The Most Overcrowded Lines-Latest Survey

[whole diary]

[Timmer]

http://www.telegraph.co.uk/news/uknews/road-and-rail-transport/8118215/Rail-chaos-the-most-overcrowded-lines.html

According to the latest official statistics the worst levels of overcrowding are to be found on trains operated by First Great Western. National Rail trends found that its trains were carrying 8.2 per cent more passengers than they were designed to do last year. The company, which operates trains out of Paddington, was followed by London Midland, with 5.9 per cent.

However the full scale of overcrowding is still unclear, partly because up until now the figures have only been collated for London termini.

Percentage above capacity (2009):

First Great Western 8.2 per cent

London Midland 5.9 per cent

Southern 3 per cent

South West Trains 2.3 per cent

National Express East Anglia 1.8 per cent

Chiltern 1.8 per cent

London Overground 1.6 per cent

Southeastern 1.1 per cent

C2C 1 per cent

First Capital Connect 1 per cent

[IndustryInsider]

However the full scale of overcrowding is still unclear, partly because up until now the figures have only been collated for London termini.

And in case they didn't realise, the worst overcrowding on FGW^▸ (First Great Western) services into Paddington actually occurs between Southall and Ealing Broadway as Ealing itself generates a large commuter flow.

[inspector_blakey]

Sigh. 

I wish the hacks who write these stories would actually bother to explain the basis on which those figures are calculated so that they actually mean something. Sadly it seems that the general intellectual level of many of them is such that any number with a "per cent" after it is assumed to be authoritative, so it's just unquestioningly regurgitated without giving it any context. Fair enough they reference that it's the PIXC calculation; however, I probably take more interest in this sort of stuff than most yet I have no idea how PIXC is calculated which is very important in deciding whether those figures are shocking or not.

Questions that I have in the light of that article:

1. Is "capacity" defined as the number of seats, number of seats plus an allowance for a modest number of standing passengers, or full crush loading? If it's the first of those, then 8.2% doesn't sound too bad. If it's the third then it may be much worse.
2. Are those figures averaged across the day, do they just apply to a single train, or are they calculated across peak hours? I'm sure the accountants at FGW^▸ (First Great Western) towers would be putting out the bunting if all their trains into Paddington were running with a load factor of 1.08. If that figure is just calculted for one particular train in the morning arriving at Paddington with 8.2% PIXC then it's grossly misleading to "cherry pick" the figure for that train and imply that all FGW's services are running like that.
3. Does the PIXC figure just include passengers alighting at Paddington? If so, as II mentions, may of them may have just boarded at Ealing Broadway so they'd have been standing for a very short period of time, hardly intolerable. However, if a train was running all the way from, say, Oxford or further afield at 8% above its crush loading that would be a very different matter.

Like I said, really difficult to interpret what those numbers actually mean. From the way they're presented, it could be that every single train into Paddington all day carries 8% more passengers than its crush loading throughout its journey (which is obviously not the case) or that there's one single train arrives at Paddington in the peak and sets down 8% more passengers than it has seats for (again, clearly not the case). So the real answer must be somewhere in between those two absurd extremes, but the article gives you no idea of where it is on that continuum!

[tramway]

I don't think we'll be seeing targets to reduce these figures published on the DafT website in the near future.

Anything useful that DafT could be held to account for like improving things aren't going to get anywhere near the information to be published on Department websites as promised by 'call me' Dave.

[paul7575]

On long distance trains, PIXC simply means in excess of seated capacity.  On short distance commuter routes, broadly defined as services with multiple stops within 20 mins of the main terminus IIRC^▸ (if I recall/remember/read correctly), PIXC is measured allowing for a designed standing capacity of the rolling stock.  In both cases it is measured at the height of the morning peak as it has been found that the evening peak is more spread out.

So as they publicise a combined figure for FGW^▸ (First Great Western) you immediately have to question how the figure is weighted between HSS^▸ (High Speed Services), LTV^▸ (London [and] Thames Valley) or West services, AFAICS^▸ (As Far As I Can See)

Considering just standing capacity, it was generally thought to be a standard percentage value in the Southeast, but in other areas it could be diffferent. This led to some critical headlines when Centro in Birmingham were told to recalculate using the same figure DfT^» (Department for Transport - about) used for the SE.  Clearly it was a bit daft for LM^▸ (London Midland - recent franchise) services either side of Northampton to have different rules...

It turns out the standee figure in the SE for typical commuter stock used to be considered as 30% of seated capacity, and Centro were only using 10%, which presumably helped their overcrowding case quite a lot.

However after some Googling a couple of years back I discovered DfT actually use the usable floor area divided by 0.45m^2 per passenger, it's possible this comes out to an average of about 30%.   But then in the particular case of units designed or adapted for a high standing capacity the divisor changes, and is 0.35m^2 for SE 376s, and 0.25m^2 for SWT^▸ (South West Trains) refitted 455s.

http://www.publications.parliament.uk/pa/cm200708/cmhansrd/cm080721/text/80721w0039.htm  (Towards the end of the page)

I'd assume trains like LO's 378s with longitudinal seating could be an even higher theoretical standing capacity - and I'm assuming they won't be the last of their type for inner suburban routes... 

HTH

Paul

 

[Timmer]

One thing that has been good with all the reports in the press and TV about overcrowding is not hearing anyone use that stupid politically correct term of 'crowding' to describe a train that is rammed full 

[onthecushions]

I did some comparative calculations on this some time ago and found (by measurement) that the area per customer on a bus/coach was 0.18m^2.

Much of the advantage of the railway is therefore spent (believe it or not) on passenger comfort....

LUL^▸ (London Underground Ltd) figures for full and standing are also instructive.

Personally I can forgive full trains when they are 12 car emu's and frequent. When they are short dmu's, (even 5 stds on an HST^▸ (High Speed Train)), on infrastructure capable of so much more, then I grouse.

OTC

[SDS]

Curious as to the different figures between FGW^▸ (First Great Western)'s old standard coach and the new Hi Density coach. Tables must add to that comfort figure.

[grahame]

I see (further up the thread) references to more "passengers than they were designed to carry" rather than seats. On that basis, some of the SWT^▸ (South West Trains) refits that tear out seats and give room for more people to stand will decrease overcrowding, and putting further racked seats into 125s without altering the standing capacity sill also decrease overcrowding stats, although (personal view) they'll make for a less pleasant journey at other times.

[inspector_blakey]

Sounds from Paul's post above as if on long-distance services capacity is simply defined as number of seats, so FGW^▸ (First Great Western)'s refurb will have upped capacity by 10% or so. But what isn't clear is how capacity is defined when, say, an HST^▸ (High Speed Train) works an OXF^» (Oxford - next trains) - PAD^» (Paddington (London) - next trains) service. Is that long-distance or commuter, so how is capacity defined in that situation. This all seems rather opaque, especially since it would appear that different operators are (or have been) using different definitions of capacity. I would hope that it's all codified somewhere on the DfT^» (Department for Transport - about) website so that there's a standard approach being used for all those operators whose figures are quoted in the Telegraph's article.

But this does all rather prove the point about how numbers like that are meaningless unless we actually know the basis on which they're calculated! Fourth estate take note...

[willc]

The methodology employed is as follows:

2.4 Passengers in excess of capacity
Background
Train operators provide data on passenger numbers to the Department for Transport (DfT^» (Department for Transport - about)) to allow for the monitoring of train crowding levels.  In the past DfT monitored crowding for London commuter services under a regime known as ^passengers in excess of capacity^ (PiXC) and this formed the basis of the crowding statistics published in National rail trends.  Though DfT no longer uses the PiXC regime, the PiXC measure has been reproduced in table 2.4 for 2009.  The PiXC measure considers the planned capacity of each service arriving in London, and the actual number of passengers (excluding first class) on the service at its most crowded point on the journey.  PiXC is the difference between the two.
More passenger count data are now being collected by train operators, and in future it is
planned that new and improved statistics on crowding will be made available covering more of the country. As these new statistics are not yet available, an interim PiXC table has been prepared for 2009.
Methodology PiXC applies to weekday commuter trains arriving in London between 07:00 and 09:59, and those departing between 16:00 and 18:59. The measure is derived from the number of passengers travelling in excess of capacity on all services, divided by the total number of people travelling, and expressed as a percentage. Capacity is deemed to be the number of standard class seats on the train for journeys of more than 20 minutes. For journeys of 20 minutes or less, an allowance for standing room is also made. The allowance for standing varies with the type of rolling stock but, for modern sliding door stock, it is typically approximately 35 per cent of the number of seats.
The data underlying the PiXC measure shown were collected each year during the autumn (mid‐September to mid‐December), and have been aggregated to represent a typical weekday (i.e. excluding school half‐terms and days when services were disrupted by engineering or bad weather, etc.).
The frequency of counts varies by train operator.  Where a train operator has a proportion of its fleet fitted with automatic passenger counting (APC) equipment, the number of passengers on each service will be counted several times during the autumn period.  This allows for an average to be calculated for each service.  However, some train operators have nil or limited APC fitment and so passenger numbers are obtained by manual counting.  This can mean that there is only one record for each service so caution should be used when referring to these results.
Other comments
Under the historic PiXC regime, DfT set limits on the level of acceptable PiXC at 4.5% on one peak (morning or afternoon) and 3% across both peaks.
DfT now sets a variety of performance targets for its individual franchise holders.

FGW^▸ (First Great Western) stats are done by manual counting.

Taken from the ORR^» (Office of Rail and Road formerly Office of Rail Regulation - about)'s latest National Rail Trains report on performance, see pages 18-19 at
http://www.rail-reg.gov.uk/upload/pdf/nrt-ch2-railperformance.pdf