Prospective Roane County Broadband Project Overview

EXECUTIVE SUMMARY:  This report was split from Roane County Broadband Report and Prospective Project, revision 6.

Prospective Project Overview

The vision here is a public county-wide fiber network, providing gigabit service to all residents, locally hosted web services, the creation of a local tech economy, and long term transformation of the county economy. Along with the notion our neighboring counties should each have their own networks.

What remains to be seen is if the notion of a “WV Department of Networks” (or Backbones) could be realized with a multi-county pilot project, otherwise the county co-op will have to build & maintain our internet highways. In either case, the county co-op would build out the secondary (branch-line) and tertiary (PON FTTx) internet roads.

Project Organization

Project Group:  A small working group does early research and planning, which reports to the Commission.

Non-Profit Cooperative ISP:  Cooperative structure is as defined in §31G-2, atypical ~ membership elects directions, directors as a board elect officers, officers manage regular operations and staff. County Commission has initial organizational role and ongoing oversight, perhaps with permanent director seats or similar hook.

Cooperative Officers:  President is overall project coordinator and public face. VP is operations manager, oversees network construction labor. CTO is network engineer and administrator, oversees technical details and results of network construction. Treasurer serves as bookkeeper. Secretary oversees member and public relations.

Cooperative Staff:  We estimate the following personnel will be needed for network build out: five cable crew managers/leaders, ten cable crew techs (utility pole & fiber trained), five network techs. After build out re-task them to contracting with businesses and other counties, reduce numbers by attrition as necessary.

Cooperative Community Groups:  A hundred or so members which group together by location to provide mutual first-tier technical support, local maintenance, member training, increase benefits to their local community, etc. These groups would organize volunteer work events to walk & maintain utility line rights of way every few months. These groups could take on a larger community building aspect, such as working to build community centers where needed.

Cooperative Skilled Groups:  Members grouped by skills, training, experience, etc that they are willing to volunteer to the co-op. Skills could include network tech, PC tech, cable crew, utility pole training, member tech support, equipment maintenance, business support, web development, marketing, etc.

Network Work Groups:  County-wide specialist groups, most consist of a co-op staff group leader, one or two co-op staff techs with advanced training, and upward of a dozen volunteers in rotation. Types of groups, cable group to build & maintain the network, install group to do end-user installations, training group to train volunteers, splice group to handle the extremely expensive splicing equipment, field groups that do regular customer service visits, and probably others.

Emergency Groups:  Hire area contractors on retainer to respond to disasters, with priority over other area service providers. Storm damage monitoring volunteer group.

Security Group:  Local security experts and co-op admins work together to keep the network secure, respond to attacks, find & patch vulnerable systems on the network, provide basic security related support to members, and work collectively on network security policy. Members of this group need an NDA and additional bond.

Project Progression

Here is a rough outline of project stages. Original timeline start point was the Roane County Commission meeting on June 8th, now the start date is unknown. Delays risks loss of opportunity to join the State pilot project, and increases the risk other counties will deplete available grant funding, the bill goes into effect 7/7/17.

Broadband Report and Prospective Project (this document):  Defines the lack of broadband as a problem and suggests a starting point. If the Roane Commission accepts that a non-profit county-wide broadband service is necessary to meet the needs of all residents and that the project outlined here is the best place to start further research, then the next step is to request a Project Report.

Project Report:  Preliminary planning, a step beyond brainstorming, almost actionable. Its exact focus will depend on the scope of the State pilot project, but will include a rough plan for county-wide broadband service. This research will require authorization to discuss this on the County’s behalf with interested County residents, other counties, and state representatives. This effort will need to be staffed and requires funding. As part of this effort a feasibility study will be conducted to ensure that there is long term viability for rural fiber broadband service in this county. If this report is accepted, next step is a project plan.

Draft Project Plan:  This plan is actionable, start to finish, step by step, with contingencies for problems. The time frame will depend on the scope of the project. If this plan is accepted, next step is to present it to the county residents.

Public Meeting:  Present the draft plan for public feedback at a public meeting, recruit a core community most interested in seeing this through. Could also collect preliminary (conditionally binding) commitments to join the co-op ISP, install FTTx, and volunteer on build out.

Form Cooperative:  Refine the plan based on feedback, work with volunteers on community organizing, and resolve final co-op details. When ready to end the planning phase and cross the point of no return, officially file the paperwork to form the cooperative.

First Cooperative Meeting:  This is the initial co-op membership drive, approve bylaws, elect directors, elect officers, take initial actions needed to secure funding. Shall occur within 30-days of forming co-op.

Build Up Cooperative:  Secure funding, then start building up co-op capabilities. The co-op will require an office, network equipment facility, workshop, network materials depot, garage, vehicles, other equipment, and an initial inventory of network materials. Staff will need to be recruited, trained, and certified. Volunteers will need to be recruited and trained. Backbone providers will need to begin setting up their facilities. There are also legal notice requirements to access utility poles, and other procedural steps.

Build Out Backbone Network:  Run cable along 14N, 33E, 36S, 119S, and 33W, one cable crew each. As crews complete their routes, begin running branch lines and connecting public buildings, community centers, and public institutional customers. Setup temporary network routers.

Build Out PON Network:  Run branch-line cable along nearly all public roads, deploy PON’s along branches, and begin offering basic broadband service. Setup full network facilities, data center, and service related facilities (e.g. satellite dishes and IPTV servers), begin offering “beta services”.

Complete Plan:  Probably another year before everything can be considered “done”.

Rough Financial Analysis

While crude, this analysis shows the project is probably viable, but a lot of work is needed to get accurate numbers.

Example Project Costs

This is a truncated shopping list plus labor for the project core costs. This excludes materials for Branch-lines, PON Network, and Installation. Total of items in bold below is $5,417,500.00. It assumes no state pilot project involvement, no partnership with neighboring counties, no county material support, 30 month network build out time frame, and deployment via utility poles. It is missing estimates on insurance, utility pole rent & make-ready.

Planning and Organizing:  Estimated 5 months of work by a small group, assume 2 will be directly paid, others paid by partner groups. Roughly 5/12*52 = 21.67 weeks x 40hrs/wk = 867hrs * $20/hr = $17.34k * 2 = $34.68k labor. For community organizing, incorporation, public meetings, free food, volunteer expenses, marketing, consultants, and other costs, crudely guess a budget of double labor, rounded up, total $70k.

Co-op ISP Building:  Buy a vacant building or house in/near Spencer for $100k, renovate to include following… Offices $10k, network equipment facility $10k, small data center $10k, workshop $5k, network materials storage $5k, garage $10k, cooling $5k, backup power $10k, solar panels $20k, other costs $10k. Renovation total roughly $95k, building total $195k.

Network Equipment:  Core routers 3x $10k + ~$3k installation = $33k. PON CO equipment (unknown, guess), $20k. Switches $5k. Cache $5k. Misc $5k. Fusion Splicers 2x $12k. total $92k.

Data Center:  Router $2k, switches $5k, 10x servers $20k + install $2k, SAN $10k, ~10x refurbished cabinets $2k, total $41k. Note, additional cabinets of servers for rental deployed on demand (cost neutral).

Vehicles:  During network build out a larger fleet will be needed for 5 cable crews, all purchased used, most can be sold off later. Building includes a garage for service. Bucket trucks 5x $30k = $150k, pick-up trucks 5x $12k = $60k, cable trailers 5x $3k = $15k, tech vans 5x $10k + tech gear 5x $3k = $65k, insurance (unknown/guess) 15x $1k/yr x 2.5yrs = $37.5k, total $327.5k.

Penta Route Backbone:  Cable, 93 miles at $0.50 per meter is ~$75k, junction boxes $4k, cable splicing $1k, utility pole fasteners $3k, other materials $5k, materials total $88k. Labor, 93mi/5mi/day = ~20 days = x 4 staff x 8hr/day = ~640hrs x $40/hr = $25.6k. Volunteers, ~20 days x 12 people x 8hr/day = ~1920hrs x $5/hr = $9.6k. Vehicles ~$4k. Insurance is unknown, guess $20k. Connection costs $10k x 5 lines = $50k. Backbone total $197.2k. With planning, mapping, driving route, and other costs, add another ~$20k and round off at $220k total. Buried conduit is better, needs research, maybe $400k, no ongoing pole rent costs.

Staff Costs:  30 months of network build ~ 30/12*52 = 130wks x 40hrs/wk = 5200hrs * 25 staff = 130k hrs * $10/hr = $1.3M stipend + stipend/2 expenses + stipend/2 employment costs = $2.6M.

Volunteer Costs:  30 months of network build ~ 30/12*52 = 130wks x 40 hrs/wk = 5200hrs * 72 volunteers (in rotation) = 374.4k hrs * $5/hr expenses = $1.872M. Note, optimistically high turnout.

Branch-lines, PON Network, End-User Installation:  This provides county-wide residential FTTx coverage and is most of the network. Labor costs are covered above. County-wide branch-line material cost will be divided among all members as part of their installation fee. PON network costs will be divided among members in each area (~road). Members pay full cost of install from public road to their house. ~$5M installation fee financing.

Project Funding

Federal Funding

Several Federal programs are available to help non-profit cooperatives, counties, towns, and economically depressed areas. The ones below are the few that seem viable now to get the project started. Others are either passed the deadline for 2017 or provide support at later stages. This will need more research to piece together a funding plan, along with contingencies, and use the state-backed bank loan for whatever is left. Note, many of the grant programs have ended, and its unlikely anything similar will be available soon.

NTIA Broadband USA (start here for technical assistance):

Federal Communications Commission (then here): , (admin by NTIA)

Department of Agriculture:

Economic Development Administration:

Universal Service Fund:

Other Funding

In order of descending preference.

  • Training and Job Creation:  WorkForce WV has some funding available for training employees and creating jobs, needs research.
  • Low Income Assistance:  To provide grants or no interest loans for low-income families to install broadband. FCC, USDA, and HUD had programs, needs further research.
  • Pilot Project:  If Roane joins the pilot project there might be zero interest loans from WV EDA, federal grants, cost distribution over more counties, and some costs covered by the state.
  • Roane EDA:  Hopefully they can contribute ~ grant, building, or no/low interest loan.
  • Membership Dues:  Yet to be determined, but e.g. 5k members x $20/yr = $100k/yr, ~$300k in 3rd year.
  • Service Credits:  Staff estimated $1.3M service credits. Volunteers upward of $1.25M service credits. Members need to volunteer for 20 hours per year, after that they can opt for $10/hr service credit or subtract from future years volunteering. If they don’t volunteer they pay a $20/hr fee ($400/yr), this should make up for a portion of the lost revenue. Members could also pre-pay service to support the co-op. Credits would be transferable to other members, could accumulate interest matched to average loan rates, and could be cashed out after the bank loan is paid off. Obviously this erodes early revenue.
  • Local Investors:  Private loans, bonds, or a similar mechanism should be created to enable wealthier members of the local community to fund this project.
  • Bank Loans:  HB3093 includes up to $10M in loan insurance, finance costs need research. This should be exclusively for local banks.
  • Rough Loan Example:  If the co-op uses all of a $5M loan, at 4% interest, 60 payments, is $92,082.61 per month, divided over 5,000 members that’s $18.42 each, cost is $525k.
  • FTTx Installation Loans:  HB3093 allows for financing installation costs for up to 5yrs, presumably this would include any related finance costs, insured by the co-op then the state. For example, secure a $5M line of credit, with a $2k per install loan limit, this can cover 2,500 members. At 4% interest, 60 payments, $2,000 is roughly $36.83 a month, and costs $209.80. More search is needed to determine those installation costs.
  • After Network Build Vehicle Sale:  Sell the following vehicles when no longer needed, assume worst case 20% devaluation, bucket trucks 3x ~$24k = $72k, pick-up trucks 3x $9.6k = $28.8k, cable trailers 4x $2.4k = $9.6k, total $110.4k.
  • After Network Build Contracting Services:  Offer neighboring counties experienced cable crews for FTTx deployment, fully paid staff and assistants (formerly volunteers), revenue potential needs research.
  • Property Taxes:  Could raise property taxes to pay for county-wide infrastructure, roughly $5M in bonds.

Cooperative Membership & Revenue

Its too early in planning and there isn’t enough information to guess at expenses & revenue, but the real question is could this co-op pay off that monster startup debt. Numerous case studies say yes, smaller towns have done more with less, but the debt puts a severe damper on it. Every effort needs to be made to find other sources of funding and reduce project costs via partnering with other counties, or joining the pilot project.

Government & Business Usage:  Needs more research. Not enough information to make a useful guess. Case studies indicate these two groups form a sort of “base load” which pays for the core network construction costs.

Membership Pool:  US Census data says the county has 14,208 residents and 7,389 housing units. Assuming Spencer, with 2,158 residents and 1,191 housing units, is blocked from the co-op by SuddenLink (~35% might be eligible), there is a potential membership pool of around 5,000 households.

Membership Monthly Service Rates:  Not enough information, but assume its price competitive with SuddenLink and uses that simplified package rate (not itemized cost share). $100 per month for a 1Gbps standard broadband service, plus installation loan payments ~$37 max, plus month share of non-volunteer fee $33.34, plus 6% tax $10.22, is a total of $180.56 per month. Volunteers with service credit ($10/hr x 10hrs), max loan ($37), plus $8.22 tax, could pay $45.22 per month. Its hoped non-volunteer fees will equal volunteer operating expenses, similarly tax and loan payments aren’t income, so adjusted payment is $100 per month.

Recruitment Rates:  Not enough information, needs more research. The goal is to provide FTTx to all residents but we can’t force them to accept it and pay a monthly service fee. Case studies show some projects with a focus on government, business, and utility usage, and with competing incumbent providers, are viable with residential usage rates of around one-third of households – most are either, in affluent areas with funding from property taxes, or they received federal stimulus funds. Projects where the only competition is dial-up, satellite, and poor mobile-wireless coverage see residential usage rates near 90%. Examples of monthly cash flow (for vague viability, not particularly useful without “other expenses” defined):

Optimistic 80% Take-Rate, 4,000 Members:  Standard service revenue $400k. Loan repayment is ~23% ($92.1k), payroll for 25 is ~21.7% ($86.7k), and ~55.3% ($221.2k) for other expenses.

Pessimistic 60% Take-Rate, 3,000 Members:  Standard service revenue $300k. Loan repayment is ~30.7% ($92.1k), payroll for 15 is ~17.4% ($52k), and ~51.9% ($155.7k) for other expenses.

Perseverance 40% Take-Rate, 2,000 Members:  Standard service revenue $200k. Loan repayment is ~46% ($92.1k), payroll for 10 is ~17.3% ($34.7k), and ~36.7% ($73.4k) for other expenses.

Service Credit Problem:  Use of service credits to finance labor creates revenue uncertainty. In the above examples member usage of those credits may have to be restricted to cover critical expenses.

Competition Problem:  Typically incumbent providers throw large sums of money at undermining projects like this, via lobbying, disinformation, negative advertising, and the easiest upgrades to their infrastructure to improve and extend service. Expect severe resistance from Frontier and SuddenLink.

Cooperative Expenses:  Not enough information, needs more research.