Designing Regional Transit for Southern Ontario and Quebec

So during the past few weeks of the Covid-19 pandemic, I’ve been fortunate to have had a lot of time to just do the things I’ve wanted to do. One thing that started to interest me was transit, specifically how we can design efficient public transit networks.

In a where time more and more people are moving back to cities, making public transit a confident primary means of transportation rather than an embarassing alternative is crucial in order to go about this development in a sustainable way.

Why?

Well, many people who own a car and use it to get around have probably, at some point in their driving lives, been annoyed with the amount of traffic they have to get through in order to get to wherever they are going.

Why so much traffic? Simply because people have places to go, and cars are viewed as the obvious way of getting there.

And so by this rationale, more people = more cars.

If we continue in this pattern, traffic is only going to get worse. And the inevitable things that follow like frustration, decreased levels of happiness, more time spent getting between places and emissions are also going to get worse.

Which is not what we want.

The solution? Public transit. Or, public transit that is actually effective.

So, after trying to redesign my city (Ottawa’s) public transit system, I decided that I’d try my hand at a regional transit system and chose Southern Ontario and Quebec (which contains 1/3 of the Canadian Population).

Here, you can find exactly how I did that. For more specifics as to the exact stops for each route, go ahead and take a look at: https://www.notion.so/Designing-Regional-Transit-for-Southern-Ontario-Quebec-9940b2e1f11b47a7bd22b30d029741ff.

Alright, let’s jump in.

Step 1: Find the Density

So, just like with all public transit, this inter municipality system is meant to serve density. That (in most cases) means the cities with the highest population.

But of course, when looking at things on a provincial scale, there are so many different little pockets that we might want to service. But, we can’t possibly service all of them with out things getting really complicated and/or expensive.

The way to get around this has to do with a concept I proposed called municipality clusters.

Municipality clusters are groups of towns, cities, villages and exurbs that people frequently commute between on a day to day basis.

In Canada, we have things called Census Metropolitan Areas (CMAs) which are basically just one core municipal area surrounded by a bunch of smaller communities which share the same public infrastructure and consist of over 100,000 people.

The same applies to Census Agglomerations (CAs) which are the same as CMAs only with 10,000 to 100,000 people.

While CMAs and CAs are not interchangeable with the idea of municipality clusters, they do come pretty darn close, so for the purposes of this project we’ll take them to mean the same thing.

My theory is that while each municipality with in a municipality cluster might have it’s own transit system, they should collectively extend their lines so that there is an efficient way to get between each municipality with in the cluster.

In this way, once all municipalities with in a CMA or CA are connected, all we have to do is intercept the major municipality with in each of them. By doing this, instead of having to intercept every single city, town and hamlet, we can service the most densely populated ones, but residents of a periphery community still have a way to get to the provincial line.

This also allows us to better understand the purpose of our routes. On a day to day basis, people who live in smaller communities will often commute to the closest large city for work which means that they need a day to day form of long distance public transit. But sometimes, people also commute long distances for travel and conferences. They also could use some form of public transit.

Our line can’t do both. Or, it could, but what would be better is if the first kind of public transit (day to day commutes) is done like above, using municipality clusters. Our line will be for the second kind.

Using the Canadian Census website, we can figure out that the CMAs the southern Ontario and Quebec region are,

• barrie
• belleville
• brantford
• guelph
• Hamilton
• Kitchener — Cambridge — Waterloo
• london
• montreal
• peterborough
• quebec city
• st. catharines
• trois rivieres
• windsor
• oshawa
• peterborough
• ottawa-gatineau
• Sherbrooke

Plotting them on google maps,

Now, since our transit has to service this density, to get a rough idea of how each route will look, we are going to draw lines connecting neighbouring municipalities in as direct a fashion as possible.

Step 2: Most Appropriate Roads

Now, these lines are not themselves the roads which we can drive on — some of them just go through green space. So, for each connection, we are going to find the roads that most closely follow them.

Step 3: Making Routes

Right away, we can see that about 12 of these 17 municipalities are quite close to the 401 (in Ontario) and then the Trans Canada Highway (in Quebec).

It’s important to keep in mind though that while all these cities are heavily populated or are surrounded by heavily populated areas in terms of travel and work, some are more frequented than others. These cities are Quebec City, Montreal, Ottawa, Kingston Toronto, London and Windsor.

Luckily, six out of seven of these cities lie on this 401 to TransCanada Highway corridor (Ottawa lies a short drive up from the 401 on the 416).

Naturally, it makes sense to have a transit route on the 401 to the TransCanada Highway (from Windsor to Quebec City). Overall, it hits some of the most heavily populated areas in the region, it services popular destinations and it is direct.

We’ll call this our L1 line.

Then our L2 line will be the short sprint up the 416 to Ottawa.

Now, we do need other lines that get to all the other places that are not along the 401. We’ll come back to this later.

Step 4: Choosing a Method

Now, we need to figure out how we want to travel these routes. So much of the time, cities will think about what they want their transit to look like before they figure out what they want it to do which inevitably leads them to mold the latter condition to the former condition. And because what transit looks like (i.e. the mode) limits where transit can go, this way of planning might sometimes end us up missing important places.

With inter municipal transit, this is less likely to happen just because there are fewer options available (you can’t exactly put a streetcar on a highway), but we still can make this mistake.

Treat what transit will look like as a secondary factor to what it will do.

But, now that we more or less know what directions we want to go in, we can figure out what mode. Since our L1 along the 401 and TransCanada is direct and services the most significant destinations, with few stops in between, we’ll make this a train track.

Trains are fast, high capacity and unencumbered by all the other traffic that happens on the freeway. Also, since they are not actually on the highway, we technically don’t have to use the exact 401 route, rather we’d build the track close by to it, and perhaps in a straighter fashion when applicable.

Now, trains are also significantly more expensive to install and maintain than a regular bus. Which is why they rely on their high capacity to keep them going. For our routes that don’t service these main destinations and consequently probably won’t have as much consistent demand, we’ll instead use busses.

Busses are also better suited to this job because their routes are changeable — they’re not set in stone. They can be modified as time goes on — maybe if certain routes become obsolete or are more in demand at certain times of the year.

Now, what about our L2 line? Since it’s a pretty short distance that could probably be serviced more frequently, due to the fact that it’s a municipality cluster, we’ll choose an express bus.

Step 5: Stops

In terms of where L1 will stop, obviously, it’ll stop when it intercepts each of these large cities, but it will stop when it gets close to smaller cities that are directly around it. To figure what these cities might be, let’s consider all the CAs in southern Ontario and Quebec (again using the Canadian Census website). Plotting them on Google Maps with the CMAs,

The ones done in orange are the ones that are closest to where L1 would be.

They are

• Victoriaville
• Drummondville
• salaberry de valafield
• cornwall
• brockville
• woodstock
• ingersoll
• chatham-kent

Now, we also want to have a specific stop for Ottawa. Although it doesn’t lie directly on the 401, we will definitely have a bus that shuttles people back and for between it (seeing as it is the country’s capital). So, at the intersection of the 401 and the 416, we’ll also put in a stop.

For our L2 line,

• spencerville
• kemptville

Step 6: Frequency

So, the frequency with which we run each line does in part depend on how fast each takes to travel.

We’ll start with the L1 line.

For now, we are just going to look at the frequency in terms of the main stops — i.e. Quebec City, Montreal, Kingston, Toronto, London and Windsor (and we’ll throw in the 401 x 416 for good measure).

The average commuter train travels at 160 km per hour. Combining this fact with the distance between significant stations,

• Quebec City to Montreal is 254 km which means it’ll take 95.25 minutes
• Montreal to Kingston is 293 km which means it’ll take 109.9 minutes
• Montreal to the Ottawa Exit takes 70 minutes
• the Ottawa Exit to Kingston takes 40 minutes
• Kingston to Toronto is 262 km which means it’ll take 98.25 minutes
• Toronto to London is 196 km which means it’ll take 73.5 minutes
• London to Windsor takes 191 km which means it’ll take 71.7 minutes

Now, we do need to factor in stop times. Let’s say that we have 10 minutes of stop and buffer time for each station. Now, we can workout what time we arrive at each station. Say we leave Quebec at 7:00 a.m. Then, we’ll

• get to Montreal at 9:05
• get to Ottawa Exit at 10:45
• get to Kingston at 11:45
• get to Toronto at 14:05
• gets to London at 16:10
• gets to Windsor at 17:40

Now, of course, we’re not going to have just one train each day. Say, we want a train to be leaving each of these stations every 3 hours. But, we also don’t need to have trains running 24 hours a day. Let’s say we don’t want a train leaving before 5 a.m. or after 21:00 (this doesn’t mean that a train can’t arrive at a station after 9:00.

So, we can basically just add increments of 3 hours to each of these times.

Pretty quickly, we can see that not all of our trains make it all the way. The third train leaving Quebec City at 13:00 would only make it to London. And, not all cities would get a train until half way through the afternoon (the first time Toronto gets a train is 14:05).

To fix this, we’ll segment the route — i.e. not necessarily start at Quebec City. This works the same way as above, just adding and subtracting increments of 3 hours to each station until it exceed the time limits.

We get…

Copy of Timetable — Southwestern Direction

The stars beside the time mean that that is the last stop for the train, the dashes mean the train doesn’t get there.

Repeating the exact same thing for the opposite direction,

Copy of Timetable — Northeastern Direction

Step 7: Auxiliary Routes

Now, this isn’t all the lines we need. Way back in Step 2, we identified a whole bunch of other CMAs we need to hit, and then in Step 5, a while bunch of CAs which would also be nice to service.

And these routes will be serviced by busses, so we need to identify the roads that these buses will go on.

Ideally, these lines will connect to a station that is on L1 but also to each other. In this way, the routes can be used not only as a way of getting to the train for travel purposes but also as a way of getting to different cities on a daily basis for work.

Because of this, the purpose of the bus lines won’t be to go into every nook and cranny of every residential area in a municipality, only to kind of stop by it. If that city is large enough to have it’s own transit system (even if it’s one bus), it can connect up to this spot and take people deeper into the city. If a city isn’t large enough or if it’s an exurb, well then that person can probably walk or bike.

Now, we’re pretty lucky here in that, unlike with a municipal transit network where there are several roads going in any particular direction, at the regional level, most cities lie along just one highway. So, going back up to our previous map with all the CAs and CMAs, we can highlight these highways. Google Maps displays all of them in yellow which makes them easy to identify.

L1 is highlighted in blue.

For now, we’ll call these routes the feeders. Rather than figuring out the highway each feeder goes on, we’ll instead figure out the municipalities and exurbs it will stop at.

This can be done using the extremely classy method of zooming in really close to google maps and finding noting down the places that have city like clumps of roads. If you want to take a look at the exact stops, just go into the document (link at the top).

Some of these lines could be amalgamated into one, but I tried to break them up when they changed direction on the map or reached a significant city.

Now, of course this brings up the question who is in charge of these lines?

Certain lines, like F12 (Guelph to Owen Sound) probably won’t be used so much due to the great distances between them paired with the fact that they are not popular destinations. The people who use these lines would probably be doing it occasionally. These could be managed provincially.

But then, lines like F3 (Montreal to Sherbrooke) would probably be used quite a bit. After all, 40% of Granby’s work force commutes do a different Census Subdivision, the two closest ones being Montreal and Sherbrooke which are both well with in commuting distance. The people who use these kinds of lines could be doing so more on a day to day basis. Lines like these could be managed as a partnership between the CMAs and CAs who are on them as they would see the most significant benefit.

Even further, lines like these might just be able to be extensions of existing municipal lines particularly in Montreal and Sherbrooke where they can continue on to more specialized directions once they get there.

Overall, the responsibility of the line could depend on why it’s being used and how often it’s being used.

Now, you could also say that we could add in lines from Belleville to Madoc, Kingston to Sharbot Lake, Orilla to Oshawa. This is totally true, and in fact some of the lines above might prove to be useless in practise. Adjustments like these can be made when cities get feedback on how they are performing

• should a line run more frequently?
• should it run earlier or later?
• should it stop running at a particular time of year?
• should it run at all?

Are among all the questions a city can assess when looking at their lines.

Step 8: Signage

Last, but not least the signage. Unlike a municipal bus line, where there are multiple stations in a given area and you can go multiple places from each, in this case when a person shows up at a station, it’s pretty intentional as to why they are there.

Still, the main components each station should explain

• Where the rider is now
• Where the rider can go from here
• When the transit is expected to come by

Where the rider is now can simply be established by naming each station, in this case by the name of the city it’s passing through.

Where the rider can go from here can be done through a schematic map. Now, I’m no graphic designer, but here’s an example:

• The CMAs are indicated in red bolded text
• the CAs are indicated in black bolded text
• L1 is in yellow
• L2 is in blue
• every stop has a node

And then, what time transit is expected to come by can be done in real time monitoring.

Alright, so just to recap the general steps of what we did here

1. Find the density as the most populous cities and turn them into municipality clusters, then identify the most significant municipality with in them.
2. Find the roads that most closely follow this density pattern.
3. Create routes by prioritizing cities that are the most populous but also common destinations.
4. Choose a means of transit most appropriate for your network.
5. Stop once per smaller city
6. Timetable using line segmentation
7. Fill in auxiliary routes in such a way to intercept your major ones. Keep in mind they can be changed depending on performance. They can be under provincial jurisdiction or several municipalities depending on use and frequency.
8. Create consistent signage that tells people where they are, where they can go and when transit is coming.

So, that’s my stab at creating a Regional Southern Ontario-Quebec transit system. It’s by no means an exhaustive list of what does go into such a system (trains themselves, station specifics…) but it’s an outline!