PRT: Monmouth County /
ORF 467 Final Network Report /
Julio Fredes (partner: Caroline Logue) /
1/11/2011 /
Introduction
Monmouth County, the northernmost county on the Jersey Shore, holds a 2009 population estimate of 644,105 people and maintains a population density of about 1,304-people/ square mile. With a total area of 665 square miles comprised of 29.04% water, and rivers and bays leading to the New York Harbor, Jamaica Bay, and the Atlantic Ocean, Monmouth County is ranked 42nd among highest-income counties and 53rd in highest per capita income in the United States. However, with the county’s growing population and steep cost of living, the county suffers from obsolete infrastructure.
The county is relatively flat and low-lying, and the county seat is Freehold Borough while the largest municipality by population is Middletown Township. There are a total of 53 municipalities. Monmouth County has a few relatively dense areas along its coast, but also some extremely isolated areas in the southwest. In the summer months, people flock to Monmouth County’s shore, making roads particularly congested. Some other notable attractions include Freehold Raceway – a horse racing arena, Wall Stadium, and Wall Ice Arena. On the border of Ocean County and Monmouth County are Six Flags Great Adventure, Six Flags Wild Safari, and Six Flags Hurricane Harbor. Monmouth County is also home to the Monmouth County Mall as well as Monmouth University, a private four-year university.
There are 14 NJ Transit heavy rail stations and approximately 20 Park & Ride locations. With a substantial number of commuters to New York and a seasonal influx of people traveling to the beach during the summer, Monmouth County would likely flourish with a PRT system, perhaps with the exception of its significant rural area in the Southwest.
2009 ORF 467 PRT Proposal
2008 ORF 467 PRT Proposal
2007 ORF 467 PRT Proposal
2004 ORF 467 PRT Proposal
More Information on Monmouth County
NJ DOT Rideshare
Placemark Data
Our starting point in this PRT was examining the placemark data from last year. From the offset, however, it became clear that the data needed extensive altering due to a large number of duplicate placemarks and values for workers/patrons that were obviously wrong (for example, 100,000 patrons / day at certain hospitals.) Caroline and I agreed on a general approach to cleaning up our data, which included
- Removing duplicate entries
- Fixing the coordinates that were pointing to Antarctica
- Double checking that some of the workplaces/attractions were in the right place (using Google Earth to cross reference the actual address),
- Consolidating schools, which were listed twice – under type 2: education and type 3: business.
I then looked up data for rail station usage, and decided to add 7000 train station commuters on average per train station and 800 commuters in each park and ride. This 7000 figure might seem a little high at a first glance, but can be explained by the fact that Monmouth County is home a large number of people who work in New York City.
In addition, we drastically changed patron numbers for service businesses that would not necessarily attract many people per day (for example, a Comcast office.)Instead, we decided that such services as UPS delivery people, plumbers, etc would be counted as patrons/day in the housing section. Instead of having zero patrons per day, we set patrons to 1% of the population for each census block. This addition did not drastically alter the results but certainly helped account for these services, which had not been added before.
This much improved placemark information file allowed us to make much better decisions when placing our stations.
Initial Networks
My initial 3 networks looked for dense, heavily populated places that could serve as self-sustaining parts of the county. Two of them were located in the vicinity of the coast while the other one focused on Middletown Township, the county’s largest town with a population of roughly 66,687 people.
My first initial network covers Long Branch, Monmouth University, West Long Branch, the Monmouth County Mall, and others. The Long Branch NJ Transit Station provides access to NYC, which is where a high number of people in this county work. Therefore, access to a station was crucial, and I made sure to include at least 1 in this initial network. This network contains a small number of 2-way guideway. The network is pictured below. The left image shows the network by itself, while the second image shows the network once it was attached to the entire network.
My second initial network covers Middletown Township, and includes 1 NJ Transit station , some shopping areas, and the town center. I deemed it important to have a PRT in this area, as it not only provides transportation to the people here, but is also a symbol of moving forward. Once the most important places have the technology, others will want to follow. The initial network is pictured below, along with the stations connected with the whole network included.
