In this part of our ongoing series we look at a critical set of points north of Purley station and what role they might play in a future Brighton Main Line. It is also a convenient time to look at the issue of trying to keep the line open between South Croydon Junction and Coulsdon, with the Christmas engineering works showing what happens when this is not possible.
Switching between the fast and slow at Purley
There is a set of points north of Purley station that look complicated when viewed on Carto Metro or the Open Train Times diagram, but in essence they can be split into two functions. One group is to do with the two slow lines north of the station serving four platforms (3-6) at Purley and accessing the aggregates yard and siding. These are in continual use. The other group is simply to enable up trains to move from the up slow to the up fast and down trains to move from the down fast to the down slow. These were once important but not vital and are less used now – but may have more of a role in future.
The latter-described points at Purley weren’t going to be part of our story. However, potential future issues at Windmill Bridge Junction with southbound (down) trains from Victoria make us wonder whether these points might actually become more important in the years to come. They are about to be renewed, so it is clear that they must play some role. The question is what.
The points to the north of Purley station have a particularly unfortunate history. Logically, from around 1983-84 onwards when the arrangement of fast and slow lines through Purley was changed, it would have made a lot of sense for trains going to London that called at Purley before becoming fast from Purley to East Croydon to use the fast lines.
Purley Train Crash
At the time of the Purley rail crash it was normal for trains from Horsham, having called at platform 3, to be switched from the slow line to the fast – at least in the off-peak. If all trains ran to time then fast trains would not be delayed. If they were only a little late then the following fast train might have to slow down and have restrictive signals (yellows), but would probably not have to come to a complete stand. Only if the train from Horsham was slightly delayed and the following fast train running slightly early would the fast train need to come to a complete stop.
On Saturday 4th March 1989, this situation occurred and the critical protecting signal, signal 168, at the northern end of platform 1 was set to red to tell the approaching fast train to stop. Unfortunately, the driver was unprepared for the red signal. In subsequent evidence to the inspecting officer he was adamant that he saw a double yellow, as he was expecting to, at the signal before signal 168. It was clear he was either mistaken or he lied.
The inevitable result was that the fast train ran into the back of the train from Horsham. Five people lost their lives and 88 were injured. Around that time the occasional train crash with a small loss of life was regarded as normal and unavoidable, but a combination of events meant that the Purley train crash finally called into question the ability of British Rail to run trains safely.
A succession of railway accidents
The crash occurred less than three months after the Clapham rail crash, by far the worst railway accident for three decades. Things got worse less than 48 hours later when there were further fatalities at Bellgrove in Glasgow – also due to a driver passing a signal at danger (albeit in completely different circumstances). A fear gripped the British Railways Board which urged the drivers to drive more “defensively” which, of course, had a consequential effect on reliability. Many also argued that this approach was actually counter-productive – because out-of-course working increased the likelihood of incidents and because drivers would be encountering yellow signals more often, which led to the known danger of subconscious cancellation of the Automatic Warning System (AWS) device.
The driver at Purley pleaded guilty to charges of manslaughter and endangering life. This was a highly unusual as juries are usually very reluctant to convict in “there, but for the grace of god, go I” type situations. The accident report also put the blame on the driver, but it did also recommend a banner repeater for signal 168 to give the driver further advanced warning. The report did note that there had been four previous SPADs (signal passed at danger) at this location in the previous five years, but clearly the inspecting officer thought his recommendation of a banner repeater signal would be sufficient in preventing SPADs at this location in future.
Because of the court proceedings, publication of the report was delayed in case it could adversely influence the jury. As a consequence, the fact that the location was a SPAD hotspot only came to light after sentencing. Although the driver had already served his prison sentence he decided to appeal on the grounds that had the information been available at the time he would not have accepted the advice to plead guilty – indeed the advice itself may even have been different. Lord Justice Latham agreed and whilst he considered that a jury may still have found the driver negligent, they may well have decided he had not been negligent to an extent that justified the imposition of criminal sanctions.
