Talk of cutting back on aspects of the CRL raises a lot of concern for many people, especially those on this blog and that is completely understandable. There may be a good case for delaying some of the stations which is something I think should at least be investigated but there has been more opposition on what should happen with the eastern link. The reality though is that there are some legitimate reasons for not building it and with this post I thought I would explore them to help inform the debate. I think the issues with the eastern link fall into the following groups, Demand, Grades and Future Development:
Demand
When we build infrastructure like the CRL it is really important that we design it in a way that benefits the most people. Of the stations in the CRL something like 50% of people be using Aotea while about 30% will use Britomart. Trains coming from the west will obviously have a much faster trip to Britomart and eastern line trains will be largely unaffected so that just leaves the Southern and Onehunga lines. If you set two trains off from Newmarket, one via Britomart and one via Grafton and the Eastern line then they would both arrive at Aotea at about the same time. That is important because if you are focusing on making journeys quicker for the most people you would go via Britomart to make it easier for the majority of passengers.
Demand for CRL stations
Grade
As many people know the CRL is going to be pretty steep for a rail line with some parts of it being as steep as 3.5% (the EMUs are being designed to handle this grade). As a result of the grades the Symonds St station will be pretty deep at about 40m below ground level and while the ground level does drop a bit after the station it will still be a very steep climb back out of the ground to the existing rail network.
The problem is that due to the grades it will likely make it almost impossible to grade separate the junction between the eastern and western link. Living without grade separation may be ok if we didn’t care about running lots of trains around the network but as soon as we decide to pump up our frequencies they will start to cause some real limitations. To grade separate the junction we would need to bring at least one line over the lines of the other side of the other link and that requires quite a bit of space, space we don’t really have. As an example the grade separated junction in Wellington is about 500m long before the tracks cross each other and while we could bring that down with steeper grades there is only about 450m from the the Symonds St station to the current western line. This means that we may not be able to get the grade separation finished before we hit the current western line at Mt Eden and the lines can’t diverge until that cross over happens. It is possible that we may simply not be able to grade separate these links and that could have massive implications for the rest of the network.
Even just building the project with a flat junction still results in at least $100m worth of costs however if we grade separate that junction we would probably see costs go up around the $300m mark and would likely to need to be done as a massive cut and cover box.
Future Development
This really comes down to how we think about our future rail network. If we were building the CRL as the final piece of the puzzle then we should definitely be building the eastern link but instead we want to expand the rail network to the North Shore using a cross pattern like the image below we need to think differently. Do we really want to spend perhaps up to $300m to build an eastern link that may only be in use for 20 or so years after which it may not be used except for exceptional circumstances?
There is probably quite a few things that I haven’t covered but at the very least I think there is a debate that needs to be had about how important the eastern link is to us. Does it serve people the best? is it actually physically possible without incurring potential extra costs? Does it really fit with how we would like our rail network to eventually look like?
Of course dropping it would have impacts on other things like how we would operate the network but perhaps that is a post and discussion for another day.
Processing...
The obvious first response is that without the eastern link you end up with way too many trains on the western line.
Is it an option to build the Dom Rd spur instead with the money saved…? getting too many spurs? politically too tricky?
Well this pattern would sure make for killer frequencies on the Western Line!
This is conceiving the system like a real ‘Metro’ with separate lines, and it has its appeal, not least of which it spreads cost out further so that we are spending sums more in line with growing demand and pictures a more ambitious network sooner.
I guess having a complicated running pattern out of the southern portal also means having to grade separate Newmarket at no small expense too.
What this idea requires is a major communication campaign explaining that by building it this way now is only possible as part of an ultimately bigger programme to build ‘The Cross’. Because without that future work it will clog up like Britomart is now. But still not so dumb, spend less now, get it built and grow the demand to show the need for an even fuller system. The increased demand that the CRL will bring [+co-ordinated bus interchanges, integrated fares etc] will make future expansion of the network even more popular and politically supported.
I’d like to see a post CRL running pattern now though Matt, and one that includes a good balance of Grafton services…?
