Serco for Sleeper

[Bletchleyite]

Wasn't the reason given by FGW for withdrawl that 85% of these never used ?

Quite possibly. Removing the razor isn't a bad idea as most people will prefer to use their own than rip their face open on a cheap disposable single-blade one.

 

[modernrail]

I catch the sleeper from time to time and am looking forward to the new stock arriving. I also love it when there are engineering works on the WCML as it means a smoother, straighter trip up the East Coast. Some people on here have rightly pointed out the problem of the sleepers taking up far too much space at Euston at what can be a very busy time for domestic WCML trains, especially in the morning. This got me thinking. There are some very long platforms just down the road at St Pancras international. Would they be long enough for the sleepers and if so, might it be a possibility to run them from there, and ideally then up the East Coast rather than the West Coast?

There are very few Eurostar departures at the time of night the sleeper leaves (none at the time the highland sleeper departs). There appear to be about 7 Eurostar departures in the morning in the time for the maximum time the sleeper would need to sit in the station. There are 6 platforms. Bearing in mind some of the arrivals form the departures within this period, and the 7 departures are spread over three hours, perhaps there is a load of capacity sitting there unused? Would also be a way of connecting Paris/Brussels to Scotland with easier connections. You would be subjecting sleeper passengers to an extra layer of security unless you could form an extra entrance but maybe that is no bad thing and you would also be leaving from a very nice station rather than the dump down the road. If viable - I suspect HS1 will argue they want the extra capacity incase they attract other services through the tunnel (such as DB) but still interested to see if we think this is feasible.

 

[najaB]

This got me thinking. There are some very long platforms just down the road at St Pancras international. Would they be long enough for the sleepers and if so, might it be a possibility to run them from there, and ideally then up the East Coast rather than the West Coast?

Silly question, is it easy (or possible) to get from the ECML to the Eurostar platfoms at St. Pancras?

 

[jopsuk]

Whilst there's a rail connection, there's other problems. The security thing would be a big deal for one. Possibly more important is that the International platforms are built to UIC TSI High Speed specifications- considerably lower than UK domestic modern standards. For the Olympic games Javelin shuttle services, when the Class 395s were using the International platforms at Stratford, those platforms were modified for accessibility.

 

[po8crg]

Whilst there's a rail connection, there's other problems. The security thing would be a big deal for one. Possibly more important is that the International platforms are built to UIC TSI High Speed specifications- considerably lower than UK domestic modern standards. For the Olympic games Javelin shuttle services, when the Class 395s were using the International platforms at Stratford, those platforms were modified for accessibility.

If this had been planned and agreed two years ago, the new rolling stock that CAF are building could have had external steps to access the TSI platforms. Disabled access would be with a platform lift, as it is for Eurostar. I suspect a modification now would be very expensive, however.

As for security, arrivals shouldn't be too bad an issue, and I suspect that most sleeper passengers arrive early enough for a 30 minute security check-in to be relatively painless. They'd have fun in August, though, with the shotguns.

 

[NotATrainspott]

I catch the sleeper from time to time and am looking forward to the new stock arriving. I also love it when there are engineering works on the WCML as it means a smoother, straighter trip up the East Coast. Some people on here have rightly pointed out the problem of the sleepers taking up far too much space at Euston at what can be a very busy time for domestic WCML trains, especially in the morning. This got me thinking. There are some very long platforms just down the road at St Pancras international. Would they be long enough for the sleepers and if so, might it be a possibility to run them from there, and ideally then up the East Coast rather than the West Coast?

There are very few Eurostar departures at the time of night the sleeper leaves (none at the time the highland sleeper departs). There appear to be about 7 Eurostar departures in the morning in the time for the maximum time the sleeper would need to sit in the station. There are 6 platforms. Bearing in mind some of the arrivals form the departures within this period, and the 7 departures are spread over three hours, perhaps there is a load of capacity sitting there unused? Would also be a way of connecting Paris/Brussels to Scotland with easier connections. You would be subjecting sleeper passengers to an extra layer of security unless you could form an extra entrance but maybe that is no bad thing and you would also be leaving from a very nice station rather than the dump down the road. If viable - I suspect HS1 will argue they want the extra capacity incase they attract other services through the tunnel (such as DB) but still interested to see if we think this is feasible.

