Here is the article in question:

http://www.smh.com.au/nsw/all-change-new-plan-for-faster-city-trains-20120614-20d5j.html

Clearly I was referring to increased infrastructure at Cabramatta.  An alternative would be a quad (perhaps triple) from Liverpool to Cabramatta.

Oh, and Campbelltown via Airport AND via Sydenham isn't very logical, but very Cityrail.

Sorry what increased infrastructure at cabramatta. secondly you can't quad anymore because the ssfl is going through so you can't put another 2 lines in that space.

The Campbelltown train via sydenham is an express service during peak hour only. all other east hills trains go via the airport

Oh, and Campbelltown via Airport AND via Sydenham isn't very logical, but very Cityrail.

You and I can upset the not-quite-as-left-as-us members of this forum by blaming the previous Lib government for that one .

And incidentally, that is one of the issues the metro cross harbour tunnel will/might address.

well it was done in the shadow of winning the right to host the olympics. even still it is paying dividends now with the removal of the surcharge on all the non airport stations and a seperation of the east hills line from the south coast line. maybe green square could become more of a transport interchange between trains and buses.

Sorry what increased infrastructure at cabramatta. secondly you can't quad anymore because the ssfl is going through so you can't put another 2 lines in that space.

Additional infrastructure at Cabramatta to be built? It doesn't exist - it's an idea. While you cannot quad through there, it can be tripled. One possibility is to reconfigure the entire station to be a three-track double-island with centre bi-directional reversing track, which allows trains to terminate in either direction without blocking the main line and without generating a conflicting move. I think there could be enough space to fit it in, but might require some expansion of the station, including pushing the bus stops one lane further into the road.

The Campbelltown train via sydenham is an express service during peak hour only. all other east hills trains go via the airport

And because they've blown all that money on K2RQ, it only makes sense to use it full-time to maximise usage of it. It also allows a separation of the fast Campbelltowns and the stopping Revesbys.

Also, the reason I don't like buses is because I don't fit in them. the seat spacings are that small that I take 2 seats up. I have to sit uncomfortably across 2 seats. Watson knows this but now I am telling you all. Plus I hate the bus drivers because they are rude. The ride is horribly bouncy which hurts my back.

Did I forget anything watson?

Yes, you forgot that you can sit on the wheelchair bay tip-up seats which have much more legroom - and no, milkshakes are not permitted cargo.

I find that internationally most heavy rail systems serving a distance similar to Rouse Hill use SD.

Examples include Hong Kong MTR East Rail Line, Seoul Subway Line 1 in Asia, BART and Washington Metro in US, London Crossrail and Berlin S-Bahn in Europe.
Not to mention that RER has a portion of SD trains as well.

That list is a bit selective. All these lines are technically Metro except CrossRail, and that has been built to take existing lines BR loading gauge lines into the centre of London. The key characteristic that makes them single deck is high frequency fast loading, not distance - in fact most Metros (US excepted) are short haul. I could have chosen many cities in Canada, Finland, France, Germany, India, Italy, Japan, Sweden, Switzerland or the United States. Even Hong Kong if you care to look.

In any case single or double deck is not the issue. It is more about the tradeoff between loading time and comfort.

You're right. In any case, those American systems cram in maximum seating, but some are already cannibalising that space for standing room.

Anyway, I really, really like djf01's staged diagrams, and agree they look a lot better than what we have right now.

actually you can't even run a triple island at cabramatta because the ssfl takes all the space on one side meaning the bus interchange takes all the space on the other side. there is no point in doing cabramatta up because liverpool has all the infrastructure there already and is the big city in the southwest.

So cabramatta is a waste of time even foaming about.

They must be planning to do something at Cabramatta as all the Single Deck plans (even Labors previous efforts) seem to end here rather then Liverpool, why I could never quite understand (Liverpool being the main centre for the SW region).

I guess it is possible they plan to build another station with a turnback north of the existing station,  before the South and Bankstown lines merge. There appears to be enough land there to do so.

They must be planning to do something at Cabramatta as all the Single Deck plans (even Labors previous efforts) seem to end here rather then Liverpool, why I could never quite understand (Liverpool being the main centre for the SW region).

I guess it is possible they plan to build another station with a turnback north of the existing station,  before the South and Bankstown lines merge. There appears to be enough land there to do so.

running the singles along the bankstown line is stupid. the south/bankstown line has 3 branches and not 1 dedicated line into the city.
once the line comes across the harbour into the city the sd's should head out along the eastern suburbs on a new line to la perouse where such a service is needed.

I have no idea why these plans end at cabramatta. there is space but not that much space north of the station and no practical use of the space is possible other then maybe an upgraded power station at the junction.

For the North you have no more North Sydney, Lindfield or Gordon starters with the trains either starting at Hornsby (northern line and north shore) or Berowra.

Good stuff djf01, but I understood that the Northern line was to run on the mains to Central only (Hornsby-Epping-Strath-Central) to free up the Suburbans for the west.  I'm not certain how the throat into the terminal will cope - it seems to be a problem just for the 'urbans.  On reflection maybe they can't do that until the NW kicks in to load the ECRL.

For the North you have no more North Sydney, Lindfield or Gordon starters with the trains either starting at Hornsby (northern line and north shore) or Berowra.

Good stuff djf01, but I understood that the Northern line was to run on the mains to Central only (Hornsby-Epping-Strath-Central) to free up the Suburbans for the west. I'm not certain how the throat into the terminal will cope - it seems to be a problem just for the 'urbans. On reflection maybe they can't do that until the NW kicks in to load the ECRL.

I don't think they can reasonably run 20tph from Hornsby or Berowa to Penrith on a single track pair.  There are more network issues than just the Richmond and Northern lines.  Mountains trains for starters.  Turning 20tph at a single site is a problem too.  The archilies heel of CityRail has always been the long boarding times, and a scheme that doubles the number of boardings for the same number of journey's doesn't sounds like a sensible approach to me, prima face at any rate.

Consequently I think they'll still branch main north trains at Strathfield, then refill the vacated slots with interurbans, flat junction not withstanding.

running the singles along the bankstown line is stupid. the south/bankstown line has 3 branches and not 1 dedicated line into the city.
once the line comes across the harbour into the city the sd's should head out along the eastern suburbs on a new line to la perouse where such a service is needed.

I have no idea why these plans end at cabramatta. there is space but not that much space north of the station and no practical use of the space is possible other then maybe an upgraded power station at the junction.

This will no doubt be ripped apart but IMO with the desire to also improve Cityrail as the claimed reason for integration of the DD and SD plans, there is only two options for SD on the southern side of the city without building new lines, that is the Bankstown/ Hurstville line as chosen or the the Airport line to Revesby, I think either has benefits. The SD mode is good for the airport line due to the type of patronage (space for luggage) but at least first up you will see the same number of services through the airport under each plan (20 tph) so past that I think the chosen alignment swings towards the choice they made, it allows relief on the Hurstville/ illawarra/ south Coast lines and the Bankstown (RP and Liverpool)/ south lines, SD deals well with the many stations along these lines (25 Bankstown plus 12 Hurstville vs 15 on the airport line), also the SD trains are meant to be at least 10% faster then DD so with the extra stations and their spacing it will improve travel times for both the Hurstville and Bankstown lines.

The Bankstown and Hurstville lines now have a direct connection to all the employment centres on the other side of the bridge too which is a good thing. I would imagine in time there is also the ability to run extra services to the Bankstown and Hurstville lines in time with growth since the Metro line is only running at 20 tph first up and it should be able to run at least 30 tph. This should see some good re-development along these corridors with much needed higher density housing (as will happen on the North Shore and North West corridors). Separating out the Bankstown line into its own sector also simplifies Sector 2 which is a long term aim of Cityrail. There are probably other things i've forgotten.

I don't think they can reasonably run 20tph from Hornsby or Berowa to Penrith on a single track pair.  There are more network issues than just the Richmond and Northern lines.  Mountains trains for starters.  Turning 20tph at a single site is a problem too.  The archilies heel of CityRail has always been the long boarding times, and a scheme that doubles the number of boardings for the same number of journey's doesn't sounds like a sensible approach to me, prima face at any rate.

Consequently I think they'll still branch main north trains at Strathfield, then refill the vacated slots with interurbans, flat junction not withstanding.

Yeah, I was the one that raised the simplification of the stopping patterns on the North Shore line to match those proposed for the Western line as you have the same problem (or worse) with people waiting at CBD stations for 3 or 4 trains as the North Shore line has so many stopping patterns in peak, e.g, upper North shore folks waiting for the North Sydney terminator, Northern line and Central Coast trains to go before they can get on a train home. However it wasn't mentioned directly in the article past the Northern line changes. It was raised earlier in the thread about whether you could actually turn enough trains at Hornsby or Emu Plains to run the 20 tph in peak. I said I didnt know but then suggested a split between Hornsby and Berowra and St Marys and Emu Plains. I believe that with the WEX plan they were looking to build a turnback at St Marys so I had wondered if the same would occur with this plan. As grog pointed out you would need to build a flyover to deal with the Mountains trains. 

However if you dont make the changes to the Northern/ North Shore/ Central Coast and Richmond lines then you still have the same issues as now. How do you go about achieving/ maintaining the 20 tph through the Western and North Shore lines and the platform crowding? It seems unlikely to me but I may be wrong.

actually you can't even run a triple island at cabramatta because the ssfl takes all the space on one side meaning the bus interchange takes all the space on the other side. there is no point in doing cabramatta up because liverpool has all the infrastructure there already and is the big city in the southwest.

So cabramatta is a waste of time even foaming about.

If they really wanted to do something, they could make Cabramatta a double-deck island and build viaducts to link the thing. The main reason why we're discussing this is because Cabramatta is where two track pairs merge into one. Not only is it at a flat double junction, these two track pairs are going to be pretty busy - thus making Cabramatta a prime point for congestion.

Also, I was suggesting eating the current interchange's space and shoving it out into the road, space to be obtained by stripping the other side of the road of whatever it has there.

For conversion to metro, I would argue that the Revesby via Airport line is if anything a better choice than the Bankstown line. It has more use for frequency, and gives the airport shiny trains to show off for the first time.

The metro should be able to run up to at least 30tph, and if played right could hit as high as 36tph.

The North Shore should see a simplification - but isn't it already all-stops-only during peak to Chatswood?

running the singles along the bankstown line is stupid. the south/bankstown line has 3 branches and not 1 dedicated line into the city.
once the line comes across the harbour into the city the sd's should head out along the eastern suburbs on a new line to la perouse where such a service is needed.

I have no idea why these plans end at cabramatta. there is space but not that much space north of the station and no practical use of the space is possible other then maybe an upgraded power station at the junction.

This will no doubt be ripped apart but IMO with the desire to also improve Cityrail as the claimed reason for integration of the DD and SD plans, there is only two options for SD on the southern side of the city without building new lines, that is the Bankstown/ Hurstville line as chosen or the the Airport line to Revesby, I think either has benefits. The SD mode is good for the airport line due to the type of patronage (space for luggage) but at least first up you will see the same number of services through the airport under each plan (20 tph) so past that I think the chosen alignment swings towards the choice they made, it allows relief on the Hurstville/ illawarra/ south Coast lines and the Bankstown (RP and Liverpool)/ south lines, SD deals well with the many stations along these lines (25 Bankstown plus 12 Hurstville vs 15 on the airport line), also the SD trains are meant to be at least 10% faster then DD so with the extra stations and their spacing it will improve travel times for both the Hurstville and Bankstown lines.

The Bankstown and Hurstville lines now have a direct connection to all the employment centres on the other side of the bridge too which is a good thing. I would imagine in time there is also the ability to run extra services to the Bankstown and Hurstville lines in time with growth since the Metro line is only running at 20 tph first up and it should be able to run at least 30 tph. This should see some good re-development along these corridors with much needed higher density housing (as will happen on the North Shore and North West corridors). Separating out the Bankstown line into its own sector also simplifies Sector 2 which is a long term aim of Cityrail. There are probably other things i've forgotten.

If the nwrl was still DD I would have no problem. But the reason for doing this single deck is to seperate it from cityrail. Running the single deck to bankstown or hustville means that it must replace all services on these lines to remain a seperate entity.

I have been thinking quite a bit about how to handle the Blue Mountains trains as these are the main complication with the 1 branch strategy for the Western to North Shore line. These are the things I would like to achieve:

• Single stopping pattern from Penrith - Upper Shore on a single track pair with no flat junctions and no branching (but trains can short run in the interest of terminal capacity)


• No gaps in 3 minute frequency through CBD

The problem was that even with a flyover at Homebush for 4TPH to run from the Mountains onto the Mains into Sydney Terminal, you still create a service gap through the CBD, and if you use the Northern line trains to fill that service gap, you are then branching and re-introducing some of your platform crowding problems in the CBD.

The compromise I have come up with will keep some in the mountains happy with a direct service, but not everyone will keep their direct service to the CBD. The plan is:

• Run 4tph Springwood to Hornsby via CBD using OSCARs, operated by Sydney Trains


• Run 4tph peak/1tph off peak (or whatever the required frequency is) from Lithgow/Mount Victoria/Katoomba terminating at Penrith with cross platform interchange to CBD service commencing from there, operated by NSW Trains. These trains could skip stations between Springwood and Penrith, saving 4 minutes and clawing back some of the interchange penalty.

This means that lower mountains and Emu Plains residents get an all day 4tph service direct to CBD, relieving some of the terminal capacity at Penrith, but upper mountains residents would need to change at either Springwood or Penrith. An additional platform would need to be built to terminate the mountains trains.

Not a great plan, but the only other option I can think of is just running the mountains trains all the way through the CBD to Hornsby or Wyong and the trains just won't be suited to that run.

For Sector 3, I came up with the following patterns based off djf01's extrapolation map for 2013, but I think these need to be proofed for problem areas, feasibility and investigated for coupling. The number in brackets is the tph value.

Out via Wynyard

- Hornsby via Gordon (4)
- Hornsby via Macquarie Park (4)
- Berowra via Gordon (4)
- Gordon via Chatswood (4)
- Epping via Macquarie Park (4)

Out via Redfern (Suburbans)

- Epping via Strathfield (6)
- Richmond via Parramatta (4)
- Penrith via Parramatta (10)

Out via Redfern (Mains)

- Hornsby via Strathfield (4)
- Newcastle & Central Coast (4)
- Blue Mountains (4)

Would all Western trains will run express to Granville at least, leaving whatever's on the Locals to run stopping? I expect the lower North Shore to run all stops on all services, unless the North Sydney reversers are retained to leave gaps, but is it really worth it?

I have been thinking quite a bit about how to handle the Blue Mountains trains as these are the main complication with the 1 branch strategy for the Western to North Shore line. These are the things I would like to achieve:

• Single stopping pattern from Penrith - Upper Shore on a single track pair with no flat junctions and no branching (but trains can short run in the interest of terminal capacity)


• No gaps in 3 minute frequency through CBD

The problem was that even with a flyover at Homebush for 4TPH to run from the Mountains onto the Mains into Sydney Terminal, you still create a service gap through the CBD, and if you use the Northern line trains to fill that service gap, you are then branching and re-introducing some of your platform crowding problems in the CBD.

The compromise I have come up with will keep some in the mountains happy with a direct service, but not everyone will keep their direct service to the CBD. The plan is:

• Run 4tph Springwood to Hornsby via CBD using OSCARs, operated by Sydney Trains


• Run 4tph peak/1tph off peak (or whatever the required frequency is) from Lithgow/Mount Victoria/Katoomba terminating at Penrith with cross platform interchange to CBD service commencing from there, operated by NSW Trains. These trains could skip stations between Springwood and Penrith, saving 4 minutes and clawing back some of the interchange penalty.

This means that lower mountains and Emu Plains residents get an all day 4tph service direct to CBD, relieving some of the terminal capacity at Penrith, but upper mountains residents would need to change at either Springwood or Penrith. An additional platform would need to be built to terminate the mountains trains.

Not a great plan, but the only other option I can think of is just running the mountains trains all the way through the CBD to Hornsby or Wyong and the trains just won't be suited to that run.

Might as well just run 4tph up the Main North .

Seriously, even if there is branching toward the end of the dog bone, it still largely solves the problem.  At the moment, 70-80% of PAX need to wait for "their" train, one of the 5 stopping patterns.  If you cut it to 2 or 3 patterns, and most of those still service all the main stations, then 70-80% of PAX can get any train.  The problem isn't solved, but it's largely reduced.

Also, I think there is a lack of critical thinking as to exactly what problem we are trying to solve here.  "Congenstion" comes in many guises.  A single stopping pattern reduces queing.  But at the expense of requiring more boardings across the network.  Boardings are the key driver of dwells and dwell variance.  Queing is a problem only to the extent that a large queue on a small platform like those at Town Hall has the potential to disrupt/slow boardings.  A single stopping pattern doesn't do anything to reduce boardings (in fact it increases them at interchange stations).   If you look at the capital and ongoing operational cost of running all those trains so far on Sector 3 (there will inevitibly a lot of unused capacity beyond the first interchange at Strathfield/Central, as you run partially filled trains on 90% odf sector 3, and other trains to haul changes to wherever they actually want to go), it will be a lot.  And the problem that this solves is to reduce the queue length at Town Hall and Wynyard (but increasing it at Central if main north trains terminate there).  Surely it would be cheaper just to attack this problem at source, and put in bigger/more platforms at source: at Town Hall.  Hundreds of millions, sure.  But compared with 40 new train sets and the ongoing cost of running them ....

