Vermont gets serious about FTTH

by Meghan Fuller Hanna

At press time, 23 municipalities in a region of east central Vermont known as the Upper Valley had signed a joint venture agreement that legally binds them to the East Central Vermont Fiber Network ( project, a municipal broadband initiative that represents a coverage area of almost 1,000 square miles and some 35,000 utility poles. Still in the process of securing the necessary funding, the project has nevertheless racked up pre-registration take rates greater than 50% in some areas. The folks behind the ECFiber Network are confident they can succeed�despite the recent troubles of some high-profile municipal fiber initiatives�in part because they know if they don�t bring fiber to their communities, no one else will.

The network will be deployed in partnership with ValleyFiber, a new venture from ValleyNet, an Upper Valley-based nonprofit that introduced broadband to the region via dialup service in 1994. ValleyNet managed upwards of 6,000 dialup subscribers before exiting the business in 2006 to pursue other broadband initiatives, including Valley-Fiber, which it launched in October 2007. ValleyFiber will design, build, and operate the network, which will include offering high-speed Internet, voice, and cable TV services on day one.

Throughout the process, ECFiber has looked to its predecessor to the north, Burlington Telecom (www.burlington, for guidance, even hiring Tim Nulty, former executive director/general manager of Burlington Telecom, to act as project director of ECFiber. Burlington Telecom expects to be profitable by the end of 2009, just four years after it began offering residential services.

The folks behind ECFiber expect nothing less from their FTTH network, which �started as a community effort and remains a community effort,� reports Stan Williams, board chair of ValleyNet and member of the ECFiber steering committee. That effort has been underway for several years with ad hoc broadband committees springing up in towns across the Upper Valley. When asked whether the sale of Verizon�s Vermont lines to Fairpoint was the impetus for such activity, Williams notes that it was �Verizon�s previous neglect of the infrastructure in this region� that spurred people to action.

�It�s a very grassroots effort that�s being driven by people in communities that, to quote the movie Network, where [Howard Beale] yells out the window, �We�re mad as hell, and we�re not going to take it anymore!� We�re sort of three, four, five, maybe 10 years behind in some ways,� he admits. �It�s almost Third World by now.�

�You�re talking educated people with 12k dialup,� adds Nulty. �You can�t even surf the Web with that.� In fact, people in the Upper Valley tell their friends not to send them photos, large PDF files, or PowerPoint presentations because they just clog their inboxes.

On its web site, ECFiber states that it is �the clear, unequivocal policy of the project to make service available to every household that wants it.� The fiber network will pass every home and business in the 23-town consortium with no exceptions. As Williams and Nulty point out, neither Fairpoint nor Comcast can make that claim.

At press time, 23 municipalities had joined the East Central Vermont Fiber Network: Barnard, Bethel, Brookfield, Chelsea, Granville, Hancock, Hartford, Montpelier, Norwich, Pomfret, Randolph, Reading, Rochester, Royalton, Sharon, Stockbridge, Strafford, Thetford, Tunbridge, Vershire, West Windsor, Williamstown, and Woodstock.

The ECFiber Network will follow the same architectural model as Burlington Telecom, which is an example of �a modified homerun,� says Nulty. A complete homerun network would feature an individual fiber from every house back to the hub, with that fiber available for anyone that wants to make use of it. But such a network is both uneconomical and impractical since it involves hundreds of thousands of fibers coming back to a single location.

A modified homerun network, by contrast, features fiber links from the home to aggregation points located around the serving area. �You locate a minimum number of fiber aggregation points around the town,� says Nulty, �and it should be the smallest number you can possibly get away with because you want to put it as far upstream as you can. In Burlington, we aggregate about 3,000 customers per fiber aggregation point.� Nulty says the fiber aggregation point is an actual building�20�30 or 20�40�that houses all the electronics, including the splitters, power equipment, and distribution racks. Downstream of the aggregation point are fiber links to every premises.

Burlington, a city of 40,000, features six fiber aggregation points. In the ECFiber area, which covers 1,000 square miles and 1,600 miles of road, Nulty estimates they�ll need about 10 to 12 fiber aggregation points.

In addition to requiring less fiber, the advantages of the modified homerun architecture are two-fold, says Nulty. First, maintenance is simplified because none of the equipment is in the field; it�s all in a secure building. Second, such an architecture facilitates easier network upgrades because, again, all the equipment is housed in a single location. To move to another generation of PON, technicians would simply go to the fiber aggregation points and replace old blades with new blades. Migrating from PON to active Ethernet would be similarly pain free, says Nulty.

�You can upgrade the electronics easily over time, forever, without touching the fiber,� he explains. �This is a very important point: We have not built a PON network. We�ve built a fiber network that we have initially provisioned as a PON, but we can change that provisioning as and when it makes sense. That�s a big difference between us and Verizon.�

ECFiber has already inked a contract with Atlantic Engineering Group for a fixed-price construction contract covering 75% of the projected capital expenditure. The network will be provisioned initially as a GPON to benefit from the economies of scale that Verizon�s FiOS has created. The network will deliver 2.4 Gbits/sec from the headend; a 32-split ratio will yield between 80 and 90 Mbits/sec to each home.

ECFiber hopes to break ground on the project in Spring 2009 and connect the first customers by the end of that year.

Like Burlington Telecom, the ECFiber Network is designed to be financially self-sustaining. The project will be funded with a 15- to 20-year non-recourse municipal capital lease. The network will be owned by the leasing company and leased to the consortium of towns until the lease is paid off. Williams says this form of financing is common for other revenue-generating municipal projects like parking garages. Participating towns will not be obligated to pay any of the expense of the network; the amount of the lease will cover the capital cost of the infrastructure as well as initial losses and interest until subscriber revenue can sustain the network on its own.

According to a recent white paper issued by the Fiber to the Home Council North America (www.ftth, �Municipal Fiber to the Home Deployments: Next-generation Broadband as a Municipal Utility,� the typical FTTH business plan usually requires a 30% to 40% take rate to break even with payback periods. At press time, the pre-registration take rate in the ECFiber territory was 15% overall, but in towns where there is not already some kind of coverage, that number escalates to 50% of the total population.

�We can go right up to Burlington and use their comparables,� says Williams. �In the parts of town that have been open for two years, they�ve got 40% plus penetration and average revenue that approaches $100. If we get close to those numbers, even in our area with our less dense towns, then we�re getting pretty close to the level that we need to break even,� he says. Moreover, he adds, some areas of Burlington already had 100% cable modem and 85% DSL coverage, and Burlington Telecom is successfully wooing those subscribers away.

For his part, Nulty believes that ECFiber will see take rates of �virtually everybody� in those areas in which there are no other providers. In fact, he believes ECFiber may have to change its installation process as a result. �Every FTTH company I�ve ever heard of�including us�runs the [fiber] pass down the road and then goes along afterward and hooks people up only when they call. But if you have 80% or 90% already presigned, you�re going to drop to everybody,� he says, �and that actually reduces your costs quite a bit.�

If the ECFiber network is a success, ValleyNet plans to help other municipalities and municipal consortiums in both Vermont and New Hampshire. There is a similar effort underway in the Granite State�the West Central New Hampshire Consortium (�that is currently looking at active Ethernet technologies. The WCNH is proposing a wholesale model similar to those followed by two municipal consortiums in Utah, iProvo ( and UTOPIA (Utah Telecommunication Open Infrastructure Agency�

Both iProvo and UTOPIA have had their share of troubles. At press time, iProvo had just been sold for $40.6 million to Broadweave Networks, a move that effectively privatizes the once wholesale network. And UTOPIA was in the midst of refinancing to retire earlier debt that carries a higher interest rate. The refinancing will give UTOPIA an additional $11 million to finish deploying its network and boost its marketing efforts.

The folks behind the ECFiber Network are not concerned about sharing a similar fate. Nulty is quick to point out the distinction between a wholesale network and an open access network. �Open access can apply where the city owns the network and is offering its own retail services, but anybody else is free to use the network too on a nondiscriminatory basis,� he says, �Burlington is open access, but it�s not wholesale only.�

�We will start out on day one providing all three services [voice, video, and data] because that�s what makes the project financeable,� adds Williams. �We�re out in the market now looking for $70 million in municipal capital leases, and people are talking to us. They wouldn�t talk to us if we were a wholesale network because on day one, where is the guarantee to the banker that there�s going to be sufficient revenue to support the project?� he mused.

In short, Williams, Nulty, and the community behind the ECFiber Network believe it is their mission to bring fiber-enabled broadband to east central Vermont�and they are not going to wait around for Comcast or Fairpoint to do it. �Our view is that DSL is going to be inadequate,� says Williams. �If it�s not inadequate now, by the time they roll it out fully in 2012, it�ll be kind of like dialup is now. It�ll just be creaking and groaning. So if we don�t do it, who�s going to do it?�

Meghan Fuller Hanna is senior editor at Lightwave.

More in FTTH/B