Dec. 2, 2008 Paul Grogan
Flexibility and Design Options Analysis of a Streetcar System in Madison, WI
Paul Grogan
Massachusetts Institute of Technology
In partial fulfillment of the requirements for ESD.71
December 2, 2008
Abstract
The city of Madison, Wisconsin has recently been investigating the construction of a streetcar system to better serve the downtown area and possibly increase economic activity. Similar successful streetcar systems have been created in similar cities throughout the United States. To date, the city has completed an initial feasibility study that outlines possible routes and ridership estimates.
This application portfolio analyzes the feasibility of both a fixed streetcar system with three routes and a flexible streetcar system with an initial route with the option to expand after ten years. The analysis uses a decision tree approach to account for different ridership levels throughout the forty year lifespan of the project. A lattice-based analysis is also performed to value the option to close operations if the ridership fares cannot cover operating expenses.
The conclusion of this study is that a streetcar system could not recoup initial investment and would generate large losses under current ridership estimates. Therefore, unless it is discovered that the ridership levels are actually much higher than expected (on the order of two or three times) or the capital expense of a streetcar system is much lower than previously quoted, the streetcar system should not be pursued further.
Table of Contents
Abstract 3
Table of Contents 4
Introduction 5
System Definition 5
Design Levers and Variables 6
Ridership Uncertainties 7
System Model Description 10
Construction Cost Spreadsheet 10
Operational Cost Spreadsheet 11
Cash Flow Spreadsheet 11
Route Design and Options 12
Option A: Fixed, Large-Scale Design 12
Option B: Flexible, Phased Design 12
Decision Tree Analysis of Flexibility and Expansion 14
Model Assumptions 14
Decision Tree Analysis Results 14
Additional Ridership Analysis (Thought Experiment) 16
Lattice Analysis of Closing Option 19
Lattice Analysis Setup 19
Lattice Analysis Results 20
Path Enumeration and Valuation 22
Conclusions 25
Regarding the Streetcar System 25
Regarding the AP and Analysis Tools 26
Introduction
Over the past few years, the city of Madison investigated the creation of a streetcar system to serve the downtown area. With a history of streetcars and a thriving downtown area, Madison is seen by many to greatly benefit from a streetcar system.
The idea of such a system (often called a “pedestrian accelerator”) is to provide sustainable commuting options, increase property value along the routes, and encourage efficient development patterns. Similar cities including Portland, Little Rock, and Tampa have successfully implemented similar solutions.
The initiative, headed by the Mayor Dave Cieslewicz, was developed and taken though a preliminary feasibility study, but was ultimately abandoned due to a lack of general public support in August 2007.[1] The final report by the feasibility study, however, describes the construction of a streetcar system in a positive light.[2] The concluding points state that Madison is a good candidate city for streetcars, streetcars would complement existing public transportation, ridership would start at strong levels, and the economic impact would be significant from such a project. The final report also outlines several options for an initial-phase construction, including cost estimates and demand models.
Even though there is not presently general public support, the preliminary analysis of the Madison streetcar system indicates that such a project could potentially be further investigated in the future. A more thorough analysis of the performance of the system under uncertainty would help to emphasize the possible benefits of a streetcar system using a flexible design methodology.
In this application portfolio, two kinds of flexibility will be investigated. The first flexible option deals with the phasing of routes. Building a large-scale streetcar system may be less expensive due to economies of scale, but the risk associated with the large capital investment may yield less profit than a phased-in design. The phased design also would benefit from discounted capital expenses as the decision to expand is pushed into the future. This option would tend to limit downside risk under poor conditions, but may also limit the upside potential due to a lower capacity with a phased design. The second flexible option deals with the option to close operations under unprofitable conditions. This would limit the downside risk of system operation under bad circumstances, but would not affect the upside potential under good circumstances.
System Definition
Under the analysis in this application portfolio, the Madison streetcar system is defined to include the following components:
· Construction decisions and costs of routes
· Maintenance costs for sustaining operation
· Ridership projections and revenue
This application portfolio will not investigate the effects on the following components:
· Financial availability and restrictions
· Effect on other public transportation beyond basic ridership projections
Under the data presented in the feasibility study final report and additional information from the Madison Metro Transit 2007 Annual Report, a simple model will be created that estimates construction costs, operation costs, and revenue for a streetcar system.[3] This model will be used throughout this application portfolio to investigate the impact of various uncertainties.
Design Levers and Variables
The most important topic in the planning of a streetcar system is the design of the routes. In the final report of the feasibility study, three routes were described in detail as initial routes to test out the system on a smaller scale before going into a city-wide presence.