All this may initially seem irrelevant, but one of the truths of modern railway operations is that if you get SPADs at a signal, especially ones with potentially serious consequences, you do something about it. A subsequent SPAD at the same location two years later – despite the addition of the banner repeater – clearly meant that more had to be done.
The solution was a procedure normally only used for royal trains – “Grove Arrangements” – whereby trains are stopped at the signal prior to the one normally used for protecting junctions or preceding trains. This solved the SPAD problem, but it naturally meant that using the crossover greatly increased the chance of delaying a following train. This, perhaps combined with the knowledge that the points were worn out, meant that the crossover only saw limited use. Typically this was for trains in the early hours when engineering work forced them to switch tracks rather than call at platform 1 or 2 at Purley. Platform 1 has its own separate entrance which is normally locked. It also probably has the least used lift on the rail network. Platform 2 at Purley is now protected by an anti-suicide fence dividing platforms 2 and 3, so it is also inconvenient to stop trains there when it can be avoided as it involves opening up the access gates.
TPWS to the rescue
The accident report acknowledged the fact that the British Railways Board had already agreed to proceed with the development and installation of “an ATP system” (Automatic Train Protection). For many years now we have had this in the form of TPWS (Train Protection and Warning System). TPWS renders the post-accident precautions at Purley redundant – but it took a long time, some cost, and (no doubt) a thorough safety case report to remove the additional restricting protection.
The future use for the crossovers at Purley
As indicated in our articles on East Croydon, the proposed arrangements for the reorganisation of the junction at Windmill Bridge will considerably reduce, if not eliminate, conflict in a northbound direction, but could actually make things worse southbound for the few peak hour fast trains from Victoria. It may be that in future southbound fast services from Victoria that stop at platform 4 at Purley will be routed on fast lines southbound between Windmill Bridge and Purley. This is, in fact, what used to happen until a few years ago. It is also possible, but less likely, that fast trains to Purley from Victoria that continue to Caterham or Tattenham Corner could, in the next decade, use the fast line from Windmill Bridge to just outside Purley and then call at platform 5 or 6.
Depending on the final arrangements for Windmill Bridge Junction, a move to route the few southbound fast trains from Victoria that stop at Purley could make sense in future. Existing restrictions (such as limited fast platforms at East Croydon) will have been resolved and the move can be made without disrupting northbound trains on the up slow line too much as they would either be slowing down to call at Purley or waiting in a platform anyway. The worst that would happen would be a short delay to the departure of the northbound train.
If the southbound trains cross from fast to slow then, if it fits in with the timetable and the service can be run sufficiently punctually, it would make some sense for the relevant northbound trains to cross from slow to fast at the same time.
Problems in keeping the railway open
South of South Croydon Junction the Brighton Main Line ‘only’ has four tracks. North of Windmill Bridge there are four tracks to Victoria and four to London Bridge. Through East Croydon there are six tracks and from there to South Croydon there are five. With five tracks, even if you need to be working on two of them and need two more closed for the safety of workers, you should usually be able to keep one line open. That’s enough to allow services to continue throughout the night whilst work is taking place. Once you get down to four though then things start to become problematic. Going further south, beyond Stoat’s Nest Junction just south of Purley, the route divides into two before rejoining south of Redhill. At Purley the Caterham and Tattenham Corner branches leave the main line, so you could argue that South Croydon Junction to Purley station is the most critical part of the railway – and a section where you really, really do not want to have to do major engineering work on all lines.
On a more positive note, the line in this area has less of an issue as regards capacity now we are further from central London and some services have already diverged away from the main line. The fast lines are generally not hampered by trains calling at stations and so between East Croydon and Gatwick they can pretty much handle anything thrown at them. The slow lines reduce from three to two tracks at South Croydon Junction, but then it also has a reduced number of trains to support there. There is not a fundamental capacity issue on the slow lines between South Croydon and Purley, although a mixture of trains stopping and not stopping at Purley Oaks can lead to slight delays. This is not helped by Purley Oaks slow platforms being only eight cars long so SDO needs to be used on longer trains.
