Kirkland-to-UW and Kenmore-to-UW Rise to the Top in Passenger-Only Ferry Study

Passengers on Colman Dock with a ferry docked.
The active Colman Dock passenger ferry facility. (Photo by the author)

The Puget Sound Regional Council (PSRC) has analyzed seven passenger-only ferry routes and issued results this month. Two Lake Washington routes–Kirkland to the University of Washington (UW) and Kenmore to UW–rose to the top, showing significant time savings and ridership potential with relatively low operating cost expected. Commissioned by the Washington State Legislature in 2019, the report evaluates demand and costs for new Puget Sound passenger ferry routes. The Legislature authorized PSRC to lead the study due to the agency’s experience with a 2008 passenger-only ferry study, whose identified near-term routes will all be in operation by the end of 2020.

Passenger ferries in the region can be traced back to Native American water transportation and the Mosquito Fleet that operated steam-powered small craft in the late 1800s and early 1900s. Passenger services were converted to vehicle ferries in the Great Depression era and as car infrastructure ramped up. Passenger-only service was reintroduced to the region in 1986 and today they operate in King and Kitsap counties, offering faster trips than more lumbering vehicle-laden ferries.

This round of research reviews the 12 county area–Clallam, Island, Jefferson, King, Kitsap, Mason, Pierce, San Juan, Skagit, Snohomish, Thurston, and Whatcom–surrounding Puget Sound, with a focus on evaluation of new routes, landing sites, potential demand, estimated capital and operating cost, and recommendations on electrification. The process for the planning and feasibility-level study has now gone through a four-step screening process to evaluate, rank, and whittle down a field of 45 routes down to 36, 18, and now seven.

Evaluating the Routes

When The Urbanist last reported on the study, PSRC had only screened the numbers of routes to just eight with the eighth route still to be decided between Suquamish to Seattle or South Whidbey to Everett. The agency has since selected the second option for the draft study, and eliminated Gig Harbor to Seattle from further analysis. The remaining routes were put through the latest set of analysis and assessment that include travel time savings, capital and operations costs estimation, and ridership demand analysis.

Routes analyzed in most depth are:

  • Tacoma to Seattle
  • Bellingham to Friday Harbor
  • Whidbey to Everett
  • and four Lake Washington/Lake Union Routes
    • Kenmore to the University of Washington
    • Kirkland to the University of Washington
    • Renton to the University of Washington
    • Renton to South Lake Union

Mock route profiles are developed for all seven of the PSRC’s chosen routes to generate figures for cost and ridership, and detect hurdles and opportunities. Outside of Bellingham-Friday Harbor, which is framed as a seasonal recreational route operating seven days a week with four roundtrip a day, all the routes are given a hypothetical commute-focused, year-round, and five days a week in operation with six departures a day service.

For PSRC’s hypothetical routes, financial analysis found annual operating cost ranging from two to four million dollars. Operating costs are a reflection of route length that causes higher operating, maintenance, and fuel costs. High capital costs for Bellingham-Friday Harbor and Tacoma-Seattle can be attributed to the need for a larger vessel for seaworthiness and passenger comfort. Routes like Tacoma-Seattle and Renton-South Lake Union also need an extra vessel due to the length of the routes, further increasing capital costs. Overall, initial investment costs for each route start in the $10 million range and can reach in to the high tens of millions, as terminal costs are given a Rough Order of Magnitude estimate of $5 million to $35 million if a replacement or new build is needed.

While not studied in depth enough to develop terminal costs, the agency does explore hypothetical landing sites for infrastructure and regulatory needs. Multimodal connections and local context are also inspected. For example, the UW stop that serves three of the routes in the study would need to have its existing dock by the Waterfront Activities Center be replaced to support services. A replacement would have to be compatible with existing uses, the crew team, and recreation boating. Proximity to the light rail and bus service seems like the main attraction of this potential terminal, accessibility improvements are noted as a need for those connections to be realized.

(Courtesy of PSRC)

For ridership, numbers do not project induced demand, and only approximate existing ridership for the corridor and the share expected to be captured by a passenger ferry route. PSRC also expects ridership numbers to increase with development in the vicinity of potential terminals. This is important context for some of the less impressive routes for projected ridership. Especially given their capital and operating costs, Kenmore-UW and Kirkland-UW stand out as potentially capturing the most annual ridership at some of the lowest cost among the routes. Some crude math puts those two routes at $15 and $12 of theorized operating cost per passenger, respectively. The other routes range from $40 to $120 per passenger.

