Professional Documents
Culture Documents
Projet Transport Commun PDF
Projet Transport Commun PDF
Projet Transport Commun PDF
Written by
Arianne Robillard – 1963449
Étienne Léger – 1639754
Hedie Keshavarz – 2116552
Tarcisio Neto – 2111477
3. METHODOLOGY ...........................................................................................................................3
5. OPTIMAL NETWORK......................................................................................................................6
7. BIG NETWORK............................................................................................................................ 10
REFERENCES ...................................................................................................................................... 21
ii
1. Introduction
This work aims to propose four optimization scenarios for public transit network of the city of Drummondville.
Drummondville is a city in the Centre-du-Québec region located in the north-east of Montréal with the Saint-
François River passing along it. The city of Drummondville has a population of 71,716 people with an area of 260,10
𝑘𝑚2 [1]. It should be noted that there are currently 41 225 households in Drummondville. In terms of public
transit, there is no metro and no municipal train or tramway system in the city of Drummondville. The only
passenger train station in Drummondville is used by VIA Rail for intercity trips mainly to and from Montreal and
Québec City. The city is served by 8 bus routes by the total fleet of 9 minibuses and a service of taxi bus. Among
bus routes, 5 routes work from Monday to Friday (7a.m. to 23p.m.) and 1 route is working Monday to Friday
(5:30a.m. to 12:30a.m.). These routes are all active from 8a.m. to 6p.m. on Saturdays and 9a.m. to 6p.m. on
Sundays. Two other bus routes that covers mostly the industrial neighbourhoods are active in working peak hours
(5:44a.m. to 8:14a.m., 14:14p.m. to 16:44p.m., and 22:14p.m. to 23:44p.m.) during weekdays and are inactive
during weekends.
In this work, four different scenarios with different required budget will be proposed. A comparison analysis will
be done between the proposed networks along with the current one in terms of annual operation cost,
accessibility to important landmarks, operation time, and headway. All the scenarios are modelled in Transition
platform.
1
2. Objectives and requirements
2.1 Objectives
When designing a public transit network from scratch, one can quickly become overwhelmed by the complexity
of the task at hand, even if the city of interest is relatively small, like Drummondville is. Therefore, it was crucial
for our team to set a list of objectives considering city’s budget that would help guide the initial structure of the
network. From there and in conjunction with the actual requirements, the rest would follow more naturally, or at
least logically. Thus, before drawing any line, we have brainstormed about what those main objectives would be
and come up with the following:
1. Create a network that any trip requires no more than one bus transfer, an exception made to very long
trips coming from more distant suburbs, where two transfers could be acceptable.
2. Ensure the best service for students by allocating more units to the lines going around the CEGEP, UQTR
and other main High schools in the city
3. Provide a direct access to the hospital to a maximum of households
4. Target a maximum headway of 30 minutes for all lines, at all times (No shortage of service in non-peak
hours)
5. Target an inter-stop distance between stations around 300 or 400 metres except for very low density
areas in outer suburb
6. Connect underserved residential neighborhoods to important landmarks
The reason for limiting the number of transfers by one is because it was determined that in a mid-sized, car-
centric, North American city like Drummondville, no resident will ever use public transit if they have to wait for a
bus on three occasions (wait in origin stop, wait at first transfer, and wait at the second transfer) for one trip.
Because public transit is most likely a lot slower than automobiles, so it should at least be somewhat convenient.
It was agreed that one transfer is the maximum a user would accept.
Secondly, there are many reasons why a transit agency would want to ensure a good service for students. One of
them is that the probability that they own a car, even though it is really high in Drummondville, is still lower than
for older citizens. Therefore, it is an opportunity for the agency to capture their usership and form mobility habits
for the young generation. Moreover, they should also enjoy the experience so that when the time comes to choose
to buy a car or not, they have good experience of public transport and maybe think twice before buying a car. One
of the challenges in connecting students to CEGEP was that it is surrounded by the river and railway in both sides
so buses should take a specific rout to reach there.
Thirdly, it was thought that maximizing accessibility to the hospital is crucial because health services are a central
part of establishing a good social safety net. Even people with lower income who do not own a car should be able
to reach health services easily.
Then, it was decided to target the headway of around 30 minutes in the Optimal network based on the city budget.
It was considered a reasonable compromise between sufficient service in wider area and not to be expensive. For
the bigger networks, headway was reduced to 20 minutes on some lines close to CEGEP and the hospital.
Regarding the inter-stop distance, it was decided to target a distance between 300 and 400 metres so that a
comfortable walking distance for public transit users can be offered. Since, by increasing the walking distance, the
2
transit ridership will drop noticeably. Thus, suppose the walking speed of 5 𝑘𝑚 ⁄ℎ, it is agreed that walking time
of 4 to 5 minutes will be acceptable by users.
Lastly, since the current network does not serve all residential areas in the city (ex: the residential area in south-
east). It is agreed to build the proposed networks in a way that connects underserved residential neighbourhoods
to different opportunities.
2.2 Requirements
In addition to the objectives listed above, a few requirements were provided, that further helped define what kind
of network would be offered. The requirements from the city are as follows:
1. Offer a service that includes 16 to 20 units in its optimal form, so the cost does not exceed city’s budget
2. Design lines that are bidirectional and symmetrical
3. Propose a service that is available all weekdays and weekends and provide constant operating hours every
day without cutting the service in low demand hours.
3. Methodology
Following the list of objectives and requirements that were presented above, the first step was to identify different
points of interests of Drummondville. It’s important to make sure that at least one line of the network passes in
front of those points of interests (POIs). These points of interest come from different sources. Starting from
important land uses, educational, commercial, residential, health, and main shopping services were identified by
looking at the city on Google Maps. Then, it was possible to get two datasets: the first containing all the major
employers in the city and the second, a geodata file containing all households in the city. From these, it was
possible to select 35 POIs. A second screening was done among those 35 POIs and finally the list was shortened
to the 10 most important landmarks in the for the undertaking the analysis in this work (Fig 2).
Légende
Figure 2 - Location of the top 10 POIs on the map (from Google Maps)
The second step was to analyze the actual bus network of Drummondville and identify the strength and
weaknesses that are presented in table 1, from the network illustrated at figure 2.
3
Table 1 - Strengths and weaknesses of the actual bus network of Drummondville
Strengths Weaknesses
The lines reach the points of interests identified Lines form long loops in one direction only
The lines give access to the densest residential areas There are a lot of residential areas that are not covered
The lines reach all the major employment areas The lines are not effective for the students who want to
go to CEGEP or university (lots of detours, lots of
transfer, long time of travelling).
Smaller budget is sufficient to run the service There is shortage of service in non-peak hours and there
are some lines with no service offered in weekends
The third step was to identify the highest density residential areas to make sure the new lines give access to these
people. Figure 3 presents the residential heat maps. Heat map was set with a 250m radius and a maximum number
of flats of 200.
Also, it is worth noting that different stop times are considered depending on the proximity of the stations to
different landmarks. Since in some stations such as the ones close to CÉGEP or hospital, the number of entries and
egresses will be higher, a longer stop time is required for the service to be safe and efficient. On the other hand,
in outer suburb areas, it is probable that there is nobody in a bus station so that the bus will not even stop in some
stations. Also, in case there are people to pick up in outer suburb areas the load will not be high, so a short stop
is enough. In suburb regions where the demand is low, a stop time of 5 seconds is considered by assuming that
4
the bus will stop in half of the stops on average. To fulfill the network the rest requirements, an average of 10
seconds stop time is established for the whole network. For important landmarks with relatively higher demand
such as shopping centres, schools, sports centres, etc., 20 seconds is considered. And finally, a 30 second stop
time is established on the stations close to CEGEP and the hospital due to higher demand and possible mobility
issues.
In terms of fleet choice, two types of buses with different capacities are selected. Minibus with the capacity of 25
and midibus with capacity of 45-50. Midibuses were chosen for the lines close to CEGEP and hospitals that higher
demand is estimated. Also, the service will run for 18.5 hours in each day from 5:30 a.m. to 24 p.m. so that people
who start working around 6:00 can benefit from the public transportation system.
According to these remarks, it was possible to develop a network different network that respects the targeted
objectives and addresses the weaknesses of the actual network. In order to meet project objective in terms of
frequency of service (average 30 minutes headway) the lines have been shortened or lengthened or the number
of units has been adjusted. To test the efficiency of service in terms of accessibility, service polygons were
generated from the 10 principal points of interest with 3 isochrones of 15 minutes (15 min, 30 min, 45 min) and
the number of dwellings reached were considered as an indicator to evaluate accessibility. Moreover, based on
the number and type of fleet used for each network, an analysis of daily, weekly, and yearly operation cost, and
an analysis of annual operation cost per km have also been undertaken. Finally, different network will be
compared based on what is measured above.
4. Actual Network
The actual bus network of Drummondville is illustrated on Figure 4. There are 8 lines in service and almost all of
them form loops in one direction. Details of the lines are presented in Table 2. It was not possible to simulate the
exact schedule and headway, because the information about the stopping times, operation speed and terminal
times are unknown. However, for the sake of simplicity and to conduct a before-after analysis, a simple scenario
with the same daily 18.5 hours of operation time is considered for the actual network. The accessibility analysis is
then run with the default parameters of Transition. Since the fleet only includes 9 minibuses, by the assumptions
made above, the total annual operation cost can be estimated 4,848,480 $.
5
Table 2 - Summary of the lines for the actual bus network of Drummondville
5. Optimal Network
In this section, the optimal network is proposed for the city of Drummondville. It is called optimal network as it is
the network that respects the city’s budget in its best. Also, the network respects the requirements of the city and
the targeted objectives. It focusses on giving a good accessibility to people to the important landmarks.
∑𝐿
1 𝑜∗𝑉𝑜
1 Commercial speeds of every network were calculated as follows: where L is the number of lines of a network, 𝑜 is the
∑𝑜
number of trips and 𝑉𝑜 is the operation speed.
6
considered for the 5th line (lime green) and the 6th line (red) to offer a better capacity to the students that are
going to CEGEP and UQTR. For the other lines, minibuses with a capacity of 25 places will be running, as it is now.
4 Minibus 24 min 2 Rue de la Sainte-Anne/ Rue Victorin Terminus (Édifice Francine Ruest-Jutras)
5 Midibus 26 min 2 Rue Armand/ Boulevard Foucault Terminus (Édifice Francine Ruest-Jutras)
7
5.2 Costs
The summary of the estimated operation costs is presented in Error! Reference source not found.. As mentioned b
efore, 18.5 hours of operation time was considered for each day from 5:30 to 00:00. The operation costs are
established at $80.00/hours for the minibuses and $120.00/hours for the midibuses. The costs are reported per
day, per week and per year. The annual operating cost for this network is approximately $11,851,840.00.
Table 4 - Summary of the costs for the optimal network
Type of bus Number of units Cost ($/h) Hours of operation Total cost per day
Error! Reference source not found. present the cost per vehicle-km, an important parameter that will be c
ompared with other networks in section 0. To estimate this parameter, total daily vehicle traveled (VKT) for each
line is calculated by multiplying the number of trips in each day in line length. Then, a sum total of VKT for each
network is calculated. Finally, total operation cost/day of the network is divided by total daily VKT to reach
operation time per kilometer.
Table 5 - Summary and Data for calculation of Cost per km for the optimal network
Line Number of units Vehicle type Cost ($/h) D p (km) Number of trips Line cost ($/day) VKT
1 4 Midibus 100 14.7 84 $ 7,400.00 1234.800
2 3 Minibus 80 13.004 80 $ 4,440.00 1040.320
3 2 Minibus 80 10.122 63 $ 2,960.00 637.686
4 2 Minibus 80 6.166 94 $ 2,960.00 579.604
5 2 Midibus 100 6.056 86 $ 3,700.00 520.816
6 2 Midibus 100 7.861 71 $ 3,700.00 558.131
7 2 Minibus 80 5.64 85 $ 2,960.00 479.400
8 3 Minibus 80 13.217 74 $ 4,440.00 978.058
Cost per km $ 5.40
6. Small Network
6.1 Description of the network
Figure 6 below represents the second scenario developed to improve public transport network in Drummondville.
This network was originated from the optimal network, however, with certain restrictions in the transport units
to reduce operation costs, but in the same pre-set interval of 30 minutes. The total length of the network 63.768
Km and the commercial speed of the network is 18.3 km/h.
In this scenario, it is aimed to improve accessibility to greater number of residents and a better connection to a
greater number of points of interest. Consequently, there will be an expansion of the provision of the service to
8
new areas in Drummondville compared to current network. In addition to reaching important and new regions to
the north of the city with line 4, there was an expansion in the East area, with 2 lines (line 1 and line 2) that leads
to downtown, but there were also extensions and changes in lines in commercial areas around the West zone of
the city with 4 lines (lines 1, 2, 3 and 7) serving these regions, where there was not good public transp ort coverage.
The passenger distribution was expanded to important points in the city through lines 5 and 6, such as CEGEP and
the hospital by improving the service frequency.
Table 6 shows a summary of the lines that integrate the network. The restriction of an average interval of 30
minutes has been met, and the entire network is operated by 16 units with required smaller operation budget.
Table 6 - Summary of the lines for the small bus network
6.2 Costs
As this scenario represents a small network with smaller budget, 16 minibus units were chosen to circulate on all
lines. Today in Drummondville, the transport agency offers 9 minibus units, which means that in addition to
9
improvements in the design of the route of the lines, there will be an increase of more than 77% in the service
fleet.
In Error! Reference source not found. there is the report costs, with an operation of 18.5 hours per day, and an a
verage cost of 80 $/h per minibus unit, the total operation cost per day is estimated to be 23,680.00 $. The annual
cost of the service has an estimated value of $8,524,800.00. Table 8, provides information on operation cost per
kilometer for this network which is estimated to be 5.16 $/km. The same calculation as the previous network is
done for this estimation.
Table 7 - Summary of the costs of the optimal network
Type of bus Number of units Cost ($/h) Hours of operation Total cost per day
Minibus 16 $ 80.00 18.5 $ 23,680.00
Midibus 0 $ 100.00 18.5 $ -
Total $ 23,680.00
Total per Week $ 165,760.00
Total per Year $ 8,619,520.00
Table 8 - Summary and Data for calculation of Cost per KM for the small network
Line Number of units Vehicle type Cost ($/h) D p (km) Number of trips Line cost ($/day) VKT
1 3 Minibus 80 12.144 74 $ 4,440.00 898.656
2 3 Minibus 80 10.066 68 $ 4,440.00 684.488
3 2 Minibus 80 9.906 66 $ 2,960.00 653.796
4 2 Minibus 80 8.523 74 $ 2,960.00 630.702
5 2 Minibus 80 6.366 86 $ 2,960.00 547.476
6 2 Minibus 80 5.648 85 $ 2,960.00 480.080
7 2 Minibus 80 11.115 62 $ 2,960.00 689.13
Cost per Km $5.16
7. Big Network
7.1 Description of the network
The Figure 7 shows the third network developed by the team. This network is an enlarged version of the optimal
network, with the total length of 98.367 km and the commercial speed of the network: 19.6 km/h. In this network
a few changes have been made, like some lines have been extended in length and some lines have increased
frequency. The goal here was to cover more residential areas especially the ones in less dense suburb areas. The
other goal was to increase the frequency of the service specifically in the lines close to CEGEP. To fulfil these goals,
it was necessary to have 6 to 8 units more than the optimal network. There are five new areas covered with this
network where service had not been offered before. They are shown in orange on Figure 7.
First, the green line has been added. It is believed that adding a line in this area would be beneficial for the
residents because in other networks the simply have none. And the density along the line is similar to the pink
line on the other shore of the river. Second, the pink line now reaches a residential area that is quite significant.
The goal for this area was to capture the residents who had never experienced service in any other scenario. The
10
same logic was applied with the extension of the violet line. For that one, not only does it give access to public
transport to the neighbourhood where the new terminal is, but it also allows people along the blue line to take
the violet line instead if they want to reach CEGEP or the industrial area in the south-west with one transfer less.
Then, the extension of the red line was important to the team because the area is quite densely built and no
service is actually given to them with the optimal scenario or the small one, even though they have access to the
green line on the actual real-life network. For these residents, it is also hard to justify that they live so close to the
city centre but have no service to go there. The blue line has also been extended to the east, with the terminal
located further from downtown. This allows the line to reach a few neighbourhoods more and therefore increase
accessibility to those residents. Finally, one change that is not visible is the improvement in frequency for the lines
that go around CEGEP. In real life, that line is the one that has the most passengers, therefore it was logical to add
more frequency there.
Total units 28
Table 9 above shows the summary of units used per line in the big network. The total of the unit is much greater
than in the two scenarios presented now (with 8 more units compared to Optimal network), but it is like that
because now many more residential areas are served. The frequency has also been increased in lines close to
11
CEGEP and in the less fortunate neighbourhood covered by the yellow line. The reason the dark green line has
been kept at 30-minute headway is because it is serving an industrial area therefore it is assumed that the ridership
on that one only happens in peak in the morning and in the afternoon. As shown, the maximum headway is still
30 minutes, but many more lines have shorter headway compared to the optimal and small networks presented
earlier. This gives much more enjoyable experience to users as they have to wait less during transfers which means
shorter trips.
Table 10 - Summary of the costs of the big network
Type of bus Number of units Cost ($/h) Hours of operation Total cost per day
Minibus 22 $ 80.00 18.5 $ 32,560.00
Midibus 6 $ 100.00 18.5 $ 11,100.00
Total $ 43,660.00
Total per week $ 346,500.00
Total per year $ 18,018,000.00
Table 10 above presents the summary of operation costs for the big network. Of course, now that the number of
units has increased significantly, the budget also has to be adjusted in consequence. But even though the cost is
much higher than the optimal and small networks, it has to be said that with this one, the number of passengers
is assumed to be also greater because of improved accessibility to residential areas. Unfortunately, this is a study
that is not possible to do within the scope of this project but that would be interesting to compare the cost per
passenger.
7.2 Costs
Table 11 - Summary and Data for calculation of Cost per km for the big network
Line Number of units Vehicle type Cost ($/h) D p (km) Number of trips Line cost ($/day) VKT
1 5 Minibus 80 20.749 83 $ 7,400.00 1722.167
2 4 Minibus 80 15.498 93 $ 5,920.00 1441.314
3 5 Minibus 80 15.629 106 $ 7,400.00 1656.674
4 3 Minibus 80 8.785 106 $ 4,440.00 931.210
5 2 Minibus 80 9.062 79 $ 2,960.00 715.898
6 4 Midibus 100 10.142 117 $ 7,400.00 1186.614
7 2 Midibus 100 5.64 97 $ 3,700.00 547.080
8 3 Minibus 80 12.862 74 $ 4,440.00 951.788
Cost per km $ 4.77
In Table 11 above, the cost of each line has been calculated, along with the vehicle -km per line, per day. This
allows calculating a total cost per kilometre for the entire network, shown to equal 4.77$/km. For a direct
comparison of cost per kilometre between each network, see section 9.
12
8. Big With Loops Network
8.1 Description of the network
In this section another big network (fig 12) with a completely different design compared to previous networks will
be proposed. The total length is 137.192 km and the commercial speed of the network is 20.2km. The objective
here is to give access to public transit to all residential neighbourhoods even the ones located in outer suburbs by
the pre-set 30-minute headway. For example, the white line will cover all households located in the north shore
of Saint-Francois River. The light blue line covers all residents in the two neighbourhoods in the north-west.
Orange, crimson and light green lines are bringing east side residents to the city centre. The red line is designed
in a way to connect the underserved south-west residential zones to the city centre while making it possible to
transfer to pink or pine green lines to reach CEGEP. Pink line is designed in a way to give access to CEGEP and city
centre. Purple line will bring students from south residential areas to CEGEP and also intersects with both pine
green and light green lines to enable transfers for other destinations. Blue line is designed in a way to circulate in
the more industrial neighbourhood by bringing people from different locations through transfers from pine and
light green lines. However, the network design will include a lot of detours in residential neighbourhoods which
make the trips be longer and the network to be less efficient in terms of cost. The total fleet of 33 which is more
than all the scenarios presented before is required to run this network with 10 lines and with the target headway
of 30 minutes. Since the light green line is a line with most transfers as it is crossing almost all other lines, and the
pink one is carrying a high number of students to CEGEP, it is agreed to use midibus with higher capacity for these
two lines. The rest will operate with minibuses. The details of the fleet number and headway in each line are
presented in table 12. The network is promising to the maximum 30-minute headway, there are also lines with
headway close to 20 minutes.
13
Table 12 - Summary of the lines for the big bus network with loops
Total units 33
8.2 Costs
As mentioned above, two types of vehicles are chosen to circulate on this network: 25 minibuses with operation
cost of 80 $/h and 8 midibus units with operation cost of 100 $/hr. With an operation time of 18.5 hours per day,
the total operation cost of 51,800$ per day and an annual budget of 18,855,200$ is estimated for this network
(table 13). Not surprisingly due to increase in number of units and improving frequency in some lines , the total
yearly estimated cost is higher than all other scenarios.
Table 13 - Summary of the costs of the big network with loops
Type of bus Number of units Cost ($/h) Hours of operation Total cost per day
Minibus 25 $ 80.00 18.5 $ 37,000.00
Midibus 8 $ 100.00 18.5 $ 14,800.00
Total $ 51,800.00
14
Table 14 - Summary and Data for calculation of Cost per km for the big network with loops
Line Number of units Vehicle type Cost ($/h) D p (km) Number of trips Line cost ($/day) VKT
1 3 Minibus 80 16.159 86 $ 4,440.00 1389.674
2 5 Minibus 80 18.129 78 $ 7,400.00 1414.062
3 4 Midibus 100 19.207 74 $ 7,400.00 1421.318
4 3 Minibus 80 10.21 97 $ 4,440.00 990.370
5 3 Minibus 80 10.9 86 $ 4,440.00 937.400
6 4 Midibus 100 12.008 93 $ 7,400.00 1116.744
7 2 Minibus 80 7.156 79 $ 2,960.00 565.324
8 3 Minibus 80 15.673 77 $ 4,440.00 1206.821
9 3 Minibus 80 10.676 86 $ 4,440.00 918.136
10 3 Minibus 80 9.948 97 $ 4,440.00 964.956
9. Comparison Analysis
In this section, the actual network and the four new proposed networks are compared together regarding
operating costs and accessibility to destinations from households.
• The number of households reached from 10 selected POIs in 15, 30, and 45 minutes by public transit
• The area reached by from 10 selected POIs in 15, 30, and 45 minutes by public transit
• Annual operation cost of each network
• Operation cost per kilometer of each network
• Commercial speed
Error! Reference source not found. regroup the 10 most important points of interests for Drummondville and i
ndicates the area covered in km2 and the number of households reached in 15, 30 and 45 minutes, by public
transportation only, for each developed network. The accessibility calculation was done for an arrival time of 8:00
a.m. The goal was to maximize both indicators. Error! Reference source not found. presents the ranking of the n
etworks according to the number of households reached. For every point of interests, a ranking of the five
networks was made according to the number of households reached and for every isochrone. The network that
reached the largest number of households was given 5 stars. The final ranking represents the average ranking for
each point of interest. Obviously, the big network is on the top step of the podium, because it reaches the most
households on average for every isochrone. The optimal network is on the second step of the podium which allows
concluding that it confers a good accessibility, from the households to the destinations, compared to the other
networks. Not surprisingly, the big network with loops stays in the third step after the optimal one, due to the
existence of a lot of loops, the 45 min access by public transit is hard to reach with such a design. And finally, we
have small and actual network with 2 and 1 star respectively. In terms of commercial speed, we can see that in
average the speeds are the same in all networks with a bit of increase in bigger ones.
15
Table 15 - Number of households reached in 15, 30 and 45 minutes for 10 points of interests
4.96 1529 5.26 2610 5.54 3476 5.05 1364 4.92 1225 15 minutes
Hopital Ste-Croix 15.73 11846 23.04 18245 25.3 19953 18.39 15400 12.67 10004 30 minutes
48.91 27663 63.83 35345 64.63 34353 48.47 30526 29.71 20689 45 minutes
4.91 831 4.98 582 5.05 440 4.96 238 4.91 93 15 minutes
Cégep de
14.27 13632 16.87 15143 14.81 12377 13.22 11032 10.65 8281 30 minutes
Drummondville
41.09 26940 45.14 27965 42.07 27865 37.34 23993 23.72 19202 45 minutes
4.95 142 4.96 219 5.56 1742 5.06 560 4.97 60 15 minutes
UQTR 13.52 6985 16.58 9999 17.53 11693 13.42 9030 9.44 4614 30 minutes
43.72 19022 45.57 23097 37.3 22745 33.95 21638 21.5 14734 45 minutes
4.92 259 4.99 791 4.95 699 4.92 64 4.93 75 15 minutes
Promenades
13.05 7694 16.27 10658 15.75 10266 13 6961 11.66 6497 30 minutes
Drummondville
42.87 22026 45.12 22877 45.59 23100 38.02 21965 27.36 17774 45 minutes
4.93 172 4.9 219 4.91 54 4.91 54 4.91 16 15 minutes
École secondaire
11.9 6912 12.34 8363 13.16 9697 11.15 7601 9.33 3517 30 minutes
Marie-Rivier
28.43 17965 37.82 24311 42.49 24435 34.9 21274 18.27 14140 45 minutes
4.92 227 4.96 306 4.94 660 4.95 511 4.95 229 15 minutes
Centrexpo Cogeco 13.55 9052 16.17 11537 15.23 11205 16.55 10898 13.3 10754 30 minutes
45.64 26685 45.38 23370 45.03 23827 44.82 25115 31.38 18773 45 minutes
4.9 268 4.9 254 4.91 379 4.91 1 4.92 0 15 minutes
Autobus Girardin 14.46 3444 10.61 2336 10.72 2240 10.23 635 9.82 747 30 minutes
35.05 17600 27.62 14591 28.82 15001 26.5 14704 19 5193 45 minutes
5.1 545 5.35 562 5.05 177 5.1 379 4.9 0 15 minutes
Centre Jeunesse-La
15.44 10583 19.99 12300 14.07 6482 14.17 6582 9.48 1588 30 minutes
Maurice
41.96 23410 52.74 27126 40.46 23152 35.87 22281 17.03 8702 45 minutes
5.31 265 5.13 98 5.04 77 5.01 11 4.9 0 15 minutes
Centre commercial
22.47 5836 19.51 6187 14.46 4734 13.18 3219 8.73 0 30 minutes
Charpentier
57.73 18625 52.19 17739 29.29 15882 26.39 13901 8.79 367 45 minutes
4.92 0 5.29 416 4.91 0 5.04 22 4.94 0 15 minutes
Soprema 13.24 1920 16.49 5394 11.06 2018 15.01 3728 10.99 303 30 minutes
34.94 17306 42 17201 28.49 10649 32.81 14614 21.28 4548 45 minutes
4
Number of stars
3
15 minutes
30 minutes
2
45 minutes
0
Big with Big Optimal Small Actual
loops
16
Optimal Big
Actual
The accessibility maps in Figure 10 above show the area that can reach CEGEP with an arrival time of 8:00am. It is
possible to see that the actual network does not offer a really good service in the outskirts of the city, the service
being more focused on city center. All scenarios developed by the team cover a larger area and the reason is
simple, we have more units and shorter headway. But this comes with increased operation costs. It is also possible
to notice a difference between the proposed networks. For example, the big network is by far the one that covers
the most area. The main gains are obtained in the eastern suburbs, where none of the other networks reach in 45
minutes. This observation also leads to the conclusion that the big network with loops is in fact trying too hard to
reach each residential areas and it loses efficiency because of that. In other words, even though with that network,
many – if not all – households have a relatively close access to a bus stop, it takes so long for the bus to complete
the loops and 45-minutes accessibility is not improved. Also worth noting, although the optimal network is not
covering eastern suburban areas as well as the big network, it does serve those in the center and north better.
That is due to the fact in the big network, one of the two lines circulating on the north shore actually deviates
farther to the east whereas on the optimal, both lines stay close to the UQTR campus, improving the service there.
17
Optimal Big
Actual
Again, looking into the accessibility map of the hospital, the actual network covers more the center areas of the
city and not outskirt neighborhoods. At a glance we can see that optimal network is providing access to the
hospital for almost all residential neighborhoods even the ones located in suburbs. Not unexpectedly, the big
network does better in terms of offering access to residentials even the ones located in far east, but the
accessibility maps shows that big and optimal network are very close. Only the 30-minute access isochrone is a bit
wider in south neighborhoods in the big network. The network with a lot of loops provides less access to hospital
to city residents since making a lot of detours make this network to work slower than others. Looking into the
north residential area, we can see a better service coverage in 30 minutes by optimal network. This shows that
separating two lines (yellow and green) at the top is not a very good idea and we would have a better coverage
by having two lines in the north residential area rather than one going to north-east. Since we know north-east
residents are more of high-income people who rely on private cars rather that public transit but people who leave
in the north are more of low-income people who should be given more priority in providing access to public transit.
18
Annual Operation Cost of each Network
$20,000,000.00 $18,855,200.00
$18,018,000.00
$18,000,000.00
$16,000,000.00
Cost per Year ($)
$14,000,000.00
$11,851,840.00
$12,000,000.00
$10,000,000.00 $8,619,520.00
$8,000,000.00
$6,000,000.00 $4,848,480.00
$4,000,000.00
$2,000,000.00
$-
Big with Loop Big Optimal Small Actual
Looking into figure 13, we can see that the operation cost per kilometer ($/km) for the optimal network has the
highest value at 5.40 $/km. This is because the optimal network has a shorter length and smaller number of trips
which makes VKT much smaller than the two bigger networks.
However, in figure 14, where the cross comparison of operation cost analysis and the amount of household
reached in 45 minutes (from CEGEP) is represented, we can see that there is a plateau in the number of households
reached in the big networks and the optimal one, but the annual operation cost for optimal network is way smaller
(close to half) of the big ones. This leads to the conclusion that the optimal network, by having relatively lower
annual cost, is the best choice. As we observe, it manages to serve almost the same number of households in 45
as minutes and is more efficient.
4.74 4.77
4.8
4.6
4.4
Big with Loop Big Optimal Small Actual
19
Comparison Cost / Household reached in 45 minuets by Public
Transit (POI: CEGEP)
$20,000,000.00 30000
NB OF HOUSEHOLD REACHED
26940.00 27965.00 27865.00
23993.00 25000
$15,000,000.00
20000
19202.00
$10,000,000.00 15000
10000
$5,000,000.00
5000
$- 0
Big with Loop Big Optimal Small Actual
20
References
[1] Affaires municipals, Regions et Occupation du territoire, Quebec. Retrieved on April 11th, 2022 from
https://web.archive.org/web/20120512010442/http://www.mamrot.gouv.qc.ca/repertoire -des-
municipalites/fiche/municipalite/49058
[2] drummonville.ca, public transport map and schedule 2022, Retrieved on March 18th 2022, from
https://www.drummondville.ca/wp-
content/uploads/2022/03/Carte_horaire_transport_en_commun_mars_2022.pdf
21