Soled and Heeled too many times
Around thirty years ago a British Rail manager commented that the problem with the Brighton Line was that it had been soled and heeled too many times. The line was changed from double track to quadruple track in the 1890s and it is probably fair to say that when it comes to the civil engineering infrastructure and permanent way it has been almost constantly patched and repaired since then. It has certainly never properly been reballasted from the base of the formation upwards – something now known to give a good 25 years of life in the track base (not the track itself) without major work being necessary, as long as it is done properly. A “scrape” is quicker, of course, but it may well only give you ten years before having to repeat the process.
Doing things properly, however, takes time. One of the causes of the problems at King’s Cross last Christmas was a commitment to do the job properly and not revert to scraping (as was in the contingency plan) when it was realised that the schedule was slipping.
The reluctance to do a job properly on the Brighton Main Line has almost certainly been at least in part due to a considerable reluctance to close it for any length of time – understandable when you serve Gatwick Airport. In recent years there has been a change in attitude as the acceptance grows that with a 7-day railway there is simply no real quiet time during the week to do maintenance. With the recognition that any substantial engineering works is going to be very disruptive, support has been moving towards big blockades at quieter times of the year which will at least sort out the job properly.
The Christmas 2013 big blockade
Between Christmas 2013 and New Year 2014 Network Rail put the idea of a big blockade into practice when renewing a total of 16 points at Stoat’s Nest Junction just south of Purley. Despite absolutely appalling weather (remember what happened at Dawlish a couple of months later) the work was completed on time although it very nearly got called off at the last moment. The reasons for considering postponement have never been given, but the need for workers to fix problems elsewhere, the closure of a diversion route due to a landslip and the fear of not being able to complete the works on time may have all been factors.
It was hard to fault the rail replacement service arrangements made in 2013. These worked extremely well despite being challenged by the closure of the line from Horsham to Dorking due to a landslip – a limited train service from London to Gatwick via this roundabout route had initially been planned. A further challenge presented itself after Christmas when a planned reopening of the line from Gatwick Airport southwards to Three Bridges was delayed because some simultaneous work to bring extra platforms into use at Gatwick was delayed by the awful weather. Nevertheless, this too was successfully surmounted.
The rail replacement plan was clever, well thought out, well publicised and well executed with plenty of staff available to assist. The East Grinstead route was used as the primary route for Gatwick services and beyond with rail replacement coaches non-stop from East Grinstead to Gatwick. As an alternative there was a quarter hourly fast service to Coulsdon Town where buses could take the A23 and M23 to reach both Gatwick and Three Bridges non-stop from outside the station. All arrivals from London involved stepping off platform 2, walking a few yards and then generally straight onto a rail replacement bus with plenty of luggage space. Rail departures weren’t always quite as convenient, with half the trains departing from platform 1 and half from platform 2. There was also a replacement stopping service via Redhill from a dedicated stop in Coulsdon High St – the relevant traffic order had been sought and the parking bays closed. Everything, including the destination of replacement buses, was clearly indicated and well publicised in advance.
Christmas 2015 – New Year
This Christmas, the sets of points north of Purley station are being replaced. With the success of two years previous one might think this would be easy. Again there are 16 points due to be replaced. More inconveniently, the opportunity has been taken to replace a bridge from 30th December to 4th January (more on this later) on the Brighton Main Line south of Purley, which means that all engineering trains need to come from and depart to the north during this period. Nevertheless, from the engineering perspective, there is no reason to believe that this work will not be as successful as it was two years ago.
What’s going on?
Where things may go wrong is with the alternative arrangements. Someone unfamiliar with the area might think that it shouldn’t be too bad with similar arrangements in place as before. Unfortunately this is not the case as Purley station will not see any passenger trains.
With Purley unable to have a railway service this means that both the Tattenham Corner and Caterham Branches can’t run. Even if they could run to Purley without getting in the way of the engineering works (and assuming Southern could find the staff to run them) the bridge reconstruction would mean that from December 30th the trains would be isolated with no means of getting to or from any depot.
