San Francisco’s Fiber-to-the-Premises Study Released

The city of San Francisco has released its long-awaited Fiber-To-The-Premises (FTTP) feasibility study, estimating buildout costs of over a half billion dollars. After a long and enlightening conversation with Joanne Hovis, president of study author CTC, we have a few initial insights to report.

For those of you new to the story, San Francisco’s supervisors voted unanimously back on October 5, 2004, to adopt a resolution urging certain city departments to consider installing broadband facilities primarily for use in city operations. That eventually led to the fiber feasibility study, which not only encompasses city operations, but “services to businesses and residents.”

Some key points:

– The report reviews a range of potential benefits for ubiquitous fiber, which we’d suggest is required reading for anyone questioning the potential business, personal, and civic value of true broadband. (If you’re going to use the “broadband” label for a digital delivery service, my motto is: If you gotta wait, it ain’t.)

– The report recommends an initial pilot in the city’s “enterprise zone,” in the heart of the business district, where city offices and business need are most common. That’s an intelligent way to start, getting the biggest bang for the buck in high-density areas.

– There’s an overview of internal city requirements for high-bandwidth communications, including the Department of Telecommunications and Information Services (which ultimately sponsored the study), and departments such as transportation, public safety & health, and utilities  again, good reference material for others looking for ways in which broadband can enable smart cities.

– A fiber buildout will require boatloads of money if you have to rip up your streets, so any “conduit assets” reasonably uncrowded and undamaged pipes designed to carry cable ‚ are like getting a “bank error in your favor” card in Monopoly. It turns out that in a prescient move rare for any city, San Francisco’s municipal railway department dropped miles of conduit in many places where its ubiquitous electric buses and trolleys run. Ka-ching.

– That doesn’t mean, however, that a large fiber diet is free. Running all that cable through conduit is just one step; linking it all together, getting all the way to the premises of major office and apartment buildings, and filling in the connections to individual homes  wired or wireless ‚ all takes time and money.

– To ensure a more extensible businesss model  and to reduce the City’s risk, a the study recommends checking out ways to initiate public-private partnerships that would allow the City to own the network, in an “open access” model that would encourage multiple providers to manage the infrastructure and sell layered services. That’s in contrast to a retail model, where the City would sell services itself, competing directly against “broadband” incumbents and newcomers – and which the report projects could cost 50% more to build.

– Looking at the reference case studies in the report, it’s clear that San Francisco is once again pioneering something new. There are woefully few domestic buildouts of any scale to compare to (save Utah’s in-progress Utopia), so it’s necessary to look around Europe for successful fiber initiatives.

– The ultimate bill? The study says it projects around $564 million if an open access model is used. But since Murphy’s Law could have been written into California’s state charter, every major public initiative here in recent memory has cost far more than initial projections, let’s call it a cool billion, obviously spaced out over a long time, if you’re going to fiberize all of San Francisco.

Is that a lot? I’d say no, given all of the value that a blindingly-fast fiber backbone and premises connections would bring. We can’t even imagine the kinds of applications this kind of network could support (though looking at Seoul’s high-bandwidth infrastructure should provide some interesting ideas).

Why, you might be tempted to ask, if San Francisco wants to run fiber everywhere, the City would install a wireless network first? Simple: Fiber takes time. Lots of time. Wireless is comparatively far easier and cheaper to install. And when a fiber network is finally in place, both networks are completely complementary, since fiber can’t do mobility, and local mesh wireless can’t cost-effectively do multi-gigabits per second (yet).

We’ll offer more posts on this groundbreaking initiative in the future.


  1. Kimo Crossman says:

    Watch the video (windows media streaming) of the presentation on Fri 1/26/07 here:


    Further hearings to be scheduled.

  2. Luther Ismaila says:

    Does anyone have the city of San Francisco fiber to subscribers studies report? Can you tell me where to get a hold of the report?