Ferry Boat Feasibility Study

2010

For the year 2000, the bus service characteristics are defined using data from various service providers including:

• The Washington Metropolitan Area Transit Authority (WMATA),

• The Potomac Rappahannock Transportation Commission (PRTC), and

• Private carriers in the City of Fredericksburg, Spotsylvania County, and Stafford County.

Travel times from outlying park and ride lots to downtown bus stops are taken from public timetables. Trip frequencies are determined by inspection of the schedules. Fares, taken from the public timetables, assume use of multiple trip discounts and reflect constant dollars between 2000 and 2010.

Rail Service

Rail service includes VRE trips originating in Virginia and Metrorail trips originating in Virginia and Maryland.

The boarding and alighting stations used by VRE riders, train travel times, and fares are assumed to be the in constant dollars for 2000 and 2010. The train frequency for 2000, based on an analysis of peak period schedules, is determined to be 28 minutes. For 2010, a 20-minute frequency, based on planned VRE enhancements, is used for most analyses.

For 2000, rail trips originating in Maryland are assumed to drive to the park and ride facility at the Anacostia Metro Station. For 2010, access to Metrorail is assumed to occur at the Branch Avenue Metro Station. Year 2000 travel times and fares from the Anacostia Station are taken from published WMATA materials. Year 2010 times and fares from the Branch Avenue Metro Station are also obtained from WMATA. Fares are assumed to remain unchanged and in constant dollars.

A multi-step process was designed to develop efficient operational ferry routes and services for testing that would be attractive to potential riders. Numerous docking locations were examined including Georgetown, Maine Avenue, and the Navy Yard in the District of Columbia, the Pentagon, Woodbridge, Neabsco Creek, and Cherry Hill in Virginia, and Indian Head and Fort Washington in Maryland. The size of potential travel markets, the achievable dock-to-dock travel times, and the times for competitive services are quantified.

The route analysis also considered the publicly documented plans of potential private ferry service operators.

As evidenced by the analysis in Chapter 3, two ferry routes emerge for further testing:

• Fort Washington to the Pentagon, and

• Woodbridge to the Navy Yard.

Travel times are estimated for each route. Two sets of one way fares are developed:

• $7.00 – based on a review of other commuter ferry service proposals, and

• $4.40 – similar to the VRE fare from Woodbridge to DC.

For each of the three transit modes – bus, rail, and ferry – it is necessary to estimate the time and cost of the trip on the primary mode, and the time and cost of access at the residential trip end and egress at the work trip end.

It was assumed that passengers would drive from residential areas to the nearest park and ride lot (bus), station (VRE and Metrorail), or dock (ferry).

Egress in the metropolitan core is more complex. For each workplace zone, the transit services were examined to determine how a worker would travel from the primary mode to the workplace. Workers could walk or use another mode (generally Metrorail). For example, a worker traveling by bus from Prince William County to the Pentagon can walk directly from the bus stop to his or her workplace, however, a worker traveling to Rosslyn would need to get off of the bus at the Pentagon and transfer to Metrorail, incurring additional travel time and expenses.

For each workplace area, the appropriate trip paths and associated times and costs were examined.

The ferry patronage forecasts were developed by using a model that estimates the share of transit travel that will be attracted to each of several available transit modes. The process is described in greater detail in Appendix 4-2. In general terms, the overall mode choice model used a nested logit formulation. The submode choice portion of the model, in which the proportion of riders using each of the available transit modes is estimated, was structured as a multinomial logit model. The resulting transit mode market shares were then multiplied, for each area-to-area pair, by the transit trip forecasts obtained from MWCOG.

The mode choice model is an extension of a procedure originally developed for VRE ridership forecasts. To ensure that the model is working properly, a year 2000 "Null" case was analyzed considering only the allocation of transit trips between rail and bus. Estimates of total transit trips were obtained from MWCOG. Travel time and cost data for bus and rail were developed as described above, and the model was applied.

The resulting bus and rail patronage estimates for travel originating in Virginia were compared to actual ridership data gathered in conjunction with a 1998 study of travel in the I-95/I-395 corridor. This comparison illustrated that the model yielded a reasonable estimate of the relative use of the modes.

The ferry patronage forecasts were performed for 2000 and 2010. Fourteen different forecasting scenarios were developed and tested in order to examine the sensitivity of alternative ferry system characteristics on ridership projections. These scenarios are based on different combinations of the following four parameters:

• Ferry Speed – Currently, various locally designated speed restrictions exist along portions of the proposed ferry routes. These speed restrictions intend to control vessel wake impacts for certain areas of the river. Projected ferry travel times are significantly affected by the speed restrictions. For example, ferry travel time from Woodbridge to the Washington Navy Yard, including maneuvering and docking, would be 70 minutes with the existing speed limits. The same route would take 45 minutes if the restrictions were lifted. The forecasting effort analyzed both "restricted" and "unrestricted" ferry speeds.

• Ferry Fare – As discussed previously, two fares were tested: a $7.00 one-way fare based on other commuter ferry services, and a $4.40 one-way fare similar to the VRE fare from Woodbridge to downtown DC.

• Ferry Headway – Forecasting was performed for two different ferry headways (a 30-minute ferry headway and a 20-minute ferry headway for 2000 and 2010).

• VRE Headway – The existing VRE services result in an effective headway of 28 minutes. This headway was used for all year 2000 forecasting scenarios. However, VRE projects for year 2010 operations would yield a 20-minute headway. To understand how ferry patronage might be affected if the VRE plans are not implemented, both 20- and 28-minute VRE headways were tested for the 2010 forecast scenarios. No changes to PRTC bus services were assumed.

Different combinations of the aforementioned factors were constructed to form testing scenarios for 2000 and 2010. Six scenarios were tested for 2000 and eight scenarios for 2010 (the additional two scenarios in 2010 investigate the 28-minute VRE headway). A summary of the forecasting scenarios is found in Table 4-4. In addition, the projected ferry patronage for the two ferry routes is also listed. More detailed data on the results of each of these scenarios is located in Appendix 4-3.

As shown in Table 4-4, the projected ferry patronage depends on the assumed conditions. Forecasted ferry ridership is low when the ferry travel time and fare are higher than travel time and fares for the competing modes. For example, under Scenario 1, the VRE headway and fare are superior to that of the ferry; and the ferry travel time is longer under the "restricted" speed conditions in this scenario. One would not expect many persons to shift to a mode that would lengthen their commute time and increase their commuting costs. In addition, HOV and transit service to and from certain areas is very attractive. However, with ferry service improvements in both fare and travel time, the projected ferry patronage increases with respect to other forecasting scenarios.

One key conclusion of the forecasting effort is that, for the ferry to be competitive, it would need to be exempted from the speed restrictions on the Potomac River.

Table 4-4 depicts a notable decline in ferry patronage between 2000 and 2010 for the Fort Washington to the Pentagon route. This is due to the anticipated HOV enhancements along the Route 210 corridor and improved Metrorail service in Prince George’s County with the extension of the Green Line to the Branch Avenue Metro Station.

As long as the speed restrictions for a vessel operating on the Potomac River remain in place, a ferry service targeted to Virginia residents commuting to the metropolitan core cannot be "high-speed" and, therefore, would not be competitive with existing passenger transportation services. Similarly, a one-way fare of $7.00 is more costly and is not competitive with other transit services, which would result in few riders choosing to travel by ferry. For these reasons, only scenarios with unrestricted ferry speeds and a one-way fare of $4.40 were considered to be reasonable options.

For 2000, only Scenario 3 (with a 30-minute ferry frequency) and Scenario 5 (with a 20-minute ferry frequency) meet these criteria, yielding estimated daily ferry patronage of 135 and 975 persons, respectively.

For 2010, the headway for VRE services is estimated to be 20 minutes. Under that condition, the projected daily ferry ridership is 335 for Scenario 3 and 925 for Scenario 5. The reduction in projected ferry patronage, in spite of a growth in the total number of commuting trips, is a result of the VRE service improvements. Scenario 7 for 2010 is comparable to Scenario 5, except that VRE services remain as they are in 2000. Under these conditions, a ferry operating every 20 minutes, with no speed restrictions and a one-way fare of $4.40, could attract 1,450 daily riders.

Scenario 5 for 2000 and Scenario 7 for 2010 represent the most favorable conditions for the proposed ferry service.

Figure 4-3 portrays the ferry patronage results for these scenarios. This graph illustrates the modes from which the person trips using the ferry are projected to be drawn. A majority (approximately 85%) of persons projected to use the ferry would shift from other transit modes (VRE or bus) under all scenarios shown. In terms of total ferry patronage, an estimated 350 daily trips are projected under the "realistic" conditions in Scenario 3 for both study years. The "optimal" scenario improvements for daily ferry patronage are approximately 1,000 trips in 2000 and 1,450 trips in 2010.

The scope of this study include an examination of the effects of ferry service on automobile trips that traverse two critical bottlenecks in the region:

• The Woodrow Wilson Bridge, and

• The I-95/I-395/I-495 Springfield Interchange.

As shown in Figure 4-3, the impacts of ferry service on auto travel are minimal. Based on the forecasting methods utilized, under the "reasonable" conditions in Scenario 3, it was estimated that less than 200 auto trips would be shifted off the roadways during the AM peak period. This number includes all automobile person trips in the study area, most of which currently use HOV-3+ lanes on I-95. Therefore, the trips through the two areas of interest would only represent a portion of this number.

Lastly, it is estimated that the most promising scenario in 2000 would result in less than 500 daily person trips shifting from auto to ferry service. Similarly, 2010 estimates indicate that less than 750 daily auto person trips will shift from auto to ferry service. Only a portion of these trips would pass through either the Woodrow Wilson Bridge or the Springfield Interchange, and most of them currently use the HOV lanes.