2018 SkyTrain Procurement

[Express691]

http://dailyhive.com/vancouver/translink-new-skytrain-cars-2024

As per the Mayors 10 year plan, the current order for 7 Mark 3s has been updated to...

14 trains 56 cars in 2018/2019
11 trains 44 cars in 2024

ALSO: Mark 1s start retiring in 2026

[Large Cat]

So to get things straight, all 100 of these cars will be acquired independently of the Broadway extension? There will be more cars coming when the Millenium Line gets longer, on top of these? If so, awesome. If not, nawesome.

[Express691]

1 hour ago, Large Cat said:

So to get things straight, all 100 of these cars will be acquired independently of the Broadway extension? There will be more cars coming when the Millenium Line gets longer, on top of these? If so, awesome. If not, nawesome.

as far as I know these are independent from the Broadway Extension. However I have no word if the procurement for the extension will even come with new cars.

[dpogue]

The last paragraph seems to indicate that the Broadway Extension will include a separate order of new cars:

Quote

The Phase Two plan, which includes the Millennium Line extension under Broadway from VCC-Clark Station to Arbutus Street, will contain a budget for an additional order of SkyTrain cars to support the service expansion and resulting systemwide ridership growth.

 

It will be interesting to see how they phase out the Mark I vehicles. Currently the original 114 cars from 1985/1986 are in refurbishment, but there are still 36 newer cars that have not been refurbished. Since we've already ordered the next batch of Mark IIIs, it seems logical to replace those unrefurbished 36 cars first, and then look at phasing out the older-but-refurbished original cars.

[cprted]

1 hour ago, dpogue said:

The last paragraph seems to indicate that the Broadway Extension will include a separate order of new cars:

 

It will be interesting to see how they phase out the Mark I vehicles. Currently the original 114 cars from 1985/1986 are in refurbishment, but there are still 36 newer cars that have not been refurbished. Since we've already ordered the next batch of Mark IIIs, it seems logical to replace those unrefurbished 36 cars first, and then look at phasing out the older-but-refurbished original cars.

Retirement of rolling stock has more to do with condition of the frame, body, heavy mechanical components. Refurbished or not, the original car bodies are still several (8-9?) years older than the second batch that are unrefurbished. 

Honestly I don't think the MkIs are going anywhere, at least not for another 10+ years.

[Dane]

Well, they just said nine years today so your ten year estimate is probably good lol

 

Great SkyTrain news today though - happy to see this acceleration of capital procurements. Particularly for Canada Line - such a shame so many corners were cut on that project. 

[maege]

48 minutes ago, Dane said:

Great SkyTrain news today though - happy to see this acceleration of capital procurements. Particularly for Canada Line - such a shame so many corners were cut on that project. 

One thing I haven't really understood is why 2 trains are being left unused during peak times, and is over a 50% increase in potential capacity needed for the Canada line right now?

On the one hand, yes, it can get quite crowded during peak times, so why are there 2 trains being held back? On the other, if there are still 2 trains left unused, ordering 11 more on top of the 20 that they currently have (only 18 of which are used during peak service) would increase theoretical capacity by 72% over what is currently used during peak (assuming all new trains and the 2 unused would all be used). Is that really necessary? Especially since the trains are intentionally being run slower than they can be, allowing for a virtually free way to increase capacity somewhat right now.

I'm not saying the new cars shouldn't be ordered, but there just seems to be a gap in logic from all the information that has been put out.

I agree about the cutting corners, especially with platform length and station design (and station placement in Richmond and downtown).

[GORDOOM]

12 minutes ago, maege said:

One thing I haven't really understood is why 2 trains are being left unused during peak times ...

1. Using all of the rolling stock for peak service leaves you absolutely no spare trains to account for breakdowns, etc. Any train that can't run for whatever reason will result in reduced service.
2. Using all of the rolling stock for peak service restricts your ability to do regular preventative maintenance on the trains, which will have a massive deleterious impact on their reliability and longevity.