My third initial network covers the Neptune and Asbury Park Areas (along the shore), 2 NJ Transit stations, Brookdale Community College, and Bradley Beach. This network pays particular attention to travel time, and includes 9 interchanges to help minimize travel time and provide easy of travel, avoiding nonsensical trip routes. As before, the initial network is pictured below.
Building the Rest of the Network
When I began the project, my goal was to build a network that would grow over time and eventually serve 90% of the people. However, I quickly realized that this would not be achievable given that many parts of the county are not dense at all.
For my final network, I decided to look for the most logical places to place stations. Thus, I added stations in places where the placemarks added up to a reasonable number of people. I looked to stay over 1,000 people for each station, a task that became increasingly difficult once I reached the southwest part of the county. My general approach was as follows:
- Keep the initial networks as being built in year 1. Then look to connect these networks by adding stations in year 2.
- Work my way out from these initial networks, moving west and south as necessary, and increasing the years I got further away from the initial networks.
Following this approach, I was able to cover approximately 80% (79.65%) of the population by placing 681 stations, 252 interchanges, and 712 miles of guideway throughout the county. The next step was connecting this large set of stations. This part required significantly more analysis in order to make sure that the network made sense. For example, I looked to add grid-like behavior in heavily populated areas in which crossing guideways would connect to main guideways. Below is an example of one area in which this worked successfully.
In other cases, however, loops were necessary due to the sparse population being served by the PRT. These loops would be connected by to other sections of the county by one set of guideway coming in and another going out, as pictured below:
On a year by year basis, my network looks to be built mostly on years 1 and 5. This would seem to make sense, since after a few years, people would become more used to the idea of a PRT network and to her areas would begin to demand the service. On the other hand, if the project wasn’t successful, a five year period would be sufficient time realize this and not build another large section of the network. However, this is an area in which my network could be improved next year, as it is not always spread out in the best possible way in terms of the year that each section was built. The statistics below show the network. However, there seems to be a bug in the code that generated this data (for example, interchanges information is wrong,) so only stations built per year should be given any serious consideration.
Financial Analysis
Because of the shortcomings of the year by year analysis that was given to us, we decided to follow the spreadsheet provided by Professor Kornhauser for our financial analysis. This spreadsheet required the trip information for each station, total number of stations, interchanges, guideways, and cost of capital.
For the trip ends served per day, the program simply added up all the trips for all of the stations. For the total trips per day, however, the spreadsheet implemented a different formula, which equaled .5*total trip ends served*(ratio of trip ends served to total trip ends.) Since not all trips were served, this resulted in lower total trips per day than total trip ends served. The peak hour trips were defined as 15% of the total trips per day, and the fleet size defined as (peak hour trips / 10) * 1.1. The average trip length was 5 miles,average vehicle occupancy of 2, and fare of $3/ride. These assumptions are summarized below:
Under these assumptions, the basic costs, revenues and profits for Monmouth County are:
In other words, we see that under this set of assumptions (which accounts for the entire network being built in year 1), the profit per year is about $357 million. Below we plot a graph showing the Profit & Loss (P&L) for fare values between $1 and $5 per ride, and we see that the breakeven point is at a fare of approximately $2.20/ride.
Given the large capital requirements to build such a project, it is comforting to find out that under our set of assumptions, the network would be profitable at such a reasonable rate, which we assume people would be willing to pay. Unfortunately, this analysis assumes that the PRT would not be in competition with cars and other modes of transportation. This is a large shortcoming of our analysis, and one that we would have looked much more closely into had we had more time to work on this.
Ridership Information
As noted above, our statistics are somewhat unrealistic due to high ridership assumptions. In this section, we offer two graphs which show the ridership per station.
This first graph shows us a breakdown of how many trip ends are served per station per day. We see that the largest spike occurs at around 37,000 riders per day. This number may seem somewhat high. Upon closer inspection, however, we note that it serves an area that includesthe Freehold Raceway Mall and Freehold Horse Track. Thus, it would make sense that the ridership would be high in this area. See picture below for visual explanation.