I would have to question the demand issue. Are we really certain that Britomart and Parnell together on the east side will have significantly more patronage than grafton, newton and K rd on the west side? Probably initially, but over time I could see it swing in favour of the uptown route with development and intensification.
The fact that Newmarket to Aotea is equidistant by either path is an absolute blessing, especially when you pair it with the need to run two lines through Newmarket. We can just run one line each way and give people the option to transfer at Newmarket if the want a quick trip the other way.
While I do like the cross pattern as a long term option we need to make the CRL work fully and efficiently for the existing network.
It is simply ludicrous to say its ok to build a two billion dollar CRL that doesn’t work efficiently, because one day we might build a second crosstown CRL leading to a new harbour tunnel and on to a north shore rail network. That’s saying we can make the CRL work real good if we just spend another six or eight billion dollars some time in the mid 21st century.
A the end of the day not having the eastern link means the western line on one side of the tunnel and everything else on the other, instead of the perfectly efficient half and half.
The only outcome possible there is a lot of trains that have to bounce in through the tunnel and back straight out again, making two passes in the tunnel to do one useful trip. That’s an inefficient waste of trains and service in the first instance, but more importantly it limits the effective capacity of the CRL significantly. Every one train you have to send back down to Britomart near empty is one to can’t run through on a service pattern that can meet actual demand. There are ways to sneak around that, you can store up trains at a stabling yard at the southern end, or piggyback them out of servie to other service runs back to the other side… but in any case that’s a lot of dead running. At any one point in time a significant fraction of trains would be empty and not able to carry passengers, so we’d need more trains and more drivers to do the same amount of work. That makes the network more expensive which consequently means we get less service for the same amount of money.
In a similar problem going in and out along the full length of the tunnel means it would take considerably longer to get a train back out to the end of the southern lines, thats a big issue when you want to get more than one peak run out of a train. With an eastern link a train that has just completed a run delivering to the tunnel stations can immediately carry on out to one of the suburban lines. Without and eastern link most would have to detour back through the full length of the tunnel again to get back to the suburban lines. That could add a good 15 minutes or more to each run, which is very significant delay for peak operations.
This its the same dilemma we have at Britomart today except the tunnel gets a bit longer, everything has to go in then out again the same way.
The only difference with a western-link-only CRL is that a portion of trains can run through to and from the west. Effectively this means that the capacity gains we could achieve would be only equal to what we could realistically run on the western line. Without the eastern link our CRL will only gain us maybe 10 or 12 trains an hour across the network.
To me this sort of concept focusses only on peak commuter flows, and ignores the longer term role as an all-day every-day rapid transit system.
At the end of the day any of the above service patterns can be done with an eastern link, but without an eastern link there is a huge restriction on the service patterns we could run. If the eastern link costs $300 million then so be it, it is an essential part of the infrastructure and savings would have to be made elsewhere, it’s needed to ensure we don’t shoot ourselves in the foot from day one.
Must be a good week if I am echoing Nick’s comments twice.
I can understand NOT building the East Link, but are we going to stump up the cash now (when we are struggling with just the CRL currently) to get the North Shore Line built to allow the optimum pattern of rail transit.
However a dosage of reality might kick in here.
Something tells me that I (that being my Generation (and I am only 27)) will be carrying the “can” for the 2032-2040 (so last life of The Auckland Plan) projects which at the moment all things considered includes the North Shore Line. So I have to think long term here.
North Shore Line won’t be until at least 2032 (at best) and so I have around 10 years (assuming CRL opens 2022) of CRL patterns that would work something like this (scratching the North Shore bit): NO, NO, NO! When the CRL tunnel is built, the east-west link connecting the CRL to the Western Line must be built – no half-arsing with just building a west link. To build just a west link from the Western Line to the CRL line would eliminate all redundancy of the inner circuit rail system (The Inner Circuit being any rail line between Otahuhu and New Lynn (so basically the boundaries of the City Monthly Pass)). You need an East Link from the CRL to the Western Line to allow back up if the Newmarket-Britomart section of track is down (break downs or tunnel collapse/blockage at Parnell or Britomart Tunnel Entrance) and maximum running flexibility of services (such as North Shore (edit: change that to Orakei, to Britomart) to Aotea Station, Grafton via CRL, Newmarket, Onehunga, Airport, Puhinui, Manukau, Botany,(fail Manukau and Botany, the train would loop round to Papatoetoe, Otahuhu then) to Glen Innes, Britomart and finally back to the North Shore (edit: repeat loop with no North Shore Line present)). In short the East Link is just as much-needed as the West Link. Auckland Council and Auckland citizens should make it a bottom line condition with the CRL that the East-West Link IS BUILT – nothing else half-baked!
That was quoted (with changes) from a post I recently did. Ok it also assumes the airport and Botany Line are also built but you get the picture one way or the other.
10 good solid years where the East Link would be used and probably a few more after that when the North Shore Line opens. In any case I like redundancy capacity which our rail network so lacks – the East Link being that = the redundancy line.
As Nick said:
At the end of the day any of the above service patterns can be done with an eastern link, but without an eastern link there is a huge restriction on the service patterns we could run. If the eastern link costs $300 million then so be it, it is an essential part of the infrastructure and savings would have to be made elsewhere, it’s needed to ensure we don’t shoot ourselves in the foot from day one.
It seems strange to me to have all these unkowns ie how trains will ciculate, what lines go where, and frequency and capacity conctarints unkown or ignored – but we have a project?!
Is this the right route anyway? how do we know if we have all these unkowns? At a high level CBD has 75,000 workers – with a fast rush to waterfront/shortland st – half will be well servered by Britomart, other half get nothing from CRL ie Wynyard/Fanshaw st. CRL does little for most of the 50,000 University/AUT student. Aotea as an entertainment area creates little demand, and all outside peak. And rest of CBD is largely student apartments.
So why not Fanshawe/ponsonby/Mt Eden? or smaller loop fanshawe/victoria st/universit/parnell? I know other options were looked at – but how can they be valid, if it was unkown how the network would actually operate. Minutes saved is meaningless when you still have a limited capacity system. And catchements are only useful, if they understand them properly
Mark you’re right that the physical design cannot be divorced from the running pattern design, that’s what this post is about, and is the core of Nick’s counter argument. This is what we now need from Matt; how would you run on this ‘skinny’ design?
But I disagree with your low views about the utility of the CRL route. Aotea is more than close enough to the Universities, I work there and my bus (when I don’t ride) stops at the back of the Aotea centre and that is a fine and well used walk up the hill. All PT involves a walk, it is rarely point to point like a car, this is one of the reasons it has such a big impact on health outcomes. It is a mistake to insist on a car-like convenience because to achieve that will mean a sacrifice of speed and frequency, by wobbling a route all over town.
Which brings me to an even more important point about the CRL: The real step change that it brings is not to do with the CBD at all. By linking up and through routing the existing lines the CRL unlocks the latent capacity in the whole of the network, it’s effects reach from Swanson to Papakura and are not just about the CBD. And because of the revolution in coordinating the bus network to interchange stations in places like Man City, Panmure, New Lynn, the catchment will be greatly expanded beyond these places. But this is only possible with the higher frequencies and further reach that the CRL will provide. As well as a real integrated fares system too.
It is important that the new Link runs directly through the city in order to add directness to this process but also to keep cost within reach. More coverage will be added later, especially west at the Tank Farm. And this will be easier to achieve on the back of the millions of extra journeys that the CRL will provide.
The CRL will achieve a huge amount, but no it doesn’t go everywhere. And it isn’t just, or even mainly, about the areas above the new stations and line, but about the vast out of reach capacity sitting wastefully out reach in the existing rail Right-Of-Way.
thanks – I’m finding all this interesting re how it could work. But still don’t see how they can go ahead or even compare against other projects. My main concern is it seems to hinder things like Western line capacity, on the previous posts – and we can’t get better than 10 min, and capacity on west is used by 2016. I know there is some debate re headways etc.
But my gut feel is that Avondale to Southdown should at least be properly compared. It frees up more of western capacity, and provides an alternative freight route north, and connects to new job areas. It also may help with Onehunga line and the 2 carriage impacts on Britomart.