As you might expect, time savings are found with every route chosen to make it this far into the analysis. Compared to bus service, passenger ferry service for the Tacoma-Seattle route and the Lake Union/Lake Washington routes reduced travel times by more than 15 minutes, amounting to a 20% time saving for longer routes like Tacoma-Seattle and Renton-South Lake Union. For a shorter route like Kirkland-Seattle, it nearly cuts travel time by 50%. Non-central Puget Sound routes see much greater time savings, cutting travel times by about 60%.

Lastly, PSRC studied electrification for each route. Only the shorter routes, Whidbey-Everett, and Lake Union/Lake Washington routes are noted as candidates for electrification. Batteries are far less dense than liquid diesel fuel, so the longer routes would require much larger batteries that would be costly and heavy. There is discussion to lower speeds or reduce capacity to make a battery powered ferries more viable, but that damages the main appeals of a ferry service alternative.

Next steps and implementation

PSRC’s December report ends on implementation considerations and next steps. Due to 2006 legislative direction, passenger ferry implementation is to be left to local entities; the agency recommends that the first step toward implementation is incorporating route, associated facilities, and last- and first-mile connections into local and regional planning frameworks. For the examined King County only routes, the duty would likely go to King County Metro that operates ferry services with the water taxi. Routes that travel to Pierce, Island, Snohomish, San Juan, and Whatcom Counties will need to identify what agency to put in charge, be it a port, transit agency, or another body.

Even though the Legislature won’t implement the routes, the state can still support policy work and provide funding options of local jurisdiction to implement passenger ferries. Additional space for the state may also exist in conducting an analysis of cumulative impacts of ferry traffic on other marine traffic and uses and the environment.

The general purpose of the study is to provide information, so PSRC included a second step to develop a business and implementation plan to refine and develop more data for regional passenger ferry service. A full financial plan would be a part of that plan, which would develop a fare structure, identify and develop agreements for landing sites location, and build a realistic implementation schedule.

A final draft of the Puget Sound Passenger-Only Ferry Study is due to the Washington State Legislature in 2021. PSRC is continuing work on the study and you can follow their work on the study website and sign-up to receive updates. The agency is also open to feedback, which can be provided to by December 28, 2020.

We hope you loved this article. If so, please consider subscribing or donating. The Urbanist is a 501(c)(4) nonprofit that depends on donations from readers like you.

Shaun Kuo is a junior editor at The Urbanist and a recent graduate from the UW's Jackson School. He is a Seattle native that has lived in Wallingford, Northgate, and Lake Forest Park. He enjoys exploring the city by bus and foot.

Inline Feedbacks
View all comments

What does the Growth Management Act say about new commuter service that connects rural areas to urban job centers? The existing connection between Clinton and Puget Sound job centers via the Mulkiteo Sounder station has been poor, with WSDOT and ST not coordinating schedules for an efficient transfer, but I never had a major issue with this as Whidbey to Seattle commuters is a fringe use-case for both agencies. I don’t begrudge Whidbey residents good jobs, but expanding the central Puget Sound commute development into a nearby rural county seems like somethings the GMA would regulate.

Daniel Thompson

Ferries make sense when there is no bridge. Still they operate at a loss and require a public subsidy, but the subsidy is less than the cost of a bridge. 520 with tolls operates at a small subsidy (far less than the 80% subsidy buses receive). The plan to electrify the ferry fleet is economically unwise. If ferries made any sense across Lake Washington a private operator would be running them.


“If ferries made any sense across Lake Washington a private operator would be running them.”

I think that statement is a little bit too strong. If ferries could be operated at a per-passenger subsidy level similar to buses, I would say “fine”. After all, if turning a profit were required, we wouldn’t have any buses either. But, the point is that the per passenger subsidies are vastly higher with a ferry than with buses, while offering essentially no travel time advantage.

“The plan to electrify the ferry fleet is economically unwise.”

I know governor Inslee had plans to electrify the car ferries across Puget Sound. For that, I’d have to see a full cost/benefit analysis. You have to consider the cost to electrify each boat, whether the diesel engines are nearing their end-of-line and would have to be replaced anyway, the savings in operating costs by using electricity instead of diesel fuel, and the cost of installing rapid chargers at the ferry docks. Then, you’d have to take the effective “net cost” of ferry electrification and compare that against what else could be done with the money to reduce carbon emissions from some other sources. For example, the money could be used to make buildings more energy efficient, or be heated with electricity instead of natural gas. It could also be used to electrify cars, trucks, or buses. Or even, simply building more wind and solar power to sell to other states that are still generating electricity with coal.

If the ultimate goal is staving off climate change, the state should be focusing on maximizing the ratio of carbon emissions saved to taxpayer dollars spent – whether the emissions are on state property or private property doesn’t really matter because, at the end of the day, the earth only cares about the total. My intuition is that the most efficient way to reduce carbon emissions is not electrifying the car ferries, but I do think it’s at least worth crunching the numbers and see what the cost/benefits are. To an extent, fighting climate change has some “low hanging fruit” elements where, once the most efficient ways to reduce carbon emissions are all tapped out, further, more costly reductions may still be necessary to get the total emissions below the necessary threshold. So, even if electrifying the ferries doesn’t make the cut today, it might in 10-15 years if the easier stuff has all been done and it turns out to not be enough.


The transit travel times look very suspect to me. For example, they list 47 minutes from Kenmore to the UW. Google puts it at 38 minutes (during rush hour, peak direction).

But that isn’t the biggest flaw with the studies. For the trips within Lake Washington, they failed to look at real alternatives. The savings largely come from eliminating stops. An express bus from Kenmore to the UW would be significantly faster than the boat. The same is true from Kenmore.

But that begs the question — why isn’t there an express? The simple answer is, it isn’t worth it. It would be great for a handful of riders, but very expensive. It ends up favoring a small group of riders, while hurting others. This is essentially a plan for a very expensive, very slow set of express buses. If you can’t fill express buses, then you won’t fill boats.

Ferries are most successful when they have a huge geographic advantage over alternatives. That simply isn’t the case with Lake Washington. The other situation in which they can be very effective is when you have massive numbers of people close to each dock. Again, that isn’t the situation here. The only dock that has lots of people very close to it is downtown Seattle. The UW has lots of people, just not by the eastern shore (where the ferry would dock). The bigger problem is that there aren’t that many people in Kenmore or Kirkland. For perspective, this is North Vancouver, by the ferry: The ferry docks right in the heart of downtown Vancouver, and connects very easily to SkyTrain. Obviously there are a lot more people there than the docks in Kenmore or Kirkland. They can’t possibly fill the boats, which is why they don’t run express buses. There just aren’t enough people.

You can see this flaw in what they considered the market area (page H11 and H12). For both Kirkland and Kenmore, they considered areas that are clearly not within normal walking distance. I’m not sure what the assumption is — are people supposed to take a shuttle bus? Why would they do that, instead of taking a bus right to the UW?

There just isn’t the demand for these trips. Partly because there aren’t that many people who would walk to the docks and partly because it doesn’t offer a huge time savings over alternatives.


Great point – the ‘no build’ option isn’t the existing transit service; it is a point-to-point express using the existing road/bridge network.


Can apply the same logic to the Whidbey to Everett ferry. For the cost of a new boat, just run shuttle service from the Mulkiteo ferry station to nearby job centers. Less convenient if you are heading to downtown Everett, but not much different if you are heading to the Paine Field industrial center, where there are more jobs. If Island Transit wants to coordinate with major employers or ET/CT so that Island County residents have better access to good jobs, that could be a good public investment for WSDOT or Island County, but none of this requires a new ferry route.


The 255 gets you from Kirkland Transit Center to UW Station in about 15-20 minutes, with an HOV lane to bypass traffic on the freeway. Soon, there will be an HOV ramp to bypass congestion at the Montlake exit ramp, also.

I once rode from the Kirkland Marina to Lake Union in a boat with some people from work, and it took at least 15-20 minutes, if not more, to reach the Montlake bridge. So, I’m having trouble seeing where the time savings is coming from, outside the worst traffic days. And that’s not even accounting for the difference in wait time, since the ferry can never run as often as the 255 with remotely reasonable operating costs.

The estimated operating cost of $12-$15 per passenger boarding is also well above the operating cost per passenger on the 255. Only extremely long Sound Transit routes, like the 594, have an operating cost per passenger in this range (and even then, only during low-demand periods when the bus isn’t full).

From a climate perspective, running a diesel-powered Kirkland->UW ferry is not helping things, with the pollution from a typical passenger ferry being no less than every person on board driving the route in a separate car (even if the ferry is full). The buses pollute too, but they’re running anyway, and the marginal climate impact of filling in empty seats is essentially zero. In 10-15 years, the 255, along with most other Metro routes, will be electrified anyway.


Exactly. Just buying the boats is $12M each route (+TBD dock construction cost). Funding the Lake City-Kenmore RapidRide and adding a 255 express is way cheaper than $24M. Operationally, those are probably 1/3 the cost of ferries as well.

Plus, the report only recommends 6 round trip ferries a day only on weekdays and the ferry is only 1 min faster UW-Kenmore and 3 min faster UW-Kirkland than bus.

This is an awful idea.


I agree. It simply doesn’t add up. I will say, though, that the article suggested electric ferries. But again, it isn’t worth the money. If your goal is to cater to a handful of riders in Kirkland and Kenmore, then you should just run express buses. I see no reason why we should single out those areas though, for special treatment. There are much better ways to spend out transit money. This is a bad idea.