Ferries For The San Francisco Bay Area; New Paradigms From New Technologies


Chris Barry (Member), Bryan Duffty (Member), and Paul Kamen (Visitor)


Presented at the joint meeting of the Society of Naval Architects and Marine Engineers, Asilomar, Pacific Grove, California


May 31 2002



Appendix A:

A Detailed Plan for a Berkeley to San Francisco Ferry Service


"Even a journey of a thousand miles begins with a two-hour schlep to the airport"

Chinese proverb (slightly modified)


Figure A-1    Jacob's Landing in the late 19th Century





Ferry service from Berkeley to San Francisco began in 1851 when James H. Jacobs built the wharf at the mouth of Strawberry Creek, and ended in 1956 when the last scheduled ferry left the end of the Berkeley Municipal Pier. The Bay Bridge (1936), the proliferation of the private auto, and finally transbay BART service (1973) all made ferries redundant and obsolete.


Temporary ferry services have operated during a BART strike (1979) and during the closure of the Bay Bridge following the Loma Prieta earthquake (1989).


But the boats were old and slow, the docking arrangements inside the Berkeley Marina were time consuming and the ridership could not be maintained.


Even in the weeks following the Loma Prieta earthquake with the Bay Bridge closed, there were never more than 500 morning commuters taking the ferry.



What the ferry can and cannot do


It is recognized that a Berkeley ferry is not likely to significantly alleviate vehicular congestion or improve regional air quality. It is primarily an amenity and an alternative to other forms of private and public transportation. However, because it will be possible to operate this ferry service without significant public subsidy, funds will not be redirected from other modes of public transportation. The probable lack of measurable impact on Bay Area traffic congestion is therefore not a valid argument against the ferry service proposed here.


What the Berkeley ferry will do is increase mobility. Ferry service has become desirable again because of the chronic congestion on the Bay Bridge and the saturation of the BART system. For example, for those who do not live within walking distance of a BART station, after about 8:00 am when the BART parking lots fill up there are essentially no options but to drive to San Francisco. Crossing the bridge by car on a weekday morning (or Saturday evening) is unpleasant and very slow, and parking is expensive. The loss of the ability to move about the inner Bay Area freely and comfortably has a high economic cost, and ultimately adds to the pressure for suburban sprawl.



The Boat


Design evolution begins with the selection of the number of passengers. See figure A-2, which shows passengers per crew as a function of number of passengers. The first local maximum, at 149 passengers, is chosen as the only economical point that is consistent with modest ridership projections. Also, a significant increase in first cost is incurred at larger capacities due to more rigorous construction standards imposed by the Coast Guard for vessels carrying 150 passengers or more.


Figure A-2   Passengers per crew v. number of passengers


The Coast Guard allows a crew of two for a single deck design, but one additional crew is required for each additional passenger deck. Therefore a single-deck design is strongly indicated. There is also motivation for large deck area to accommodate bicycles, dogs and possibly electric scooters.  


Speed is determined by the distance and the required transit time. 17 knots is sufficient for a single boat to provide hourly service over the 5.6 mile route, with 20 minute transit time and ten minute turn-around at the terminals. 18 knots is the specified service speed, to allow for adverse tides.


Figure A-3 shows the sensitivity of required speed to terminal turn-around time. This implies multiple boarding points, whether docked bow-in or side-tied, to facilitate very rapid loading and unloading.


Figure A-3:   Required speed and power v. turnaround time


The ride must be quiet and comfortable, with vessel motions comparable to larger vessels. The seating arrangement should be spacious, allowing most passengers the choice of a work table or a good view.


Fuel efficiency must be extremely high and emissions must be extremely low.


To summarize the primary requirements:


Single-level passenger area

Large deck area

Modest speed

Spacious interior

Ride comfort

Fuel efficiency


All these requirements point to a multihull with very long and very slender hulls operating in displacement mode.


An asymmetrical proa configuration is attractive because of the longer overall length for reduced motions and better fuel efficiency, and for reduced cost due to the consolidation of machinery into a single engine and driveline. However, figure A-4 shows that in the power range under consideration, there is no significant machinery cost saving available by using one large propulsion system instead of two smaller ones. This is because the economies of scale have made diesel engine in the 300 hp range very inexpensive compared to smaller and larger power ratings.


Figure A-4   Cost per brake horsepower (BHP) v. horsepower



The vessel proposed is a 149-passenger single-deck catamaran (with seats for 180) designed to operate at a speed of 18 knots. It will be powered by two diesel engines of approximately 350 hp each and operate with a crew of two.

18 knots is relatively slow by modern standards, but it is fast enough to cover the 5.6 mile route in less than 20 minutes. This is barely enough time to buy a latte, turn on the cell phone and open up the computer. That is, a faster trip will not be any more saleable than the proposed 20-minute timing.


Fig A-5   The proposed conceptual ferry design


More important is efficient ticketing and very fast loading and unloading. This will have a greater effect on trip time than a few knots of additional speed. This requirement dictates the open side decks, in addition to the deck areas forward and aft of the passenger cabin. Whether moored bow-in or on a side tie, there will be deck area to accommodate multible gangways.


Low hull volume also dictates the very wide beam, necessary to meet stability requirements.


Emissions controls will be similar to those used on modern busses. Because of the low speed, the horsepower per seat is comparable to that found on a city transit bus.



Location of the terminal


This proposal calls for a ferry service that more closely resembles the service of 100 years ago. Rather than use the existing ferry dock inside the Marina, there would be a new terminal in the open water alongside the Berkeley Fishing Pier. This location has several compelling advantages:


1) It is the closest to San Francisco. The 5.6-mile route will only take 20 minutes at a modest speed of 17 knots. This is about a mile closer than other suggested locations at the foot of Gilman Street or Fleming Point (behind the race track).


2) There will be no wasted time maneuvering in and out of the Marina.


3) Little or no dredging is required, unlike the foot of Gilman Street or at Fleming Point that would require extensive dredging projects.


4) The Berkeley Pier is at the end of the AC Transit 51 M bus line.  Because this is the terminus of a major trunk line, there is very frequent service (about every 20 minutes all day), yet ridership is low because it is near the end of the line. It is a natural location for an intermodal transfer. Rather than requiring new bus service, this location would make use of existing excess capacity at the end of an existing route.


5) There are 410 parking spaces in the parking lot serving Hs. Lordships restaurant and along Seawall Drive south of the Pier. These are mostly unused during the week. Other overflow parking areas are nearby in the Marina. Peak demand hours for ferry parking generally miss the peak demand periods for Marina parking, so this would also take advantage of existing infrastructure rather than requiring new facilities.


6) There is existing nearby commercial activity (unlike the proposed Gilman Street location, where commercial development is in question). Commercial activity makes the ferry terminal a more attractive place to wait or to meet, and the presence of the ferry is likely to increase public revenue from the nearby businesses.


7) The ferry route would not have to traverse any part of the Eastshore State Park. Sierra Club and other groups have opposed commuter ferry service to any location in Berkeley or Albany on the grounds that it would not be compatible with the Eastshore State Park. The Berkeley Pier location is furthest from the park boundaries, and does not traverse any park tidelands.


Figure A-6   The East Bay shoreline in 1883, showing the ferry terminal at Jacob's Landing at the foot of University Avenue.



The Politics


The Sierra Club is on record as strongly opposing a commuter ferry operating from anywhere along the Berkeley or Albany waterfront. This opposition appears to be based primarily on the prospect of a larger and faster ferry operating from the foot of Gilman Street, an area that the Sierra Club hopes will eventually be acquired by the State and become part of the Eastshore State Park.


These considerations should have no relevance to the this proposal, which is well removed from the Eastshore State Park land and tidelands.


Sierra Club has, however, raised the objection to ferry-related traffic transiting the Eastshore State Park on University Avenue. The numbers, however, don't support the claim that this would cause a detectable qualitative change in the traffic level. With a maximum capacity of 149 passengers and three departures every morning, it is hard to imagine how the four lanes of University Avenue would become congested. Considered in the context of the number of seats in existing waterfront restaurants, activities at other Marina businesses and offices, and the number of boat berths served by those same roads, the traffic argument becomes specious.


If we conservatively assume a capacity of 500 cars per hour per lane, and three full ferries with every seat representing another car, then we have only used 26 minutes of road capacity for the entire morning commute.


On the other hand, several interest groups can be expected to offer strong support for a Berkeley ferry.


The bicycle groups - East Bay Bicycle Coalition and Bicycle Friendly Berkeley - are ferry advocates. BART is closed to bicycles during commute hours, while a ferry with a large bicycle deck promises the efficiency of a true dual mode personal/public transportation system.


Handicapped advocacy groups are expected to show favorable interest. Ferries are spacious and easy to use, and offer safe mobility for people with a wide range of disabilities.


People who like to travel with their dogs are likely to be strong ferry advocates. There is no reason to preclude dogs from the outside deck areas of a ferry operating on a short route.


As we have seen from various public planning processes, the combination of bicycles, dogs, and the handicapped comprises a formidable lobbying force. Even in Berkeley, the Sierra Club would be ill advised to oppose it.



The Fare Structure


If we make some very conservative assumptions about costs, the back-of-the-envelope economic analysis suggests that the required break-even no-subsidy ticket price is $6.28 each way. This covers vessel construction and operation, and assumes 2/3 full on the forward commute, 1/3 full on the reverse commute, seven round trips per day, and about $400/hour for total operating expenses.


$6.28 is steep, but this is less than the ticket price on the Vallejo ferries and comparable to the $6.75 charged for the Sausalito and Tiburon ferries. It compares very favorably with the cost of parking in downtown San Francisco for the day, or the cost of a trip on an airport shuttle. As long as the boat is small and the service is not too ambitious, there is probably a sustainable market at this price point.


However, the market-rate ticket exposes the ferry proposal to charges of elitism, and the criticism that the service is designed to serve only the rich. There is some truth to these charges, so the strategy proposed here is to implement a policy analogous to that in effect on the major toll bridges. On the bridges, carpools go free. The rationale for this policy is that they tread lightly on both the environment and the transportation infrastructure, and help alleviate the congestion of single-occupancy vehicles.


On the ferry, anyone arriving by bicycle and bringing it aboard would ride free. The rationale is the same as on the bridges. This would insure that the ferry remains accessible to the widest possible range of income levels, without relying on public funds or complicated and inefficient subsidy schemes. It would also immensely simplify ticketing and boarding for bicyclists.


This scheme is of course experimental, and its success would depend on both its popularity and the level of abuse that could not be prevented by the obvious countermeasures. The free ride might have to be replaced by a deep discount after some operational experience is obtained.


Discounts would also be offered for the "reverse" commuters, SF to Berkeley in the morning and westbound in the afternoon. Parking is scarce and expensive at the San Francisco ferry terminal, so the discount could be justified on similar environmental/infrastructure grounds, if not on simple supply and demand. Reverse commute discounts would be extremely valuable to a significant number of students at U.C. Berkeley, as well as some City employees



The Schedule


Because of the low subsidy level, the schedule must be driven by cost and revenue considerations. As long as there is a place to park at BART, the ferry can never hope to compete with BART for level of service as measured in convenient access to the terminals or stations.


But BART parking lots fill up early in the morning. They are generally full by 8:00 am on weekdays. Although a small number of spaces become available at 10:00, and can be reliably accessed between 10:00 and 10:05, for all practical purposes BART is not an easy option after 8:00 am. Certainly there are many people who can walk, bike, or bus to the BART station. And while we acknowledge that these people hold the moral high ground, for the majority of riders the only viable means of travel to the station is access by private auto.


There is no point, then, in scheduling a ferry departure before 8:00 am. With 20 minute transit time an hourly schedule is feasible using one boat, so the proposed morning departures would be at 8, 9, and 10.


After the 10 am departure it is assumed that the demand will fall below the level that would support an hourly schedule. The problem becomes one of keeping the boat busy during the mid-day lull. Assuming the crew's work day has started at 7:30, the crew has only been on watch for three hours when the ferry is finished unloading at the Ferry Building at 10:30.


We propose the route topology shift from part of the 'hub and spokes" system centered around SF to a circular route around the central bay, keeping the legs short enough to be compatible with the relatively slow speed.


After leaving the Ferry Building, the ferry would begin a clockwise circuit to Sausalito, Tiburon, Larkspur and Richmond, returning to Berkeley in time for a 1:00 pm departure.

The next loop would include a stop at Treasure Island, then SF, Sausalito, Tiburon, Larkspur and Richmond, returning to Berkeley for a 4:30 pm departure.

The 4:30 pm Berkeley departure would be in SF in time for a 5 pm return commuter run straight back to Berkeley. Hourly departures would continue at 6, 7, and 8 pm.


The boat is finished for the evening at 8:30 pm, and the crew could probably punch out at 9. This is a 13.5 hour day, presumably done as one 7:30 am  1:00 pm shift of 5.5 hours, and one 1:00 pm - 9:00 pm shift of eight hours.


The mid-day circular routes provide one trip from the East Bay to Treasure Island, and two trips each to the three terminals in Marin. This would fill an important gap in transit mobility, as there is currently no easy way to travel from the East Bay to Marin.


Note that there are four evening commuter return trims, but only three morning trips. This apparent imbalance is to account for two factors: 1) a significant number of morning commuters will be pulled away by informal carpools, attracted to the HOV lane to the Bay Bridge. Although other transit agencies have resisted and discouraged this practice, the ferry service should recognize the value of the informal carpool phenomenon and accommodate it as much as possible. 2) There is typically a much wider distribution of return times than departure times. People work late, run errands, and have other reasons to delay their return trip, suggesting the four-trip three-hour return window.

Here is what the tabulated schedule might look like:



          Treasure   San                                       

Berkeley   Island  Francisco Sausalito Tiburon Larkspur Richmond  Berkeley


 8:00                8:30                                          9:00

 9:00                9:30                                         10:00

10:00               10:30    11:00     11:30   12:00     12:30     1:00

 1:00       1:30     2:00     2:30      3:00    3:30      4:00     4:30

 4:30                5:00                                          5:30

 5:30                6:00                                          6:30

 6:30                7:00                                          7:30

 7:30                8:00                                          8:30



On the other hand, the public might be better served by a simpler schedule, with hourly service all day:


Berkeley   San Francisco   Berkeley



 8:00         8:30           9:00

 9:00         9:30          10:00

10:00        10:30          11:00

11:00        11:30          12:00

12:00        12:30           1:00

 1:00         1:30           2:00

 2:00         2:30           3:00

 3:00         3:30           4:00

 4:00         4:30           5:00

 5:00         5:30           6:00

 6:00         6:30           7:00

 7:00         7:30           8:00







First cost of ferry: $4 million (very conservative - other estimates are below $1.5 million for a ferry that meets these capacity and speed specifications).


Operating cost (crew, fuel, maintenance, admin): $400/hour (also conservative, probably based on a crew of three).


Port fees in San Francisco: $20,000/year.


Terminal construction in Berkeley: $2 million (this is a wild guess, but probably conservative if no breakwater construction is required).



2/3 capacity (99) full fare passengers on seven forward commutes.

1/3 capacity (50) full fare passengers on seven reverse commutes.

1/3 capacity (50) full fare passengers on one mid-day circle route.

12.5 operating hours/day, 255 commute days/year.




Total fares/year = 255 x (149 x 7 x 2/3 + 149 x 7 x 1/3 + 149 x 1/3)

  =  278,630 full fares/year.


Capitalization = $6 million x 0.1 = $600,000/year.

Operation = 255 x 12.5 x 400 + $20,000 = $1,150,000/year.

Total = $1,750,000/year.


Required fare for unsubsidized break-even = $1,750,000/278,630

  = $6.28.



Closing quote


 "Nothing is impossible for the man who doesn't have to do it himself."

A.H. Weiler