[roger1818]

Applying gravity model:
for Kitchener - Toronto CMAs
(575,000)*(6,202,000)/(93.78)^2 = 405 489 067.697

for Montreal - Toronto CMA's

(4,300,000)*(6,202,000)/(508.87 )^2 = 102 987 975.53

Montreal-Toronto has1/4 the theoretical demand of Kitchener -Toronto.

But curiously when ever someone talks about improvements for the Kitchener line - these are brushed away as a waste of money.

A user on UrbanToronto recently argued that there was no need for a larger Kitchener central station.... um if one thinks that Kit-Tor doesn't need capacity, then toronto - montreal is way down the priority list. Like If a Montreal- Toronto project came up I would support it but there are better ways to plan

The gravity model estimates overall demand, not just HSR demand. To get the potential demand for HSR, you need to compensate for the competition from other modes of travel. CityNerd does a good job of explaining this in his video Top 10 Places to Build High Speed Rail In the U.S. At about the 8:00 mark, he shows the following chart showing travel time vs. distance (I added conversion of the distances to kilometers). At a distance of only 94 km, HSR between Kitchener and Toronto CMAs is going to have difficulties competing with driving. Now traffic will play a role, but I would argue that KW-TO is in the range that regional rail would be more appropriate.

 

[kEiThZ]

At a distance of only 94 km, HSR between Kitchener and Toronto CMAs is going to have difficulties competing with driving. Now traffic will play a role, but I would argue that KW-TO is more of the range that regional rail would be more appropriate.

I'm not sure his model translates well here. His model does not account for traffic or higher fuel prices here. And traffic, cost of gas and parking are highly relevant in this example. At rush hour (8AM), from Kitchener GO to Union, it's 1:50 hrs to drive and 1:43 hrs on the train according to Google. GO RER is supposed to reduce travel times by 20%. That should take it down to 1:22 hrs. The proposed Ontario HSR would have cut that down to 48 mins. The biggest competition to any HSR proposal from Kitchener will then be GO RER. It will essentially be a trade-off between a longer ride and a higher fare with people falling along the spectrum. I suspect Toronto-Kitchener would still have similar ridership numbers to Toronto-Montreal.

 

[TRONto]

So I decided to investigate this. I looked at the schedules for Friday, May 26th 2023 on Air Canada, Westjet and Porter, for departures from YYZ and YTZ to YOW and YUL. Wanted to look far enough out to discount the current issues. I then took the scheduled aircraft and the seat counts from Wikipedia to get a rough idea of how many seats are offered on each city pair. Here's what the schedules look like:

(Excuse the formatting. I have no idea how to do up tables here.)

	AC	WS	PD
YYZ-YOW	17 x A223	5 x Q400
		1 x B738
YTZ-YOW	4 x Q400		12 x Q400
YYZ-YUL	1 x E175	5 x Q400
	15 x A223	1 x B73G
	1 x A321
	1 x A333
YTZ-YUL	14 x Q400		12 x Q400

Here's what the total seat counts look like on each city pair:

	AC	WS	PD
Toronto-Ottawa	2641	564	936
Toronto-Montreal	3710	524	936

If you look at the above, you'll notice that the airlines are scheduling about 25% more seats to Montreal than Ottawa. But it's specifically one carrier that is Montreal focused: Air Canada. They have over 52% more flights to Montreal and 40% more seats. But a large chunk (about a quarter) of the 40% more seats they send to Montreal is literally one aircraft: the A330-300 they are sending there. That's highly likely a repositioning flight that they want to make some money on. For both Westjet and Porter, Ottawa and Montreal are equally important in both the number of flights and the number of seats offered.

I also look at the number of departures on VIA from Union to Ottawa and Montreal. VIA is planning on sending 8 trains to Ottawa and 6 trains to Montreal that day. So even here, Ottawa is actually slightly more important than Montreal.