The first option, “East Isthmus – Campus Alignment,” is approximated at $83 million and positions a route between the east side of the city and the eastern edge of the campus. The second option, “Park Street Alignment,” is approximated at $94 million and positions a route between the southern edge of the campus and the south side of the city, through several hospital areas. The final option, “Central Loop Alignment,” is approximated at $61 million and positions the route between the campus and the capital.
These three options will form the basis of the analysis in this application portfolio and mix a combination of the following items:
· Initial construction cost
· Annual operational cost
· Initial and projected ridership values
For this application portfolio, I will investigate these three options under uncertainty to select the best initial strategy for a streetcar system.
Ridership Uncertainties
The ridership is a fundamental portion of the success for a streetcar system. The preliminary feasibility analysis final report performed a basic survey of potential ridership levels in the region targeted for streetcar service. The streetcar ridership was calculated by taking a percentage of person trips (in three proximity categories) currently utilizing auto, walking, and bus transportation.[4]
Table 1: Year 2000 Ridership Levels
Trip Mode / Trip Proximity / Total Person Trips / Shift to Streetcar / Streetcar RidersAuto / Interior / 27,385 / 1.50% / 411
Cross-Border / 248,631 / 0.10% / 249
Exterior / 116,057 / 0.00% / -
Total / 392,073 / 659
Walk / Interior / 35,814 / 4.00% / 1,433
Cross-Border / 23,266 / 2.00% / 465
Exterior / 1,502 / 2.00% / 30
Total / 60,582 / 1,928
Bus / Interior / 3,402 / 3.50% / 119
Cross-Border / 16,181 / 0.75% / 121
Exterior / 1,723 / 0.00% / -
Total / 21,306 / 240
System Total / 473,961 / 2,828
Estimates for 2030 were also presented, taking into account future urban expansion in the city from committed development. The increase of 557 riders between 2000 and 2030 represents a ridership growth rate of 0.6%. This is rather small in comparison to the recent growth in the Madison Metro bus system, which achieved a growth of 5.3% between 2006 and 2007.[5]
Table 2: Year 2030 Ridership Levels
Trip Mode / Trip Proximity / Total Person Trips / Shift to Streetcar / Streetcar RidersAuto / Interior / 31,814 / 1.50% / 407
Cross-Border / 302,346 / 0.10% / 302
Exterior / 144,532 / 0.00% / -
Total / 478,692 / 780
Walk / Interior / 44,163 / 4.00% / 1,767
Cross-Border / 25,738 / 2.00% / 515
Exterior / 1,517 / 2.00% / 30
Total / 71,418 / 2,312
Bus / Interior / 4,259 / 3.50% / 149
Cross-Border / 19,262 / 0.75% / 144
Exterior / 1,991 / 0.00% / -
Total / 25,512 / 294
System Total / 575,622 / 3,385 (0.6%/year)
Since the Madison streetcar project is still in its infancy, it is important to compare these ridership values to those from established streetcar systems. In the Portland streetcar project, the projected ridership target was 3,500 when it opened in 2001, which was immediately exceeded. By 2005, the ridership had grown to over 9,000 on weekdays and has further grown to over 11,900 per weekday during the winter of 2007/2008.[6]
A similar success story is echoed in Tampa.[7] Their “Vintage Trolley” streetcar system opened in October 2002 with an expected ridership of 950. By February and March 2003, the daily ridership had risen to over 1,500. Between 2003 and 2007, the ridership levels have increased at an average of 1% per year.[8]
One reasoning for the low ridership growth in the Madison analysis is actually addressed as well in the feasibility analysis report. It seems that by adding the streetcar system in the city, it actually spurs additional economic growth, thus inherently increasing ridership. With these additions to ridership levels, the annual growth rises to 1.5%.
Table 3: 2030 Ridership Projections with Redevelopment
Trip Mode / Trip Proximity / Total Person Trips / Shift to Streetcar / Streetcar RidersAuto / Interior / 36,540 / 1.50% / 548
Cross-Border / 319,824 / 0.10% / 320
Exterior / 151,991 / 0.00% / -
Total / 508,355 / 868
Walk / Interior / 47,925 / 5.00% / 2,396
Cross-Border / 26,725 / 2.50% / 668
Exterior / 1,557 / 2.00% / 31
Total / 76,207 / 3,096
Bus / Interior / 5,242 / 4.50% / 236
Cross-Border / 22,087 / 1.00% / 221
Exterior / 2,334 / 0.00% / -
Total / 29,663 / 457
System Total / 614,225 / 4,420 (1.5%/year)
Considering the wide range of ridership values between different cities, the two uncertainties I will investigate are the initial ridership values and the ridership growth rate.
