Meeting Notes: 11-16-04, 5pm
Board members present: Bill Leslie, Roberta Kennedy, Steve Williamson, Don
Hokanson, Scott Brown, Jeremy Pietzold, Dick Stamm, Joe Johanesen. Also present:
Carol Johanesen, Garrett Lang, Scott Lazenby
After introductions, Lazenby summarized the role of the newly-appointed board. Council
member Leslie thanked the board members for their willingness to serve, and stressed the
importance of their advice to the council.
The first policy issue was whether SandyNet should continue to be a wholesale customer
of Verizon (see meeting memo, attached). Board members urged SandyNet to move away
from wire-based service and move to wireless.
Lazenby summarized projections of different financial options for future service. A
wireless site roughly equals the net present value cost of DSL service if there are 32 or
more wireless customers. Adding WaveRider sites to fill existing coverage shadows
would be expensive. The board noted that if the payback period is too long (4 to 5 years)
the technology may be obsolete.
Board members noted that if DSL service is available, there isn’t a compelling reason for
a SandyNet investment in wireless. But there are DSL line problems on some areas of
Dubarko (in Knollwood subdivision) and in parts of Cascadia Village. An antenna site on
a ham radio tower near Bornstedt Road could serve both the rural (unserved) area and
Pietzold noted that Charter Cable claims they will offer cable modem service in 2006. He
also urged consideration of higher frequency wireless technology that provides higher
Stamm recommended that we contact Intel to offer SandyNet as a guinea pig for their
new Wi-Max technology.
There was discussion on the fundamental role of SandyNet: should we offer service only
where no other broadband (i.e., for now, Verizon DSL) is available, or should we position
ourselves to be the primary telecommunications utility in Sandy?
Board members made the following comments:
SandyNet shouldn’t undercut private companies on price; that isn’t the role of
SandyNet has already made a big difference in broadband service to the
community by placing competitive pressure on Verizon to upgrade their DSL
System expansion should focus on areas where there is no current service provider
(mostly rural areas surrounding Sandy).
A city-owned telecommunications utility can be a draw in economic development,
and in supporting home-based businesses.
There are possible community and education benefits where residents are tied
together in a network; Internet access isn’t the only use of our network.
The board agreed to meet at 5pm on Wednesday, December 8 to discuss the issue of
pricing our service. The meeting was adjourned at 6:40pm.
Notes by Scott Lazenby
City of Sandy Memorandum
DATE: November 2, 2004
TO: SandyNet Advisory Board
FROM: Scott Lazenby
RE: POLICY ISSUES FOR NOVEMBER 16 MEETING
1. The future of SandyNet DSL service.
In order to provide DSL service, we must purchase certain wholesale services from
Verizon. They consist of:
a. A $100 one-time fee, per customer, to connect the subscriber’s line to our
b. A flat monthly fee of over $600 for space rental, DC power, and air conditioning
for our equipment in the Verizon “central office” (the switch building in Sandy).
Our “Interconnect Agreement” with Verizon lists a tariff-based monthly charge of $14.36
per subscriber for the use of the line to get our DSL signal to the home or business. Until
recently, the FCC has prohibited Verizon from billing this fee because the subscribers are
already paying for their phone lines. In a recent ruling, however, the FCC changed its
mind and basically said that Verizon could charge whatever they want.
After Verizon closed us out of new orders in early October, we were forced to sign a
“standstill agreement” whereby we pay 25% of the tariff rate (or $3.59/month) for each
subscriber we started serving as of October of 2003. This buys us time—until February
2005—to negotiate a new line-sharing rate with Verizon.
I have had one conference call “negotiating session” with Verizon staff in Virginia and
Washington DC. Their initial proposed terms (these are confidential according to our
standstill agreement) are to bill us XX/month per subscriber for all customers signed up
before October 2004, and $X/month/line for subscribers signed up after that. They also
proposed to increase the one-time fee for hooking up a customer to $XXX (the only
service they provide for this is the time it takes to connect two wires; their on-line
ordering database is so full of errors that it is almost useless).
My understanding is that if we refuse to agree to these kinds of charges, they can refuse
any new hookups, but they can’t bill us for existing customers.
There are much bigger players than SandyNet (e.g., ATT, Sprint, Covad) that are battling
Verizon on this issue. I have heard that Covad has negotiated a $5/month fee for the
shared line, so this seems to be the going rate.
The decision we face is to:
a. Accept whatever fees we negotiate with Verizon, and continue to offer DSL
b. Discontinue providing new DSL connections after February 2005, and refer
potential SandyNet customers to Verizon for DSL service; or
c. Fill in the holes in our wireless coverage, and after February, provide new service
exclusively via the wireless network.
I have done some financial projections for these alternatives, but the results are sensitive
to the assumptions you use in each model (e.g., how we price our service, and the
percentage of homes that use it). We can explore them at the meeting.
A basic philosophical issue is whether we want SandyNet to hold onto a dominant role in
providing broadband service, at least within the City of Sandy itself. I think there are
some economic development benefits to doing so, and ultimately some community
benefits if we can reduce barriers to Internet access (e.g., for lower-income residents).
