The Silly Argument Over BRT and Rail

[nfitz]

And because it's on the Internet it's true?

How can anyone possibly believe that you can get 160 passengers on an 18-metre vehicle. That would be like getting over 100 people on the current 12-metre vehicles.

At best, this would be an extreme crush load, that you'd only ever see if you spend a long-time loading the vehicle from completely empty. It doesn't allow for any movement on the vehicle as people get on and off. And it would assume that there are no wheelchairs or strollers.

More importantly, you can not use such extreme crush numbers for route capacity planning as you were trying to do. It would be technically incompetent for a professional to do so.

The error in your calculations has been made clear by a few people here, and yet you refuse to listen. You've calculated that to move 7000 people an hour you'd need to run 39 buses an hour, which would mean an average load of 179.5. This calculation is extremely flawed. First of all you CAN'T use the absolute crush load for such a calculation, because once you exceed the planning number the TTC uses (something like 100 for such a vehicle) the increased dwell times slow the average vehicle velocity and reduce route capacity. Second, the number you use appears to grossly exceed the crush capacity actually observed on TTC vehicles.

 

[gweed123]

It should be obvious that if there are documents stating 160 people in an 18.5 metre bus, that this is not reasonable, and well beyond crush loading. Even 115 would be pushing it.

And since when does one use crush loading for transport planning?

I agree. For every calculation I've done for BRT using artics I've always used 100 passengers as the load. Simple math gives you 4,000 pphpd at 90 second frequencies. You can push it up to 6,000 pphpd if you do 1 minute frequencies, but at that point I would think that the operating costs don't justify it. I've always stated that I think 4,500 is about the upper limit for a single BRT route. When you get into overlapping routes on a the same corridor, that's a different story though. You can increase the capacity of the corridor beyond 4,500 if you have local routes running every 15 minutes or so 'filling in the gaps' in the BRT schedule. It doesn't really add to the operating cost of the BRT route, because the local route would be running regardless of if the BRT was there.

And although howl's numbers were off, I think the meat of what he was trying to say holds true. There's a point at which BRT becomes viable, a point at which it is optimal, an area in which BRT and LRT overlap (where both would work), and a point where BRT is no longer the preferred choice. I do think it was a flaw for Transit City to not include BRT, as there are several routes in Transit City that would be much more suited to BRT compared to LRT. This would have freed up substantial money to build more of the network off the get-go.

 

[howl]

double

 

[howl]

And because it's on the Internet it's true?

How can anyone possibly believe that you can get 160 passengers on an 18-metre vehicle. That would be like getting over 100 people on the current 12-metre vehicles.

At best, this would be an extreme crush load, that you'd only ever see if you spend a long-time loading the vehicle from completely empty. It doesn't allow for any movement on the vehicle as people get on and off. And it would assume that there are no wheelchairs or strollers.

More importantly, you can not use such extreme crush numbers for route capacity planning as you were trying to do. It would be technically incompetent for a professional to do so.

The error in your calculations has been made clear by a few people here, and yet you refuse to listen. You've calculated that to move 7000 people an hour you'd need to run 39 buses an hour, which would mean an average load of 179.5. This calculation is extremely flawed. First of all you CAN'T use the absolute crush load for such a calculation, because once you exceed the planning number the TTC uses (something like 100 for such a vehicle) the increased dwell times slow the average vehicle velocity and reduce route capacity. Second, the number you use appears to grossly exceed the crush capacity actually observed on TTC vehicles.

Two points: First. There is absolutely NO error in that calculation and no one has pointed out anything of the sort. Your statement is an outright lie.

Second. At least one of those references is on the Volvo corporate website, what more do you want? Maybe you want me to buy us both plane tickets to Shanghai so we can count how many people are getting on and off each bus.

When you’re doing comparisons of this nature you always have to use the most aggressive numbers you can find. If I had of based my analysis on 120 people per bus BRT advocates would have jumped all over it saying there are plenty of examples of BRT with 160, 180 or even 200 people per vehicle and therefore your argument is invalid. When Transportation Planners are looking at a future system (something that will not be operation for five to ten years) they have to use the most aggressive numbers they can find on the assumption that if Shanghai and Santiago can do it, so can Toronto. To use today’s current standards and numbers would be professionally suspect.

My simple analysis shows that above 4,000 people per peak hour BRT becomes less attractive and above 7,000 people per peak hour it becomes next to impossible in a centre-of-street configuration. It becomes next to impossible at 10,000 people-per-hour in a dedicated right-of-way configuration. These are numbers that are generally consistent with real world systems (Ottawa is at about 10,000). You can complain about the details all you want, but it's not going to make a huge different in the resulting numbers.

 

[new.flyer.408]

The capacity of 160 on an 18 metre bus still seems bloated. Do interior photographs exist of those Volvo 7800 units? It seems like they'd be completely perimeter seating with storage/empty areas to increase standee numbers, like some spaces in the 7700.

To be fair, the Calgary report may be considering both articulated and bi-articulated buses.
[Edit] Yeah, reading further into the report, they cite North American articulated buses to have a capacity of 110, while multiply-articulated units can have up to 270 passengers. The articulated 'buses' they cite as having up to 200 passenger capacity are the electric trams (LRVs with tires).

North American specification 18m buses max their capacities around 112 (Nova LFS-A/LFX) and 100 (Van Hool AG300). I can't find capacity numbers for the New Flyers, but their seating capacity is 64 (and the Nova is 62) so their number would proably be around 112 as well.

I'm going to have to agree with nfitz on this - a Nova articulated bus (which seems to be the front-runner Canadian artic bus) for the TTC would probably be specced as to maximize the capacity to 120, if they employ some TR measures like the foldable seats. The TTC would probably like to maximize movement space too (which the Van Hool doesn't do very well at, riding VIVA every day).

 

[gweed123]

Two points: First. There is absolutely NO error in that calculation and no one has pointed out anything of the sort. Your statement is an outright lie.

Second. At least one of those references is on the Volvo corporate website, what more do you want? Maybe you want me to buy us both plane tickets to Shanghai so we can count how many people are getting on and off each bus.

When you’re doing comparisons of this nature you always have to use the most aggressive numbers you can find. If I had of based my analysis on 120 people per bus BRT advocates would have jumped all over it saying there are plenty of examples of BRT with 160, 180 or even 200 people per vehicle and therefore your argument is invalid. When Transportation Planners are looking at a future system (something that will not be operation for five to ten years) they have to use the most aggressive numbers they can find on the assumption that if Shanghai and Santiago can do it, so can Toronto. To use today’s current standards and numbers would be professionally suspect.

My simple analysis shows that above 4,000 people per peak hour BRT becomes less attractive and above 7,000 people per peak hour it becomes next to impossible in a centre-of-street configuration. It becomes next to impossible at 10,000 people-per-hour in a dedicated right-of-way configuration. These are numbers that are generally consistent with real world systems (Ottawa is at about 10,000). You can complain about the details all you want, but it's not going to make a huge different in the resulting numbers.

Ottawa gets ~10,000 pphpd by having 3 routes each with around 3,500 pphpd overlap going through downtown. It's the corridor that's carrying that capacity, not the actual route. The capacity of any one BRT route in North America is around 4,500 pphpd. The corridors themselves can physically handle more, because in most cases you have express buses, and multiple routes overlapping.

I've rode OC Transpo artics quite a bit, and you'd be hard pressed to pack in any more than 100 people on those buses. The most packed I've ever been on those buses was heading to and from downtown on Canada Day. There was literally no room to even move, because the bus was packed at Baseline, so the driver pretty much bypassed all the stations and went straight to downtown. If there were more than 100 people on that bus, it wasn't by much.

 

[howl]

Well regardless of the current typical North American standard there are no errors in the analysis. BRT's CAN work at that level. The only question is could Toronto ever have a BRT system as urban as Santiago or Shanghai? I would argue that in ten or fifteen years it could and therefore the numbers I have used are reasonable. If you choose to argue that Toronto could never have a system as aggressive as Santiago's or Shanghai’s have today that's fine, but that's a different argument than "the calculations are flawed".

It also doesn't change the gist of my analysis which was BRT's work fine for low demand routes and LRT's work better for higher demand routes. The only point of contention is where is the line between the two? Or rather how big is the grey area between the two where either may be feasible?