Well National wants to close the Northland Line so John Key can watch Bankers on gardening leave ride expensive mountain bikes past the back of his electorate office… well really as yet more welfare for their mates in big trucking. So that’s not a pressing issue, but also Avondale-Southdown does nothing to turn the bottleneck at Britomart into a through route. I know you don’t think much of the CBD but it is still and will remain the biggest destination and drawcard for PT users, and that deadend is the whole system’s biggest problem.
Why do you say 10 min headways is a limit on the Western? At the moment we just don’t have the rolling stock but by 2016 that issue will be well solved, but not the Britomart restriction.
Patrick – RFP out now for NI Ports review – out of that we will see a better use of Northland and Tauranga.
Everything I can find out has 4800 per hr am peak as westernline capacity, based on 10 min, is best we get out of Western line with CRL. I looked at the various options on the previous post re use of CRL and merging western/southern on 5min. So western gets 10min frequency – which is what govt and AT seem to have used in teh analysis as well.
It’s not a matter of “not think much of the CBD” – I personally did the CBD strategy, and have sat on the CB Board. i did the research on it’s chnaging face back in 2003-04 – and all trends have been confirmed. So feel I have a better understanding of direction and use, that transport people. I still have export education, and commercial/property owner contacts, and feel my predications are well based.
We currently may have 11-12% of employees (remeber also higher use of fulltime FTE than say fringae and other areas ie oart-time workers. So % of transport movement is less that a 11-12% share).
This 11-12%, IMHO is only going to drop. So we need to think about the 90%. A shuttle from Onehunga/penrose picks up a very large working population. The corridor is there and protected, with a shortish tunnel under Hillsborough to deal with gradient.
So all I’m saying is there needs to be a functioning plan of network operation/capacity constraints, and only then can you compare and rank projects.
Mark, none of the post-CRL analyses have accounted for the obvious: buy more trains, run them more frequently. 10 minutes is not even vaguely close to the capacity limit of any part of our rail network except the single-track spur lines (which are constrained by how quickly a train can do the return journey). The signalling upgrades that have been carried out allow, I believe, four minutes headway on any section of track.
The 10 minute frequencies are based on the number of EMUs being purchased, not any kind of limitation imposed by the track configuration. If we had best-quality signals and well-behaved passengers it’d be possible to run two minutes, or even 90 seconds. Short of that, though, there’s still significant capacity beyond the 10 minute frequency restriction imposed by the number of available EMUs. Buy more trains and it’s possible to operate more services.
Mark – AT and its consultants already have prepared operations plans to help decide on the best use of the infrastructure and how to get the most patronage but they haven’t been released publicly yet. There will be far more trains that one every 10 minutes on the western line regardless of if the eastern link is built or not.
There are also other studies going on to determine things like just how much capacity the roads have in the CBD so there is plenty of information being fed into the project. Most of that we will have to wait for and AT will have to release as part of the designation process.
I have also heard that the reason why the eastern link was included in the first place is not because it is necessarily a good idea (I’m on the fence with it) but for political reasons as people higher up could only understand the concept of trains going around in circles and didn’t understand how that would happen without it.
Mark I live in hope for any kind of ports strategy, at the moment there is none what so ever, that being virtual communism apparently. But then that there might be a rational one- just imagine. In which case it might, should, include more action at Marsden Pt and the necessary investment in the NAL…. but there’s a bit to do there besides the Avondale line. Anyway agree that there is likely demand West-South which is why any running pattern should include good services down the western-Grafton-Newmarket- south. And ideally the Avondale line, although how to run that isn’t straight forward either!
Now as to CBD demand, I don’t doubt your expertise but it seems you are only looking at one kind of traveller: the two universities there have over 1.7 square kms of floor space and are still growing, more than just employees need to travel. But also, I repeat, the CRL is vital for the whole network and is not just dependent on CBD sourced demand- it really is about fixing a critical gap in the line that cripples the whole system, and happens to be in the CBD. Someone going from Papakura to Penrose will benefit from it because it means so much more frequency on their route even though they may never see the inside of Aotea station…. it’s a network fix, and that fix has to happen where the break is.