Owning the network is already providing benefits to local government (city hall, fire hall,
and community center are part of the network), and others—such as police, library, and
even the Mt. Hood Forest Headquarters—could benefit in the future. It could also reduce
our large telemetry (phone) bill for operating the water and sewer systems.
But on the other hand—unlike water or sewer service—there is no natural monopoly here,
and we can’t be guaranteed a financial return. I believe we can manage an ISP at least as
well as large bureaucracies like Verizon and Comcast, but if they follow a strategy of
pricing their service well below cost, we might not be able to outlast them.
Probably in response to pressure from Comcast and similar cable companies, Verizon has
reduced their monthly cost throughout the Northwest to $29.95. This is only eight dollars
less than SandyNet service, but it “feels” like a ten dollar savings. In the past we have
benefited from Verizon’s marketing: it got people interested in high speed access, and
when Verizon couldn’t deliver, they turned to SandyNet. But now we’re seeing the
opposite: people look into SandyNet, and then choose Verizon for price (and for some
misleading marketing, such as “we can deliver your DSL modem tomorrow!”).
Some municipal ISPs, particularly in the Midwest, intentionally price their service low in
order to serve as many homes and businesses as possible. For example, Buffalo MN,
which uses a WaveRider system like us, charges $9.95/month but then recovers their up-
front cost by having the customer buy ($475) or lease ($10/month for 55 months) the
equipment. Still, this results in a net charge of less than $20/month.
Chaska MN, which operates a wi-fi (802.11) mesh, charges a flat $15.99/month.
I have surveyed the costs of several other municipal broadband providers (wireless and,
like Ashland and Tacoma, cable-based), as well as the largest private “WISPs” (wireless
Internet service providers):
MONTHLY BROADBAND PRICE
As you can see, we are on the high end of the spectrum. To remain competitive, it seems
it would be a good idea to drop our price to $25 in-city and $35 out-of-city. The major
unknown is price-elasticity: If we cut our cost by 25%, will we get an increase in our
customer base of more than 25%? We can try to explore this by including a survey in our
next utility bill.
Especially for wireless service, almost all our costs are capital costs: the initial investment
in the network (around $250,000), the Access Point radio and antenna (about $10,000 per
site) and the customer equipment ($475 each). Our monthly cost for access to the Internet
itself is $2,000, and while this does depend on the number of customers, it does so in
large quantum steps (our current combined capacity of 4.5 Mbps can handle more
customers). Except for that, our marginal ongoing cost per customer is nearly zero.
For simplicity, we have evolved to a single monthly rate ($38) for service, which provides
data speeds of around 900mbps. Some providers provide different tiers of service at
different costs (Verizon initially did this with DSL, but has apparently abandoned the
strategy). For example, the $9.95 service in Buffalo MN guarantees bandwidth of only
192 kbps, or about a quarter of our speed (it is simple to set the “grade of service” in the
WaveRider equipment to limit the customer’s speed). The only operational advantage in
limiting the customers’ speed is to prolong the time it will take us to buy another
increment of access to the Internet itself; our own microwave and fiber optic networks are
well under capacity.
Another strategy is to offset our high up-front cost with a one-time fee, to essentially
allow the customer to “buy down” their monthly cost. Buffalo’s pricing options (i.e., buy
vs. lease the equipment) are intriguing. If you can convince a customer to buy their
equipment (whether a satellite dish in the case of TV service or a WaveRider modem for
SandyNet) they tend to become very loyal customers because they need to see a return on
3. Wireless Technology
Within the city limits, we have four major holes in our wireless coverage: the Sandy Bluff
subdivision, the Tickle Creek (Dubarko) valley, the Gary Street/Langensand area, and
Cascadia Village. There is a possibility of using a 100ft+ ham radio tower in the
Bornstedt area to provide coverage for a large area, including Cascadia, Bornstedt
Village, and Gary Street.
Large areas (but probably not all) of the Tickle Creek valley (including Nicolas Glen)
could be served by an antenna behind Suburban Auto Group. There would still be
“shadows” in areas like Orr Circle and the east end of Knollwood and Reich Court.
There is no apparent antenna site to serve Sandy Bluff. Some WISPs have used tree-
mounted antennas, and it might be worth trying one at Jonsrud Viewpoint; this would
have the added advantage of providing some service into the Sandy River valley. An ideal
spot would be the cell phone tower on the high school property, but the monthly rental
cost would probably rule this out.
Another alternative, though, is to try a pilot project using a wi-fi mesh. Chaska, MN, uses
a system supplied by Tropos, where wi-fi access points are mounted on street light poles
and form a self-routing mesh. They feature relatively powerful antennas, so they can be
placed at ¼ mile intervals. Based on this spacing, 5 access points each could cover Tickle
Creek/Double Creek, Cascadia Park, and Sandy Bluff neighborhoods (a total of 15), and
around 7 for Knollwood/Nicolas Glen. I am researching the cost (including installation)
of these access points. The big advantage of this system is that we would incur no cost for
customer equipment (vs. $475 for WaveRider equipment), since the customers would use
readily-available wi-fi cards.
This kind of system only makes sense inside the city; distances between homes are too
large in the rural area for it to work. We will do a full analysis of the rural pilot project
later in the fiscal year, but my guess is that we will need to wait for the long-promised
“wi-max” technology to get Internet service up the Mt. Hood corridor.