New rolling stock planning in the Netherlands

[Bemined]

In the next couple of weeks, the NS will be experimenting with a different type of rolling stock planning. Historically the NS optimised the rolling stock planning per train, resulting in a lot of different formations on the same route, especially on long distance services.

The problem however is that this planning is nearly impossible to maintain when there is a disruption. Often this leads to overcrowded trains because the long formations are on the wrong side of the disruption. In the upcoming experiment the NS will reduce the number of formations.

They put a short video on their website with an animation of what they want to achieve (on the bottom of the article): https://nieuws.ns.nl/succesvolle-proef-geen-kleine-reparaties-op-drukke-forensendagen/

NS will run trains more often in a fixed composition. NS will shorten long trains in this trial, for example from 10 to 8 carriages, and extend short trains, for example from 6 to 8 carriages. This reduces the chance that a train is far too short in the event of a disturbance. The trial will ultimately make it clear whether this actually has the desired effect.

Looking at the rolling stock planning for next week though, it seems they made an additional change not mentioned in the article. A lot of train types come in two sizes. On several routes they limited the rolling stock planning to one size, for example all 3-car ICM units are on the Schiphol - Groningen/Leeuwarden service while the 4-car ICM units are on the Den Haag - Leeuwarden and Amsterdam - Almere / Deventer services. Some services still mix 4 and 6 car units, but with the 6-car units running all day and the 4-car units being the additional peak-only unit.

 

[AdamWW]

In the next couple of weeks, the NS will be experimenting with a different type of rolling stock planning. Historically the NS optimised the rolling stock planning per train, resulting in a lot of different formations on the same route, especially on long distance services. The problem however is that this planning is nearly impossible to maintain when there is a disruption. Often this leads to overcrowded trains because the long formations are on the wrong side of the disruption. In the upcoming experiment the NS will reduce the number of formations. They put a short video on their website with an animation of what they want to achieve (on the bottom of the article): https://nieuws.ns.nl/succesvolle-proef-geen-kleine-reparaties-op-drukke-forensendagen/

Looking at the rolling stock planning for next week though, it seems they made an additional change not mentioned in the article. A lot of train types come in two sizes. On several routes they limited the rolling stock planning to one size, for example all 3-car ICM units are on the Schiphol - Groningen/Leeuwarden service while the 4-car ICM units are on the Den Haag - Leeuwarden and Amsterdam - Almere / Deventer services. Some services still mix 4 and 6 car units, but with the 6-car units running all day and the 4-car units being the additional peak-only unit.

Interesting.

This presumably means they are now planning to run trains which they know will be overcrowded.

Maybe they will need a move to a more demand-based pricing system.

 

[philg999]

So they’re thinking this will make it more pleasant during disruptions, whilst making it more miserable during the normal service. Surely that’s the wrong way round? Most of the time the service is normal so they should be maximising the number of pleasant experiences!

 

[DanielB]

So they’re thinking this will make it more pleasant during disruptions, whilst making it more miserable during the normal service.

Not necessarily. The way they explained it to the public is that they'll make short trains slightly longer and long trains slightly shorter. Actually those very short trains are generally the most miserable ones as they cannot cope with sudden increases in demand (for example due to nice weather) from the average.

The former planning optimized by demand per train made sense when the number of passengers was sort of predictable, but you could have odd situations. For example the The Hague - Enschede service had one ICM-consist in between all the high capacity DDZ-consists in the afternoon peak. As of next week it seems to be all DDZ there, so not only more convenient in case of disruption but also more robust should for example a group of passengers miss their connection and end up in the train with less room instead of their planned one.

I think a more consistent rolling stock planning is actually also an improvement in normal service, as NS used to exaggerate a bit in demand adjustments: in the time I was still commuting by train they every once in a while tried to run my train just before the morning peak with a single 4-car ICM unit on Friday, leading to an overcrowded train on the quietest morning peak of the week.

 

[Bemined]

Also it should be noted that rolling stock planning was already very unpredictable on some lines, partially because the stock rotation was always messed up, but also because the planing was different each day of the week and changed every few months. They introduced crowd indicators in the journey planner in the hope that people would adjust their work times to avoid the busiest trains, but I don't think that works when it constantly changes which trains are the busiest ones. If the rolling stock planning is more stable and reliable perhaps more people will start taking the less busy trains.

 

[NCT]

I must admit that it surprises me that for a network that's so metro-like in timetabling has up to this point been so un-metro-like in rolling stock planning.

 

[dutchflyer]

Hmm- on metro´s (subways for Norte-Americano´s)-several of which in the Rotterdam area are in fact rebuilt former railwaylines-the same may apply. Train length adapted to expected demand. Which effectively mans sets coupled/uncoupled at some points/times. A long time ago there were even explanations that this was also done to do regular preventive maintenance of sets in between the busy peaks.
For those older Brits: did not that former Southern Region of BR-the 3d rail area- not do the very same? Keep trains at peak-hour full length for the whole day? This was fairly often mentioned as being the main example for NS to do their planning.

 

[The exile]

Hmm- on metro´s (subways for Norte-Americano´s)-several of which in the Rotterdam area are in fact rebuilt former railwaylines-the same may apply. Train length adapted to expected demand. Which effectively mans sets coupled/uncoupled at some points/times. A long time ago there were even explanations that this was also done to do regular preventive maintenance of sets in between the busy peaks.
For those older Brits: did not that former Southern Region of BR-the 3d rail area- not do the very same? Keep trains at peak-hour full length for the whole day? This was fairly often mentioned as being the main example for NS to do their planning.

Splitting full length sets to run shorter ones gradually fell out of favour for several reasons. The ones that occur to me at the moment are: more opportunities for things to go wrong; increasingly less practical at the London end as service frequencies increased - fewer paths available to get the bit you don’t want back to the depot - not to mention shortage of staff to do it.

 

[NCT]

Indeed. Thameslink, Northern City, and South Western Railway suburbans have all gone / are going to fixed formation. When the off-peak frequency isn't that far off the peak frequency - e.g. 12-16tph, there aren't enough paths to run portions into and out of the depot between the peaks. Sometimes it might be possible to dump a portion at the 'country end' where there's spare platform capacity or easily accessible sidings.

Where splitting down is done, it tends to be with a relatively homogeneous fleet and a simple pattern. 8-cars running inter-peak strengthened by a further 4 cars in the peaks. You wouldn't have different formations and different rolling stock types tailored to specific journeys.

Something like 6-car VIRMs running all day strengthened with 4-car VIRMs would one service group would be in line with London & South East practice. We wouldn't do things like targeting single 4s, single 6s, 6+4s and 6+6s for specific services on different days of the week.

 

[Austriantrain]

Hmm- on metro´s (subways for Norte-Americano´s)-several of which in the Rotterdam area are in fact rebuilt former railwaylines-the same may apply. Train length adapted to expected demand. Which effectively mans sets coupled/uncoupled at some points/times. A long time ago there were even explanations that this was also done to do regular preventive maintenance of sets in between the busy peaks.
For those older Brits: did not that former Southern Region of BR-the 3d rail area- not do the very same? Keep trains at peak-hour full length for the whole day? This was fairly often mentioned as being the main example for NS to do their planning.

Strengthening/weakening services during the day has fallen out of fashion elsewhere too. On the Vienna S-Bahn, 6-car (two units) used to „lose“ one set outside of the peaks, especially in the evenings, Sundays etc; sometimes also towards the extremities of the network. For the reasons mentioned elsewhere and also because demand has strongly grown off-peak, they will now run as 6-car continuously, including evenings and weekends (well, that is if they have enough serviceable units to actually run all sets as a double).