift car has failed to stop. The LU lift spec required an auto close mechanism to close the landing doors though I am aware of at least one occasion where this did not happen. The European Lift Spec (EN81) did not have this requirement though this may have change in the last few years.

    People don’t seem to worry about pre-opening in this case.

    Of course with slam door train pre-opening was quite common with people jumping out as the train slowed. I would think a few people were killed falling between the train and the platform. Also I believe this was the reason for some of the deaths when a train hit the buffers at Cannon Street.

  92. John M
    Electrostars are also programmed with the stopping pattern that prevents the doors being opened at stations where the train is not scheduled to stop.
    What happens if there is an emergency or other out-of-course event, that requires the doors to be opened?
    I assume there is an override procedure?

  93. @John M

    “people were killed falling between the train and the platform….I believe this was the reason for some of the deaths when a train hit the buffers at Cannon Street.”

    Neither of the two deaths at the Cannon Street buffer stop collision (8th January 1991) were caused by doors being open. Indeed, as the carriages telescoped together (the underframe of one carriage cutting through the superstructure of its neighbour) those who had already left the train before it crashed were safer than those still on board.

    The large number of standing passengers contributed to the large number of injuries (542), but at 0844 on a Tuesday morning any arrival at Cannon Street will have a large number of people standing. Even those who did have seats would be getting up as the train drew into the terminus. They still do so on sliding-door trains.

  94. @Greg T 27th Jan
    I can’t say whether the feature is common to all electrostars, but their is a method for emergency door release in circumstances which require it on 378’s. It’s surprisingly long winded, but it’s also designed to prevent the driver in a high stress situation releasing doors of carriages where passengers might be placed in even greater danger.

    Once ‘Emergency Release’ is pressed on the Train Management Control System (also known as MITRAC), the software “asks” you to select one by one which carriages you wisb tbe doors to released on. Each enabled carriage will release both sets of doors on the side on which the driver presses the door open button. An example might be if a passcomm has been pulled as a train departs the platform and the first set of doors is in a tunnel. The driver will enable emergency release of all doors except for those in the first carriage. In a really serious emergency it is envisaged that passengers will pull egress handles. In the event of controlled evacuations this can be the most effective method for the driver to use to as they can control the evacuation of passengers by pulling an egress on one set of doors and then directly supervise the evacuation.

  95. Two points from me:

    1. The issue with having guards is also a timetabling/rostering issue. A DOO train only requires one member of staff to depart. A non-DOO train requires two. Whenever there is perturbation, it is always easier to find one replacement person rather than two. I also fully agree, that staff other than a driver should be focused on revenue protection and ‘keeping the peace’ (for want of a better phrase) amongst passengers. To this end, I much prefer the approach taken by the TfL Rail concession – platform staff at all stations and DOO trains with groups of roving ticket inspectors regularly patrolling trains outside the peak.

    2. The issue of wheelchair accessibility on the GB railway is – to my mind – solved very badly. Compare and contrast the humble London Bus and a sleek new Class 345 Crossrail train. The bus has had to be low floor for two decades now, and is equipped with an electrically-operated ramp which opens and retracts in around 10 seconds. A hand crank is provided in case the motor fails, but in my 13 years of living in London I have yet to witness a ramp malfunctioning when it was needed. And in pretty much all of the cases I have seen, getting a wheelchair user on or off the bus doesn’t even require the driver to leave their cab.

    Many other railway undertakings outside of the UK have provided similar equipment on their trains for donkey’s years. As an example, the Bombardier-built double-deck coaches in use throughout Germany had electric ramps since the introduction of the 4th generation model in 1997.

    In the UK – and especially across the South East – we instead have hand-operated ramps at stations (different models for different trains!) and wheelchair users have to rely on outside help to board the trains that sometimes does come, and sometimes doesn’t. The process of getting wheelchair users on or off the train is laborious, too! Picking up the ramp, walking over to the correct door, letting people get off, installing the ramp, checking it’s in place, getting the wheelchair user on or off, dismantling the ramp. Phew!

    Here’s a thought, then: why can’t we equip trains with electric ramps in the UK? These could even be operated by the public (additional bodyside button like on buses). Even on the most complicated EMUs trailer coaches don’t tend to have so much equipment under the carriage so as to preclude adding ramps underneath? And seeing as we are moving towards DOO, why can’t we ask drivers to get out of their cabs and assist those wheelchair users who have trouble getting on or off (even if it means paying them a grand a year extra)? With a trend towards longer-formation EMUs (8, 10 or 12 coaches long) it should be possible to design compartments for disabled passengers in – say – the leading or second vehicles to shorten the time required for the driver to walk to provide assistance?

  96. A big difficulty with a driver assisting a wheelchair user is the time required to close down and lock the cab, go to the appropriate place, and the reverse procedures afterwards. It plays havoc with the schedules. The cost of such delays is sure to far exceed any extra money paid to the drivers.

  97. @Malcolm: In my 13 years of living in London I must have seen a bus driver get out to assist a wheelchair passenger once. Out of however many tens or hundreds of occasions I’ve been on a bus with a wheelchair passenger. And bear in mind bus ramps are exposed to far more mud, salt and various other muck than train ramps would be.

    Also, what are the chances the ramp is going to fail precisely on a DOO train without a second member of staff present AND at an unstaffed station?

    PS – could the webmasters please rethink the layout of the site? Right now I have to open the article, scroll right down to the end of the comments, read them, then find the middle of the layout where the comment box is and type. Unless you deliberately want to put people off from commenting, could you please put the comment box at the end of the comments section?

  98. P.S. – regarding closing down cabs before getting out: surely in the age of software-driven trains simplifying the procedure of securing the cab so that the driver can leave it for a short period should just be a question of a software update?

  99. @Straphan – There are two problems with automatic ramps for wheelchair users. Firstly, it would be fairly mechanically complex as it needs to serve both sides of a carriage, or you would need two separate ramps which also increases the complexity. It would also need to cope with a range of platform heights and distances. This all adds additional kit to the train, which the operators will not want to buy or maintain.

    There is a bigger issue though – how do you ensure that automatic ramps don’t hit anyone on the platform waiting for the train? At busy time, people will crowd around the door waiting for it to open, so there is nowhere for a ramp to go.

    A better approach would be a closer match of the door height to the platforms, with a simple extending step to cater for the gap. This can be done by lowering the floor of the unit (either for the whole unit or part of it) and raising the height of the platforms. The new Merseyrail class 777 units will have level boarding using lower floors, although some Merseyrail stations have had to have the platforms adjusted a bit to ensure they are all the same.

    This just needs a bit of joined up thinking across the railway, and will help a lot of people – many accidents happen to able-bodied people stepping down from a train. The problem now is that after the recent splurge of new units, there is likely to be a lean period with very few new trains for the next few years, so it looks like the manual ramp it here to stay for a while.

  100. @Jimbo:

    Yet again it would seem that problems that exist on the railway don’t seem to exist anywhere else – i.e. on the buses. Ramps on buses fold out with a loud buzzer and do so slowly enough for people to step away. They also work with a range of distances to/from the kerb, as well as different kerb heights.

    Adjusting platform heights is all well and good, until you come to a railway line which is shared with freight trains – or platforms on curves for that matter (hello, Clapham Junction!).

  101. In East Anglia we now have platform level boarding with the Stadler class 745 and 755 trains.Bliss for wheel chair users as detachable ramps for each class of train not needed for access.Even for able bodied people no steps up or down to get into the trains.This is because of large double width doors in the centre of each carriage at platform level and retractable ramps to bridge the gap.Level boarding like this also helps reduce dwell times.

Comments are closed.