Yes, the Eurostar platforms are long enough, but the various stakeholders involved (Transport Scotland writing the new Sleeper ITT, HS2 Ltd, Network Rail, HS1 Ltd, Eurostar... ) seemingly didn't think it would be worth their while. If HS2 Ltd had pointed out that it would be unreasonably expensive to keep having the Sleepers at Euston while the station is rebuilt then I think we would have seen the St Pancras option investigated more, as it really would be possible with only minor changes to how St Pancras is set up in a non-railway sense.

 

[najaB]

As for security, arrivals shouldn't be too bad an issue, and I suspect that most sleeper passengers arrive early enough for a 30 minute security check-in to be relatively painless.

Currently they don't - most rock up about 15 minutes before departure - but I suspect that is because they can. If they had to arrive earlier, then they would.

Since it's the international platforms we're speaking about, there's also the (slight) risk that arriving sleeper passengers in the morning could find themselves a way onto the early departures to Paris.

Thinking about it some more, would that mean that security checks would have to be done at Edinburgh, Glasgow, etc. since they would be arriving into the 'secure' platform side of the station?

 

[NotATrainspott]

Currently they don't - most rock up about 15 minutes before departure - but I suspect that is because they can. If they had to arrive earlier, then they would.

Since it's the international platforms we're speaking about, there's also the (slight) risk that arriving sleeper passengers in the morning could find themselves a way onto the early departures to Paris.

Thinking about it some more, would that mean that security checks would have to be done at Edinburgh, Glasgow, etc. since they would be arriving into the 'secure' platform side of the station?

The more sensible solution is that one of the Eurostar platform islands would be semi-separated from the rest of them, in the same way that the Eurostar platforms at Lille can be separated off. When the sleepers are there, they would be open, with access from the concourse level rather than from the secure undercroft, so that no security is necessary. When they've left, the platforms would be checked and then returned to the secure zone. That arrangement might sound convoluted but was designed into Curzon Street, Birmingham Interchange and Old Oak Common for the sake of the HS1-HS2 passengers.

 

[Airline Man]

Arriving passengers into St Pancras would either have to be security cleared when they boarded the sleeper or physically separated from departing security cleared passengers. You can't have cleared passengers mixing with non cleared.

 

[button_boxer]

Possibly more important is that the International platforms are built to UIC TSI High Speed specifications- considerably lower than UK domestic modern standards.

And don't the international platforms use a French signalling system rather than the NR standard TPWS, so the locos would need extra kit and the drivers extra training.

 

[po8crg]

It's one of those "if they'd planned it before the new trains were ordered" exercises, isn't it?

It could be done, for a reasonable amount of money, if it had all be planned five years ago:

A small adaptation of St. Pancras station to allow two platforms to be isolated from the rest (so the sleepers can exist outside the secured zone and the platforms revert to secured at the appropriate time).
CAF specified to include retractable steps on the rolling stock so they can access TSI platforms.
KVB signalling installed on locos hauling trains into and out of St Pancras, or AWS/TPWS on the track (around Calais, where the same decision had to be made, they chose to dual-signal track with AWS/TPWS).

But Serco have ordered the rolling stock now, and changing the specification would be expensive.
The Class 92s have TVM signalling installed, so KVB shouldn't be a huge issue, if there's time to design and test is properly. Adding AWS/TPWS into St Pancras would be do-able if there's time to plan it carefully if that's cheaper/easier than changing the locos and retraining drivers.
Reconfiguring St Pancras security / platform isolation would be technically possible, but would be the sort of thing that you'd need to plan years in advance.

If someone had got behind this four or five years ago, then it could have been done in time for the arrival of the new CAF LHCS stock, and - given the relief it would offer to Euston - would probably be a good idea (not needing platforms longer than 280m would make Euston reconstruction much easier).

But it would be an expensive pain in the backside to implement now: Serco and CAF would have to be compensated for the design change and for any delivery delay that was caused, changes to St Pancras would have to be designed and implemented in a hurry, as would the signalling (whether on the loco or in the station), all of which is asking for trouble.

 

[sprinterguy]

In an article published at the start of the month (but I don't think it has been posted here), it states that the sleepers' PPM scores had dropped to 88.8% in April, the first month of operation by Serco, and 91.7% for May, compared to 98.2% from a year previously under First Scotrail (assuming that the latter figure was pertaining solely to the sleeper services and not to Scotrail services as a whole?).