We tracked down similar spikes, and they all seem to make sense. Indeed, serving 37,000 trip ends as our maximum sounds quite reasonable. Below we include a sorted graph by trip ends served per day. Our results are quite reasonable once again.
Dijkstra’s Algorithm (section written together with Caroline)[1]
We implemented Dijkstra’s algorithm in MATLAB to measure how efficient our networks are. To do this, we used the “matrices.html” link provided by John and Nathan and followed the simple instructions to obtain the distances traveled and the paths from each stations to each other. In the gravity method we implemented in the My City assignment we assumed that distances between two zones were just 1.2 times the Cartesian distance of their centroids. We see now, after applying Dijkstra’s algorithm to our network, that in our last assignment we made a gross simplification. Although each of our counties are 916 square miles (Ocean County) and 665 square miles (Monmouth County), the top ten distances between two stations running along network guideway can reach as much as 108 miles for Ocean County and 69.71 for Monmouth.
The table below shows the top 20 largest station-to-station trips using this code:
From this, we clearly see that station 407 is not very well linked to the rest of the network. Tracing what station this is, though, proved very difficult, so we did not have time to trace what the cause of this was. However, it seems obvious that something went wrong with this station and requires particular attention.
These figures highlight the trade off we made when creating our networks. In an effort to conserve money, we purposely avoided constructing stations and, consequently, guideway that would produce fewer than 1,000 trips per day; this cost consciousness caused us to serve fewer than 90% of trips as per the initial goal. Our conservative approach to laying stations inevitably produced networks that were less efficient.
What we deem most advisable would be to begin construction on the conservative end that we have upheld in our final network design, and hold off on investigating technical inefficiencies until the network has proven itself sustainable and profitable in the long run.
Appendix 1: MATLAB Code for Dijkstra Implementation
A = csvread('JF40djmat.csv');
C = csvread('JF40guidewaydist.csv');
LoS = csvread('JF40-matrices-los.csv');
[cost,path] = dijkstra(A,C);
fori = 1:length(cost)
for j = 1:length(cost)
if cost(i,j) == Inf
cost(i,j) = 0;
end
end
end
maximums = zeros(3,10);
fori = 1:20
[r_valr_ind] = max(cost,[],1);
[col_valcol_ind] = max(r_val);
maximums(1,i) = r_ind(col_ind);
maximums(2,i) = col_ind;
maximums(3,i) = cost(r_ind(col_ind),col_ind);
cost(r_ind(col_ind),col_ind) = 0;
end
Appendix 2: Financial Analysis Previously Conducted on Initial Networks.
Please disregard these results when comparing to the ones highlighted above. This was simply a preliminary analysis on the initial networks.
Network 1. Covers Long Branch, Monmouth University, West Long Branch, 1 NJ Transit Station, Residential Areas, 1 Mall, 1 Hospital
Assuming 203,394 Trips
Cost of One-way Guideway / $77,300,000Cost of Two-way Guideway / $3,000,000
Cost of Stations / $24,000,000
Total Building Cost / $104,300,000
Upper Bound Total Building Cost / $114,730,000
Annual Maintenance / $2,294,600
Average Trip Length / $8
Annual Operating Cost / $59,242,570
Annual Interest Cost / $6,883,800
Total Annual Cost / $68,420,970
Annual Fares Revenue / $222,716,430
Annual Rent / $432,000
Total Annual Revenue / $223,148,430
Anticipated Annual Revenues/ Anticipated Annual Costs / 3.26
Total Revenue after 30 years / $6,694,452,900
Total Cost after 30 years / $2,167,359,111
Revenue less Cost after 30 years / $4,527,093,789
Number of Trips / 203,394