Typically, you want to maintain about an 18% spare ratio, meaning that if you need (say) 100 trains for peak book-out, the entire fleet should consist of about 118 trains. Using 18/20 for peak service is an 11% spare ratio, which is extremely low to the point of being unsustainable in the long term.

16 minutes ago, maege said:

... and is over a 50% increase in potential capacity needed for the Canada line right now?

Given the extent of pass-ups on the Canada Line during rush hour, and the new ridership that will be generated by bus improvements, and the latent demand that exists because of the crowding on the Canada Line, I'd say it's badly needed.

[maege]

3 minutes ago, GORDOOM said:

1. Using all of the rolling stock for peak service leaves you absolutely no spare trains to account for breakdowns, etc. Any train that can't run for whatever reason will result in reduced service.
2. Using all of the rolling stock for peak service restricts your ability to do regular preventative maintenance on the trains, which will have a massive deleterious impact on their reliability and longevity.

Typically, you want to maintain about an 18% spare ratio, meaning that if you need (say) 100 trains for peak book-out, the entire fleet should consist of about 118 trains. Using 18/20 for peak service is an 11% spare ratio, which is extremely low to the point of being unsustainable in the long term.

That's fair. I was just thinking about Skytrain cars where everything not in refurbishment/maintenance is running at peak and has been for the last while. When there is more demand than capacity during peak, it would seem to make sense to run those extra trains, rather than having them sit. If maintenance is needed on a train, then pull it from service. The lack of one train would have a very minimal impact on wait times and wouldn't be noticeable to most people, other than being slightly more crowded. 

Of course over the long term (as you mention) this is not what you want, but until more trains arrive, why not? It would improve service for most people most of the time.

5 minutes ago, GORDOOM said:

Given the extent of pass-ups on the Canada Line during rush hour, and the new ridership that will be generated by bus improvements, and the latent demand that exists because of the crowding on the Canada Line, I'd say it's badly needed.

Hence part of the question as to why the don't run the extra 2 trains that they have, similar to Skytrain right now. And why don't they increase the train speed to what they are capable of running during peak? The combination of those two would provide a significant amount of extra capacity and relief. 

Thanks for pointing out the spares, the number of 11 more trains makes sense when spares are accounted for. 

[dpogue]

4 hours ago, maege said:

Hence part of the question as to why the don't run the extra 2 trains that they have, similar to Skytrain right now. And why don't they increase the train speed to what they are capable of running during peak? The combination of those two would provide a significant amount of extra capacity and relief. 

Thanks for pointing out the spares, the number of 11 more trains makes sense when spares are accounted for. 

I recall hearing that part of the Canada Line P3 contract also states how many trains they are allowed to run during rush hour, with monetary penalties for exceeding that number. The number of permitted trains increases at set intervals of years, so it should increase around the time that the new trains arrive (or else they're going to have to negotiate something to avoid paying penalties).

[maege]

2 hours ago, dpogue said:

I recall hearing that part of the Canada Line P3 contract also states how many trains they are allowed to run during rush hour, with monetary penalties for exceeding that number. The number of permitted trains increases at set intervals of years, so it should increase around the time that the new trains arrive (or else they're going to have to negotiate something to avoid paying penalties).

And that is why I hate P3 contracts for transit with a passion.

[Xtrazsteve]

In Ontario, they're using P3 for every new LRT line being built. For lines outside Toronto, they have service plans for the entire 30 years. How do they even know what ridership a line with have 20 years down the line? They could just have repeat the Canada Line ridership miscalculation. It only protects against capital overrun cost. 

[maege]

13 hours ago, Xtrazsteve said:

In Ontario, they're using P3 for every new LRT line being built. For lines outside Toronto, they have service plans for the entire 30 years. How do they even know what ridership a line with have 20 years down the line? They could just have repeat the Canada Line ridership miscalculation. It only protects against capital overrun cost. 

Of course there are statistical ridership projections, but as most other projections, it can be counted on that they will be wrong. The only question would be wrong by how much and in which direction. They can serve to provide a vague idea, but it is virtually impossible to get an accurate ridership projection. I don't think the Canada Line projections were necessarily bad (conservative estimates are usually not a bad problem to have), the fact that stations weren't designed to easily allow for extra growth, no matter if that growth was within 10 years or 30 is the issue.

P3's don't inherently protect against capital cost overruns. They definitely can be written up to do that, and often are, but it isn't an inherent trait. Similarly, regular design-build and similar contracts can be written to put cost overruns on the contractor(s), and they are sometimes. It all just depends on how it is written in the contract.

 

[nname]

On 8/18/2017 at 4:43 PM, GORDOOM said:

Typically, you want to maintain about an 18% spare ratio, meaning that if you need (say) 100 trains for peak book-out, the entire fleet should consist of about 118 trains. Using 18/20 for peak service is an 11% spare ratio, which is extremely low to the point of being unsustainable in the long term.

Not to be picky, but 18/20 is 10% spare...

And people complain how "unreliable" the Expo/Millennium line service is, but they don't realized that for much of the time before 2009, the MkIIs were running at a spare ratio of just 3%, and the MkIs had only a spare of 8% even though there were nearly 25 years old at that point!

[GORDOOM]

3 hours ago, nname said:

Not to be picky, but 18/20 is 10% spare...

No, it's actually (# spares) / (# vehicles required for service). So in this case, (2 spare sets) / (18 sets peak requirement) = about 11%.

3 hours ago, nname said:

And people complain how "unreliable" the Expo/Millennium line service is, but they don't realized that for much of the time before 2009, the MkIIs were running at a spare ratio of just 3%, and the MkIs had only a spare of 8% even though there were nearly 25 years old at that point!

[ThatBusGuy]

1 hour ago, GORDOOM said:

No, it's actually (# spares) / (# vehicles required for service). So in this case, (2 spare sets) / (18 sets peak requirement) = about 11%.

Could be looked at either way. The ratio of spares to total number of vehicles is 2 to 20, aka 10%, while the ratio of spares to required vehicles is 2 to 18, aka 11.1111...%

But like, ew, math 

[nname]

2 hours ago, GORDOOM said:

No, it's actually (# spares) / (# vehicles required for service). So in this case, (2 spare sets) / (18 sets peak requirement) = about 11%.

That's not how TransLink calculate it actually... If you see the old fleet order strategy inside this zip file, under 2007 column in table 4:

MkI: 68 sets used, 75 total, 7 spare... 7/75 = 9.33% spare ratio

MkII: 29 sets used, 30 total, 1 spare... 1/30 = 3.33% spare ratio

 

[Xtrazsteve]

11 hours ago, maege said:

Of course there are statistical ridership projections, but as most other projections, it can be counted on that they will be wrong. The only question would be wrong by how much and in which direction. They can serve to provide a vague idea, but it is virtually impossible to get an accurate ridership projection. I don't think the Canada Line projections were necessarily bad (conservative estimates are usually not a bad problem to have), the fact that stations weren't designed to easily allow for extra growth, no matter if that growth was within 10 years or 30 is the issue.

P3's don't inherently protect against capital cost overruns. They definitely can be written up to do that, and often are, but it isn't an inherent trait. Similarly, regular design-build and similar contracts can be written to put cost overruns on the contractor(s), and they are sometimes. It all just depends on how it is written in the contract.

 

Something like the Canada Line needs a better estimate than a typical LRT line. The cost to dig up a station is far more difficult than pouring more concrete on a surface stop. I think this needs to be taken in account with transit infrastructure design.

P3's do give taxpayers a bit more protection than a public agency or the government managing the project. I believe that the last two major projects in Vancouver, the Evergreen Line and Canada Line were done with a P3 as the BC government has little expertise in managing some complicated projects. A private project manager would be liable for overruns cost by mismanagement (bad timing and staging) as it's not the government's fault. Cost calculation errors done by the consortium would also be their fault. If a project is tender traditionally with the government as manager, they might have to assume some of those cost. P3's are less likely to mess up as they have more expertise, is in business for a profit and to earn a good reputation.

A second way a project could sustain overrun is by changing the project halfway. Not even a P3 would protect against this. Changing a project would be a lot easier if the project was tendered traditionally which make it easier to to accumulate more overruns especially when the political wind changes. 

The delay and overrun with the subway extension in Toronto is good indication that a P3 would have done better.  

[maege]

10 hours ago, Xtrazsteve said:

Something like the Canada Line needs a better estimate than a typical LRT line. The cost to dig up a station is far more difficult than pouring more concrete on a surface stop. I think this needs to be taken in account with transit infrastructure design.

Tunnelled lines should allow for easy expansion far beyond any estimated ridership in 30 years or whatever. Tunnels will be around for 100+ years, nobody will be able to even remotely project ridership levels in 100 years. When a tunnel is built, the stations should allow for easy expansion and lengthening, whether it is 15, 50, or 90 years from now. Allow extra room in the station area for additions and expansions when the time comes, because it will come. I don't necessarily have an issue with how long the platforms were made initially, but they should have allowed for easy lengthening when needed, and they didn't. So I somewhat agree with you, but take a slightly different perspective.

For the Canada line specifically, there was a lot of questioning and derision saying the ridership projections were far too optimistic. It probably wouldn't have been very beneficial or productive to have higher estimates as there likely would have been even more opposition saying they were completely unreliable and challenging the whole process and project as incompetent. Stupid politics.

13 hours ago, Xtrazsteve said:

P3's do give taxpayers a bit more protection than a public agency or the government managing the project. I believe that the last two major projects in Vancouver, the Evergreen Line and Canada Line were done with a P3 as the BC government has little expertise in managing some complicated projects. A private project manager would be liable for overruns cost by mismanagement (bad timing and staging) as it's not the government's fault. Cost calculation errors done by the consortium would also be their fault. If a project is tender traditionally with the government as manager, they might have to assume some of those cost. P3's are less likely to mess up as they have more expertise, is in business for a profit and to earn a good reputation.

Yep, Canada Line was a long-term P3 (DBFOM - Design, build, finance, operate, maintain) and the Evergreen a short-term P3 (DBF).

Two notes: 

I am not suggesting that Translink or others should build the lines themselves, but that a simpler fixed-cost contract with requirements for work and quality can achieve the same thing as a P3 (I view a project as a P3 when a private company finances at least some of the project). 

Second, I'm really not against P3's as a whole, but more precisely, that the operations and maintenance should not be part of the P3.

Lastly, I agree, politics should have very minimal to zero involvement in transit planning and construction. Politics and political motivation is what screws over transit agencies much of the time and makes them much less efficient. 

[ThatBusGuy]

Bumping this thread - the first set of new Skytrain cars are on their way from Ontario. Based on the image tweeted by Translink, the cars are numbered 429-432 so it looks like they'll be continuing the 400 series numbering with these ones.

https://buzzer.translink.ca/2018/09/new-skytrain-cars-are-on-their-way-to-vancouver/

[Express691]

9 hours ago, ThatBusGuy said:

Bumping this thread - the first set of new Skytrain cars are on their way from Ontario. Based on the image tweeted by Translink, the cars are numbered 429-432 so it looks like they'll be continuing the 400 series numbering with these ones.

https://buzzer.translink.ca/2018/09/new-skytrain-cars-are-on-their-way-to-vancouver/

Something's not gonna click if we get car FOURFOURFOUR

[anyfong]

Here are some of the cars seen in Regina:

 

[dover5949]

3 hours ago, anyfong said:

Here are some of the cars seen in Regina:

 

Arnold Brothers Transport is hauling the cars out. I'm surprised there is no pilot car following them for the long load?

[translink]

33 minutes ago, dover5949 said:

Arnold Brothers Transport is hauling the cars out. I'm surprised there is no pilot car following them for the long load?

It’s probably in front of the convoy. Every province has different rules for long loads. Some need pilot cars, some don’t.

[maege]

Saw a truck with one of these new cars on it while I was on the Expo line this evening!!! It was coming up Southridge Dr to the Edmonds OMC. Sadly no pictures