Worse still the plan seems to involve running replacement bus services all the way to East Croydon. Not only is this around four miles from Coulsdon Town (three miles from Purley) it is going to be along a congested route badly affected by the traffic caused by the post-Christmas sales. No attempt has been made to focus on getting Caterham passengers to Upper Warlingham on the East Grinstead line (still open) where there will be a service. The intention seems to be to bus them all the way to East Croydon.
East Croydon is not the ideal location for replacement bus services to start or finish and the set down and pick up points are not very convenient. The drop off point for passengers laden with luggage is in Cherry Orchard Road which, despite its charming name, is not really a user-friendly place to dump people who need to get to East Croydon station. At least it does separate passenger flows by having arriving passengers using the old station foot overbridge with ramps at the south end of the station and departing passengers using the new bridge which has lifts.
There are a couple of thoughtful innovations in the rail replacement plan. Buses to East Croydon will not go through a time-consuming diversion to serve Purley station – in the other direction they will stop close to the station but there is very little time penalty in doing this. Replacement buses from Purley to Tattenham Corner will continue to Epsom Station for direct trains to and from London.
Curiously, passengers to East Croydon from the Tattenham Corner branch are advised by journey planner to change at Coulsdon Town for a second bus to East Croydon but in the other direction they are advised to change at Purley. It is hard to know whether this is just an artefact of journey planner (which suggests you change at the first convenient place) or whether it has actually been thought through. It is highly likely that it would be better if Coulsdon Town was used in both directions, but with details of what the exact arrangements are rather sketchy it is hard to be sure. Then again, changing at Purley means more choice of buses with the East Croydon – Caterham buses being an alternative to the Redhill buses so who knows where the best place to change is?
Publicity or lack thereof
What is concerning is the lack of publicity for this closure and, until very recently, the total absence of any real detail – such as where the replacement buses stop – except for those at Purley and East Croydon. For the most part the stations involved have not had any publicity in the form of either posters or replacement timetables to hand out. The relevant timetables with the latest timetable change on December 13th make no mention of this work starting less than two weeks after that timetable commences. The situation is improving but, as appears to be modern practice, there is very little detailed written or online information and a total reliance appears to be place on National Rail’s Journey Planner.
As a public service announcement we can tell you what seems to be impossible to find out online and tell you that the proposed replacement bus services (according to the posters that Network Rail have placed on the fencing at their work site) will be:
- Gatwick – East Grinstead (12 buses per hour)
- Three Bridges – East Grinstead (12 buses per hour)
- East Croydon – Dorking via Purley, Coulsdon Town and Redhill (2 buses per hour)
- East Croydon – Epsom via Purley, Coulsdon Town and Tattenham Corner (2 buses per hour)
- East Croydon – Caterham via Purley (2 buses per hour)
In fact that is slightly wrong because the Epsom service will start from Purley.
So far the public have not been given any indication of which services serve which stations but it is presumed that they mimic the rail services they replace. The buses do appear on National Rail’s journey planner but do not include calling points so it is hard to work out exactly what is going on.
It is not unusual for advantage to be taken to do engineering works at other sites already affected by major works. In this case Network Rail have decided to deal with a bridge that is long overdue for replacement. This is the rail-over-road bridge at Old Lodge Lane very close to Reedham station. There is another bridge, very close by, that takes the Tattenham Corner branch over the road but this is unaffected by the works.
Because the Brighton Main Line was generally built double track and then widened to quadruple track, most bridges over roads are two (or more) separate structures. This makes replacement slightly easier. Furthermore the older one in such instances is generally masonry and the newer one steel. It is thus usually only the masonry one that needs replacing. In many cases, as here, the need to replace it is partly down to repeated bridge strikes. The steel ones, having more clearance, tend to suffer less from this.
In the case of Old Lodge Lane bridge, the construction is unusual with the original masonry bridge being extended widthways by a steel construction. This means the whole bridge needs replacing. What has been disappointing is the lack of advance publicity given before closing a fairly major road in order to facilitate this. Worse still, the 455 bus route has been diverted with absolutely no publicity on any of the many stops not currently being served.