Number of Miles of One-way Guideway / 15.46
Number of Miles of Two-way Guideway / 0.5
Number of Stations / 12
Average % of Guideway Ridden / 50%
Cost per mile of One-way Guideway / $5,000,000
Cost per mile of Two-way Guideway / $6,000,000
Cost of Station / $2,000,000
Operating Cost per Mile / $0.20
% Maintenance Cost on Capital / 2%
Threshold Upper Bound Capital Cost / 1.1
Average Number of Riders in Vehicle / 2
Average Fare Paid / $3.00
Rent Paid Per Month Per Station / $3,000
Interest Rate on Bond / 6%
Network 2. Covers the biggest township in the county, Middletown Township, 1 NJ Transit Station, Shopping Areas (Including 1 Target Store), High School, Town Center
Assuming 121,931 Trips
Cost of One-way Guideway / $60,600,000Cost of Two-way Guideway / $6,000,000
Cost of Stations / $26,000,000
Total Building Cost / $92,600,000
Upper Bound Total Building Cost / $101,860,000
Annual Maintenance / $2,037,200
Average Trip Length / $7
Annual Operating Cost / $29,195,159
Annual Interest Cost / $6,111,600
Total Annual Cost / $37,343,959
Annual Fares Revenue / $133,514,445
Annual Rent / $468,000
Total Annual Revenue / $133,982,445
Anticipated Annual Revenues/ Anticipated Annual Costs / 3.59
Total Revenue after 30 years / $4,019,473,350
Total Cost after 30 years / $1,222,178,759
Revenue less Cost after 30 years / $2,797,294,591
Number of Trips / 121,931
Number of Miles of One-way Guideway / 12.12
Number of Miles of Two-way Guideway / 1
Number of Stations / 13
Average % of Guideway Ridden / 50%
Cost per mile of One-way Guideway / $5,000,000
Cost per mile of Two-way Guideway / $6,000,000
Cost of Station / $2,000,000
Operating Cost per Mile / $0.20
% Maintenance Cost on Capital / 2%
Threshold Upper Bound Capital Cost / 1.1
Average Number of Riders in Vehicle / 2
Average Fare Paid / $3.00
Rent Paid Per Month Per Station / $3,000
Interest Rate on Bond / 6%
Network 3. Covers Neptune and Asbury Park Areas, 2 NJ Transit Stations, Brookdale Community College, Several Heavy Residential Areas, and Bradley Beach
Assuming 206,215 Trips
Cost of One-way Guideway / $60,050,000Cost of Two-way Guideway / $1,500,000
Cost of Stations / $26,000,000
Total Building Cost / $87,550,000
Upper Bound Total Building Cost / $96,305,000
Annual Maintenance / $1,926,100
Average Trip Length / $6
Annual Operating Cost / $46,139,575
Annual Interest Cost / $5,778,300
Total Annual Cost / $53,843,975
Annual Fares Revenue / $225,805,425
Annual Rent / $468,000
Total Annual Revenue / $226,273,425
Anticipated Annual Revenues/ Anticipated Annual Costs / 4.20
Total Revenue after 30 years / $6,788,202,750
Total Cost after 30 years / $1,711,624,255
Revenue less Cost after 30 years / $5,076,578,495
Number of Trips / 206,215
Number of Miles of One-way Guideway / 12.01
Number of Miles of Two-way Guideway / 0.25
Number of Stations / 13
Average % of Guideway Ridden / 50%
Cost per mile of One-way Guideway / $5,000,000
Cost per mile of Two-way Guideway / $6,000,000
Cost of Station / $2,000,000
Operating Cost per Mile / $0.20
% Maintenance Cost on Capital / 2%
Threshold Upper Bound Capital Cost / 1.1
Average Number of Riders in Vehicle / 2
Average Fare Paid / $3.00
Rent Paid Per Month Per Station / $3,000
Interest Rate on Bond / 6%
Comparison Chart:
Network 1 / Network 2 / Network 3Main Areas Served / Monmouth University / Middletown Township / Brookdale Comm. College
1 NJ Transit Station / 1 NJ Transit Station / 2 NJ Transit Stations
Monmouth Medical Center / Target / Neptune
Long Branch / Middletown High / Asbury Park
Monmouth Mall / Holmdel Town Center / Bradley Beach
Trips / 203,394 / 121,931 / 206,215
Guideway Miles / 15.96 / 13.12 / 12.26
Stations / 12 / 13 / 13
30-year Profit / $4,527,093,789 / $2,797,294,591 / $5,076,578,495
1
[1] See appendix for MATLAB code