Operating Margins

TOK - eat your heart out: you have competition! ;D ;D ;D ;D

I'd be quite interested to know the Jubilee's operating margin, as it is pretty tight!

As a passenger I've noticed a huge improvement in Piccadilly reliability (particularly in the East to West direction) since the tph was reduced a few years ago.

The main question you ask stephenk, is quite difficult to answer in a straighforward, one answer fits all reply.

All lines and all parts of each line are different.

Aldgate as you mention has a 25 second (it's actually 26 seconds ) window. But even this is only during the peak, and is not the consistent window. Sometimes the window grows to in excess of 2 minutes.

The Jubilee varies from station to station. Some areas only require a 30 second additional dwell time to cause the train behind to be delayed (if only by a few seconds) mainly as a result of the lack of signals. The section between West Hampstead and Swiss Cottage is a prime example of this. A train cannot leave West Hampstead until the train ahead has left Finchley Road.

However, on a different part of the line, running the same service, the same additional dwell time would not affect the following train.

Dramatic as some of these windows can seem, remember that the timetable is written around them, and whilst delays can happen in the areas mentioned, recovery is given at other less strained locations.

Difficult to shed any extra light on this in general. You appear to have calculated many areas anyway (us railway staff tend to leave it up to the well-paid management people to do the working out ;D )

Strictly speaking, Aldgate East is a District Line station, so speak to their management about the service that station gets. I am paid by the H&C and worry only about the windows on that line ;D ;D

Harsig once posted a diagram of Finchley Road (MD) prior to the 1984 resignalling and creation of Baker Street SCC.

www.geocities.com/idsmesig/Signals/Finchley_Road_MD.JPG

A modern high-tph layout might look like this:

216.55.161.203/theonekea/underground/diagrams/Finchley_Road_MD.txt

True - it's designed to show what sort of layout _could_ have been built. I do realize that the Jub will get moving block signalling and that fixed signal aspects would become unnecessary, but just the same, the above diagram shows what could have been...

I would expect that moving block would allow for station run out run in times of approx 50-55secs (based on observations on other moving block systems and adjusting the time to take into account for train length). This is about 20-30 seconds faster than the present time that it takes at Finchley Rd. So I would expect that moving block would increase the throughput substantially.

The train seperation distance on moving block is worst case braking distance, plus positioning error distance. The worst case braking distance is based on the braking distance with 75% of braking power (although this is user selectable by the metro company), taking into account track grades. The positioning error probably depends upon the system used, but I would guess would be about 25m on the system which the Jubilee is getting (as that is the distance between IL loops). On most existing moving block systems, the trains will keep a 50m safety distance distance from a stationary object in front (points, another train), however the train will creep slowly up to this 50m mark as can be observed often on the DLR.

I wouldn't quite go for 50tph, not even Moscow can manage that! In my opinion, the maximum realistic tph in London with moving block would be about 33-34tph, which happens to be what the infracos are aiming for after the various line upgrades. This is also the frequency that used to be run with speed signal/multiple home arrangements. However with moving moving block and high performance trains, this frequency can be achieved with faster run times - conventional signalling prefers 20-25mph station run in speeds, moving block prefers 30-35mph station run in speeds for optimum throughput. (As a side point station run out run in times, and journey times would be more consistant). This, combined with faster acceleration and decceleration (the latter being the most important in throughput), and possibly higher line speeds, allow for higher capacity and faster run times with the same number of trains.

As for running too many trains, the frequency should match the required capacity. It would be overkill to run 34tph on the Bakerloo, yet 34tph still wouldn't be enough on the Victoria! Likewise, 34tph may be suitable for the peak periods, but off peak there is no need to increase capacity from the present. So the main benefits of moving block in terms of capacity would only be useful for a few hours each day. The benefits of moving block in terms of equipment reliability may benefit the line throughout the whole day, however this is still open to question!

Concerning line reliability, I would still like to see decent operating margins rather than trying to get as many trains as possible down the line. This is why I think 33-34tph would be the upper limit, rather than the theoretically possible 40tph. As we can see from the Piccadilly Line, the higher the frequency is not always better! Where there are large operating margins, care needs to be taken to ensure headways run evenly, as opposed to trains bunching. Lines running with minimal operating margins quite often regulated themselves!

Concerning Baker Street, yes it does have a high capacity. Moving block would maybe only decrease run out run time by about 10secs. However, the rest of the line cannot presently match it's capacity, and thus should have it's capacity increased to match that of Baker Street.

The only way 34tph would be useful on the Bakerloo is if it were extended south, somewhere towards the central area of the Southern Region suburban services. That oft-mentioned extension to Camberwell could be expanded into an extension to somewhere like Herne Hill.

I have heard that 34tph is the absolute limit, i.e. there is absolutely no way to stuff any more trains down the northern half of the Circle once you have reached 34tph, because of the signalling currently in place. Is this true?



But surely the route north of Baker Street is equally bad, right? From personal observation there aren't many signals on the Met slow lines north of West Hampstead at all...

Oh goody.



It seems to me then that improving the signalling to theoretically allow more than 34tph, without actually running more than 34tph, would be the best option - the existing peak service would be more 'stable' because of the latent unused capacity...



True. Still though, when the job is up the wall has it ever hampered recovery?

I had heard that the Jubilee upgrade was only going to result in 31tph, but hey lets not worry about 2tph (or approx 5secs!).

The original plan for the Jubilee (when the JLE was planned) was 36tph. Personally I think that was a bit too optimistic!

I don't think the Jubilee needs 24tph off-peak at present, I reckon approx 18tph would suffice. But maybe with continued growth of Canary Wharf?