There is a whole branch of mathematics devoted to queing theory, and one of the results of that that a single queue served by multiple servers (all trains go to the same destinations) will produce shorter queues and queing time than multiple queues for any one server.

One idea I've had in this regard is a new platform to the west of the Shore UP (difficult top build because of the need to relocate the main support pillars for the station - or else build station screening) allowing platform 3 at TH to be an arrivals platform for sector 3, and platforms 2 and 3&1/2 to be departures.

I think it is one of the strengths of the Sydney system is such a high portion of journey's can be done with a single trip.  We did in the 30s with our system what Paris did in the 60s/70s with the RER, and what Britain is doing now with their crossrails projects.  If you can dfistribute your disembarcation load across 2 or more platforms closer to journey's end, you increase capacity *and* reduce the need/cost of providing the secondary service.  This attacks the queing problem by minimising the number of required boardings, rather than trying to decrease delays to which boardings are achieved. 

Applying this philosophy to Sydney you run all interurbans as OSCARS through the city (capacity permitting of course, but that's achieved by running the main Nth on sector 2 and using the vacated mains for sector 3) and you eliminate the 10-15K boardings an hour required there from changing passengers, distributing this load across TH, Wynyard and Nth Sydney.  Not that I'm advocating this.

The conspiracy theorist in me says that if you increase the interchanges, you double number of trips required per actual journey, and so your "cost per trip" halves and it looks like you've increased efficinecy.  The fact that total end to end journeys delivered (and thus public benefit) remains the same while your costs increase is beside the point, the KPI on which you are judged is improved, and that's the main thing right? .

Hong Kong MTR East Rail Line

A fine example. It was originally a suburban railway, run with single-deck trains with plenty of transverse seating - quite comfortable. Then the development came, and three doors per car went to five, the seats went to the walls and suddenly the comfy train became a sardine can.

When you claim that SD trains with plenty of transverse seating is quite comfortable, why not use them on NWRL?

And for the record, East Rail is a line I've used a lot - it's good, but it's Hong Kong, not Australia. If passengers here are made to stand for an hour, they'll start moving back into their cars - which defeats the entire purpose of the line.

No passengers on the SD train will need to stand for an hour.

With the ultra-low density in the Northwest, I think everyone boarding the inbound train before Epping is guaranteed a seat.

Just because Asian commuters (including me) put up with long-distance sardine tinning doesn't mean Australian commuters will. Just because I've lived overseas and can and will stand from Causeway Bay to Sha Tin and make three connections on the way doesn't mean the average Sydney commuter will.

12tph SD trains will be clearly less "sardine" than 4tph DD!!!

I find that internationally most heavy rail systems serving a distance similar to Rouse Hill use SD.

Examples include Hong Kong MTR East Rail Line, Seoul Subway Line 1 in Asia, BART and Washington Metro in US, London Crossrail and Berlin S-Bahn in Europe.
Not to mention that RER has a portion of SD trains as well.

That list is a bit selective. All these lines are technically Metro except CrossRail, and that has been built to take existing lines BR loading gauge lines into the centre of London. The key characteristic that makes them single deck is high frequency fast loading, not distance - in fact most Metros (US excepted) are short haul. I could have chosen many cities in Canada, Finland, France, Germany, India, Italy, Japan, Sweden, Switzerland or the United States. Even Hong Kong if you care to look. In any case single or double deck is not the issue. It is more about the tradeoff between loading time and comfort.

NWRL is also technically Metro. That's why SD trains perfectly suit the purpose.
High frequency and fast loading is excellent.

Also, I don't think most Metros are short haul, especially those high-capacity subway systems.
Examples of high-capacity subway systems going well into the suburbs include Tokyo Subway, Seoul Subway, Beijing subway, Guangzhou metro, Shanghai metro, London Underground, etc.

Not to mention that long haul journeys can be well served by SD trains. As an extreme example, the proposed high speed railway between Beijing and Hong Kong will be operated by single deck CRH trains over a distance in excess of 1000km!

When you claim that SD trains with plenty of transverse seating is quite comfortable, why not use them on NWRL?

Because those trains can go from fairly comfortable to a sardine can. Yup - same physical rolling stock. KCR ripped out all the seats and even cut out an additional two doors per side.

No passengers on the SD train will need to stand for an hour.

With the ultra-low density in the Northwest, I think everyone boarding the inbound train before Epping is guaranteed a seat.

That's what you think. What happens when you're wrong, and people have to stand? Oh, they'll be happy.

In their cars.

12tph SD trains will be clearly less "sardine" than 4tph DD!!!

Let's just say the North West was due to get 8tph DD, as some of the Upper North Shore services would be diverted to balance the load.

Seemed to work.

When you claim that SD trains with plenty of transverse seating is quite comfortable, why not use them on NWRL?

Because those trains can go from fairly comfortable to a sardine can. Yup - same physical rolling stock. KCR ripped out all the seats and even cut out an additional two doors per side.

DD trains can get into a sardine can as well.
In fact, many AM peak services are ALREADY like that.

No passengers on the SD train will need to stand for an hour.

With the ultra-low density in the Northwest, I think everyone boarding the inbound train before Epping is guaranteed a seat.

That's what you think. What happens when you're wrong, and people have to stand? Oh, they'll be happy.

In their cars.
This is unlikely to happen in next couple of decades. 12tph SD clearly provides more than sufficient capacity.

Not to mention that plenty of passengers are ALREADY standing on other lines and they don't get back to cars.

12tph SD trains will be clearly less "sardine" than 4tph DD!!!

Let's just say the North West was due to get 8tph DD, as some of the Upper North Shore services would be diverted to balance the load.

Seemed to work.
No. Upper North Shore will be under-serviced and you are basically pushing passengers on Upper North Shore services into cars.

Not to mention that 12tph SD trains with transversal seating clearly have higher seating capacity than 8tph DD.

NWRL is also technically Metro. That's why SD trains perfectly suit the purpose.
High frequency and fast loading is excellent.

Err - no.  It it nothing except a proposed rail link yet.  That's what this debate is about.

Here is the article in question:

http://www.smh.com.au/nsw/all-change-new-plan-for-faster-city-trains-20120614-20d5j.html

Thanks for that Rails.  But I don't think that is the announced plan.  That was a bit of a hair brained scheme someone thought was getting too much traction, probably someone on the financial side of things who could see how much it would cost.

I think perhaps running 4tph Cumberland line trains to Richmond interlaced with 4tph "locals" from the city to Schofields would achieve largely the same thing.  It's still in the passenger's interest to get on the first western line train that shows up.  It doesn't matter if they change at Westmead, Seven Hills or Schofields to get to Richmond.  And in the case of PAX to between Blacktown and Schofields, they *could* wait for the next direct train, but if you just missed it the trailing Penrith train would most likely catch and pass the Schofields train and they could change to it at Seven Hills.  And for any of the other trains they would still get the Penrith trains and change to a Cumberland line train at Seven Hills. 

Unless of course the new fare structure penalised you for doing so .

DD trains can get into a sardine can as well.
In fact, many AM peak services are ALREADY like that.

But retain high seating levels, which don't seem to be going anywhere. It's very tempting to rip out seating in SD stock because there are real benefits to doing so for the operator.

This is unlikely to happen in next couple of decades. 12tph SD clearly provides more than sufficient capacity.

Not to mention that plenty of passengers are ALREADY standing on other lines and they don't get back to cars.

That's because they don't drive at the moment, unlike people in the Hills District.

No. Upper North Shore will be under-serviced and you are basically pushing passengers on Upper North Shore services into cars.

Not to mention that 12tph SD trains with transversal seating clearly have higher seating capacity than 8tph DD.

Run the numbers.

No. Upper North Shore will be under-serviced and you are basically pushing passengers on Upper North Shore services into cars.

Not to mention that 12tph SD trains with transversal seating clearly have higher seating capacity than 8tph DD.

Run the numbers.
The first version of SD rolling stock on Washington Metro had 82 seats per car.
Each set has a capacity of 82*8=656.
Then 12tph provides 656*12=7872 seats

As a comparison, 8tph A set has 896*8=7168 seats.

When trains get congested, the number of seats may be reduced in further train models.
However, I don't see the need to do so for the first couple of decades.

Also, it has the advantage of not crowding out the capacity on upper North Shore Line.

DD trains can get into a sardine can as well.
In fact, many AM peak services are ALREADY like that.

But retain high seating levels, which don't seem to be going anywhere. It's very tempting to rip out seating in SD stock because there are real benefits to doing so for the operator.

Nothing stops the operator from reducing seats in DD stock either!!!