Broadband Case Studies Report

EXECUTIVE SUMMARY:  Both proof of concept, and examples to follow, from other counties and municipalities which built fiber networks.

Broadband Case Studies

Here are four case study summaries with notes and comments.

Roane Notes:  Area 484 sq.mi., population 14,208 (‘16 est), 7,389 housing units (~5,000 outside of Spencer).

Citizens Take Charge: Concord, Massachusetts, Builds a Fiber Network

Source:  Harvard University, February 2017, 20 pages.
PDF Download:

Summary:  Municipal project, population ~16,500 (mostly affluent), over ~26 sq.mi, mostly buried cable. Started in 2009 unusually as a $3.9M 4-year smart grid project with 100-mile fiber backbone. Added internet facilities in 2013 for $600k, FTTx installations started 2015. By end of 2016 the town saves $108k/yr in telecomm costs, generates $88k leasing revenue from several large data users, and internet services are nearly at break even with ~750 internet customers (~12% penetration of ~6k power customers, notes that includes ongoing construction costs and small net loss due to one-time expenses). Next it will enable regional utility control over smart grid for power load-leveling to generate $125k/yr. Additional benefits, savings, and revenue are expected from fiber network over 30 year service life.

Incumbent Provider Fight:  Comcast provided horrible service, opposed town internet, hired think tank to create propaganda (much of it lies), and ran a campaign against FTTx, but failed.

Smart Grid Costs:  Original estimate $4.5M, final cost $3.9M. Labor $2M, network materials $1.4M, smart grid wireless nodes $250k, smart grid equipment (e.g. remote thermostats) $180k, engineering services $125k, substation renovation $100k, technician’s vehicle & equipment $100k. This was financed with a general obligation bond, with annual payments of $418.6k collected from customer electric bills (avg. $4/bill).

Smart Grid Network:  Roughly 100 miles of fiber, fiber tech is GPON, network equipment from Calix, network designed by Uptown Services, Nexgrid provided smart grid hardware, Nextgen constructed the network, uses 80 fiber-connected wireless gateways on utility poles, uses 600 wireless repeaters in streetlights, and 878 smart meters (so far). With this it is able to remotely switch 141 electric heaters, 360 electric hot water heaters, and more, ~3MW total. This can generate $125k/yr for regional power load leveling and will help manage expanded solar panel deployment (have 1.7MW and 4.5MW arrays, are planning rooftop next).

Internet Service:  Was planned after smart grid, offers symmetric data service, study concluded triple-play (phone-TV-internet) cost ineffective due to Comcast competition. Expected “take rate” is 2k of 6k customers, build out is 15 customers per month, estimated installation cost is $600 to $1.4k (plus any conduit per ordinance). Startup cost of $600k covered by town bond ($1M avail.). For 2016, ~750 customers produced $560k revenue, at a cost of $583k including some one-time costs, they have a staff of 5. Residential rates 25Mbps $50/mo, 50Mbps $65/mo, 100Mbps $75/mo, 200Mbps $90. Commercial rates 50Mbps $75/mo, 100Mbps $90/mo, 200Mbps $150/mo, 300Mbps $200/mo, 1Gbps call for quote.

Remarks:  Without triple-play and with symmetric service, use of GPON is inefficient or plainly a poor choice. EPON hardware is significantly less expensive (upward of ~half for 1Gbps, some claim 90% savings in niche high-density lower-bandwidth cases, like 25Mbps per unit in apartment buildings) and is far more energy efficient (their project focus). If FTTx were built out first and faster they could have avoided some of their wireless hardware expenses.

Roane Notes:  County and towns potential savings on telecom costs needs further research. Ask Mon Power and AEP about interest in smart grid deployment. Propose county and towns study potential smart grid related benefits, as well as solar, wind, and storage.

Also, for smart grid ROI also see


All Hands on Deck: Minnesota Local Government Models for Expanding Fiber Internet Access

Source:  Institute for Local Self-Reliance, September 2014, 87 pages.
PDF Download:

Summary:  This collection of twelve short case studies from across Minnesota is representative of most issues faced around the nation. A few cases of the most interest here include: Scott County with fiber for economic development. City of Windom FTTH project, very high take rate and popularity led to expansion into rural areas, but a lot of debt. Lac qui Parle county built fiber to the farm via cooperative but skipped county seat (Madison) due to opposition from incumbent providers, years later rural areas have great broadband while city service is still poor.

