True Nick – it is a balancing act given Britomart and Newmarket are the 2 biggest patronage stations, but the principle remains true, that junctions are where transfers should happen. You yourself reinforce this by noting that no station in Auckland (not even Britomart) has capacity for more than 6-car length, so beyond that you can only increase frequencies, which brings the bottleneck problem in.

In special cases, like Newmarket-Britomart, it is worth continuing all current routes (West, East & South) into Britomart because ARTA patronage figures show all 3 routes (dis)embark roughly 1/2 their passengers at Britomart. That is, route duplication (West & South trains) over the Newmarket-Britomart line, and triplication over the Britomart East entry tunnel (West, East & South trains) works (up to a point) because the duplicated sections are short enough to minimise delays compared to the inconvenience of passengers transferring for just traveling 1 more station.

But things are different when the network expands. Bring Onehunga and Manukau routes into Britomart and even your Mt Eden-Quay Park track doubling would not give enough capacity (5 bidirectional routes at 6tph gives 60tph required total). And that’s without any North Shore line, which the busway is slated to become if a Waitemata harbour tunnel is built.

An example; Prague – capital city of the Czech Republic and 1 million people, has 3 main subway lines that converge in a triangle at their CBD. There is no 1 station that receives all trains, and this avoids the bottlenecking.

I guess all I’m saying is a Western CBD rail tunnel will solve short-term growth, but not medium-term growth issues. And there are faster and cheaper short-term solutions; what about a 1 stage railbus (using rail fares/discount cards with revenue to rail operator) between Kingsland and Britomart via western CBD.

It would be well set up for Eden Park/RWC 2011, and faster for West rail passengers, especially if a 24-7 bus lane is provided. If it sucked enough Britomart patronage off the West trains, they could terminate at Newmarket, letting East & South (& West) trains to go to 5 min frequency. Just thinking out loud.

If that doesn’t swing then fall back on stage two, which is doing just the Quay St section and leaving the extension to the western line for a future date.

The last thing I would want to see is four tracks to Britomart killing the chance of a CBD tunnel, but the second to last thing I’d want to see would be a two track city tunnel feeding into the same Britomart platforms and throat tunnel.

Otherwise, the logical outcome is to not have any through running trains, but to terminate tunnel trains at platform 1 and 5 and Quay Park trains on 2,3 and 4 to maximise capacity.

The way I see it is that there are three main lines at the core of the network, the two new branches or any other routes or variations will simply feed into to these three lines (except well into the future if they build a harbour tunnel for rail, that would add a fourth main line).

A well managed twin track rail line can handle up to 18 trains per direction per hour, or one every 3 or 4 minutes. This means that the core network of three lines can handle 54 trains each way per hour. 54tph is enough to run nine routes with a train every ten minutes on each of them. With a CBD tunnel from the western line to Quay Park Junction, the central capacity would be three lines like the suburban capacity (i.e. three times what it is now, while the CBD tunnel to Britomart would only double capacity, less actually as trains coming in to terminate at Britomart would steal slots from those running through the tunnel).

However right now all three lines feed into just one line leading to Britomart, this is the bottleneck. So the whole network is limited to 18tph, or the equivalent of three lines running at ten minute headways. This obviously isn’t enough to run those five routes at a high level of frequency, so we are left with the main lines every 15 minutes and maybe a half hour service on the two branches. But doubling the central capacity would give more than enough to run those five routes at fast headways.

Onehunga rail should bring more patronage, but it will be a slow growth at 30min frequency (which is what they are planning to start with). And given it takes Veolia 5 months to make a new timetable (according to them), we won’t get rapid jumps in train frequency unless demand swarms like a horde of angry bees.

So the South trains should cope for all but the AM peak, and even then, the most crowded trains are the expresses, which don’t stop at Penrose anyway.

Your Britomart ideas are practical and a cheap, fast way of expanding capacity! But…. we still have the bottleneck problem if we insist all trains must run into Britomart. You say:
“…you end up with 50% more capacity in the end which is a level of capacity which is exactly matched to the capacity of the wider network…”
but we already have 3 routes, and 2 more (Onehunga & Manukau) planned for the next year or 2. So that makes 5 routes coming into Britomart…. We need to shift the train operators to a mentality that supports fast transfers at junctions. Sometimes – as you say – capacity is sufficient to justify an extra partially duplicated route, but mostly not.

I think the best solution is to start the CBD tunnel by building a new cut and cover link from Quay park to a pair of platforms under Quay St, connected to the existing station via the concourse level. This would double the capacity of Britomart fairly cheaply (I’m assuming it is cheaper to build a new link under the road corridor than somehow try to quadruplicate the entrance tunnel to Britomart). Then you have effectively three sets of double track coming in from the suburbs feeding into two sets of double track at the terminal, twice the current capacity.

Then when that approaches capacity, you continue the new link along the CBD tunnel route (with the proposed CBD stations) to the western line. Then you have three double track routes coming in from the sububurbs feeding into three sets of double track at the core.

By building a CBD tunnel in two stages and running it alongside Britomart rather than through it, you can stage the capacity expansion to meet growth (rather than having to do the whole tunnel to get anything), you could have the first capacity increase in 3-4 years rather than a decade, and you end up with 50% more capacity in the end which is a level of capacity which is exactly matched to the capacity of the wider network (not a central capacity which is only 2/3 of the wider network). It also allows you to purpose-design metro style platforms for the tunnel and leave the existing terminal station to do what it was designed for (rather than converting part of the terminal station into a through station so all the EMU tunnel trains and the terminating diesels have to jostle for space in the existing throat tunnel).

Jarbury – I agree we would prefer more than 2 tph frequency on any Onehunga or Airport line, but the point that ‘extending the length of a rail route does not increase the demand on Britomart’s bottleneck tunnel’ still holds true. The point being, that shifting the electrification cash to line extensions could bring in more cash and passengers, which gives more cash to grow/improve the network, …

Nick R – I spoke to Veolia’s GM (Akld) a few years back, and they were planning on a max of 6tph for East and South routes, and 4tph for West route. The bi-directional running signal improvements would allows them to squeeze in 2tph Onehunga, but they were scratching their heads over fitting in the proposed Manukau City-Britomart trains. So that is 18tph total, plus Manukau trains. Of course, West trains are likely to go to 6tph once double tracking is completed, so that puts more pressure on the Onehunga and Manukau train frequencies. Veolia & Ontrack were working on track block length to see if they could fit more trains in, I think.

That is one reason why I have suggested running Onehunga-Penrose and Manukau-Puhinui as shuttles rather than duplicate services between Penrose-Britomart and Puhinui-Britomart respectively. That, and alternating services confuses passengers (get on South train and go via Newmarket when they wanted to go to GI). But shuttles only work if they are timetabled to meet mainline services, and they stick to that!

A key related issue people often ignore is that Britomart entry tunnel is a bottleneck because 3 routes all run into it (West, South & East). If Onehunga trains run all the way to Britomart, then the whole Newmarket-Britomart section becomes a 3 route bottleneck (while the entry tunnel becomes a 4 route bottleneck!). Manukau-Britomart makes it a 5 route bottleneck in Britomart tunnel! Aaargh…

Maybe back to the dwarves…

Regarding amplification of the eastern approach – the worry is that if that was done then the CBD Rail Tunnel project would be put off for a few more decades. Perhaps if a start was made on the CBD Rail Tunnel and the first few hundred metres were used to stable trains, that could improve Britomart’s efficiency as not every train that comes in during the AM peak would have to leave again. Also, in the PM peak you could stack the tracks in the tunnel between 1pm and 4pm in preparation for the rush.