 

[rbt]

Ottawa gets ~10,000 pphpd by having 3 routes each with around 3,500 pphpd overlap going through downtown. It's the corridor that's carrying that capacity, not the actual route. The capacity of any one BRT route in North America is around 4,500 pphpd. The corridors themselves can physically handle more, because in most cases you have express buses, and multiple routes overlapping.

While true, OCTranspo also has between 4 to 8 mixed-traffic lanes (depending on how you count) through downtown. A BRT on Finch will be lucky to receive 2 dedicated lanes; in all likely hood it will be 2 mixed traffic (HOV) lanes for the majority of its length.

A full OCTranspo style BRT on Finch, including grade separated intersections, passing lanes at stations, stations with space for multiple buses to load/unload probably isn't going to happen; and as a result the capacity will be well below what Ottawa manages.

 

[nfitz]

Two points: First. There is absolutely NO error in that calculation and no one has pointed out anything of the sort.

??? There is very clearly an error. You've a) used an ultimate crush load in a capacity calculation for peak hour. This is an error. This has been pointed out to you by more than one person. Second, your ultimate crush load is erroneously high.

Your statement is an outright lie.

Look, just because you are wrong, doesn't mean you should be hostile. Just fix your calculations and move on. Oh, and apologize, for being so unnecessarily rude.

Second. At least one of those references is on the Volvo corporate website, what more do you want?

The number defies reality. Other sites use a much lower number for the same vehicle. Clearly the number is, for whatever reason, wrong or fabricated. It certainly isn't applicable to travel in North American metropolis. And even if it was possible to shoehorn that number of people on a bus, why would you be using an ultimate crush load for peak hour calculations? Such a fundamental flaw in the calculation should be blatantly obvious.

If you design a BRT route to handle 7,500 per hour, all your going to do is create failure, unless you can somehow design the route to use trains of 24-metre articulated buses.

 

[gweed123]

While true, OCTranspo also has between 4 to 8 mixed-traffic lanes (depending on how you count) through downtown. A BRT on Finch will be lucky to receive 2 dedicated lanes; in all likely hood it will be 2 mixed traffic (HOV) lanes for the majority of its length.

A full OCTranspo style BRT on Finch, including grade separated intersections, passing lanes at stations, stations with space for multiple buses to load/unload probably isn't going to happen; and as a result the capacity will be well below what Ottawa manages.

But there's the curbside bus lane section on Woodroffe that, based on calculations I posted on here about a month ago, is carrying just over 3,500 pphpd. There's the Transitway route, about 3 local routes, and a handful of express routes that use them, and they move without a problem. This is likely the type of setup that would occur along Finch. I don't think comparing it to downtown is a realistic comparison.

And there is no real "OC Transpo-style BRT", because the number of different configurations in runs in throughout the City is quite extensive. There's full grade-separated dedicated ROW, at-grade dedicated ROW, shoulder lanes on expressways, curbside lanes on arterials, and even mixed traffic. You pick whichever one is most appropriate for the corridor, given available funds, predicted ridership, and urban design.

 

[gweed123]

??? There is very clearly an error. You've a) used an ultimate crush load in a capacity calculation for peak hour. This is an error. This has been pointed out to you by more than one person. Second, your ultimate crush load is erroneously high.

I agree. Designing a system to barely work with crush load isn't the way to go about things. Yes, the buses should be nearly full, but not "face pressed up against the glass omg my hand is on someone's ass but I can't stop it because I can't move my hand" full.

The number defies reality. Other sites use a much lower number for the same vehicle. Clearly the number is, for whatever reason, wrong or fabricated. It certainly isn't applicable to travel in North American metropolis. And even if it was possible to shoehorn that number of people on a bus, why would you be using an ultimate crush load for peak hour calculations? Such a fundamental flaw in the calculation should be blatantly obvious.

If you design a BRT route to handle 7,500 per hour, all your going to do is create failure, unless you can somehow design the route to use trains of 24-metre articulated buses.

You COULD design a BRT that could carry 7,500 pphpd. However the corridor would need to have a number of express routes running it, with some pretty big stations. Once you get up to that point, you'd be better off going with grade-separated LRT, because you'd need grade-separated BRT to even hit those numbers. Side note: I'm pretty sure Tunney's Pasture station in Ottawa gets over 7,500 pphpd, however a decent percentage (maybe 20%, just a guess though) of that number is express buses that don't stop in the station.