The work involved is a big job and Network Rail must hope there are not too many more like this on such a critical section of the Brighton Line. The steel section of the bridge will require a crane to remove and there needs to be a crash deck on the highway for when the masonry arch is reduced to rubble. This means that a suitable base has to be built for the crane and the water supply has to be temporarily diverted by the local water company in case the reconstruction causes a cracking of the iron water pipe. The new bridge will also be craned in, which means continuous working from late December 30th to the early hours of January 4th.
At least, after the work is done, it should be possible for the 13′ 6” height restriction to be raised by at least 12 inches which would mean that London Buses would be able to run double-decker buses on route 455, though at this end of the route it is very quiet.
London Reconnections rolling news – almost
In a departure from our normal practice, we hope to be able to prove updates to this story should anything relevant and significant happen in the next two weeks. In 2016 we will be diverting attention away from the Brighton Main Line but still intend to continue with a handful of articles during the year in order to cover the line down to Brighton. 2016 should also be the year when we finally get beyond the London boundary with this series.
Update: Of Mice and Men
Despite earlier foreboding it was clear by Sunday the 27th December that the replacement buses between East Croydon and Redhill/Caterham and Purley – Epsom were working tolerably well if little used. In general, there were plenty of staff to assist at Purley, and, doubtless, at East Croydon and Redhill but elsewhere passengers were left to fend for themselves. By the 28th the expected delays on the road started appearing with bus schedules becoming a work of fiction.
It is clear that the best laid plans often go astray. The A23 Purley Way and the Brighton Road direct between Croydon and Purley merge at Purley. About a mile south of Purley the Coulsdon Bypass means that there is once again an alternative route – either on the main A23 or through Coulsdon town centre. South from the Coulsdon Bypass much of the road is dual carriageway before most of the traffic leaves to join the M23 at Hooley. So, probably the last thing you would want is a blocked main road midway between Purley and Coulsdon.
Southern (Gas Networks) delays passengers
It is therefore most unfortunate that Southern Gas Networks decided that they needed to do some emergency gas repairs at precisely such a location as described above. The only saving grace was that they only closed half the road and installed temporary traffic lights. Nevertheless this could mean delays of at least 10 minutes southbound and generally much longer northbound. As traffic was moving slower than walking speed it would have been possible for passengers without luggage to get off the rail replacement bus, start walking and have a good chance of catching up with the previous bus.
On occasions like this it must be very frustrating for the organisers of rail replacement buses to have an incident completely outside their control make the situation much worse than it need to have been. They are probably unaware of this but, at the junction in question, Southern Gas Networks (SGN) are often digging up the road but generally at this location they do this at night and reinstate it in time for the morning peak period. It is hard not to be cynical as to how much of an emergency this was and that it didn’t happen on Christmas Day or the day after when it would have been far less inconvenient for the travelling public if not for SGN and their employees. To put it another way, it is hard not to be suspicious and be convinced that this was a “planned emergency” and that Network Rail are not the only ones who schedule disruptive work to be carried out between Christmas and the New Year.
What, of course, makes this so much worse is that this is not factored into the planning – not that much could be done about it. It would at least be possible to warn passengers (and indeed drivers) that their journey could take even longer than planned. Rail replacement bus schedules could have been amended to be more realistic and some alternative options considered or reconsidered. As it is there is nothing to tell passengers to also allow time for delays due to this emergency work. According to the TfL traffic website listing details for the A23 the work by SGN will take from Monday 28th to Wednesday 30th. You can get also a map of the specific incident and even click on it to look at the traffic cameras in the area. Needless to say National Rail Journey Planner does not reflect this and allows just 8 minutes for a rail replacement bus along the Brighton Road from Purley to Coulsdon Town.
What has happened is unfortunate but it is almost certain to cause passengers to dread, even more than they do now, the spectre of the rail replacement bus service.