Caledonian Sleepers' Performance and Safety Manager states that "it can only take a few small incidents to quickly knock trains beyond the PPM targets", but it seems to me that many of the incidents that have occurred to the sleeper since Serco took over can hardly be regarded as small, particularly from the passengers' perspective when they wake up at Crewe having left London the previous night, or Slochd summit the day after leaving Inverness.

The article does also give some interesting insight into the issues that were affecting the class 92s at the start of the franchise:

http://www.railtechnologymagazine.c...performance-figures-and-the-role-of-new-locos
01.07.15
Caledonian sleeper: Poor performance figures and the role of new loco's
Source: RTM Jun/Jul 15

Since becoming a separate franchise, the punctuality and performance for Caledonian Sleeper have dipped substantially. But David Simpson, performance & safety director at Serco Caledonian Sleeper, told RTM’s Sam McCaffrey the figures aren’t what they seem.

The latest PPM figure for the Serco-run Caledonian Sleeper franchise, for Period 2 (3 May to 30 May 2015), is 91.7%, which is significantly lower than the 98.2% posted a year ago when the services were a part of ScotRail. However, it is an improvement on the 88.8% from April.

According to performance and safety director David Simpson, it’s usual for the numbers to fluctuate from period to period, especially as the franchise operates so few services.

“We measure seven trains a night which is 168 trains a period on a PPM target of 92%, which rises to 92.5% at the end of this control period. It doesn’t take many trains to be late in a period to miss that target,” he said.

“With all the small numbers it can only take a few small incidents to quickly knock trains beyond the PPM targets, and I think that’s what happened in periods 1 and 2 since we took over from ScotRail.”

One of the issues that can cause Caledonian Sleeper’s PPM figures to plummet is a delay to one of their services that splits into multiple trains serving different destinations.

Of the trains that run to and from London each night, one splits into two parts to serve Glasgow and Edinburgh, and another splits into three to serve Fort William, Inverness and Aberdeen.

“So if there’s a delay on the London to Edinburgh section on the train that splits into three parts that means the train can be late at all three destinations – which is three of our quota of 15 late trains per period just from one incident,” Simpson said.

Last period, 15 Sleeper trains missed their PPM targets, with the causes evenly split between Network Rail issues, delays caused by other operators and Serco’s own problems.

Of the operator’s own problems, the one contributing most to missed PPM targets is an issue with the new Class 92 locomotives, which haul the Mark 2 and Mark 3 British Rail coaches.

When it took over, Serco elected to move on from DB Schenker, which had been hauling the Sleeper coaches for more than a decade, and awarded the contract to GB Railfreight instead. This also meant a switch from Class 90s to GB Railfreight’s Class 92s and – from October on the non-electrified sections – a switch from Class 67s to Class 73/9s.

Under the new arrangement, refurbished Class 92 dual-voltage electric locomotives haul services south of Edinburgh and Glasgow, and failures with these have contributed to 20% of Caledonian Sleeper’s total missed PPM targets.

The problem with the locos is caused by two pieces of safety equipment meant to detect stray electrical currents and protect against their impact.

Simpson said: “The limits [for] those two pieces of kit are set quite tight, for good reason, and what we’ve found is that when the locomotives travel through neutral sections in the OLE [overhead line equipment] some of that protection has been tripping out, because it has recognised interface issues between the OLE and the loco. The procedure there is straightforward – the driver can reset the kit and travel on as normal – but that can cause delays and has and caused PPM miss at destination.”

The Class 92s were built for use through the Channel Tunnel and Serco believes that the levels at which this equipment is set is higher than is necessary for use on the main line network. The company is working with GB Railfreight at relaxing the level at which the protection is set.

“Now that has to go through Network Rail to be validated, which is absolutely right and proper, but there is a strong case for that and it should give us some additional resilience as we go forward,” Simpson said.

He added that so far in period 3, as of 17 June, they have had no PPM failures due to the issue so it appears it may be resolved. "I’m confident that the 92s’ reliability has been growing. We do still have a Class 90 from DB Schenker as a backup that we use from time to time, but overall the Class 92 performance has been growing to where it needs to be to hit these demanding punctuality targets.”

 

[al.currie93]

The article does also give some interesting insight into the issues that were affecting the class 92s at the start of the franchise:

Interesting read - thanks