HomeMy WebLinkAboutCOW 2005-01-10 Item 2A - Presentation - Deployment of Tukwila Fiber Network for High Speed Connectivity COUNCIL AGENDA SYNOPSIS
ti tS�' Initials ITEMNO.
0 1 I F LEL elleeting Date I Pnpard 1 Mayor's renew 1 Council ncim
1/10/05 I FI I S sub- I f-c/ I
rsoa
I I I I
ITEM INFORMATION
CAS NUMBER: 05-004 I ORIGLNAL AGENDA DATE: JANUARY 10, 2005
AGENDA IITMTITLE Special Presentation Connectivity Project
CATEGORY Disetutvrn Motion Resolution Ordinanx Bid Award Public Hearing Other
Attg Dote M(g Date 31tg Date Mtg Date MUg Date AUg Date !g Dare 12 /13/01
ISPONSOR Council !vigor Adm Svcs DCD Finance Fire Legal Pe t Police PTY/
SPONSOR'S Asset Analytics will brief the Council on the proposed deployment of the Tukwila Fiber
SIDDIARY Network (TFN). TFN will provide Tukwila residents and businesses with access to high-
speed connectivity. Briefing will assist Council in making an informed "go" or "no -go"
decision on the project.
REVIEWED BY COW ?..Mtg. CA &P Cmte F &S Cmte Transportation Cmte
Utilities Cmte Arts Comm. Parks Comm. Planning Comm.
DATE:
RECOMMENDATIONS:
SPONSOR /ADMEN.
CoN1lTrEE
COST IMPACT FUND SOURCE
EXPENDITURE REQUIRED AMOUNT BUDGETED APPROPRIATION REQUIRED
Fund Source:
Comments:
1 MTG. DATE RECORD OF COUNCIL ACTION
12/13/04 Special Presentation Connectivity Project
1/10/05
MTG. DATE ATTACHMENTS
1/10/05 Information Memo from Asset Anatytics dated January 5, 2005
1
AAt
IJ.uaA~, '+-e.
7
INFORNIATION alEMO
From:
Mayor Mullet
Asset Analytics, Inc.
January 5, 2004
Connectivity Report to Council
To:
Date:
Subject:
ISSUE
Provide Council with final report regarding Tukwila Fiber Network (TIN). Discuss deliverables for
upcoming Council Retreat (Attachment I).
BACKGROUND
The City has contracted with Asset Analytics to acbieve the City's goal of deploying a metropolitan area
network and providing gigabit Ethernet connectivity to Tukwila citizens and businesses. The City has
requested Asset Analytics to analyze the affect of deploying fiber in the Central Business District first to
gain initial customers and to reduce costs in lieu of deploying the network throughout the City. There are a
number of tasks (Attachment 2) specific to the completion of this assignment.
DISCUSSION
The following is a summary of each task assigned to Asset Analytics:
CBD Business Model
From December Report
First, the Central Business District had to be carved out as a singular deployment event, along with a
connection from the CBD to the Sabey Data Center. While there is an assumption that the network will
stage from Sabey's Data Center, it does not have to and indeed could be deployed from one of the City's
buildings in the CBD. The advantage with Sabey is that it is a sophisticated network center, with physical
security and redundant backup systems in the event of any kind offailure. Businesses using the Tukwila
Fiber Network, will want systems such as Sabey's to ensure their connections are of the bighest quality.
Next a Bill of Materia Is (BOM) was developed to reflect the CBD carve out from the design previously
completed for the network. The vendors for each of the components have been contacted and we are
collecting price information from them. Additionally, we are considering a wireless solution to the
customer from the fiber located in the street. This may be a less expensive way to deliver data to business
customers rather than through the delivery of fiber from end to end. This is primarily due to the last
connection point costs in the CBD. Trenching, drilling, and micro-channel cuts are needed for a full fiber
deployment and may require owner permission. Wireless requires some infrastructure but may ultimately
cost less. Wireless c",!,abilities however, are not up to those of fiber in terms of data transmission rates,
clarity of signal and security. Additionally, wireless may be considered for some deployments as simply a
stop-gap m= from a cost standpoint. Theoretically, ifwireless is used and the customer requires a fiber
connection, the City could deliver it as an upgrade. To be resolved is whether there would be prohibitive
signal interference with the wireless deployment. Some have expressed concern that both Boeing and the
Homeland Security Building may cause substantial interference.
Business Model Findings
There are many variables that make up the business model. Take rates, network costs, personnel, average
payment to the city for a customer and ongoing maintenance are but a few of the costs that must be
considered. AAI has continually used conservative estimates in its modeling of the network. Conservative
in this case means bigher deployment cost estimates, lower take rate estimates, average sized payments to
16004 Tualatin-Sberwood Rd, Suite 505, Sherwood. OR 97140
Oient Confidential
AA(
A-uaA~, 1_,
the City by content vendors and a three-year time fuyne for the customer base to accrue to the model
assumptions.
AN ran four scenarios to test the elasticity of the model to changes as well as to quantify different
deployment methods. Full documentation will be provided in the Council Retreaz packet while only
summarized (bottom line) values are contained in this report. All scenarios assumed a 45% take rate (608
customers), the average gross revenue of $100 per month per customer with 20% of the revenues (540)
going to the City. The model assumes (2) City FIE's as well to manage the network. Co-location of the
electronics could add another 51,500 per month, the connection to Sabey will cost about 550,000 using
TFN conduit
Options 1&2: These options assume full deployment of a fiber-based system. The difference is a change in
the cost of the customer premise electronics (CPE) - the electronics deployed in the customer's location.
We know a data only device (no video and marginal telephony) can be purchased for around 5300 (possibly
lower in a formal RFP). Therefore, we ran the model using a CPE price ofS400 in Option I and 5300 in
Option 2.
Option I Net Revenue Results: -$197 Monthly or -53571 Annually (a small loss)
Option 2 Net Revenue Results: 5I39 Monthly or SI,676 Annually (positive cash flow)
Options 3&4: These options consider the use of wireless as part of the solution. In Option 3 we run fiber all
the way to the nodes (190 of them placed in fiont of major business locations) and in lieu oftrenchinglsaw-
cutting fiber to the building, we use a wireless solution from towers located in the CBD. This is a lower
cost option in preparation for when we ,,-ant to deliver fiber-based solutions over time. In Option 4 we
consider an entirely wireless solution but deploy the core (backbone) fiber for future use.
Wireless has advantages. It is considerably less expensive to deploy. It has disadvantages. It is less secure,
subject to weather or other interference and most importantly, it does not have the bandwidth capability
fiber does ranging to about 75 megs max today, versus 10 to 15 times that speed with fiber based solutions
(delivering gigabit speed).
Option 3 Net Revenue Results: 53649 Monthly or S43,789 Annually (positive cash flow)
Option 4 Net Revenue Results: 55,693 Monthly or 568,3 19 Annually (positive cash flow)
Consider that the more money you save, the less satisfactory the deployment (its slower). This may affect
take rates as well since the allure of speed up to a gigabit is attractive to businesses as well as the security
of a fiber based active system. The fiber only system deployment cost is about 52.1 million; the hybrid
about 51.6 million and wireless only is about 51.3 mil/ion (or much lower ifno fiber is utilized).
Recommendation: Try to make the complete fiber based system work financially. The loss is small enough
that increased advertising sales, higher take rates or other variables could easily push the model into the
positive. An RFP may pare prices down. All other options will be more costly in the future when upgrading
them to a fuIl fiber system. The alternative would be to charge the customer to upgrade. However this may
run afoul of the Goal IV mandate to provide gigabit Ethernet to all in Tukwila. Last, what is not considered
here is the positive attraction a first class network will have to busincsses. Hebert Research's econometric
model predicted just under a 10% increase in sales volume as a consequence of the network in Tukwila.
That was about 5300 million more in sales, given that there is S4 billion in gross sales for the area each
year. The City can use the TFN to attract clean, upper scale companies who need bandwidth. As an
economic development tool, it is hard to find one that would have such a considerable impact and cost-
maybe - 53,500 a year (the net loss each year). Employment and housing needs will increase as well.
Determine Revenue Opportunities
From December Report
With respect to revenue generation, we have assembled a list of potential content provider resources and
are working through the companies on the list to determine their qualifications, what interest - if any - they
16004 Tualatin-Sberwood Rd, Suite 505. Sbernood. OR 97140
Oient Confidential
AAt
A-w:JA~, r~,
would have in delivering content over the City's infr9structure and whether they would be willing to pay
the City for the right to do so.
Revenue Opportunities Findings
One of the key issues with respect to building the network is whether or not there are content providers who
would be interested in paying the City for access to business customers via the City's network. We know
fiom previous polling that there are small lSP's (data providers) out there who definitely are interested,
ho\\'ever partnering with a major provider seemed to provide a greater seme of viability and ongoing
revenue. Since the last report, we have found that Comcast, Verizon and Qwest all have interest in using
the TFN as a resource to deliver services to the Central Business District. Specific terms have not been
negotiated, but there is open interest in engaging with the City should it deploy a fiber network in the CBD.
There are enough data I voice vendors out there to ensure multiple partners utilizing the City's network.
This puts the City in a better negotiating position as well as ensuring a constant source of revenue fiom the
vendors signed to provide services across the network. Video providers remain elusive in terms of numbers.
However the large providers, Verizon and Qwest have both announced intentions to provide video as well
as voice and data. Their capability may come on stream as the City's network is completed. Verizon
especially is gearing up quickly to access markets over fiber. No specific negotiations where conducted
with respect to the content providers. The discussions were hypothetical in nature.
The conclusion is that there are vendors available for data and voice services who have expressed interest
in negotiating with the City to use its network to gain access to the Central Business District businesses.
Multiple vendors on the system not only create competition for payment to the City, but may increase the
take rate above what a single vendor could provide due to product differentiation.
Use of City's Street Crews
From December Report
Three members of the Public Works department have been trained to do fiber related activities such as
fusion splicing, fiber testing and termination. In an initial meetiog with Public Works, those attending the
meetiog did not think they any available resources to be utilized in the TFN project. There was additional
discussion regarding the logistics of how (or if) to structure the "on caU" fiber maintenance from the City
or whether to out-source it. This may be an issue to discuss with respect to the union as well. Regardless~
the consensus fiom the City personnel present was that there was no capacity for additional work with
respect to these three. Training another may be a possibility. Outsourcing the construction and maintenance
of the fiber can be done as well. If the TFN network requires roughly 4,000 splices, it will cost the City
about SI20,000.
Use of City Crew Findings
After the meeting with the Public Works it was clear that under current staffing requirements and
assignments that no personnel could be utilized for the TFN deployment. The business model was adjusted
to include outsourcing of the fusion spicing and termination activities that could have been done in-house if
resources were available. The approximate cost for the Central Business District fusion splicing of the fiber
is between Sloo,ooO to S130,000. lfthe City needs another Public Works FIE, possibly one of the fiber
trained Public Works personnel could help with the splicing of the fiber on the TFN. There are
approximately 4,000 splices required to complete the CBD network. At one half hour per fiber this equates
to about 330 man-days of work (given a six hour active work day). Spread over a two - three year
deployment would imply a workload of about one third of a workday average.
Evaluate Potential Efficiencies with City Traffic Signal Projects
From December Report
The City has two traffic signal projects to be completed. One, in conjunction with surrounding
municipalities, is substantially completed. The other is about a year off. It was agreed that using the same
conduit could be a savings as well as using the same fiber bundle. The Traffic Signal project would always
have its own fiber as would the TFN. Some electronics could be common, but will ultimately depend upon
the vendor selected for the TFN project and ensuing interoperability (working together). The next step is to
16004 TuaJatin--Shervwod Rd. Suite 505, Sherwood, OR 97140
Oient Confidential
AA(
A-uaA~, (.-.e,
overlay the CBD TIN design over the traffic design JO see what physical efficiencies can be gained. In the
first project the fiber is mostly down, so there is no gain there. In the second, there could be common
bundles of fiber used and therefore pulled at the "ame time.
Gaining Efficiencies Findings
There may be some marginal cost savings available if the TIN is deployed. The current signal project is
just about completed and using common conduits could be the best chance for cost savings there. The
upcoming signal project may also have some cost savings if fiber is pulled (and spliced) concurrently with
the TIN. While hard to quantify, combining two deployments into one will usually save money. Fiber
could conceivably be shared as well. Separate bundles, but on the same cable would be the configuration.
Repair for either signal or consumer related data would then occur in separate fiber groups ensuring the
security of each from the other's activities. Overall, the cost savings would most likely be in the low five
figures, but could be had nonetheless.
Finalize List or Vendors ror System
From the Deamber Report
As mentioned above, the Central Business District was separated as a project fiom the City as a unique
deployment Then the component parts were identified and a list of vendors for the network was created.
Some of the vendors will only be contacted in the event the City goes to bid, others have been contacted to
get illustrative pricing for the business model.
List of Vendors for the System Findings
AAl has accumulated a list of numerous vendors for each necessary component of the network. This sets
up the possibility of price competition as long as quality and confonnity to design specificatioIl5 are not
affected. The business models use relatively conservative estimates of costs as provided by selected
vendors and presumably serious competition in an RFP would further reduce deployment costs. If Council
wants to move forward, an RFP for the network and for the electronics would be recommended.
Determine Financial Resources and Funding Strategies
From the December Report
Previously, the City has considered borrowing short-term funds at still fairly low rates to construct the
network and then bond afterward for long tenn amortization. We are investigating other sources that may
be available to the City. This wiI! be the last issue to be resolved. First the cost of the network must be
worked out (including optiOIl5), the general viability assessed and then payment strategies can be
addressed.
Financial Resource and Strategies Findings
During the course of AAl's engagement, DynamicCity Metronet Advisors approached the City to discuss
the possibility ofTukwilajoining a cOIl5omum of Seattle area cities interested in deploying fiber networks.
This is similar to the work done by DynamicCity (DC) in Utah under the name of Utopia. Essentially an
intergovernmental agreement (IGA) was struck between the interested small ruraI cities ostensibly to share
costs and an LLC is formed and run by the participants. DC is then engaged by the LLC to do feasibility
studies, business models and design initially; then help with the financing upon agreement to move forward
and finally to manage the network operationally after deployment The review of Utopia is mixed.
Originally there were 18 small cities involved, some dropped out Salt Lake City joined and then dropped
out It is unclear at this writing what Provo's position is, AAI is awaiting a response to a number of
questiOIl5 fiom Provo regarding their take on Utopia. The remaining Utopians are deploying a system.
In conversatioIl5 with DC, they indicated that the only way to attract the big companies was to have
population mass. The inference being that Tukwila is too small to attract the "big players". Significantly,
both Qwest and Verizon have expressed interest in utilizing the TIN without any concern as to market size.
Additionally the big guys aren't necessariIy the only answer to gain success. In some ways, nimble,
reactive, entrepreneurial small companies may be more attractive. The history of poor customer support,
monopoIistic pricing, lack of technology upgrades and legacy thinking don't auger support for the baby
Bells or large MSOs such as Comcast or Charter. The key is to have a mix. Interestingly, in DC's case,
16004 Tuahitin-SherYr"ood Rd, Suite 505, Sherwood, OR 97140
Client ConfidentiaJ
AAf
A1M1A~, (w..
AT&T, the big player they brought in to Utopia, will not provide triple play, refusing to bring video in.
Interestingly, DC's answer for video was a small lSP who decided to get into the video business as well.
Additionally the business model for a Utopia in Seattle would necessitate TukwiIa forming an LLC with
other interested cities. To date there are no others, DC is trying to leverage TukwiIa as the first and
springboard from there. The LLC owns the fiber - not Tukwila. TukwiIa would also have to pay in to the
LLC its pro-rata portion of contribution should there be a capital need. The network under Utopia would be
delivered at once, so Tukwila would lose its market advantage in fIrst to market deployment, which it still
has in the Seattle area, if it deploys and lights the network in the CBD.
For financing, DC then takes the LLC to its Bank of America loan rep and finances the loan to the LLC.
However, the City must still guarantee its portion of the loan proceeds and (presumably) help one of the
others in the event there is a partial default.
There are still questions that need to be answered, but to date we have not been able to validate cost savings
attributable to joining a Seattle Utopia, market advantage for joining or financing advantages for joining.
AAI at the City's direetion will continue to discuss this with DynamicCity Metronet Advisors, but to date
they have not proven to have added value. They are good marketers of what they do and they create a cloud
of doubt that if you don't join you will lose out on contracting with big content providers, but they haven't
proven that to be true as of yet. We recommend further study, with the idea that DC must pass some
fundamental test of value for Tukwila to consider moving forward with them.
PROJECT RECOMMENDATION
The project as modeled is --at the worst - close to neutral in tems of cost without going to a fomal bid. We
recommend adopting a resolution that the network be deployed subject to financial milestones for costs
being met, and subject to quality/design considerations. We recommend talking more to DynamicCity,
while continuing to proceed to gain initial market surprise and advantage.
Respectfully submitted by Asset Analytics, Inc.
1600-4 TuaJatin-Sberwood Rd, Suite 505, SheJ"'ffood, OR 97140
Client Confidential
COUNCIL RETREAT PACKET ITEMS (Attachment 1)
. AAI Summary of work to date completed regarding the TFN
. Network Electronics Proposal Documentation
. AAI CBD Conduit Gap Report (Conduit Needed To Complete The CBD Network)
. Bond Council Opinion Letter With Respect To Financing The Network
. AAI Tukwila Fiber NetwOIx Business Case
. Central Business District Key Businesses Contact List
. AAI Recommended Central Business District Deployment Process
. Competitive Local Exchange Carriers (CLECs) Regulated by the WUTC (for voice and data)
. AAI TFN Cost Breakdown
. Hebert Research Business and Residential Connectivity Survey Results
. AAI Network Pro Forma
. AAI Network Operations Model
. Tukwila Census Numbers
. Hebert Research Take Rate Rationale
. AAI Cash Flow Model
. AAI Survey ofTukwila Voice, Video and Data Monthly Rates
. Network Electronics and Related Equipment Vendor List
EXHIBIT A (Attachment 2)
,
Statement of\Vork
Work described in this Exhibit shall not commence without a wntten "Notice to Proceed"
provided by the City. Contractor shall submit to the City a written schedule that outlines
when the following tasks will be completed. Schedule shall be submitted to the City,
ATTN: Public Works Director, not later than 15 days after receipt of "Notice to
Proceed".
1. TASK A: SUBMIT A BUSINESS MODEL
Contractor shall re-evaluate the most recent business model associated with the proposed
deployment of the Tukwila Fiber Network (TFN). Update and submit a formal written
business model that considers the fiber deployment costs of only the Commercial
Business District (CBD) and as an option, a connection from the CBD to the Sabey Data
Center (CBD). The new business model shall include assumptions used in developing
the model such as take rates, network architecture selection, service providers, operations
and maintenance, bill of materials and other pertinent data.
2. TASK B: DETERMINE REVENUE SOURCES
Asset Analytics shall explore and determine from content providers, revenue
opportunities to defray or recapture deployment costs. Provide a written report to the
City, summarizing by content provider, the services and the amount of revenue that the
City can expect to receive.
3. TASK C: USE OF CITY'S STREET CREWS
In collaboration with City Staff, Contractor shall factor use of Public Works Department
Street Crews that have recently been trained and certified to work on fiber optic projects.
Contractor shall S1lmm~rize in written form, the advantages and disadvantages of using
trained City Crews and assess whether or not use of trained City Crews would lower fiber
deployment costs.
4. TASK D: EVALUATE POTENTIAL EFFICIENCIES WITH CITY TRAFFIC
SIGNAL PROJECTS
In collaboration with City Staff, Contractor shall explore possible efficiencies that could
be achieved with respect to the TIN Project and the City's traffic signal program (Green
River Signal Project and the CBD Signal Interconnect Project). Provide a written report
that outlines how TIN would interface with the two projects, and the savings that could
be obtained.
5. TASK E: FINALIZE LIST OF VEl\'DORS FOR SYSTEM COl'fIPONENTS
Contractor shall coordinate and develop a list of vendors that would provide equipment,
and service. Contractor shall provide to the,city a list summarizing the names of vendors
and the equipment/services that they could provide.
6. TASK F: DETER1\fiNE FINANCING/FUNDING STRATEGY
In close coordination with City's Finance Director, determine the optimal short and long-
term financing/funding resources. Coordinate with commercial entities (QWEST,
Verizon, COMCAST, Dynamic City, and other companies that may become potential
funding partners. Provide a ",Titten summary of companies contacted and an assessment
of their interest in becoming partners to the Project.
7. TASK G: MID-POINT REPORT
On or before November 15,2004, Contractor shall brief and submit report to the City
Council on progress related to TASKS A-F.
8. TASK H: OTHER TASKS DEEMED NECESSARY BY CITY
At the direction of the City, Contractor shall perform tasks related to this Statement of
Work. These tasks include but not limited to meetings with City Staff, attendance at
Utilities Committee Meetings, Connectivity Committee Meetings, and Council Meetings.