The other case studies are also worth reading, it forms an overarching perspective that local ownership and control, by either county or cooperative gives the best returns. Cases with the best results used federal funds. Cases with unmet expectations relied on either, large bond financing, or large non-local companies to build, maintain, and operate the network. Note parts of this dates back a decade, some information is obsolete, in particular, fiber and some equipment is much less expensive now.

Scott County:  County project, population ~130k, area 368 sq. mi. Project started january 2007, was for a major overhaul of public safety communications across the county, including $4M for a 90 mile fiber ring, and a $3.5M bond. Estimated savings were $500k/yr from county fiber replacing a large number of telco leased lines, but debt payments and apparently unexpected operating costs reduced it to $35k/yr savings. Fiber ring linked all county buildings, state backbone, two neighboring counties, and eventually a national fiber backbone hub in Minneapolis. Was built and is maintained by contractor, Zayo. Economic development efforts resulted in attracting a $70M plant with 500 jobs from Emerson Process Management, the deciding factor between Scott County and Chihuahua Mexico was Emerson saving $1.1M to 1.7M over 20 years via use of Scott’s fiber network. Shutterfly also located there, adding 329 jobs plus 200 seasonal jobs, in part due to Scott’s fiber network.

Windom City:  Municipal project, population ~4,600, had local cable TV network but no broadband, jumped on FTTH with buried fiber as best long term solution. Started project in 2004 with $9.5M bond funding, expected 1500 customers, received 2000 requests so added debt to build more, in 2007 a local business near their network demanded service so they added debt to build there too, repeated with another factory, then hospital, and saw significant returns in economic development. With federal stimulus funding in 2012, they created Southwest Minnesota Broadband Services (SMBS) to expand into 8 nearby smaller towns, and further. From there the text goes on defensively about the large economic benefits, jobs saved & created, money saved by residents from lower service fees vs telco, and justification for high startup costs that burdened the town with bond debt until it was refinanced onto the ISP. Project debt is around $11.2M, which was expected & they can certainly repay, eventually.

Lac qui Parle County:  Cooperative project, county is 778 sq.mi, population ~7,200, ~4000 spread across communities of under 100 residents, ~3,000 split between towns of Madison (county seat) and Dawson, 25% are over 65. Frontier and Mediacom service both towns with high speed internet. Farmers Mutual Telephone (Farmers), a co-op founded in 1904, serves ~1,000 members the northern part of the county with higher quality service, deploying DSL in the late 90’s, then fiber in 2007 to 2010 at a cost of $5.5M. In 2010, ~3,200 residents in the southern half of the county were still on dialup or satellite. The county EDA tried to get Frontier to expand south but was refused, so the EDA asked Farmers to do it, they struggled but ultimately completed it in 2014 for $10M. Sign up rate was suppressed by Frontier demanding $250 to $300 in termination fees based on contracts that didn’t exist (no mention of a lawsuit). Farmers offered DISH instead of cable TV, but had a problem near both towns where Frontier & DISH had an exclusivity contract, Farmers partnered with a IPTV provider to offer TV. End result was Farmers’ rural customers have superior FTTH broadband while Madison turned into a “doughnut hole” or “reverse oasis” of slow internet.


SandyNet Goes Gig: A Model for Anytown USA

Source:  Institute for Local Self-Reliance, November 2015, 8 pages.
PDF Download:

Summary:  Municipal project, population ~10,000, over 3.14 sq.mi. A small town neglected by incumbent providers, to the point of not being able to get DSL to City Hall in 2001, tries a wifi mesh plus DSL but finds that lacking in 2008, then replaces it with FTTx to downtown businesses in 2010-12, expanded it to the entire city 2013-15, and sees excellent results. Their fiber project started with ~175 wifi & DSL customers and grew to 1,500 FTTx customers at a ~60% take-rate, and is planned to expand outside of town up to ~2,000 customers. They tried to do a public-private partnership but they couldn’t find a good partner, so they financed it with $7.5M in revenue bonds, giving them a minimum take-rate of 35%. A contractor, OFS designed and built the network. Prices are $39.95 for 100Mbps and $59.95 for 1Gbps plus $350 install fee, phone is extra, and TV is a work in progress, no data caps or contracts. Their backbone is a county fiber ring. The downtown business loop was done as part of an urban renewal project and waved installation fees to get smaller businesses to join. Study includes several statements of support from businesses, citing significant savings vs previous data services and increased productivity from more bandwidth. Incumbent providers remain indifferent despite being nearly driven from the market, but Google was very impressed.

Project Planning:  They have a good IT director and finance director, and used an experienced contractor to design a network tailored to the town’s needs. This overlapped with an urban renewal project, improving project efficiency, and lowering cost to downtown small businesses.

Demographic Note:  Census data cited on Wikipedia says Sandy experienced high population growth. 1,544 in 1970, 2,905 in 1980 (+88.1%), 4,152 in 1990 (+42.9%), 5,385 in 2000 (+29.7%), 9,570 in 2010 (+77.7%), and 10,644 in a 2015 estimate (+11.2%). In 2010 median age was 32.7. So its a younger population in a rapidly growing town, this makes accomplishing big things easier, but also makes the incumbent provider neglect more unusual. Note, Spencer has a population of 2,271 in 1970 and 2,195 in 2015 (est), with a 2010 median age of 40.2.

Service Prior To FTTx:  While the case study doesn’t mention this, offering wifi and DSL prior to FTTx probably had a significant benefit to the smooth progression of FTTx deployment.

Roane Notes:  Consider combining parts of network construction with other infrastructure projects, such as installing conduit with gas or water lines.


Florida Fiber: Martin County Saves Big with Gigabit Network

Source:  Institute for Local Self-Reliance, June 2012, 13 pages.
PDF Download:

Summary:  County project, population ~150k, ~27% over 65, area ~550 sq.mi. A county with a lot of leased dark fiber from Adelphia was facing a rate increase from $10.5k/mo to $98.26k/mo from Comcast following the bankruptcy of Adelphia, instead they built a county fiber network. Initial project estimate was $9.8M, but a joint conduit venture with FDOT saved $2.5M reducing the cost to $7.3M, a partnership with a local independent telco to swap 43 miles of fiber saved $3.16M reducing the cost to $.4.14M, the school district paid $1.3M for 30% ownership stake, and the county financed $3.12M for 15 years at 3.25% with annual payment of $270k. Then the network was extended to a neighboring county and lease signed with an area hospital network, no figures but they said it “put them in the black” vs what they previously paid Adelphia. Then they connected 3 public safety towers to save another $73k/yr in fees. The school district recouped its investment after 5 years, saving them a total of $568.49k through the sixth year vs Comcast sky-rocking service fees. Later expansion and connections to other counties saved even more. State law makes it extremely difficult for the county to offer communications services directly to residents and businesses, dark fiber is not considered communications.

Partnerships Pay:  Cooperation between the county and FDOT, the local school district, and area hospitals, combined with savings, paid for the project.

Network Redundancy:  Comcast’s fiber is all on poles, vulnerable to hurricanes, and no redundancy. The county fiber network is all buried, and the network is arranged in interconnected rings for redundancy to a cut cable, but at a much, much higher material cost than Comcast.

Good CIO:  Kevin Kryzda was the catalyst for the project. Whether its an individual or small team, this role of project organizing is critical to success.

Roane Notes:  Investigate partnerships with large data users and neighboring counties. Consider eventually connecting the far ends of the Penta-Route together into 5 rings, add full redundancy to the government network, and network-wide emergency traffic (e.g. phone service) redundancy via wireless.


Links to Resources, More Case Studies

There are a lot more resources available ~ news, case studies, academic reference, industry sites, manufacturer's websites, and websites of the communities in the case studies. A lack of information won't be a problem…

Institue For Self Reliance:

Broadband Communities Magazine:

Fibre to the Home Council Europe:

Wilson's Greenlight Leads North Carolina in Fast Internet
PDF Download:

Breaking the Broadband Monopoly
PDF Download:

Broadband At the Speed of Light: How Three Communities Built Next-Generation Networks
PDF Download:

Santa Monica City Net: An Incremental Approach to Building a Fiber Optic Network
PDF Download: