Sunday, August 27, 2017

The Monorail Project–Build Out Monorail Ridership Daily Trip Distribution

With the Skyliner gondola transit system announced to connect Hollywood Studios, Epcot, and some nearby resorts, I figured it is time to do some more work on the monorail project.

I have calculated the estimated demand for every hour of each segment of my preferred alternative monorail networks. However, to understand what they mean, it is important to understand how many people an hour a monorail can realistically carry.

The design capacity of any grade-separated transit line is determined by several factors:
  • total passengers per car
  • number of cars per train
  • headway (the time between train departures), which is the sum of:
    • minimum train separation in seconds
    • dwell time at station
To keep things simple, I am assuming that the total number of passengers per car will remain the same as in current Walt Disney World Monorail trains, at 60 people per car. I am assuming that the approximate dimensions of the monorail cars are limited by the width and height restrictions of the monorail where it goes through the Contemporary Resort.

Likewise, the Contemporary Resort also forms the overall maximum length of trains. Currently, the monorail station is the northern half of the building. But it is conceivable that the monorail station platform could be extended to run the length of the building (approximately 430 feet). If the monorail cars remain the same dimensions as they are currently, this creates a maximum train length of 13 cars.
Unfortunately, determining the headway of the monorail requires more than some simple extrapolations of the current monorail system.

From the beginning, I have been assuming that the monorail trains and stations would need to be reconfigured to allow for platform-level boarding. Platform level boarding is when the station platform is level with the floor of the train so passengers in wheelchairs can board without the use of a ramp. This eliminates the requirement for a cast member to be present at all stations to operate the ramp, and can dramatically reduce the required station dwell time.

I also assume that all the bugs in the automatic train control can be worked out to allow for the most efficient train spacing.

Because both of these characteristics are not present in the existing monorail system, I have to look to other transit systems to identify a realistic headway time.

Some metro systems have been known to operate with headways of as little as 70 seconds, but that is on metro systems that primarily serve commuters, people most likely to know where they are going and how things work. Perhaps a better comparison might be airport people movers such as the ones at the Orlando International Airport. An airport people mover's load and unload characteristics are similar to the monorail because they have platform level boarding, automated train control, and platform doors. Airport people movers are not as familiar to riders as urban metro systems, because they are not a part of the daily commute for most passengers. Additionally, the riders of such systems frequently have rolling suitcases which may take up space during passenger flows through doorways in a fashion similar to strollers upon exit and entry of the monorail at Walt Disney World.

An airport people mover system can have headways as fast as 90 seconds, but according to the Transportation Research Board's Transit Capacity and Quality of Service Manual, a more typical headway for such a system is 120 seconds.

Once the design capacity was determined using the numbers above, I calculated the achievable capacity. This is the capacity that a transit line can be reasonably assumed to achieve given the variability in demand across an operating hour. The Transit Capacity and Quality of Service Manual recommends a peak hour factor of 0.8 to be used. This means that if the design capacity for a specific segment is 1,000 people in an hour, it should be assumed that only 800 people in an hour will actually be carried before unacceptable train loads or queuing would be experienced.

Monorail Achievable Capacity for Two Minute Headway Operation by Train Length

Cars Per Train
6 (Existing)
7
8
9
10
11
12
13
Train Capacity
360
420
480
540
600
660
720
780
Achievable Train Capacity (Peak Hour Factor of 0.8)
288
336
384
432
480
528
576
624
Design Capacity (people per hour)
10,800
12,600
14,400
16,200
18,000
19,800
21,600
23,400
Achievable Capacity (people per hour)
8,640
10,080
11,520
12,960
14,400
15,840
17,280
18,720

Now that these capacities have been determined, we can look at the estimated hourly capacities of each network to determine what each network would look like (and whether it would even work at all).

When I first calculated these numbers, I wanted to create some GIFs that would show the estimated ridership on each segment for each hour of the day. Unfortunately, after spending my free time on this task for the last several months (I determined the ridership numbers in February 2017), I have come to the conclusion that either Data Driven Graphics in Photoshop defy logic or (more likely) that I bit off more than I could chew. So, I will instead present some charts showing the maximum hourly demand for each line and maps showing the peak hour passenger flows for each network. If anyone out there knows anything about Data Driven Graphics in Photoshop and wants to help, feel free to comment on this post. Because I still think it would be a helpful way to visualize the passenger flows on each network.

Network 1094 Characteristics

Line
Max. Hourly Passengers
Peak Hour
Peak Segment
Cars Per Train
Headway (minutes)
Cycle Time (minutes)
Trains Required
Park Connector B
22,139
10:00AM – 11:00AM
Hollywood Studios - Epcot
13
1.7
34.7
21
Existing Resort
2,163
10:00AM – 11:00AM
Grand Floridian - Magic Kingdom
13
8.0
15.9
2
Epcot Loop B
3,378
10:00AM – 11:00AM
Port Orleans French Quarter - Epcot
13
9.9
9.9
1
Epcot Loop E
8,193
10:00AM – 11:00AM
Saratoga Springs - Disney Springs West Side
13
4.0
31.9
8
Studios Loop A
4,559
10:00AM – 11:00AM
Boardwalk - Hollywood Studios
13
5.3
15.9
3
Animal Kingdom Loop A
6,392
10:00AM – 11:00AM
All Star - Animal Kingdom
13
4.7
14.1
3

The monorail network with the highest score per mile, Network 1094, has one link with a passenger demand greater than the maximum achievable capacity of a monorail at Disney World given my original constraints. This link, connecting Hollywood Studios to Epcot, would require the longest trains possible to operate at headways of only 1.7 minutes. Unfortunately, this would not be possible to achieve on a consistent basis, so this network must be rejected.

Before we completely reject this network, I do want to discuss a possible confounding factor that may make this a more feasible option. When I started working on this project four years ago, there wasn’t even the slightest hint that the Skyliner gondola transit system would be built connecting Hollywood Studios to Epcot. It remains to be seen exactly how this system will operate and how it would be integrated with any monorail network expansion. But this gondola transit system is being built in the area where my network model suggests there would be the greatest demand. So potentially a network like this may still have a chance of being technically possible.

Network 838 Characteristics
Line
Max. Hourly Passengers
Peak Hour
Peak Segment
Cars Per Train
Headway (minutes)
Cycle Time (minutes)
Trains Required
Park Connector B
16,775
10:00AM – 11:00AM
Hollywood Studios - Epcot
13
2.2
34.7
16
Existing Resort
1.391
10:00AM – 11:00AM
Grand Floridian - Magic Kingdom
13
15.9
15.9
1
Epcot Loop B
2,580
10:00AM – 11:00AM
Old Key West - Epcot
13
7.8
15.5
2
Studios Loop A
2,701
10:00AM – 11:00AM
Yacht & Beach Club – International Gateway 
13
8.0
15.9
2
Animal Kingdom Loop A
3,912
10:00AM – 11:00AM
All Star – Animal Kingdom
13
7.0
14.1
2


Monorail network 838 was investigated because it has the least amount of new monorail track out of any of the 25 top-ranked networks. As you can see in the graphic above, my passenger demand model does not call for greater capacity than a monorail is technically capable of providing, given my initial constraints. This Network would carry an estimated 208,000 people a day--well over the 78,100 people per day that the Miami Metro Rail carries on its 24.4 miles of track (source). In fact, the ridership would be almost identical to that on Atlanta’s MARTA Rail and would carry more people than the Los Angeles Metro Rail. If this network was built, it would be the 9th busiest rapid transit system in the United States (source).

This network would call for 13 car trains that would arrive on 2.2 minute headways for the Park Connector B line. To provide operational flexibility, I assume that the 13 car trains would be used on the other lines, even though they could achieve an acceptable capacity with shorter trains. The table above indicates that the Epcot Loop C, Studios Loop A, and Animal Kingdom Loop A lines could have headways of approximately 7 or 8 minutes, so riders would have an average wait time of 3 or 4 minutes, comparable to the Existing Resort line and much better than the average headways on many major city’s rapid transit lines.

The Existing Resort line, however, would be able to operate with a single train and would have a headway of almost 16 minutes. I would consider this unacceptable, and although enough capacity could be provided with a single train, I think two trains would be necessary to maintain service quality and keep wait times within the limits described at the onset of this project.

Network 1112 Characteristics

Line
Max. Hourly Passengers
Peak Hour
Peak Segment
Cars Per Train
Headway (minutes)
Cycle Time (minutes)
Trains Required
Park Connector B
34,520
11:00AM - 12:00PM
Hollywood Studios - Epcot
13
1.1
34.7
33
Existing Resort
3,619
11:00AM – 12:00PM
Grand Floridian - Magic Kingdom
13
8.0
15.9
2
Epcot Loop B
5,080
11:00AM - 12:00PM
Port Orleans French Quarter - Epcot
13
4.9
9.9
2
Epcot Loop E
14,905
3:00PM - 4:00PM
Disney Springs West Side - Typhoon Lagoon
13
2.5
31.9
13
Studios Loop A
6,840
11:00AM – 12:00PM
Boardwalk – Hollywood Studios
13
3.8
11.3
3
Sports Loop A
11,597
11:00AM - 12:00PM
Art of Animation - Hollywood Studios
13
2.6
12.8
5
Animal Kingdom Loop A
9,611
11:00AM - 12:00PM
All Star – Animal Kingdom
13
3.5
14.1
4

Network 1112 was investigated because it was the highest ranked alternative to have a station at the ESPN Wide World of Sports. This network would carry a whopping 329,000 riders per day and would be the 6th busiest rapid transit system in the United States (behind San Francisco’s BART system which moves 35% more people on 140% more track (source)). However, this network would most definitely not work. At the peak hour in the morning with 13 car trains, the Park Connector B would need to operate at a headway of only 1.1 minute. My research suggests that some rapid transit systems in Eastern Europe may be able to operate at nearly that frequency, but the train doors open before the trains come to a complete stop, the riders are generally commuters who know where they are going and don’t have strollers, the weather is colder so people get out of their seats faster, and the liability and insurance requirements work very differently there than in the United States. In short, this network could not work. Not even the Skyliner gondola system can save it.

To make this system work, I would need to connect the Animal Kingdom side of the Park Connector B to Magic Kingdom and make two separate Park Connector lines running in loops in opposite directions. Maybe one day I will sketch out what that could look like. But for now, pretend this network doesn’t exist and we can move on to discuss something that could work.

Network 453 Characteristics


Line
Max. Hourly Passengers
Peak Hour
Peak Segment
Cars Per Train
Headway (minutes)
Cycle Time (minutes)
Trains Required
Park Connector A
10,382
10:00AM - 11:00AM
Epcot – Hollywood Studios
8
2.1
40.6
19
Existing Resort
2,163
10:00AM - 11:00AM
Grand Floridian - Magic Kingdom
8
8.0
15.9
2
Epcot Loop B
3,378
10:00AM - 11:00AM
Port Orleans French Quarter - Epcot
8
4.9
9.9
2
Epcot Loop E
5,626
10:00AM - 11:00AM
Saratoga Springs - Disney Springs West Side
8
4.0
31.9
8
Studios Loop A
4,559
10:00AM - 11:00AM
Boardwalk – Hollywood Studios
8
4.0
15.9
4

Network 453 was selected because at 23rd, it was the highest ranked alternative to use Park Connector A. This network would be able to operate eight car trains with a 2.1 minute headway on the Park Connector A line and headways of between four and five minutes on all of the resort loops except for the existing Resort Line. The existing Resort Line would be able to operate with only two trains, but the headway would be eight minutes, and would probably be noticeably longer than the waits on the other lines. So, because I can make up whatever I want to, I would recommend having three trains operating on the existing Resort Line so the headways would be in the neighborhood of five minutes like the other resort loops.

This network would carry 178,000 people per day and would be the 9th busiest rapid transit system in the United States (source). The eight-car trains in this network would dramatically reduce the amount that the existing stations would need to be modified, and in most cases would not even require a replacement of any existing track in station areas.

This transit system would require an additional 22 miles of track to build, 36 new eight car trains, 17 new stations and major reconstruction of all the existing stations. This system or any other system expansion described in these posts would optimistically cost hundreds of millions of dollars, and at the end of the day would only increase Disney’s bottom line by the amount that a premium could be charged to the hotel rooms that would be gaining monorail service access. As a fan of the Walt Disney World Monorails and theme parks in general, I am not sure whether I would want Disney to spend that kind of money on something like this instead of, say, a major expansion of an existing theme park.

The construction of such a system would cause major disruption to the existing monorail lines. And once it is built, a minor mechanical problem on a single train would cause incredible disruption resort-wide.

On the other hand, it would be separate guests traveling between parks and hotels from the growing surface traffic congestion across the resort. Wider roads and new bridges aren’t cheap either. Potentially, a monorail expansion could be taken on by the Reedy Creek Improvement District to take advantage of its ability to utilize tax free municipal bonds like what was done with the Disney Springs parking garages.

I am working with a friend who has done some cost estimation for large scale transportation projects so we can try to put together an order of magnitude cost estimate for one of these networks.
If you have any feedback or questions about my methodologies here, please feel free to comment on this post.

About the author: Ben Lytle is an AICP Certified Urban Planner living in Orlando, Florida. He is currently working at AECOM on the architectural facility design of a major theme park attraction in Central Florida and previously worked on the preliminary ground transportation infrastructure design for a theme park resort in Asia. The views expressed in this blog do not represent those of his employer, his employer’s clients, Walt Disney World, or potentially even his own. The information contained in this post is purely speculative and is derived from publicly available information as described elsewhere in this blog. The author does not make any claims to the accuracy of any information presented here. It has not been verified by anybody, and should not be taken any more seriously than any other Armchair Imagineering you read online. So there.