And yes, for math purposes, use 100 as the expected load.

 

[howl]

??? There is very clearly an error. You've a) used an ultimate crush load in a capacity calculation for peak hour. This is an error. This has been pointed out to you by more than one person. Second, your ultimate crush load is erroneously high.

Look, just because you are wrong, doesn't mean you should be hostile. Just fix your calculations and move on. Oh, and apologize, for being so unnecessarily rude.

The number defies reality. Other sites use a much lower number for the same vehicle. Clearly the number is, for whatever reason, wrong or fabricated. It certainly isn't applicable to travel in North American metropolis. And even if it was possible to shoehorn that number of people on a bus, why would you be using an ultimate crush load for peak hour calculations? Such a fundamental flaw in the calculation should be blatantly obvious.

If you design a BRT route to handle 7,500 per hour, all your going to do is create failure, unless you can somehow design the route to use trains of 24-metre articulated buses.

Dude. Nowhere in my analysis does it state that it is North American specific. It is an analysis of what BRT and LRT can potentially achieve. I have provided evidence that the numbers are correct, and that they can and are being reached in other parts of the world.

Go back and look at the last 25 posts. I provided the analysis. You said you didn’t believe such a vehicle existed. I provided an example. You denied the example existed and wanted links. I provided links. You got all flustered and said my calculations had errors (without stating what the errors were). I provided further evidence that such numbers are achievable. Now you’re saying the error is it doesn’t work with current North American standards. . . . but that was never implied in the original analysis.

If you want a discussion about what is possible in North America today, you’re free to discuss that, but that doesn’t change my initial analysis. Those numbers show what is technically possible.

 

[nfitz]

You COULD design a BRT that could carry 7,500 pphpd. However the corridor would need to have a number of express routes running it, with some pretty big stations.

Yes, I guess you could ... you'd need overtaking lanes in stations ... like you see in Ottawa and Seattle - and complete grade separation ... or at least some pretty sophisticated traffic signals.

Side note: I'm pretty sure Tunney's Pasture station in Ottawa gets over 7,500 pphpd, however a decent percentage (maybe 20%, just a guess though) of that number is express buses that don't stop in the station.

You might well ... though both Ottawa (and Seattle) hit the point where they wanted to upgrade their BRT to LRT.

 

[nfitz]

Nowhere in my analysis does it state that it is North American specific.

???? In your original analysis, you took your 7,500 and 39 buses, and then said "f you are trying to move 10,000 people an hour at peak times you need 56 BRTs running one every minute or 25 LRTs running every 2.4 minutes. In a dedicated right-of-way configuration like Ottawa's this is the limit of the BRTs". This clearly implies you think that Ottawa is getting 180 people on each bus. I can assure you when I've ridden the transitway in Ottawa, that there is no where near 180 people on each bus - I'd be surprised if it is 100 - unless something has dramatically changed!

It is an analysis of what BRT and LRT can potentially achieve. I have provided evidence that the numbers are correct, and that they can and are being reached in other parts of the world.

The numbers are blatantly incorrect. They may be published, but there is no way that any agency is averaging this number of people on an 18-metre bus.

Those numbers show what is technically possible.

It's not technically possible using the vehicles and frequencies you've used. I'm not sure why you don't understand this.

 

[gweed123]

Yes, I guess you could ... you'd need overtaking lanes in stations ... like you see in Ottawa and Seattle - and complete grade separation ... or at least some pretty sophisticated traffic signals.

Building any type of dedicated ROW BRT station without having it be 4 lanes through the station is a waste of time. I can't think of a single station along the dedicated ROW sections of the Transitway that aren't 4 lanes through the station. But yes, you're right. The higher the predicted ridership, the more design elements you need to incorporate in order to make it work.

You might well ... though both Ottawa (and Seattle) hit the point where they wanted to upgrade their BRT to LRT.

And I personally think this is the way to go. Make a moderate investment in BRT, grow the ridership, and then upgrade key sections to LRT. And while Ottawa is upgrading to LRT, they will also be building 2x as many KMs of BRT in the next 20 years than they will LRT.