The idea that Montreal is so much more important as a destination for travelers from Toronto is a myth. It's on par with Ottawa. And the difference in seat counts between Montreal and Ottawa certainly doesn't line up with the population differential between those two cities. It's highly likely a large chunk of the flights and extra seats that Air Canada offers to Montreal is a function of them being a network carrier and needing to shuttle travelers back and forth between YYZ and YUL to make connections at those hubs. It's not entirely about origin-destination demand between Toronto and Montreal.



The routine whining here about Toronto-Montreal travel times certainly leaves that impression that there's people who think that way.



What's the problem with doing that? Nobody is suggesting that this is a primary goal of HFR. Broadly improving travel times is a goal. But from the beginning they've always had other goals that were paramount. Reliability, capacity and frequency are at the top of the list. A secondary benefit of efforts to improve those things is faster travel time. And nobody has suggested that this secondary benefit will be evenly applied.

I don't think you can use this airport comparison accurately as many MTL-YYZ trips are stopovers, they are not final destination.

 

[kEiThZ]

I don't think you can use this airport comparison accurately as many MTL-YYZ trips are stopovers, they are not final destination.

I mentioned that in my post. And it only makes demand to Montreal even lower than normally assumed. My point was to show that traveling to Ottawa is just as important as traveling to Montreal, from Toronto.

We need to get over the idea that because Montreal is so much larger than Ottawa-Gatineau (almost 3x in population) that it generates substantially more demand to and from Toronto. Maybe that was true decades ago when Ottawa was a lot smaller. Not today.

 

[roger1818]

I'm not sure his model translates well here. His model does not account for traffic or higher fuel prices here. And traffic, cost of gas and parking are highly relevant in this exactly. At rush hour (8AM), from Kitchener GO to Union, it's 1:50 hrs to drive and 1:43 hrs on the train according to Google. GO RER is supposed to reduce travel times by 20%. That should take it down to 1:22 hrs. The proposed Ontario HSR would have cut that down to 48 mins. The biggest competition to any HSR proposal from Kitchener will then be GO RER. It will essentially be a trade-off between a longer ride and a higher fare with people falling along the spectrum. I suspect Toronto-Kitchener would still similar ridership numbers to Toronto-Montreal.

Those certainly are valid points. At that distance though, I have to wonder how much all day demand there would be for HSR. There certainly would be some demand (especially in the morning and evening), but how much extra would people be willing to pay to save 36 min as compared to GO RER (using the travel times you quoted). Then there is the issue of wait time. I gather GO RER will operate every 30 minutes from/to Kitchener during peak periods, which is an average wait of 15 minutes. With hourly HSR, you end up with an average wait time of 30 minutes, which eats away at a significant portion of that 36 minutes of travel time savings. GO RER will be a huge improvement and go a long way to improve the situation. It will make it easier to see when demand reaches a point where HSR might make sense.

 

[kEiThZ]

I gather GO RER will operate every 30 minutes from/to Kitchener during peak periods, which is an average wait of 15 minutes. With hourly HSR, you end up with an average wait time of 30 minutes, which eats away at a significant portion of that 36 minutes of travel time savings. GO RER will be a huge improvement and go a long way to improve the situation. It will make it easier to see when demand reaches a point where HSR might make sense.

The OHSR proposal was basically more like regional rail. I believe they proposed 30 min headways. It was effectively acting as an express service overlaid on GO.

 

[SashaLemon]

No. But this is a whole different topic from your supposed prioritization.

Let me guess, if the feds announced tomorrow that Banff-Calgary-Edmonton and Toronto-Kitchener-London were their lead HSR projects, you'd be here saying that it's dumb because Toronto-Montreal should be first.

The shorter projects are attractive because they are far less risky. 200-300 km is the absolute sweet spot. Too close to have any competition from air. Too far too make driving regularly comfortable. Yet close enough that HSR could actually create an exurban commuter market. Nobody will ever commute between Toronto and Ottawa or Toronto and Montreal with HSR. But you will absolutely get commuters on Ottawa-Montreal and Toronto-London. Possibly even on Calgary-Edmonton and Montreal-Quebec City depending on speed. This is where speed returns a disproportionate return in ridership.