I will range the initial ridership values between -20% to +50% of the stated values in the feasibility report. This accounts for the potential “over-selling” of the streetcar by the feasibility study as well as the potential for similar cases as in Portland and Tampa where the initial ridership exceeds expectations. The three levels (-20%, 0% and +50% from stated numbers) are assigned best-guess probabilities of occurrence. The raw ridership numbers from the reports also have to be modified to account for the different portions of the region covered by the three routes, as approximated by a population use factor from personal experience or population density and local travel patterns.
Table 4: Initial Ridership Ranges
Route / Population Use Factor / Expected RidershipP=60% / Low Ridership
(-20%)
P=10% / High Ridership
(+50%)
P=30%
East Isthmus – Campus / 40% / 1,131 / 905 / 1,697
Park Street / 45% / 1,272 / 1,018 / 1,909
Central Loop / 30% / 848 / 679 / 1,272
Also, I will also range the ridership annual growth rates between a worst-case 0.5% per year, expected 1.5% per year, and as high as 2.5% per year, independent of route. In addition to varying the growth rates, I could also experiment with variable growth rates, by selecting a different growth rate each year with the following cumulative distribution based on the ridership levels for the Madison metro bus line between 1971 and 2007.[9]
Figure 1: Ridership Growth Rate CDF and PDF
System Model Description
The system model is used to generate cash flows and valuation of a particular design over the course of its forty year lifespan. The system model has three main components: a spreadsheet to estimate construction costs, a spreadsheet to estimate operation costs, and a spreadsheet to tabulate annual cash flows.
Construction Cost Spreadsheet
Construction costs are taken directly from the feasibility study final report and take into account estimates for construction items, construction soft costs, construction contingency costs, engineering and administration, and vehicles. For each of the three routes considered, the estimated construction costs are taken directly from the feasibility study. For the large-scale route, the construction costs of the constituent routes are summed and some line items are slightly discounted for economies of scale.
Table 5: Example Construction Cost Spreadsheet (East Campus – Isthmus Route)
Item / Unit Price / Units / Qty / Total PriceTrackwork - Track Slab (single) / $ 425 / lf / 28,100 / $ 11,942,500
Trackwork - Turn/Track Crossing Installation / $ 160,000 / ea / 2 / $ 320,000
Catenary Poles and Overhead Wire / $ 200 / lf / 28,100 / $ 5,620,000
Traffic Signals - New (or Full Replacement) / $ 200,000 / ea / 3 / $ 600,000
Traffic Signals - Modified / $ 100,000 / ea / 21 / $ 2,100,000
Civil/Roadway - Asphalt Pavement Overlay / $ 75 / lf / 35,300 / $ 2,647,500
Civil/Roadway - Concrete Pavement Reconstruction / $ 125 / lf / 10,550 / $ 1,318,750
Civil/Roadway - Bridge Reconstruction / $ 150 / sf / 0 / $ -
Utilities - High Allowance / $ 600 / lf / 6,900 / $ 4,140,000
Utilities - Medium Allowance / $ 300 / lf / 11,300 / $ 3,390,000
Utilities - Low Allowance / $ 150 / lf / 9,900 / $ 1,485,000
Drainage Allowance / $ 50 / lf / 28,100 / $ 1,405,000
Stop Platforms - Street Side / $ 75,000 / ea / 21 / $ 1,575,000
Stop Platforms - Median / $ 110,000 / ea / 1 / $ 110,000
Stop Platforms - Upgrade / $ 45,000 / ea / 3 / $ 135,000
Substations / $ 500,000 / ea / 5 / $ 2,500,000
Train Signaling Systems / $ 550,000 / ea / 0 / $ -
Maintenance Facility Allowance / $ 4,000,000 / ls / 1 / $ 4,000,000
Construction Subtotal / $ 43,288,750
Construction Soft Costs / 18% / $ 7,791,975
SUB-TOTAL CONSTRUCTION COST / $ 51,080,725
Construction Contingency Cost / 15% / $ 7,662,109
TOTAL ANTICIPATED CONSTRUCTION COST / $ 58,742,834
Engineering and Administration Cost / 15% / $ 8,811,425
Vehicles / $ 3,000,000 / 5 / $ 15,000,000
Right of Way / $ -
GRAND TOTAL / $ 82,554,259
Operational Cost Spreadsheet
Operation costs are tabulated using the following quantities from the feasibility study final report: