Reopening Cullompton and Wellington stations (merged topic)

[whole diary]

[Red Squirrel]

On it goes:

Councillor Mark Edwards, deputy leader at Taunton Deane Council, will speak with Mid Devon District Council and Devon County Council about working together towards re-establishing rail links between Exeter and Taunton.

This could mean the re-introduction of stations in Wellington and Cullompton.

source: Somerset County Gazette

[JayMac]

That's the same Taunton Deane Borough Council who, along with West Somerset District Council, have been trying to reopen the rail link between Taunton and Minehead. They, and their predecessors, have only been trying to achieve that since, oh, approximately 12:01am on the 3rd January 1971. Of course, WSDC can do little these days but talk about the line to Minehead. What with them having only about tuppence ha'penny in the bank.

Sticking block has often been their overlords, Somerset County Council. One minute they blow hot about rail enhancement projects, the next they blow cold. No mention of any input from SCC^‡(resolve) regarding Welly and Clumpton in the County Forget article.

[TonyK]

It only takes one or two councillors to raise a bit of a rumour within  the local media.. The art is to get a bit of momentum going, at least in Zummerset. Cullompton is in Devon, which gives it a bit of an inbuilt advantage. Devon County Council seem much better organised with rail projects, having seen the benefits almost before rail became cool again.

Wellington's politicos are, I think, looking to Tavistock for inspiration, and looking to link housing development to the station - can't see anything further coming prior to 2019 though.

[Red Squirrel]

It only takes one or two councillors to raise a bit of a rumour within  the local media..

It's almost as if people resented having their trains taken away by Marples and Castle, and would like them back..!

I try to suppress it, but I can't stop thinking that you could reopen a lot of missing links for ^40,000,000,000 (or is it ^50,000,000,000? Oh and some trains too... ^60,000,000,000)

[TonyK]

I try to suppress it, but I can't stop thinking that you could reopen a lot of missing links for ^40,000,000,000 (or is it ^50,000,000,000? Oh and some trains too... ^60,000,000,000)

Now, now! No need for that. I didn't say they would actually achieve anything!

Being slightly more serious, I can't see any reason why it can't be done for the cost of platforms and signals. The line between Taunton and Exeter is a lot quieter than that passing through Lawrence Hill and Stapleton Road, and trains stop at those stations. And I keep reading about the stock soon to be cascaded because of electrification.

[grahame]

And I keep reading about the stock soon to be cascaded because of electrification.

But remember too the stock we keep reading about that has to be withdrawn in 2019 because it will no longer be legally fit for purpose.

I seem to recall seeing a DfT^» (Department for Transport - about) document showing rolling stock plans looking forward 30 years and predicting requirements, but alas I've also seen documentation than bases future plans and growth on conservative (low) estimates - said to be to avoid spending money on promoting things that will be underutilised, but in practise if the current trends continue it would mean a system running beyond the capacity it has for decades to come.

[Network SouthEast]

And I keep reading about the stock soon to be cascaded because of electrification.

But remember too the stock we keep reading about that has to be withdrawn in 2019 because it will no longer be legally fit for purpose.

Porterbrook (who FGW^▸ (First Great Western) lease their Pacer trains from) have a plan to make them compliant with the PRM^▸ (Persons with Reduced Mobility) TSI requirements. Only Angel Trains have said they are scrapping Pacers that they lease out to their customers in 2019.

Here's a link to Poterbrook's plan for compliance: http://www.porterbrook.com/downloads/brochures/14x%20Brochure.pdf

[John R]

They'll only do it if they get a guarantee of use for the units over a period that covers the cost of the work.

[TonyK]

They'll only do it if they get a guarantee of use for the units over a period that covers the cost of the work.

TOCs^▸ (Train Operating Company) are quick to blame shortage of rolling stock for overcrowding and not increasing services. They should, in thoery be overjoyed, and qqueuing for any spare units. Although I have said before that their profit is greater from a 125 full to bursting point with disgruntled passengers than it would be from two trains with everyone able to get a seat. That said, GWR^▸ (Great Western Railway) made good use of the small number of extra units they were able to get a couple of years back.

[John R]

Agree, but remember that we are looking forward to guaranteeing usage after 2019. By that time, the displacement of the vast majority of the Thameslink fleet, the Valleys' electrification, together with the current North West electrification (excluding Class 185's which are already accounted for) will release enough DMU^▸ (Diesel Multiple Unit)'s to cover the 144 odd Pacers in service. 

In addition, there's a clear steer that MML^» (Midland Main Line. - about) will now be complete by the end of CP5^▸ (Control Period 5 - the five year period between 2014 and 2019) (ie 2019), which will cascade a lot of DMU's at the top end of the fleet specification. Scottish central belt electrification similarly, not to mention trans-Pennine wiring and a few other smaller schemes* which will each release a handful. That should release enough to cater for the current suppressed demand that you allude to, together with some room for growth in the next 6 years.

So, if I were a shareholder being asked to sign up to a significant investment programme, I'd want a very good business case that a) there would be work for the units and b) the lease charge I could make covered the costs of making the units compliant. If the cost of leasing goes up too much then clearly they will look less attractive to operators, particularly given the loss of seating for the accessible toilet will give them a very low capacity

* I can think of Gospel Oak to Barking (8?), Rugeley branch, Glasgow suburban branches, Windmere now likely, plus any more that emerge in the next few years.

[grahame]

TOCs^▸ (Train Operating Company) are quick to blame shortage of rolling stock for overcrowding and not increasing services. They should, in thoery be overjoyed, and qqueuing for any spare units. Although I have said before that their profit is greater from a 125 full to bursting point with disgruntled passengers than it would be from two trains with everyone able to get a seat. That said, GWR^▸ (Great Western Railway) made good use of the small number of extra units they were able to get a couple of years back.

The business of an overcrowded train being more profitable, even though potential travellers are put off, has interested me for a while - looking at how the arithmetic works.  So I put a program together to try out a few things / figure and see what I got.

Potential passengers - 300
1 carriage ...  1.89 142 429.0 4.00
2 carriages ...  1.70 255 1110.0 2.00
3 carriages ...  1.22 275 675.0 1.33
4 carriages ...  0.95 285 120.0 1.00
5 carriages ...  0.78 291 -483.0 0.80
6 carriages ...  0.66 295 -1110.0 0.67

Figures on each line ...
... actual loading
... physical number of passsnegers carried
... profit made
... potential loading if everyone who wanted to travel did so

So there's 9 times more profit to be made running a 2 coach train full to bursting point than there is running a 4 coach train with a handful of spare seats, even though your 2 coach train has put off 30 people to the extent that they're not your customers at all, and have driven/ taken the bus etc.

If your potential customer pool has increased - let's say to 450 people, then it is worth adding an extra carriage.

Potential passengers - 450
1 carriages ...  1.89 142 429.0 6.00
2 carriages ...  1.90 285 1470.0 3.00
3 carriages ...  1.70 382 1959.0 2.00
4 carriages ...  1.35 405 1560.0 1.50
5 carriages ...  1.11 418 1041.0 1.20
6 carriages ...  0.95 427 474.0 1.00

Here are the inputs I used to my model:

Code:

pricePerSeat = 9.00			# How much it costs to provide a seat
incomePerPassenger = 12.00		# How much each passenger pays to travel
seatsPerCarriage = 75			# Number of seats per carriage
crewCosts = 600.00			# Crew cost for running the service
goldenLoad = 0.5			# Loading below which no-one is put off
physicalLimit = 1.9			# Loading above which it's physically impossible to board
mustGo = 0.5				# Proportion of people who simply have to travel
rampScale = 0.1				# Scaling by how much other people are put off by overcrowding

And of course the results will change based on the numbers which are educated guesses on my part (any expert like to come up with real numbers for me?)

The next complication comes in the need to add in the cost of the return working ... if the potential passenger count on that is only a half peak direction working, the results are all the more stark.   With 300 on the outward leg and 150 on the return leg, a 2 coach train turns in a profit of 864, but a 4 coach train results in a loss of 1380. 

Then you have partial journeys and the effect of fare levels and frequencies on the numbers of potential passengers.  And then you need to conside the effect of promotional offers to get people using seats which would be running empty otherwise, and how to avoid those offers tempting people who would travel at full price to travel more cheaply ... I wouldn't want to run a TOC

For the more techical reader - my algorithms, coded in Python:

Code:

def getPassengers(potential, seats):
	loadfactor = float(potential) / float(seats)

	if loadfactor < goldenLoad:
		return potential,loadfactor

	putoff = (loadfactor - goldenLoad) * rampScale
	if putoff > mustGo: putoff = 1.0 - mustGo

	willgo = int(potential * (1.0 - putoff))
	willtake = int(seats * physicalLimit)

	if willgo > willtake: return willtake,loadfactor
	return willgo,loadfactor
	
# generate result tables from 50 to 450 passenger

for potential in range(50,500,50):

	print "Potential passengers -",potential
	for carriages in range(1,7):

		seats = carriages * seatsPerCarriage
		passengers,plf = getPassengers(potential,seats)
		alf = float(passengers) / float(seats)

		incomings = passengers * incomePerPassenger
		outgoings = seats * pricePerSeat + crewCosts

		profit = incomings - outgoings

		print carriages,"carriages ... ","%.2f" % alf,passengers,profit,"%.2f" % plf
	
	print

[trainer]

I don't understand any of the figures at all, but I do understand Graham's words and it seems to me that the point is as true as it's ever been: passenger railways don't (or rarely) make profits.  As was recently re-iterated in the Beeching Night programmes on BBC^» (British Broadcasting Corporation - home page) Parliament, only British governments seem to have the (to me) perverse ideology that railways must pay for themselves.  Most of Europe understands that the railways, like roads, are an integral public service which must be reasonably subsidised from general taxation.

The fact that a profit can only be made from squeezing people in like cattle shows what the basic attitude to people against profits is in this business. Governments are elected and I understand that asking for more money will not be an election winner these days in Britain, but we certainly get what we don't pay up for.

Local councillors ought not to build up false hopes: they know more than anyone how empty the coffers are and how their voters would react if asked for more money to make their railway aspirations come true.

[TonyK]

Nice work there, Red Squirrel, although I wouldn't be surprised to find that train operating companies and DafT already have (and use) such numbers. I agree with trainer too. The drive for profit everywhere by Government, even at the cost of resentment, reminds me of the story of a former Chancellor at a funeral, who whispered to the man next to him "Put that thing on wheels, and you can make five men redundant".

Yes, I was aware of the cascading that will happen with electrification. I hope the end result is even more profound, with rolling electrification after the initial programme is complete. If that aim is taken as a discrete project, rather than having to go the "whole promise it three times cancel it twice" that rail projects always have, then so much the better.

When the bit through Wellington and Cullompton is electric, the should-we-shouldn't-we question will be informed by different economics. Having electric vehicles calling at stations with short distances between them is surely much more economic (or probably less uneconomic) than trying to do the same with DMUs^▸ (Diesel Multiple Unit). We've got a long way to go, though. But when you think that the high output electrification train is said to be capable of doing a mile in a 6-hour possession, Bristol to Penzance becomes a 6-month job. I hereby call upon the Transport Minister to take my idea on board, and for Network Rail to draw up a schedule covering the period from 1 January 2019 to end of task, somewhere around 2036, quicker if we leave Scotland to get on with their own bit.

[Red Squirrel]

Nice work there, Red Squirrel...

Kind of you to credit me. but I think Graham should get at least some of the kudos... and I hope Graham will forgive me, but I'm afraid I would have done it in php.

[grahame]

... and I hope Graham will forgive me, but I'm afraid I would have done it in php.

And I would have done it in PHP too if I was giving a PHP course at the moment - but I haven't done one of those in 10 days     ... the week just gone was Python, C and C++ ...

On a serious note, it was fascinating to try the effect of changing the relative cost of a seat, takings per seat, and fixed costs (I've used a crude differentiator with nothing between the two extremes).  Also great fun playing with the potential passenger flexibility - at what point will some stop travelling because the train's too crowded, and what proportion will try and cram in no matter what.   Where the algorithms get to be really fun, and I'm sure that First and the DfT^» (Department for Transport - about) have modelling software, is where they're extended to cover full services (which changes such as more shorter or fewer longer trains), and time (where customer / potential customer reaction over short and long term change the results).