I will support any rail line in the current news. Build anything! That line between Calgary and Edmonton is also a painfully obvious project.

 

[Urban Sky]

You know/knew your philosophy teacher; I did not; so I shall not make assumptions on what they intended.

No offence taken, in the slightest! Sometimes we don’t explain in sufficient detail what we mean and sometimes we misinterpret what other people meant (see below)…

What are you SAYING? Am I a crazy person? I keep saying focus on Toronto - Ottawa - Montreal!

I honestly don’t know what went wrong here, but I clearly wasn’t responding to the point you were making. My apologies!

The gravity model estimates overall demand, not just HSR demand. To get the potential demand for HSR, you need to compensate for the competition from other modes of travel. CityNerd does a good job of explaining this in his video Top 10 Places to Build High Speed Rail In the U.S. At about the 8:00 mark, he shows the following chart showing travel time vs. distance (I added conversion of the distances to kilometers). At a distance of only 94 km, HSR between Kitchener and Toronto CMAs is going to have difficulties competing with driving. Now traffic will play a role, but I would argue that KW-TO is in the range that regional rail would be more appropriate.

View attachment 417758

You are correct, if you blindly use the gravity model for estimating the ridership potential for fast intercity rail services rather than for that of any passenger rail service, we may wrongly conclude that Toronto-Oshawa is the most viable HSR corridor in this country!

Another common misconception I see here in the debates about the results from the gravity model that the respective values of certain routes determine the (relative) ridership potential. However, no single O-D will justify the construction of a HFR/HSR segment; it‘s the sum of all O-Ds which overlap over the same segment, as shown in the middle part of the following table I already posted back in April:

Re-post from: Post 12,221

Even if the gravity model suggests that Montreal-Quebec might be the O-D with the largest ridership potential, there are no other ridership O-Ds which would add significant ridership potential onto that corridor, as Trois-Rivières is much smaller than QBEC and OTTW is almost twice as far from it and much smaller than MTRL. Similarly, the viability of Toronto-London suffers from the low ridership potential beyond Kitchener, as Kitchener is about one-tenth of the size of Toronto…

I will support any rail line in the current news. Build anything! That line between Calgary and Edmonton is also a painfully obvious project.

I‘m afraid it isn‘t as „painfully obvious“ as you (and many other passenger rail enthusiasts) believe, as its ridership potential is about one mesure of magnitude smaller than Montreal-Ottawa-Toronto, if you check in the table above. If we don’t focus on the most biable corridors, we won‘t get anything built…

 

[kalis0490]

I'm not sure his model translates well here. His model does not account for traffic or higher fuel prices here. And traffic, cost of gas and parking are highly relevant in this example. At rush hour (8AM), from Kitchener GO to Union, it's 1:50 hrs to drive and 1:43 hrs on the train according to Google. GO RER is supposed to reduce travel times by 20%. That should take it down to 1:22 hrs. The proposed Ontario HSR would have cut that down to 48 mins. The biggest competition to any HSR proposal from Kitchener will then be GO RER. It will essentially be a trade-off between a longer ride and a higher fare with people falling along the spectrum. I suspect Toronto-Kitchener would still have similar ridership numbers to Toronto-Montreal.

The model he uses is works for metro areas with little commuting population - ex london- Hamilton.

Not related but :There would a very high commuting ridership from Kitchener - guelph

 

[SaugeenJunction]

 

[SashaLemon]

No offence taken, in the slightest! Sometimes we don’t explain in sufficient detail what we mean and sometimes we misinterpret what other people meant (see below)…


I honestly don’t know what went wrong here, but I clearly wasn’t responding to the point you were making. My apologies!


You are correct, if you blindly use the gravity model for estimating the ridership potential for fast intercity rail services rather than for that of any passenger rail service, we may wrongly conclude that Toronto-Oshawa is the most viable HSR corridor in this country!

Another common misconception I see here in the debates about the results from the gravity model that the respective values of certain routes determine the (relative) ridership potential. However, no single O-D will justify the construction of a HFR/HSR segment; it‘s the sum of all O-Ds which overlap over the same segment, as shown in the middle part of the following table I already posted back in April:

Re-post from: Post 12,221

Even if the gravity model suggests that Montreal-Quebec might be the O-D with the largest ridership potential, there are no other ridership O-Ds which would add significant ridership potential onto that corridor, as Trois-Rivières is much smaller than QBEC and OTTW is almost twice as far from it and much smaller than MTRL. Similarly, the viability of Toronto-London suffers from the low ridership potential beyond Kitchener, as Kitchener is about one-tenth of the size of Toronto…


I‘m afraid it isn‘t as „painfully obvious“ as you (and many other passenger rail enthusiasts) believe, as its ridership potential is about one mesure of magnitude smaller than Montreal-Ottawa-Toronto, if you check in the table above. If we don’t focus on the most biable corridors, we won‘t get anything built…

It is 300 km of flat nothing, with an already existing right-of-way, with three million people. Don't give me this gravity model/city pair thing. You are not factoring in the ease to build the line, which is much greater here then in T-O-M. (Again, I want it all!)

 

[nfitz]

What's the problem with doing that? Nobody is suggesting that this is a primary goal of HFR. Broadly improving travel times is a goal. But from the beginning they've always had other goals that were paramount. Reliability, capacity and frequency are at the top of the list. A secondary benefit of efforts to improve those things is faster travel time. And nobody has suggested that this secondary benefit will be evenly applied.

The rule of thumb used to be 3 to 4 hours - where you need to get travel times under, to have much impact.

Five-hour travel times to Montreal isn't going to do much. And they can do four on the current route, with regulatory changes.

Look at the envelope in Rogers picture. 5 hours is well out of it, and that kind of distance.

VIA is planning on sending 8 trains to Ottawa and 6 trains to Montreal that day.

Years ago, it used to be about 4 to Ottawa and 6 to Montreal. But the Montreal trains were typically much longer. How does that compare now? (to me, the Montreal trains are shorter than they used to be - but load factors were higher.

 

[kEiThZ]

Five-hour travel times to Montreal isn't going to do much. And they can do four on the current route, with regulatory changes.

This is basically an assumption on your part. And one that assumes there are no economic impacts to disrupting freight by just imposing passenger priority.

 

[crs1026]

The rule of thumb used to be 3 to 4 hours - where you need to get travel times under, to have much impact.

Five-hour travel times to Montreal isn't going to do much. And they can do four on the current route, with regulatory changes.

Look at the envelope in Rogers picture. 5 hours is well out of it, and that kind of distance.

Years ago, it used to be about 4 to Ottawa and 6 to Montreal. But the Montreal trains were typically much longer. How does that compare now? (to me, the Montreal trains are shorter than they used to be - but load factors were higher.

The old CN and VIA had a much larger fleet available, and priced accordingly. Trains ran very long on weekends and holidays certainly, but not so much on midweek. Red white and blue fares were the primitive demand management tool. Nobody cared about fleet utilization so having a fleet that only ran at peak was not seen as wasteful.
When VIA acquired the LRC fleet and its older fleet reached end of life, pricing changed. Morr trains were added - two four car trains is a more marketable proposition than one eight car train. The pricing was adjusted to assure loads 7 days a week.
One of the concerns I have with HSR is that it risks veering to an airfare comparator for the very reason that it delivers equal or greater value than road travel. The model needs to price to attract road users, not encourage them to trade off driving time and effort to save money.

- Paul

 

[robmausser]

This is basically an assumption on your part. And one that assumes there are no economic impacts to disrupting freight by just imposing passenger priority.

Another important thing to note that I think people forget is that CN used to be a crown corporation, owned by the government. Of course in this time they would work better with VIA Rail, another crown corporation. Now that they are private, they have an obligation to their investors, to turning a profit, and they simply don't care as much about VIA, nor have government connections to them etc.