I know I keep harping on this, but yet again it feels like the Southern Line is getting treated as a second-class part of the network. Without the eastern link there is no easy way to run services between the south and the east. Those of us between Otahuhu and Britomart already have very, very substandard service frequencies (I do not consider a 42-minute gap between services before 8pm to be in the slightest bet acceptable), and being limited to only circulating trains between the Southern and Western Lines will not improve that situation.
Also, without the eastern link Grafton may as well be closed entirely since it’ll be completely divorced from any kind of frequent service. If we couple that with staging stations we’ll cut off the entire hospital/Grafton catchment from rail services. If the eastern link is not constructed immediately (which quite possibly means it’ll never be built) it must be a decision made in conjunction with a commitment to build the Symonds St station before the CRL is opened.
“The problem is that due to the grades it will likely make it almost impossible to grade separate the junction between the eastern and western link”
This doesn’t make any sense, as grade separation requires lower levels, not higher. Ground level, and two levels below that. The gradient of the tunnel is not only not an issue, but in fact it’s greatly eased by grade separation.
The only potential issue is that it means more volcanic rock to excavate.
Yes that occurred to me too, but not being a rail buff wondered if I was missing something… Surely all that is needed is a height difference, above or below makes no odds…? As well as more square metres.
The ‘all the way back to Grafton’ line above is not credible either surely. But by going lower ealier would help remove the level crossing east of the current Mt Eden station, no?
I think impossible is the wrong word, rather it could be very expensive. At a guess you might end up having to tunnel one link all the way to Grafton station, or even lower Grafton or relocate it to get the length needed.
Possible, but expensive. Would be interesting to see a cost-benefit evaluation on that if it means the difference between say 40 trains an hour and 60, as has been suggested.
What’s needed is analysis of three options isn’t it?
1. No Eastern Link
2. At Grade Eastern Link
3. Grade Separate Eastern link
with possible running patterns, for now and the decades ahead, costs, limitations and future implications….
The APB&B concept design for the CBD rail Link had grade separated junctions of both the Western and Eastern junctions at Mount Eden.
I don’t see any need for a tunnel back to Grafton either. It would be something like this:
Geoff the links in your diagram are ridiculously short. As noted above Newton station is 40m underground. The trains can manage a grade of about 1 in 33, so for a track to climb one metre in height needs a ramp 33m long. If you do the math on that then you’re looking at at least one link going several hundred meters further. Again it’s not impossible by any means, but not cheap and easy either.
What is ridiculously short? The yellow lines are what is already proposed. The links I have added (on two lower levels) would only require a 7 metre height difference with each other level (I.e., top level, then 7m down, then 14m down).
At your suggested 1 in 33, that’s only a distance of 230m for the middle level to reach the surface, and 460m for the lower level. My drawing is actually more generous than that.
The 40m depth of Newton station makes the middle and lower tunnels very easy to build. Of the four tunnels, the steepest would be the two that are actually proposed currently. My additional two tunnels would be on lesser grades.
Just to clarify, I envision Newton station to be on the two track section north of the various tunnels.
As Nick says, the Newton/Symonds St station will be 40m deep and part of the reason for that is if it was any higher then the tunnel wouldn’t get under the motorway with enough clearance to stop it being damaged (we don’t want the ground level above the tunnel dropping do we). There is simply no way that the trains could get back to surface level in the space you suggest
I knew it! Shut down the motorway and we’ll be sweet.
The red and magenta grade separated links are both about 1800m long.
At 3.5% this could give an elevation change of 63m.
Newton station is 46m above sea level, the connection points could both therefore be up to 109m above sea level.
As drawn, they would be far less than this, meaning the grade could be eased from 3.5% and/or the links shortened, and Newton station could be single level.
That should have read, each link woul dbe 900m long, elevation change 31m, total 77m above sea level. Still no problem.
Geoff’s idea of an upper and lower tunnel is a good one. To have adequate grades, Newton station would need to have a lower (outbound) and upper (inbound) platform, with the eastern and western tracks diverging shortly after the station. As a 2-level station, transfers would need to be by lift or escalator. (If someone wants a cross-platform connection, they can wait until K Road.) There is one such station in Los Angeles (Wilshire/Vermont) where the Red and Purple lines diverge. The diverging sections are in bored tunnel, so there is no need for an expensive box.
There is scope for substantial cost reductions with a 2-level station that has only one platform face at each level, because access to the station could be by several large lifts that go directly to street level. This design is not possible with an island platform located below the street. The station would then be like a shopfront with lifts going directly to platform level. There is then no need for inclined tunnels, escalators, nor pedestrian forecourts beneath the street.
The fact that no one has being able to come up with a creditable operating pattern for the network post the CRL being built suggest to me that the project is basically flawed and should not proceed. Consideration should be given to an idea that has being advanced in the past. The line from Newmarket to Britomart could branch off just past the proposed Parnell station and then proceed across the gully and tunnel under the university to the site of the proposed Aotea station from there it could proceed to Britomart. So you have still got your loop but you dont have the awkward junctions at Mount Eden which leaves Grafton and Newmarket out on a limb. Obviously no Kroad or Newton stations but it sounds as though we cant afford them or the Eastern entry to the tunnel.
There are lots of credible operating patterns so I don’t know where you get that idea from. One of the biggest benefits of the CRL is that it speeds up the travel time from the west considerably so doing what you suggest would not fix that. Also while the stations may be delayed, that isn’t the same as not even allowing them to be built at all.
If that was built, wouldn’t it mean quad tracking Newmarket-Parnell, including the tunnel under the domain?
The section of line from Kaiwharawhara to Tawa Tunnel number 1 is only 1:130, so it a lot longer thn what could be achieved with 1:28…
The Kaiwharawhara – No1 Tunnel gradient was made deliberately shallow to allow freight trains of up to 2400 Tons to ascend this incline. This is in no way representative of gradients which might be found on a passenger-only route such as the CRL. There is absolutely no impediment to grade-separating the eastern junction in terms of achieving acceptable gradients.
I can’t understand the suggestion that there should not be an eastern link at Mt. Eden.
Whilst many trains would be through trains heading from say south to west or north via Britomart a large number would as they do in Melbourne, come in from south go clockwise or anticlockwise through the tunnel and back out south exiting the loop at Newmarket. The same would apply for western trains and north when built.
Is this not common sense or have I missed something? No eastern link would be as daft or worse than the crazy layout at Newmarket for western trains
“…a large number would as they do in Melbourne, come in from south go clockwise or anticlockwise through the tunnel and back out south exiting the loop at Newmarket.”
Should avoid doing a Melbourne with the CRL at all costs. Hugely complicated, confusing and inefficient – an example of what not to rather than something to copy. Simple through routing is the way to go …
Through routing would mean making one or two transfers to get to your station. For example, GI to Parnell would involve a transfer to Newmarket via Mt Eden, then another transfer to Parnell.
Much easier to have all trains do a loop, acting as distributors/collectors for the entire CBD area. It’s not complicated at all, it’s probably the simplist way of doing things.
Huh? GI to Parnell: GI to Britomart, Britomart to Parnell, why would you go to Newmarket? All lines go both ways…..One transfer…. through routing over looping in my book
In a grade separation surely the east to north track would pass under the north to west tracks, since it gives the option to steepen the grade for downhill trains.
The diagram, though not quite clear, seems to show an eastern chord of 945m at a grade of 1:28, thus a fall of about 34m.
For the east to north track to fall 7m more to allow clearance, it would have to start diving 28×7 thus about 200m further east than whatever endpoint of the eastern chord the diagram refers to. This would presumably be a trench on the south side of the line from Grafton – not a big deal surely.
The trench could be shorter if a steeper grade was allowed for downhill trains.
North of the flyover point the two northbound tracks would then need to come to the same level before the Newton platform (assuming a simple island platform design). With the lower track rising** and the upper track falling at the ruling grade, this would require a run of 100m, to which add about 50m for the turnout and possibly something for transitions at the change of grade.
** I realise this messes up the figures I have given just above, but you get the general idea.