How are timings calculated?

[Future]

When reading timetables, I often wonder how they are written in such a way that they are a) accurate to each timing load and b) written so that a train can exactly follow them. How are times calculated in the first place? Is a test train run to work them out or is it a computer program?

 

[The Planner]

When reading timetables, I often wonder how they are written in such a way that they are a) accurate to each timing load and b) written so that a train can exactly follow them. How are times calculated in the first place? Is a test train run to work them out or is it a computer program?

Both, as well as cab rides, stop watching, on train data (OTMR) etc

 

[chrisdoward]

Both, as well as cab rides, stop watching, on train data (OTMR) etc

Indeed, just over 33 years ago during the first week of the May 1990 timetable I was sent out to chack on the timings for the newly electrified lines between Portsmouth,
Southampton and Eastleigh to see how they were. My instructions were to remain inconspicuous, so that there was no influence on the train crew, that was a bit awkward when I knew one of the guards. From memory, I think the schedules were ok.

In the South West Division (Southern Region) Train Planning Unit there used to be a big A3ish size book full of schedules from various routes and timing loads which if there wasn't a comparable train in a Working Timetable I referred to on a few occasions. I don't know if they still exist

 

[Horizon22]

In the South West Division (Southern Region) Train Planning Unit there used to be a big A3ish size book full of schedules from various routes and timing loads which if there wasn't a comparable train in a Working Timetable I referred to on a few occasions. I don't know if they still exist

It may well do, but I would suggest such a book is now digitised!

 

[zwk500]

It may well do, but I would suggest such a book is now digitised!

You'd just use Bplan nowadays, which is the Link-by-link breakdown of SRTs for each timing load.

 

[AF91]

Brand new SRTs can be modeled in advance using acceleration and braking curves being applied over the top of the speed limits and other geographical information. Once SRTs are in place, GPS data, signal berth data and manual stopwatch timings are combined to ensure the timings are correct.

 

[DelW]

When reading timetables, I often wonder how they are written in such a way that they are a) accurate to each timing load and b) written so that a train can exactly follow them. How are times calculated in the first place? Is a test train run to work them out or is it a computer program?

I always assumed that Heath Robinson had provided the definitive explanation:

(Cartoon by W Heath Robinson titled "Ingenious plan for fixing the approximate times for trains to cover the required distances, used successfully in the compilation of the first timetable".)

 

[Railperf]

On top of what was mentioned, additional time for speed restrictions due to future or ongoing maintenance are added in to the timetable as 'engineering' allowances.

And where a faster train is likely to catch up a slower train and be slowed down, or where a train might have to wait at a junction for a train to cross its path ahead, then additional time is added to the running time and these are called 'performance' and / or 'pathing' allowances.

Of course the running times in dry conditions can be quicker than when the rails are wet and slippery as drivers have to accelerate and brake more cautiously to avoid wheelspin and wheel slide. So the general running time has to reflect this too.

Some of the longer station dwell times of over 1 min allow drivers to catch up time where the actual time needed to load / unload passengers can be as little as 30 seconds or so.

But it seems that to standardise the timetable, dwell times of a minute or longer allow for slower boarding during the rush hour. I have noted that many Elizabeth line station stops are booked for a full minute, but late running trains are moving again in 30 seconds to try and recoup delays.

There is also the need to regulate running times so that trains can get a clear path at the many flat junctions and stations.
On the WAML, the overall running time between Liverpool Street and Cambridge - for example - has to be regulated by the need to path trains at the various junctions - Bethnal Green, Hackney Downs, Clapton, Broxbourne, Stansted South and Shepreth Jn into and out of Cambridge.

You could plan a much faster timetable than at present up to Shepreth Jn but then the train might have to stop there for several minutes waiting there for GN and Thameslink trains to clear the junction ahead.

Even if you could get a clear run through there, there might not be an unoccupied platform at Cambridge!! So slower running tunes allow junctions to offer a clear route and station platforms to be vacated.

 

[The Planner]

Of course the running times in dry conditions can be quicker than when the rails are wet and slippery as drivers have to accelerate and brake more cautiously to avoid wheelspin and wheel slide. So the general running time has to reflect this too.

It generally isn't, apart from specific leaf fall timetables. Otherwise you would be standing time everywhere as SRTs would be over inflated.

 

[Railperf]

It generally isn't, apart from specific leaf fall timetables. Otherwise you would be standing time everywhere as SRTs would be over inflated.

The fastest recorded times in the RPS archives suggest that working timetable net sectional running times are slower than what is achievable given a clear flat-out run - which gives drivers a time buffer to help them stay close to their schedule when railhead conditions are poorer. Of course in really slippery conditions, trains generally accumulate time loss which proves the net SRT's are not based on wet weather running.

 

[zwk500]

The fastest recorded times in the RPS archives suggest that working timetable net sectional running times are slower than what is achievable given a clear flat-out run - which gives drivers a time buffer to help them stay close to their schedule when railhead conditions are poorer. Of course in really slippery conditions, trains generally accumulate time loss which proves the net SRT's are not based on wet weather running.

SRTs have a 5% allowance on the theoretically achievable times in the rounding, but this is not specifically for railhead conditions.

 

[Railperf]

SRTs have a 5% allowance on the theoretically achievable times in the rounding, but this is not specifically for railhead conditions.

Yes. But with working times calculated to the nearest half minute..the 5% allowance - where it exists - is not accurately applied to each section.
Here's an example - Stratford to Shenfield - csn be achieved in 12.5 min by a Class 321. Add 5% = 3 seconds per minute run time = 37.5 seconds. So a run time rounded up to the nearest 15 seconds of 13.25 min - or 13.5 min if rounded up to 1/2 min.
But the net time currently is 14 min for stock that accelerates and brakes better than the 321's with a minute extra engineering allowance thrown in for good measure. So we end up with 100mph trains dawdling along at 70mph on 90mph track to dissipate a 'spare' 2.5 mins And if you get a press- on driver, tben you are likely to catch up the preceding train at Shenfiekd if it is booked 3 mins in front..because its driver may well be taking it easy to dissipate the 'spare' 2.5 min it has over the same leg - a distance of 16.25 miles - feels slow to the passenger and is not exactly the most efficient use of the resources. You could probably get to Colchester close to 10 min faster if some of the inefficiencies were ironed out - engineering allowances spread more evenly as an additional percentage to each SRT rather than minute blocks. And rounding times to 1/4 minutes.

Interestingly - many driver schedules are displayed in full minutes - usually losing the half minute where it might exist and rounding it up to the next minute.

 

[louis97]

SRTs have a 5% allowance on the theoretically achievable times in the rounding, but this is not specifically for railhead conditions.

This isn't everywhere, this is done instead of applying engineering allowance within the schedules. Couple of examples of where it is applied: Passenger services on the Far North line and Electric routes in Scotland and Passenger Electric services in Kent.

 

[The Planner]

The fastest recorded times in the RPS archives suggest that working timetable net sectional running times are slower than what is achievable given a clear flat-out run - which gives drivers a time buffer to help them stay close to their schedule when railhead conditions are poorer. Of course in really slippery conditions, trains generally accumulate time loss which proves the net SRT's are not based on wet weather running.

SRTs have a 5% allowance on the theoretically achievable times in the rounding, but this is not specifically for railhead conditions.

That is due to rounding to half minutes, some SRTs are rounded up, some down on the cumulative journey/route, there is nothing added for wet weather. If a run of SRTs were 1 minute 12 seconds each then the first is normally 1 and the second 1½ etc.
I know the Southern add a percentage as they don't use engineering allowances, but I have never known it added elsewhere, just rounding.