Update to the Update: the 1999 San Francisco Regional Ferry Plan Update, reviewed in the context of a new plan for a Berkeley ferry

This is the Berkeley portion of the 1999 San Francisco Regional Ferry Plan Update, prepared for MTC by Pacific Transit Management Corporation. It was published in Bay Crossings in February 2000, and can be found on the web at www.baycrossings.org.

While the report is in general supportive of a Berkeley to San Francisco ferry, some of the assumptions and conclusions need to be modified in the light of the different approach taken by the low-speed 149-passenger proposal.

My comments are indented and preceded by initials [pk] .

The SF Regional Ferry Plan Update will be indicated by the initials [rfpu] .


[rfpu] The 1992 Regional Ferry Plan prepared for MTC evaluated 17 potential new ferry routes throughout the Bay Area. Of the potential new routes, four stood out as the most feasible and were recommended to receive further analysis. The 1999 San Francisco Bay Area Regional Ferry Plan Update re-evaluated three of these potential routes - the fourth (Alameda Harbor Bay Island to San Francisco) has been implemented. Once again, a ferry from Albany or Berkeley to San Francisco stood out as a route with a high potential for successful transit service.

The proposed route would operate between Berkeley and San Francisco. The Regional Ferry Plan investigated three sites: the Berkeley Pier, the Berkeley Marina, and the foot of Gilman Street near Golden Gate Fields. Distance is about six nautical miles (a nautical mile is 1.15 miles) from the pier, six plus miles from the marina, and just under seven nautical miles from Gilman Street. All terminals will require a dredged channel for approximately two nautical miles.

Terminal Location, Access, and Facilities

Terminal options described in the 1992 Regional Ferry Plan were considered in relation to the City of Berkeley's adopted Waterfront Plan, the existing traffic conditions to and from the Berkeley waterfront, and the waterside constraints and opportunities.

Dredging would be needed for each option, to create a channel 10 feet deep at mean lower low water and about 75 feet wide. Dredging requirements, which would be somewhat higher at Gilman Street, have been reviewed with Bay Conservation and Development Commission staff that did not foresee conceptual problems. Some amount of maintenance dredging would be required, which would constitute ongoing maintenance expenditure.

[pk] There is no need for a channel this deep. Ferries of the type contemplated by the low-speed proposal (18-knot 149 passenger multihull or monohull) typically require about four feet of water. Even allowing for the very infrequent -2.0 ft tide levels, six feet of depth at the MLLW tide datum is all that is required.

The Berkeley Marina location already meets this requirement, with controlling depth of 6.7 feet according to a recent survey. Some dredging would be required at the foot of Gilman Street, but this is at least an order of magnitude smaller in scope than the suggested ten foot deep channel. Note that a ten foot channel would have to extend a mile or more into the Bay to reach water of similar depths. The six foot deep channel, even originating from Gilman Street, only has to go about three thousand feet.

Other technologies could be brought to bear to reduce the draft requirement further. (See my paper about Surface Piercing Propellers at http://www.well.com/user/pk/SPAprofboat.html)

[rfpu] In 1978-79 and 1989, emergency ferries operated from a terminal in the southeast corner of the marina -- at the intersection of University Avenue and Marina Boulevard. This terminal has parking limitations, with much of the paved area used for parking for existing uses. Operationally, ferries operate slowly within the marina to prevent damage to other vessels and to limit wake.

While the Berkeley Pier was historically used for ferry service, it is now used for fishing and recreational uses that would be incompatible with ferry service.

[pk] Not clear why these uses are incompatible. The pier is 3,000 feet long, and a ferry terminal couldn't possibly occupy more than a small portion of it.

[rfpu] The distance from available parking would also be more than 500 yards.

[pk] A ferry terminal at the base of the Berkeley Pier, on the south side, would be about 200 yards from the north end of the parking lot serving Hs. Lordships restaurant. This parking lot is the only parking area in the Marina with excess capacity, and it is notably empty during weekday commute hours. Capacity of this lot, including the parking on Seawall Drive south of the pier, is 410 cars. There is also a considerable amount of overflow parking serving the South Sailing Basin, Skate's restaurant, and docks L, M, and N. Peak demand hours for this parking generally miss the peak demand hours for ferry parking.

[rfpu] At Gilman Street, automobile parking could be accommodated on the existing overflow parking for Golden Gate Fields, south of Gilman. Each of the sites would require a new floating dock and gangway to operate ferry service.

[pk] The need for floating docks depends on the design of the ferry. For a modest service proposal, there is very little duplication of hardware if the gangway is on the boat instead of on the dock. (As the number of ferries increases, it becomes more economical to put the gangway at the terminal and build a more expensive float.)

[rfpu] Waterfront Plans. Berkeley is currently updating its Marina Master Plan. The document will recommend and plan for upgrading and enhancing existing facilities, trails, signage and boating facilities. The current Berkeley Waterfront Plan calls for limited development throughout the area. Specifically, the policy calls for development only in parcels adjacent to Gilman Street. However, much of the Berkeley Marina privately owned property is being purchased by the East Bay Regional Park District on behalf of the State Park System B to be included in the Eastshore State Park. The Park District expects to also develop a master plan for the property it is purchasing but has not yet initiated such a study due to a lack of funding.

[pk] This is all in a state of rapid flux as of September 2001. Two examples: 1) There is underutilized commercial property in the Marina that would very likely be revitalized by a nearby ferry service (Hs. Lordships Restaurant, Dock of the Bay building). 2) Possible accommodation of the parking demands of the Eastshore State Park might result in fairly large lots able to serve ferry passengers during weekday commute hours.

[rfpu] Traffic and Access Issues.

Of the three sites, the Berkeley Pier and the marina have the same access routes, while Gilman Street serves a slightly different catchment area and has different access.

Marina/Pier Access

Primary access is via University Avenue and West Frontage Road, and the I-80 interchanges at University and Ashby Avenues. University Avenue, the primary access route to a marina/pier terminal, has generally poor levels of service in both peaks, but the afternoon period is worse than the morning. University Avenue westbound traffic on weekends exhibits especially poor conditions. The marina is well served by AC Transit with Line 51M operating every 20 minutes during weekdays.

Gilman Street Access.

Primary access is via Gilman Street. The Gilman Street I-80 interchange provides direct access from vehicles entering and exiting I-80 in both directions. There is currently no transit service to the proposed Gilman Street ferry terminal, although AC Transit Lines 9 and 52 operate within a few blocks of the site.

[pk] If the primarily target market is the driver-commuter, then any evaluation of local traffic effects should include the impact of taking cars off the Bay Bridge and approach routes.

Making the conservative assumption that the Marina access road (lower University Ave., two lanes in each direction) can accommodate 1,000 cars/hour, then each departure of a 149 passenger ferry uses the capacity of the access road for only nine minutes.

[rfpu] Market Potential

The model, when analyzing other routes, predicted a 25 percent increase in patronage as service levels doubled, and 50 to 60 percent patronage increase as service levels tripled. Using these same formulas results in the patronage shown below.

On racetrack days, total patronage could be 1,500 to 1,600 passengers with peak trips every 20 minutes. The basis for the model's predictions is the 1990 census data, which while several years old is still the most useful information available and is consistent with MTC's current Berkeley to San Francisco work trip estimates. Peak period patronage is assumed to be primarily composed of North Berkeley and Albany origins, with some Richmond and El Cerrito residents also likely to use the service.

[pk] The 1990 census data misses more than a decade of West Berkeley yuppification, the upscale development of Fourth Street, and other factors that make the market analysis conservative. On the other hand, race track patronage has declined sharply in recent years, possibly due to the advent of online betting alaternatives.

[rfpu] It would appear that based on the existing transportation market, ferry service could be competitive. Ferries would travel to San Francisco in about 20 minutes, compared to 25 minutes from North Berkeley BART and 29 minutes from El Cerrito Plaza. Bus travel times range from about 25 to 30 minutes from various areas of Albany and North Berkeley. When access times are included, all the modes have about the same travel time to San Francisco.

[pk] Note that the 18-knot ferry proposal implies a 30 minute trip from the Berkely Marina (20 minutes at speed, ten minutes for docking).

[rfpu] After the Loma Prieta earthquake, the Berkeley ferry carried up to 500 passengers during the morning commute period and about 1,600 passengers daily with a 22 trip schedule. However, when the Bay Bridge reopened, patronage fell to about 500 to 700 trips daily, and then to 400 to 500 trips daily. As a result, the Berkeley ferry was eventually discontinued. The slow speed of the emergency service (12 knots) and travel time (45 minutes) provided a poor model of what ferry service could be with high-speed vessels and a visible terminal.

With its outstanding freeway access and potential large parking lot, a Gilman Street ferry terminal could also be used to provide service to Treasure Island for both the permanent development at the site and for special events, and to China Basin for Giants baseball games and special events. Recent special event ferry service to Treasure Island overwhelmed parking capacity at Jack London Square, so a second convenient ferry terminal in the inner East Bay with parking capacity would be beneficial.

Vessel Type

The short six to seven mile route length between Berkeley and San Francisco requires a vessel to travel at approximately 25 knots (about 30 mph). Increasing speed to a 30 knot vessel (35 mph) would only shorten the sailing time by about three minutes. There is a large fuel consumption penalty associated with higher speed vessel operation which is not warranted given the small time differential associated with higher speed on this potential route.

[pk] For the same reasons given, it can be argued that speed above 20 knots is not necessary. Lower speeds are increasingly attractive considering renewed challenges to ferry proposals based on environmental and energy considerations.

[rfpu] There are several satisfactory vessel types and the vessel selected depends upon service frequency and demand. Should an hourly service be recommended, then one 250 passenger catamaran would be an appropriate selection. This could be a vessel such as the M.V. Bay Breeze, a 29 meter 25 knot vessel now used on the Harbor Bay Isle route. Such a vessel would cost about $4-4.5 million.

[pk] This suggests that the estimated $4 million for a single 18-knot 149 passenger ferry is extremely conservative. It allows enough margin for a spacious boat with passenger amenities (work tables with AC power) and ample room for bicycles.

[rfpu] Another alternative is the purchase of several 150 passenger vessels. This would allow more frequent service. Under this scenario, either catamarans or fast monohull vessels could be purchased.

[pk] 149 passengers, that is. Coast Guard fire resistance regulations, among other things, create a dramatic bump in construction cost when the 150 number is crossed.

[rfpu] Financial Analysis

Different service alternatives were considered for a Berkeley/Albany ferry. One scenario anticipates one 250 passenger vessel providing hourly service to San Francisco. Another scenario anticipates 30 minute service frequencies using two 150 passenger boats. A third scenario anticipates 20 minute service in the peak periods, with 30 minute service in the off-peak. As service increases, patronage increases, but so do costs.

[pk] Seems optimistic, considering that only 500 morning commuters could be found during the weeks following Loma Prieta. Proposed capacity appears to be at least a thouand commuters each morning.

Note that the 18-knot 149 passenger proposal assumes only 300 paid westbound commuters spread over three morning departures to make the economics work. Capacity is only 447 morning commuter seats.

[rfpu] While the project planning stage would determine the appropriate level of service, for this analysis it is assumed that the most expensive capital plan would be used, requiring the purchase of three 25 knot, 149 passenger vessels costing $3-3.5 million each. In addition, terminal facility costs, including dredging, docks, gangways and parking improvements, are projected at about $4 million. This does not include the purchase of land for the terminal and parking lot, assuming it would be provided by the property owner or State Park. Approximately $1.5 million is allocated for planning, design, and contingency. Thus, total capital costs would be approximately $15.5 million.

[pk] Most of these expenditures, beyond the boat itself and some minimal improvements to the existing ferry dock, are not necessary with the single boat 18-knot 149 passenger proposal.

[rfpu] This analysis assumes a fare of $3.50 and hourly operating costs ranging between $325 and $400 (crew, fuel, insurance, management, etc.).

[pk] This is consistent with the estimate of $400 per hour for the 149 passenger ferry - although fuel and engine maintenance costs would be lower due to the reduced speed and power.

[rfpu] The results indicate that hourly service would require a subsidy of about $590,000 annually, 30 minute service would require a subsidy of about $1.3 million annually, and peak service every 20 minutes would require a subsidy of about $1.6 million annually.

[pk] At $9 ticket price, the 18-knot proposal requires no subsidy for construction or operation. The elasticity of demand with respect to this ticket price is an open issue. Note that the Sausalito and Tiburon routes attract riders at a ticket price of $6.75, and the Vallejo route is popular at $9. It would only require a 25% subsidy to get the ticket price of the 18-knot Berkeley ferry down to parity with the Tiburon and Sausalito services.


INDEX --- POLICY AND PLANNING --- WATERFRONT COMMISSION --- MARINA INFO --- YELLOW PAGES -- - CALENDAR